Professional Documents
Culture Documents
Volum Afases 2008 1 PDF
Volum Afases 2008 1 PDF
HENRI COAND
I.S.B.N. 978-973-8415-56-0
16 17 MAY 2008
BRAOV
ORGANIZING COMMITTEE
Chairman:
Col.prof. Stelian PNZARU, PhD
Vicechairman:
Lt.col.associate prof.eng.
Marian PEARSIC, PhD
Members:
Associate prof.eng. Horia TRZIU, PhD
Lecturer eng. Liliana MIRON, PhD
Associate prof.eng. Lavinia GAVRIL, PhD
Lt.col.eng. Mihai MIRON, PhD
Lecturer Mihaela SMEAD
Cpt.cmdr.eng. Marian BOBE
Assist.prof. Magdalena SCAFARU
Vol. I
AIR FORCE
Silviu BLU, Marian PEARSIC
TENDINE PE PLAN INTERNAIONAL PRIVIND REALIZAREA SISTEMELOR DIN
CLASELE SHORAD I VSHORAD
Ion CHIORCEA, Paul Mugurel POLEANSCHI
LUPTA AERIAN, COMPONENT CRITIC A LUPTEI NAVALE
Liviu COEREANU , Cornel PLEA, Daniel TURCANU
TARGET RECOGNITION ENHANCEMENT THROUGH EDUCATION
Dumitru DINU
ASIGURAREA SUPORTULUI PSIHOLOGIC AL MILITARILOR DIN FORELE AERIENE
PARTICIPANI LA OPERAIILE MILITARE DE STABILITATE
Daniel HURUBEAN
HIPOXIA DE ALTITUDINE: REACII LA NIVELUL ORGANISMULUI I EFECTE
GENERATE ASUPRA PERFORMANELOR PROFESIONALE ALE PILOILOR MILITARI
Gabriel-Florin MOISESCU
ACADEMIA FORELOR AERIENE HENRI COAND PREZENT I PERSPECTIV
Marius-Adrian NICOARA
APARITIA, DEZVOLTAREA SI PRINCIPALELE CONCEPTE CARE AU GUVERNAT
AERONAUTICA ROMNA
CONSIDERAII CU PRIVIRE LA FUNCIONAREA I VALIDAREA MODELULUI
FORTELOR AERIENE ROMANE N PERIOADA CELUI DE-AL DOILEA RZBOI
MONDIAL
Mihai NICU, Titica VASILE
UN MODEL MATEMATIC AL MICRII PROIECTILULUI N TUNUL
ELECTROMAGNETIC CU BOBIN
Cornel PLEA, Liviu COEREANU , Daniel TURCANU
ANALYZING OF AIRCRAFT TARGET PREDICTION USING A PROBABILISTIC MODE
Ionel POPA , Adrian POPA , Marian RISTEA
CONSIDERATIONS ON RESCUE OPERATIONS BY AIR
Marcel PREDA, Eugen BOAMB
PLANIFICAREA NTREBUINRII N LUPT A RACHETELOR SOL - AER
Marcel PREDA
MODALITI DE NTREBUINARE N LUPT A UNOR SISTEME DE RACHETE
ANTIAERIENE
Marius ERBESZKI
PERSPECTIVELE PARTENERIATULUI PUTERE AERIAN PUTERE TERESTR
Georgic SLMNOIU, Gabriel VLADU, Adrian CIUCULIN, Sorin PERICLEANU ,
Vasile IONACU
SISTEM DE CONTRAMSURI ELECTRONICE PENTRU AERONAVE, NAVE I
BLINDATE
ANS N PLUS PILOTULUI
Adrian STOICA, Corneliu AXENTE
CONSIDERATIONS ABOUT POSITION AND ATTITUDE DETERMINATION OF A UAV
USING INS AND GPS
Adrian STOICA, Ctlin MORARU
ELECTRONIC EQUIPMENT FOR COMBAT SEARCH AND RESCUE USED BY NATO
FORCES
Florin- Radu SUCIU
UN NOU MODEL MATEMATIC AL FENOMENULUI TRAGERII NTR- O GUR DE FOC
12
20
25
33
42
49
55
60
66
74
80
87
92
98
104
115
124
130
136
141
147
160
166
174
178
183
188
194
202
210
215
223
232
238
246
251
259
267
275
283
290
297
305
313
322
327
334
340
346
351
359
363
367
377
384
393
399
Aurelia TNSOIU
ASUPRA REZISTENEI I FIABILITII AMORTIZORULUI CE ECHIPEAZ APARATUL
DE TRACIUNE A VAGOANELOR DE CLTORI
ASUPRA REZISTENEI I FIABILITII ARCURILOR METALICE ELICOIDALE
24,4mm I 31mm UTILIZATE LA SUSPENSIA BOGHIULUI Y25 Lsdi CE ECHIPEAZ
VAGOANELE DE MARF
Radu RULESCU, Stelian RULESCU
SENZORI ULTRASONICI FOLOSII LA NAVIGAIA ROBOILOR MOBILI
Horia TARZIU, Doru LUCULESCU
METODE GRAFICE CU MIJLOACE CAD
Filis TULEI
ORGANIZAREA UNUI TERMINAL DE CONTAINERE: POSIBILITI DE CRETERE A
PRODUCTIVITII SISTEMULUI DE DESERVIRE AL NAVELOR DE DIMENSIUNI MARI
ELECTRICAL AND ELECTRONICAL ENGINEERING
Cornel Ovidiu BADEA
ASPECTE ALE REALIZRII CU FIBR OPTIC A SISTEMELOR DE TRANSMITERE DE
DATE PENTRU ANTENE ORIENTABILE CU COMAND NUMERIC
Ioan BELE, Liviu Sevastian BOCI , Ioan STRINESCU, Valentin MLLER, Mihaela POPA
CARACTERISTICILE DE FUNCIONARE PENTRU ECHIPAMENTELE DE ACIONARE
A TROLEIBUZULUI CU MOTOR ASINCRON
Marius Dan BENA, Csilla FARKAS, Petre Nicolae MOGAN
MSURRI MICRO I NANOMETRICE AL SUPRAFEELOR MATERIALELOR
METALICE MSURATE LA INTERCIUNEA SENZOR-SUPRAFA
Marius Dan BENA, Viorel ENE
METODE MODERNE DE MSURARE I PRELUCRARE A DATELOR APLICATE N
INGINERIA SUPRAFEELOR
Octavian-Ioan BOGDAN
POSIBILITI DE REALIZARE A FILTRELOR ACTIVE CU MOD DE LUCRU N CURENT
DIN FILTRE DE ACELAI TIP CU MOD DE LUCRU N TENSIUNE
Gigi-Daniel BUDARIU , Constantin GROZEA , Mihai MARIN
EVALUAREA STATISTIC A PROPRIETII DE DIFUZIE A ALGORITMILOR DE TIP
BLOC
Aurelian CRCIUNESCU, Valentin NVRPESCU, Cornel Ovidiu BADEA
AVANTAJELE INTRODUCERII COMENZII NUMERICE LA COMANDA ANTENELOR
ORIENTABILE
Eleonora Darie, Emanuel Darie, Garibald Popescu
THE DISTURBANCE IN THE AIRCRAFT ELECTRIC SYSTEM
405
410
418
424
433
440
448
456
461
446
474
478
486
492
500
506
512
516
524
532
540
546
554
562
567
572
580
588
594
600
606
612
619
625
630
635
650
656
662
669
674
682
690
695
699
705
711
716
724
732
740
747
755
761
769
775
781
787
Gabriela MOGOS
QUANTUM ENCRYPTION USING PRIVATE KEYS
QUANTUM ORACLES
Bogdan Gheorghe MUNTEANU
NOUA APROXIMARE A DISTRIBUTIEI TIMPULUI DE VIATA A UNUI SISTEM
tefan NEDELCU, Mircea LUPU
INFLUENTA STRUCTURALA A FORTELOR ASUPRA STABILITATII SISTEMELOR
DINAMICE(CU APLICATII) IN CAZUL MODELARII LAGARELOR IN REGIM HIDROAERODINAMIC
Gabriel NEPOTU
AN INTRODUCTION IN THE ASYMPTOTIC STUDY OF NON-LINEAR STOCHASTIC
DIFFERENTIAL EQUATIONS IN THE FIELD OF OSCILLATIONS
Maria-Carmen PNGLEANU
OPTIMIZING QUERIES
Dinu PASCALE, Tulei FILIS
DEZVOLTAREA SISTEMELOR INFORMATICE PORTUARE INTEGRATE O
NECESITATE STRINGENT PENTRU CRETEREA PERFORMANELOR PORTUARE
Alexandrina Maria PROCA
APPLICATIONS OF EXTREMAL VALUE THEORY TO RELIABILITY. SIZE EFFECT ON
STRENGTH OF PAPER STRIPS
Adrian-Gelu RADU, Anca HARABAGIU
GENERAREA FRACTALILOR DE TIP PLASMA I AFIAREA ACESTORA NTR-UN
MEDIU 3D
RENDERIZAREA FRACTALILOR JULIA PRIN VIZUALIZAREA PROPRIETILOR 4D
Gheorghe RADU, Gheorghe ANTON
AN OPERATIONAL METHOD FOR SOLVING LINEAR DIFFERENCE EQUATIONS
Ioan-Gheorghe RAIU
CONCEPTE DE BAZ PRIVIND SISTEMELE DE BAZE DE DATE
Adela SASU
K NEAREST NEIGHBOR PENTRU PREDICTIA SERIILOR DE TIMP
Petru SCURTU
METODE DE CALCUL A FUNCIEI DE INTERCORELAIE BIDIMENSIONAL
Ioan Tiberiu SOCACIU-LENDVAI, Ioan MAXIM
COMPLEXITATEA UNOR ALGORITMI IN REZOLVAREA PROBLEMEI GENERALIZATE
A TURNURILOR DIN HANOI
Alexandru STEFAN
CREATING YOUR E-LEARNING 2.0 STRATEGY UNDER THE WEB 2.0 DEVELOPMENT
Nadia Elena STOICUA
IMPLEMENTAREA N MATLAB A ALGORITMULUI METODEI RADIALE REGRESIVE
PENTRU TEORIA DE CURGERE J2 PENTRU CAZUL ECRUISRII IZOTROPE
NELINIARE
Miroslav TOLPA
TESTING OF NETWORKS SECURITY DEVICES
Florena Violeta TRIPA
CONJUGATA FUNCIILOR CONVEXE
Cezar VASILESCU
ESTIMAREA I PREDICIA FIABILITII SISTEMELOR DE COMAND I CONTROL
PE BAZA PROFILELOR OPERAIONALE
APLICAREA METODOLOGIEI DE ESTIMARE I PREDICIE A FIABILITII PE BAZA
PROFILELOR OPERAIONALE LA SISTEMELE C4ISR
792
796
804
808
816
819
827
836
842
850
859
865
875
879
884
890
895
903
909
913
920
AFASES - 2008 -
I. Aspecte introductive
Destinaia sistemelor SHORAD-VSHORAD o reprezint aprarea antiaerian
nemijlocit a aerodromurilor, elementelor de dispozitiv operativ, punctelor obligate de trecere,
precum i altor obiective, mpotriva unui inamic aerian ce acioneaz la nlimi mici i medii,
folosind intens manevra i bruiajul radioelectronic, n orice condiii meteo, ziua i noaptea. De
asemenea, pot fi folosite pentru acoperirea golurilor din zonele de angajare a sistemelor de
rachete sol-aer cu btaie medie sau mare. Raza de aciune eficace a sistemelor de tip
VSHORAD este de circa 4-5 km i, respectiv de aproximativ 10-15 km pentru sistemele de
tip SHORAD.
Dezvoltarea componentelor de lovire rachete de croazier, rachete balistice tactice,
avioane de lovire fr pilot (UCAV), muniii inteligente i rachete de aviaie cu raz de
aciune de peste 10 Km precum i perfecionarea procedeelor de aciune (procedeele de atac
STAND OFF -lansarea muniiilor din afara zonelor de aciune a sistemelor de arme sol-aer i POP-UP - apropierea de obiectiv la nlimi mici cu folosirea acoperirilor din teren i
lansarea muniiilor dup o intrare scurt n zona de angajare) impune realizarea capabilitilor
de angajare att a platformei purttoare ct i a muniiei. Principalele elemente caracteristice
ale mijloacelor actuale de lovire / atac din aer sunt:
- aciunea la nlimi mici sau foarte mici, sub acoperirea elementelor caracteristice ale
terenului i / sau bruiajului;
12
AFASES - 2008 -
Fig.1: Tipuri de inte combtute i anvelopele de aciune caracteristice sistemelor din clasele
SHORAD / VSHORAD.
Cerinele generale de capabilitate ale acestor sisteme sunt urmtoarele:
nimicirea intelor aeriene pe cile nemijlocite de atac la obiectiv n condiiile
ntrebuinrii contramsurilor radioelectronice (ECM, CHAFF, FLARE) i a
rachetelor antiradiolocaie (ARM);
angajarea simultan a unui numr mare de inte aeriene ce atac pe aceeai
direcie sau direcii apropiate;
timp redus de reacie;
integrarea n sistemul de comand control aerian (ACCS);
identificarea activ i sigur a intei (aparatur IFF);
aprare omnidirecional (360 0);
transportabilitate aerian;
mobilitate tactic;
cost operaional redus (personal redus).
II. Structuri de sistem integrat SHORAD-VSHORAD
Combinarea a dou sisteme, unul de tip SHORAD i altul de tip VSHORAD, ntr-o
entitate integrat sub aspect tactic i operaional, reprezint una dintre tendinele notabile care
caracterizeaz evoluia acestor sisteme de aprare. Astfel, un sistem integrat de arme
SHORAD - VSHORAD este definit prin ansamblul de componente SHORAD i VSHORAD
integrate acional i funcional prin intermediul unui centru de conducere a focului (FDC) ntr-
13
AFASES - 2008 -
o entitate de tragere care s dispun de capabiliti de conectare cu sisteme de comandcontrol aerian (ACCS). Apariia i dezvoltarea de sisteme integrate SHORAD -VSHORAD
este influenat n pricipiu de urmtorii factori:
pe timpul desfurrii unitii SHORAD, trebuie asigurat o aprare imediat prin
dislocarea iniial de uniti VSHORAD (cu rachete i/sau tunuri AA) avnd
capabiliti de tip C2 (comand-control);
posibilitile de lupt ale componentelor sistemulor trebuie s se completeze reciproc,
n scopul creterii eficienei aciunilor mpotriva inamicului aerian, ziua i noaptea, n
orice condiii meteo.
n cele ce urmeaz vor fi trecute n revist posibilitile de a realiza astfel de structuri
de sisteme integrate SHORAD VSHORAD.
2.1 Structuri de sistem integrat avnd componenta SHORAD realizat cu
rachete, iar componenta VSHORAD -cu tunuri antiaeriene
Asemenea structuri sunt cunoscute n principiu sub denumirea de sisteme antiaeriene
mixte tun-rachet. Realizarea acestei structuri este determinat n principiu de faptul c, sub
aspectul probabilitii de nimicire, sistemele de tunuri i cele de rachete prezint o
complementaritate. Astfel, componenta VSHORAD de tip tun este eficient la distane foarte
mici (0-1000 m), distane la care racheta nu poate aciona, ea avnd o zon moart cuprins
ntre 200 i 1000 m, n funcie de tipul rachetei, datorit timpului necesar a intra n etapa de
dirijare. n acelai timp, tunul prezint avantajele:
proiectilul de tun nu poate fi bruiat, spre deosebire de racheta dirijat;
tunul este mai ieftin dect instalaia de rachete dirijate.
n cazul forelor antiaeriene din cadrul alianei NATO sunt cunoscute urmtoarele
sisteme integrate de tunuri antiaeriene cu instalaii de rachete SHORAD:
sistemul Skyguard-Sparrow, avnd n compunere: subsistemul radiotehnic -tip
Skyguard, tunul antiaerian cu 2 evi tip GDF-005 i instalaia de lansare a
rachetelor tip Sparrow;
sistemul Skyguard Aspide similar cu sistemul Skyguard-Sparrow, cunoscut
sub numele de Spada (n Italia) i respectiv Toledo (n Spania); acest sistem
folosete rachete Aspide;
sistemul Skyshield-ADATS, compus din sistemul de tunuri Skyshield (tip
Revolver gun) i platforma de lansare rachete ADATS.
Fig.2: Platform ADATS (a) de lansare rachete i integrarea acesteia (b), prin intermediul
unitii de comand a focului (FCU ), n cadrul sistemului de tunuri antiaeriene Skyshield
14
AFASES - 2008 -
Rusia, spre deosebire de aliana nord-atlantic, s-a orientat spre realizarea de astfel de
sisteme autopropulsate, precum:
- sistemul PANTSYR-S1, montat n principiu pe un asiu auto (tip Ural);
- sistemul TUNGUSKA, montat pe un vehicul enilat.
15
AFASES - 2008 -
16
AFASES - 2008 -
17
AFASES - 2008 -
Folosirea de ctre sistemele de dirijare a rachetelor, funcie de zona n care se afl inta
naintea lansrii, a dou moduri de operare: LOBL (Lock-On-Before-Launch) i
respectiv LOAL (Lock-On-After-Launch), cum este cazul sistemului Spyder;
Creterea substanial a probabilitii de nimicire a intelor n tragerile executate cu
antiaeriene, prin folosirea muniie AHEAD cu subproiectile;
18
AFASES - 2008 -
Bibliografie
[1] ACTTM: Sistem antiaerian mixt tun-rachet studiu de concept, 2003.
[2] ACTTM: Sistem antiaerian integrat de arme SHORAD-VSHORAD studiu de concept,
2005.
[3] Janes Land-Based Air Defence 2002, London
[4] Land-based Vshorad and Shorad Systems Armada International, nr.2/2002;
[5] www.army-technology.com/projects/
[6] http://www.defence-update.com
[7] www.gbad.org
[8] www.israeli-weapons.com/
Silviu Blu, CS III dr. ing., Agenia de Cercetare pentru Tehnic i Tehnologii Militare, Bucureti,
tel. 0214233058/263, e-mail: sbaluta@acttm.ro
Marian Pearsic, confereniar univ. dr. ing., Academia Forelor Aeriene, Braov, Mihai Viteazu nr. 160, email: marianpearsica@yahoo.com.
19
AFASES - 2008 -
20
AFASES - 2008 -
http://www.fas.org/man/dod-101/sys/smart/agm-84.htm
http://www.fas.org/man/dod-101/sys/smart/agm-84.htm
2
21
AFASES - 2008 -
http://www.mbda-systems.com/mbda/site/FO/scripts/siteFO_contenu.php?lang=EN&noeu_id=105
http://www.defenseindustrydaily.com/ships-ahoy-the-harpoon-missile-family-02718/#missiles
6
http://www.missilethreat.com/cruise/id.104/cruise_detail.asp
5
22
AFASES - 2008 -
http://www.tulamash.ru/en/prod_kashtan.htm
23
AFASES - 2008 -
Cpt. cdor. Ion CHIORCEA, lector universitar doctor, Universitatea Naional de Aprare Carol I, Bucureti,
Str. Panduri nr. 68-72, sector 5, tel. centrala:319.48.80, e-mail: chiorcea44@yahoo.com
Lt. cdor. Paul Mugurel POLEANSCHI, masterand la Universitatea Naional de Aprare Carol I,
Bucureti, Str. Panduri nr. 68-72, sector 5, tel. centrala:319.48.80, e-mail: mpoleanschi@yahoo.com
24
AFASES - 2008 -
I. INTRODUCTION
Thermal cameras can be nowadays considered as standard military equipment in
surveillance applications for air force, navy, ground troops or paramilitary organizations like
borders guards, or police. Effectiveness of these imaging systems depends significantly on
their operators. There are usually no special problems with training military operators or
pilots for use of TV cameras as these systems generate images similar to images generated by
human eyes. However, training of the operators in use of thermal cameras is much more
difficult and time consuming because of four basic reasons.
First, because of different spectral range of thermal cameras in comparison to the
spectral range of human eyes, the thermal image differs very significantly to the visible image
of the same scenario.
Second, at least at present, thermal cameras are not stereoscopic systems like human
eyes, field glasses or some image intensifier goggles and is difficult to visually percept
different objects seen in the thermal image.
Third, there are no shades in thermal images even when the scenario is illuminated by
the Sun, the moon or other artificial sources.
Fourth, training operators of thermal cameras is costly due to high price of
consumable blocks like the cooler and the high costs of time consuming field training.
Due to the reasons mentioned above, the interpretation of thermal images is often
difficult for thermal cameras operators or pilots. Novice operators often are not sure whether
25
AFASES - 2008 -
they see real military targets of interest in the obtained thermal image or only typical natural
objects. They have also difficulties with recognition and identification of the detected objects
or to determine the distance to these objects. Due to military demands it became a constant
need to develop hardware/software image generators and to perfectly match training
guidelines suitable for air force, navy, ground troops or paramilitary organizations.
All these conclude to a probability definition for a target in the optoelectronic field of
view to be resolved. This aspect will be discussed bellow, from the theoretically point of
view.
Talking about education as a key factor in target recognition enhancement, means to
describe an entire process, with its hardware/software parts, trainers, subjects, instruction
periods with respect to thermal cameras, well defined laboratory conditions and guidelines to
be followed by the subjects being trained.
Generally, a training program should enable simulation of any factor presented in
diagram shown in Fig. 1.
Sun/moon/stars
clouds
target
Atmosphere
Fog/rain/snow/
dust
Thermal camera/operator
26
AFASES - 2008 -
27
AFASES - 2008 -
28
AFASES - 2008 -
stand out. In combat, soldiers and leaders should engage targets at the farthest range possible,
maximizing their weapons stand-off capability. The literature suggests that soldiers can
identify vehicles at tactical distances, particularly with flank and oblique views. But the most
efficient training is on close-up images first, particularly with soldiers who have low initial
skills. Advanced skill training should involve vehicles at greater and greater distances.
Therefore, the image database must contain images of vehicles at different ranges. We think
that 4 ranges will be OK, but 5 or 6 would work as well. More ranges would probably yield
unnecessarily redundant information with each increment in range; fewer ranges would
probably provide a change too great in the thermal signature with each increment.
2.3 Image fidelity
Image fidelity depends on resolution, contrast and luminance. A compromise must be
found between these characteristics in order to achieve an image with a good fidelity.
Sensor polarity
Most thermal sensors have reverse polarity, where the hotter spots can be displayed as black
(black-hot) or as white (white-hot). Switching from one polarity to the other helps the soldiers
to discern critical features. By using both polarities, soldiers frequently are able to discern
different pieces of information, which, collectively, help them, identify the vehicle.
2.4 Observing time
Observing time is also an important characteristic that has to be correctly evaluated. The
minimum observing time is calculated as the time necessary to scan the whole image. The
value obtained can be an acceptable value for the observing time. Experiments revealed that a
value for the observing time greater than 20s is useless and a plot will be presented bellow as
example of how the probability of resolving the depends with observing time.
Image scanning techniques
The human eye searches a display by aiming the fovea at different locations in an
image, rapidly moving from one fixation point to another. Each fixation lasts about a third of
a second on the average. The large jumps are called saccades; a saccade plus a fixation is
termed a glimpse. The saccades last only a few milliseconds and can reach speeds of 1000
deg/s. During the saccadic motion, vision is greatly reduced.
The speed and length of the saccade are dependent on the task, being faster for
situations where, for example, the allowed time to perform the search is limited to a short
interval, and being physically shorter when the density of information in the image is higher.
For more complicated scenes, the fovea will tend to interrogate areas of high information
content. The choice of the areas to interrogate will be governed in part by the observer's
expectations and by the information desired from the image. Obviously the characteristic
motion of the eye will play a central role in the ability of an observer to perform visual search;
therefore, models of visual search must explicitly or implicitly account for some aspects of
typical eye motion.
There are numerous factors which influence an observer's ability to acquire a target.
Only some of the major considerations are discussed here.
For competition search, as would be expected, the more the target differs from the
nontargets in contrast, shape, or size, the faster the target will be found. This assumes that
other factors such as the number of targets and the target density are kept constant. Any other
factors that would act to increase the target discriminability would also lead to decreased
mean search times.
Also, as might be expected, increasing the number of false targets, their homogeneity,
or the size of the search field will generally result in increased mean search times. Any change
29
AFASES - 2008 -
to the displayed image that would require an observer either to examine more objects, or to
examine the objects in greater detail, will lengthen the search process.
Some authors found that an observer, while first training on a set of targets (in their
case, letters of the alphabet), will scan more slowly while searching through a list for one of
many targets. Others point out that after learning the target set thoroughly an observer can
scan through a list looking for one of many targets in a target set just as fast as when scanning
for a single target. There are also authors who state that even though the scan rate for multiple
targets may be the same after learning, the error rates increase. Practice also can improve the
performance of tasks such as threshold detection that do not involve learning the shapes or
characteristic features of targets.
If the observer has been pre-briefed so that he or she knows probable positions or
states of the target, mean search times will be shorter. On the other hand, if a target does not
appear very much like the mental image of the target that has been learned by the observer,
the search process will be more difficult. If the target is camouflaged or partially occluded, the
target discriminability from background decreases and the search process becomes
correspondingly more difficult.
Knowledge of thermal technology
The better subjects understand the nature of FLIR and the parameters that affect what
is seen through a thermal sight, the more adept they become at interpreting thermal signatures.
This knowledge will also facilitate transfer from night to day images, to new tactical images,
to interpreting objects in the terrain, and to detecting targets.
Knowing thermal features of vehicle or aircraft classes
It is obvious that vehicles or aircraft within a certain class or certain type have
similarities (tanks are tracked, have a main gun and a turret and typically the engine is in the
rear). In addition, there are some distinctions among classes or types. Basic instruction on
vehicle or aircraft types or classes should precede training on naming individual vehicles or
aircrafts, as knowledge of these features helps narrow a soldiers choices considerably.
2.5 Presentation of specific vehicles or aircrafts and their cues
When soldiers are first exposed to a set of vehicles or aircraft, they need to be presented
with images that help them internalize the thermal representations of those vehicles or
aircraft. There are many ways to accomplish that goal. Critical to this instruction are the
following visual displays for each vehicle:
Comparison of thermal and visible (color) images
Displays that allow a 360 degrees view
Displays showing changes in thermal signature with distance
Videos of the vehicle moving at tactical speeds
Side by side visible and thermal displays help soldiers relate daylight to thermal
images. These displays provide connection between the known and the unknown.
Another technique can be to mix the images: a visible image of the vehicle or aircraft is
shown, which turns then into a series of white-hot and black-hot thermal images, and then
back to the original visible image.
The program (software) should point out the specific thermal features as the images
are displayed. Conduct this instruction with near images. Cues common to most vehicles
(engine, tread friction and exhaust heat) and cues associated with vehicle classes (large,
prominent turrets on tanks) should be pointed out for each vehicle. And finally, the features
that give each vehicle its unique thermal signature must be stressed. This signature is
impacted by the vehicles shape, its hot spots, the arrangement of these hot spots and the other
temperature differences on the vehicle that are typically less distinct, yet integral to its
signature.
30
AFASES - 2008 -
31
AFASES - 2008 -
Bibliography
[1] Michael C. Dudzik, Electro-Optical Systems Design, Analysis, and Testing, SPIE
Optical Engineering Press, Bellingham, Washington USA, 1996
[2] E.D. Turcanu, C. Plesa, L. Cosereanu, Training A Key Factor in Target Recognition
Enhancement, D&D Symposium, Mons, Belgia
Col. conf. univ. dr. ing. Coereanu Liviu , Mr.dr.ing. Plea Cornel , Lt. drd. ing. Turcanu Daniel - Agenia de
Cercetare pentru Tehnic i Tehnologii Militare, Str. Aeroportului nr. 16, CP19 OP BRAGADIRU, jude Ilfov
cod 077025, liviuc@acttm.ro, cplesa@acttm.ro, edturcanu@acttm.ro
32
AFASES - 2008 -
33
AFASES - 2008 -
restricionarea;
legitimitatea;
perseverena;
utilizarea forei militare n mod selectiv;
aciunea decisiv pentru prevenirea escaladrii;
nelegerea potenialului pentru consecine nedorite ale aciunii individuale ale
unitilor mici;
etalonarea capacitii de a folosi fora militar.
34
AFASES - 2008 -
stabilitate, accidente grave, etc.) nu ar ridica probleme sau dificulti majore, pentru care este
necesar o pregtire prealabil.
De aceea, n aceste ipostaze, militarii aparinnd F.A.R. se pot confrunta cu:
instruire insuficient pentru situaii limit (salvare, evacuare, transport i ngrijire a
rniilor, etc.);
informaii incomplete i, de multe ori, contradictorii;
aprovizionare i sprijin precare sau chiar lipsa total a acestora;
modificri drastice i instantanee ale condiiilor de via (hrnire, ngrijire i igien
corporal, somn, etc.);
condiii de mediu modificate sau dificile;
periclitarea sntii i a vieii proprii;
salvarea i/sau evacuarea muribunzilor, precum i evacuarea unui numr mare de
mori;
imagini dezolante, oripilante ale locurilor incidentelor critice.
Astfel, pregtirea psihic a militarilor aparinnd Forelor Aeriene atrage o serie de
cerine circumscrise domeniului, cum ar fi:
creterea profesionalismului lupttorilor, promovarea valorilor i lucrului n
echip, crearea unui climat de munc raional i onest;
dezvoltarea calitilor intelectuale, psihice i fizice;
formarea trsturilor psihomorale pentru a aciona oportun n situaii limit;
formarea, la ntreg personalul F.A.R., participant la operaii de stabilitate
multinaionale, a unor convingeri bazate pe recunoaterea contient a necesitii
ndeplinirii misiunilor , creterea ncrederii n forele proprii, asigurarea nsuirii i
respectrii valorilor i normelor militare;
cunoaterea i respectarea de ctre ntreg personalul aparinnd Statului Major al
F.A.R.
participant
la
misiuni
multinaionale
a
prevederilor
acordurilor/memorandumurilor i tratatelor ncheiate n baza crora se desfoar
misiunea (Memorandumul de nelegere MOA; Aranjamentul tehnic; Acordului
privind Statutului Forelor Armate - SOFA), a prevederilor regulilor de angajare
(Rules of Engagement - ROE), a legislaiei internaionale privind conflictele
armate, a dreptului internaional umanitar, a regulilor de comportare n teatrul de
operaii, etc.;
contracararea aciunilor de influenare psihologic;
realizarea coeziunii i operaionalizarea structurilor militare aparinnd
F.A.R. participante la misiuni de stabilitate multinaionale;
realizarea unui climat de exigen i corectitudine n relaiile interumane.
n linii generale, pe timpul operaiilor de stabilitate, trebuie s se realizeze pacea (prin
negocieri i medieri), dup care trebuie s se construiasc pacea (prin schimbri sociale,
dezvoltare economic i reconstrucie). De aceea, operaiile de stabilitate se desfoar pe
baza principiului consensului sau cererii naiunii gazd.
Caracteristici psihosociologice ale operaiilor militare de stabilitate
n urma studiilor efectuate, precum i a observaiilor rezultate n urma acestora privind
misiunile la care au participat efective ale Armatei Romniei, se poate afirma c, indiferent de
tipul acestora, din punct de vedere psihologic, militarii s-au confruntat cu aspecte de cele mai
multe ori asemntoare, reaciile lor de rspuns fiind n consecin.
35
AFASES - 2008 -
36
AFASES - 2008 -
- intimidarea, prin acte de cruzime extrem, rpiri, asasinate sau alte forme ale
violenei oarbe (exploziile, incendierile, distrugerea unor edificii publice, sabotarea
mijloacelor de comunicaii fiind fapte cotidiene).
Trstura principal rmne, astfel, teroarea, climatul de team, pericolul general,
teama de necunoscut ce mrete senzaia de insecuritate generat de orice aciune terorist i
sporete astfel ocul psihologic.
Implicaiile psihologice sunt deosebite, situaia concret fiind n aceste misiuni
aproape o stare lipsit de normalitate , n care se deregleaz att sistemul prin care oamenii
sunt informai corect i la timp asupra celor ntmplate, ct i cile prin care au acces la datele
necesare supravieiuirii, schimbrile rapide de situaii ducnd de la stri depresive la extaz sau
invers, deopotriv n rndul celor implicai direct, ct i a celorlali militari aflai n misiune.
n concluzie, se poate aprecia c operaiile de stabilitate desfurate ntr-un mediu i
context specific, presupun o mbinare armonioas de msuri de ordin politic, administrativ,
economic, militar i, nu n ultimul rnd, psihologic, precum i relaii de comunicare eficiente
pentru a anihila rapid efectele crizei i a scurta timpul de implicare a forelor armate.
Efectele sau urmrile misiunii
Efectivele participante la astfel de misiuni nregistreaz, de regul, alturi de rniii
somatici i tulburri psihice sau rnii psihic, care sunt victime ale teatrului de aciune,
victime ce sunt marcate, mai mult sau mai puin, de unele evenimente produse.
Evaluarea dimensiunii psihosociale a evenimentelor, inventarierea diferitelor
probleme psihice prezentate de cei implicai, examinarea reaciilor colective (panica,
inhibarea colectiv, exodul, zvonul, etc.), cercetarea reaciilor emoionale ale factorilor de
decizie i de salvare pe timpul gestionrii unor crize ne oblig s subliniem principalele
aspecte cu care se pot confrunta efectivele F.A.R. care pot participa la executarea unor
misiuni de stabilitate n cadrul unor fore multinaionale. Trebuie cunoscut c n asemenea
situaii putem avea de-a face cu:
rniii somatici (ari, strivii, traumatizai, mutilai, tec.) care, n plus, sunt
afectai i psihic datorit stresului post-traumatic la care au fost supui;
victimele psihice pure (fr rni somatice), ce prezint reacii emoionale mai
mult sau mai puin durabile, reacii nevrotice, isterii, etc. n astfel de cazuri, stresul acumulat
pn la salvarea acestora poate duce la manifestri neateptate, chiar necontientizate, de cele
mai multe ori, de victim (agarea de salvatori, lacrimi, proteste nejustificate, indiferen la
toate eforturile salvatorilor, nereceptivitate la toate ncurajrile acestora sau chiar refuzul, din
fric, la cooperare cu cei venii s-i salveze, etc.). Oricine este aflat n centrul unui dezastru
este victima unui stres psihic intens, raportat la spaim, surprindere sau imposibilitatea de a
nelege ce s-a ntmplat. n plan emoional, spectacolul morii celor apropiai,
eventualul contact cu cadavrele lor, cnd este constrns s supravieuiasc sub drmturi sau
blocat n epavele unor aeronave prbuite, disperarea de a nu fi salvat, n ciuda tuturor
apelurilor fcute, precum i convingerea iminenei decesului epuizeaz forele i ruineaz
rezistena psiho-fizic;
implicaii care, fr s prezinte manifestri patologice n mod expres, nu sunt n
starea lor normal. Fiind tulburai de prezena haosului, nelinitii, n cutarea celor apropiai
sau disperai de pierderea bunurilor personale, acetia vor prezenta, mai trziu, sechele posttraumatice nsoite de comaruri, regresia curajului, marcate de amintirea experienei
traumatizante trite;
leziunea colectiv (fenomene colective de agitaie i panic), observat n
catastrofele brutale, neateptate, este o reacie deosebit de contagioas i violent, ce pune n
pericol, de cele mai multe ori, operaia militar. Consecinele unor astfel de accidente
mobilizeaz toate contiinele asupra unei preocupri urgente, de moment, care genereaz o
deviere a comportamentelor de la schemele complementare de salvare i de ajutor reciproc,
37
AFASES - 2008 -
genernd impulsuri arhaice de fug panicat, de agitaie violent, de inhibiie colectiv, sau de
ncremenire;
lovirea sistemelor vitale care asigur supravieuirea societii umane: sistemul
de hran; sistemul de alimentare cu ap potabil; sistemul de sntate public; sistemul de
nvmnt i educaie; sistemul de meninere a ordinii publice; sistemul de liber circulaie a
persoanelor i bunurilor; sistemul de informare i comunicare, etc.
Salvarea vieii capt valene deosebite i triri intense, care solicit un consum
mare de energie, ncrctura emoional fiind foarte puternic. Tablourile dezolante sau
macabre (numr mare de mori, desfigurai, rnii aflai n agonie, etc.), constatarea decesului
sau rnirii unor cunoscui care produc uneori aa-zisele cderi nervoase, duc n final la
stri depresive, imposibilitatea de a aciona, neputin, accidente datorate scderii ateniei sau
preciziei n executarea unor operaiuni. Aceste fapte impun realizarea cu responsabilitate a
pregtirii psihice i fizice a efectivelor participante la astfel de misiuni .
Efectul psihologic al aciunilor teroriste este deosebit de puternic, deoarece se poate
genera urmtoarele efecte:
deteriorarea funciilor cognitive ale psihicului, instaurarea unui sentiment profund
de team i fric la gndul iminenei unor atentate teroriste;
emoie dominat de frica de necunoscut;
existena unui anumit control asupra indivizilor i mulimilor de ctre terorist
prin producerea (exploatarea) unor fenomene psiho-sociale;
indivizii, grupurile i mulimile supuse terorii sunt deturnate ctre aciuni care
nu sunt conforme cu propriile interese, motivaii i voine;
este afectat normalitatea psiho-social a grupului, producndu-se deseori
schimbri de judeci, atitudini, credine, cei speriai avnd o reacie excesiv i o
atitudine de defetism;
lipsii de ajutor, au ateptri exagerate din partea celorlali i o vulnerabilitate
deosebit la dezamgire.
Pentru aceasta, pot spune c, din punct de vedere psihologic, militarii implicai n
operaii multinaionale de stabilitate trebuie s dea dovad de:
neutralitate politic, impus prin lege militarilor, ceea ce creaz o echidistan
strict necesar i, totodat, compatibilitatea de a interveni n mod profesional i
specializat;
pstrarea stpnirii de sine pentru a nu se lsa dominat/ acaparat de emoie;
consecven n aplicarea i susinerea msurilor specifice;
solicitudine n ascultarea cu mult rbdare i atenie a tuturor celor implicai,
pentru soluionarea panic i operativ, fr violen, a diferendelor i pentru
ndeplinirea misiunii militare;
caliti deosebite de negociator i mediator, etc.
Militarii romni sunt recunoscui de ctre comunitatea internaional ca parteneri de
baz n efortul internaional pentru stingerea focarelor de conflict i intrarea n normalitate a
vieii economico-sociale din regiunile afectate.
Militarii din cadrul F.A.R. au demonstrat c pot participa, n condiii foarte bune, la
misiunile multinaionale, la nivelul standardelor cerute.
Executarea unor astfel de misiuni oblig efectivele participante s gndeasc mai mult,
s dea dovad de abilitate n aciune i comunicare, s demonstreze c nu sunt brute
narmate, ci adevrai soldai ai pcii, n condiiile n care misiunile reale, indiferent de
natura lor, reprezint un test absolut necesar pentru un militar profesionist al secolului XXI.
38
AFASES - 2008 -
39
AFASES - 2008 -
religiei din zona de operaii, etc.), sau pe baza unor edine practice (de negocieri i medieri,
sau de aplicarea regulilor de angajare).
n aceast etap, militarilor li se perezint Codul/Ghidul de conduit (Code of
conduct) sau Regulile de comportament (Rules of behaviour), pentru teatrul de operaii n
care urmeaz s fie dislocai. Aceste documente, care sunt puse la dispoziia fiecrui militar
din cadrul Forei Multinaionale, au la baza ntocmirii lor studiile, evalurile i estimrile
efectuate de elemente specializate ale Forei, cu privire att la zona de responsabilitate, ct i
la zonele de interes ale acesteia.
Documentele pot fi realizate n form grafic i text, nu sunt voluminoase, pot fi purtat
n buzunar de ctre fiecare militar i cuprind ntr-o form sintetizat principalele reguli i
norme de conduit pe care militarii forei multinaionale trebuie s le respecte atta timp ct
sunt dislocai n acea zon de operaii, precum i alte informaii care sunt considerate a fi
utile.
Activitile de suport psihologic nemijlocit vor urmri ndeosebi: cunoaterea
documentelor n baza crora a fost instituit misiunea; contientizarea personalului
detaamentelor asupra importanei i specificului misiunii; diversificarea i amplificarea
motivaiilor pentru participarea la misiune, dincolo de interesele materiale (financiare, n
special); nelegerea rspunderii ce le revin ca reprezentani ai Statului Major al Forelor
Aeriene Romne; cunoaterea riscurilor previzibile i a specificului zonei de dislocare i de
operaii; formarea deprinderilor necesare pentru a aciona oportun, n situaii limit;
cunoaterea tradiiilor i a obiceiurilor locale; identificarea elementelor distinctive ale altor
armate participante la misiune; nsuirea unor tehnici de confort i control psihomoral.
La acestea se adaug instruirea privind asigurarea igienei personale i colective,
protecia faunei i a florei, precum i a msurilor ce trebuie luate n zonele ce nu ofer condiii
civilizate de via sau sunt greu accesibile.
Asigurarea suportului psihologic pe timpul executrii misiunii n teatre de
operaii militare multinaionale
Pe timpul executrii misiunii, principalele activiti sunt menite s asigure o bun
coeziune i un moral ridicat, controlul i reducerea stresului. Aceste activiti ar putea s fie:
ntrirea i meninerea coeziunii detaamentului;
meninerea moralului detaamentului la un nivel ridicat;
contracararea propagandei i aciunilor ostile;
managementul stresului i intervenia pentru reducerea acestuia, dac este cazul,
prin desfurarea urmtoarelor activiti:
realizarea unor emisiuni radio n ar (tiri de ultim or din domeniile
de interes, politic, muzic, sport, nouti, etc.) special pentru militarii
aflai n misiune i transmiterea acesteia ctre zona de dislocare a
detaamentului;
asigurarea unor canale de comunicare rapid i facil a militarilor cu
familiile din ar (servicii potale, legturi telefonice, etc.);
analiza organizaional i intervenia corectiv.
Asigurarea suportului psihologic dup ntoarcerea din misiuni executate n teatre
de operaii militare multinaionale
Dup ntoarcerea din misiune, este recomandat s se desfoare un set de activiti
specifice care vizeaz n mod special: reacomodarea militarilor care au participat la misiune,
reintegrarea lor familial i reinseria profesional n unitatea din care au plecat; executarea
unor controale medicale amnunite; discuii de grup despre problemele ntmpinate i
soluiile care s-au adoptat pentru rezolvarea acestora; ncurajarea implicrii active a tuturor
40
AFASES - 2008 -
41
AFASES - 2008 -
Daniel HURUBEAN *
ASPECTE GENERALE
Fiecare celul vie din organism, trebuie s primeasc n permanen oxigenul necesar
metabolismului propriu i s elimine dioxidul de carbon pe care acesta l produce. Acest
proces este asigurat prin funcia respiratorie.
Organismul funcioneaz printr-un sistem de schimb de gaze, care comport 3 etape:
ventilaia, circulaia i schimburile tisulare. Mecanismele puse n joc sunt active: convecia
ventilatorie i circulatorie, i pasive: difuzia tisular. Debitele de gaze schimbate cu mediul
ambiant trebuie s satisfac cererile energetice, ceea ce implic o capacitate de adaptare
aplicat regimului de funcionare a ntregului sistem. n termeni cantitativi, aprovizionarea cu
oxigen este msurat prin presiunea sa parial (P1O2) care reprezint aproximativ numrul
de molecule de oxigen disponibil n unitatea de volum.
Conform legii lui Dalton, ntr-un amestec de gaze, presiunea total este egal cu suma
presiunilor pariale a gazelor ce compun amestecul, cunoscnd presiunea aerului (760mmHg
la niv. mrii) i volumul procentual al oxigenului (21%), se poate calcula presiunea parial a
oxigenului in aer, dup formula:
P oxigenului = (76021) : 100 = 160 mmHg
Altitudine
( metrii )
0
2500
4500
5500
6100
6700
7600
Presiunea barometric
( mmHg )
760
565
429
380
350
321
282
42
Pres. parial a O2
( mmHg )
159,2
118,3
89,9
79,5
73,4
67,2
59,1
AFASES - 2008 -
CLASIFICAREA HIPOXIEI
P. Dejours a propus o clasificare a diferitelor stri de hipoxie n funcie de cauzele de
apariie. Aceast clasificare distinge clar mecanismele care pot fi implicate la fiecare nivel al
sistemului de schimburi de gaze. In primele 5 tipuri de hipoxie, presiunea parial a
oxigenului din sngele arterial i saturaia hemoglobinei sunt sczute, iar aceste stri sunt n
mod obinuit atribuite hipoxiei hipoxice i hipoxemice. Aceste tipuri de hipoxii sunt: hipoxia
prin scderea presiunii pariale a oxigenului, hipoxia prin hipoventilaia alveolar, hipoxia
prin perturbarea raportului dintre ventilaie i perfuzie, hipoxia prin perturbarea difuziunii
oxigenului, hipoxia prin contaminare venoas, hipoxia hemoglobinic, hipoxia prin
ischemie, hipoxia tisulara prin edem, hipoxia tisular histotoxic, hipoxia tisular prin
utilizare excesiv.
Pragurile de reacie cauzate de hipoxie
Metri
Praguri
Presiunea
oxigenului atm.
67
101
HbO2 (%)
6700
Prag critic
3600
Prag de
tulburari
129
86-98
2100
Prag de reactie
160
96-98
< 65
65-86
Denumire
Zona mortal
Zona
compensrii
incomplete
Zona
compensrii
complete
Zona indiferent
43
AFASES - 2008 -
Fiziologic:
Ameel
Grea
Dureri de cap
Furnicturi
Comportamental:
Agresivitate
Vizual:
Vedere neclar
Reducerea cmpului vizual
periferic
In condiii de laborator, efectele date de hipoxie sunt diferit percepute de ctre observator i
persoana supusa hipoxiei. Manifestrile cele mai evidente sunt cele de la nivel mental i
comportamental.
44
AFASES - 2008 -
Creterea ventilaiei pulmonare n funcie de atitudine, la un subiect care respir aer i a crui
hipocapnie nu este compensat
Altitudine ( metri)
Creterea ventilaiei
( n procente )
0
4
14
24
33
50
1500
3000
4000
5000
6000
7000
45
AFASES - 2008 -
subiectul care ii pierde complet judecata i autocritica. El poate avea de asemenea senzaia c
judecata sa este perfect clar, dei el ntreprinde aciuni lipsite de logic i impotriva propriei
securiti.In egal msur s-au observat pierderi de atenie i de memorie care agraveaz
consecinele necoordonrii neuromusculare.
Aciunea hipoxiei asupra motricitii
Hipoxia provoac asupra masei musculare o prelungire a duratei i amplitudinii contraciei, o
prelungire a perioadei refractare i o diminuare a excitabilitii. Dup o uoar exagerare a
reflexelor tendinoase, observat ntre 2000-2500 m, urmeaz o diminuare a reflexelor pn la
aproximativ 6000 m. Dincolo de aceast altitudine, acestea se accentueaz din nou.
Dintre numeroasele teste de rezisten la hipoxie amintim: testul Omega, testul scrierii, pegmoving-test. De menionat c testul scrierii este un excelent test cu privire la pierderea
judecii, subiectul fiind ignorant n ceea ce privete deformarea scrierii lui, considerndu-l
firesc.
46
AFASES - 2008 -
repaus
activitate
moderat
activitate
moderat
7000
10 min
5 min
5 min
2 min
8000
5 min
2 min
2 min
1 min
8500
2 min
1 min
1 min
30 s
9000
1 min
45 s
45 s
20 s
10500
45 s
30 s
30 s
15 s
12000
30 s
20 s
20 s
15 s
13500
15 s
12 s
15 s
12 s
16000
15 s
12 s
15 s
12 s
20000
12 s
12 s
12 s
12 s
Acesti timpi reprezinta timpii de autosalvare pe care ii are pilotul la dispozitie dupa survenirea
accidentului, pentru luarea masurilor necesare corectiei si salvarii.
BIBLIOGRAFIE:
[1] J. Colin, J Timbal Medecine Aerospatiale, Expansion Scientifique Publication,1999
[2] Marian Macri, Constantin Raduica Curs de medicin aerospaial, Editura Militara
[3] Bert P. La pression barometrique. Recherches de physiologie experimentale. Masson,
Paris, 1979.
[4] Brendel W., Zink R.A. High altitude physiology and medicine. New-York, SpringerVerlag, 1982
47
AFASES - 2008 -
Cpt.medic Drd. Daniel HURUBEAN, Academia Fortelor Aeriene Henri Coanda, Brasov
48
AFASES - 2008 -
AFASES - 2008 orientrii i consilierii s abordm cteva concepte specifice actului formativ-educativ promovate
n instituia noastr, care i propune ca la finele anului universitar 2012 s ating standarde
superioare n pregtirea absolvenilor.
Educaia reprezint un fenomen social, aprut de la nceputurile omenirii, prin care
generaiile adulte transmit noii generaii experiena lor practic i de cunoatere, asigurnd, astfel,
continuitatea i progresul vieii sociale.
Din multitudinea de definiii, vechi i noi, ale educaiei, multe sunt orientate spre funcia
social a acesteia; ea are ca obiectiv s provoace i s dezvolte la individ un numr oarecare de stri
fizice, intelectuale i morale pe care le reclam de la el att societatea politic n ansamblul ei, ct i
mediul special cruia i este cu deosebire destinat. 1
Unii specialiti n domeniu sunt de prere c educaia trebuie neleas ca o experien de
via, iar, din aceast perspectiv, cea mai bun definiie n momentul actual i care se potrivete
punctului nostru de vedere ar fi urmtoarea: Educaia este o experien social, n contact cu care
individul se descoper, mbogete raporturile sale cu ceilali dobndete bazele cunoaterii i
aciunii. 2
Exist i alte moduri de a defini educaia, ns, fr a putea cuprinde multitudinea lor, le
menionm pe cele considerate a fi mai importante de ctre pedagogii contemporani 3 :
educaia este un fenomen specific uman; numai omul dispune de potenialul nativ care
face posibil educaia i permite umanizarea ;
este un proces de dezvoltare global, adic pe toate laturile personalitii individului;
este un proces continuu, desfurat pe parcursul ntregii viei a individului;
este un proces bipolar, implicnd interrelaia dintre factorii de educare i cei supui
influenei educative. Aceast interrelaie presupune reciprocitate, chiar dac rolurile
difer fundamental, fiind asimetrice.
Din perspectiva carierei, educaia profesional reprezint un ansamblu de activiti
pedagogice organizate, care-l pregtete pe individ pentru succesul n viaa profesional, centrat
mai ales pe formarea lui n spiritul unei etici a muncii, bazat pe profesionalism, responsabilitate,
cooperare i competiie.
n pedagogia actual termenul de educaie este deseori substituit cu conceptul de formare,
n dou moduri relevante: fie n sensul de educaie ca proces, fie n sensul de rezultat al educaiei. El
reprezint, ntr-un sens destul de larg, pregtirea tinerei generaii pentru via, pentru viitoarea
profesie i continuarea acesteia pe parcursul vieii.
Orientndu-ne asupra formrii profesionale a indivizilor ntr-un anumit domeniu, aceasta
reprezint pregtirea iniial a cadrelor necesare pentru acel domeniu, care se realizeaz prin
structuri instituionale, programe i strategii pedagogice, perfecionarea acestei pregtiri dureaz pe
parcursul ntregii viei profesionale.
ncercnd s sintetizm, considerm c EDUCAIA individului n Forele Aeriene
reprezint un proces de dezvoltare general i continu, ce urmrete formarea intelectual i
profesional a tinerilor n spiritul responsabilitii, profesionalismului i cooperrii.
Academia Forelor Aeriene Henri Coand este o instituie de nvmnt superior militar,
continuatoarea tradiiilor educaionale a mai multor instituii de nvmnt ce au funcionat de-a
lungul timpului, cu o istorie de peste 96 de ani.
Educaia n aceast instituie de nvmnt superior din Forele Aeriene este orientat pe
pregtirea universitar a tinerilor studeni, precum i pe pregtirea profesional iniial a viitorilor
ofieri aviatori, artileriti i rachetiti antiaerieni, radiolocatoriti, n raport cu solicitrile i comanda
beneficiarilor din Ministerul Aprrii, Ministerului Internelor i Reformelor Administrative, etc..
Cu alte cuvinte, Academia asigur educarea global a unui tnr liceniat n Managementul
organizaiei, domeniul tiine Militare i Informaii i formarea sa iniial, modelarea unui
produs profesionist generator de securitate naional i euroatlantic.
1
50
AFASES - 2008 Pornind de la acest statut, de la misiunea pe care o are nvmntul din Academia Forelor
Aeriene, planurile i programele, celelalte documente de nvmnt, cu alte cuvinte curriculumul
educaional, se impune a fi elaborate n acord total cu nevoile specifice de pregtire a ofierilor din
Forele Aeriene, a celorlali beneficiari, dar cu respectarea reglementrilor naionale n domeniu i
corelate cu cerinele U.E. i ale N.A.T.O., precum i cu cele ale Organizaiei Internaionale a
Aviaiei Civile, pentru compatibilizarea ntregului nvmnt cu coninutul, metodele i procedurile
de instruire existente n armata rilor aliate.
n acest context, prioritatea instituiei noastre o reprezint realizarea unei concordane
depline ntre domeniul de licen, coninutul propriu zis al nvmntului universitar i modelul
absolventului Academiei Forelor Aeriene, prin realizarea unei instruiri corelate cu standardele
euroatlantice, dar i cu nevoile specifice structurilor operaionale ale Forelor Aeriene i ale
celorlali beneficiari, pe aeronave i echipamente militare moderne, prin nsuirea procedurilor i
tehnicilor de lupt comune, inclusiv prin formarea unui sistem de gndire i a unei mentaliti
compatibile cu cele ale partenerilor de alian.
Cine suntem i ce am realizat?
n acest moment Academia Forelor Aeriene Henri Coand este o instituie militar de
nvmnt superior, parte a sistemului naional de nvmnt, care funcioneaz n baza
prevederilor legale specifice domeniilor de activitate, care respect i promoveaz conveniile
europene i recomandrile Uniunii Europene privitoare n statutul instituiilor de nvmnt
superior.
Obiectivul fundamental al activitilor de nvmnt n Academia Forelor Aeriene Henri
Coand, instituie acreditat prin Hotrrea de Guvern nr. 410/25.04.2002, l constituie formarea
ofierilor liceniai n domeniul tiine Militare i Informaii, specializarea Managementul
organizaiei pentru studenii de la cursurile de zi i fr frecven, pentru Forelor Aeriene i ali
beneficiari interni, continuarea implementrii standardelor, procedurilor i tehnologiilor de instruire
i evaluare ale armatelor partenere ale alianei, devansarea prin aria curricular a transformrilor i
schimbrilor care se produc la nivelul structurilor operaionale ale Forelor Aeriene.
Organizarea nvmntului n Academia noastr este analizat astfel nct curriculum-ul
universitar s respecte standardele de acreditare universitar, Concepia de formare, dezvoltare
profesional i utilizare a ofierilor din Armata Romniei i Modelul absolventului, susinute
prin obiective didactice generale i operaionale, cuprinse n planurile de nvmnt i programele
analitice.
Planurile de nvmnt, aprobate de ealoanele n drept, sunt valabile pentru seriile de
studeni n curs de colarizare i cuprind discipline fundamentale, discipline de specialitate n
domeniu specialitile Pilot pe avion cu motor reactiv, Pilot pe elicopter, Controlor trafic
aerian operaional, Artilerie i rachete sol-aer, Artilerie i rachete antiaeriene, cu durata
colarizrii de 3 ani (180 de credite transferabile) i discipline complementare, grupate la rndul
lor n discipline obligatorii , opionale i facultative, n conformitate cu cerinele normative stabilite
la nivel naional.
Disciplinele de studiu i coninuturile de nvmnt asigur articularea raional-pragmatic
a categoriilor de pregtire i corespund profilului i specializrii viitorilor ofieri. Ele sunt prevzute
ntr-o succesiune logic i au n vedere definirea i delimitarea precis a competenelor generale i
de specialitate, compatibilitatea cu domeniul i specializarea pentru studii universitare de licen,
precum i cu nevoile i prioritile de formare n specialitile militare specifice.
Ponderea fiecrei categorii de pregtire n totalul activitilor didactice, precum i
disciplinele pe baza crora se ndeplinesc obiectivele didactice generale, sunt stabilite n
concordan cu standardele naionale de evaluare academic, adaptate la specificul instituiilor de
nvmnt superior militar.
51
AFASES - 2008 Managementul resurselor umane participante la procesul de nvmnt vizeaz, n primul
rnd, asigurarea acoperirii disciplinelor din planurile de nvmnt cu personal didactic de
specialitate, stabilirea perspectivei reale n carier a cadrelor didactice, sprijinul acestora n vederea
pregtirii continue, creterea participrii profesorilor i studenilor la activitile extracademice i la
cele cu caracter internaional, tot ce este posibil a fi fcut pentru creterea performanelor
studeneti.
Posturile didactice universitare legal constituite sunt acoperite n totalitate cu cadre didactice
universitare, din care 45,4% sunt ncadrate cu confereniari i profesori universitari.
Majoritatea cadrelor didactice continu s-i perfecioneze propria pregtire prin doctorat ,
cursuri postuniversitare, lucrri de cercetare i participri la sesiuni de comunicri tiinifice
naionale i internaionale.
Dinamica cifrei de colarizare este n cretere, iar conducerea academic se preocup pentru
promovarea imaginii instituiei, precum i pentru motivarea adecvat a studenilor.
n prezent, activitatea de cercetare tiinific a cadrelor didactice se axeaz, n principal, pe
participarea la competiia de granturi, elaborarea de manuale, cursuri interne i publicaii de
specialitate, obinerea de titluri tiinifice, mbuntiri ale procesului de nvmnt i ale bazei
materiale a acestuia.
Continu editarea revistei Academiei Forelor Aeriene Henri Coand, revist de informare
tiinific de categoria B, cu anse de trecere la categoria B+, cu apariie semestrial, n care sunt
publicate rezultatele notabile ale cercetrii tiinifice efectuate de cadrele didactice i studenii
instituiei noastre. ncepnd cu semestrul al II-lea a acestui an universitar, a aprut i revista Noi,
revist a studenilor, cu o frecven de apariie de trei ori pe an.
Ca urmare a nivelului tiinific ridicat al lucrrilor publicate i a demersurilor efectuate,
editura academiei a fost recunoscut de ctre Consiliul Naional al Cercetrii tiinifice n
nvmntul Superior.
Academia dispune de spaii de nvmnt i cercetare care corespund domeniului i
specializrii pentru studii universitare de licen, de spaii destinate cazrii studenilor i cursanilor,
de spaii destinate activitilor sociale, culturale i sportive.
S-a avut permanent n atenie dotarea cu aparatur i dispozitive moderne a laboratoarelor i
slilor de specialitate, completarea fondului de carte a bibliotecilor universitare, asigurarea pieselor
i materialelor necesare funcionrii n bune condiii a laboratoarelor i slilor de specialitate,
realizarea de machete, montaje, panoplii, plane pentru dotarea cabinetelor, dotarea spaiilor de
nvmnt cu tehnic de calcul, conform cerinelor, dezvoltarea bazei sportive a Academiei i
modernizarea condiiilor de lucru n Tabra de instrucie.
Ce ne propunem s realizm?
Concepia reorganizrii nvmntului din Ministerul Aprrii se apropie de final, iar
Academia Forelor Aeriene Henri Coand se pregtete s fie inspectat n vederea reacreditrii.
n sperana c vom ncheia cu bine i acest an universitar, nu putem s uitm emoiile ce
ne-au stpnit n ultimii ani din perspectiva continuitii funcionalitii instituiei noastre. n
consecin, ne propunem ca Academia Forelor Aeriene Henri Coand s devin mai puternic,
mai modern, mai bine adaptat la cerinele i exigenele comenzii sociale, precum i ale
nvmntului naional i european.
Obiectivul primordial de perspectiv n formarea viitorului absolvent se prezint oarecum
diferit ntruct se va urmri ca n cadrul ciclului de studii universitare de licen s li se ofere
cunotine tiinifice solide, n vederea atingerii nivelului de cunoatere necesar pentru a putea
continua formarea n cadrul studiilor de masterat, precum i competene i aptitudini practice
solicitate de beneficiari i considerate ca fiind necesare i suficiente pentru ocuparea primei funcii
de ofier.
Una din principalele preocupri ale managementului academic, n dorina de a ne asigura
progresul i stabilitatea instituional, va fi de a dezvolta oferta educaional, att pentru mediul
52
AFASES - 2008 militar romnesc i strin, ct i pentru mediul civil naional. n acest sens, intenionm s
dobndim autorizarea provizorie de funcionare i, ulterior acreditarea pentru nc dou noi
specializri:
- pentru absolvenii piloi specializarea Managementul aeronautic;
- pentru aviatorii nenavigani i radiolocatoriti specializarea Managementul traficului
aerian, att pe filier militar, ct i pe filier civil, n colaborare cu coala de Aviaie
Bneasa.
n sperana unei colaborri fructuoase cu instituia mai sus menionat, dar i cu ali
parteneri civili, dorim s nfiinm cel puin o facultate, structur de baz a oricrei universiti i
academii, la nevoie dou, una cu orientare militar, iar cea de-a doua civil.
Un alt obiectiv prioritar n domeniul funcionalitii instituiei l reprezint nfiinarea n
academie a studiilor universitare de masterat cu durata de 2 ani, n domeniul Managementului
spaiului aerian, susinut de un proces continuu de recrutare i promovare de cadre didactice cu
titluri universitare de confereniar i profesor, competente n domeniul studiilor de masterat.
Tot din perspectiva educaional intenionm ca, pn n anul 2010, s adoptm o serie de
sisteme de pregtire E-learning, ca procedee moderne de extindere a formelor de organizare i
desfurare a procesului educaional i tot pn n acel an s acreditm Centrul de Cercetare
tiinific pentru specializarea Managementul organizaiei.
Managementul resurselor umane, angajate n procesul educaional i de cercetare tiinific,
va urmri promovarea unei politici care s atrag personal didactic de perspectiv, s ncurajeze
excelena, evoluia profesional, ataamentul i loialitatea fa de instituie.
n intenia de a transforma Catedra de Specialitate i Instrucie Militar ntr-o catedr
universitar, intenionm s sprijinim pregtirea prin doctorat a ct mai multor instructori, dar i s
selecionm i s promovm ofieri cu experien din uniti operaionale, dar mai ales, dintre cei
care au participat la misiuni n teatrele de operaii.
Trebuie s fim convini c deficitul de ofieri tineri nu poate fi soluionat adecvat prin
formarea acestora pe filiera indirect, aa c ne ateptm s fim sprijinii spre a crete dinamica
cifrei de colarizare n instituia noastr.
n scopul unei selecii ct mai bune la admitere intenionm s promovm n mod activ
imaginea Academiei Forelor Aeriene Henri Coand, s oferim condiii ct mai bune de studiu i
de formare profesional pentru cei care se pregtesc aici.
Conducerea academiei va avea n atenie i va aciona pentru consolidarea logisticii, a
nivelului de asigurare financiar, pentru dezvoltarea relaiilor de colaborare inter-universitar i cu
societatea civil, necesare att creterii calitii nvmntului i cercetrii tiinifice precum i a
prestigiului naional i internaional al academiei.
n acest sens, se va urmri proiectarea unei noi infrastructuri a reelei de calculatoare care s
conduc, pn n anul 2009, la definitivarea informatizrii procesului educaional n academie,
dezvoltarea sistemelor INTERNET i INTRANET, modernizarea bazei materiale a nvmntului
n concordan cu tendinele tehnologice i ergonomice existente.
Se va avea n atenie modernizarea editurii i tipografiei, precum i relocarea, reorganizarea
i modernizarea bibliotecii universitare, inclusiv prin dezvoltarea bibliotecii virtuale, toate menite s
satisfac ct mai bine solicitrile utilizatorilor i beneficiarilor.
Acestea sunt doar o parte din coordonatele strategice pe care urmrim s acionm n
perspectiv, n organizarea i derularea procesului educaional n Academia Forelor Aeriene Henri
Coand, coordonate care vor urmri s menin cu succes instituia noastr pe firmamentul
nvmntului universitar romnesc i internaional, care s consolideze renumele ctigat n cei 13
ani de existen.
n cele menionate mai sus, am operat cu dou concepte, pregtirea continu i adaptarea
permanent, ntre care apreciem c exist o relaie determinist, primul reprezentnd cauza, iar cel
de-al doilea efectul. Numai printr-o pregtire continu ne putem adapta n permanen la cerinele i
exigenele nvmntului universitar naional i european. Dar, performana cere, din punctul
53
AFASES - 2008 nostru de vedere, i mai mult, ntruct, dac durata de adaptare este mare, va crete direct
proporional i clivajul dintre ceea ce se cere i ceea ce putem oferi.
Care ar fi soluia de adoptat n aceast situaie ?
Apreciem c atingerea i meninerea nivelului de performan cerut, nu se poate face dect
printr-o aciune concertat, desfurat de toi membri comunitii academice, n direcia pregtirii
continue a personalului coroborat cu aciuni menite s scurteze timpul destinat componentelor
adaptrii, asimilarea i acomodarea i implicit ntregului proces al adaptrii.
Dar conceptele trebuie transpuse n practic, cu oamenii i pentru oameni!
BIBLIOGRAFIE:
Mircea tefan Lexicon Pedagogic, Editura Aramis, Bucureti, 2006;
***
- Gndirea Militar Romneasc, Revist de teorie i tiin militar editat de
Statul Major General al Armatei Romniei, Nr. 5/2007;
***
- Gndirea Militar Romneasc, Revist de teorie i tiin militar editat de
Statul Major General al Armatei Romniei, Nr. 3/2007;
***
Revista Academiei Forelor Terestre, Editura Academiei Forelor Terestre,
Sibiu, Nr. 1/2007.
54
AFASES - 2008 -
Comandor av. Alexandru Sahini, Arma Aerului n Arta Militar, Biblioteca Academiei Romne,
Bucureti, 1943, p. 3
* Giulio Douhet (1869 1930) general italian, adept i promotor nflcrat al principiului conform cruia arma
aerian are un loc decisiv i capital n conducerea i desfurarea rzboiului, solicitnd n consecin, o
putere aerian independent, concretizat n armata aerian de sine stttoare. n lucrarea Il domino dell
aria (Stpnirea aerului) aprut n 1921, pleac de la ideea c pretutindeni unde doi oameni se pot ntlni e
inevitabil o lupt i a pledat pentru capacitatea ofensiv a avionului, apreciat ca fiind att de mare nct
duce la o consecin absurd n sine: pentru a te apra de o ofensiv aerian e nevoie de mai multe fore dect
pentru a ataca. Concluzia de fond a concepiei lui era c a nvinge nseamn, a cuceri supremaia aerian.
Am convingerea ferm scria el - c viitorul nu m va dezmini i ca urmare, flota aerian va nregistra o
cretere nsemnat, reducndu-se n aceeai msur rolul armatei de uscat i al marinei.
55
AFASES - 2008 -
56
AFASES - 2008 -
10
57
AFASES - 2008 -
scurt, iar unei baterii de artilerie (terestr n.n.) spre a arunca acest tonaj i trebuie circa
24 ore. 14
Aeronautica Regal Romn a folosit n practica sa din cel de-al doilea
rzboi mondial, pe rnd, cele dou concepii de ntrebuinare. Trecnd n revist
principiile artei militare interbelice romneti, devenite i ale artei militare aeriene,
prin logica experienei cptate, prin studiul i activitile personalitilor din aeronautic,
sau a practicii frontului, se considera c aeronautica romn respecta legile i
principiile militare fundamentale comune, ns cu particularitile specificului armei.
Aceste principii fundamentale adaptate n ntregime n strategia i tactica
aerian, dup caracterul, volumul i natura acesteia erau: a) principiul moralului, b)
principiul aciunii, c) principiul economiei forelor (principiul massei), d) principiul
conservrii forelor, e) principiul iniiativei, f) principiul siguranei. Analiza lor
din punctul de vedere general i al aeronauticii, a format suportul pe care s-a
sprijinit materia artei militare aeronautice, pstrnd logica general a procesului.
Calitile superioare tuturor celorlalte arme, fceau din aeronautic primul
instrument tactic i strategic. Toate aceste proprieti i incomparabile caliti prin
care excelau Forele Aeriene Regale Romne*, formau prin nsumarea lor ceea ce astzi
numim putere aerian.
Strategic, aviaia era considerat arma principal ofensiv i independent,
care n mna comandamentului suprem, era primul mijloc politic, militar, strategic i tactic,
pentru impunerea asupra voinei adversarului. Aciunea strategic de ansamblu a
forelor aeriene cerea ns coordonarea concepiei operaiilor generale cu aciunea
forelor terestre i navale, operaia aerian conducndu-se i desfurndu-se
separat, dup regulile proprii ale marilor uniti aeriene.
Tactic, aviaia se constituia n grupri aeronautice destinate aciunilor militare
aeriene dar i n folosul marilor uniti terestre sau navale, exploatndu-se
proprietile sale ofensive. Cooperarea aeroterestr sau aeronaval, forma capitolul cel
mai delicat al ntrebuinrii aeronauticii, pentru c pretindea suprapunerea i
coordonarea perfect n timp i spaiu ntre aer i sol, activitate foarte greu de
realizat, mai ales pe timpul desfurrii luptei.
Adevrul istoric nepolitizat, cu privire la aceast categorie de fore armate, ne
permite s urmrim i modul n care s-au micat i au evoluat marile idei ale artei militare
aeronautice romneti i mondiale. Anticipnd putem spune c dup cel de-al doilea rzboi
mondial aeronautica a fost desfiinat i fragmentat pe armele din compunerea sa conform
doctrinei militare sovietice, i abia dup evenimentele din decembrie 1989 s-a revenit la
matca normal a Forelor Aeriene Regale Romne, de ntrebuinare conform doctrinelor
militare aeriene occidentale, ca element central de structur militar, o categorie de for
armat care prin armele din compunere sa, s pun n practic noiunea de putere aerian.
* Generalul comandant aviator Gheorghe Jienescu s-a nscut la 16 septembrie 1895, n localitatea Rast, judeul
Dolj i a murit la data de 3 aprilie 1971 la Bucureti. A fost comandantul Forelor Aeriene Regale Romne pe
timpul Campaniei din Est a Armatei Romne n cel de-al doilea rzboi mondial. Arestat pe 23 august 1944, la
29 ianuarie 1945 a fost inclus pe lista celor 65 de persoane bnuite de dezastrul rii i deferite acuzatorilor
publici spre cercetare i instrucie. A primit o pedeaps de numai 20 de ani temni grea deoarece are
circumstane atenuante pentru c a dispus ca medicii romni s dea asisten i populaiei civile de peste
Nistru. A fost nchis n perioada 12 februarie 1946 - 18 aprilie 1964, n penitenciarele Aiud, Vcreti,
Rmnicu-Srat i Gherla. A fost
eliberat cu puin nainte de mplinirea vrstei de 70 de ani. n ultima perioad a vieii, bolnav, evacuat din
propria cas, i-a dus cu demnitate viaa alturi de familia pe care o iubea i care la sprijinit pn la sfrit.
14
A.M.R., Fond 1376, Dosar 1375, f. 4
* Dei nu am gsit un document oficial de atribuire a sintagmei de Forele Aeriene Regale Romne, aceasta
mpreun cu formularea Aeronautica Regal Romn erau denumirile uzitate n toate formulrile oficiale ale vremii.
58
AFASES - 2008 -
Putem concluziona c dei Forele Aeriene Regale Romne au fost desfiinate dup
ultima conflagraie mondial, prin clauzele Tratatului de Pace ncheiat de Romnia cu
Naiunile Unite la Paris, n data de 10 februarie 1947, 15 ele i dovediser eficiena i
eficacitatea. Dei modelul organizatoric al acestei categorii de for armat a fost apreciat
n vreme pentru rezultatele sale n lupt, acesta a fost abandonat sub tutela noului aliat
sovietic, dar a revenit in actualitate dup aproape 47 de ani, prin renfiinarea la 30
noiembrie 1993, a Statului Major al Aviaiei i Aprrii Antiaeriene (ulterior Statul Major
al Forelor Aeriene) n conformitate cu Ordinul General al Ministrului Aprrii Naionale
nr. 24/1993, ca o confirmare a justeii drumului interbelic pe care se fixase aeronautica
romn, i o rennodare a acestui drum, pe coordonatele concepiei de ntrebuinare a
aeronauticii occidentale.
Capitan comandor dr. NICOARA MARIUS-ADRIAN eful catedrei de aviaie din SAFA
15
X.X.X., Institutul de istorie i teorie militar, Istoria militar a romnilor (Culegere de lecii), volumul II,
Editura Militar, Bucureti, 1992, p.249.
59
AFASES - 2008 -
60
AFASES - 2008 -
61
AFASES - 2008 -
3 Ibidem, f 323
4 Comandor (r) Dan Stoian, Pumnul tarii., Editura Modelism, Bucureti 2000, p. 181
5 X.X.X., Istoria aviaiei romne, Editura tiinific i enciclopedic, Bucureti, 1984, p 305
6 G.V. Zimin, Piloi de vntoare, Editura Militar, Bucureti, 1990, Vol 1, p 174
7 Ibidem, p. 175
62
AFASES - 2008 -
8 1bidem, p 191
9 Ibidem, p. 192
10 A.M.R., Fond 1377, Dosar 854, f 324
11 Hlihor Constantin, Op. Cit., p. 37 - 40
12 Doar dou asemenea grupuri specializate n misiuni aeriene de asalt au existat n ntreg cel de-al doilea rzboi
mondial, pe froturile din Europa i Africa, unul german i cellalt roman, vezi lucrarea lui Cornel Marandiuc,
Cavalerii bimotoarelor de asalt, Editura Modelism, Bucureti, 2000, p 19
13 Cornel Marandiuc, Op. Cit. Editura Modelism, Bucureti, 2000, p 33
14 A.M.R., Fond S.M.A, Dosar 136, f 10
15 A.M.R., Fond 359, Dosar 18/1944, f. 209 212
16 A.M.R., Fond 359, Dosar 20/1944, f. 112
63
AFASES - 2008 -
iarn, ploaie sau noroi cnd nici mcar nu se putea zbura. Alt lucru comun era studierea
minuioas a adversarului aerien i terestru.
De asemenea, pstrarea n secret a aciunilor aeriene fceau ca ambele comandamente
s lucreze i pe baz de ncredere i secret militar. Incidentul reretabil din 26 aprilie 1945, n
care piloii romni de asalt au atacat, trupele sovietice devastnd un regiment de infanterie ce
a fost scos din linia de lupt deoarece hrile piloilor romni nu erau actualizate cu situaia
frontului, greeal a comandamentului sovietic recunoscut cu curaj i sim al responsabilitii
nct piloii romni au fost considerai nevinovai,17 este un exemplu de corectitudine. Aceste
schimbri n bine sunt consemnate i n revista Sburturii Romniei, numerele din anul
1945.
Din punct de vedere al concepiilor i terminologiei folosite, n cooperarea romno
sovietic, misiunile aeriene pentru ntreaga perioad analizat pot fi grupate n patru categorii:
a) misiuni aeriene mpotriva potenialului aerian al inamicului, b) misiuni de sprijin aerian
apropiat mpotriva forelor militare terestre i fluviale ale inamicului, c) misiuni de informaie
aerian, d) misiuni de legtur i transport aerian.
Marea unitate aeronautic romn a fost o for restrns numeric, dar capabil s
dezvolte aciuni aeriene eficiente n sprijinul armatelor romne i sovietice terestre. Generalul
aviator Emanoil Ionescu, aprecia c personalul care a ncadrat unitile era de prim
calitate, echipajele rzboite n condiii din cele mai grele i cu o bogat experien, moralul
ns era cam sczut din cauza strii de nesiguran.18
Superioritatea aerian sovietic pe frontul sovieto - german a nceput n vara anului
1944, astfel nct datorit numrului mare de avioane disponibile, deseori sovieticii foloseau
procedeul interzicerii aviaiei inamice pe direciile principale, sau blocarea aviaiei germane la
aerodromuri prin trimiterea aviaiei de vntoare pe aliniamente apropiate frontului,19 aciuni
aeriene numite n teoria aeronautic sovietic tactic de ntmpinare.
Analiza istoric obiectiv relev i faptul c relaiile militare cu comandamentele
sovietice nu au fost caracterizate doar de suspiciuni i conflicte. Relaia de ansamblu a
Aeronauticii Regale Romne cu aliatul sovietic, depind momentele grele ale doborrii fr
logic a unor avioane romne i dincolo de amenirile directe sau subnelese cu moartea a
personalului aeronautic romn i a familiilor acestora, relaiile interumane pot fi considerate
acceptabile pentru ambele pri. Aciunile militare desfurate de Corpul 1 Aerian Romn, n
ar sau dincolo de frontierele rii n Campania din Vest, au fost succese evidente finalizate
cu nfrngerea armatelor germane i maghiare.
Strategia general a sovieticilor de a impune regimul sovietic i factorii sovietizani n
Romnia au alterat relaiile militare de pe front i au destructurat trupele romne rmase n
ar. Ulterior nchisorile comuniste au fost umplute i cu militari romni a cror vin a fost
aceea de a fi luptat n Campania din Est. Am amintit deja din rndurile aeronauticii, cazurile
emblematice ale generalilor comandani avatori Gheorghe Jienescu i Ermil Gheorghiu ce au
fost nchii la Rmnicu Srat sau Gherla, la care putem aduga alte spee precum cutarea
nverunat de ctre sovietici a cpitanului aviator Horia Agarici* pentru a fi judecat ca i
criminal de rzboi, nchiderea cpitanul Ioan Di Cezare la Ploieti,20 fuga din ar a
cpitanului aviator n rezerv, prinul Constantin (Bzu) Cantacuzino, excluderea cpitanului
64
AFASES - 2008 -
aviator Mircea T. Bdulescu din rndurile aviaiei romne dei era cavaler al Ordinului Mihai
Viteazul, i exemplele ar putea continua.
Aeronautica Regal Romn a fcut dovada unei structuri funcionale viabile, capabile
s rezolve situaii de lupt dintre cele mai dificile, fiind verificat i maturizat n condiiile
grele ale rzboiului. Activitatea de pe front a fost ns influenat i de clauzele Conveniei de
Armistiiu a Romniei cu Naiunile Unite (12 Septembrie 1944), de negocierile Tratatul de
Pace (semnat la Paris n 10 februarie 1947) i de situaia Forelor Aeriene Regale Romne din
ar.
Capitan comandor dr. NICOARA MARIUS-ADRIAN eful catedrei de aviaie din SAFA
65
AFASES - 2008 -
Abstract:
The theoretical and practical limits reached in the classic guns development impose major
changes regarding the projectile propulsion. A new propulsion system can be the
electromagnetic propulsion, which uses the electricity as an energy source. In this paper it is
elaborated the mathematical model of the projectile motion, on the basis of the physical
principles that lay the foundation of the electromagnetic coilgun. By the solving of the
mathematical model it is obtained, finally, the variation of projectile velocity versus time. The
validation of the mathematical model is done by calling the experimental data met in the
international literature of specialty, are compared with theoretical results.
Capitolul I
Cu privire la tunurile electromagnetice cu bobin
1.1 Propulsia electromagnetic
Limitele teoretice i practice la care s-a ajuns impun schimbri revoluionare n ceea ce
privete propulsia proiectilelor. O soluie de viitor pentru creterea vitezei iniiale a proiectileor o
constituie propulsia electromagnetic, un sistem propulsiv total diferit de cel al tunurilor clasice.
Propulsia electomagnetic folosete electricitatea ca surs de energie, ns forma n care
energia este transferat proiectilului difer funcie de tipul constructiv al tunului electromagnetic.
n acest context sunt examinate mai multe direcii:
a. tunuri electromagnetice
- tunuri electromagnetice cu ine, n terminlogia occidental, railgun;
- tunuri electromagnetice cu bobin, numite coilgun n literatura de specialitate
occidental;
66
AFASES - 2008 -
Proiectil
Bobin conductoare
Direcia de lansare
Tub de ghidare
67
AFASES - 2008 -
Bobina conductoare
Proiectil acionat de
fora electromagnetic
k2
u
u0
R
Re
IGBT
Sursa de alimentare
Punte redresoare
Baterie de condensatori
68
AFASES - 2008 -
Capitolul II
Ecuaiile de micare ale proiectilului n tunul electromagnetic bobin cu reluctan
2.1 Forma ecuaiilor lui Lagrange
Pentru determinarea ecuaiilor de micare ale proiectilului se realizeaz un model
dinamic, asupra cruia se aplic metodele calcului variaional (lagrangean). Pentru stabilirea
lagrangeanului sistemului electromecanic [2] se identific mrimile caracteristice schimburilor
energetice n cmp electromagnetic, respectiv energiile cinetice i poteniale, care guverneaz
ansamblul condensator echivalent bobin armtur (proiectil).
Lagrangeanul sistemului electromecanic este o funcie scalar dat de diferena dintre
energia cinetic i energia potenial a sistemului [14]:
L = Tp + Tm Wq Wm .
(2.1)
Primii doi termeni ai ecuaiei (2.1) alctuiesc energia de micare, unde Tp reprezint
energia cinetic pe care o capt armtura n micare n interiorul bobinei, Tm energia cinetic
electromagnetic aprut ca efect al energiei de interaciune dintre bobina parcurs de curent i
cmpul magnetic. Energia potenial a sistemului este constituit din energia nmagazinat n
condensator Wq i Wm energia magnetic a bobinei.
n funcie de poziia armturii, lagrangeanul sistemului (2.1) va avea forme diferite i
pentru prima poziie la intrarea n bobin se demonstreaz:
1
Q2
l + 2 2
L = m v z2 + a 2 N Q& B
(2.2)
B + a2 S z
a 2
2
2C
2 0
De la o poziie la alta a proiectilului, forma lagrangeanului este diferit, ca rezultat al
evoluiei forei electromagnetice n cursul micrii. Prezint interes practic numai prima poziie
fundamental a proiectilului, i anume, poziia proiectilului la intrarea n bobina conductoare.
Rezultatele stabilite permit obinerea ecuaiilor de micare ale armturii cu ajutorul unui
sistem de ecuaii cu derivate pariale de ordinul doi, sistemul ecuaiilor lui Lagrange [1]:
d L L
= fj ,
(2.3)
dt x& j x j
unde x j reprezint una din coordonatele generalizate ale sistemului ( B , z sau Q ) i f j
semnific orice for extern generalizat (de exemplu, forele de frecare sau pierderile rezistive).
Prin diferenierea lagrangeanului (2.2) n funcie de coordonata generalizat aferent i
substituind pentru forele externe disipative, n funcie de poziia armturii se obin ecuaiile
micrii. Pentru poziia armturii aflat la intrarea n bobin [2]:
1 B
B
(2.4)
N Q& = ( l + 2 ) z M 0 tanh r
0
r 0 M 0
Q
R Q& = a 2 N B& +
(2.5)
C
(2.6)
m &&
z = a2 S .
Dup cum se observ n ecuaia (2.6) fora exercitat asupra proiectilului este
proporional cu seciunea transversal a proiectilului a i cu densitatea energiei magnetice S n
69
AFASES - 2008 -
interiorul armturii. Termenul densitate a energiei magnetice S are o comportare similar unei
presiuni)[2]. Valoarea seciunii transversale a proiectilului se stabilete constructiv, iar S = S (B ) :
S=
1 B
r 0 2
M 0 ln cosh 0
.
r 1
r 0 M 0
(2.7)
300
X: 0.0016
Y: 226.2
250
200
X: 0.0055
Y: 169.1
150
100
50
0.001 0.002
0.003 0.004
0. 008 0.009
0.01
70
R
t
L
(2.8)
AFASES - 2008 -
Acest lucru se ntmpl dup t f = 0, 0055 s, conform modelului experimental folosit drept
comparaie. Dup cum se observ, ulterior momentului t f , graficul are o cdere exponenial
puternic [4].
2.4 Acceleraia i viteza proiectilului n interiorul bobinei conductoare
Pentru a calcula acceleraia proiectilului n lansator,
a2 S
&&
.
(2.9)
z=
m
trebuie stabilit dependena de timp a presiunii magnetice S exercitate asupra armturii [2], n
baza ecuaiei (2.7).
ntreruperea alimentrii cu energie electric se alege funcie de momentul n care
proiectilul ajunge cu vrful aproape de ieirea din bobin. Dac nu s-ar ntrerupe alimentarea cu
energie electric, fora electromagnetic i inverseaz sensul, devenind contrar vitezei
proiectilului i acionnd n sensul atragerii acestuia spre mijlocul bobinei conductoare.
Din ecuaia acceleraiei proiectilului (2.9), prin integrare ntre limitele corespunztoare, se
obine viteza la gura evii lansatorului:
a2 S
(2.10)
t.
v0 =
m
Relaia (2.10) permite calculul vitezei proiectilului n interiorul bobinei conductoare sau la
ieirea din lansator, prin stabilirea corespunztoare a timpului dup care se face integrarea. n
figura 4 este ridicat graficul vitezei proiectilului n interiorul bobinei conductoare.
12
X: 0.0055
Y: 11.68
10
3
Timp (s)
6
-3
x 10
71
AFASES - 2008 -
2.5 Concluzii
Prezentul studiu elaboreaz un model matematic al micrii unui proiectil, ntr-un tun
electromagnetic bobin cu reluctan. Analiza micrii s-a fcut n baza unor ipoteze
simplificatoare, alese astfel nct s reduc din complexitatea fenomenelor fizice implicate, dar
care s permit prin modelul creat o aproximare ct mai realist a micrii.
Pentru validarea rezultatelor analitice obinute pe modelul teoretic ales, s-a fcut apel la
un model similar experimental existent n literatura de specialitate occidental.
Graficul intensitii curentului prin bobina conductoare este ridicat n prezena unor
proiectile confecionate din materiale feromagnetice diferite. Se observ similitudinea graficelor
pe poriunea de ascenden spre valoarea maxim, corespunztor unei durate aflate ntre 1 la 2
milisecunde. Similitudinea se produce deoarece proiectilul nc nu a nceput deplasarea sau s-a
deplasat foarte puin n interiorul bobinei. Dup ce intensitatea curentului ajunge la valoarea
maxim, graficele ncep s se separe, proiectilul avanseaz n interiorul bobinei conductoare i
ncep s se fac resimite influenele saturaiei magnetice. Odat cu apariia micrii proiectilului
ncepe s se extrag energie din cmpul magnetic la o rat semnificativ [4]. Armtura din ferit
uor saturat are o eficien mai redus, se distinge o scdere uoar a curentului prin nfurrile
bobinei. n schimb, la proiectilul din oel i cel nesaturat rata de extragere a energiei magnetice
este mare, cderea curentului este mult mai mare, pn n momentul n care se decide
ntreruperea alimentrii cu energie electric. Acest moment este ales n mod diferit pentru fiecare
proiectil n parte, n scopul obinerii unei viteze ct mai ridicate a proiectilului la gura evii.
n figura 5 sunt redate graficele experimentale ale vitezelor realizate n eava tunului
electromagnetic bobin cu reluctan, obinut pentru trei tipuri diferite de materiale ale
armturii.
72
AFASES - 2008 -
Bibliografie
[1] Novacu, V., Bazele teoretice ale fizicii, vol.I, Editura Tehnic, Bucureti, 1990
[2] Slade, G., W., A Simple Unified Physical Model for a Reluctance Accelerator, IEEE
Transactions Magnetics, vol. 41, nr. 11, 2005
[3] Slade, G., W., Notes on FEA modeling of Coilgun Mass Accelerator,
http://users.skynet.be/BillsPage/MassAccelerators.html, 2006
[4] Slade, G., W., Fast Finite Element Solver for a Reluctance Mass Accelerator, IEEE
Transactions Magnetics, vol. 42, nr. 9, 2006
[5] Wang Ying, Richard A. Marshall, Cheng Shukang, Physics of Electric Launch,
Science Press, Beijing, 2004
Col. (r) doctorand Nicu Mihai, tel 0723 728168, e-mail: mihain53@yahoo.com
Gl.bg.(r) prof. dr. ing. Vasile Titic, Academia Tehnic Militar , B-dul George Cobuc, 81-83, Sector 5,
Bucureti, tel 021 3354660 int. 157, fax 021 3355763, e-mail: titica@mta.ro
73
AFASES - 2008 -
74
AFASES - 2008 -
G As
P1 = 1 e
(2)
where, at is target area, As is area that will be analyzed and is the visual field area in the object
plan, t is the observation time, G is the congestion coefficient having a value from 1 to 10 for
majority targets of interest.
The display object observation time is calculated on the basis of the following
assumptions: the approximate time for one looking (integration time) is from 0.1 to 0.3s, the
instantaneous visual field one looking is 5o.
The observation time for a display that is seen in a visual field 16o 16o:
16 o 16 o
t = o o 0.3 = 3.1s
(3)
5
5
The probability P2 for threshold contrast Cp is by definition 50%. The probability P2
could be calculated using the following equation for the other value of threshold contrast [5]:
C
4.2
1
Cp
1 1
1 e
(4)
2 2
Where the sign minus is used for C < Cp.
The factor from eq. (1) is:
1 e (S / Z 1)
S /Z 1
(5)
=
S /Z <1
0
Where S/Z is signal/noise ratio.
Therefore, using eq. (1), (2), (4) and (5), the equation that gives us the mathematics
expression for detection probability of a target is:
P2
700 at
4.2
1
1 1
C
G
A
Pd = 1 e s
1 e p
2 2
1 e ( S / Z 1)
(6)
(7)
where t is target reflection coefficient and b is the background reflection coefficient.
The contrast for thermal devices is:
T Tb
C= t
(8)
Tt + Tb
where Tt is the mean temperature of the target and Tb is the background average
temperature.
When a night vision device with image intensifier is used, the resolution that must be
attained by the optoelectronic system with the certain probability is given by [2]:
R d [lp / mm] =
S fk t Ta To E t t
S / Z FA
10 3
(9)
75
AFASES - 2008 -
time, is the minimum size /maximum target size ratio, Sfk is the photocathode sensibility of
the image intensifier, Et is the target illumination, and FA is the image intensifier noise factor.
The limit range, for the detection probability is given by eq. (10):
H
(10)
Rd f ob
Ld =
M
where fob is the focal length of the objective.
If you have a thermal device than maximum detection range will be given by
eq. (11):
Ld ,TIR =
(11)
Ad f
where is the Boltzman constant, is the target emissivity, At, Aob Ad are the areas of the
target, objective and detector, D* is the sensor detectivity and f is the bandwidth.
III. NUMERICAL SIMULATION
The variation of the detection probability (eq.2) with observing time is calculated and
plotted in MathCad (see fig.1 and 2):
G1 := 1
G2 := 5
G3 := 7
G4 := 10
a := 4 m
A := 2000 m
t := 0 , 0.1.. 50 s
P1( t) := 1 exp
700 a
P2( t) := 1 exp
700 a
P3( t) := 1 exp
700 a
P4( t) := 1 exp
700 a
G1
G2
G3
G4
A t
A t
A t
A t
76
AFASES - 2008 -
1.05
0.84
P 1( t )
P 2( t )
0.63
P 3( t )
0.42
P 4( t )
0.21
12
15
18
21
24
27
30
Fig. 1 Detection probability vs observation time and no range variation but with
different congestion coefficient
1
0.9
P 940( t )
P 890( t )
0.8
0.7
P 840( t )
P 790( t ) 0.6
P 740( t )
P 690( t )
P 640( t )
0.5
0.4
P 590( t )
P 540( t ) 0.3
P 490( t )
0.2
0.1
10
11
12 13
14
15 16
17
18 19
20
21 22
23
24 25
26
27 28
29
30
The VIS and IR (for example, see fig.3) images were projected on a 20 inch display,
1280x1024 pixels resolution and optimum luminance. We approximated the optimum
luminance to 1.5-3.0 cd/m2 measured at the observer. The separation angle of the eye is taken
to be 1 minute.
The images are skipped decreasing the distance to target, at t time each. Each observer
notifies the moment when a discrimination task is fulfilled.
77
AFASES - 2008 -
Results:
The aircraft target prediction could be realistic estimate with the help of the equation
(6), and, also, is useful to show the detection range of the target. This equation was developed
using both experimental measurements and numerical simulations and it is based on a
probabilistic model. Detection range of an aircraft target could be negatively influenced by
the atmospheric conditions represented in the eq. (10) and (11) only through atmospheric
transmission.
Bibliography
[1] H. H. Bailey, Target Detection Through Visual Recognition: A quantitative Model, Rand
Corporation, Santa Monica, Ca., February1970.
[2] C. Plea, E. Creu, "Considerations on the implementation of a recognition model for a
target for the modern optoelectronic systems" (in Romanian), Journal of the Military
Equipment and Technologies Research Agency (in Romanian), Iss.2/2003.
[3] C. Plea, The use of infrared radiation for thermal signatures determination of ground
targets", Communication to the 5th international Balkan Workshop on Applied Physics,
Constanta, Romania, July 5-7, 2004.
[4] C. Plea, Contributions to the calculation and construction of night vision systems with
image intensifiers, Ph. D. Thesis, Military Technical Academy, 2006.
78
AFASES - 2008 -
Mr.dr.ing. Plea Cornel , Col. conf. univ. dr. ing. Coereanu Liviu , Lt. drd. ing. Turcanu Daniel - Agenia de
Cercetare pentru Tehnic i Tehnologii Militare, Str. Aeroportului nr. 16, CP19 OP BRAGADIRU, jude Ilfov
cod 077025, liviuc@acttm.ro, cplesa@acttm.ro, edturcanu@acttm.ro
79
AFASES - 2008 -
Abstract
Helicopters are now quite common throughout the world, especially in the
vicinity of capital cities. In routine rescue operations police or emergency
service helicopters are those most likely to be called to assist in search and
rescue operations, and they are the ones with which FRB crews will need to be
most familiar. In disasters/emergencies the aircraft of the Armed Forces may be deployed to
support the civil community, particularly as a result of floods or cyclones.
I. INTRODUCTION
Rescue operations often require crew and passengers to work and travel in
unfavourable conditions. Boats operate in sun, wind and rain and those travelling in boats are
in the open with little or no shelter. There is always the possibility of crew, passengers or the
person being rescued being affected by exposure to the elements. The conditions most likely
to have an effect are:
a. heat;
b. cold
c. moisture
1.1. HEAT
In hot weather those persons in an open boat will be subjected to direct
sunlight, light reflected from the water, and any light coloured surface such as
the boat hull. If the boat is stationary, the effect will be most obvious. While
moving the effect is less obvious because of the breeze created by the boat's
movement. In either situation, the end result can be similar; after a short period,
people can suffer sunburn and/or heat exhaustion.
A person suffering from heat exhaustion will often become pale and cold, have
temperature variations, be dizzy and nauseous. First aid for this condition is to
80
AFASES - 2008 -
81
AFASES - 2008 -
cool the casualty by providing shade and rest. A cold drink with added salts may assist.
When operating in hot weather, all crew should wear shady hats or caps and
sunglasses and cover up in loose fitting clothing, preferably of cotton. Minimise
exposure to the sun and heat, drink plenty of fluids and wear a good sunblock
lotion.
1.2. COLD
When it is raining, during winter when strong winds are blowing, or when
operating at night, cold will have an effect. The most common effect of cold
conditions
is discomfort,
but
if not checked,
discomfort can
extend to
hypothermia.
Hypothermia is basically a loss of body core temperature. The human body has
a body core temperature of about 37C. With prolonged exposure to cold,
people may tend to lose body heat faster than the body can cope. With this loss
of body heat will come a loss of efficiency, injury or even death. A loss of as
little as 2C in body core temperature will cause serious hypothermia.
Where there is any wind, a 'wind chill' effect can be created even when the still
air temperature is quite high. An example of wind chill effects is shown in
Figure 1.
Figure 1 Wind Chill GraphEquivalent Temperature Curves (Source, National SAR Manual)
Possible signs of hypothermia include extreme shivering loss of coordination,
mental rigidity, uncaring or flippant attitude, confusion and amnesia muscular
rigidity, collapse and unconsciousness.
The first aid for hypothermia involves gently rewarming the casualty by the
provision of shelter, rest, warm clothing, warm (never hot) drinks (no alcohol or
cigarettes), and huddling together to share body heat.
In cold weather operations, crew should wear layered clothing which can easily
be adjusted to changing temperatures. This clothing should be preferably of
wool, with a light windproof outer layer, gloves and wool hat or balaclava.
82
AFASES - 2008 -
1.3. MOISTURE
Crews are required to wade, work in the rain or mist and are frequently splashed by
waves or spray. When combined with cold and wet conditions, this can be extremely
uncomfortable and exhausting and will greatly increase the likelihood of
hypothermia.
Another problem relating to wet conditions is immersion. Persons trapped in
flooded buildings, marooned by flood water or the survivors of sunken boats
may have been immersed in water or exposed to cold for long periods. These
conditions may affect the survivability of the victim/s.
Studies have shown that an average person wearing a life jacket and
immersed in water has a life expectancy under various conditions, as shown in
Figure 2.
Figure 2Water Chill without Anti-exposure Suit (Source, National SAR Manual)
The above figures are intended as a guide only. There are many variables
relating to individuals and the effect of clothing.
Generally the precautions for wet conditions are similar to cold. Clothing
requirements are much the same with the exception of waterproof clothing.
Crews are advised to carry a change of clothing.
83
AFASES - 2008 -
84
AFASES - 2008 -
85
AFASES - 2008 -
Popa Ionel, sef lucr. univ. dr. ing., Academia Navala Mircea cel Batran Constanta, Str. Fulgerului, nr.1, email: ionelpopaac@yahoo.com
**
Popa Adrian, asist. univ. drd. ing., Academia Navala Mircea cel Batran Constanta, Str. Fulgerului, nr.1
***
Ristea Marian, prep. univ. drd. ing., Academia Navala Mircea cel Batran Constanta, Str. Fulgerului, nr.1, email: ristea_k_marian@yahoo.com
86
AFASES - 2008 -
1
2
87
AFASES - 2008 -
Zona de
nimicirere
Patriot
Avantaje
+ putere de foc circular
+ descoperire 3600
+ suprapunere ntre PATRIOT i HAWK
+ inexistena flancurilor descoperite
Dezavantaje
- inexistena triangulaiei la bruiaj
- putere de foc limitat mpotriva unei ameninri
dintr-o direcie
- imposibilitatea ntrebuinrii mpotriva TBM
Zona de
nimicire
Hawk
Zona de
nimicire
Patriot
Zona de
nimicire
Hawk
88
AFASES - 2008 -
Direcia
de atac
Zona
de de descoperire
Distane
PATRIOT
14.000m
HAWK
ROLAND
110km
3.000m
16km
150m
30m
89
170km
60m
AFASES - 2008 -
HAWK
40
Distane de nimicire
(n km)
25
ROLAND
80
50
PATRIOT
90
AFASES - 2008 -
acesta definete, n mare msur, poziia sistemelor PATRIOT pentru aprarea obiectivului
(FOOTPRINT).
Dac un sistem de rachete antiaeriene PATRIOT este planificat pentru aprarea
antiaerian mpotriva TBM, trebuie orientat pe direcia principal de ateptare a adversarului
PTL 6 . O surs de informaii extern, care ar putea transmite datele necesare va duce la
mbuntirea capacitii de reacie a sistemelor PATRIOT mpotriva TBM.
n cazul ntrebuinrii de ctre inamic a luptei electronice, la planificarea aprrii
aeriene cu baza la sol n sistem CLUSTER trebuie avut n atenie, ca cel puin 2 uniti de foc
PATRIOT s aib zonele de nimicire suprapuse, pentru a asigura o corelare a acestora. La
folosirea capacitii de triangulaie a PATRIOT trebuie ca cel puin 3 uniti de foc s aib
zonele de nimicire suprapuse, n msura n care limitrile n zona de descoperire a radar sunt
susinute.
Pentru sistemul de rachete sol-aer PATRIOT trebuie stabilit pe lng PTL i cte o
7
STL .
3. Concluzii
Cu toate c sistemele de rachete sol-aer au n compunere elemente ce asigur
descoperirea, identificarea, urmrirea i combaterea intelor aeriene, precum i instalaiile de
conducere necesare ducerii luptei independente cu inamicul aerian, acestea sunt ntrebuinate,
de regul, ntr-o combinaie de mai multe sisteme de foc de acelai tip sau diferite. Prin
aceasta se pot obine o serie de avantaje, precum: creterea puterii de foc prin folosirea
sinergetic a diferitelor capaciti, compensarea deficienelor/limitelor tehnice ale unui sistem
cu posibilitile altuia, mbuntirea capacitii de coordonare a conducerii, crearea unei
nalte flexibiliti, obligarea inamicului aerian de a se ntrebuina suplimentar, precum i
aprarea reciproc a diferitelor sisteme.
BIBLIOGRAFIE
1. Planificarea zonei de ntrebuinare n lupt a rachetelor sol-aer (Planificarea
CLUSTER) / Lernunterlage Planung der Flugabwehrraketeneinsatzzone
(CLUSTER-Planung) Fhrungsakadenie der Bundeswehr, decembrie 2000
2. Conceptul conducerii i ntrebuinrii n lupt a unitilor de rachete sol-aer ale
forelor aeriene / Konzept fr Fhrung und Einsatz der FlaRak-Verbnde der
Luftwaffe Bundesministerium der Verteidigung, FL III 2 AZ 31-01-01, 1994
3. Instruciuni pentru planificarea ntrebuinrii n cadrul zonelor de responsabilitate
a rachetelor sol-aer / Anweisung fr die Einsatzplanung in der FlaRak-Einsatzzone
LwFKdo A 3 II, octombrie 1996
Marcel Preda, locotenent colonel, Academia Tehnic Militar, Bucureti, Bd. George Cobuc nr.81-83, tel.
021 335 46 60 / 186, e-mail: spanish_mark@yahoo.com
Eugen Boamb, colonel, Academia Tehnic Militar, Bucureti, Bd. George Cobuc nr.81-83, tel.
021 335 46 60 / 212, e-mail: eboamba@mta.ro
6
7
91
AFASES - 2008 -
I.
Supremaie aerian capacitatea forelor aeriene proprii de a aciona nestingherit asupra inamicului, iar forele
aeriene ale acestuia nu mai au posibilitatea de a periclita propriul potenial
Suveranitate aerian capacitatea, ca ntr-un spaiu de timp bine definit sau ntr-o zon bine delimitat a
spaiului aerian, deasupra teritoriului naional, forele aeriene proprii s poat desfura operaii fr o aciune
consistent a inamicului aerian, iar pe de alt parte, acestuia s-i fie limitat libertatea de aciune.
92
AFASES - 2008 -
Aprarea aerian activ cuprinde toate msurile i activitile de aprare direct sau de
atenuare a atacurilor aeriene executate de ctre inamic. n cadrul aprrii aeriene active sunt
ntrebuinate urmtoarele mijloace de rzboi aerian: sistemul de conducere, mijloacele
desupraveghere a spaiului aerian, mijloacele de lupt aerian (aviaia aer-aer) i mijloacele de
lupt aerian bazate la sol (GBAD 2 ).
n cadrul aprrii aeriene active, un rol deosebit l au mijloacele de lupt aerian
bazate la sol sistemele de rachete antiaeriene.
Misiunea de baz a sistemelor/unitilor 3 de rachete antiaeriene, ca fore principale de
aprare antiaerian, este contribuia la aprarea spaiului aerian n cadrul conceptului aprrii
antiaeriene integrate.
II. Caracteristici ale unor sisteme de rachete antiaeriene de producie vestic
1. Sistemul de rachete antiaeriene PATRIOT
Sistemul de rachete antiaeriene PATRIOT (fig.1)
este un sistem modern i foarte mobil, de generaia a
treia, dezvoltat n anii 70, care acoper, ntr-un sector
bine delimitat, o gam variat de nlimi 4 : de la cele
joase, pn la foarte mari. Acesta este dislocabil cu
ajutorul mijloacelor aeriene de transport strategic i
capabil de a duce aciuni de lupt n orice condiii de
stare a vremii. Totodat dispune de protecie, ntr-o
foarte mare msur, mpotriva bruiajului electronic.
Sistemul dispune de un radar multifuncional Fig.1 Sistemul de rachete
antiaeriene PATRIOT
(cutare, descoperire, identificare, urmrire i iluminare a
0
intei). Sectorul de cutare este n plan orizontal de 90 , iar urmrirea intelor aeriene i
dirijarea rachetelor n timpul procesului de combatere sunt posibile n plan orizontal ntr-un
sector (Track-Sector) de 1200.
Cu cele 32 de rachete (o unitate de foc 5 ) ale ncrcturii maxime pentru cele 8
instalaii de lansare ale unei uniti, pot fi combtute concomitent 9 inte. Suplimentar
sistemul dispune de nc o unitate de foc (32 de rachete), ca ncrctur de rezerv n cadrul
dispozitivului de lupt.
Sistemul de rachete antiaeriene PATRIOT poate s descopere i s combat, n
condiii limitate, n modul de lucru ATBM 6 , rachete balistice tactice. De asemenea are
posibilitatea combaterii rachetelor antiradiolocaie (ARM 7 ).
Pentru a putea combate TBM, se activeaz un sector special de cutare TBM, cu
dezactivarea sectorului de descoperire pentru distane mari (fig.2). Sectorul de cutare TBM
93
AFASES - 2008 -
acoper n nlare o zon cuprins ntre cca. 20-900 i n azimut este posibil un cmp ntre +/300 pe direcia principal de aciune.
132 Km
89
TBM
sector de
Search
cutare
Sector
15
25
Limita
LongLimita
Range ABT
apropiat
Short
Range
ndeprtat
87 Km
170 Km
94
AFASES - 2008 -
probabilitate de nimicire mai mare de 90%, o rachet antiaerian ROLAND are o raz de
aciune a exploziei la int de aproximativ 6 metri.
Un punct de comand antiaerian ROLAND poate conduce pn la 10 instalaii de
rachete antiaeriene i poate acoperi cu ajutorul radarului suplimentar o zon de descoperire de
pn la 60 Km.
Sistemul de rachete antiaeriene ROLAND poate uor transportabil cu ajutorul
avioanelor de transport C-130/C-160.
Combaterea rachetelor cu aripi de ctre rachetele antiaeriene ROLAND nu este
posibil.
4. Sistemul de rachete antiaeriene STINGER
Sistemul de rachete antiaeriene STINGER (fig.5)
este un sistem antiaerian portabil, care este ntrebuinat
pentru aprarea obiectivelor de dimensiuni mici
mpotriva atacurilor din aer, n gama de nlimi foarte
joase i joase. n plus acesta ofer premise ideale pentru
protecia antiaerian a unitilor pe timpul deplasrilor
(manevrelor).
Fig.5 Sistemul de rachete
antiaeriene STINGER
Instalaia STINGER este orientat optic, iar racheta
se lanseaz pasiv pe inta aerian dup principiul cutrii
termice i se dirijeaz dup lansare independent n int.
La o distan ntre 0,4 i 6 Km. instalaia dispune de o posibilitate de lovire foarte
mare, iar prin folosirea unui dispozitiv de vedere nocturn, devine uor de ntrebuinat i pe
timp de noapte. O posibilitate de identificare electronic, precum i capacitatea de
ntrebuinare n orice condiii de stare a vremii nu sunt totui posibile.
5. Sistemul de rachete antiaeriene al noii generaii MEADS
MEADS 11 este un sistem de rachete antiaeriene foarte mobil i uor dislocabil pe cale
aerian, care ar putea combate n viitor ntreg spectrul de ameninri aeriene. Modularitatea i
arhitectura deschis a sistemului fac posibil o alctuire flexibil din sensori, tipuri de rachete
i elemente de conducere a luptei. De aceea acest sistem este potrivit n special pentru
ntrebuinarea n situaii limit pe timpul situaiilor de criz.
Proiectul de nzestrare trinaional MEADS (SUA, Italia, Germania) ar trebui s
nglobeze ntr-un sistem toate capacitile propuse n jurul anului 2012, potrivit planificrii
fcute.
MEADS va mbunti considerabil capacitatea de aprare mpotriva rachetelor
balistice tactice i ar trebui s constituie baza pentru un sistem de aprare aerian al noii
generaii, ai crui parametri s uneasc actualele 3 sisteme de rachete antiaeriene PATRIOT,
HAWK i ROLAND.
III. Modaliti de ntrebuinare n lupt a sistemelor de rachete antiaeriene
Rachetele antiaeriene sunt mijloace de lupt cu reacie rapid ale aprrii antiaeriene,
care pot fi meninute pe perioade de timp mari ntr-o trapt ridicat de pregtire pentru lupt.
Pentru c sistemele de rachete antiaeriene PATRIOT i HAWK, datorit
caracteristicilor specifice de performan (zon eficace n plan orizontal i vertical, timpul de
11
95
AFASES - 2008 -
reacie i gradul de mobilitate), nu pot acoperi toate gamele de distane i nlimi sau toate
categoriile de ameninri, acestea sunt, n principiu, ntrebuinate mpreun n zona de
ntrebuinare n lupt a rachetelor antiaeriene sau aa-numitul CLUSTER sau pentru
aprarea antiaerian a grupelor de obiective. n funcie de situaie pot fi ntrebuinate
suplimentar sistemele de rachete antiaeriene ROLAND i STINGER, pentru protecia
(aprarea) obiectivelor izolate sau completarea zonelor de descoperire i de nimicire ale
sistemelor PATRIOT i HAWK.
ntrebuinarea eficace a sistemelor de rachete antiaeriene n CLUSTER, respectiv
aprarea antiaerian nemijlocit a grupelor de obiective presupune:
- eficacitate circular cu concentrare pe direcia principal de ameninare, de la
nlimi de zbor ale aeronavelor inamice joase pn la cele mari, n toate condiiile
de stare a vremii, precum i sub influena bruiajului electronic sau a armelor de
nimicire n mas;
- probabiliti de nimicire ridicate;
- capacitate de combatere simultan a mai multor inte aeriene;
- mobilitate ridicat;
- capacitatea, ca la pierderi s poat lupta coordonat cu prile intacte ale sistemelor.
Conducerea grupelor de rachete antiaeriene i meninerea interoperabilitii ntre
sisteme se realizeaz n principiu prin intermediul unui punct dee comand ala rachetelor
antiaeriene SAMOC 12 .
Sistemul de aprare aerian de tip CLUSTER poate fi ntrebuinat, n funcie de
situaia creat, pentru aprarea antiaerian a unor raioane importante, obiective sau grupe de
obiective i a operaiilor forelor terestre. Folosirea rachetelor antiaeriene n cadrul operaiilor
mobile ale forelor terestre ridic mari probleme n ceea ce privete conducerea unitilor de
aprare antiaerian respective.
Pe lng ntrebuinarea n cadrul CLUSTER sistemele de rachete antiaeriene
ROLAND poate fi ntrebuinat i de sine-stttor pentru aprarea antiaerian a unor puncte
dinainte stabilite (fig.6), pentru aprarea antiaerian nemijlocit a unor obiective de
dimensiuni mici (fig.7) i pentru aprarea antiaerian a coloanelor pe timpul manevrelor
(fig.8).
max. 30km
cca.40km
max. 30 km
cca. 40km
96
AFASES - 2008 -
Marcel Preda, locotenent colonel, Academia Tehnic Militar, Bucureti, Bd. George Cobuc nr.81-83, tel.
335 46 60 / 186 e-mail: spanish_mark@yahoo.com
97
AFASES - 2008 -
PERSPECTIVELE PARTENERIATULUI
PUTERE AERIAN PUTERE TERESTR
Marius erbeszki
Aciunea ntrunit, privit cel puin ca un ideal, este greu de neluat n seam ns,
nainte de a decide nceperea unui parteneriat aer-sol, trebuie gsit rspunsul la o ntrebare
care s pun n discuie dac o asemenea integrare este fezabil. Obstacolele, din acest punct
de vedere, sunt substaniale. Categoriile de fore trebuie s renune la tradiionala definire a
relaiei de comand n termenii comandant sprijinit i comandant care sprijin. Mai
mult dect att, adevrata integrare necesit comandani versai n operaii aeriene, terestre
i ntrunite, beneficiind de o pregtire, educaie i experien ce puini ofieri o au n prezent.
Susintorii pasionai ai puterii aeriene, de regul, nu vorbesc sau scriu prea multe
despre rolul acesteia n atacurile mpotriva forelor terestre. Cu cteva excepii, superioritatea
aerian i atacurile strategice constituie temele dominante n lucrrile ce abordeaz teoria
puterii aeriene. Cu toate acestea, nimicirea forelor militare ale inamicului a constituit, de
regul, o condiie necesar pentru obinerea victoriei militare i, deseori, a fost i suficient.
Atacurile mpotriva forelor terestre inamice nu reprezint singura modalitate prin care
puterea aerian poate contribui la nfrngerea inamicului obinerea superioritii aeriene are
o importan critic pentru aciunile desfurate n spaiul de lupt i reprezint, de regul, o
condiie necesar pentru asigurarea eficacitii operaiilor aeriene mpotriva forelor de
suprafa, iar atacurile strategice, la rndul lor, pot diminua substanial capacitatea de lupt a
inamicului. Chiar dac nu sunt privite cu mult entuziasm de ctre aviatori, atacurile aeriene
mpotriva forelor terestre au jucat, n majoritatea cazurilor, rolul principal n obinerea
succesului militar. Misiunile de interdicie aerian i sprijin aerian apropiat executate de aliai
n cel de-al doilea rzboi mondial au contribuit substanial la eliberarea Europei de vest. n
Vietnam, puterea aerian a oprit Ofensiva de Pati din anul 1972. De asemenea, puterea
aerian ndreptat mpotriva forelor terestre a neutralizat rapid armata irakian n 1991,
forele talibane din Afganistan n 2001 i, din nou, armata irakian n 2003, de fiecare dat
conducnd la obinerea unor rezultate decisive. nfrngerea Serbiei n anul 1999 constituie o
excepie de la acest mod de aciune, excepie ce pare s ntreasc regula.
n unele cazuri, forele terestre inamice pot reprezenta singurele inte - acceptabile din
punct de vedere politic mpotriva crora pot fi executate atacuri aeriene sau singurele inte
care s merite executarea unor atacuri pe scar mare. De asemenea, politicienii i strategii
98
AFASES - 2008 -
militari tind s opteze pentru astfel de atacuri, considerndu-le acceptabile din punct de
vedere politic i legal. Legile conflictelor armate interzic atacurile asupra intelor civile, iar
atacurile executate mpotriva unor obiective cu dubl utilizare, civil i militar, pot atrage
dup sine efecte negative semnificative la nivel politic. O astfel de problem devine i mai
acut n cazul rzboaielor de coaliie la care particip mai multe state. Fr ndoial, atacarea
forelor militare inamice este considerat att moral, ct i legitim, ceea ce i determin pe
liderii politici s le autorizeze cu uurin i s delege conducerea executrii lor comandanilor
militari. Totui, n unele circumstane, chiar i n cazul atacurilor executate mpotriva forelor
militare inamice pot apare unele consideraii de ordin moral, n mod particular atunci cnd
aceste fore sunt formate din personal recrutat cu fora de ctre regimuri dictatoriale.
Abilitatea puterii aeriene de a nimici forele terestre a crescut de la generaie la
generaie, pe msur ce s-au perfecionat aeronavele, senzorii i muniiile. ncepnd deja din
timpul celui de-al doilea rzboi mondial, puterea aerian a reprezentat o for de temut
mpotriva unitilor terestre. Cu toate acestea, capabilitile puterii aeriene n ceea ce privete
atacarea intelor terestre erau limitate din multe puncte de vedere. Spre exemplu, pe timp de
noapte sau n condiii meteorologice dificile, forele terestre erau imune la atacurile din aer,
vehiculele blindate puteau fi atacate din aer ns erau greu de nimicit, iar nimicirea forelor
terestre inamice cu un anumit grad de precizie necesita ca avioanele de atac la sol s se
apropie destul de mult de acestea ceea ce le fcea extrem de vulnerabile la focul artileriei
antiaeriene, provocnd pierderi n rndul forelor aeriene, pierderi ce astzi ni s-ar prea
inacceptabile.
n prezent, situaia este mult diferit n sens pozitiv; cu toate acestea, puterea aerian
are nc limitri atunci cnd este ntrebuinat mpotriva forelor terestre. Capabilitile de atac
pe timp de noapte au crescut foarte mult odat cu dezvoltarea unor senzori perfecionai.
Astfel, putem afirma c intele terestre sunt uneori mai vulnerabile la atacurile din aer
executate pe timp de noapte dect cele executate ziua, iar capabilitile de atac n condiii
meteorologice dificile au fost mult mbuntite odat cu apariia muniiilor dirijate prin
satelit. Muniia dirijat cu precizie a fcut ca atacul unor inte punctiforme - cum ar fi
vehiculele blindate - s fie mult mai eficient i, totodat, a permis aeronavelor s execute
atacuri de la altitudini i distane mari, reducndu-se astfel pierderile datorate aciunilor
mijloacelor de aprare aerian ale inamicului, precum i executarea unor atacuri strategice cu
grad ridicat de discriminare.
Desigur, n prezent i forele terestre dispun de un potenial de lupt sporit. Spaiul de
lupt a devenit un loc mult mai periculos datorit capacitilor mbuntite de supraveghere,
mijloacelor moderne de ochire i sistemelor de arme perfecionate. Puterea aerian rmne
nc departe de a fi omnipotent, n special cnd se dorete lovirea unor inte greu de detectat
i identificat sau cnd se acioneaz n mediul urban sau n teren complex. Totui, eficacitatea
puterii aeriene mpotriva forelor terestre a crescut n asemenea msur nct deseori este
posibil s vorbim de numr de obiective nimicite ntr-o ieire avion n loc de numr de
ieiri-avion pentru nimicirea unui obiectiv, cel puin atunci cnd se acioneaz mpotriva
forelor terestre convenionale ale inamicului, cum ar fi unitile de blindate. Aceasta se
datoreaz faptului c, pe de o parte, abilitatea puterii aeriene de a ataca forele terestre inamice
a crescut mult mai repede dect abilitatea forelor terestre de a supravieui unui atac din aer,
iar pe de alt parte, abilitatea puterii aeriene de a nimici forele terestre a crescut mult mai
rapid dect abilitatea forelor terestre de a nimici forele terestre inamice.
Aceste transformri au produs o schimbare fundamental a rolului jucat de puterea
aerian mpotriva forelor terestre. Unitile din adncimea teritoriului inamic pot fi nimicite
chiar dac sunt dispuse n poziii defensive ntrite. Executarea cu eficien a misiunilor de
interdicie aerian poate preveni declanarea de ctre inamic a atacurilor sau contraatacurilor
mpotriva forelor proprii, reducnd astfel nevoile de executare a sprijinului aerian apropiat.
99
AFASES - 2008 -
Totodat, aceste schimbri au transformat sprijinul aerian apropiat ntr-o misiune mult mai
eficient dect a fost vreodat. Viitoarele doctrine i regulamente de lupt ale forelor terestre
i aeriene vor trebui s ia n considerare aceste schimbri.
Capabilitile crescnde ale puterii aeriene relativ la puterea terestr nu implic, totui,
c forele terestre au devenit inutile. Ele pot executa misiuni pe care aviaia nu are
posibilitatea s le execute sau chiar execut unele misiuni mult mai bine dect ar putea s le
execute aviaia.
Misiunile ce pot fi executate numai de forele terestre sunt cele care implic contactul
uman cu populaia local. Exist o gam mare de misiuni de acest gen ce se ntinde de la
aciuni de scotocire (tip caut i distruge) mpotriva formaiunilor de gheril sau
insurgenilor la administrarea teritoriului ocupat, de la colectarea de informaii (HUMINT) la
desfurarea unor activiti cu caracter constructiv, fie pentru ctigarea ncrederii populaiei
locale, fie n scopuri umanitare. Activitile care necesit contact uman tind s devin cele
mai importante pe timpul executrii misiunilor de contrainsurgen, stabilizare, meninere a
pcii, reconstrucie etc.
Pe lng aceste activiti, forele terestre dispun de avantaje semnificative fa de
puterea aerian, n special fa de aviaie (avioane cu arip fix). Cel mai semnificativ avantaj
l constituie capacitatea de a executa rapid focul mpotriva intelor aprute prin surprindere,
cum ar fi focul mpotriva artileriei inamice, atunci cnd artileria proprie dispune de raza de
aciune necesar. n aceste cazuri, cnd este necesar un timp de reacie sczut (de regul, dou
minute sau chiar mai puin), artileria poate da rezultate mai bune dect puterea aerian,
exceptnd cazul n care bazele de decolare ale aviaiei sunt dispuse n apropiere sau
mijloacele de aviaie execut patrularea de lupt n zona de aciune a forelor terestre. Spre
exemplu, pe timpul campaniei americane din Vietnamul de sud, misiunea de a executa foc de
rspuns rapid n sprijinul patrulelor surprinse n ambuscade a fost preluat de ctre un sistem
de subuniti de artilerie dislocate n teritoriu, renunndu-se n mod progresiv la meninerea
aeronavelor n misiuni de patrulare de lupt pentru asigurarea sprijinului aerian apropiat la
chemare. De asemenea, plasarea unor lupttori teretri proprii n apropierea inamicului este
deseori necesar pentru localizarea i indicarea intelor. Astfel, forele terestre au o contribuie
important la succesul atacurilor aeriene. Pe scurt, forele terestre rmn nc indispensabile
n spaiul de lupt modern, asigurnd anumite capabiliti ce nu pot fi nlocuite prin aciunile
puterii aeriene.
Pentru a spori mobilitatea strategic i tactic a forelor i a reduce necesarul de sprijin
logistic al acestora, armatele moderne i schimb structura de fore, punnd un accent mai
mare pe organizarea unor fore de dimensiuni medii - ce sunt mai vulnerabile la focul
inamicului dect forele blindate tradiionale i reducnd ponderea artileriei din organica
acestora. Aceste fore de dimensiuni medii i compenseaz puterea de foc redus prin
ntrebuinarea unor sisteme ce asigur un grad sporit de supraveghere i cercetare a spaiului
de lupt, folosind sisteme de arme cu precizie mare i exploatnd n avantaj propriu puterea
de foc asigurat rapid de puterea aerian. De asemenea, se urmrete ca aceste fore s poat
manevra mult mai rapid i s acioneze dispersat pe arii mult mai extinse dect n mod
tradiional i avnd o linie de front ce separ forele proprii de cele inamice mult mai slab
definit, un concept ce s-a pus deja n practic pe timpul invaziei Irakului din anul 2003.
Eficacitatea operaional a acestor structuri de fore terestre va depinde foarte mult de modul
n care ele vor coopera cu forele aeriene. n operaiile ntrunite, ele se vor baza mai mult ca
oricnd pe puterea de foc i protecia oferite de puterea aerian.
n timpul celui de-al doilea rzboi mondial, puterea aerian a avut o eficien
devastatoare mpotriva forelor terestre inamice, dar numai n anumite circumstane: atunci
cnd aciona pe timp de zi, n condiii meteorologice favorabile, iar inamicul se afla n
deplasare sau i deconspirase poziiile. n prezent, forele aeriene moderne au capacitatea de a
100
AFASES - 2008 -
ataca obiective pe timp de noapte i chiar i n condiii meteorologice dificile. Ele pot lansa
muniia de la bord - supunndu-se unui risc acceptabil de la distane mai mari dect raza de
aciune a artileriei antiaeriene sau a armamentului uor individual. Cu toate acestea,
descoperirea i identificarea forelor inamice rmne o problem dificil de rezolvat dac
acestea au posibilitatea de a se ascunde i masca (aa cum s-a ntmplat n Kosovo). De aceea,
o cretere a eficacitii aciunilor mpotriva inamicului se poate obine numai cu aportul
forelor terestre proprii care s foreze inamicul s-i deconspire poziiile i s ajute la
indicarea intelor.
n trecut, lansarea muniiei n apropierea forelor proprii era constrns de lipsa
informaiilor privind dispunerea inamicului i lipsa de acuratee a muniiei n sine. Grupele de
control aerian tactic i controlorii aerieni n poziii naintate aveau doar o idee general asupra
localizrii trupelor inamice i nu dispuneau de mijloace de indicare precis a poziiilor
acestora (altele dect marcarea cu grenade fumigene sau prin transmiterea ctre piloi a
poziiei inamicului prin referire la elemente de teren din apropiere). Mai mult dect att,
poziiile forelor proprii nu puteau fi transmise cu precizie, n special n situaiile des
schimbtoare din cmpul de lupt. De asemenea, lansarea muniiei se solda de multe ori cu
erori mai mari chiar i dect raza de aciune a acestora. Pentru a preveni fratricidul, distana
de siguran la care se lansa muniia trebuia s fie foarte mare, permind inamicului
manevrarea forelor sale i ncercuirea forelor proprii. n prezent, progresul tehnologic a
permis reducerea acestei distane la cteva sute de metri, iar controlorii aerieni n poziii
naintate au acces la informaiile provenite de la o serie ntreag de senzori, incluznd satelii,
aeronave fr pilot i chiar aeronavele care execut atacul propriu-zis, acestea fiind echipate
cu senzori i aparatur de ochire performante (pentru a permite aceasta, toi senzorii trebuie s
fie nglobai ntr-o reea comun care s permit diseminarea informaiilor n timp real).
Controlorii aerieni n poziii naintate dispun n prezent de posibilitatea ntrebuinrii
sistemelor GPS 1 pentru a determina coordonatele poziiilor forelor proprii i ale celor
inamice, precum i a marcatorilor i indicatorilor laser pentru a marca/indica intele inamice.
n Afganistan i Irak, forele coaliiei transmiteau constant informaiile privind poziiile
proprii ealoanelor superioare, reducnd astfel riscul producerii fratricidului. Hrile actuale
au devenit mult mai precise, rspunznd rapid cerinelor lupttorului prin distribuirea lor
electronic. Pentru a completa imaginea, lansarea muniiilor moderne se execut cu mult mai
mult precizie. Pe timpul Operaiei Anaconda (Afganistan), controlorii aerieni n poziii
naintate au aprobat lansarea muniiei n apropierea forelor proprii, la distane de cteva zeci
de metri, fr a se produce incidente fratricide.
Importana crescnd a operaiilor mpotriva forelor de suprafa nu scade din
importana atacurilor strategice, ns previne planificatorul realist s se bazeze pe o abordare a
rzboiului ce trateaz atacarea trupelor inamice din spaiul de lupt ca irelevant. Pot exista
situaii n care victoria militar s fie obinut folosind astfel de strategii (spre exemplu,
Operaia Desert Storm, 1991), iar aviatorii trebuie s fie pregtii pentru aceasta; totui,
situaiile respective vor rmne mai degrab o excepie de la regul dect regula n sine.
Letalitatea crescnd a puterii aeriene mpotriva trupelor terestre nu implic faptul c
forele aeriene nu ar trebui ntrebuinate niciodat pentru a augmenta puterea de foc a
acestora; totui, ea ne conduce la ideea c perspectiva de artilerie zburtoare a puterii
aeriene nu va maximiza capabilitile acesteia n operaiile mpotriva forelor terestre inamice.
Capacitatea puterii aeriene de a aciona n principal n misiuni de sprijin aerian
apropiat, sprijinit ocazional de subuniti de fore terestre, va fi limitat de contra-msurile
luate de inamic, cum ar fi dispersarea forelor i amestecarea acestora cu populaia civil.
Abilitatea puterii aeriene de a domina puterea terestr va fi, de asemenea, afectat de natura
conflictelor viitoare; operaiile de contra-insurgen, cele de sprijin al pcii i conflictele de
1
GPS = <engl.>, (abv.) Global Positioning System sistem de poziionare global bazat pe informaiile primite de la satelii
101
AFASES - 2008 -
mic intensitate necesit un rol major jucat de forele terestre, iar subordonarea acestora
puterii aeriene va fi rareori indicat. Pe de alt parte, reorganizarea forelor terestre n uniti
uoare, mult mai agile, ce vor opera n condiii des schimbtoare i fluide va face ca focul
asigurat de puterea aerian att n ofensiv, ct i n aprare s fie att de important nct
ntrebuinarea forelor aeriene exclusiv n rol de sprijin al schemei de manevr a trupelor
terestre va duce la diminuarea eficacitii echipei aero-terestre.
Perspectiva unui parteneriat cu drepturi egale ntre puterea aerian i cea terestr nu
va putea fi aplicat tuturor operaiilor ntrunite viitoare, dar ea rmne n prim-plan din cel
puin dou motive. n primul rnd, ea poate fi rapid ajustat pentru a permite creterea rolului
uneia sau alteia dintre cele dou puteri. n al doilea rnd, ea s-a bucurat de puin atenie din
partea teoreticienilor i practicienilor militari datorit dificultilor inerente punerii sale n
practic. n cele din urm, ea nu asigur superioritatea unei puteri fa de cealalt i, astfel, nu
servete intereselor particulare, limitate, ale unei singure categorii de fore.
n prezent i, foarte probabil, n viitor, operaiile mpotriva forelor terestre inamice la care vor participa Forele Aeriene romne - se vor desfura, de regul, n mod ntrunit i n
cadrul unor fore NATO sau al unei coaliii multinaionale. Fr a beneficia de prezena unor
fore terestre proprii care s asigure informaii precise despre inamic, s-l oblige pe acesta si deconspire poziiile i s exploateze rezultatele atacurilor aeriene, operaiile mpotriva
forelor terestre inamice se vor putea dovedi ineficiente. De aceea, caracterul ntrunit al
aciunilor de lupt trebuie s coboare la nivelele inferioare ale ierarhiei militare. Necesitatea
planificrii integrate a aciunilor aeriene i terestre se manifest de la cel mai nalt nivel pn
la cel mai mic. n prezent, grupele de control aerian tactic sunt prevzute a se nfiina pe timp
de rzboi la nivelul batalioanelor i brigzilor ns, n viitor, prezena lor ar putea fi necesar
chiar la nivelul companiilor sau al plutoanelor pentru a avea astfel posibilitatea de a transmite
la timp informaii actualizate despre spaiul de lupt i de a permite exploatarea la maxim a
capabilitilor puterii aeriene.
Chiar i n prezent, unele tipuri de uniti din armatele moderne desfoar deja aciuni
de lupt cu caracter ntrunit. Unitile de fore speciale nu au posibilitatea susinerii pe timp
ndelungat a capacitii de lupt proprii, bazndu-se pe puterea aerian pentru asigurarea
supravieuirii lor. Puterea aerian le transport la obiectiv, le aprovizioneaz, le asigur cu
informaii i cu putere de foc i le extrage dup executarea misiunii. n condiiile n care
unitile terestre i reduc dimensiunile n favoarea creterii mobilitii, este necesar ca acestea
s acioneze ntrunit cu forele aeriene. Pe timpul aciunilor de lupt din Afganistan, cel mai
mare ealon de fore terestre convenionale a fost reprezentat de o brigad american avnd n
subordine trei batalioane de dimensiuni reduse. De asemenea, n Irak (2003) au fost angajate
n lupt uniti de nivel brigad care au trebuit s-i integreze aciunile proprii cu atacurile
aeriene, nu numai pentru a preveni fratricidul, ci i pentru a minimiza pierderile n rndul
civililor, obinnd efectele dorite mpotriva trupelor irakiene.
Lupta n mediul urban impune, de asemenea, implementarea caracterului ntrunit al
aciunilor de lupt la nivelul ealoanelor inferioare. Mediul urban fragmenteaz spaiul de
lupt n spaii reduse ca dimensiuni, limitate uneori la cteva cldiri i intersecii de strzi,
comandanii avnd nevoie de echipe mici de arme ntrunite pentru a-i ndeplini misiunea. De
regul, aceste echipe includ tancuri, armament de asalt, mijloace de geniu i lupttori de
infanterie sprijinii de mijloacele de foc cu lovituri indirecte i elemente ale puterii aeriene.
Ele dispun de un cmp de vedere redus asupra terenului, uneori numai pn la urmtorul col
al unei cldiri, avnd nevoie de sprijin rapid - inclusiv cu informaii - pentru a reduce punctele
de rezisten inamice sau pentru a mpiedica ncercrile inamicului de a-i ntri poziiile.
Mijloacele aeriene pot zbura deasupra cmpului de lupt din mediul urban, contribuind la
transmiterea oportun a informaiilor privind manevrele inamicului i atacndu-l rapid atunci
cnd situaia la sol o cere. Pe timpul luptei n mediul urban, aeronavele de atac au posibilitatea
102
AFASES - 2008 -
Cpitan-comandor Marius erbeszki, lector universitar dr., Universitatea Naional de Aprare Carol I,
Bucureti, os. Panduri, nr. 68-72, tel. 021 319 48 80, int. 282, e-mail mserbeszki@unap.ro,
serbeszki@yahoo.com
103
AFASES - 2008 -
I Introducere
Inducerea n eroare a inamicului prin aciuni de dezinformare este un procedeu foarte
des ntrebuiat de forele beligerante, desfurndu-se pe plan politic, economic i militar
nclinnd astfel balana n favoarea dezvoltrii unor muniii de contramsuri electronice ct
mai performente astfel c se impune amplificarea cercetrilor n domeniul bruiajului pentru
dezvoltarea unui sistem ct mai performant de muniii de contramsuri electronice pentru
protecia obiectivelor de interes.
Muniiile de contramsuri electronice au capabilitatea de a materializa n cmpul tactic
capcane pentru radiolocaie (RL), capcane cu emisivitate preponderent n infrarou (IR) i
perdele de mascare (aerosoli fumigeni - AL) n mod rapid, eficient i coordonat pentru
bruierea sistemelor de detecie, urmrire i de dirijare a radarelor, a rachetelor cu cap de
dirijare specializate n acest domeniu.
Noile cerine tactice, operative i strategice au impus extinderea ariei cercetrii i n
domeniul bruiajului pentru executarea acestuia n spectrul electromagnetic de interes. n acest
sens Centrul de Cercetare tiinific pentru Forele Navale (C.C..F.N.) Constana din cadrul
Ageniei de Cercetare pentru Tehnic i Tehnologii Militare (A.C.T.T.M.) a pus un accent
104
AFASES - 2008 -
105
AFASES - 2008 -
106
AFASES - 2008 -
107
AFASES - 2008 -
108
AFASES - 2008 -
109
AFASES - 2008 -
110
AFASES - 2008 -
111
AFASES - 2008 -
Grenada reactiv cal.81mm pentru bruiaj mixt RL + IR pentru protecia blindatelor fig. 16 face parte din categoria muniiilor pentru contramsuri electronice combinate (att n
domeniul infrarou ct i n domeniul radiolocaiei) necesare proteciei blindatelor.
Grenada reactiv calibru 81mm pentru bruiaj mixt RL + fumigen Al pentru protecia
blindatelor - fig. 17 face parte din categoria muniiilor pentru contramsuri electronice
combinate (att n domeniul aerosolilor pentru mascare ct i n domeniul bruiajului de
radiolocaiei)
necesare
proteciei
blindatelor.
112
AFASES - 2008 -
V Concluzii
Muniia de contramsuri electronice de bruiaj IR, RL i AL realizat de C.C..F.N. i
prezentat n aceast lucrare asigur compatibilitatea i interschimbabilitatea operaional cu
muniia similar produs pe plan mondial ndeplinind standardele NATO pe linia siguranei n
exploatare, depozitare, manipulare i transport.
Bibliografie
[1] Lt. col. dr. ing. Florentin Moraru, (1976). Manual de balistic exterioar. Ed.
Militar,Bucureti.
[2] Col ing. Stelian Constantin, col. dr. ing. Tudor Niculescu, (1988). Confruntarea pe calea
undelor Ed. Militar, Bucureti.
113
AFASES - 2008 -
[3] Col. ing. Costin Ioan Janica, (1980). Construcia, exploatarea, efectele i calculul
muniiilor. Academia Militar.
[4] Stan Petrescu, (2003). Arta i puterea informaiilor. Ed Militar, Bucureti.
[5] Lt. col. Ion Marcu, (1980). Mecanica zborului rachetelor dirijate. Academia Militar.
114
AFASES - 2008 -
Abstract
Preparing and conducting military actions these days is unconceivable without the
large scale use of radioelectronic counteraction, becoming at present a prime component of
modern military actions.
The increase in quantity and amplification of the parameters of radioelectronic
means, as well as the necessity of knowing the discovery and neutralization of these,
simultaneous with the application of the protection means for our own means has imposed a
rigorous analysis of the tendencies for development and improvement of the way of
deployment and use of electronic countermeasures ammunitions, in that particular
electromagnetic spectrum, leading to the increased efficiency, operativity and precision of
radioelectronic combat actions conducted.
In the activity area of radioelectronic control, the Research Center for Navy has
developed an electronic countermeasures system (ammunition type CHAFF&FLARE) for
aircrafts type IAR-99 OIM, MiG 21 LanceR, PUMA and so on) that is capable of creating IR
and RL decoys in the tactical field in a fast, efficient and coordinated way (passive jamming
decoys - RL clouds and infrared-emitting decoys IR) for jamming the detection, tracking
and guidance systems of air-to-air or ground-to-air missiles.
I Introducere
Creterea cantitii i amplificarea parametrilor mijloacelor radioelectronice, precum i
necesitatea cunoaterii posibilitilor de cercetare i neutralizare a acestora, concomitent cu
aplicarea msurilor de protecie a propriilor mijloace a impus analiza temeinic a tendinelor
de dezvoltare i perfecionare, a modului de desfurare i ntrebuinare a muniiilor pentru
contramsuri electronice, n spectrul electromagnetic stabilit, conducnd la creterea
eficienei, operativitii i preciziei aciunilor de lupt radioelectronic desfurate.
Pregtirea i ducerea aciunilor militare este de neconceput, n zilele noastre, fr
ntrebuinarea, pe scar larg a mijloacelor de contraaciune radioelectronic devenind n
prezent o component de baz a aciunlior militare moderne.
115
AFASES - 2008 -
116
AFASES - 2008 -
117
AFASES - 2008 -
Dup lansarea T1
Momentul iniial T0
118
AFASES - 2008 -
119
AFASES - 2008 -
120
AFASES - 2008 -
121
AFASES - 2008 -
Fig. 16 Efectul
capcanei IR
asupra rachetei
122
AFASES - 2008 -
VI Concluzii
Muniia de contramsuri electronice pentru protecia aeronavelor este muniie de
categoria V cu o singur utilizare i numrul strilor posibile de funcionare este de tipul I cu
dou stri posibile de funcionare funcionare sau nefuncionare) face parte din categoria
muniiilor necesare proteciei obiectivelor de interes i este necesar pentru asigurarea
contramsurilor electronice mpotriva rachetelor cu dirijare/autodirijare n infrarou,
radiolocaie n mod rapid, eficient i coordonat. Aceste produse nu necesit costuri de
ntreinere i nici reparaii capitale pe durata de funcionare, n condiiile respectrii normelor
de depozitare.
Muniia de contramsuri electronice de bruiaj pentru aeronave realizat de C.C..F.N.
i prezentat n aceast lucrare asigur compatibilitatea i interschimbabilitatea operaional
cu muniia similar produs pe plan mondial ndeplinind standardele NATO pe linia
siguranei n exploatare, depozitare, manipulare i transport.
Bibliografie
[1] Col ing. Stelian Constantin, col. dr. ing. Tudor Niculescu, (1988). Confruntarea pe calea
undelor. Ed. Militar, Bucureti.
[2] Stan Petrescu, (2003). Arta i puterea informaiilor. Ed Militar, Bucureti.
[3] Lt. col. Ion Marcu, (1980). Mecanica zborului rachetelor dirijate. Academia Militar.
[4] *** Raport testare evaluare de dezvoltare , C.C..F.N. Constana dec. 2007.
[5] *** Strategia de securitate naional a Romniei, Bucureti, noe. 2001.
123
AFASES - 2008 -
124
AFASES - 2008 -
Fig. 1: The overall structure navigation, guidance and control loop in UAV
III. The Attitude Heading Reference System
In figure 1 the AHRS is represented by a GPS receiver combined with a low-power
consummation controller in the navigation loop. The controller has implemented an algorithm
based on Kalman filtering in order to calculate the attitude of UAV. In our study case was
chosen an Extended Kalman Filter (EKF), due to his simplicity in implementation.
One of the most difficult parts to estimate attitude is obtaining a measurement. A
number of different solutions to the attitude measurement problem have been proposed and
implemented. One popular method is to use the carrier phase of GPS signals. This involves at
least three GPS antennas with a known geometry. Once the phase ambiguity is resolved,
phase differences between antennas can be calculated, and a good estimate of attitude is
made. The solution improves as the baseline between antennas increases, making
implementation on small UAVs very difficult.
Perhaps the most straight-forward way to measure attitude is to use the accelerometers
as an inclinometer to give a measurement of roll ( ) and pitch ( ) with the direction of GPS
velocity used to measure heading ( ). Complications arise, however, due to the aircraft
acceleration in the reference frame. This means that the accelerometers will not measure the
gravity vector, rather they will measure the aircraft apparent gravity ( g a aircraft ).
Measurements of and are made by measuring the acceleration in the body frame
and relating it to a reference acceleration vector in the Earth Centered Earth Fixed (ECEF)
T
frame (usually the gravity vector g = [ 0 0 g ] , where g is the gravitational constant). The
reference vector is constructed such that when the reference vector is measured in the body
frame, this will correspond to zero roll and zero pitch angles.
125
AFASES - 2008 -
To relate the measured acceleration in the body frame to the reference vector in the
ECEF frame, a mathematical relationship for transformation between the two frames should
be formalized.
The rotation matrix from the ECEF frame to the body frame is given by (1).
CECEF body
cos ( )
0
sin ( )
(1)
This can be use in the transformation of acceleration from the ECEF frame into
acceleration in the body frame with (2).
ax
ay
az = CECEF body rx
ry
rz ,
(2)
rx ax + rz rx 2 + rz 2 ax2
rx 2 + rz 2
r ax
= tan 1 x
rz
r = rx sin() + rz cos()
=
(3)
ry a y + r ry 2 + r 2 a y2
ry 2 + r 2
ry + a y
= tan 1
r
z
The procedure for obtaining an UAV attitude measurement using a low-cost GPS
receiver and three axis accelerometers is now addressed. While many GPS receivers can be
configured to output velocities as well as positions, it will be assumed that the GPS receiver
outputs only position information at frequency 1 Hz.
To make a good measurement of UAV attitude the following steps are necessary:
1. Obtaining three consecutive GPS position measurements;
2. Make the difference of the GPS consecutive measurements to obtain two velocity
measurements;
3. Average the velocity measurements to give the average velocity over two steps (over
2 seconds);
4. Calculate the heading from velocity:
X&
= tan 1
&
Y
(4)
126
AFASES - 2008 -
5. Make the difference of the GPS calculated velocities to obtain a GPS acceleration
measurement aGPS ;
6. Using average of the accelerometers over the same 2 seconds as the GPS velocity is
calculated a in the body frame;
7. Calculate roll and pitch using the accelerometers and GPS acceleration rotated
by using equations (3).
A non-linear model with 6 degree of freedom was simulated. The simulated aircraft
performed an 11 minute flight, with multiple coordinated turn and climb maneuvers. Figure 2
shows the roll angle over the simulated flight (pitch and yaw angles show similar noise
characteristics). The solid line is the true roll angle and the dashed line is the estimate from
the AHRS.
This system computes position, velocity and attitude of the vehicle with respect to the
reference frame by numerical integration of the accelerations and angular rates. The AHRS
and INS are in a cascaded (serial) format. In this format, the AHRS and INS filters can be
designed and tuned completely independent of each other.
A cascaded structure also reduces computation. In other words, the cascade
implementation allows for increased flexibility, is easier to implement and tune, and is
dramatically less computationally expensive, with only a small performance loss.
The general UAV equations of motion for navigation are given by (5) and (6) equations.
X&
&
T
Y = DCM
Z&
127
u
v
w
(5)
AFASES - 2008 -
u&
v& = DCM
w&
0
p
u
1
0 + F q X v
m
g
r
w
(6)
Where DCM is the Direction Cosine Matrix, u, v, and w are the body velocities out the
nose, the right wing, and the belly, respectively; X and Y are the inertial coordinates of the
UAV in the Earth Centered Earth Fixed (ECEF) frame; (Z) is altitude; g is the gravitational
constant; m is the mass of the UAV; F is the translational forces acting on the UAV; p, q, and
r are the roll, pitch, and yaw rates, respectively.
To simplify the calculations the following assumptions will be made: (I) attitude is
available from a preceding AHRS filter; (II) v and w are much less than u and close to zero
(this is equivalent to assuming small angle of attack and very little sideslip); (III) a
measurement of altitude is available even when GPS lock has been lost (say from an absolute
pressure sensor).
With these assumptions in place, the cascaded INS and AHRS take the scheme given in
figure 3, where Vp is the airspeed.
Fig. 3 The general scheme of the flight control and stabilization system
Considering the attitude estimate from the AHRS and the UAV airspeed as an input to
the INS, the state-space model of position is given by relation (7).
V p
x& = DCM 0
0
T
(7)
where the DCM is the rotation matrix that maps vectors from the reference frame to the body
T
frame (purely a function of attitude) and x = [ X Y Z ] .
Two flight tests were simulated. First of them where GPS was available for the full time
of flight, and the second where GPS information was ignored for the remainder of the flight
after the filter had run for a short time.
Figure 4 (a) shows the EKF results with the corresponding GPS measurements in the
case when GPS connection is uninterrupted. A magnified portion of the data is shown to
verify that the INS correctly smoothes through GPS measurements, providing an estimate of
position at every instant of time.
Figure 4 (b) shows the case when GPS lock is lost halfway through the flight. As can be
seen, even after 5 minutes of no GPS lock, the INS only strays a maximum of 22 meters from
the true position. In these figures the solid line is the true position and the dashed line is the
estimated position from the INS.
128
AFASES - 2008 -
The performance of the INS is very good considering the quality of the inputs (standard
deviation of attitude angles of 5o and on airspeed 4 m/s). It is concluded that the INS is robust
to both input noise and GPS loss.
V. Conclusions
The increased interest for GPS and INS control systems is justified for UAV
applications and a good choice for laboratory works can be the integrated micro-pilot
MP 2028 and the associated software Horizon.
The results obtained in simulations confirm the theory about the robustness of INS
guidance and navigation systems, and the algorithm proposed, based on Kalman data filtering
offer a good way to obtain the attitude of UAV in test flying.
These INS and GPS navigation systems are very versatile, no so expensive (few
hundred of dollars). They can be easy adapted for ground vehicles or for small vessels, which
can be used in military or civil applications.
Bibliografie
[1] Grewal M. S., Weill L. R., Andrews A P., Global Positioning Systems, Inertial Navigation
and Integration , S.U.A. New Jersey, Editura John Wiley and Sons Inc., 2001.
[2] Grewal M. S., Andrews A P., Kalman filtering: Theory and practice using MATLAB, second
edition, S.U.A. New Jersey, Editura John Wiley and Sons Inc., 2001.
**
129
AFASES - 2008 -
130
AFASES - 2008 -
131
AFASES - 2008 -
action to realize signalling between isolated personnel and rescue team with respect to
the limits and restrictions determined by capabilities of signalling devices, details of
terrain, weather and visibility conditions, medical status, activities of hostile forces or
environment
mark the denied territory and mantraps providing evidence to friendly forces and rescue
forces in the battlefield or in the hazardous area.
PROPER
UNIT
ISOLATED
PERSON
OTHER
UNITS
JSRC
COMPONENT
RCC
For RCC or JSRC it can be a distress indicator one or more received reports such as:
mayday alert, overdue visual or radio contact, non return from mission of persons or craft,
emergency beacon transmissions, reports of personnel being isolated by hostile environment or
forces activities, sighting a vessel or aircraft going down or crushed, sighting or receiving of
emergency coded signals used by survivors or escapers from enemy's prisons.
III. Standard search methods
CSAR methods include visual and electronic surveillance and search of specified area
focusing on inland, coastal, water surfaces and under water environment.
The basic environment, hostile forces activities, proper operations and purposes,
weather factors, search and rescue assets features, the CSAR capable and available teams, all
of these plays an important role in selection of search and intervention methods.
There are five methods for search of survivor caller, starting usually with electronic
search based on sensors in various domains of electromagnetic spectrum. After that it must
continue with the inland search and /or water surface search (usually named maritime
search).
In this cases it can be used the boundary method for search in every square or
rectangular area described by two latitudes and two longitudes. Also, it can use the corner
method for search in polygonal areas described by sequence of pair latitude-longitude for each
corner. Another way is the center point method for search on radial axis divergent from
specified latitude-longitude of starting point and the step and sense of angular deviation. The
trackline method presumes the values of coordinates for start point and the law of movement
and the width of coverage for search mission.
These methods can be used for underwater or in space search missions, but
additionally it must be specified the depth or the altitude. The coastal search is a mixed
method with inland and maritime searches with the same procedures, practicable in the first
time of near water searches.
The search mission of crashed aircraft and aircrews in danger can be significantly
aided and simplified by existence and functionality of Emergency Locator Transmitters
(ELT). There are four types of standardized ELT:
132
AFASES - 2008 -
133
AFASES - 2008 -
low level visibility, no radio transmission, hard and adverse weather as cover rather than
combat fighting or the use of firepower during CSAR mission.
The CSAR Task Force method (CSARTF) is necessary in free space of desert or plane
terrain, lighting condition and good visibility all around, when the hostile forces are waiting for
CSAR activities and can not be surprised.
CSARTF can help the recovery vehicles by navigation assistance, armed escort,
authenticating isolated persons and units in distress, surveillance the enemy activities in the
region of interest for CSAR mission, provide electronic warfare actions counter enemy's
sensors and communications.
The Special Operation Forces Option (SOF) in recovery operations is an expensive
method, used to extract and to place isolated person (VIP or special agent) in company with a
highly trained units, as soon as possible, and to transport them in a friendly or controlled area.
Method of Evasion and Recovery Teams (E&R) is based on indigenous friendly groups
and/or SOF infiltrated in the hostile territory, which act together or separately to search and
recovery isolated persons, to help and rescue them.
Figure 2 present a guide to select the right assets in CSAR missions.
Level of Threat
High density
Air Defense
NON RECOMMENDED
TO USE
Low density
Air Defense
Small Arms
Day or Night
SAFE
Overland
GOOD
TO
USE
SOF
CSARTF
Aviation or ground
combat element
Required
CSAR assets
134
AFASES - 2008 -
Bibliografie
[1] Joint Pub 3-50.2, Doctrine for Joint Combat Search and Rescue, 1996.
[2] STANAG 3650 SAR Essential SAR location equipment and associated characteristics
(Aircraft).
[3] STANAG 7007 SAR Search and Rescue electronic Systems (SARES).
*
**
135
AFASES - 2008 -
Capitolul I
Introducere
Fenomenul tragerii este constituit din totalitatea proceselor care se desfoar n eava
gurii de foc din momentul iniierii arderii pulberii care formeaz ncrctura de azvrlire pn
la momentul ieirii proiectilului din eav. Astel, n urma iniierii unei amorse printr-un
procedeu oarecare(mecanic, electric sau de alt natur) se produc n camera de ncrcare gaze
i particule incandescente. Prin transfer termic, temperatura suprafeei elementelor pulbere
crete i, dup depirea unei anumite valori, se produce iniierea aprinderii acesteia. n cazul
real, dat fiind geometria camerei de ncrcare i modul de dispunere a ncrcturii de
azvrlire, aprinderea elementelor de pulbere nu are loc simultan. Presiunea amestecului
format din gaze i elemente de pulbere care ard la volum constant crete pn se realizeaz
valoarea presiunii de forare, cnd ncepe micarea proiectilului. Acesta se deplaseaz sub
aciunea rezultantei forelor de presiune, de frecare dintre proiectil i pereii evii i de
rezisten a aerului din canalul evii. n condiiile creterii volumului dinapoia proiectilului,
presiunea gazelor continu s creasc pn la o valoare maxim, dup care scade pn la o
anumit valoare cnd proiectilul prsete eava gurii de foc.
n literatura de specialitate se ntlnesc mai multe modele matematice care reflect
fenomenul tragerii, deosebirea dintre acestea constnd n ipotezele adoptate.
136
AFASES - 2008 -
sp (l + l ) = f
qv 2
L6 ,
(1)
unde:
l
bf
T T
CM S0
L6 = s pt dl ; f = f 1 ;
=
0.7774
T1 T1
(l0 + l )v vdl
l0 vk
= z (1 + z + z 2 ) ;
(2)
= 1 + 2z + 3z 2 ;
(3)
u = u1 p ;
(4)
AFASES - 2008 -
d
= u1 p ;
dt
e1
(5)
dv
dv
= qv
= s( p pt ) .
dt
dl
(6)
n modelul matematic propus unele din ipotezele admise n mod frecvent au fost
nlocuite cu altele care reflect mai bine realitile fenomenului tragerii.
Capitolul III
Rezolvarea sistemului de ecuaii care alctuiesc modelul matematic
Caracteristici
Presiunea maxim
[daN/cm2]
Date calculate
2349,47
685
Date experimentale
2419,52
719
138
AFASES - 2008 -
Pe baza datelor rezultate din programul de calcul s-au realizat urmtoarele grafice:
- variaia presiunii gazelor de pulbere funcie de timp (Fig. 1) i funcie de spaiul parcurs de
proiectil (Fig.2)
3000
3000
2500
2500
2000
Prsiune
Presiune
2000
1500
1000
1500
1000
500
500
0,002
0,004
0,006
Tim p
0,008
0,01
10
Spaiu
20
30
8000
8000
7000
7000
6000
6000
Viteza (dm/s)
Viteza (dm/s)
- variaia vitezei proiectilului funcie de timp (Fig.3) i de spaiul parcurs de proiectil n eav
(Fig.4)
5000
4000
3000
2000
5000
4000
3000
2000
1000
1000
0
0
0,002
0,004
0,006
0,008
0,01
Tim p (m s)
10
20
Spaiu (dm)
i variaia scderii relative a temperaturii gazlor funcie de spaiul parcurs de proiectil n eav
(Fig.5)
0,025
Variaia temperaturii
0,02
0,015
0,01
0,005
0
0
0,002
0,004
0,006
0,008
Tim p
Fig.5 Variaia relativ a tem peraturii funcie de tim p
139
0,01
30
AFASES - 2008 -
Capitolul IV
Comentarii
Bibliografie
[1]. CIURBANOV, E.V. - Vnutrenniaia Ballistika, Izdatelstvo VAOLKA, Leningrad,
1975.
[2]. SEREBRIAKOV, M.E. Internal ballistics of guns and solid propellant rocket
engines, vol. 1 and 2, Military Academy Publishing House, Bucharest, 1970.
[3]. VASILE, T. Internal ballistics of guns, vol. 1 and 2, Military Academy
Publishing House, Bucharest, 1993 and 1996.
[4]. VASILE, T., MOCANU, R. Influena pierderilor prin transfer de cldur
asupra principalelor caracteristici ale instalaiilor de impuls, Revista Termotehnica
Nr.2/1994, pg.138-141;
[5]. VASILE, T., MOCANU, R., - Studiu privind micarea proiectilului, n eava gurii
de foc, innd seama de tierea treptat a brului forator in ghintuiri, A XXVIII-a sesiune
de comunicri tiinifice a Institutului de Cercetare-Dezvoltare al Armatei, Buletinul sesiunii,
vol.I, 7 noiembrie 1996, Bucureti, pg.137-140;
[6]. VASILE, T. - Studies and researches regarding the fundamental problem solving
of interior ballistics for weapon with grooved barrel, PROCEEDINGS of III-rd International
Armament Conference on Scientific Aspect of Armament Technology, October 11-13, 2000,
Waplewo, pg.157-165.
[7]. VASILE, T. - Contributions to the fundamental problem solving of interior
ballistics, PROCEEDINGS of the II-nd International Conference Artillery Barrel Systems,
Ammunition, Means of Artillery Reconnaissance and Fire Control, 27-29 October 1998,
Kiev, pg.440-445;
140
AFASES - 2008 -
under action of powder gases pressure inside of a gun with grooved barrel, using physical
law of powder burning. On the bases of the mathematical modeling was elaborated an
interior ballistic software, which allows to study the variation of the powder gases pressure
and the projectile velocity versus its displacement inside of gun and versus time. For an
extant gun, the theoretical results obtained with the aid of this soft and the experimental data
are compared.
Chapter I
INTRODUCTION
The mathematical modelling of projectile motion under action of powder gases
pressure inside of the barrel of a gun with grooved barrel it is realised, in many cases,
accepting a series of simplified assumptions. Thus, in many works of specialty literature [1-4]
it is admitted, among others, the geometrical law for powder burning, proposed by French
researcher Vieille and a constant index of burning rate law.
141
AFASES - 2008 -
In the elaborated mathematical modelling, for the powder burning was utilized the
physical law which reflects better the reality of firing phenomenon, and uses the experimental
results obtained during the powder burning in the closed bomb at densities much bigger than
the densities from ballistic system. The index from the burning rate law is variable. It is
determined as a function of the pressure using the pressure in closed bomb P() , which are
presented in figure 1.
The elaborated mathematical modelling is consisted of the differential and algebraic
equations. By solving the differential and algebraic equations system, in numerical way with
the aid of the computer, it is obtained the variation of gases pressure and projectile velocity
versus its displacement within the barrel and versus time since the beginning of powder
burning, which defines the content of fundamental problem of interior ballistics.
The signification of the used parameters is that established in specialty literature [1-4].
Chapter II
MATHEMATICAL MODELLING
The elaborated mathematical modelling is based on the main differential and
algebraical equations of interior ballistics [1-3].
The differential equation of displacement is obtained with the aid of the fundamental
equation of interior ballistics and of the equation of projectile translation motion, thus:
dl
l +l
q
f
vdv
2
+ v
(1)
2
vlim
where:
f
l = l 0 (a b ) ; a = 1 ; b = 1 ; = 1 1 / + 1 .
b P
For the obtaining of velocity differential equation, it is admit that in closed bomb and
in ballistic system has been burned the same thickens of powder grain and it is took into
account the equation of projectile translation motion, thus:
dv
dI
= s P
q p
(2)
The differential equation for time has got the following form:
dt = dl .
v
(3)
These equations have been adequately transformed in order to allow the utilizing of
experimental data, obtained during the powder burning in closed bomb, at the solving of
fundamental problem of interior ballistics [4-7].
So, it is adopted as an independent variable , defined as ratio between impulse value
of powder gases pressure at certain moment I and impulse value of powder gases pressure at
the fragmentation moment I s , in the case of multiperforated powders. In the case of simple
142
AFASES - 2008 -
shape powders, impulse I s is changed with impulse I k , i.e. the impulse at the end of powder
burning.
In order to increase level of generality of these equations are utilized following
dimensionless variables:
= v ; = l ; = t ,
l0
v lim
t lim
where:
vlim =
2 f
q
q v lim
P
; t lim =
s
(4)
where
E=
a b +
2
Taking into account that dI = I s d and v = vlim , the equation (2) gets the form
d
=
d
B P
2 p
(5)
The differential equation for time (3), after the using of the dimensionless variables
and the changing of d d with its expression from equation (4), becomes
d
d
l0 E
t lim v lim
=2 1
B P
2 p
(6)
f
.
E
(7)
It is eliminated the pressure from equations (4), for displacement, from (5), for
velocity, from (6), for time and, finally, it is obtained:
d
d
d
=
d
d
d
2 B E EP
f
B EP
2 f
(8a)
(8b)
I
= 1 s EP ;
t lim P f
(8c)
p=
f
.
E
143
(8d)
AFASES - 2008 -
For the index from burning rate law, it was taken a variable value that is calculated
on the base of experimental data from closed bomb, only for two densities of charge, with the
aid of relation
= 1
dI
log 2
dI1
P
log 2
P1
d1
d 2
P
log 2
P1
log
(9)
The differential and algebraical equations system (8) can be used in first period, as
well as in the second period of the firing phenomenon.
The differential and algebraic equations system of projectile motion within grooved
barrel of gun (8) represents the new mathematical modelling which allows to study the
variation of gases pressure and projectile velocity versus its displacement and versus time in
conditions closer of the reality of the firing phenomenon.
Chapter III
INTEGRATION OF EQUATIONS SYSTEM
The differential and algebraic equations system (8) was integrated using the RungeKutta IV method, with the aid of following initial conditions:
= 0 0 ; = 0 = 0 ; = 0 = 0 ; = 0 0 ,
in which
0
I
0 = 0 = 0
s
Is
Pd
Pd
; 0 = 0 = 2.303
t lim
t lim
I k (1 )
p
log( 0 ) .
f
pa
The first period, in the case of multiperforated powder grains has two phases. In the
first phase, from the moment of the inflammation of powder until the moment of breaking of
powder grains, when s and 1 , it is integrated the equations system (8). In the second
phase, from the moment of powder grains breaking until the moment of the end of powder
charge burning, when s 1 and = 1 , the system (8) becomes:
d
=
d
2 B E ;
(10a)
d
=
d
B
2
(10b)
d
d
I E
= 1 s ;
t lim f
p=
f
E
(10c)
(10d)
In the equations of this system is not included the pressure P from manometrical
bomb. However, the solving of this system is done with the aid of the curve P( ), because the
values for from E relation are calculated on the base of experimental data.
In the second period, from the moment of the end of powder charge burning until the
moment when projectile leaves the barrel, = 1 and = 1 , the expression for E becomes
E = a b +2 .
1
144
AFASES - 2008 -
The initial conditions for the integration of differential and algebraic equations system
(10) are just values of ballistic magnitudes at the moment of end of powder charge burning.
If the solving is effectuated for the powder that burns without breaking and in the
conditions of the burning rate law u = u1 P , in which = 1 , than it is utilized the system (10),
taking into consideration the particularities of the second phase of the powders that burn with
breaking of the grains.
Chapter IV
RESULTS AND CONCLUSIONS
In the Fig. 2 and Fig. 3 it is presented the variation of gases pressure versus
displacement and time obtained with the aid of the interior ballistics software that was
elaborated on the bases of proposed new algorithm, as well as usually algorithm in which was
used geometrical law for powder burning, for an extant ballistic system.
From the analysis of the data that are presented in the diagrams from Fig. 2 and Fig. 3,
can be concluded that there are some differences in the case of the fundamental problem
solving of interior ballistics on the bases of physical and geometrical law of powder burning.
250000
200000
150000
Serie1
100000
Serie2
50000
0
0
10
20
30
Serie1
100000
Serie2
50000
0
0.000
0.002
0.004
0.006
0.008
0.010
145
AFASES - 2008 -
From the analysis of the data, one of them are presented in the diagrams from Fig. 2
and Fig. 3, can be concluded that this way of the fundamental problem solving of interior
ballistics is closer of firing phenomenon intimacy with gun that uses multi-perforated
powders.
For an extant gun, the experimental and theoretical data are presented in Table I.
Table I
Way of determinations
Experimental data
Theoretical data
Magnitudes
Maximum pressure
[daN/cm2]
2400
2344,6
Muzzle velocity
[m/s]
840
847,1
From the comparing of the experimental data with the theoretical values obtained in
the case of the proposed mathematical modelling, had resulted that the difference between the
muzzle velocity values is approximately 0,8% and between the maximum pressure values is
approximately 3%.
REFERENCES
VASILE Titica, professor dr. eng., Military Technical Academy, Bucharest, 81-83, George Cosbuc Av., tel.
0040 21 33 54 660, fax 0040 21 33 55 763, e-mail: titica@mta.ro;
**
SAFTA Doru, professor dr. eng., Military Technical Academy, Bucharest, 81-83, George Cosbuc Av., tel.
0040 21 33 54 660, fax 0040 21 33 55 763, e-mail: DSafta@mta.ro;
***
BARBU Cristian, professor dr. eng., Military Technical Academy, Bucharest, 81-83, George Cosbuc Av.,
tel. 0040 21 33 54 660, fax 0040 21 33 55 763, e-mail: barbucr@mta.ro.
146
AFASES - 2008 -
ZGREABN CRISTIAN *
Abstract New methods for safety and securityof flight improvement
Safety of air transportation is of major concern to the aviation industry, to
governments and to the public. When compared with other transportation industries
(maritime, rail or road transportation) the aviation industry enjoys a superior safety record.
Safety consciousness and resources devoted to safety within the civil aviation community are
among the reasons of this record. David R. Hinson, Administrator at the U.S. Federal
Aviation Administration ascribes the observed safety enhancement to four major factors:
the improvement in airborne engine's reliability, the increased airborne automation, the
extensive usage of flight simulators and the continuous advances in Air Traffic Control. The
first step in this process is the identification of potential hazards: for example, potential
hazards could be identified during the investigation of an incident, safety surveys or the
Functional Hazard Assessment of airborne System. The second step, risk analysis, analyses
the risk involved in identified hazards, i.e. estimates the probability of occurrence of hazards
and evaluates their degree of severity. The third step, risk assessment, aims at determining
whether the organization is prepared to accept that risk.
CONCEPTUL DE SIGURAN A ZBORULUI
Obiectivul central al organizaiilor aeronautice romneti i internaionale l constituie
adoptarea unor programe de securitate care s permit evitarea accidentelor i incidentelor de
zbor.
Astfel, programul Aviation Security Programme (AVSEC) elaborat de Organizaia
Internaional a Aviaiei Civile (I.C.A.O.) precizeaz c obiectivul principal al programului
de securitate al ICAO este asigurarea c pasagerii, echipajul, personalul terestru, publicul
larg nu vor fi afectai de incidente sau accidente de zbor cauzate de motive tehnice sau
umane sau de accesul ilegal la bordul aeronavelor .
Din punct de vedere secvenial, conceptul de sigurana zborului pornete de la starea
psiho-fizic a personalului participant, continu cu pregtirea tehnicii i a personalului, cu
dirijarea i conducerea zborului i se finalizeaz cu asigurarea proteciei personalului.
Pornind de la cele afirmate, rezult c sigurana zborului depinde de o multitudine
de factori care acioneaz n toate etapele existenei aeronavei, variabile ce pot determina
evenimente, incidente sau accidente uneori cu urmri foarte grave privind pierderile materiale
i umane.
147
AFASES - 2008 -
148
AFASES - 2008 -
149
AFASES - 2008 -
Ateptrile nerealiste
Factori aleatorii de producere a accidentului
Defectarea (indiferent de natura ei: material, uman, organizatoric) poate coincide cu
alte defectri, sau poate determina alte defectri. Astfel de combinaii nedorite pot cauza
defectri n cascad ale altor pri sau subsisteme.
n sistemele complexe, aceste posibile combinaii sunt practic nelimitate;
Un sistem complicat are propria sa inteligen, cu conexiuni invizibile, redundane
neutralizate, protecii si condiii de exploatare pe care proiectanii nu pot sa le anticipeze sau
planifice corespunztor.
Defectrile n cascad pot accelera pierderea controlului, confuzia operatorilor umani i
incapacitatea lor de a proceda pentru repunerea n funciune a sistemului.
150
AFASES - 2008 -
efectul unui eveniment este cunoscut, dar apariia evenimentului este nesigur;
Ce poate merge ru ?
151
AFASES - 2008 -
MANAGEMENTUL RISCULUI
Managementul riscului este un proces ciclic ce conine trei faze distincte, i anume
identificarea riscului, analiza riscului i reacia la risc.
IDENTIFICAREA RISCULUI
n faza de identificare a riscului se evalueaz pericolele poteniale, efectele i
probabilitile de apariie ale acestora pentru a decide care dintre riscuri trebuie prevenite.
Practic, n aceast faz se identific toate elementele care satisfac condiiile (1) i (2).
Totodat, se elimin riscurile neconcordante, adic acele elemente de risc cu probabiliti
reduse de apariie sau cu un efect nesemnificativ.
Identificarea riscurilor trebuie realizat n mod regulat lund n considerare att riscurile
interne ct i pe cele externe. Riscurile interne sunt cele pe care echipa managerial le poate
controla sau influena, n timp ce riscurile externe nu se afl sub controlul acesteia.
Riscul poate fi identificat folosind diferite metode:
9 ntocmirea unor liste de control care cuprind surse poteniale de risc, cum ar fi: condiii de
mediu, rezultatele ateptate, personalul, modificri ale obiectivelor, erorile i omisiunile de
proiectare i execuie, estimrile costurilor i a termenelor de execuie, etc.;
9 analiza documentelor disponibile n arhiv, pentru identificarea problemelor care au
aprut n situaii similare celor curente;
9 utilizarea experienei personalului direct productiv (efi de secii i de echipe) prin
invitarea acestora la o edin formal de identificare a riscurilor. De multe ori personalul din
teren este contient de riscuri i probleme nesizate de cel din proiectare i de aceea
comunicarea eficient este una dintre cele mai bune surse de identificare i diminuare a
riscurilor;
9 identificarea riscurilor induse din exterior.
ANALIZA RISCULUI
Faza a doua ia n considerare riscurile identificate anterior i realizeaz o cuantificare
aprofundat a acestora. Pentru analiza riscului se folosete un instrumentar matematic divers,
mergnd de la analiza probabilistic la analiza prin metoda Monte Carlo. Alegerea aparatului
matematic trebuie s fie adaptat necesitilor analizei i s in seama de acurateea datelor
disponibile.
Cea mai simpl metod de cuantificare a riscurilor este aceea a valorii ateptate
(VA), care se calculeaz ca produs ntre probabilitile de apariie ale anumitor evenimente i
efectele acestora:
VA(a) = P(a) x E(a) (3)
unde: VA(a) = valoarea ateptat a evenimentului (a)
P(a) = probabilitatea de apariie a evenimentului (a)
E(a) = efectul apariiei fenomenului (a)
Avnd n vedere faptul c estimarea probabilitilor este un proces cu un grad mare de
subiectivitate, rezultatele obinute prin metoda valorii ateptate sunt de obicei utilizate ca date
de intrare pentru analize ulterioare.
Simulrile constituie o metod avansat de cuantificare a riscurilor. care utilizeaz un
model matematic al unui sistem pentru a analiza performanele sau comportamentul acestuia.
Pentru proiecte complexe se folosete frecvent simularea Monte Carlo a programului de
execuie i a costurilor asociate activitilor.
Aceast tehnic simuleaz realizarea obiectivelor de un numr mare de ori furniznd
o distribuie statistic a rezultatelor.
152
AFASES - 2008 -
Valoarea ateptat a unei decizii = suma valorilor ateptate ale tuturor efectelor
rezultnd din acea decizie
153
AFASES - 2008 -
Reacia la risc este faza de aciune din cadrul ciclului managementului riscului, n care se
ncearc s se elimine riscurile i s se reduc sau s se repartizeze riscurile.
REACIA LA RISC
Aceast a treia etap i propune eliminarea riscurilor identificate i analizate
anterior. Pe cale de consecin, echipa managerial poate sau nu s iniieze o anumit
activitate, care s acopere riscurile.
Cele mai multe dintre opiunile care i propun s elimine total riscul au dezavantajul
c tind s transforme organizaia ntr-una necompetitiv. O organizaie cu aversiune prea mare
fa de risc nu va supravieui mult timp i va trebui s-i investeasc capitalul n alt zon.
Diminuarea riscurilor se poate realiza printr-o serie de instrumente cum sunt:
instruirea. Multe riscuri sunt legate de securitatea muncii, care influeneaz productivitatea
i calitatea lucrrilor. Probabilitatea producerii accidentelor i efectul acestora se poate reduce
prin programe de instruire i contientizare n domeniul securitii muncii.
154
AFASES - 2008 -
Identificare
Documente
Comunicate
Analiza
Urmrire
Plan
Fig.1
IDENTIFICAREA RISCULUI
Conceptul de identificare a riscului.
Prima etap n procesul MCR este identificarea riscurilor (tehnice si programatice)
specifice proiectului, att n termeni de accidente /evenimente neateptate ca i a consecinelor
acestora. n plus, identificarea riscului include cunoaterea informaiilor necesare pentru a
ncadra riscul n contextul programului/proiectului.
Identificarea riscului trebuie sa se continue n ntregul ciclu de via, aciune ce trebuie
realizat ntr-o manier clar, astfel nct s poat fi neleas de oricine, chiar dup trecerea
timpului.
Riscul poate fi considerat identificat, dac se rspunde cel puin la ntrebrile:
155
AFASES - 2008 -
156
AFASES - 2008 -
a.
Expertize tehnice;
b.
Analize statistice ale nregistrarilor anterioare;
c. analize de incertitudine (constnd n modelarea sistemului investigat, cu modificrile
inerente aprute n ciclul de via)
d.
Evaluarea probabilistic a riscului; analiza arborelui de defect, analiza modurilor de
defectare i a efectelor defectrilor (v. cap 7)
e.
Ordonarea dupa nivelul de risc;
f.
Comparaii cu sisteme analoage.
nainte de a stabili gradul de prioritate, riscul trebuie clasificat i grupat funcie de
similitudinea cu alte riscuri.
Scopul clasificrii este:
a. nelegerea naturii riscului si gruparea acestora, pentru adoptarea unei scheme cu raportul
cost-eficien ct mai favorabil;
b. Identificarea riscurilor identice, echivalente sau o combinaie a acestora;
c. Eficientizarea urmririi i monitorizrii
Scopul stabilirii gradului de prioritate este sortarea/separarea riscurilor care nu pot fi
acceptabile, dintre riscurile posibile. Pentru stabilirea gradului de prioritate se pot folosi
metode cantitative sau calitative cum ar fi numrul de voturi sau de atribute
Consecine
Consecina reprezint o evaluare a unei situaii potenial credibile aprut ca urmare a
unui risc, iar unitatea de msur difer funcie de riscul specific. De exemplu, riscul privind
costul se exprim n bani (dolari, euro, lei) sau n procente din bugetul
programului/proiectului.
Consecinele se clasific n:
Consecine catastrofale
Consecine critice
Consecine moderate sau marginale
Consecine neglijabile
a. Clasa I - Consecine catastrofale. O situaie care poate cauza:
9
moarte sau infirmitatea unor persoane;
9
distrugerea aeronavei;
9
distrugerea unor faciliti la sol;
9
nendeplinirea unei misiuni vitale;
9
depirea costurilor cu peste 50% din costurile planificate.
b. Clasa II Consecine critice. O situaie care poate cauza:
9
rnirea grav sau boli profesionale
9
avarii grave ale aeronavei;
9
avarii ale facilitilor de la sol
9
depirea costurilor cu peste 15% pn la 50% din costurile planificate
c.
Clasa III Consecine moderate. O situaie care poate cauza:
9
rniri sau boli profesionale minore;
9
distrugerea minor a facilitilor;
9
depirea costurilor cu 2% pn la 15% din bugetul proiectului
d. Clasa IV - Consecine neglijabile. O situaie care poate cauza:
9 rniri minore care pot necesita primul ajutor, dar nu afecteaz sntatea sau sigurana
persoanelor;
9 avarii ale aeronavei, sistemelor sau echipamemnelor care nu sunt mai mari dect cele
produse prin uzur;
9 depirea costurilor planificate cu mai puin de 2%.
157
AFASES - 2008 -
Probabilitate de producere
Probabilitate de producere este posibilitatea de a se produce un risc identificat.
Exemple de categorii pentru probabiliti de producere a riscurilor identificate sunt:
a.
Probabilitate de producere A.
Posibil s se produc (de ex. probabilitate > 0.1).
b.
Probabilitate de producere B.
Probabil se va produce (de ex. 0.1 > probabilitate > 0.01).
c.
Probabilitate de producere C.
Poate s se produc (de ex. 0.01 > probabilitate > 0.001).
d.
Probabilitate de producere D.
Puin probabil sa se produc (de ex. 0.001 > probabilitate > 0.000001).
e.
Probabilitate de producere E.
Improbabil (de ex. 0.000001 > probabilitate).
Matricea de risc.
Matricea din figura 2 prezint aplicarea claselor de consecine i probabilitate de producere
pentru determinarea calitativ a riscului ridicat, mediu, slab.
PROBABILITATE DE PRODUCERE
CONSECINE
CLASA
A
B
C
D
I
1
1
2
3
II
1
2
3
4
III
2
3
4
5
IV
3
4
5
6
figura 2 Matricea de risc - codurile de evaluare a riscului
E
4
5
6
7
BIBLIOGRAFIE
1
DECISION NO. 2003/1/RM OF THE EXECUTIVE DIRECTOR OF THE AGENCY
of 17 October 2003 on acceptable means of compliance and guidance material for the
airworthiness and environmental certification of aircraft and related products, parts and
appliances, as well as for the certification of design and production organisations (AMC and
GM to Part 21)-- www.easa.eu.
2
Normal Accident Theory/ The Changing Face of NASA and Aerospace/ Hagerstown,
Maryland/November 17, 1998/ Dr. Michael A. Greenfield/Deputy Associate Administrator
Office of Safety and Mission Assurance
3
SAFETY AND MISSION ASSURANCE ROLE IN ACQUISITION/ NASA Assurance
Technology Conference/May 25-26, 1999 /Michael A. Greenfield, Ph.D Deputy Associate
Administrator Office of Safety and Mission Assurance (nasa.gov)
4
CHALLENGES IN THE BETTER, FASTER, CHEAPER ERA OF AERONAUTICAL
DESIGN, ENGINEERING AND MANUFACTURING / Earll M. Murman, Myles Walton,
Eric Rebentisch/ Massachusetts Institute of Technology, Cambridge, MA 02139 USA
(esd.mit.edu)
5
Fault Tree Handbook witth Aeroos pace Applications Prepared for NASA Office of
Safety and Mission Assurance NASA Headquarters Washington, DC 20546/ August, 2002
6
Risk-Based Design Overview /Thomas A. Zang / ASCAC Methods Development Peer
Review November 27-29, 2001(nasa.gov)
7
Aerospace Risk Management/ Bristol Consortium University of Bristol 27th 29th
April 2004
158
AFASES - 2008 -
8
NASA FAR Supplement (NFS) Federal Register: July 20, 1999 (Volume 64, Number
138)] [Proposed Rules] [Page 38880-38884] From the Federal Register Online via GPO
Access [wais.access.gpo.gov] [DOCID:fr20jy99-33]
9
NPG 7120.5 NASA Program and Project Management Processes and Requirements
10
ESA Alert System OPERATIONAL PROCEDURE Doc. Ref. Q/EAS/PROC/2
/December 1998
11
Probabilistic Risk Assessement Guide for NASA Managers and Practitioners/ Office of
Safety and Mission Assurance NASA Headquarters/ Washington, DC 20546 August, 2002
12
Risk-Based Acquisition Management (R-BAM) / Office of Safety and Mission
Assurance NASA Headquarters Washington, DC 20546 August, 2002
13
Risk Balancing Profile Tool / Dr. Michael A. Greenfield, Deputy Associate
Administrator for Safety and Mission Assurance / National Aeronautics and Space
Administration / Washington, D.C. 20546
14
Risk Management Guide for DoD Acquisition / DEPARTMENT OF DEFENSE
DEFENSE ACQUISITION UNIVERSITY DEFENSE SYSTEMS MANAGEMENT
COLLEGE
15
Risk Management Risk As A Resource Langley Research Center/ 1998/ Dr. Michael
A. Greenfield Deputy Associate Administrator/ Office of Safety and Mission Assurance
159
AFASES - 2008 -
160
AFASES - 2008 -
trecerea timpului, unii consumatori i-au restructurat activitatea, alii chiar au disprut de pe pia,
fapt care a determinat scderea consumului de energie electric, dar i sub form de energie
termic (n special abur industrial i apoi energie termic pentru ncalzire i preparare ap cald
de consum). O reducere a produciei de energie va avea loc n viitoarea perioad i datorit
msurilor ce se vor adopta privind reducerea consumurilor de energie i utilizarea eficient a
energiei si resurselor energetice. n aceast situaie se impune respectarea normelor de mediu,
redimensionarea instalaiilor n vederea satisfacerii noilor consumuri i producerea energiei la
preuri competitive impuse de pia. Analizele care se vor efectua pentru fiecare producator n
parte vor impune adoptarea unor soluii radicale privind combustibilii utilizai, putndu-se ajunge
la renunarea la actualele soluii de combustibili. De cele mai multe ori, aplicarea soluiilor
tehnice necesare integrrii instalaiilor actuale n normele de mediu impuse, implic cheltuieli
foarte mari, dar aceste msuri nu au i efecte pozitive asupra eficienei i a randamentului global.
De aceea, n multe situaii se constat c este mai convenabil construirea unor instalaii noi care
sa nglobeze tehnologii moderne, de ultim or, dect modernizarea celor actuale.
Soluiile moderne de combustibili prevd i utilizarea biomasei pentru producerea de
energie electric i termic. Pe plan mondial, [2] biomasa are o utilizare larg n ponderea
combustibililor pentru producerea energiei necesare: n SUA 3,2% din consumul primar,
Danemarca 8%, Austria 11%, Suedia 19%, Finlanda 21%. La ora actual n Romnia, biomasa se
utilizeaz n special n instalaii mici, preponderant n gospodrii individuale, sau n localiti mai
mici. Un proiect de succes Rumegu 2000 a fost dezvoltat n patru localiti (Vatra Dornei,
Huedin, Gheorghieni si ntorsura Buzului), unde, n colaborare cu guvernul danez s-au realizat
instalaii de nclzire centralizat, n care se arde rumeguul.
Biomasa [2] este ansamblul materiilor organice nonfosile. La fel ca i energiile obinute
din combustibilii fosili, energia produs din biomas provine din energia solar nmagazinat n
plante, prin procesul de fotosintez. Rezervele de biomas sunt n special deeurile de lemn,
deeurile agricole, deeurile menajere i culturile energetice. Producerea de biomas nu reprezint
doar o resurs de energie regenerabil, ci i o oportunitate semnificativ pentru dezvoltarea
durabil. Romnia este o ar cu un potenial energetic ridicat de biomas de aproximativ 8000
mii tep/an, ceea ce la nivelul anului 2000 reprezenta 19% din consumul total de resurse primare.
Cea mai mare pondere o reprezint deeurile lemnoase, constituite din reziduurile exploatrilor
forestiere (crengi, vrfuri etc), din reziduurile proceselor de prelucrare ale lemnului (rumegu,
resturi de la decupri sau alte prelucrri), sau din alte utilizari ale lemnului (demolri ale
cldirilor construite din lemn, deteriorarea pieselor si structurilor confecionate din lemn, etc.).
Aceste deeuri pot fi arse direct, sau mai eficient, dup tocare. n privina utilizrii crengilor
rezultate n urma exploatrilor forestiere, pot apare controverse, deoarece silvicultorii din
Romnia recomand ca acestea s fie utilizate pentru regenerare i protecia solurilor. Deeurile
agricole sunt constituite din paie, tulpini de porumb, sau reziduuri ale proceselor de prelucrare ale
produselor agricole (coji de semine, etc). Arderea paielor n instalaiile energetice poate genera
stoparea fenomenului de incinerare a lor n camp, cu efecte negative in privina polurii, a solului
i a siguranei privind incendiile. Deeurile urbane pot fi incinerate n instalaii de ardere, sau pot
fi exploatate n vederea obinerii biogazului. Dei n Romnia nu s-au utilizat pn n prezent,
culturile enrgetice au o larg utilizare n Italia, Danemarca, sau Suedia. Ele constau din cultivarea
unor specii arboricole care au o viteza de cretere foarte mare, cu un potenial energetic ridicat
fat de alte soiuri, pe terenuri denivelate, puin fertile sau improprii altor culturi. Cel mai des n
astefel de cazuri se procedeaz la dezvoltarea culturilor de slcii, dar si ali arbori pot da rezultate
benefice (plop, rchit, eucalipt, etc), n funcie de condiiile i specificul zonei. n prezent, n
Romnia nu s-au dezvoltat tehnologii de valorificare complet a tuturor deeurilor (de exemplu,
161
AFASES - 2008 -
Paie uscate
Paie verzi
Lemn
Crbune
Gaz natural
Coninut de ap
10-20
10-20
40
12
Volatile
>70
>70
>70
25
100
Cenu
0,6-1,5
12
Carbon
42
43
50
59
75
Hidrogen
5,2
3,5
24
Azot
0,35
0,41
0,3
0,9
Sulf
0,16
0,13
0,05
0,8
MJ/kg
14,4
15
10,4
25
48
Putere caloric
Din datele prezentate comparativ n tabelul nr. 1, se poate observa c paiele i lemnul au
un potenial energetic ce poate permite utilizarea lor n instalaiile energetice, dar i c cenua
rezultat n urma arderii lor este mult mai mic dect n cazul crbunilor. Acest fapt determin
realizarea unor depozite pentru cenu mult mai mici, pentru cazul n care aceasta nu este
utilizat n agricultur.
162
AFASES - 2008 -
Arderea biomasei are loc pe grtare, n strat. De regul, schema instalaiei de ardere este
similar celei prezentate n figura 1. Combustibilul transportat cu ajutorul mijloacelor auto, este
deversat ntr-un buncr de stocare, de unde este preluat cu ajutorul unor instalaii de ncarcare n
buncrii individuali de consum. De aici, prin cdere natural sau cu ajutorul benzilor de
alimentare, este deversat pe grtar, unde are loc arderea n strat. Pentru a preveni arderea
combustibilului n afara camerei de ardere, banda de alimentare este prevazut cu o instalaie
automat care controleaz temperatura combustibilului sau prezenta flacarii m aceast zon.
Grtarele au viteze reduse de deplasare pentru a putea permite arderea n totalitate i degajarea
ntregii cantiti de cldur continut n combustibil. Arderea volatilelor coninute are loc n
camera urmtoare a cazanului. n cele mai multe cazuri, n a doua camer are loc i arderea
gazului metan suplimentar, pentru a se putea usca i ridica parametrii finali ai aburului rezultat.
Resturile metalice i cenua sunt colectate la partea inferioar, cenua este separat de resturile
metalice care sunt trimise la reciclare, iar cenua la depozit sau pentru utilizare n agricultur.
Gazele rezultate sunt filtrate n cicloane i electrofiltre. n funcie de noxele coninute, gazele pot
fi supuse i altor procedee de curare.
163
AFASES - 2008 -
164
AFASES - 2008 -
n Romnia [1] pentru anul 2008, productorii calificai de energie electric i termic din
surse regenerabile sunt n numr de 33, din care 10 din surse eoliene, 23 hidro, iar pentru biomas
nu se nregistrez nicio poziie. Acest fapt trebuie s constituie o serioas problem, deoarece
pn n 2010 Romnia trebuie sa asigure 33% din consumul de energie din resurse alternative.
Pentru realizarea acestor investiii se pot atrage fonduri europene prin programele de coeziune sau
prin credite carbon. Un alt avantaj al producerii de energie electric i termic din biomas o
constituie i faptul c pentru energia produs din surse regenerabile Romnia trebuie s aloce cote
anuale obligatorii pentru achiziia certificatelor verzi, astfel: n 2008 cota este de 5,26%, n 2009
de 6,78%, iar n 2010 de 8,3%. Aceste certificate verzi au o valoare de 25 /CV n Suedia, 103
/CV n Belgia regiunea Valonia i 139,1 /CV n Italia (pentru anul 2006). n Romnia, preul
mediu pentru anul 2006 a fost de 42 /CV.
Se poate observa c n toate cazurile prezentate, instalaiile satisfac n totalitate condiiile
de mediu impuse, dar n acelai timp satisfac i necesarul de energie electric i termic al zonei.
Acest lucru ne permite s tragem concluzia c soluiile prezentate pot fi o alternativ viabil
pentru producerea n cogenerare a energiei electrice i termice.
BIBLIOGRAFIE
1.
2.
3.
4.
Ciprian Bncil, drd. ing., Agenia Pentru Protecia Mediului Braov, Politehnicii nr.3, tel 0268419013, e-mail:
director.executiv@apmbrasov.ro
**
Dumitru Cuibar, drd. ing., S.C. Energy Saving Agency S.A., Braov, Olte nr. 11, tel 0268422025, e-mail:
dcu@esa.com.ro
165
AFASES - 2008 -
166
AFASES - 2008 -
which there is a difference between the ship speed and the speed of the water particles relative
to the ship is also termed wake. Generally speaking, the wake is a zone not investigable
theoretically (analytically), due to very complex, aleatory flow character within it. In ships
propeller theory, a distinctive importance is having only the incipient part of the trail (wake),
located immediately in the front of the propeller disk plane. The movement from this zone is
called wake movement or simply wake. The wake movement can be investigated or in the
presence or in the absence of the propeller, when is taking the attribute of the effective wake or
the nominal wake, respectively. However, the wake movement of interest is only that from the
plane where the propeller follows to be situated. The flows average velocity from that plane
is termed wake speed VW, and is in general smaller then ships speed VS, relative to infinite
upstream water. If the water is moving in the same direction as the ship, the wake is said to be
positive. Then
For adimensionalization, the precedent relation can be divided by either VW or VS leading to
two wake factors
- Froude wake factor = wF = (VS VW) / VW;
(2.1)
- Taylor wake factor = w = (VS VW) / VS; ;
(2.2)
Besides, this general effect of the ships hull, there will be local perturbations due to the shaft,
shaft modeling or shaft brackets and other appendages. These effects combined lead to the so
called relative rotative efficiency (RRE), defined by:
RRE = hR = efficiency the ship hull / efficiency of propeller in open water (at speed
VW);
Always, but especially in present circumstances, propeller cavitation reducing, overall
propulsive efficiency and stability improving, represent important challenges for researchers
from ship hydrodynamics field and not only. As already mentioned previously, the dynamics
of a cavitating propeller depends on the system environment in which it is operating: such as,
the flow field within a propeller mounted behind of a ship hull is very different from that one
in an open water test or in a section of a cavitation tunnel. Thus, a propeller that is very
efficient in open water can not be suited for a certain kind of stern shape architecture. For this
reason, the wake distribution in the propeller disk plane represents a key element for
designing of a ship hull stern form. A uniform wake distribution from an immediately
upstream propeller parallel plane disk can lead to the formation of propeller cavitation
decreasing (having as indirect consequence on the noise and vibration level induced on board
and in the hull stern structure, lowering), the propulsive efficiency increasing (for minimum
energetic consumptions obtaining). Therefore, obtaining of a good nominal wake distribution
is an important objective of all naval architects. In addition, the global directional
hydrodynamic stability improving by using a special kind of stern having certain architecture
(more appropriate), can not be but favourable.
Of propeller behind Wake = VS VW
3. Up-to-date level in the field
Present-day tendency in maritime transportation industry is represented by designing and
building of bigger, faster, more energy-efficient and stable ships but simultaneously having
stricter noise and vibration levels for stern hull structure. A modern ships hull lines are
designed to minimize forward resistance, to reduce propeller cavitation, to improve
propulsion
performance
and
to
increase
global
hydrodynamic
stability.
167
AFASES - 2008 -
Since the apparition of the first ships, the naval architects ceaselessly were racking
their brains how to improve the existing hull forms. As a general recently accepted opinion,
the ships of the future will be designed and built only on the basis of some new devised
concepts. It is well known that the stern flow problem is rich in complexity and poses many
challenges. Ships hulls stern flows have received much attention these last years, in particular
with regard to their 3odeling and design principles. As a state of the art in the field, the most
recently known industrial achievements, focused on flows improvement in the stern region,
which consist in symmetrically flattening of the stern lateral surfaces towards the central plane.
This concept has resulted in a huge amount of inconveniences almost in all practical
applications to real ships (unsuitable placing of equipments, lack of necessary spaces for
inspections, repairs, etc.). For a long time, the present projects author thought how to
redesign the two systems hydroframe system and propulsion system very important
(critical) for a ship, so that the hydroframe may meet the propulsion and the propulsion may
meet the hydroframe in an optimal way.
4. Scientific researchs objectives
total forward resistance reduction, propulsive efficiency increasing (for minimum
energetic consumptions obtaining);
- propeller cavitation reduction (for level of noise and vibration induced on board and in the
stern structure decreasing);
- development of a numerical parameterized model;
- design sensitivity analysis of fields generated;
- optimizations;
- original concepts and ideas validation;
- new methodologies establishing.
It is hoped that the successful solving of the above mentioned objectives will contribute to the
top new knowledge accumulation and progress, in a very important, actual and complex
scientific field as contemporary ship hydrodynamics is.
I have proposed (intuitively, based on experience), a new stern hydrodynamic concept
of streamline tube type, (having quasi-cylindrical increasing sections), which starts from front
propeller disk and stretches until hull cylindrical region (Fig. 1). In devising of this new
design concept, the author referred (as a supplementary basic background) to two very well
known and simple existing theories: - the streamline tube theory (the water particles axial
velocities distribution at entrance in the propeller disk can be configured favourably homogenized - by comprising the radial crenellated - corrugated stern sections in a stream
tube that also comprises the propeller disk); - the Bernoulli Effect (increasing of water
particles axial velocities in the regions within which the water pressure is decreased). Taking
into account the streamline tube theory and the Bernoulli effect, we can estimate that the 3D
spectrum of flow generated around and outside of a classical stern hull having practiced
transversal crenellated-corrugated stern sections can be substantially improved by an
architectural optimization in the sense of axial velocities from a propulsion propeller
immediate front plane uniformization (Fig.2).
-
168
AFASES - 2008 -
Fig.1
The directions of the crenellated-corrugated sections teeth crests and troughs longitudinal
curved lines, will be those of the stern natural streamlines (which can be established
experimentally in a flow visualization test) for vortices turning up avoiding and for a
minimum forward resistance obtaining.
Fig.2
Comparison between experimental wake obtained for the model with initial stern shape design
(left) and for the model with modified stern shape in conformity with the new concept design
(right).
Finally, the most important, until now, proved result, is the reducing of propeller cavitation
(working in the simulated nominal wake of the hull using the new stern hydrodynamic
concept with radial crenellated - Fig.3).
169
AFASES - 2008 -
Fig.3
Simulated nominal wake testing, in 850x850 mm section of the cavitation tunnel at 25 rps
rotative speed (it can be remarked lack of cavitation).
Unfortunately, this cavitation decreasing (lack of cavitation) is associated with a total
forward resistance (of the ship) increasing (approximately 4-5%) due to initiation and
movement of some multiple increased vortices (Fig. 4), resulted from the separation (although
a low one Fig. 5) of the boundary layer (destruction of an important part of fluid mechanical
energy, pressure decreasing on the down stream part of the ship stern body, etc).
Fig.4
Vortex initiation and separation FLUENT 6.3
( left - the model with initial stern shape design - simple vortex; right- the model
with modified stern shape in conformity with the new concept design- multiple vortices)
170
AFASES - 2008 -
Fig.5
Limit streamlines on stern surface- FLUENT 6.3
( left - the model with initial stern shape design; right - the model with modified stern shape
inconformity with the new concept design)
Therefore, it would be necessary (a much more) reducing or even complete separation and
multiple vortices phenomena (within the turbulent boundary layer) avoiding. In this direction I
thought that I should try to use the inverse piezoelectric effect (electric current highfrequency generator piezoelectric driver made of certain ceramic material Fig. 6), which
induces an elliptical vibratory movement (high frequency over 20 kHz), into the elastic side
plates (15 mm thickness) in the streamlines direction (of the external flowing water).
Fig. 6
Principle scheme of an ultrasonic vibrator
Piezoelectricity is the ability of crystals and certain ceramic materials to generate a
voltage in response to an applied mechanical stress. Piezoelectric effect was discovered by
Pierre and Jacques Curie in 1880.
The basic principle would be the following: certain piezoelectric ceramic materials can be
used to convert electrical energy into mechanical energy in the form of vibrations of an elastic
body (ship hull stern plates), whose surface points perform an linear elliptic motion (in the
streamlines direction of the external
Water particles (from within the ship hull stern turbulent boundary layer Fig.7), are
pressed against vibrating steel plates reducing the interface (hull - water), skin friction drag.
It is hoped that such a combination of devices can reduce ship forward resistance due to hull
skin-water friction reduction by controlling the inside turbulent boundary layer
flow characteristics.
171
AFASES - 2008 -
Fig. 7
Boundary-layer flow regions
I consider as interesting the realization of a:
parameterized geometrical model streamline tube type, (including the effects of new stern
design having quasi-cylindrical increasing radial crenellated-corrugated sections on
inside propeller flow);
- design sensitivity analysis of the new stern fields generated.
In these cases different geometries (as necessary form, width and depth, along hull distances,
for flow separation avoiding), should be studied theoretically, numerically and
experimentally. Design sensitivity analysis consists in determining derivatives of a systems
response with respect to its design parameters xi. In the context of design optimization (of the
new hydrodynamic stern concept proposed), the response is expressed in terms of objective
and constraint functions, and accordingly the overall aim of design sensitivity analysis is to
find the gradients of these functions. However, since any such problem function depends
explicitly on the dependent variables of the considered problem, sensitivity formulations in
essence aim at the calculation of the derivatives / xi. In other words, the changes in flow
field resulting from a given change in design must be predicted. After the determination of
these flow field sensitivities, it is a matter of straight forward calculus to compute the design
-
172
AFASES - 2008 -
Interdisciplinarity degree
The main disciplines from which the present project is taking its sap are:
- mathematical-physics (partial differential and integral equations);
- physics applications; technical physics;
- materials physics;
- modeling and simulation;
- hydraulics and fluid mechanics;
- naval hydrodynamics.
of any problem function
References
[I] G. K. Batchelor, F.R.S. An Introduction to Fluid Dynamics, Cambridge University,
1991;
[2] C. A. J. Fletcher, Computational Techniques for Fluid Dynamics, Springer-Verlag,
1991;
[3] FLUENT 6.3, Users Manual, Fluent Incorporated, Lebanon, NH, 2006;
[4] J. P. Ghose, R. P. Gokarn, Basic Ship Propulsion, ALLIED PUBLISHERS Pvt.
LIMITED, New Delhi, Mumbai, Kolkata, Lucknow, Chennai, Nagpur, Bangalore,
Hyderabad, Ahmedabad, 2004;
[5] S. K. Godunov, V. S. Reabenki, Finite Difference Computing Schemes, Bucharest,
1977;
[6] Sir Horace Lamb, Hydrodynamics, Cambridge University, 1952;
[7] J. N. Newman, Marine Hydrodynamics, Massachusetts Institute of Technology, 1982;
[8] W. H. Press, S. A. Teukolsky, W. T. Vetterling, B. P. Flannerly, Numerical Recipes in
Fortran The Art of Scientific Computing, Cambridge University, 1992;
[9] K. J. Rawson, E. C. Tupper, Basic Ship Theory, Longman Group Limited, 1995
[10] M. G. Salvadori, M. L. Baron, Numerical Methods in Engineering, Bucharest, 1972;
[11] H. E. Saunders, Hydrodynamics in Ship Design, The Society of Naval Architects and
Marine Engineers 74 Trinity Place, New York 6, N.Y. 1957
[12] N. Tanasescu, Numerical Methods in MATLAB, MATRIXROM, Bucharest, 2002;
[13] H. C. Tanasescu, Al. A. Vasilescu, Fluid Mechanics, Course for engineers, University
of Galati, 2000;
[14] H. C. Tanasescu, Numerical Analysis, Course for master of science engineers,
University of Galati, 2001;
[15] H. C. Tanasescu, Contributions concerning surface ships forward resistance studied by
numerical simulation, thesis, University of Galati, 1998;
[16] Al. A. Vasilescu, Hydromechanics Course, University of Galati, 1962
173
AFASES - 2008 -
1. Introduction
There are situation when the performing possibilities of the model impose deviations
from the complete geometrical similarity obtaining distorted models which have horizontal
lengths and vertical lengths reduced on different scales.
Generally, the distorted models are imposed either when the practical achievement
possibilities of models make impossible the strict observance of geometrical similarity
between the model and prototype, or when the evolution of the phenomenon on the model
performed on a single scale should lead to a laminar motion instead of a turbulent motion
what should make more difficult all the experiments.
By using the distorted models the water rate and depth in desired limits are provided
on these models removing the above mentioned difficulties.
Any phisical phenomenon can be expressed in the most general way by a function of
more physical magnitudes and the establishment of a conection between them is made ( when
the number of phisical magnitudes n > 5 ) by means of theorem.
Any homogeneons function of more physical magnitudes determining a physical
phenomenon can be always reduced to a relation between complex non- dimensional
magnitudes under the form of :
(1 , 2 ,........, n k ) = 0
(1.1)
In similarity theory, this function is called a criterion equation and its establishment
represents the first phase of study on the model of a phenomenon ..
2. Establishment of Criterion Equation of Water Flow Phenomenon Around the Ship
We propose on to establish the criterion equation in the case of advance resistance for
a ship model performed on two scales.
For constant speed motion, the advance resistance is equal and of opposite direction
with the thrust force of the propeller.
174
AFASES - 2008 -
(2.1)
m
or
The homogeneity condition by which all the terms must have the same size leads to
x1 + x6 + x11 = 0
3 x1 + x 2 + x3 x5 + x6 + x7 + x8 + x9 + x10 + x11 = 0
x 2 x 4 x5 2 x6 2 x11 = 0
the equation system: (4) :
(2.4)
It is built up the dimensional matrix of the variables:
g L B T R
v d n
Kg
m
s
1
0 0 0 1
0 0
3 1 1 0 1 1 1
0 1 0 1 1 2 0
0
1
0
0
1
0
0 1
1 1
0 2
175
(2.5)
AFASES - 2008 -
1
0 0
(2.6)
3 1 1 = 1
0 1 0
As all the the fourth order minors are null (these minors are formed by adding a zero
line to the dimensional matrix) it results that the main determinant is of the third order and
the matrix rank is r = 3. The system (4) is solved by Cramers rule and the matrix of the
system solution is obtained:
x1 x 2
x 3 x 4 x 5 x 6 x 7 x 8 x 9 x10 x11
v
d
n
g
L
B T
R
1
2
0
1
1
1
1
1
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
3
4
0
0
2
0
1
1
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
5
6
0
0
0
0
1
1
0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
(2.7)
7
8
0
0
1 0
0
0
0
0
0
1
0
1 2 2 0
0
0
0
0
0
0
1
The number of lines in the solution matrix represents the number of non dimensional
complexes interfering in the phenomenon studied. In the present case, the non dimensional
complexes are:
1 = v 1 d 1 n1 =
nd
;
v
2 = 1 v 1 d 1 1 =
2
1
1 gd
; 3 = v d g =
;
vd
v2
L
1
1 B
1
1 T
; 5 = d B = ; 6 = d T = ;
d
d
d
R
7 = d 1 1 = ; 8 = 1 v 2 d 2 R = 2 2
d
v d
4 = d 1 L1 =
(2.8)
nd
gd L B T
R
,
,
, ,
, , ,
)=0
v vd v 2 d d d d v 2 d 2
(2.9)
In criterion equation is explained with regard to the last term obtaining the expression
:
R=
v 2 d 2 1 (
nd
gd L B T
,
,
, ,
, , )
v vd v 2 d d d d
176
(2.10)
AFASES - 2008 -
3. Conclusion
After faking knowledge of the non dimensional complexes entering the criterion
equation, before building the model of the phenomenon studied, we shall establish the
connection relations between the scales of physical magnitudes determining these complexes,
that is we shall establish the law of the model.
Knowing the law of the distorted model, we can transpose the results of the
measurement obtained on the model to the ship prototype.
REFERENCES
[1]. Vasilescu Al., A., Analiz dimensional i teoria similitudinii, Editura Academiei,
Bucureti, 1969.
[2]. Reynolds, A., J., Curgeri turbulente n tehnic, Editura tehnic, Bucureti, 1982.
[3]. Bella, D., I., Dinamica della nave, vol I, Editore Briano, Genova, 1950.
177
AFASES - 2008 -
Victor BENCHE *
Radu RULESCU **
Rezumat
n aceast lucrare se prezint o succint informare privind realizri ale unor
instituii, universiti, agenii etc i tendine n construcia telescoapelor viitorului cu oglinzi
lichide.
Autorii prezint un studiu cu privire la obinerea unor parabolizi de revoluie prin
centrifugarea unor lichide cu fa liber n mediu controlat. Se evideniaz factorii de
influen (de proiectare i exploatare).
Aplicaia numeric obine caracteristicile principale definitorii (adncimea, poziia i
volumul parabolidului, viteza unghiular i turaia de antrenare, masa i greutatea lichidului
dizlocat etc.).
I. Introducere
Echilibrul relativ al lichidelor coninute n vase n micare de rotaie prezint
numeroase aplicaii tehnice industriale [1]:
Turnarea centrifugal a unor materiale, metale i aliaje topite, ca de exemplu
tuburi cilindrice din azbociment, piese din sticl, din bazalt topit etc; buce, inele,
cuzinei de arbori ai motoarelor cu combustie intern, bandaje de roi de vagoane
feroviare din fonte aliate etc. Turnarea centrifugal este echivalent cu turnarea sub
presiune n forme rotative.
Principiul separatorului centrifugal, al filtrului centrifugal, al unui motor hidraulic
cu jeturi reactive [3].
Principiul generatoarelor hidro-gazo-aero-dinamice centrifugale (pompe radiale,
turbocompresoare, turbosuflante, ventilatoare).
Principiul tahometrului hidraulic.
Apariia portanei n lagrele verticale.
Autorii cunosc realizarea unor lentile i oglinzi parabolice de mari dimensiuni prin
rotirea lent i rcirea ndelungat a unor materiale topite, coninute n vase/forme cu
suprafaa deschis. Topirea se poate realiza cu ajutorul unor cureni de nalt frecven (CIF).
n [2], sub titlul Oglinzi lichide pentru telescoapele viitorului se prezint informaia
de mai jos.
178
AFASES - 2008 -
Celebrele fotografii luate de telescopul spaial Hubble vor fi surclasate de cele pe care
le va transmite LLMT (Lunar Liquid Minor Telescope). Un proiect lansat de echipa
cercettorului canadian Ermanno Borra de la Centrul de optic, fotonic i laser al
Universitii Laval din Quebec i prezentat Ageniei NASA. E vorba de un telescop cu
oglinda lichid cu un diametru de 100 de metri, ce va fi construit pe Lun. Este de cinci ori
mai ieftin dect costul unui telescop clasic.
Instrumentul va putea detecta un furnicar" de peste 1000 de astre, asta n timp ce
succesorul lui Hubble, prevzut pentru anul 2013, nu va fi capabil s observe mai mult de 10
astre. O mare provocare pentru cosmotogi, care vor putea detecta obiecte spaiale formate n
timpul n care Universul avea mai puin de 6% din vrsta actual, adic 750 de milioane de
ani. S-ar putea, aadar, identifica galaxiile formate n prima etap a formrii Universului i,
probabil, prima generaie de stele.
Proiectul viznd telescoapele cu oglind lichida se afla n atenia cercettorilor de 15
ani. Conform legilor fizicii, fora centrifug imprim lichidului deversat pe un disc aflat n rotaie o curbur similar cu cea a unui telescop, far a se recurge, aadar, la lefuire. Viteza de
rotaie a oglinzii trebuie s fie extrem de precis, far ca suprafaa s prezinte cea mai mic
abatere/eroare. A fost ales ca lichid, mercurul (spre disperarea aprtorilor mediului) care se
evapora lent dar care formeaz rapid un strat de oxid ce blocheaz vaporii toxici.
Ideea a fost susinut i de cercettorii Universitii din Columbia-Britanic (Canada)
care au realizat un telescop cu oglind lichid avnd un diametru de 6 metri (tot cu mercur),
care se rotete cu o vitez de 5 km/h.
Centrul de optic din Quebec, n colaborare cu Observatorul european austral (ESO),
va construi un astfel de telescop cu diametrul de 4 metri, n Silla (Chile).
II. Elemente de calcul
n absena unei documentaii/memoriu de calcul, la ndemna specialitilor, privind
proiectele semnalate mai sus, autorii prezentului studiu elaboreaz i prezint cteva relaii i
calcule de hidrostatic adecvate, elementare/de baz.
Condiia echilibrului static al fluidelor este exprimat prin relaia vectorial:
r 1
f gradp = 0
(1)
r
In care f este rezultanta forelor masice care acioneaz asupra unitii de mas (fora
masic unitar, cu dimensiune i semnificaie de acceleraie), este masa specific a fluidului,
iar grad p este gradientul presiunii.
Raportat la referenialul triortogonal, cu planul xOy orizontal i axa vertical Oy de
rotaie, Fig.1, relaia (1) se transform/proiecteaz n trei ecuaii cu derivate pariale (Euler):
1 p
f x x = 0,
1 p
= 0,
(2)
f y
y
1 p
=0
f z
z
179
AFASES - 2008 -
Fig. 1.
n cazul lichidelor este justificat izodensitatea lor practic, = ct. n cmp
gravitaional, cu acceleraia gravitaional g, fz = -g. n cazul micrii de rotaie cu vitez
unghiular = .n/30, turaia n, fr = r.2 este componenta radial, orizontal, avnd proieciile
fx = x.2 i fy = y.2. Raza curent r [0, R=D/2], r = x 2 + y 2 .
n cazul turnrii centrifugale, la componenta hidrostatic a presiunii pst = gh se
adaug presiune radial, centrifugal:
r 2 2
[Pa]
(3)
pr =
2
Relaia (3) arat creterea ptratic a presiunii radiale cu raza, deci suprapresiunea
realizat are o distribuie parabolic.
Suprafaa liber a lichidului, n repaus, este plan, orizontal pentru c fz = -g, fr = 0.
n echilibru relativ de rotaie fz = -g, fr = r.2 i suprafaa liber echipotenial, izobar,
izoterm, devine, din punct de vedere geometric, o cuadric: un paraboloid de
rotaie/revoluie, de tipul X2+Y2-Z=0, (fig. 1).
Fa de planul orizontal iniial (n repaus) vrful paraboloidului coboar pe axa Oz cu
z, iar pe peretele cilindric al vasului lichidul se ridic tot cu:
R 2 2
z =
[m]
(4)
4g
Se consider adncimea paraboloidului pe axa Oz:
R 2 2
[m]
(5)
h p = 2 z =
2g
Volumul paraboloidului (volumul de lichid dizlocat) este:
1
Vp = R 2 h p =
R 4 2 [m3]
(6)
2
4g
Avnd masa:
M p = Vp = R 4 2 [kg]
(7)
4 g
180
AFASES - 2008 -
min
R 2 2
0,098m , volumul
Adncimea paraboloidului suprafeei libere h p =
2g
Vp =
R 4 2 = 1,4m 3 i masa dizlocat M Hg = Hg Vp = 19000kg .
4g
0
0,098
0,5
0,095
1
0,087
Tabel 1.
1,5
2
2,5
3
0,0735 0,0545 0,0294
0
IV. Concluzii
[1] Benche, V., Mecanica fluidelor i maini hidraulie, Editura Universitii Transilvania
din Braov, 1978;
[2] Benche V., rulescu R., rulescu S., Reactive centrifugal rotor the analytical study
of two applications, Proceeding of The Symposium CIVIL Engineering 2007,
Ovidius University Annals Constanta, 2007;
[3] Mran D., Oglinzi lichide pentru telescoapele viitorului, Revista Magazin sptmnal
cultural-tiinific, nr. 46 (2609), 15. 11. 2007, pag. 3, Bucureti.
181
AFASES - 2008 -
Victor BENCHE, Profesor Univ. Dr. Ing, Universitatea Transilvania din Braov, Facultatea
de Inginerie Mecanic, Catedra de Termodinamica i Mecanica Fluidelor, Braov, 500036,
Bdul Eroilor 29, Tel.: 0268412921 / int.193, Fax : +40 268 474761.
**
Radu RULESCU, Asistent Univ. Ing, Universitatea Transilvania din Braov, Facultatea
de Inginerie Mecanic, Catedra de Termodinamica i Mecanica Fluidelor, Braov, 500036,
Bdul Eroilor 29, Tel.: 0268412921 / int.193, Fax : +40 268 474761, E-mail
radu.tarulescu@unitbv.ro
182
AFASES - 2008 -
183
AFASES - 2008 -
Flags Great Adventure, de lng New York. Se numete El Toro i are nlimea
maxim de h = 55 m, ct un imobil cu 18 etaje, altitudine de la care viteza maxim de coborre ajunge la 113 km/h. Odat nceput marea aventur, miile de grinzi ncep s trosneasc
i s scrie din toate ncheieturile, melodie" deloc optimist, dar care e ntotdeauna
acoperit de urletele pasagerilor.
vt 2
Ceea ce provoac o reducere a energiei cinetice specifice
, cu v 2 = 0,9025 , de 9,75%,
2g
deci randamentulmecanic al vehiculului n instalaie este de 90,25%.
b) Cdere n vid
Impresionant alura lui El Toro, dar performanele sale nu se compar cu cele ale lui
Kingda Ka, din acelai parc american. Cu cei 139 metri nlime, este cel mai nalt montagne
russes din lume, coborrea fcndu-se practic pe vertical, cu calea de rulare doar puin
spiralat. Dup ce e urcat n vrful instalaiei, trenuleul e catapultat astfel nct n numai 3,5
secunde viteza lui de cdere ajunge de la 0 la 206 km/h. Un adevrat sadism din partea
constructorilor, corpul uman, dup calculele noastre, fiind supus la acceleraia:
206
a=
= 16,4 = 1,67 g [m/s2]
(3)
3,6 3,5
Un om avnd masa corporal n repaus de m = 75kg, capt, n cteva secunde, o for de
inerie de 125 kgf.
184
AFASES - 2008 -
185
AFASES - 2008 -
Fig. 4. Insanity
Cu o raz de R = 15m, dup calculele noastre, se asigur o vitz periferic:
60
u = R =
= 16,7 [m/s]
(6)
3,6
cu o vitez unghiular de rotaie = 1,11 s-1. Pasagerii sunt supui unei acceleraii unghiulare,
centrifugale, de R.2 = 18,5 =2.g m/s2.
III. Concluzii
186
AFASES - 2008 -
Victor BENCHE, Profesor Univ. Dr. Ing, Universitatea Transilvania din Braov, Facultatea de Inginerie
Mecanic, Catedra de Termodinamica i Mecanica Fluidelor, Braov, 500036, Bdul Eroilor 29, Tel.:
0268412921 / int.193, Fax : +40 268 474761.
**
Stelian RULESCU Asistent Univ. Ing, Universitatea Transilvania din Braov, Facultatea de Inginerie
Mecanic, Catedra de Autovehicule i Motoare, Braov, 500036, Bdul Eroilor 29, Tel.: +40.68.413000 Int. 173,
E-mail s.tarulescu@unitbv.ro
187
AFASES - 2008 -
Remus BOBOESCU *
Ion SPOREA **
Abstract: Sudarea laser a metalelor este un domeniu cu larg aplicabilitate.Puterea
laserului, viteza de sudare , intensitatea la suprafa sunt parametrii reglabili care
influeneaz calitatea suprafeelor cordoanelor. Sunt prezentate rezultatele experimentale
legate de studiul procesului. Ca funcie de rspuns s-a folosit lungimea conturului valului. Se
folosete metoda expermentelor factoriale pentru obinere unui polinom de regresie. Se
discut aspecte fizice i de calitate asupra proceselor.
I Introducere
Sudarea laser a materialelor metalice este n prezent un procedeu industrial larg
rspndit. Laserii cu mediu solid Nd:YAG pot fi utilizati la sudarea laser a aluminiului i
oelului titanului i aliajelor acestuia.Transmiterea radiaiei prin fibr optic i ataarea
capului de sudare la un sistem robotizat asigur flexibilitatea procesului. Cercetarea n
domeniu proceselor de prelucrare laser a cunoscut n ultimul timp o dezvoltare datorat
apariiei laserilor cu mediu solid care pot opera la puteri mari (3-4kW) n regim continuu i
apariia laserilor cu discuri care permit dezvoltarea unor puteri mari. n industria aeronautic
sudarea laser este utilizat pentru sudarea componentelor obiectelor interioare din aeronave
cum ar fi spre exemplu dulapuri din aluminiu sau umplutura din pereii avionului de tip
fagure de albine. Ca poziie de sudare se folosete sudarea prin suprapunere a tablelor subiri
i sudarea cap la cap cu sau far adaos de material. Aceste procedee tehnologice sunt
apropiate de experimentele realizate.Cercetarea n sudarea laser nregistreaz dou direcii de
dezvoltare una strict tehnologic iar cealalt ndreptat ctre nelegerea fenomenelor fizice
care au loc n timpul procesului. Iradierea cu o surs termic concentrat conduce la apariia
unor fenomene particulare legate de topirea i vaprizarea metalului topit.
Problema studiat n lucrare se refer la suprafaa cordonului sudat. Ea a fost analizat prin
studiul unor linii de fuziune i al unor iradieri pe plci de oel. Linia de fuziune prezint trei
caracteristici i anume o supranlare (sau o coborre) fa de suprafaa plcii, valuri regulate
datorate
solidificrii
topiturii
n
micare
i
ridictui
neregulate.
188
AFASES - 2008 -
Obinerea unor cordoane a cror calitate a suprafeei s fie constant i predictibil poate
constitui un obiectiv tehnologic. n lucrare sunt prezentate dependena unora din
caracteristicile suprafeei n funcie de parametri procesului. Cercetarea valurilor produse pe
suprafaa cordonului sudat a fost abordat n lucrrile [1] [2] i [3]. Valurile produse la
suprafaa cordoanelor apar ca oscilaii ntre fora grvitaional i tensiunea superficial care
apar la suprafaa topiturii. Ele apar acolo unde metalul lichid are o vscozitate ridicat
(datorit rcirii).Timpul ct dureaz unda fiind de acelai ordin de mrime cu timpul de
solidificare face ca valul s fie prins de solidificare n micare.
II Realizarea experimentelor
Experimentrile au fost efectuate la Institut de Trataiment et materiau Le Creusot France n
cadrul unui program finanat de Consiliul Regional Bourgogne. S-a folosit n experimentri o
sursa laser HL 3006D produs de firma TRUMPF care prezint urmtoarele caracteristici :
Este un laser cu mediu solid Nd :YAG avnd lungimea de und de =1064 nm. Pompajul
optic se realizeaz cu ajutorul lmpilor.Poate s emit att n regim continuu ct i n regim
pulsat Puterea maxim la suprafaa piesei este de 3kW.Parametru
de calitate al
fascicolului(BPP este egal cu 25 mm*mrad. Diametrul mediului fibrei optice de ieire este
de 600 m. Ca material s-a folosit oel carbon cu limit ridicat de elasticitate ( limit ridicat
de curgere) avnd o concentaie apropiat de 0.16%., tabla avnd o grosime de 10 mm.
Iradierile s-au efectuat n atmosfer protectoare de Ar. Pata generat de radiaia laser pe
suprafaa piesei a avut diametru de 600 m. A fost studiat variaia a trei parametrii : putrea
medie P, viteza de sudare v, poziia punctului focal fa de suprafaa piesei . Laserul a
funcionat n regim continuu. S-a urmrit cercetarea ntregului domeniu experimental prin
planuri clasice i planuri factoriale. Iradierile staionare au urmrit variaia timpului de
expunere i a poziiei punctului focal fa de suprafaa piesei. Figura prezint 1 exemple de
linii de fuziune iar figura 2 prezint exemple de iradierii staionare.
189
AFASES - 2008 -
Se observ din figura 4 c exist o tendin cresctoare cu puterea laserului pentru lungimea
valului. Coborrea punctului focal n interiorul piesei aigur o lungime a valuluir relativ
contant la puteri mici dar are un efect punernic la puteri mari. La puteri medii poziia
punctului focal ( defocalizarea) nu influeneaz lungimea valului. Din figura 5 se observ o
scdere a lungimii valului n funcie de vitez . Ea este mai pronunat pentru focalizarea la
suprafa i mai puin pronunat pentru focalizrile n interiorul piesei. Pentru acestea apar
diferite tendine de instabilitate. n mod suplimentar s-au efectuat cercetrii privind variaia
lungimii valului funcie de energia liniar ( raportul dintre putere i vitez i energia
volumic( mrime arbitrar definit ca fiind raportul dintre intensitatea la suprafa,
dependent de poziia punctului focal ,i vitez) . Aceste mrimi sunt date de formulele (1) i
P
[j/cm]
(1)
(2) El =
v
I ( )
Ev =
[j/cm3]
v
Introducerea acestor mrimi au cumulat datele din experimentarea clasic
Series 1
Series 2
6.5
Series 3
= 0 mm
= -2mm
= -4mm
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
P(kW)
0.5
1.5
2.5
190
AFASES - 2008 -
Folosind un plan de experimentare factorial s-a obinut urmtoarea funcie de corelaie ntre
mrimi:
lungimea valului = 3.461 + 1.58375 A 0.69375B 0.02375 C 0.27375 AB + 0.22125 AC +
+ 0.25375 BC + 0.16375 ABC
unde s-au introdus urmtoarele transformrii pentru parametrii variaii :
8.5
LW(mm)
Series 1
= 0mm
Series 2
=-2mm
Series 3
=-4mm
7.5
(3)
7
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
v(m/min)
-0.3
0.3
0.6
0.9
1.2
1.5
1.8
Series 1
Series 2
Series 3
=0mm
=-2mm
=-4mm
El(kJ/cm)
1
191
(4)
(5)
AFASES - 2008 -
(6)
l(mm)
Series 1
f(x)=1.6442499*ln(x)+2.1949488; R=0.493
8
7.5
7
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
-0.5
Ev(kJ/cm3)
2
10
12
14
16
18
20
22
24
26
28
192
AFASES - 2008 -
care este la prelucrrile mecanice. Pentru lungimea conturului valului se observ scderea
dependenei de vitez n raport cu cea pentru limea cordonului.Acest lucru face can studiul
cordoanelor de sudare s putem deosebi dou tipuri de mrimi i anume cele care au n
vedere dezvoltarea topiturii n material dup o singur direcie spaial i cele care au n
vedere dezvoltarea topiturii dup dou s-au trei direcii spaiale. Rezultatele experimentale au
artat c odat cu considerarea unei mrimi studiate care se dezvolt dup mai multe direcii
crete ponderea de influen a puterii asupra rezultatului final. Lungimea valului vizeaz
dezvoltarea topiturii dup dou direcii spaiale i se nscrie n acest tip de comportament
observat.Cercetarea valurilor impune i studiul altor mrimi. Este vorba despre frecvena
valurilor. Aceasta nu a putut fi msurat cu precizie deoarece fenomenul este afectat de
supranlare astfel c rezultatele nu sunt comparabile ntre cordoane diferite. Valurile se
unesc pe partea suprnlat de la mijlocul cordonului astfel c nu poate fi msurat cu
precizie frecvena lor. La marginea cordonului topitura curge liber i astfel iari se unesc
valurile.Exist o zon intermediar pe latura cordonului ntre centru i margine unde ar putea
fi msurat frecvena. Aici ns este dificil att prin msurtoare direct ct i prin
msurtoare indirect (pe fotografi). Cu toate acestea s-a stability ca o distan medie ntre
valuri apropiat de 0.1 mm care pare a nu depinde de parametrii variai. Un alt tip de
fenomene observate este instbiliti mari i anume ridicturi, valuri dublate sau ngroate care
sunt reproductibile la distane mari de ordinul a mai muli milimetri sau chiar centrimetri.
Spranlatrea este una din aceste mrimi . Ea se stabilizeaz dup sudarea a 5-6 cm. Astfel
nclzirea materialului naintea capului de sudare are un rol important n cadrul formrii
formelor de suprafa. Desi cordoanele au avut 10 cm ele nu au fost suficient de lungi
pentrupredicia unor astfel de mrimi. O alt problem o reprezint cantitatea de topitur. S-a
observat c aceasta este mai mic n cazul unei iradieri staiononare dect cea obinut n
cadrul procesului de sudare dei timpul de expune a fost acelai. Aceasta reconfirm
importana cumulrii cldurii n material i efectului acesteia asupra formei cordonului de
sudare.Prin fenomenele puse n eviden se arat c sudarea laser este un domeniu n care se
poate pune n eviden fizica topirii i vaprorizrii libere a materialului n condiiile n care
sursa termic este concentrat i nu aduce material strin n zona de interaciune. Astfel exist
un dublu interes n ceia ce privete studiul fenomenelor care au loc n sudarea laser din punct
de vedere al fizicii i din punct de vedere al tehnologiei.
V Concluzii
Bibliografie
[1] Prokhorov Laser Hearting of Metals International edition
[2] N. Rykalin, A Uglov, A Kokora Laser Machining and Welding Mir Publishers 1978.
[3] Olivier Perret Tese doc. Soutenu a Universite de Bourgogne
**
193
AFASES - 2008 -
I. Introducere
Transmisiile mecanice se utilizeaz practic n toate mainile fiind considerate ca cele
mai sofisticate componente. De funcionarea transmisiilor mecanice n cea mai mare parte
depinde fiabilitatea mainilor n general. Analitii scientologi consider, c n domeniul
tiinelor tehnice la scar mondial odat n 20-25 de ani se inventeaz un tip principial nou
de transmisie mecanic. Astfel, n anul 1923 inginerul german L. Braren inventeaz
transmisia cicloidal CYCLO, n 1944 inginerul rus A. Moskvitin - transmisia armonic cu
friciune, 1959 inginerul american C. W. Musser transmisia armonic cu angrenaj. n anul
1979 Ion Bostan a elaborat un nou tip de transmisie mecanic, care a intrat n circuitul
terminologic internaional sub denumirea de transmisie planetar precesional (TPP).
Transmisia planetar precesional se deosebete de cele clasice prin principiul nou de
transformare i transmitere a micrii i sarcinii i anume prin utilizarea micrii sfero
spaiale a satelitului i profilului convex concav variabil. Datorit acestor particulariti
novative n transmisia planetar precesional multiplicitatea angrenrii este de 100% (n cele
clasice 3-7%) fapt ce asigur capacitate portant sporit, gabarite i masa reduse, diapazon
cinematic extins 10... 3599 (n transmisiile armonice 79...300), precizie cinematic sporit
etc.[1]. Transmisiile planetare precesionale au fost publicate n peste 500 de articole
tiinifice. Soluii tehnice originale sunt protejate cu cca. 160 brevete de invenie i
implementate n domeniul mecanicii fine, utilajului tehnologic special, n complexele
robotizate pentru explorarea concreiunilor feromanganice de pe fundul oceanului (conceptul
U.R.S.S.), n tehnica cosmic de zbor etc. Realizrile n domeniu au fost prezentate la peste
50 de Expoziii Internaionale i apreciate cu 43 de medalii de aur i 20 de argint. n aceast
perioad autorii au elaborat: modele matematice ale cinematicii i dinamicii transmisiilor
planetare precesionale, n baza crora au fost cercetate caracteristicile funcionale ale acestora
194
AFASES - 2008 -
Z1 Z
Y1
R
K2
D
K
Y
7
K1
2
sunt
(1)
Z Cm = RC sin cos ,
( )
YCm = RC2 Z Cm
punctului
X Cm = 0;
Coordonatele
urmtoarele:
unde este unghiul de nutaie egal cu unghiul dintre axele OZ1 i OZ.
Micarea punctului Dm n raport cu sistemul de coordonate mobil legat rigid de
semifabricat se descrie cu ajutorul formulelor:
195
(2)
AFASES - 2008 -
X 1mD = X Dm cos
Y1mD = X Dm sin
+ Y Dm cos
Z1
Y Dm sin
Z1
Z1
Z1
(3)
Z 1mD = Z Dm .
Z Cm = RC sin sin ;
Z Cm m
Y = m ZC ;
YC
m
C
(4)
m
1D
= X cos
m
D
m
Y1D
= X Dm sin
Z1
Z1
m
D
X sin
X Dm cos
Z1
Y sin
m
D
+ YDm cos
Z1
Z1
YDm cos
YDm sin
Z1
.
Z1 Z1
Z1
Z1 Z1
Z1
Cunoscnd ecuaiile parametrice ale traiectoriei p. D i utiliznd elementele de transformri
din geometria analitic descriem traectoria micrii punctului E pe sfer, care, de fapt,
constituie profilul pe sfer al dinilor modificai. Traiectoria micrii p. E pe sfer se exprim
prin urmtoarele ecuaii parametrice:
m
m
= k2m Z 1E
+ d 2m ;
X 1E
m
d1m ;
Y1Em = k1m Z1E
m
=
Z1E
(5)
( k1m d1m k2m d 2m ) ( k1m d1m k2m d 2m )2 + ( k1m2 + k2m2 + 1 ) ( RD2 d1m2 d 2m2 )
,
k1m2 + k2m2 + 1
unde:
m
m
m
m
m
m 2
m
X 1D
X
X
Y
Y
Z
X
+
+
m
m
1D
1D
1D
1D
1D
1D
k1mY1D
+ Z 1D
m
m
k1 =
; k2 =
;
m
X 1D
m
m
m
m
m
Z 1D X 1D Y1D Y1D X 1D
d =
m
1
m
RD2 cos X 1D
; d
m
2
(R
=
2
D
m
cos + d 1mY1D
).
X
m m
m
m
X 1D Y1D X 1D Y1D
196
m
1D
AFASES - 2008 -
unde m =
AX
D
,
m
+ BY1Em + CZ1E
m
1E
(
B = (X
C = (Y
D = (Y
YEm = m Y1Em ,
m ,
Z Em = m Z1E
(6)
iar
A = Z E 2 Z E 1 nY Y E 2 Y E 1 n Z ;
E2
E2
E1
)
)n
X E1 n Z Z E 2 Z E1 n x ;
(
)n
Y E1 n X X E 2 X E1
Z E 2 Y E 2 Z E1
+ X E 2 Z E 1 Z E 2 X E1 nY + X E 1 Y E 2 X E 2 Y E 1 n Z ;
n X = Y E1 Z E 2 Y E 2 Z E1 ;
nY = X E 2 Z E1 X E 1 Z E 2 ;
n Z = X E1 Y E 2 X E 2 Y E1 .
n care: XE1, YE1, ZE1 i XE2, YE2, ZE2 sunt coordonatele punctelor minime ale profilului
dintelui iniial E1 i final E2; nX, nY, nZ proieciile vectorului n coliniar cu vectorul vitezei
punctului D al profilului n arc de cerc.
Profilul modificat al dintelui se descrie de ecuaiile
m = X E m cos
;
+ R D cos ( + + ) + Y E m sin
Z1
Z1
= X E m sin sin
Z1
+ R D sin ( + + ) + Z E m cos .
tg ( + +
unde s in =
cos
Z1
+ tg
cos
( +
;
+
cos =
cos
Z1
+ tg
(7)
Z1
2
( +
.
+
Fig. 2. Profilul dintelui modificat (a) i modelul computerizat 3D al danturii roii dinate (b).
197
AFASES - 2008 -
198
AFASES - 2008 -
199
AFASES - 2008 -
a.
b.
Fig. 3. Profilogramele dinilor roilor dinate centrale: a Z1=27; b Z4=20.
a.
b.
Fig. 4. Modelul computerizat 3D al roilor dinate: a) Z1=20; b) Z1=27.
Reductor precesional cu roi dinate din materiale plastice i pulberi metalici. n
rezultatul analizei materialelor plastice recomandate sub aspectul rezistenei i propirietilor
antifriciune pentru fabricarea roilor dinate a fost aleas masa plastic Hostaform produs
de Compania Ticona, Germania. Pentru fabricarea roilor dinate a fost elaborat, proiectat i
fabricat forma de turnare. n fig. 6,a,b este prezentat prototipul industrial al reductorului
precesional cinematic cu raportul de transmitere i=-144 destinat pentru transmiterea
Fig. 6. Reductor precesional cinematic (a) cu bloc satelit din mas plastic (b) i din
pulberi metalici (c).
200
AFASES - 2008 -
momentului de torsiune T=18 Nm. Pentru transmiterea unor sarcini mai mari a fost elaborat,
proiectat i fabricat prototipul industrial al reductorului precesional cinematic cu roi dinate
executate prin sinterizare din pulberi metalici, prezentat n fig. 6,a,c. n acest scop a fost
elaborat, proiectat i fabricat forma de presare. Pentru fabricarea roilor dinate a fost
selectat pulberea metalic SUMIRON 4100S. Pentru mbuntirea proprietilor antifriciune pulberea conine adaos de lubrifiant solid MoS2. Reductorul precesional realizeaz
raportul de transmitere i= - 72 i asigur transmiterea unui moment de torsiune T = 56 Nm.
Modul electromecanic precesional de acionare din componena staiei sovietice
interplanetare VEGA 6. n baza sarcinii rezultatelor obinute i a sarcinii tehnice formulate
de ntreprinderea Cometa din or. Moscova a fost elaborat, proiectat i fabricat modulul
electromecanic precesional (fig. 7) alimentat de la baterii solare, avnd n componena sa
Bibliografie
[1] Bostan I. Prechessionnye peredachi s mnogoparnym zachepleniem. Chiinu, tiina,
1991, 356p.
[2] Bostan I., Dulgheru V., opa M., Vaculenco M. Brevet nr.1886 (MD). Angrenaj
precesional i procedeu de realizare a lui. U.T.M. I.cl.: F16 H1/32. Publ.31.03.2002. Bul. nr.3.
[3] Bostan I., Dulgheru V., Sochireanu A. Brevet nr. 2821 (MD). Transmisie planetar
precesional. U.T.M. I.cl.: F 16 H 1/ 32. Publ. BOPI - 2005. Nr.7.
[4] Bostan I., Dulgheru V., Grigora . Transmisii planetare, precesionale i armonice. Atlas.
Editura Tehnic-Bucureti, Editura Tehnica UTM., 1997, 200P.
[5] Bostan I., Dulgheru V., Sochirean A. Brevet nr. 3058 (MD), I.cl.: B60j1/17. Mecanism de
acionare. U.T.M. - Nr. 2005 0137. Publ. BOPI - 2006. Nr. 5.
*Ion Bostan, academician, Universitatea Tehnic a Moldovei, Bdul tefan cel Mare 168, Chiinu, Republica
Moldova, 373 22 23 78 61, e-mail: bostan@adm.utm.md
**Valeriu Dulgheru, prof.univ.dr.hab., Universitatea Tehnic a Moldovei, Bdul tefan cel Mare 168, Chiinu,
Republica Moldova, 373 22 509939, e-mail: dulgheru@mail.utm.md
201
AFASES - 2008 -
I. Introducere
Intruct literatura de specialitate nu ofer ntotdeauna date suficiente pentru verificarea
rezultatelor teoretice, s-a impus efectuarea de determinri experimentale proprii. n cercetarea
experimental au fost utilizate aparate i dispozitive aflate n dotarea Laboratorului de
Mecanica Contactului al Facultii de Inginerie Mecanic din cadrul Universitii tefan cel
Mare Suceava.
Lucrarea prezint un studiu experimental privind comportarea la contact, la solicitare
static a unui inel din cauciuc (Di = 6,18 mm, De = 12,65 mm), folosit la dispozitivul de
acionare de la stingtoarele de incendiu. Studiul experimental folosete inelul din cauciuc n
mai multe moduri: simplu ori montat n dispozitivul metalic aa cum este folosit la
stingtoarele de incendiu, cu lubrifiant sau fr lubrifiant. Se urmrete evoluia elementelor
contactului liniar (limea ariei inelare de contact, apropierea) n funcie de fora de ncrcare
(F) i timp (t).
Standul experimental folosit este compus din mai multe pri componente, ce vor fi
descrise n lucrare. Iniial, pentru vizualizarea ariei de contact s-a folosit un microscop de
cercetare MC 6, dar pentru c deformrile sunt mari, imaginea contactului este nregistrat cu
ajutorul unei camere video digitale (pentru deformarea n timp) i cu ajutorul unui aparat foto
digital (pentru diferite valori ale forei de ncrcare). La ncrcarea - descrcarea static, s-au
luat imagini pentru aria de contact la fiecare valoare a forei (4,81 N 48,16 N). Pentru
studierea variaiei ariei de contact n timp, la aceeai valoare a forei, s-a filmat cca. 5 min. la
fiecare determinare, la o for de 48,16 N i datele au fost nregistrate n calculator sub form
de videoclip, n sistem avi. Cu ajutorul unui program, videoclipurile au fost transformate n
202
AFASES - 2008 -
fotografii jpg (25 foto / s). Astfel, s-au obinut numeroase puncte experimentale ce vor fi
analizate, aflndu-se variaia dimensiunilor ariei de contact n timp sub aciunea unor fore
statice de ncrcare - descrcare diferite.
3
2
203
AFASES - 2008 -
6
4
5
10
11
9
Fig. 2: Dispozitivul de acionare (pri componente)
Iniial, pentru vizualizarea contactului, s-a folosit un microscop de cercetare MC6.
Deoarece deformaiile sunt foarte mari (inelul fiind din cauciuc), aria de contact nu poate fi
vizualizat i s-a folosit o camera video/foto digital.
2.2. Dispozitivul de realizare a contactului liniar i a ncrcrii - descrcrii
statice
Dispozitivul este fixat pe o mas deplasabil n coordonate X-Y, aa cum se observ n
figura 1, pentru poziionarea central a contactului n cmpul vizual al obiectivului camerei
video-foto digitale [1].
204
AFASES - 2008 -
(a) fr lubrifiant
(b) cu lubrifiant
(a) fr lubrifiant
(b) cu lubrifiant
Fig. 5: ncrcare descrcare inel din cauciuc montat n dispozitivul metalic (48,16 N)
205
AFASES - 2008 -
k th = 9,602 10 3 N / m
iar cea experimental:
kexp = 9,633.103 N/m, eroarea fiind 0,323%.
Fora se determin cu relaia:
F = pkexp
(1)
Filmul unei determinri, fiind un videoclip de cca. 5 min., este stocat n calculator ca
fiier n format avi, prin intermediul programului Windows movie maker. Cu ajutorul
programului de prelucrare Frame Shots, fiierele avi sunt prelucrate n fotografii jpg (25
foto/s). Pe aceast imagine jpg se poate determina aria de contact cu ajutorul programului
Autocad 2006.
206
AFASES - 2008 -
Deplasarea
[mm]
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
Forta
[N]
4.81
9.63
14.44
19.26
24.08
28.89
33.71
38.53
43.34
48.16
43.34
38.53
33.71
28.89
24.08
19.26
14.44
9.63
4.81
cu lubrifiant
fr lubrifiant
L1
[mm]
L2
[mm]
L1
[mm]
L2
[mm]
0.5
0.6
0.75
0.88
0.98
1.06
1.12
1.21
1.28
1.44
1.31
1.27
1.24
1.16
1.1
1.05
0.97
0.82
0.56
0.33
0.59
0.76
0.91
1.03
1.22
1.42
1.36
1.42
1.49
1.51
1.5
1.28
1.24
1.13
1.07
1.01
0.84
0.53
0.47
0.59
0.73
0.78
0.93
1
1.08
1.1
1.2
1.29
1.13
1.08
1.01
0.87
0.81
0.68
0.51
0.44
0.25
0.23
0.29
0.59
0.79
0.79
0.85
0.93
1.02
1.08
1.12
1.05
1.05
0.96
0.92
0.88
0.8
0.7
0.55
0.21
L2 cu lubrifiant
L1 cu lubrifiant
60
50
50
40
40
F [N ]
F [N ]
60
30
30
20
20
10
10
0
0
0.5
1.5
0.5
1
L [mm]
L [mm]
(a)
(b)
207
1.5
AFASES - 2008 -
L1 fr lubrifiant
L2 fr lubrifiant
60
50
50
40
40
F [N ]
F [N ]
60
30
30
20
20
10
10
0.2
0.4
0.6
0.8
1.2
1.4
0.2
0.4
L [mm]
0.6
0.8
1.2
L [mm]
(c)
(d)
Fig. 7
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Timpul
[s]
1
2
3
4
5
6
7
8
9
10
60
120
180
240
300
Forta
[N]
cu lubrifiant
48.16
fr lubrifiant
L1
[mm]
L2
[mm]
L1
[mm]
L2
[mm]
1.25
1.38
1.37
1.29
1.22
1.28
1.29
1.26
1.31
1.29
1.32
1.25
1.33
1.27
1.27
1.06
1.09
1.03
1.09
1.19
1.23
1.08
1.03
1.11
1.1
1.15
1.12
1.21
1.17
1.25
1.24
1.35
1.34
1.3
1.32
1.36
1.34
1.41
1.37
1.3
1.3
1.33
1.35
1.3
1.29
1.35
1.33
1.31
1.3
1.36
1.35
1.3
1.34
1.3
1.33
1.33
1.35
1.4
1.37
1.29
L2 cu lubrifiant
L2 fr lubrifiant
1.4
1.2
1
0.8
0.6
0.4
0.2
0
(a)
L1
L1
1.42
1.4
1.38
1.36
1.34
1.32
1.3
1.28
50
100
150
200
250
300
350
(b)
t [s]
50
100
150
200
t [s]
208
250
300
350
AFASES - 2008 -
L1 cu lubrifiant
L1 fr lubrifiant
1.45
1.4
1.4
1.35
L1
L1
1.35
1.3
1.3
1.25
1.25
1.2
1.2
0
50
100
150
200
250
300
350
t [s]
50
100
150
200
250
300
350
t [s]
(c)
(d)
Fig. 8
IV. Concluzii
n vederea aprecierii comportrii garniturilor inelare din cauciuc la contact, s-au efectuat
ncercri experimentale noi. Acestea vizeaz att variaia limii fiei de contact la un
ciclu de ncrcare descrcare ct i evoluia acesteia n timp sub sarcin constant.
ncercrile de ncrcare descrcare evideniaz prezena histerezisului mecanic.
ncercrile sub sarcin constant arat tendina sistemului de a se stabiliza printr-o serie de
oscilaii n jurul valorii finale.
Sunt necesare noi ncercri i investigaii teoretice pentru nelegerea deplin a
fenomenului.
V. Bibliografie
[1] Ciornei, F.C., Research upon Dynamic, Hertz like Circular Contacts in Linear
Viscoelastic Domain, in Romanian, Ph.D.Thesis, University of Suceava, (2004).
[2] Greenwood, J. A., and Jhonson, K. L., The mechanism of adhesion of viscoelastic
solids, Phil. Mag. Of Roy. Soc. Lond., A43, 697 711, (1981).
[3] Jhonson, K. L., Contact mechanics, Cambridge University Press, (1985).
[4] Ciutac, F., Diaconescu E., and Pintilie D., Experimental investigations of contact radius
between a rubber ball and a flat, 10 th International Conference on Tribologie, ROTRIB 07,
November 8 10, 2007, pp. RO-047-1 - RO-047-4.
*
Cpt. Drd. Ing., Inspectoratul pentru Situaii de Urgen Bucovina al judeului Suceava, ROMNIA
Prof. Univ. Dr. Ing., Universitatea tefan cel Mare Suceava, ROMNIA
**
209
AFASES - 2008 -
210
AFASES - 2008 -
Fora de
traciune
lateral
Dispozitiv de
admisie
Jet direcionat
211
AFASES - 2008 -
(4)
(5)
Pins=Nmot Pmd
(6)
(7)
(8)
(9)
(10)
P1,P2,P3 puterea dezvoltat pentru traciunepentru cele trei cazuri: elicopter clasic Dauphin,
un elicopter cu jet MD520N, MD520N cu grinda propus (fig.1).
212
AFASES - 2008 -
Pmax.
P3
90
P2
P2
P3
85
P1
P1
80
P1
P2
P3
75
70
C max. 100
C1
C1
C2
C3
C2
C3
80
C1
60
C2
40
C3
20
0
213
AFASES - 2008 -
BIBLIOGRAFIE:
1.Revista TOP GUN nr.6din1999
2.Revista TOP GUN nr.1din 2000
3.Mecanica clasic , Dumitru Luca IAI, 2007
4.Teoria, performanele i construcia elicopterelor, Adrian Postelnicu, Cluj - Napoca, 2001
5.Proiectarea structurilor de aeronave,Augustin Petre, Bucureti, 1999.
Crciu Ionic, instructor ef- Academia Forelor Aeriene ,,Henri CoandBraov, str.Mihai Viteazu nr.160,
tel.423421, e-mail:circiuionica@yahoo.co.uk
214
AFASES - 2008 -
Abstract:
Within an informatics based society, data traffic globalisation gives a multinodal
system that covers the whole industrial activity, including the foundry type economic agents.
Computer stands for the basic part of a data administration system; metallurgist engineers
must become efficient users of the computer informatic system and relearn to design and
manage technological processes with these dedicated applications.
215
AFASES - 2008 -
General usage systems, which help solving a large range of problems from
several fields;
Local systems, programmes necessary to process data when data are on one
calculation system;
Network systems, the system works within a computer network, in which case,
data and programmes can be distributed to several workstations that are part of
that network.
Depending on data localization and the processing place, there are different informatic
systems:
1. With centralized data, data are on one calculation system only;
2. With distributed data, data are distributed on several computers within the
network;
3. With centralized processing, data processing is being done on one workstation
only, regardless of the number of stations that host data to process;
4. With distributed processing, several computers process data come from one or
more computers of the network.
After the hierarchical level of informatical systems in the societys organizatorical
structure, there are:
Informatic systems for activities management within economic units;
Organisation system with group structure;
Territorial informatic systems;
Branch, subbrench and national economic systems;
General use systems.
II. DATA BASES MANAGEMENT SYSTEMS (SGBD)
An applicaion (an achievable programme) is a programme system designed to
perform a well determined set of operaions on data bases; the application consists of:
1. the very programmes (main programme, coordinator, menu programmes,
screens, interrogations, indexes, actualisations, reports);
2. data bases.
Applications performance takes place at two levels:
Interpretative when a programme called interpreter takes each enunciation
of application, translates it into an internal code and performs error analysis, executes it, then
passes to the next enunciation;
Compilative when the whole application is being translated by the compilator
programme into an intermediate code, memorised on the disk, called object code, submitted to
suplementary processing by the connection editor in order to get the final achievable form of the
application. The performing is done under the operation systems control.
Achievement of a data base implies: analysis of the system for which data bases are
being built; designing the base structure; loading data into database; data base exploitment
maintenance.
Effective achievement of an application implies: assign task, task analysis, application
design; application coding, modulus testing, implementation of application, application
maintenance.
1. Assign task. Task is assigned by the application beneficiary, in accordance with
the activities that are about to be modeled.
2. Task analysis. This operation implies identification of data types, connections
between them, necessary operations for their management. Task analysis ends with the
assignment of four data multitudes, whose rigour and exactitude influence the veracity of
216
AFASES - 2008 -
applications results: description of input data; description of data stored in the database; list
of performed processings on data; description of reports data.
3. Application design. At this stage, we achieve data and programmes structure
design. Designing programme structure implies detailing the modulus necessary to
application achievement: modulus for folders creation, data implementation, processing and
result abstraction, error correction etc. These modulus are controled and coordinated by the
main programme, that has the following structure: main programme; global picture
declaration; global variables initialization; environment state saving (general iniialization).
4. Application coding. If at the precedent level called initial designing of
application the detailing level has a pseudocode type, at this stage application is written in
a specialized language, respecting the rule imposed by that one.
5. Modulus testing. At this stage, modulus are being checked, possible errors are
detected and corrected, extreme cases analysis is being performed, tests are being designed.
6. Applications implementation. We build the final form of application by gradual
integration of tested functional modulus.
7. Application maintenance. We must remove the errors pointed out by the user
during the warranty period, modernise and upgrade the application.
Database users can be:
Non specialist users (conversational) that are offered a form of communication
with database, close to current speech;
Specialist users who know the structure of database;
Database administrator is a special user, who defines database exploitations
objectives, shares users access rights, elaborates database design conception, is in charge of
all the activities and operations refering to database, helps defining users demands etc.
Databases languages. Within Databases Management Systems (SGBD), data
declaration and and manipulation functions are being done with different languages:
a. Data defining languages (LDD). LDD functions are: performs entities and their
attributes definition by names, memorising form, length, specifies the connections between
data and access strategies to them, assigns confidential different criteria, defines used data
automatic validation criteria.
b. Languages for data manipulation (LMD). Operations performed on databases
require a specialized language, where commands are expressed in sentences that describe
actions upon the base. A command has the following structure:
- operation, that can be an arithmetical or logical calculation, editation,
extraction, opening-closing, manipulation (introduction, add, delete
etc.);
selection criteria (for, while, where etc.);
access mode (secvential, indexed etc.);
edit form;
c. Data control languages (LCD). A database control implies: assuring data
confidentiality and integrity; data saving in case of damage; getting performance; solving
competition problems;
d. Universal languages. An universal language is rarely used to manage a database.
The interface between user and SGBD is achieved in two ways:
Using a calling mechanism inserted in the application programme. This
mechanism can be a CALL or another key word. A system that allows this type of mechanism
is called a SGBD wih a host language;
Using a special command, used independently. n this case, the managing
syistem is called autonomous SGBD. Still, there is a special interface, which is capable to
interprete the commands of the sollicitation language.
217
AFASES - 2008 -
III. USINGE THE COMPUTER FOR DESIGN AND RESEARCHDEVELOPMENTTYPICAL USAGE DOMAINS
Within the general context of optimizing the functioning of the industrial systems, the
computer is used to: 1. the so called designing; 2. achievement of some data bases; 3.
shaping the technological processe; 4. leading the processes towards working places; 5.
achievement of Expert Systems = SE.
The first stage of the fabrication conception/designing, must anticipate the technictechnological evoluion frame and be at a time level with < + 2-4 units >, in an execution
report.
Accordingly, the usage of computer and some programmes is implemented and <
authenticated > during the designing process; implicitely, designing was the first <step> even
in the case of number command units, and the peforming processor was subsequently
implemented.
At international level, there have already been established a series of principles of the
fundamental discipline, the theoretical base, including already known terms elements.
Our country confrunts itself with the same dilemma as in other domains (longlasting
development ecology, globalization, natural resources limitation, energetical alternatives
etc.): < taking the same stepsi, including the failures already known > or < selection of the
essential points and framing at optimal levels only >; chosing the variant is conditionned by
the difference of the total costs. In this regard, we may think about the < mechanizationautomatization > stages or integration by < robotization >.
We must give a series of explanations as far as terminology is concerned.
The designing working technique, in connection to computers usage, through the
appliance of some apprpopriate programmes and equipments, is defined as CAD = Computer
Aided Design. The concatenation of the designing programmes with data processing systems,
in the engineering science, is generically called CAE = Computer Aided Engineering. The
integration of an enterprise into a complex and complete system of data monitoring,
processing and administration is called CIM = Computer Integrated Manufacturing.
The CAD applications, with the designing assistance programmes systems are the
most famous and are mainly made up of the scheme domains from Figure 1, as it follows:
1. a communication domain organizes data input and output to and from the
designing department;
2. a method domain contains the working modulus corresponding to shaping,
informing and calculation;
3. a data administration and management field, namely an integrating system of the
data bank organizes all the data transfers and stockings between the method algorithms and
communication domain on the one hand, and the data banks networks or specialized or
standard individual folders, on the other hand;
4. the designing data base containts all the stocked geometrical and nongeometrical
data, which are necessary both to the designing methods and to the communication between
the projecting operator and the CAD system.
A fundamental characteristic of the CAD systems is made up of the RID concept =
Geometrical Objects Internal Representation. This system is built on a real object; through
the abstract process, one can get a mental virtual object, from which one can get an
informational object, through formalization and multiple filtrations in fundamental specific
languages. By a series of transformations and transposings based on the binary code, one can
constitute an internal model in the computers memory = RIM (see. Fig. 2).
In Figure 3, we give the very simple scheme of the tridimensional representation
means essence of objects technical bench-marks and of biunivocal transformations from an
218
AFASES - 2008 -
informational model (with discrete sizes, quantizable in volumes, surfaces, contours and/or
points) to an internal model of the computer, RIM type.
An object representation can be achieved through a volume, surface or line model (the
so called the edges models or the wire models).
The volume model (spatial) can be recomposed in a way oriented towards bodies or
surfaces. For calculations and analytical integrations one needs the internal generation of a
tridimensional body, therefore, the construction of a spatial model oriented towards bodies.
Having a surface model, one is able to sufficiently describe volumes for constant thickness
pieces only (e.g. tin pieces) or for rotation symmetry pieces.
Fig.3. RIM
principle general
scheme of
geometrical
objects internal
representation
performing, in the
computers
memory
Fig.2.
Abstractization
models of technic
objects.
219
AFASES - 2008 -
Plane and rotation symetru pieces can be clearly described only by the bidimensional
representations = < D Models>, surfaces or lines. These models are appropriate not only for
the internal representation of views and sections; for the normal-classic drawings execution,
this patterns is sufficient (v. fig. 4).
From < 2D Views >, hand designed, scanned and introduced into the computer,
through reconstruction automatic techniques < 3D Models > are automatically generated.
The macros technique is based on the usage of the already existant geometrical
models, namely stocked for simple pieces, classic sets, fundamental structures; from case to
case, these patterns = standard solutions are submitted to adaptations, adjusting,
redimensioning. It is very productive to import from the data bases stocked-archived
representations = finished elements, pieces and standardized-tipified solutions by usual
devices and cars organs.
n the CAD field, a package of programmes can contain a large series of individual
programmes and programme systems:
1. Configuration programmes, that can integrally perform a geometrical
representation, calculation and preprocessing of the designing data, in an interractive form; in
a special way, there must be possible a shaping of the tridimensional geometry of the
computers internal model, on the basis of which rapid, accurate and exact regrant of classic
projections of the usual technicalal drawing should be possible.
2. Programmes specialized in data organization (the fundamental principles of
solving standardized pieces, materials, cost data, rate-setting times values, data connected to
specific consummation); such data banks systems are easy to use with searching dialogue
interractive facility programmes;
220
AFASES - 2008 -
BIBLIOGRAPHY
1. Chira I. .a. Methodological and bibliographical rule for yearly projects in the
Shaping Theory. Litografia I.P.B.Publishing House Bucharest, 1989.
221
AFASES - 2008 -
Chira, Ion, Professor Ph.D eng., Departament of Cast Metals, POLITEHNICA University of Bucharest,
Romania, e-mail: iondanchira@gmail.com
Ionica, Ruxandra, Cpt. Drd. Mat., Serviciul de Telecomunicatii Speciale, e-mail: ruxandraionica@yahoo.com
(Nemes) Beca, Ruxandra, Drd. Eng., Inspectoratul General al Jandarmeriei Romane, e-mail:
ruxa_angel@yahoo.com
Teodorescu, Constantin Bogdan, Drd. Ec., ING WholesaleBanking Bucharest, e-mal: bogdanteo@pcnet.ro
222
AFASES - 2008 -
Abstract:
The present work stands for a brief synthesis of the last ten years accomplishments in
the field of softs dedicated to foundry technologies. This view of the main applications that
inforced themselves in the technological practice from abroad and, partially from Romania,
should frame the computer work informatic future tendencies, technological designing in the
CAD/CAE-CAM/CIM systems, the general levels and standards against which we must lead
the structure of the metallurgic national department, organization of the foundries and their
management, towards maximum efficiency and optimal interconnection to the European
circuit.
223
AFASES - 2008 -
224
AFASES - 2008 -
meanwhile assuring their structural adaptation to the functional requirements and specific
contextual situations of the particular problems to be solved.
For an industrial process, such as the High pressure pouring integrated technologic
cell (TPI) the computer conducting scheme is principially presented in Figure 2.
The thermic-technologic refining,dosing and pouring operations, in the presented
scheme can be detailed, and the principle component elements can be particularized.
The standard configuration architecture of the expert system and the soft
correspondances designing with real technical-technologic elements consist of (see Fig. 3):
1. Automatics elements: Identification of process behaviour models; Automatic
leading of the technological process, with informatic captators and interfaces;
2. Supervising: Process documents emission; Cartographic vectoring, as a complex
computer graphic representation.
225
AFASES - 2008 -
and represents a situation that imposes a new approach and interdisciplinary treatement. In
Figure 4, we presented the scheme of a technic data flow proposal within a foundry, that is in
the course of assimilating the technology designing on the computer, through shaping and
simulation of melting charge pouring-flowing, solidification-cooling in a form, or the difficult
issue of pouring network heasing and optimising.
Similarly, in Figure 5, we indicate a logical succession for the poured pieces
solidification shaping and simulation, and in Figure 6, we introduced the solidification model
setting stages.
Fig. 3 The architecture of the piloting expert system of the TIP pouring step, in the
integrated technological cell.
III. PRINCIPLES IN TECHNOLOGIC SOFTWARE ELEMENTS
DESIGNING
3.1 The metallic melting charge flow during the pouring process.
The numeric simulation of the pouring processes, therefore the hydraulic elements and
metallic melting charges flow are recent history due to the NAVIER-STOKES flowing
situation and the nonliniar thermic transfer equations complexity in a tridimensional system.
Solving these equations using numerical methods, the finished differences and finished
elements methods were possible together with hardware and operating system development.
The programmes meant to solve the fluids flowing and mixtures solidifications
problems appeared only in the 1980s due to the complex problems generated when solving
fluids free surface behaviour. At the present time, much significance has been given to
the programmes that start from the finished differences method because of the rapidity
with which they solve the flowing problems and, so, due to their efficiency in industrial
applications where the response time is essential.
226
AFASES - 2008 -
The mixtures flow under gravitational, viscid and pressure gradients forces is
decribed by the speed and pressure field whose evolution is given in the Navier-Stokes
equation, the continuity equation and the free surface conditions.
3.2 Solidification of metallic melting charges during cooling in the pouring form.
The thermical transfer during the flowing and solidification processes is ruled by the
FOURIER EQUATION thermic conductibility. This equation is accompanied by the initial
and limit conditions, Cauchy Type (when we consider the external environment temperature
and the global coefficient of heat exchange between environment and the analysed system, as
programme input data; this case appears when the heat exchange conditions are given to the
systems external part, through thermic convection), Dirichlet (when we consider as input
data the temperature of the ingot surface or of the poured piece ) and Neumann (when we
consider as input data the thermic flow on the border segment considered, as in the case of
radiations or external heat source). The specific equations that we mentioned apply to the
participant domains.
227
AFASES - 2008 -
228
AFASES - 2008 -
229
AFASES - 2008 -
To solve the physical shaping we must use different techniques and methods; the
special literature shows numerous case studies in this field. For example, with the SOLA VOF = Solution Algotithm - Volume of Fluid method, based on an explicit finished
differences scheme, applied on a crossing 3D graphic grating, we can solve a series of
problems with regard to steel ingots pouring.
IV. DEDICATED TECHNOLOGICAL SOFTS
230
AFASES - 2008 -
BIBLIOGRAPHY
1. Chira I. Paper review: Htte Engineers Guide. Basis. Translation from German,
following the 29th edition. Tehnica Publishing House Bucharest,. In: Foundries Magazine,
No. 1, 1998, p. 32-33, ISSN 1224-2144.
2. ***Site.Web(servers.www=WorldWideWeb...format.http://=HyperTextTransferProtoc
ol)/Homepage(interconnected.network.address.URL=Uniform.Resource.Locator):..bizoo.ro/f
irma/metalprogres.soft;cad.magazine.ro/maxcad/templates/detalii;computer.world.ro/;
csit.sun.pub.ro/research/history/;cyco.com;inas.ro/content/view/;greentwinkie.com/blo
g/utilitare.optimizare/utilitar.de.monitorizare;info.legal.ro/;math.md/imisite/ro/progbd.sh
tml;networked.ro/;networking.info.europa.ro/;pc.forum.ro/view.topic;pc.world.ro/in
dex;quartz.matrix.ro/sections/automatizari.industriale/conducere.monitorizare/;sirom.kap
pa.ro;synchro.ro/.
3. Bratu Ctin. and Popa I. Solidification coefficient assessment through numerical
shaping. In: Foundries Magazine, No. 1, 2000, p. 37 - 39, CZU 621.74 (051), ISSN 1224 - 21
44.
4. Chira I. - Relansin Contract No. 720/2000: Functional pattern achievement of a
technological cell (inclusively soft and hard) for under pressure pouring.
Chira, Ion, Professor Ph.D eng., Departament of Cast Metals, POLITEHNICA University of Bucharest,
Romania, e-mail: iondanchira@gmail.com
Ionica, Ruxandra, Cpt. Drd. Mat., Serviciul de Telecomunicatii Speciale, e-mail: ruxandraionica@yahoo.com
(Nemes) Beca, Ruxandra, Lt. Drd. Eng., Inspectoratul General al Jandarmeriei Romane, e-mail:
ruxa_angel@yahoo.com
Teodorescu, Constantin Bogdan, Drd. Ec., ING WholesaleBanking Bucharest, e-mail: bogdanteo@pcnet.ro
231
AFASES - 2008 -
Capitolul I. Introducere
Lucrarea prezint un studiu experimental asupra elementului elastic metalic, arc
elicoidal, cu diametrul barei de 24,4 mm, utilizat la echiparea boghiurilor Y25Lsdi destinat
vagoanelor de marf. Au fost efectuate ncercri la oboseal, determinri ale deformaiilor
relative, respectiv a tensiunilor i de durat de via [1], [2], [5].
Pentru verificarea tehnologiei de fabricaie i a tratamentelor termice i mecanice
aplicate a fost ncercat un set de arcuri la oboseal n scopul depistrii cauzelor ruperilor n
exploatare. Obiectivul impus de cercetarea pe care am efectuat-o este acela de a reliza arcuri a
cror rezisten la oboseal s fie la nivelul prevzut de cerine (4000000 cicluri n trei
regimuri de ncercare), cerine ce asigur o rezisten la oboseal optim n exploatare impus
de normele internaionale ale UIC (Uniunea Internaional a cilor ferate Europene).
Setul de arcuri are caracteristicile mecanice ce indic valori ale limitei de curgere de
1489-1551 N/mm2 (fa de min. 1370-1670 N/mm2 prescris) i valori ale alungirii la rupere de
10,8-12,8% (fa de min. 6% admis). Diametrele iniiale ale barelor au valori de 24,33 24,34
mm ncadrndu-se n intervalul prescris (24,40,1 mm). Materialul din care a fost
confecionat arcul a fost 50CrV4 conform DIN 17221.
232
AFASES - 2008 -
Figura 1.
Figura 2.
233
AFASES - 2008 -
Figura 3.
Capitolul III. Determinarea deformaiilor relative respectiv a tensiunilor
n seciunea de mijloc a arcului au fost determinate deformaiile relative utiliznd
traductori tensometrici rezistivi de tip 6/120RY11-HBM, traductori tip rozet 0 o/45 o/90o. Au
fost aplicai la interior i exterior cte un traductor conform schemei de amplasare din fig. 4
[3], [4].
Figura 4.
Aplicnd forele conform valorilor din tabelul nr. 1 pentru cele trei trepte de ncercare
au fost obinute rezultatele experimentale coninute n tabelele 3 i 4.
Tabelul 3.
Treapta 1
Treapta 2
Treapta 3
Rozeta interior
c [m/m] 1 [N/mm2]
F [kN]
a [m/m]
b [m/m]
33.790
14.090
35.607
13.754
37.141
-705.00
-419.59
-712.74
-394.16
-759.65
-3885.50
-1655.00
-4153.70
-1502.20
-4276.60
388.50
171.20
422.45
155.96
429.67
12.796
-358.50
-1399.60
148.88
234
2 [N/mm2]
[o]
550.38
210.50
598.82
188.41
609.76
-643.52
-283.60
-684.25
-258.51
-706.87
40.83
39.54
40.97
39.38
40.89
ech [N/mm2]
1034.99
429.46
1112.00
388.63
1141.26
178.23
-239.92
39.46
363.44
AFASES - 2008 -
Tabelul 4.
Treapta 1
Treapta 2
Treapta 3
F [kN]
33.790
14.090
35.607
13.754
37.141
a [m/m]
85.40
61.11
87.85
56.43
89.32
b [m/m]
-1908.40
-706.50
-2063.90
-626.10
-2127.30
12.796
48.40
-585.90
Rozeta exterior
c [m/m] 1 [N/mm2]
-15.50
318.34
45.85
136.17
-19.38
342.66
59.97
125.56
-21.77
352.50
59.66
117.31
2 [N/mm2]
-297.76
-104.70
-322.51
-91.31
-332.62
[o]
-44.26
-44.71
-44.27
44.93
-44.27
ech [N/mm2]
533.66
209.19
576.14
188.60
593.42
-85.51
44.75
176.36
Log a
m=3
12,901
12,751
12,601
12,451
12,301
12,151
12,000
11,951
11,701
11,551
11,401
11,251
11,101
11,001
m=5
17,036
16,786
16,536
16,286
16,036
15,786
15,536
15,286
15,036
14,786
14,536
14,256
14,036
13,386
235
Tensiunea de oboseal
[N/mm2]
117
104
93
83
74
66
59
52
46
41
37
33
29
26
AFASES - 2008 -
Avnd ciclurile de solicitare aplicate n fiecare caz a fost stabilit durata de viaa, n
numr de cicluri de solicitare conform excedentului dinamic i = a pentru fiecare caz.
Rezultatele obinute pentru durata de via sunt trecute in tabelul 6. La calculul de durat de
via s-a utilizat curba avnd clasa de concentrare 160 i tensiunea admisibil la oboseal 117
[N/mm2].
Tabel 6.
2
Nr. treapt de solicitare
a [N/mm ] Durat de via [nr. de cicluri]
interior
169,940
1622232
Treapta 1
exterior
91,085
10535603
interior
205,205
921374
Treapta 2
exterior
108,540
6226307
interior
215,765
792605
Treapta 3
exterior
117,545
4895910
a[N/mm2 ]
1000
500
Tensiunea admisibila
de oboseala
1
100
50
m=5
10
10
160
140
125
112
100
90
80
71
63
56
50
45
40
36
m=3
10
2
6
10
10
N(cicluri)
10
Figura 5.
Se constat c tensiunile cele mai mari apar la fibra interioar unde durata de via
estimat este aproape de cerinele impuse pentru fiecare treapt de solicitare. Exist diferene
mari ntre durata de via estimat pentru fibra interioar fa de cea exterioar datorit
diferenelor mari a deformaiilor relative respectiv a tensiunilor msurate experimental.
Estimarea duratei de via utiliznd curbele Wohler dau rezultate foarte apropiate de numrul
de cicluri prevzut pentru treptele de solicitare impuse ce s asigure o bun rezisten n
exploatare.
Capitolul V. Concluzii
Ca i concluzii finale, se impun urmtoarele:
1. Estimarea duratei de via prin utilizarea curbelor Wohler i a determinrilor n
regim static a deformaiilor relative pot conduce la concluzii elocvente n cercetarea
experimental. Rezultatele obinute pot fi suficient de concludente oferind elemente de
considerabil ncredere ntr-o reproiectare ulterioar.
236
AFASES - 2008 -
**
237
AFASES - 2008 -
D (t ) = x1 (t ) x 2 (t ) = v1 (t )dt v 2 (t )dt
0
(1)
t12
238
(2)
AFASES - 2008 -
Fig.1
Evoluia n timp a vitezelor v1, v2 i a spaiului parcurs x1 , x2 pentru vagonul de tamponare "
1 ", respectiv tamponat" 2 ", n cazurile de tamponare cu m1 = m2 i sensurile vitezelor
conform situaiilor A , B , C (A. vitezele vehiculelor sunt de sens contrar ; B. viteza unui
vehicul este nul v2 = 0 ; C. vitezele vehiculelor au acelai sens).
Acest fenomen se produce ntruct la timpul t*12 izolatorii de oc ai vehiculelor
prezint nc o deformaie ( contacie remanent ) care pe intervalul de timp ( t*12 - t2 ) se
anuleaz . Astfel , creterea spaiului dintre vehicule n fiecare moment al intervalului
de timp ( t*12 - t2 ) este compensat de revenirea contraciei izolatorilor de oc . La momentul
t2 , care marcheaz finalul procesului de tamponare, izolatorii de oc ( tampoane, amortizori
cupl central ) revin la poziia iniial , corespunztoare momentului t = t1 = 0 ( debutul
procesului de tamponare ) .
De la acest moment t2 , creterea distanei dintre vehicule se manifest n mod diferit
la vehiculele echipate cu tampoane fa de cele echipate cu cupl central i amortizor cupl
central . Astfel :
- la vehicule echipate cu tampoane i aparat de traciune apare o distan ntre talerele
tampoanelor . Dac vehiculele nu sunt cuplate distana se mrete n timp , dac sunt cuplate
239
AFASES - 2008 -
dup anularea jocurilor dintre vagoane apare o for de traciune la nivelul aparatului de
traciune care provoac o contracie a acestuia ;
- la vehiculele echipate cu cupl central i amortizor cupl central apare o distan la
nivelul contactului dintre vehicule care , anulnd jocurile existente la nivelul cuplei centrale ,
provoac o for de traciune n cupla central , respectiv o nou contracie a amortizorilor
vehiculelor .
4. Procesul de transformare a energiei poteniale inmagazinate n energie cinetic,
declanat la momentul t = t12 , se termin n momentul t = t*12, cnd vehiculele ating
vitezele v*1 , respectiv v*2 . La acest moment , izolatorii de oc prezint o contracie
remanent , deci practic i o energie potenial de deformaie nmagazinat , care ar putea fi
transformat n energie cinetic . De la momentul t = t*12 , energia cinetic a vehiculelor
rmne constant pe intervalul de timp ( t*12 - t2 ) , ntruct n fiecare moment al acestui
interval suma dintre energia potenial nmagazinat i cea disipat este egal cu energia
potenial disipat la momentul t2 .
n consecin, valoarea contraciei maxime " Dmax " i suprafaa " S ", care
reprezint valoarea contraciei maxime sunt :
*
t12
*
t12
t12
t12
(3)
(4)
aa nct :
v1 (t ) = v2
m2
(v2 (t ) v2 )
m1
(5)
a1 (t ) =
m2
a2 (t )
m1
(6)
240
AFASES - 2008 -
Fig. 2
241
AFASES - 2008 -
Evoluia n timp a acceleraiei " a2( t ) ", a vagonului tamponat" 2 ", determinat experimental
i a parametrilor derivai " v1( t ) ", " v2( t ) ," x1( t ) " i " x2( t ) ", pentru tamponarea A A .
Fig. 3
Evoluia n timp a acceleraiei " a 2( t ) ", a vagonului tamponat" 2 ", determinat experimental
i a parametrilor derivai " v1( t ) ", " v2( t ) "," x1( t ) " i " x2( t ) ", pentru tamponarea C C
.
Avnd diagrama experimental a2 = f( t ) , prin integrare s-a obinut evoluia n timp a
vitezei vagonului tamponat v2( t ) n procesul de tamponare n intervalul de timp ( 0 - t2 ) :
v2(t) = a2(t)dt
(7)
n mod similar s-a obinut spaiul parcurs de vagonul tamponat x2( t ) n procesul de
tamponare n intervalul de timp ( 0 - t2 ) :
x2(t) = v2(t)dt
(8)
242
AFASES - 2008 -
Prin integrare s-a determinat evoluia n timp a vitezei vagonului de tamponare v1( t ) :
m2
v1(t) = - a2(t)dt
m1
(9)
x1(t) = v1(t)dt
Astfel au fost trasate evoluiile n timp ale parametrilor de micare rezultai , v2( t ) , x2( t
,
v
)
1( t ) , x1( t ) , pentru ambele cazuri studiate n fig. 2 , respectiv fig. 3 , folosind un program
adecvat pe calculator .
Parametri determinai sunt trecui n tabelul nr. 1.
Tab.1. Parametri de micare i energetici rezultai n procesul de tamponare pentru cele dou
cazuri considerate
Nr.
Parametri
UM
AA
CC
crt.
01
m1
kg
80,000
80,000
02
m2
kg
90,000
80,000
03
v1
m/s
2,897
3,028
04
v2
m/s
0
0
05
v*1
m/s
0,411
0,953
06
v*2
m/s
2,210
2,075
07
v12
m/s
1,363
1,514
2
08
amax
m/s
40,000
22,150
Cap. III. Analiza rezultatelor experimentale
(11)
243
AFASES - 2008 -
x1(t12) - x2(t12) = D = D1 + D2
Fig.4 Evoluia n timp a forei medii " F med" [ MN ] transmise i a contraciei medii " D med"
[mm] la tampon categoria C
Cap. IV. Concluzii
244
AFASES - 2008 -
BIBLIOGRAFIE
1.- Bradley W. " Introducere n msurarea ocurilor i vibraiilor . ocuri i vibraii ".
Editura Tehnic , Bucureti - 1968 .
2.- Bendat J.S.; Piersol A.G. " Mesurement and ananlysis of random data ". John Wiley and
Sons Inc. New York , London , Sydney - 1966 .
3.- Biderman V. " Teoria mehaniceskih kolebani " . Vaia Skola , Moskva - 1980.
4.- Buzdugan Ghe. ; Blumenfeld " Tensometria electric rezistiv ". Editura Tehnic ,
Bucureti - 1966 .
5.- Buzdugan Ghe.; Mihilescu E.; Rade M. " Msurarea vibraiilor ". Editura Academiei ,
Bucureti - 1979 .
6.- Buzdugan Ghe. " Izolarea antivibratorie a mainilor ". Editura Academiei , Bucureti 1980 .
7.- Broch J.T. " Mechanical Vibation and Shock Mesurements ". Brel - Kjaer Naerum 1976 .
8.- Dally W.J.; Riley F.W. " Experimental stress analysis ". Mc. Graw - Hil Book Comp. New
York - 1966 .
9.- Keil S. " Zur Spanmungsanalyse mit Dehnungs-mestreifen bei elasticher Werk
Stoffbeanspruchung ". HBM VD 83002 , nr. 24 - 1983 .
10.- Mocanu D.R. ; .a. " Analiza experimental a tensiunilor ". Editura tehnic , Bucureti
- 1976 .
11.- Mocanu D.R. ; .a. " Determinarea experimental a eforturilor unitare ". Editura
Transporturilor i Telecomunicaiilor , bucureti - 1966 .
12.- Peters M.; Wilkening G. " Przisionsmessung in der Kraftmetechnik ". ( VD 81.001 )
HBM Messtehnische Briefe 16 - 1980 .
13.- Rubin S. " Metode de analiz a rezultatelor msurrii ocurilor . ocuri i vibraii " ( trad.
l. englez ) . Editura tehnic , Bucureti - 1968 .
Ion Copaci, prof. dr. ing. Universitatea Aurel Vlaicu Arad, ioncopaci@gmail.com
245
AFASES - 2008 -
246
AFASES - 2008 -
element de mcinare implic antrenarea acestuia de ctre elementele aflate n micare, ceea ce
duce la deteriorarea mai multor elemente de mcinare sau chiar la avarierea morii, creterea
duratei de indisponibilizare a morii i implicit creterea cheltuielilor cu reparaiile.
Pentru remedierea defeciunilor cauzate de avarierea elementelor de mcinare sunt
necesare cantiti mari de materiale i manoper, iar timpul de indisponibilizare crete foarte
mult, n unele cazuri putndu-se ajunge chiar pn la cteva luni.
Fig. 1 Rotorul morii MV 50, dup 1200 ore de fucionare a paleilor de mcinare
Deteriorarea elementelor de mcinare este cauzat de uzarea de abraziune, iar rolul de
element abraziv l are crbunele mcinat. Uzarea morilor ventilator se manifest n special la
paletele rotorului, dar i la blindajele statorice. Pe lng caracteristicile crbunelui mcinat,
prezint importan i urmtoarele caracterstici:
rezistena la uzare a metalului din care sunt executate piesele de mcinare ale morii,
construcia elementelor de mcinare,
comportarea pieselor de mcinare funcie de procesul de mcinare.
Pentru confecionarea elementelor de mcinare se utilizeaz de regul oelurile aliate cu
mangan, deoarece manganul confer duritate, o bun clibilitate i implicit o bun rezisten la
uzare. La aceste oeluri manganul este cuprins n domeniul 11,5 i 14,5 %. Oelul manganos poate
fi utilizat pentru confecionarea pieselor supuse uzrii atunci cnd n timpul funcionrii este
supus la presiuni i la ocuri. Oelurile manganoase au o rezisten la uzarea abraziv de
aproximativ 10 ori mai mare dect a oelurilor carbon care au aceeai rezisten mecanic.
Pentru morile ventilator paletele inferioare se pot executa din OT 45, iar paletele
exterioare se pot executa din oel manganos T20Mn14. n ultima perioad se ncearc utilizarea
unor materiale ale cror preuri de achiziie s fie ct mai reduse, dar care s confere o bun
comportare la uzare n timpul mcinrii (spre exemplu OL 52).
La alegera unui anumit material este necesar s se ia n considerare:
247
AFASES - 2008 -
248
AFASES - 2008 -
Ultimele dou metode sunt noi aplicate i nu au putut fi studiate pentru a se putea trage
concluzii n privina comportrii n funcionare i a evoluiei uzrii. n literatura de specialitate
sunt trecute urmtoarele mrimi caracteristice ale procesului de uzare [1]:
uzura specific, exprimat n [g/t], sau [g/kWh],
variaia debitului de crbune mcinat n funcie de creterea uzurii.
Aceste marimi au fost determinate n condiii de laborator. Pentru a putea determina
aceste mrimi n condiii normale de funcionare, trebuie s se msoare cantitatea de material
metalic pierdut prin uzare, cantitatea de crbune mcinat i cantitatea de energie electric ce se
consum n perioada de timp necesar mcinrii unei cantiti de crbune.
249
AFASES - 2008 -
figura 3. Pe parcursul funcionrii se vor face controale periodice pentru a se constata evoluia
uzurilor, apariia eventualelor canale prefereniale de uzare, dac apar diferene semnificative n
comportarea fiecrui palet n parte, sau s nu apar fenomene care ar putea pune n pericol
funcionarea morii. Se vor msura periodic vibraiile lagrelor morii. Dup un ciclu complet de
funcionare, sau cnd se constat c unul sau mai muli paleti rotorici prezint uzuri avansate,
nemaiputnd funciona, se va proceda la demontarea tuturor paleilor i cntrirea fiecrui palet n
parte. Pe ntreaga durat, se va msura energia electric consumat i cantitatea de crbune
mcinat de moar. n cazul n care este prevzut o instalaie de cntrire a crbunelui pentru
fiecare moar, se vor face estimri ale crbunelui consumat prin formule matematice care in cont
de viteza de deplasare a alimentatorului cu crbune i de nalimea stratului de crbune pe
alimentator. Aceste determinari se vor efectua la fiecare or, pentru a se putea obine valori ct
mai exacte.
n urma acestor determinri se va putea stabili uzura specific pentru paleii alei,
exprimat n [g/t] i [g/kWh], comparativ, n acelai timp, n aceleai condiii de funcionare,
utiliznd acelai crbune, dar materialele din care sunt confecioni acetia fiind diferite i n plus,
muchiile de atac (zonele cele mai expuse la uzare) fiind protejate, respectiv neprotejate. Se vor
putea trage concluzii despre comportarea celor dou materiale (OL 52 i T20Mn14) la uzare, dar
i despre compararea rezultatelor obinute n cazul muchiilor de atac ntrite prin cordoane de
sudur sau neprotejate. Se vor putea trasa cubele care stabilesc dependena dintre variaia
debitului de crbune mcinat de ctre moar i evoluia gradului de uzur, pentru fiecare dintre
materialele i metodele aplicate.
Dup obinerea curbelor i ale mrimilor caracteristice n condiii practice de funcionare
i compararea lor cu cele toretice, se vor putea trage concluzii privind noile metode care se vor
aplica pentru creterea duratei de via a paleilor rotorici i aplicarea unor materiale i soluii
tehnice avnd costuri ct mai reduse.
BIBLIOGRAFIE
1. PNOIU, N.; CAZACU, .C; MIHESCU, L.; TOTOLO, C.; EPURE A. Instalaii de
ardere a combustibililor solizi, Editura Tehnic, Bucureti 1985.
2. CUIBAR, D. Studii i cercetri privind posibilitatea creterii fiabilitii morilor pentru
mcinarea crbunilor folosite n CET-uri, prin nlocuirea materialelor pieselor puternic
solicitate la uzare; Tez de doctorat; Universitatea Transilvania din Braov; 2007.
Dumitru Cuibar, drd. ing., S.C. Energy Saving Agency S.A., Braov, Olte nr. 11, tel 0268422025, e-mail:
dcu@esa.com.ro
**
Ciprian Bncil, drd. ing., Agenia Pentru Protecia Mediului Braov, Politehnicii nr.3, tel 0268419013, e-mail:
director.executiv@apmbv.ro
250
AFASES - 2008 -
251
AFASES - 2008 -
252
AFASES - 2008 -
cei cu diametru mai mare, se ridic la suprafa i se elimin mpreun cu zgura, dar o bun
parte rmn n oel sub form de incluziuni. [1, 3]
ntruct n oel, pe lng fier i carbon, se gsesc i alte elemente, unele dintre ele
introduse la operaia de dezoxidare, pe lng oxizii de fier, n structura aliajului mai apar i
compui ai oxigenului cu aceste elemente, ca de exemplu: Al2O3, SiO2, Cr2O3, TiO2, ZrO2
sau incluziuni cu o compoziie chimic destul de complex ca (FeMn)O,
FeO SiO2 , Al 2 O3 FeO i altele.
Azotul apare n oel la elaborare din aerul aflat n mediul nconjurtor. Astfel, n
oelurile electrice, coninutul de azot este mai ridicat dect n oelurile Martin, iar n oelurile
elaborate n convertizor, mai ridicat dect n oelurile electrice (variaz ntre 0,01 0,03% N).
Deformarea plastic la rece favorizeaz separarea azotului din ferita suprasaturat i
apariia fenomenului de fragilizare.
Hidrogenul ajunge n oel n timpul elaborrii fie din adaosurile care conin hidrogen
i care sunt introduse n oel la elaborare (ferosiliciu, var), fie din cptueala cuptorului sau a
oalei de turnare. Pe de alt parte, hidrogenul mai poate fi absorbit de oel i n timpul
nclzirii aliajului solid, datorit fenomenului de difuzie. [5, 6]
Hidrogenul poate ptrunde n oel i n timpul decaprii, cnd el formeaz sulfuri de
decapare. Combaterea ptrunderii hidrogenului se poate realiza prin adugarea de inhibitori
n soluiile de decapare.
II. Oeluri aliate
II.1. Influena elementelor de aliere
Oelurile aliate sunt aliaje complexe ale fierului cu carbonul, care conin n mod voit
cel puin un al treilea element (de aliere) ntr-o cantitate suficient pentru a produce
modificri sensibile n proprietile fizicomecanice, chimice i tehnologice ale acestor aliaje.
Elementele de aliere folosite n mod curent sunt: Cr, Ni, Mn, Si, W, Mo, V, iar n
cazuri mai rare: Co, Ti, Al, Cu, Nb, Zr, B, N, Be etc.
Oelurile aliate pot fi aliaje ternare, cnd conin un singur element de aliere (Cr, Ni,
Mn, Si etc.), cuaternare, cnd conin dou elemente de aliere (CrNi, CrMn, CrW etc.) sau
polinare, cnd conin mai multe elemente de aliere (CrNiMo, CrMnV etc.).
Unele elemente pot exista n oel n stare liber (Pb, Ag, Cu) cnd sunt insolubile sau
foarte puin solubile, sau legate sub form de oxizi, sulfuri, silicai (elemente care au o
afinitate mai mare fa de oxigen i sulf dect fierul); importana i cantitatea admis a
acestor faze n oel sunt foarte mici. [4]
Majoritatea elementelor de aliere se dizolv n ferit (excepie C, N, O, B, S
elemente ndeprtate de fier n tabloul periodic al elementelor) i anume:
- cele aezate n sistemul periodic la dreapta fierului se dizolv n ferit i nu
formeaz carburi cu carbonul din oel (Ni, Co, Si, Cu etc.);
- cele aezate la stnga fierului se repartizeaz ntre ferit i fazele carburice, o parte
din coninut dizolvndu-se n ferit i alta n cementit sau formeaz carburi speciale.
Variaia calitativ i cantitativ a proprietilor oelurilor aliate este funcie de natura,
numrul i cantitatea elementelor de aliere adugate precum i de raportul lor cu principalii
componeni ai oelului: fierul i carbonul.
II.2. Oeluri aliate cu proprieti speciale
Aceste oeluri sunt aliate cu diverse elemente de aliere, cu diferite grade de aliere, n
scopul obinerii uneia sau a mai multor proprieti fizice, chimice i mecanice cu valori
253
AFASES - 2008 -
254
AFASES - 2008 -
255
AFASES - 2008 -
Oelurile slab aliate cu Mn, Si, V, Al pot fi utilizate pn la 80C, adugarea de 0,5
0,8% Ni, micornd mult tendina de rupere fragil. Oelurile slab aliate cu CrMo, pot fi
foloste pn la 120C, iar la adaos de peste 0,1% Ni pn la 185C.
Oelurile aliate cu Ni, cu 210%Ni, prezint o tenacitate substanial mrit ca urmare
a formrii unei structuri austenitice fine i a apariiei martensitice tenace FeNi. Pentru
temperatura de regim cuprins ntre 60C i 60C se recomand un coninut de 3,23,8% Ni,
ntre 100C i 150C 4,55,5% Ni, iar sub 200C se recomand 8,5 ,5%Ni. n funcie de
coninutul de nichel oelurile acestea au structur ferito perlitic, bainitic sau martensitic.
Oelurile aliate cu CrNi conin sub 0,1%C, 1626% Cr i 3,522%Ni i au o foarte
bun tenacitate pn la temperaturi apropiate de zero absolut (-273C). Rezistena este
sensibil mbuntit prin adaos mai ridicat de Ni i Mo sau N. Reziliena acestor oeluri este
foarte puin influenat de temperatur, caracteristicile de plasticitate fiind superioare
oelurilor aliate numai cu Ni; ele au de asemenea o bun sudabilitate.
Oelurile aliate CrMn sunt folosite n scopul nlocuirii celor CrNi; acestea conin
1322%Mn i 216%Cr, avnd structura austenitic.
Oelurile criogenice se prezint sub form de produse laminate sau turnate n piese i
sunt utilizate n construcii aeronautice, autovehicule i nave, agregate i maini frigorifice,
instalaii de lichefierea gazelor, recipieni de stocat i transportat, construcii aeronavale etc.
unde sunt supuse exploatri la temperaturi ntre -10 i -200C.
Oeluri cu proprieti magnetice deosebite
256
AFASES - 2008 -
Sunt oeluri care posed o rezisten electric ridicat, coeficient de temperatur mic
(variaia rezistivitii cu temperatura) i capacitatea de a rezista la oxidare la temperaturi
ridicate (refractaritate). Aceste oeluri au o structur omogen, monofazic, aliate cu Ni i Cr
i sunt folosite la confecionarea unor elemente de nclzire n cuptoarele i aparatele
electrice. Cu larg utilizare sunt mrcile:
- Fecral, cu maxim 0,12% C, 17% Cr, 5% Al, restul fier, care rezist pn la 1100C;
- Cromal, cu maxim 0,12% C, 26% Cr, 5% Al, restul fier, care rezist pn la 1200C;
- Oelul cu maxim 0,05% C, 25 27% Cr, 6 8% Ni i restul fier, rezistent pn la 1300C.
Aceste materiale au dezavantajul c prin nclziri ndelungate i repetate devin fragile.
Oeluri cu proprieti termice speciale
257
AFASES - 2008 -
Bibliografie
1. COLAN, H., .a., Studiul metalelor, Editura Didactic i Pedagogic, Bucureti, 1983.
2. DINESCU, I., Tehnologia materialelor, Materiale tehnologice, Editura Academiei Aviaiei
i Aprrii Antiaerine Henri Coand, Braov, 2000.
3. MITELEA, I., BUDU V., Studiul metalelor ndreptar tehnic, Editura Facla, Timioara,
1987.
4. PNZARU, S., DINESCU I., Materiale utilizate n producia special, Editura Academiei
Forelor Aeriene Henri Coand, Braov, 2002.
5. SMEAD, M., DINESCU I., Materiale convenionale utilizate n inginerie, The 11th
International Conference, Sibiu, 23-25 Noiembrie 2006.
6. SMEAD, M., MANIU Al., Materiale utilizate n tehnica de aviaie, Sesiunea de
comunicri tiinifice cu participare internaional, Braov, 11-12 Mai 2007.
Ion Dinescu, profesor univ. dr. ing., Academia Forelor Aeriene Henri Coand, Braov, Mihai Viteazu, nr.
160, tel. 0268/423421, e-mail: dinescuion@yahoo.com
**
Mihaela Smead, lector univ. drd. fiz., Academia Forelor Aeriene Henri Coand, Braov, Mihai Viteazu,
nr. 160, tel. 0268/423421, e-mail: mihaelasmeada@yahoo.com
258
AFASES - 2008 -
In the 1930s the cast iron was considered an alloy not suited to be deformed to get
pieces used in machines building.
The capacity of the cast iron to be deformed at high temperature is influenced by the
basic metallic body and the graphite shape, besides the carbon and others elements
concentration. In certain conditions of temperature and tensions, the deforming capacity of
the cast iron converges to the one of the steel. Thus, the mechanical properties are improved
and the resistance properties converge to the one of the quality carbon steel.
Key words: cast iron deformation, nodular graphite.
Pn la nceputul anilor 1930 i mai mult nc dup aceea, fonta era considerat un
aliaj Fe-C-Si care nu se preteaz a fi deformat la cald n vederea obinerii unor piese utilizate
n construcia de maini.
n conformitate cu prerile unor cercettori, referitor la capacitatea de deformare la
cald, nu exist o limit bine conturat ntre oel i font, ci mai degrab exist o trecere
continu, n sensul c n anumite condiii de temperatur i stare de tensiuni, posibilitile de
deformare a fontei se apropie de cele ale oelului. Afirmaia c, capacitatea de deformare
scade cu creterea coninutului de carbon nu este ntrutotul valabil, deoarece la fonte aceast
capacitate este hotrtor influenat de masa metalic de baz i de forma grafitului.
Cu toate c n ultimul timp s-au cules date despre capacitatea de deformare a
fontelor, s-a constatat c aceasta este mult mai mic dect a oelului. Se poate pune deci cu
justificare ntrebarea de ce s-a ncercat cercetarea capacitii de deformare a unui material,
dac el se preteaz mai puin la aceast tehnologie.
Cercetrile sistematice privitor la deformarea la cald a fontelor au nceput n anul
1930 cnd s-au fcut ncercri privind deformarea prin laminare la cald a fontelor. Pn atunci
s-au semnalat capacitatea de deformare a fontei n cercetri rzlee. PIWOWARSKY,
SCHLEGEL, WITTMOSER i alii [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11] au ajuns la concluzia c la
259
AFASES - 2008 -
260
AFASES - 2008 -
261
AFASES - 2008 -
Din analiza lucrrilor enumerate mai sus rezult urmtoarele aspecte de parametrii
metalurgici, care controleaz procesul de deformare la cald:
- gradele limit de deformare admisibile n funcie de temperatura de deformare
- influena vitezei de deformare i a temperaturii de deformare asupra rezistenei la
deformare
- procesele de recristalizare static i dinamic
- modificrile structurale i ale proprietilor mecanice n funcie de gradul de
deformare.
n Romnia primele cercetri semnalate n legtur cu deformarea la cald a fontelor
cu grafit nodular, n cuiburi (maleabil) i a fontei albe au fost efectuate de I. DRGAN [12]
de la Institutul Politehnic Cluj n ideea c matriarea unor semifabricate turnate din aceste
materiale ar putea cpta o extindere mai mare i n industria romneasc.
Matriarea multor piese ridic probleme dificile privitor la unele faze pregtitoare ale
matririi n final, iar pe de alt parte de gradul de pregtire a semifabricatului este strns
legat de posibilitatea obinerii n bune condiiuni a piesei n cavitatea final de matriare.
Pierderile de material n bavur sunt n general mari la piesele care au treceri brute
de la o seciune la alta sau la piesele n form de inel i cadru unde apar timpane de
dimensiuni mari. S-a constatat ns c se pot mbunti substanial proprietile mecanice ale
unor piese turnate i apoi matriele.
Prezumia modificrii unor caracteristici mecanice n sensul creterii lor, datorit
deformrii la cald se datoreaz urmtorilor factori:
a. micorarea mrimii greutilor
b. micorarea microsegregaiilor
c. atenuarea sau nlturarea complet a retasurilor sau a microporozitior
d. apariia unei structuri orientate datorit alungirii grunilor, a incluziunilor i a
dendritelor n sensul deformrii printr-o micorare a distanelor inter i
intradendritice.
Un rol important din factorii enumerai mai sus l constituie la piesele turnate,
microporozitile. Aceste microporoziti i microretasuri de mrimi diferite, care sunt
localizate mai frecvent n spaiile interdendritice se datoreaz bulelor de gaze prezente n
metalul lichid ca urmare a urmtorilor factori:
a. coninut ridicat de hidrogen i azot n baie
b. o defectuoas dezoxidare a bii
c. descompunerea vaporilor de ap din forma de turnare.
Este evident c prin reducerea la minim a seciunii lor transversale, sau a nchiderii
lor complete se mrete seciunea sntoas a piesei, deci cresc caracteristicile mecanice ale
ei. Acest aspect se poate urmri dac considerm modul cum se deformeaz microporozitile
sau golurile care apar n urma turnrii n condiiile unei anumite stri de tensiune, de
comprimare pe toate feele semifabricatului, fig.4 [12] cnd b0=b1 nlimea lor se micoreaz
fr a se mri limea.
262
AFASES - 2008 -
263
AFASES - 2008 -
264
AFASES - 2008 -
265
AFASES - 2008 -
Mihai Ciprian Fodor, Drd. Inginer, Universitatea Tehnica Cluj-Napoca, B-dul Muncii 103-105 Cluj-Napoca,
tel: 0740.027486, e-mail: mihai.fodor@yahoo.com
Viorel Dan, Conf. Dr. Ing., Universitatea Tehnica Cluj-Naapoca, , B-dul Muncii 103-105 Cluj-Napoca, tel:
0264/401.624, e-mail: vdan30@yahoo.com, viorel.dan@sim.utcluj.ro
George Drago, Drd. Inginer, Universitatea Tehnica Cluj-Napoca, B-dul Muncii 103-105 Cluj-Napoca, tel:
0768.972248, e-mail: ronin_katana1979@yahoo.com
Iosif Andrei Fodor, Drd. Inginer Universitatea Tehnica Cluj-Napoca, B-dul Muncii 103-105 Cluj-Napoca,
tel: 0745.307262, e-mail: fodordea@yahoo.com
266
AFASES - 2008 -
Bazele acestei metode constau n modelarea structurii ntr-un numr discret de mase
(elemente) rigide cu interconectri ntre ele.
Considernd modelul reprezentat de un sistem elastic cu N grade de libertate, se
analizeaz echilibrul unui element de mas mn+1.
Ecuaiile micrii oscilatorii, pe un element, sunt[6]:
m n+1
n+1
Kn
&x&n = 2 xn
Xn
Xn+1
Fig. 1Echilibrul unui element de mas mn+1
De aceea, urmtoarele dou ecuaii pot fi scrise sub forma:
267
(1)
AFASES - 2008 -
xn+1 = xn
Fn
1
= xn Fn
kn
kn
(2 a)
(2 b)
xn
x n+1
kn
kn
(3)
cu condiiile iniiale:
F1 = m1 2 , xi =
(4)
Matricea (3) se numete matrice de transfer. Aceste ecuaii de baz pot fi utilizate pentru
fiecare element ncepnd cu n=1, prin asumarea valorii pulsaiei 2, pn la ultimul element
al structurii.
Valoarea corect a lui 2 va determina condiiile la limit corecte la ultimul element al
structurii. n general, orice valoare aleas arbitrar pentru pulsaie va fi incorect i trebuie
utilizate iteraii succesive pentru a obine valoarea corect.
Iteraiile, n acest caz, se fac prin puncte sau mai bine zis prin urmrirea valorilor ctre
stnga (conform desenului), pn la atingerea condiiilor limit prin comparare cu valoarea
arbitrar a lui 2.
XN+1
268
AFASES - 2008
(n+1)
(n)
{ }s = [Rn ]{ }ns
(5)
(6)
n relaia (6), condiiile limit sunt ncorporate n { }sB i { }Vs aa cum a fost precizat
anterior.
Se consider modificarea unei matrice tipice de transfer aa cum a fost definit, aplicat
unui element de pal.
Fie elementul mprit n dou pri distincte:
- o mas concentrat asociat mrimilor: masa - mn+1 i momentul de inerie - In+1;
- un element elastic de mas neglijabil, ataat spre dreapta masei concentrate, notat { }dn+1
ca n fig. 5.
Cele dou pri fiind asociate una alteia, n punctul de referin al segmentului de-a lungul
palei, (conform desenelor iterarea se face de la dreapta la stnga), notaia corespunztoare
figurii fiind aa cum se arat n desen.
m n+1
{}
s
n+1
{}
{}
n+1
269
AFASES - 2008 n
Fn = mi xi 2
(7)
i =1
(8)
w sn +1 = w dn +1
sn +1 = dn +1
0
0 0 w
1
w
0
1
0 0
=
m n +1 2
0
1 0 T
T
M
I n+1 2 0 1 M n +1
0
n +1
14444
42
44444
3
matricea de transfer[Pn +1 ]
mn+12 wn+1
(9)
Tn+1
In+1 2 n+1
Mn+1
Nn
Nn+1
wn+1
Mn+1
s
T n+1
{ }dn+1 i { }ns
din figura 7.
T sn
Nn
M n+1
s
Mn
Nn
sn
d
T n+1
d
d
wn+1
n+1
wns
Ln
270
AFASES - 2008 -
(10)
= x xT i = w wT
(11)
EI NT = TT ( L ) + M T
(12)
Ecuaia devine,
unde, variaz de la 0 la L, i unde soluia ( ) trebuie s ndeplineasc condiiile la
limit:
(0) = w0 wT i (0) = 0
(13)
L
N
L
N
ch
L
T
T
1 MT
1 TT L
T = 0
+
+
[LthL]
1
chL N T chL N T
(14)
F
, iar pentru L 0 (cazul fr for axial).
EI
Aceste expresii se simplific i devin cele mai elementare expresii pentru deformaiile
barei (palei):
1
1
wT = w0 + 0 L + TT L3
+ M T L2
3EI
2 EI
(15)
2 1
2 1
T = 0 + TT L
+ MT L
EI
2 EI
Unde, =
(16 a)
T0 = TT
(16 b)
Soluiile pot fi aranjate i combinate ntr-un format matriceal, rezultnd astfel ecuaia
matriceal, i implicit matricea de transfer:
271
0
F
L
L
f
g
M
n
nf
n
n
n
4444n 4444244444444
n +1
1
3 n
matricea de transfer [Qn ]
d
(17)
unde,
shL
fn =
; g n = chLn
L n
(18)
hn = Ln EI n ; n = Fn EI n
0
[Rn ] =
2
m n +1
gn
m n +1 2 Ln f n
[I
n +1
g n Fn Ln f n
( f n + 1)
(g n + 1)
Fn
2 Ln
( f n + 1)
1 + m n +1
Fn
2
I n +1 ( g n + 1)
+ Ln f n
Fn
(g n + 1)
Ln
Fn
Ln f n
Fn
Fn
f n hn
m n +1 ( g n + 1)
Fn
[ I
n +1
f n hn + g n
(19)
Pentru c fiecare matrice de transfer este dependent de pulsaie, calculul valorilor proprii
se face cu relaia:
{ }B(butuc) = [1
Rn ][Rn1 ]K[R2 ][R1 ]{ }V (vrf )
4442444
3
(20)
R 2
[ ( )]
{ }V
w1
= 1
0
0
la baza palei
{ }N
0
0
0
TN
TN
0
M N
123
123
rotor articulat rotor nearticulat
(21)
Dac aceti vectori sunt introdui n ecuaia de transfer i dac termenii nenuli sunt
reinui, se obin urmtoarele ecuaii:
272
AFASES - 2008 -
0 = t11w1 + t121
( N ) = t21w1 + t221
TN = t31w1 + t321
(22)
(M N ) = t14 w1 + t421
unde unul dintre ei, (dar nu amndoi termenii din parantez), poate fi ales iniial zero.
De aceea, pentru fiecare tip de rotor, articulat sau nearticulat matricea explicitat va
conine dou ecuaii scalare n funcie de w1 i 1.
Dac primele 3 ecuaii sunt egale cu zero rezult:
( )
( )
t11 2
2
t i1
( )w = 0
( ) 0
t12 2
ti2 2
(23)
( )
2 =
( )
( )
t11 2
ti 2 2
( )
( )
t12 2
ti 2 2
(24)
( )
2 = 0
(25)
Deoarece diferitele (n) matrici de transfer au fost deja evaluate pornind de la butuc (n
procesul obinerii elementelor determinantului caracteristic), acestea sunt cele mai potrivite
pentru a descrie rspunsul fiecrui element n comparaie cu cel al vrfului palei.
Paii necesari pentru a gsi modurile proprii de vibraii pentru fiecare rdcin (soluie)
sunt:
1. Stabilirea condiiilor pentru normalizarea vectorilor proprii; de exemplu stabilirea
valorilor pentru w1 sau 1 egale cu valori de referin cunoscute (de obicei 1,0);
2. Rezolvarea pentru alte variabile necunoscute la vrful palei folosind relaia
t121 = t11w 1
(26)
(De notat c fora tietoare i momentul ncovoietor sunt nule la vrful palei).
Acestea completeaz elementele de definiie ale vectorului de stare la vrful palei.
Se folosete matricea de transfer individual pentru toate elementele interioare care
definesc vectorul de stare, pentru toate elementele de studiu alese de-a lungul
anvergurii(razei).
Cunoaterea vectorilor de stare pentru cteva elemente constituie o soluie pentru analiza
modurilor proprii de vibraii n termeni de deplasri i pante.
Mult mai important, este faptul c vectorii de stare pot genera informaii asupra forfecrii
i ncovoierii n elementul de pal din lungul anvergurii.
273
[1] Alecu, A., Boiangiu, M., Dinamica palei de elicopter, Editura BREN, Bucureti, 1999
[2] Bielawa, R., Rotary Wing Structural Dynamics and Aeroelasticity, American Institute
of Aeronautics and Astronautics, Washington, 1992
[3] Johnson, W., Helicopter Theory, Princeton Univ. Press, Princeton, NJ, 1980
[4] Petre, A., Proiectarea structurilor de aero-astronave, Editura Academiei Romne,
Bucureti, 1999
[5] Postelnicu, A., Deliu, Gh., Udroiu, R., Elicoptere, Editura Universitii Transilvania,
Braov, 1999
[6] Voinea, R., Stroe, I.V., Sisteme dinamice, Universitatea Politehnica, Bucureti, 1993
Lavinia GAVRIL, conf. univ. dr. ing., Academia Forelor Aeriene Henri Coand din Braov, str. Mihai
Viteazu, nr.160, telefon 0268-423.421, fax 0268-422.004, e-mail lavinia.gavrila@afahc.ro
274
AFASES - 2008 -
275
AFASES - 2008 -
276
AFASES - 2008 -
277
AFASES - 2008 -
Tabelul 1 : Tabel cu date tehnice ale gurilor de foc utilizate la probele experimentale
Guri de foc eava balistic eava balistic
M.T.B.
Date tehnice
marca Barinov
marca Levaov
Calibrul
7,62 mm
12,7 mm
14,5 mm
Lungimea evii propriu zise
755 mm
1.005 mm
1.276 mm
Lungimea total
900 mm
1.200 mm
1.480 mm
Greutatea total
6,5 kg
10,5 kg
52 kg
- foc n serii scurte ;
Modul de executare al focului
foc cu foc
foc cu foc
- foc n serii lungi ;
- foc continuu
278
AFASES - 2008 -
Rezultatele obinute n urma analizei chimice a plcii de blindaj sunt date n tabelul
urmtor, precum i configuraia elementelor din compoziia chimic a blindajelor sunt
materializate n tabelul 3 i figura 5, dup cum urmeaz :
Tabelul 3 : Rezultatele obinute n urma analizei chimice a plcii de blindaj (%)
Mn
1,37
Si
1,31
P
0,011
S
0,005
Co
0,01
Cr
0,12
Cu
0,02
Mo
0,19
Ni
0,33
W
0,02
Configuraia elementelor
din compoziia chimic a blindajului
1,37
1,31
0,33
0,19
0,12
0,02
Mn
Si
Ni
Mo
Cr
0,02
Cu
Co
279
AFASES - 2008 -
Capitolul III
Rezultatele obinute n urma probelor experimentale
n urma impactului dintre gloanele de calibru mic i placa de blindaj utilizate s-a
produs perforarea total. Cu alte cuvinte, gloanele au strpuns placa de blindaj.
Rezultatele experimentale obinute sunt prezentate n tabelul urmtor :
Tabelul 4 : Rezultatele experimentale n urma efecturii probelor
Muniia
utilizat
inta
Obtained
results
Meniuni
Gradul de
protecie al
fiecrei
componente
a vestei
antiglon
Gradul de
protecie al
vestei
antiglon
Glon
Glon perforant incendiar
perforant-incendiar
cal. 7,62 x 54 mm
cal. 12,7 x 108 mm
Lovitura 1
Lovitura 2
Vest antiglon
Vest antiglon
Vest antiglon
Plac
Pachet
Plac
Pachet
Plac
Pachet
ceramic
balistic
ceramic balistic ceramic balistic
Perforare
Perforare
Amprent Neperforare Amprent
Amprent
parial
parial
11
21
esturi
esturi
perforate
perforate
50%
100 %
50%
75%
66,7 %
58,4%
50%
36,4 %
43,2%
Glon
perforant-incendiar
cal. 14,5 x 114 mm
Vest antiglon
Plac
Pachet
ceramic
balistic
Perforare Perforare
total
total
33
esturi
perforate
0%
0%
0%
67%
280
AFASES - 2008 -
Gloanel e utilizate
la tragerile experimentale
(diametrul, exprimat n mm)
14.5
79.18
14
12
80
12.7
70
61.83
60
10
50
8
7.62
40
30
20
Gradul de protecie
al vestei antiglon
(exprimat n procente)
16
10
0
1
281
AFASES - 2008 -
Capitolul IV
Concluzii finale
Chiar dac blindajul nu ofer protecie maxim lupttorului aflat n vehiculul de lupt
blindat, acest reper prezint o importan destul de mare, deoarece, n urma impactului se
micoreaz, att masa glonului, ct i puterea de perforare a acestuia.
Se remarc importana echipamentului de protecie balistic, vest antiglon, datorit
faptului c, n cazurile tragerilor cu gloanelor perforant incendiare cal. 7,62 x 54 mm i
12,7 x 108 mm, acesta ofer protecie purttorului.
ns n cazul tragerii cu glon perforant incendiar cal. 14,5 x 114 mm, vesta
antiglon nu ofer nici un fel de protecie, avnd n vedere faptul c s-a produs perforarea
total. Cu alte cuvinte glonul perforant incendiar cal. 14,5 x 114 mm a strpuns
echipamentul de protecie balistic.
Ionescu Ctlin Eugen, cercettor tiinific, Poliia de Frontier Clrai Oltenia, Dr. Lucian Popescu,
nr. 2, tel. 0242 515798, e-mail : catalineugenionescu@yahoo.com
282
AFASES - 2008 -
Capitolul I
Prezentarea general a Metodei Elementelor Finite
Metoda Elementelor Finite este o tehnic de analiz numeric care permite obinerea
unor soluii aproximative pentru probleme de domeniu ingineresc.
n cadrul acesteia, domeniul soluiilor se mparte n subdomenii diferite ca dimensiuni
i forme, care sunt conectate n noduri (puncte) denumite elemente.
Pentru a reprezenta variabilele de cmp ale unei structuri (tensiune, deformaie,
deplasare, temperatur, presiune etc.), n interiorul elementelor se definesc funcii de
aproximare (interpolare) care exprim aceast variaie n raport cu valorile nodale ale
variabilelor.
Metoda Elementelor Finite asigur tehnici de determinare a valorilor nodale ale
variabilelor, care s asigure o apropiere mare n raport cu soluia real.
ntr-o problem cu orice dimensiune a mediului continuu, variabila de cmp are o
infinitate de valori, rezultnd astfel un numr infinit de necunoscute. Procedurile de
discretizare n elemente finite a domeniului reduc problema la una cu un numr finit de
necunoscute, datorit faptului c n interiorul fiecrui element variabila de cmp se exprim
prin funcii de interpolare, n raport cu valorile sale nodale necunoscute, dar n numr finit. n
general, nodurile sunt situate pe frontierele elementelor, reprezentnd punctele de conexiune
dintre acestea, dar pot exista i noduri interioare elementelor, n special n cazul unor funcii
de interpolare mai complicate. Cu aceast reprezentare, valorile nodale ale variabilei de cmp
283
AFASES - 2008 -
284
AFASES - 2008 -
fiecrui element. Astfel, ecuaiile generale ale structurii vor fi de acelai tip, dar vor fi mai
multe i vor conine mai muli termeni ca rezultat direct al numrului de elemente
participante. n esen metoda de asamblare se bazeaz pe ipoteza natural c valoarea global
a variabilei de cmp ntr-un nod este rezultatul conlucrrii elementelor care sunt conectate n
nodul respectiv.
e) Scrierea condiiilor la limit
Problemele guvernate de acelai tip de ecuaii cu derivate pariale sunt individualizate
de condiii la limit pentru fiecare caz n parte. Aceast etap const n scrierea valorilor
variabilei de cmp sau a derivatelor sale pe frontiera domeniului.
f) Rezolvarea sistemului de ecuaii
Rezultatul asamblrii ecuaiilor nodale este un sistem liniar de ecuaii care are ca
necunoscute valorile nodale ale variabilei de cmp. Indiferent de tipul de problem de
rezolvat, aceast etap const n rezolvarea sistemului liniar de ecuaii algebrice, folosind una
din metodele cunoscute.
g) Efectuarea de calcule suplimentare
De regul, dup determinarea funciei de cmp necunoscute, sunt necesare i alte
calcule care au ca obiectiv determinarea altor parametri importani care descriu fenomenul
respectiv.
Capitolul II
Materialele utilizate la modelarea procesului de impact
n vederea realizrii modelrii procesului de impact care face obiectul prezentului
referat, au fost necesare urmtoarele materiale :
muniia ;
blindajul
2.1. Muniia
Muniia utilizat la modelarea procesului de impact este reprezentat de urmtorul tip
de glon :
glonul perforant incendiar calibru 14,5 x 114 mm (fig. 1).
285
AFASES - 2008 -
Configuraia elementelor
din compoziia chimic a blindajului
1,37
1,31
0,33
0,19
0,12
0,02
Mn
Si
Ni
Mo
Cr
0,02
Cu
Co
286
AFASES - 2008 -
Capitolul III
Modelarea procesului de impact
3.1. Impactul glon plac de blindaj
287
AFASES - 2008 -
Fig. 4 : Etapele importante ale variaiei tensiunii efective Von Mises din procesul de impact
288
AFASES - 2008 -
Capitolul IV
Concluzii
n figurile anterioare se pot vedea cele mai importante momente ale impactului dintre
glon perforant incendiar calibru 14,5 x 114 mm i plac de blindaj de grosime 6 mm.
n cadrul ambelor figuri anterioare, la t = 0 s este exprimat starea iniial, n care se
prezint glonul i blindajul.
La t = 0,35 s se prezint momentul impactului, unde ambele elemente care-l
genereaz ncep sse deformeze.
Pentru t = 2,4 s, t = 3,2 s i t = 4 s se prezint, pe rnd, etapele perforrii plcii de
blindaj de grosime 6 mm de ctre glonul perforant incendiar calibru 12,7 x 108 mm.
Imaginile corespunztoare lui t = 5,5 s reprezint perforarea total a procesului de
impact ce face obiectul acestui referat.
Ionescu Ctlin Eugen, cercettor tiinific, Poliia de Frontier Clrai Oltenia, Dr. Lucian Popescu,
nr. 2, tel. 0242 515798, e-mail : catalineugenionescu@yahoo.com
289
AFASES - 2008 -
290
AFASES - 2008 -
avioanele din dotarea lui, rezolvnd problema prin introducerea n structura avionului a unui
rezervor suplimentar de oxigen, cu care supradoza intrarea n turboreactorul avionului nainte de
lansarea rachetei. Ca urmare a acestor incidente, s-a pus problema mbuntirii combustiei n
motorul rachet, astfel nct, combustia produselor de ardere evacuate s nu mai necesite un
consum semnificativ de oxigen, care s afecteze motorul turboreactor al avionului purttor.
II. ADITIVI FOLOSII PENTRU REDUCEREA EMISIEI DE FLACR LA
PROPULSANII SOLIZI PENTRU MOTOARE RACHET
II.1 Aspecte teoretice
Combustia unei pulberi solide din motoarele rachet este n sine un proces complex,
dependent de muli factori. Modelrile acestei combustii au constituit i constituie subiect de
studiu pentru specialiti, elabornd diverse metode de calcul a caracteristicilor termochimice i
termodinamice. Prin modelarea combustiei se obin unii parametri termoenergetici absolut
necesari n etapa de studiu pentru un combustibil nou, cum ar fi: produii de ardere, constanta
gazelor, masa molar a gazelor, temperatura gazelor etc. Dar cel mai mare neajuns al a cestei
modelri const n faptul c procesele de ardere nu sunt legate de factorul timp, datorit
complexitii acestor procese. Produsele iniiale rezultate prin ardere nu sunt stabile, ele suferind
recombinri n camera de ardere, n ajutaj i dincolo de seciunea de ieire a ajutajului unde se
transform n produse de ardere gazoase sau solide stabile. n cazul prezentat, pentru
mbuntirea combustiei era necesar s se intervin asupra reetei prin adugarea unui inhibitor
de flacr (aditiv de suprimare a flcrii). n procesul de ardere, acest inhibitor s fie vaporizat i
descompus n radicali liberi, care ulterior, dincolo de seciunea de ieire a ajutajului s
reacioneze cu produii pariali de combustie (CO,C,OH), ducnd la ntreruperea lanurilor
cinetice de postcombustie a produilor oxidai parial [4],[5].
II.2 Aspecte practice
Pentru realizarea combustibilului ce echipeaz motorul proiectilului reactiv S24 pentru
avionul MIG 29, aditivul folosit a fost un complex ce nglobeaz metale mpreun cu un oxid.
Compusul folosit este hexanitrocobaltiatul de potasiu, avnd formula chimic K3Co(NO2)6. El se
prezint sub form de particule micronizate, de culoare galben intens, bine cristalizat, sub form
cristalografic aparinnd sistemului cubic polimorf.
Distribuia granulometric a lui este:
cristalite sub 2
1,2 ppg
cristalite ntre 2 i 10
60,8 ppg
cristalite peste 10
38,0 ppg
Coninutul de cobalt din compus este de 9,5 % ppg.
Hexanitrocobaltiatul de potasiu se prepar prin sintez pe instalaii de sintez specifice
fabricaiei catalizatorilor solizi.
Din punct de vedere al materiilor prime, exist mai multe posibiliti de a sintetiza
hexanitrocobaltiatul de potasiu. Compusul strategic al acestui aditiv este cobaltul, pentru care
291
AFASES - 2008 -
materia prim poate fi metalul ca atare sau nglobat sub form de azotat. La rndul su, azotatul
de cobalt ( Co(NO3)2 ) poate fi sub form de cristale sau sub form de soluie.
n etapa de cercetare care a avut ca obiectiv stabilirea tehnologiei optime din punct de
vedere pre i timp pentru sinteza hexanitrocobaltiatului de potasiu s-a acionat pe trei variante.
La prima variant s-a utilizat cobaltul ca atare. Aceast variant s-a dovedit ca fiind
nerentabil att din punct de vedere al consumurilor ct i al costurilor. Pentru sinteza a 100 kg
de hexanitrocobaltiat de potasiu erau necesare 12 kg de cobalt de o puritate medie.
La a doua variant de sintez s-a utilizat azotatul de cobalt cu ase molecule de ap,
Co(NO3)2 6H2O.
La a treia variant s-a utilizat azotat de cobalt sub form lichid, a crei caracteristici de
baz sunt prezentate mai jos:
aspect
soluie limpede
culoare
rou-viiniu
3540 %
coninut Co(NO3)2
densitate [g/cm3]
1,31,4
pH
5,56,5
Din analiza efectuat pe baza criteriilor specificate mai sus a rezultat c varianta trei este
cea mai bun.
Oxidul folosit este oxidul de magneziu, MgO, sub form de particule cu suprafa
rugoas, dopat cu atomi de Fe3+ i Mn2+.
Compoziia aditivului este:
hexanitrocobaltiat de potasiu
oxid de magneziu
impuriti de metale alcaline
i alcalinopmntoase, altele dect K i Mg
Reeta pulberii este:
nitroceluloz
nitroglicerin
aditiv de suprimare a flcrii
dinitrotoluen
dibutilftalat
centralit II
ulei de vaselin
grafit
difenilamin
umiditate
72 %
26 %
2%
57 %
24 %
8%
4%
3%
2,5 %
0,5 %
0,3 %
0,2 %
max 0,5 %
292
AFASES - 2008 -
s-au efectuat analize de laborator n scopul verificrii compoziiei din punct de vedere procentual
al constituenilor. n urma analizei s-a constatat c aditivul se regsete n reet n proporie de
numai 25 % din cantitatea introdus n malaxor. Cu toate acestea s-a continuat procesul de
fabricaie a pulberii prin extrudarea pastei n element sub form de macaroan, la dimensiunile
cerute de documentaie. Din elemenii extrudai s-au prelevat probe pentru determinarea
repartiiei granulometrice a aditivului. n urma analizei rezultatelor s-a constatat existena unei
repartiii neuniforme a aditivului. Explicaia acestor rezultate a constat n faptul c procedeul de
introducere a aditivului la sfrit nu este adecvat. Prezena nitroglicerinei n compoziia apoas i
confer acesteia proprietile unei suspensii uleioase, n care nu se poate realiza o repartiie
uniform a unui compus solid sub form micronizat. ntr-o suspensie apoas i uleioas,
particulele micronizate ale aditivului nu au gsit un suport de care s se agae i prin urmare el
s-a dispersat aleatoriu. Mai mult dect att, 75 % din coninutul de aditiv a fost antrenat de ap n
procesul de stoarcere. Chiar i n aceste condiii s-au efectuat teste statice pe prin trageri pe banc.
Rezultatele obinute au fost considerate ca fiind promitoare, n sensul c parametrii balistici sau ncadrat n totalitate n cerinele impuse, iar din punct de vedere al aspectului flcrii s-a
constatat o reducere a ei, comparativ cu testele n care combustibilul nu coninea nici un fel de
aditiv, sau aditivi sub form de oxizi
n a doua variant, aditivul a fost introdus n malaxor, imediat dup introducere
nitrocelulozei, dar nainte de introducerea nitroglicerinei. Justificarea acestei succesiuni de
introducere a componentelor pulberii a fost susinut de structura fizic a nitrocelulozei.
Nitroceluloza are o structur fibroas. Dac n suspensia de ap, n care de fapt se i produce ca
materie prim se introduce un compus solid micronizat, n procesul de amestecare particulele
solide se aga de fibre. Mai mult dect att, durata procesului de amestecare poate determina i
gradul de uniformizare n masa nitrocelulozic al compusului solid micronizat.
Pe baza argumentelor prezentate mai sus s-a modificat tehnologia i s-a fabricat o nou
arj de pulbere. Ca i n prima variant, din pasta rezultat n urma stoarcerii s-au prelevat probe
pentru analiza cantitativ a constituenilor prezeni i compararea rezultatelor din punct de vedere
procentual cu ceea ce s-a introdus n malaxor. Referindu-ne la aditiv, acesta se regsea n masa de
past n procent de numai 50 % din ct a fost introdus. La fel ca i n prima variant, pasta a fost
extrudat, presat i prelucrat sub form de calupi. Pe calupii de combustibil rezultai s-au
efectuat teste statice pe banc. Dac la testele prezentate anterior, rezultatele au fost calificate ca
fiind promitoare, de data aceasta rezultatele au confirmat faptul c aceasta este direcia cea
bun. Analiza comparativ a rezultatelor va fi fcut i comentat mai trziu.
Pe baza rezultatelor prezentate mai sus, era clar c direcia de cercetare trebuia orientat
pe mbuntirea tehnologiei de fabricaie a pulberii. Cel mai important aspect nc neelucidat era
acela al dispariiei unei cantiti importante din masa de aditiv pe fluxul de fabricaie, de data
aceasta 50 %. Aceast diferen semnificativ trebuia s se regseasc undeva. Cel mai probabil,
diferena de aditiv se putea regsi n emulsia de ap, dar din pcate, aceast ap rezultat n urma
stoarcerii este evacuat mpreun cu celelalte ape reziduale i nu este colectat.
La aceast etap de cercetare s-a stabilit modificarea instalaiei de pulbere, n sensul
adugrii unui bazin de colectare a apelor reziduale. S-a reluat cercetarea pe o nou arj de
combustibil. Rezultatele obinute au fost foarte asemntoare cu cele prezentate la etapa
anterioar. De data aceasta ns, s-au efectuat analize pe apa colectat. Cum era i de ateptat, o
parte din aditivul lips era n aceast ap. i totui numai o parte. n aceast faz de cercetare s-a
ivit o nou ipotez, i anume aceea c, este posibil ca n anumite etape ale fluxului tehnologic de
fabricare a pulberii, aditivul s sufere o modificare, cel mai probabil s se dizolve parial.
Dizolvarea aditivului meninut un anumit timp n ap se poate produce numai dac n compoziia
293
AFASES - 2008 -
sa este prezent i sodiul, ntr-o anumit cantitate. Prezena sodiului n structura aditivului, chiar i
ca element rezidual este inerent, datorit existenei lui materiile prime folosite. Prin rafinarea
procesului de sintez, coninutul de sodiu poate fi diminuat, dar nu eliminat n totalitate.
Urmtoarea etap de cercetare a constat n sinteza unei noi cantiti de aditiv, n care s-a
obinut reducerea procentului de sodiu rezidual.
n paralel cu activitatea de sintez a unei noi cantiti de aditiv a fost luat n calcul i
ipoteza reducerii timpului tehnologic de amestecare a constituenilor, astfel nct aditivul s nu
aib timpul necesar dizolvrii. Dar modificarea unei tehnologii presupune alte implicaii. Timpul
tehnologic este impus n primul rnd de realizarea gradului de omogenizare a constituenilor.
Chiar i n aceast situaie, unele aspecte rmn neelucidate, cum ar fi faptul c n apele
recuperate se regsete aditiv care nu s-a dizolvat, dei a fost recuperat dup o perioad de timp
mult mai mare dect aceea n care a fost n malaxor.
Folosind noul aditiv, s-a fabricat o nou arj de combustibil. Rezultatele obinute au
confirmat unele ipotezele emise. De data aceasta, n masa de pulbere s-a regsit 70 % din
procentul de aditiv introdus. Testele prin trageri statice pe banc au confirmat nc o dat faptul c
prezena aditivului n reet, ct mai aproape de procentul necesar, duce la atingerea scopului
propus.
Specialitii n sinteza aditivului au continuat activitatea de cercetare pentru reducerea pe
ct mai mult posibil a existenei sodiului, dar n acelai timp au realizat i faptul c acest aspect
duce la creterea semnificativ a preului de producie. n acest stadiu de cercetare s-a pus i
problema modificrii tehnologiei de fabricaie a pulberii. Propunere a constat n nlocuirea apei
ca agent de emulsie n faza de amestecare a constituenilor cu un alt agent. Un posibil agent de
emulsie poate fi uleiul, agent n care dizolvarea aditivului este complet eliminat. n aceast
situaie apar dou aspecte noi.
Primul aspect const n faptul c n pasta stoars va rmne o anumit cantitate de ulei.
Acest lucru nu deranjeaz, avnd n vedere faptul c, oricum n reet se introduce i ulei, ca
agent emulgator pentru etapa de extrudare. Prin urmare, nu se mai introduce ulei ca un
constituent separat, ci se controleaz procentual uleiul rezidual.
Al doilea aspect se refer la costurile de fabricaie ale pulberii. Din acest punct de vedere,
apa este incomparabil mai ieftin dect uleiul dar, dac se procedeaz la recuperarea uleiului i
reutilizarea lui, este posibil ca unele aspecte privind costurile de fabricaie s se ncadreze n
limite acceptabile.
Revenind la activitatea descris mai sus, o modificare a fluxului de fabricaie presupunea
alte aspecte de ordin tehnic, financiar i de timp. Pentru continuarea cercetrilor s-a stabilit i
efectuat fabricarea unei noi arje de pulbere, la care procentul de aditiv a fost supradozat cu
diferena constatat la ultimele analize. n urma rezultatelor obinute la analizele de laborator i a
testelor experimentale au fost obinute rezultate foarte apropiate de cele stabilite ca obiectiv.
II.4 Aspecte experimentale
n Fig. 1 sunt reproduse imagini din nregistrrile video efectuate pe timpul efecturii
testelor experimentale pe banc. Din analiza lor se poate constata influena procentului de aditiv
din reet asupra flcrii.
294
AFASES - 2008 -
Fig. 1 Aspectul comparativ al jetului pentru o reet de combustibil cu procente diferite de aditiv
pentru suprimarea flcrii
III. REZULTATE I CONCLUZII
Folosirea aditivului pe baz de hexanitrocobaltiat de potasiu i oxid de magneziu n
pulberea motorului rachetei S24 a dus la diminuarea flcrii dincolo de seciunea critic a
ajutajului.
Testele dinamice, prin trageri de pe toate tipurile de avioane au demonstrat c motoarele
aeronavei nu au mai fost perturbate pe timpul lansrii.
REFERINE
[1] G. P. SUTTON: Rocket Propulsion Elements, Seventh Edition, A Wiley-Intercience
Publications, John Wiley & Sons, Inc. New York, 2001.
[2] A. DAVENAS: Solid Rocket Motor Design, vol. 170, Progress in Astronautics and
Aeronautics, AIAA, 1996.
295
AFASES - 2008 -
[3] C. BOYARS and K. KLAGER: Propellants: Manufacture, Hazards and Testing, Advances
in Chemistry Series 88, American Chemical Society, Washington, DC, 1969.
[4] V. T. TIGANESCU and L. ISTODE: Sisteme tehnice integrate. Sisteme propulsive, Editura
Univers Stiintific, Bucuresti, 2006.
[5] L. ISTODE: Propellants with Additives for Diminution of Flame, Proceedings of the XXVIth International Conference, Military Academy Publishing House, Bucharest, 1995.
Lucian Istode, lt.col(r) cercettor tiinific I drd. ing., Academia Tehnic Militar, Bucureti, George Cobuc nr. 8183, tel. 021 3354660 int. 198, e-mail: istodeluci@yahoo.com
**
Titic Vasile, gl.bg.(r) profesor univ. dr. ing., Academia Tehnic Militar, Bucureti, George Cobuc nr. 81-83,
tel. 021 3354660 int. 149, e-mail: titica@mta.ro
296
AFASES - 2008 -
297
AFASES - 2008 -
mului de micare a vaselor de extracie, sau oprirea mainii ntr-o anumit poziie a vaselor
(frnarea de manevr) i oprirea automat a mainii, independent de voina operatorului, ntrunul din urmtoarele cazuri considerate perturbaii sau avarii: dispariia tensiunii, scderea
presiunii fluidului de lucru pentru acionarea frnei, supranlarea vaselor de extracie,
depirea vitezei admise, suprasarcina etc (frnarea de siguran).
Mecanismul de execuie a frnelor const din saboii de frnare i din transmisia cu
prghii ce leag saboii de dispozitivul de acionare.
Mecanismul de execuie este comun (n majoritatea cazurilor) pentru frnarea de
manevr si pentru cea de siguran [1].
n cazul organelor de nfurare sub form de tobe, genile de frnare se afl pe
marginile interioare (fig. 4) sau pe marginile exterioare ale acestora (fig. 5).
Amplasarea genilor de frnare pe marginile interioare ale tobelor se aplic n cazurile
cnd ambele perechi de saboi au acelai dispozitiv de acionare.
Amplasarea exterioar se folosete la tobe simple, sau n cazul mainilor de extracie
mari, cnd fiecare pereche de saboi are acionare individual (fig. 6).
La roile motoare monocablu (fig.10), genile de frnare se dispun pe cele dou pri
ale canalului cablului.
La roile multicablu genile de frnare pot fi dispuse n prile exterioare (fig. 5) sau
ntre canalele cablurilor, n cazul cnd trebuie folosite mai multe perechi de saboi (fig.6).
n funcie de modul de suspendare a saboilor, organele executoare pot fi de dou
298
AFASES - 2008 -
tipuri: cu saboi fixai pe supori cu reazeme fixe i cu saboi pe supori cu reazeme articulate.
Organul de execuie al frnelor cu saboi i prghii (fig. 7) const din dou bare suport
(stlpii port-saboi) (1) articulate n reazeme (2), legate ntre ele prin tija (3) (tirantul care face
legtura dintre stlpii portsaboi) acionat prin ridicarea sau coborrea prghiei (4)
(triunghiului). Pe barele de susinere (1) sunt fixai suporii (5) ai saboilor de frn (rigid n
cazul deplasrii un-ghiulare i articulat n cazul deplasrii paralele). Pe suprafaa interioar a
suporilor se fixeaz saboii (6) care acioneaz direct asupra genilor de fr-nare. Deplasarea
saboilor n timpul frnrii este mpiedicat de reazemele (7) la extremitile suporilor (5) [1].
Diagnosticarea mecanismului de frnare a mainilor de extracie miniere const n
determinarea coeficienilor de siguran reali la aplicarea frnei de siguran i la aplicarea
frnei de manevr.
n vederea expertizrii i reglrii mainii de extracie msurtorile experimentale au
fost fcute la maina de extracie al instalaiei de extracie multicablu de la Puul auxiliar nr.1
incinta E.M. Livezeni (fig. 8), a crei prezentare se face n cele ce urmeaz.
2. Instalaia de extracie luat n studiu
Instalaia de extracie echipeaz Puul auxiliar nr.1, Incinta Livezeni din cadrul E.M.
Livezeni, care este destinat [4] pentru extracia transportul de materiale i personal din i n
subteran. Adncimea de extracie este de 300m. Cota rampelor din pu (punctul de ncrcare i
descrcare a schipurilor) este 284 (cota 300) i respectiv 584 (cota 0).
Instalaia de extracie ce deservete puul este echilibrat (cu cablu de echilibru) i este dotat
cu o main de extracie de tip MK 2,25x 4 (fig. 9), fabricat n fosta U.R.S.S. i a fost pus n
funcie n anul 1976. Maina este echipat cu dou motoare asincrone tip AKH-14-46-10 de
putere nominal 630 kW i turaia nominal de 585 rot/min i un reductor tip 2D-14 cu
raport de transmisie 6. Diametrul tobei de friciune (acionare) este de 2,25m. Sarcina static
maxim este de 328.000 N, sarcina maxim dezechilibrat este de 120.000 N.
Cablurile de compensaie n numr de dou au masa (pe metru liniar) respectiv de 7,6
kg/m i6,8 kg/m. Viteza maxim este de 11 m/s ridicat materiale, de 7 m/s cobort materiale,
6 m/s pentru transport personal i 0,5 m/s pentru revizie.
Timpul de frnare cu schipul plin ascendent la viteza de 8,64 m/s este de 3,77 s, iar
spaiul de frnare este de 7,8 m. Momentul maxim de frnare este de 264.870 Nm. Distana
reglabil ntre aiba de frnare de 2,4 m i saboi este de 1,2-1,5 mm.
299
AFASES - 2008 -
300
AFASES - 2008 -
k st e + 1 +
Td
la ridicarea ncrcturii (ncrctura pe partea moletei de deviere):
k st e 1
a 3 cr = g
[m/s2]
(10)
M'm g
e k st +
+1
Td
la coborrea ncrcturii (ncrctura pe ramura opus poziiei moletei de deviere):
e k st
(11)
a 3 cr = g
[m/s2]
M
'
g
m
e + 1+
+ k st
Td
la coborrea ncrcturii (ncrctura pe partea moletei de deviere):
e k st
(12)
a 3 cr = g
[m/s2]
M
'
g
m
e + k st +
Td
/
unde M m este masa moletelor de deviere.
Deceleraia de lucru se adopt pe baza relaiei:
(13)
a3 = (0,7 0,8) a3c r [m/s2]
4. Verificarea forelor efective din tirani i evaluarea coeficienilor de siguran reali
Pentru determinarea forelor de ntindere din tirani s-au lipit cte dou mrci
tensiometrice pe fiecare tirant (fig.10 i fig.11), diametral opuse, cu scopul de a elimina
efectul ncovoierii i cu ajutorul a altor dou mrci de compensare s-a realizat o punte
Wheatstone avnd dou ramuri active i dou ramuri pasive.
Puntea Wheatstone s-a echilibrat cu ajutorul unui compensator, diferite stri ale frnei i
s-a determinat deformaia specific a materialului.
301
AFASES - 2008 -
300
200
100
0
1
75 149 223 297 371 445 519 593 667 741 815 889 963 1037 1111 1185 1259 1333 0
-3
-100
-6
-200
-9
-300
Acceleraia
Comand frn
Fora
Spaiul
300
200
100
0
0
1
72 143 214 285 356 427 498 569 640 711 782 853 924 995 1066 1137 1208 1279
-100
-3
-200
-6
-300
Timpul *1/10 [s]
Viteza
Acceleraia
Comand frn
Fora
Spaiul
302
AFASES - 2008 -
Frna de manevr:
Fora de ntindere din tirani: F (daN)
(14)
Fs = s.Eo.Ss.10-6, Fs = 7,746x103, Fd = d.Eo.Sd.10-6, Fd = 7,404x103
Dj = 2,4
Dj diametrul jantei de frnare (m)
.i2 = 2,5
i2 raportul de amplificare parial(stlp port-sabot)
Momentul de frnare la frna stng
Momentul de frnare la frna dreapt
Fd i2 D j 2 .
Fs .i 2 D j 2...
, M Fd = 1,266 x10 4 (15)
, M Fs = 1,325 x10 4 , M Fd =
M Fs =
2
2
Moment de frnare total: Mt (daNm)
(16)
Mt = MFs + MFd, Mt = 2,591x104
cs coeficientul de siguran real la frnare determinat experimental
Mst1 = 8630
Mst1 moment static maxim (daN)
Mt
cs =
, cs = 2,967
(17)
M st 1
Frna de siguran:
Fora de ntindere din tirani: F (daN)
(18)
Fs = s.Eo.Ss.10-6, Fs = 7,902x103, Fd = d.Eo.Sd.10-6, Fd = 8,869x103
Momentul de frnare la frna stng
Momentul de frnare la frna dreapt
F
Fs .i 2 D j 2. ..
d i 2 D j 2 .
, M Fd = 1,517 x10 4 (19)
, M Fs = 1,351x10 4 , M Fd =
M Fs =
2
2
Moment de frnare total: Mt (daNm)
Mt = 2,868x104
(20)
Mt = MFs + MFd,
cs coeficientul de siguran real la frnare determinat experimental
Mst1 = 8630
Mst1 moment static maxim (daN)
Mt
cs =
, cs = 3,285
(21)
M st 1
5. Calculul deceleraiilor critice la alunecarea cablului
303
(22)
(23)
= 190
180
= 0,3
(24)
(25)
(26)
AFASES - 2008 -
- coborre de sarcin
e . k st
a crc = g .
, acrc = 2,897 .sec-2
e + k st
- deceleraia maxim admis (m / s2)
amaxc = 0,85.acrc, acrc = 2,463 sec-2
- schipuri goale:
k st .e . 1
, acrr = 4,307 m.sec-2
a crr = g
.
k st .e + 1
amaxr=0,85.acrr, amaxr = 3,661m.sec-2
Deceleraiile reale la frnarea de siguran:
Tabelul 5.1 Deceleraiile reale
V1 [m/s]
Nr.crt.
dQ [daN]
V0. [m/s]
1
0
7,78
7,78
2
0
6,66
6,66
3
0
8,64
8,64
4
0
7,64
7,64
(27)
(28)
kst = 1
(29)
(30)
tef [s]
2,91
2,39
3,13
2,9
tm [s]
0,78
0,77
0,64
0,57
a [m/s2]
2,674
2,78
2,76
2,634
Dceleraiile reale conform tabelului 5.1 sunt mai mici dect cele admise.
6. Concluzii
Coeficientul de siguran real calculat cu fora efectiv din tirant, obinut n urma rezultatelor
msurtorilor experimentale, la aplicarea frnei de siguran i la aplicarea frnei de manevr
se ncadreaz n limita admis. Deasemenea deceleraiile i timpii de ntrziere, la aplicarea
frnei de siguran, au valorile cuprinse n limitele admisibile.
Bibliografie
[1] Magyari A., Instalaii mecanice miniere, Editura Tehnic, Bucureti, 1990;
[2] S.C. TECHNOSAM S.R.L., Expertizarea i reglarea mainii de extracie tip MK 2,1x4
montat la Puul cu schip E.M. Lonea;
[3] electronic a tensiunilor mecanice, Editura Universitas, Petroani, 2005;
[4] *** Documentaie tehnic E.M. Livezeni.
304
AFASES - 2008 -
I. INTRODUCERE
Materialele compozite sunt astzi special create, cu cele mai potrivite alctuiri micro i
macro-structurale, pentru a corespunde unor exigene deosebite n privina rezistenei
mecanice, a rigiditii, a greutii specifice reduse, a stabilitii dimensionale, termice si
chimice, a rezistenei la oboseal, la oc i la uzare, a proprietilor izolatoare, a esteticii i, nu
n cele din urm, a imperativelor economice.
Materialele asupra crora voi insista sunt materialele compozite armate cu fire lungi,
iar motivaiile acestei alegeri sunt multiple. Una din aceste motivaii rezult din calitatea
acestora de a pune n valoare n mod global comportarea remarcabil la traciune a fibrelor din
componena lor, fapt care autorizeaz folosirea acestora pe scar din ce n ce mai larg la
fabricarea reperelor din industria aeronautic. Un alt motiv constituie faptul c acest tip de
materiale sunt reprezentative pentru comportamentul materialelor ortotrope.
305
AFASES - 2008 -
II.
MODELUL MATEMATIC
0
0 C44 0
0
0
0
0
0
0 C55 0
0
0
0
0 C66
0
(2.1)
Pentru un astfel de material, n axele de ortropie nu exist termeni de cuplaj ntre
tensiunile normale i lunecrile specifice din tensorul deformaiilor.
Matricea complianelor are forma:
306
AFASES - 2008 -
S11
S
21
S
[S ] = 31
0
0
S12
S13
S 22
S32
0
0
0
S 23
S33
0
0
0
0
0
0
0
0
0
S55
0
S 44
0
0
0
0
0
0
0
S66
(2.2)
0
0
1 G23
0
0
0
0
0
0
0
1 G31
0
0
0
0
0
1 G12
0
(2.3)
i, fiind simetric, ij/Ei = ji/Ej pentru i j; i,j = 1,2,3.
Spre deosebire de materialele izotrope, la cele anizotrope nu exist relaii de
legtur ale modulelor de forfecare Gij cu modulele longitudinale Ei i coeficienii lui
Poisson ij.
307
AFASES - 2008 -
limea acestor capete de prindere este i ea limitat. De exemplu, pentru captul de prindere precizat
anterior, limea este de 45 mm. Din acest motiv este dificil de prins un specimen cu o lime mai mare.
innd cont de aceste limitri, este de dorit a se elimina sau, cel puin, de a se minimiza
concentrrile de tensiune induse de capetele de prindere. O soluie des ntlnit este de a amortiza
forele de prindere folosind materiale-tampon la capetele epruvetei. Aceste plcue pot fi prinse sau
lipite de capetele epruvetei.
308
AFASES - 2008 -
309
AFASES - 2008 -
E
F
A
1
2
2
(daN)
(mm ) (m/m)
(N/mm )
MPa
1
50
24
500
70
2,083
4167
2
100
24
1050
155
4,167
3968
3
150
24
1450
260
6,250
4310
4
200
24
1500
325
8,333
5556
5
250
24
1500
475
10,417
6944
6
300
24
1200
670
12,500
10417
Din pcate, am putut consta c mrcile tensometrice s-au desprins de pe epruvet.
ntruct remarc c valorile deformaiei specifice longitudinale prezint o evoluie anomal
dup pragul de 150 daN, putem considera c desprinderea mrcii de pe epruvet s-a iniiat la
aceast valoare a forei i c valorile indicate de puntea tensometric ulterior nu mai sunt de
ncredere i pot fi omise.
Oricum, dac putem ridica semne de ntrebare asupra liniaritii relaiei ntre tensiunea
i deformaia specific pn la rupere, este cert c n pentru valori mai mici de 50% din
valoarea de rupere comportamentul este liniar. Cu alte cuvinte, primele trei valori furnizeaz o
valoare destul de bun a modulului de elasticitate:
4167 + 3968 + 4310
E1 =
= 4148 MPa .
3
Similar s-a procedat pe direcia btelii, obinndu-se o valoare de 2219 MPa.
E1
G12
(3.1)
= arctg
1
2
E (1 + 21 ) G
2
12
310
AFASES - 2008 -
pentru min. Cu acest unghi se repet experiena de ntindere, n urma creia se obine o alt
valoare a lui G12. Procesul iterativ se reia pn la stabilitatea dorit a soluiei.
Dei la prima vedere metoda pare destul de laborioas, cu multe ncercri la diferite
unghiuri, de fapt este simpl i practic deoarece numrul de ncercri este dat n principal de
valoarea de plecare unghiului iar aceasta poate fi aleas n jurul valorii de 45 valoarea
teoretic a decuplrii efectelor.
Aplicnd acest principiu,
2
1
(1 + 0,166 )
3184 = 45,019deg
(3.2)
= arctg 4148
2
1
IV. CONCLUZII
311
AFASES - 2008 -
Bibliografie
[1] Fuiorea I.
[2] Fuiorea I.
[3] Ille, V., Soare, M., V.
[4] Melin L. N., Neumeister, J. M.
312
e-mail:
AFASES - 2008 -
I. INTRODUCTION
En 1894, Reynolds a dmontr que lcoulement de fluide change d'un tat ordonn et
prvisible en chaotique et imprvisible quand le nombre de Reynolds dpasse une valeur
critique. L'tat, la turbulence, souvent rencontr dans les applications techniques, et sa
signification pratique a attir l'attention de certains des meilleurs esprits dans les domaines de
la physique, des mathmatiques et de la technologie. Le progrs vers une thorie analytique
rigoureuse a t empch par le fait que la dynamique de turbulence est fortement nonlinaire. La description la plus gnrale d'un coulement de fluide est obtenue partir du plein
systme des quations de Navier-Stokes. Ces quations non-linaire sont connues depuis plus
d'un demi sicles mais les dtails de la transition et du mouvement turbulent ne sont pas
entirement compris. La difficult principale est que le systme des quations est ferm
seulement dans un nombre trs limit de cas. En plus, le problme des quations bien pos
dans le sens classique est limit un certain intervalle des paramtres ou des intervalles courts
de temps. Par consquent on doit employer des techniques numriques pour obtenir les
solutions approximatives aux quations de Navier-Stokes.
Le CFD est un excellent outil pour tudier le comportement des quations non
linaires du mouvement et pour gagner une vision physique sur les dtails des mcanismes
fondamentaux. Dans une simulation d'un coulement de fluide, en utilisant un code CFD, tout
la zone d'coulement est connue chaque point et moment, au moins vers le bas de la
rsolution choisie. Particulirement la pression qui est trs difficile obtenir d'une voie
prcise par des expriences, mme pour un seul point, est connue chaque point dans
l'intrieur du domaine de calcul. Les donnes exprimentales, si disponibles, ne peuvent
jamais fournir cette quantit massive d'informations. De plus en plus coulements complexes
de nombre de Reynolds levs peuvent tre simuls en raison de l'augmentation rapide de la
puissance des ordinateurs, du dveloppement des schma numriques et des modles
amliors de turbulence.
313
AFASES - 2008 -
II.
L'coulement compressible autour d'une section d'aile grande vitesse est compos
par des rgions distinctes avec leurs propres caractristiques. Il y a des rgions laminaires,
transitoires et turbulentes dans le champs de lcoulement. Au bord principal l'coulement est
laminaire, caractris comme tant ordonn, en suite des instabilits descendant en
coulement commencent se dvelopper et l'coulement subit le processus de transition et
devient turbulent. Dans les coulements transsoniques il apparat des ondes chocs qui
interagissent avec la couche de limite.
(2.1)
314
AFASES - 2008 -
ou rapport spcifique chaleur constant pression, Cp, spcifique chaleur constant volume, Cv,
R est la constante du gaz et T est la temprature.
p
Pour le gaz idal
est constant. Sous forme diffrentielle, a se peut crire :
T
p T
=0
(2.3)
p
T
Lquation (2.3) comment les changements de pression, densit et temprature sont
corrl dans les coulements compressibles. Pour mettre en vidence limportance du nombre
de Mach, on considre un coulement isentropique. En appliquant le principe de la
conservation du moment, on a :
dV
p
= M 2
(2.4)
p
V
Lquation de lnergie donne :
dT
dV
(2.5)
= ( 1) M 2
T
V
En fin, en combinant les quations (2.3), (2.4) et (2.5), on obtient :
p
dV
= M 2
p
V
(2.6)
A partir de ces quations on peut voir que pour un changement fractionnaire donn de
vitesse, c.--d. pour un dV/V donn, le changement induit correspondant de la densit, la
temprature et la pression dpendra de la place du nombre de Mach. Cest clair que le nombre
de Mach est le paramtre qui dtermine l'importance des effets de compressibilit en
coulement.
D'une manire gnrale, la compressibilit a un effet relativement faible sur les
tourbillons en coulements autour de parois. Ceci apparatre pour tre vrai pour Mach nombre
jusquau environ 5, si coulement non prouver grand pression changement au-dessus un
court distance tel pouvoir avoir travers une onde de choc.
2.2. Sparation
315
AFASES - 2008 -
En revanche on peut trouver des coulements spars galement derrire les objets
mousss. Le mcanisme pour une telle sparation est fondamentalement non visqueux (bien
que la viscosit joue un rle dans le comportement final d'coulement).
La physique de la sparation d'un coulement aprs un profil d'aile peut tre explique
comme suit. Devant le point de stagnation, une lvation considrable de pression se produit
le long de la direction de l'coulement, mais l'coulement ne spare pas en raison de l'absence
du frottement de paroi. Prs du bord dattaque, l'coulement ne spare pas non plus, puisque
le fluide entre dans la direction de la pression dcroissante. un certain point la plus basse
pression est atteinte et de ce point l'coulement trouvera un gradient adverse de pression.
Compar au jet principal, l'coulement retard dans la couche de limite souffre une
dclration relativement plus grande; par consquent le moment de l'coulement prs de la
surface est petit et la capacit du fluide de se dplacer en avant contre l'lvation de pression
est galement limite. En aval, cette petite quantit de moment et d'nergie le long de la
surface de corps est utilise pour surmonter l'lvation de pression et le frottement et, en
conclusion, le fluide est apport pour se reposer. La thorie classique de sparation
u
d'coulement dfinit un tel point auquel
= 0 comme point de sparation, o u est la
y y =0
vitesse tangentielle la surface et y est la direction normale sur la surface. La tension de
u
= 0 au point de sparation est nulle, cela est la force visqueuse
frottement w =
y y =0
disparat. En aval de ce point de sparation se produira un coulement invers, en raison d'un
gradient dfavorable de pression. En raison de cette inversion d'coulement la couche de
limite paissit le considrablement. La sparation d'un coulement autour d'une aile changera
nettement les performances de l'aile. Le coefficient de trane augmentera, qui augmente la
consommation de carburant, et le coefficient d'ascenseur diminuera et dans le plus mauvais
cas la cale peut se produire.
2.3. Sparation induite par choc
316
AFASES - 2008 -
L'paississement guide la partie externe de la couche loin la paroi qui produit des
vagues de compactage dans l'coulement supersonique. Ces vagues commencent dans la
partie supersonique de la couche limite et propagent dehors dans l'coulement externe o vont
former le choc. La couche limite convertit ainsi l'lvation pointue de pression en plus
progressive qu'elle peut ngocier, condition que le choc soit faible. Cependant, mesure que
la force de choc (c.--d. la grandeur de l'lvation globale de pression) augmente, les
gradients dans cette rgion montent et le degr de dclration dans la partie intrieure
augmente progressivement jusqu' ce que l'coulement vers l'avant ne soit plus possible sur la
surface. Ainsi la sparation apparat dans une configuration se dveloppante sans interruption
d'coulement qui est dvoile plus loin et autre mesure que la force de choc augmente
progressivement.
2.4. Interaction choc - couche limite
On peut rappeler que quand un obstacle est plac dans un coulement grande vitesse,
des zones locales vitesses supersoniques peuvent se former au voisinage de lobstacle. Le
passage d'une vitesse supersonique vers une vitesse subsonique est alors gnralement
accompagn d'une onde de choc. Une onde de choc est une rgion trs fine de l'coulement
dans laquelle les variations de pression, de temprature et de densit deviennent importantes
en la traversant. Dans la couche limite, les particules fluides se dplacent des vitesses
subsoniques; le choc gnr dans l'coulement externe ne peut donc attendre cette zone.
Divers exprimentes physiques d'coulement compressible stationnaires sur plaques
plane (voir l'ouvrage de Schlichting [3]) montrent qu'au voisinages du choc, l'augmentation de
pression le long de la couche limite laminaire ou turbulente est plus graduelle que dans
l'coulement externe; la distribution de pression "diffuse" alors prs de la paroi, de plus
cause d'une rgion subsonique transversale plus importante, cette diffusion est, plus
prononce dans une couche limite laminaire qu'une couche limite turbulente. On peut
s'attendre alors parfois qu'une couche limite laminaire, qui supporte qu'une trs faible
augmentation de la pression, dcolle en amont du front de choc en prsence d'un nombre de
Reynolds modr et un nombre de Mach suffisamment important.
Les figures 4 et 5 montrent divers exemples d'interactions choc - couche limite dans
des coulements transsoniques stationnaires autour de profils d'aile. La figure 4 montre un
exemple de dcollement d'une couche limite laminaire respectivement en amont et en aval du
choc. La figure 5 montre une couche limite turbulente en amont du choc et l'absence de
dcollement; ce qui prouve qu'en prsence d'une agitation turbulente, la couche limite
supporte de plus importants gradients de pression.
317
AFASES - 2008 -
L'tude des coulements transsoniques autour de profils d'aile a fait l'objet de plusieurs
tudes tant exprimentales que numriques durant ces quelques dernires dcennies, en raison
du grand intrt que suscite cette classe d'coulements dans le domaine aronautique. Une vue
exhaustive de ces investigations est rsume dans l'article de McCroskey [4]. Cette auteur
relate plus spcialement les problmes du dcrochage dynamique intervenant dans les
mouvements de tangages d'un profil d'aile. Dans ce contexte, le caractre instationnaire mane
du mouvement oscillatoire de l'obstacle. En revanche, trs peu d'tudes exprimentales
examinent le caractre instationnaire inhrent l'coulement qui peut intervenir dans cette
catgorie d'coulements transsoniques et qui peut avoir des influences cruciales sur les valeurs
des paramtres globaux et sur l'apparition de vibrations induites. Dans ce qui suit, nous
relatons diffrents travaux concernant ce domaine dans le cas des faibles et grands nombres
de Reynolds.
Cas des grands nombres de Reynolds
Pour les grands nombres de Reynolds, on peut citer les tudes exprimentales de
McDevitt, Levy et Deiwert [5] concernant des coulements transsoniques autour d'un profil
d'aile de type arc de cercle 0 d'incidence pour des nombres de. Reynolds compris entre 106
et 1.7x107 et des nombres de Mach compris entre 0.75 et 0.78. Les diffrentes visualisations
et mesures de pression ralises, ont permis de distinguer des domaines d'coulements
stationnaires et instationnaires ainsi que des dcollements de bord de fuite (trailing edge
sparation) et des dcollements induits par onde de choc (shock induced sparation) en
fonction du nombre de Mach et du nombre de Reynolds. Cette distinction est rsume dans la
figure 6. Afin d'valuer en dtail le comportement de l'coulement dans le domaine
instationnaire, des signaux de pression ont t valus 50% et 77.5% l'extrados et
l'intrados du profil, pour un nombre de Mach variant entre 0.76 et 0.78 et un nombre de
Reynolds gale 7.3x106. Ces signaux ont montr un caractre priodique et asymtrique des
variations de pression tant exactement en opposition de phase). Les visualisations montrent
un alternement intrados - extrados entre le dcollement de bord de fuite et le dcollement
induit par onde de choc. En augmentant la valeur du nombre de Mach, les oscillations
diminuent en amplitude pour finalement cesser M=0.78; l'coulement redevient alors
stationnaire.
Les travaux exprimentaux de Seegmiller, Marvin et Levy [6] ainsi que ceux de
Seegmiller, Marvin, Harrison et Kojima [7], ont permis, par des techniques de vlocimtrie
laser, de quantifier des profils de vitesses moyennes, de l'nergie cintique turbulente et des
contraintes de Reynolds Re=11.106 dans les zones dcolles pour les diffrents rgimes
transsoniques cits prcdemment (stationnaires et instationnaires). La figure 7 montre un
exemple d'coulement stationnaire M=0.79 avec dcollement la base de l'onde de choc et
s'tendant au del du bord de fuite du profil. Des ondes de Mach semblent maner de
l'interface entre la couche cisaille turbulente et l'coulement externe. Sur cette figure sont
superposs les profils de vitesses moyennes. La figure 8 montre une visualisation d'un
coulement instationnaire M=0.76 durant quatre instants d'un cycle oscillatoire.
L'coulement commence par prsenter une srie de faibles ondes de choc environs 85% de
la corde du profil. Ces ondes se consolident, coalescent et se dplacent en amont du profil en
dcollant la couche limite et provoquant un sillage turbulent important. Lorsque l'onde de
318
AFASES - 2008 -
choc atteint le milieu du profil les vitesses locales dcroissent et le choc se dissipe. Ce cycle
se rpte priodiquement. L'coulement intrados est essentiellement identique mais est en
opposition de phase.
319
AFASES - 2008 -
IV. CONCLUSIONS
Les coulements spars compressibles autour des sections d'aile sont d'intrt
considrable pour la technologie et les raisons scientifiques puisqu'ils comportent la
transition, la turbulence proche paroi et la sparation. la vitesse transsonique, l'interaction
entre le choc et la couche limite visqueuse est d'intrt. Si le choc est assez fort, on a la
sparation induite par le choc. Dans beaucoup d'applications, telles que l'coulement grande
vitesse autour d'ailes et des lames de turbine, la sparation de couche limite est d'importance
majeure.
Divers expriences physiques et numriques ont mis en vidence la forte interaction
qui existe entre une couche limite, dpendant essentiellement du nombre de Reynolds, et des
conditions de chocs, dpendant principalement du nombre de Mach. Lanalyse de cette
interaction est complexe puisque les hypothses bien connues de la couche limite deviennent
invalides.
Bibliographie
[1]
[2]
Brilloin M.
[3]
[4]
McCroskey, W.L.
320
AFASES - 2008 -
[5]
[6]
[7]
LARCO Ciprian, cpitan inginer aviaie, Academia Tehnic Militar Bucureti , e-mail:
ciprian_larco@yahoo.com
321
AFASES - 2008 -
Rezumat
n lucrare se prezint o metod de mbuntire a proprietilor stratului superficial al
materialelor pentru sculelor achietoare folosind procedeul fizic cunoscut sub numele
Implantaie de Ioni(II). Lucrarea cuprinde bazele procedeului de implantare a ionilor de TiN i
rezultatul aplicrii procedeulu la materialele din care se execut partea aciv a sculei
acietoare tip carburi metalice i oel rapid.
I. Introducere
Primele cercetri ale tehnologiei de implantaie a ionilor s-au desfurat la nceputul
anilor 1950, iar aplicarea industrial n prima jumtate a anilor 1970.
Implantaia de ioni este unul dintre cele mai avansate procedee de prelucrare final a
suprafeelor. Cel mai mare avantaj al acestui procedeu const n faptul c TiN produs este mai
dens i mai stabil chimic dect cea produs prin metode tradiionale. Din punct de vedere al
rezistenei la uzur i coroziune, experimentele au artat c placarea realizat prin implantaie de
ioni este de 5 8 ori mai bun dect cea obinut prin metode tradiionale. De fapt, se poate
considera c aceast acoperire este de nendeprtat.
II. Principiul i caracteristicile implantatiei de ioni
Procedeul implantaiei de ioni const n introducerea (implantarea) n corpuri solide de
atomi ionizai din substaa dorit. Procesul are la baz energia mare pe care o posed ionii, situai
ntr-o incint n care exist un vacuum naintat (6 10 5 Pa) i unde sunt accelerai n cmp
electric cu potenial cuprins n domeniul 20 300 KeV. Ionii, avnd energii de la civa KeV
322
AFASES - 2008 -
pn la civa zeci de MeV, ptrund n adncimea corpului solid i datorit nucleelor atomice
intercalate n reeaua cristalin a materialului rmn implantai n aceasta. n cazul unor energi de
ionizare foarte mari, de ordinul MeV o cantitate mic de ioni este respins de materialul de baz.
O cantitate mai mare de ioni implantai se obine n cazul pulverizrii atomilor n materialul de
ionizat (pulverizare ionic).
Schema de principiu a implantaiei de ioni este reprezentat n figura 1.
Ca rezultat al implantaiei de ioni, n stratul superficial (SS), de grosime 0,01 - 1 m se
introduce o cantitate determinat de ioni, fapt care conduce la modificarea, n sens pozitiv, a
proprietilor fizico-chimice ale stratului n raport cu cele ale materialului de baz.
323
AFASES - 2008 -
la valori normale n limita de 0,2 0,3 m. Obinerea unei adncimi mai mari a implantaiei i a
cantiti de ioni implantai se poate realiza acionnd asupra pulverizrii ionilor suprafeei int
(o singur bombardare ionic triaz cteva mii de atomi).
Implantaia de ioni este nsoit de creterea temperaturii suprafeei materialului de
prelucrat. Astfel, bombardarea ionic poate determina, n zona local, o temperatur de
aproximativ 1000 0 C ntr-un interval mai mic de 10 11 sec., care determin un gradient de
0
temperatur de circa 10 14 K/sec.
Cldura materialului implantat depinde cel mai mult de energia i doza ionilor i se
determin pe baza densitii puterii de alimentare. n cazul unei densiti de putere de 10
KW/m 2 suprafaa metalului se nclzete pn la 100 0 C, pentru densiti de 100 KW/m 2
temperatura ajunge pn la 350 - 500 0 C n cteva minute, iar n cazul unei densiti de
aproximativ 6000 KW/m 2 materialul poate fi pulverizat sau vaporizat.
Bombardarea cu ioni de nalt energie poate conduce la fisurarea intern, care apare n
materialul iniial, la mrunirea structurii, precum i la omogenizarea acesteia. Tehnologiile de
baz care utilizeaz implantaia de ioni dau posibilitatea alierii materialelor respective.
Principalele procede de implantare cu ioni sunt prezentate tabelul 1:
Tabelul 1.
Procedeul
Substratul
Adncimea stratului
implantat
Material supus
implantaiei
20 1000nm
Material +strat de
metal
20 200nm
Idem
50 1000nm
Material supus
implantaiei
10nm 1m
sau
Metalele i aliajele dure lor pot fi implantate cu elemente nemetalice (N, C, etc.) sau
metalice (Al, B, TI, etc.) Scopul de baz al implantaiei de ioni este obinerea unui strat cu
duritate mare, cu rezisten ridicat la uzur i coroziune i cu proprieti superioare de ungere.
Obinerea proprietilor dorite ale stratului depinde de alegerea corect a substanei de implantat,
stabilirea dozei opime de ioni precum i a valorilor energiei ionilor i a temperaturii suprafeei de
implantat.
Relativ, uor se comport la implantaie metalele care conin azot, bor, carbon, staniu,
siliciu, crom, etc. i mai greu cele cu titan, aluminiu, etc.
Creterea duritii stratului superficial se obine, aadar, prin implantarea ionilor altor
substane, de regul toate elemente uoare (azot, carbon, bor, etc.), n stratul superficial al piesei.
Rezultatul acestui proces const n introducerea unor tensiuni nsemnate n strat, blocarea
micrii dislocailor, formarea unor faze dure de nchidere, de tip nitruri, boruri, carburi, etc., cu
o dispersie mare.
324
AFASES - 2008 -
Materialul
WC-Co
Oel de scule
WC-Co
Oel,
WC,
WC+Co
Valuri acoperite pentru cupru Oel aliat
i aluminiu
Tarozi
Oel rapid
Filiere
Oel rapid
Scule
Burghie pentru metale
Oel rapid
achietoare
Burghie pentru grafit
WC
Freze medicale
WC+Co
Freze disc
Oel rapid
Cuite pentru filetat
Oel rapid
Cuite disc profilate
Oel rapid
Elementele
active
ale
Organe de matrielor pentru mase plastice Oel rapid
maini
Carburator pentru injecie
Oel de scule
Lagre cu beriliu
Aliaj cu beriliu
Destinaie
Proteze
Aliaj cu Ti
special
Ti6Al4V
Ionii
implantai
Tabelul 1.
Creterea
durabilitii
Destinaia
C
Azot
Azot
Azot
5
2-4
5
2-5
Azot
3-6
Azot
Azot
Azot
Azot
Azot
Azot
Azot
8-10
3-4
2-6
6
3
2
5
11
Azot
Azot
Bor
Azot
2-10
10
3-5
100
400
325
AFASES - 2008 -
Gheorghe Mare* , conf. univ. dr. ing., Universitatea Transilvania din Braov, B-dl Eroilor,
nr.9, Braov.
326
AFASES - 2008 -
Rezumat
Lucrarea conine elemente referitoare la posibilitile tehnologice de prelucrare termic
a stratului superficial al suprafeelor metalice folosind laserii. Se prezint principalele
caracteristici ale urmtoarelor operaii: clire superficial, aliere, fuziune (placare normal i
placare sticloas) i durificare prin explozie.
Introducere
Aplicarea corespunztoare a tratamentului termic superficial impune alegerea unui regim
optim de nclzire-rcire i anume un regim rapid sau ultrarapid. n acest sens se recomand
realizarea unor viteze de nclzire de ordinul 102 103 C / sec , pentru regimul rapid i mai mari
0
de 10 3 C / sec pentru regimul ultrarapid. n aceste condiii se obin puteri de concentrare a
cldurii de ordinul 103W / m2 , pentru regimul rapid i de ordinul 106 1010W / m2 pentru regimul
ultrarapid. Regimurule menionate pot fi obinute folosind radiaia LASER.
n cazul dispozitivelor cu LASER izvorul de energie termic este raza (fascicolul) LASER,
n spectrul vizibil sau invizibil, coerent, practic monocromatic, limea spectrului fiind de
aproximativ 10-12m i avnd divergen mic (ex. n cazul laserului cu gaz, diametrul
fascicolului, dup un parcurs de 1 km, ajunge la aproximativ 7 cm.). Radiaia LASER se
focalizeaz ntr-un fascicol continuu sau pulsatoriu n form de impulsuri scurte, cu un diametru
pornind de la 310 m i pn la civa mm, fascicol care n zona radiaiilor continue are o mare
densitate de putere (maxim 1016 Kw/m2), suficient pentru nclzirtea instantanee local a
suprafeei (viteza maxim a gradientului de cldur este de aproximativ 106 K/s) nu numai le
temperatura de topire ci i la cea de fierbere.
327
AFASES - 2008 -
328
AFASES - 2008 -
Impulsul produs poate avea durate de 0,1 8 msec i o stuctur temporal, constnd din
vrfuri separate, scurte (~1 sec.), cu intervale de 3 10 sec. Variind parametri rezonatorului,
se poate ordona distribuia vrfurilor n timp i amplitudinea, sau se poate obine un un impuls
neted (fr vrfuri). Puterea impulsului n regimul generrii relaxate variaz n limitele
10 3 10 6 W.
n tratamentele termice cel mai utilizat este LASERUL cu mediul activ gazos
(CO2+He+N2), cu excitaie electric. Pompajul se realizeaz prin transmitere a energiei
moleculelor de CO2. Heliul se adaug pentru mrirea randamentului laserului.
n ultimul timp se folosesc laseri cu CO2, excitai cu descrcare prin scnteie transversal
i cu presiune pn la 105 Pa a amestecului gazos, aa numii LASERI TEA (Tranversaly Exited
Atmospheric Pressure).
Procesul termofizic din zona nclzit cu radiaie LASER const n cedarea energiei
cuantice a fotonilor, care alctuiesc radiaia LASER, gazului elecronic din suprafaa metalului
iradiat, iar acesta - la rndul lui cedeaz energia reelei cristaline. Temperatura gazului
electronic se uniformizeaz ntrun timp extrem de scurt, de ordinul a 1013 ...1014 sec. iar aceea
a reelei cristaline ntrun timp de aproximativ 1011 sec. astfel c radiaia LASER devine surs
termic efectiv ntr-un timp de ordinul a 109 sec. Teoretic, se demonstreaz c pentru impulsuri
cu durate mai lungi de 108 sec. i cu densiti de putere superficiale mai mici de 10 9 W / mm 2 ,
diferana de temperatur dintre gazul electronic i reeaua ionic atinge valori de ordinul a 103
0
K, ceea ce face posibil aplicarea legilor conduciei termice, n rezolvarea problemei nclzirii
metalelor cu surse de energie concentrate.
II. Procedee i metode de utilizare a laserilor nclziri rapide
Prelucrarea termic a metalelor cu LASER se poate face n urmtoarele variante de lucru:
1. continuu, unde n mod obinuit se folosete LASER cu gaze (moleculare) la care
mediul activ este CO2 cu adaos de He i N2; lungimea de und activ este de 10,6 m, puterea de
la cteva sute de wai la 20 Kw, deseori fiind de (2,55) kw iar n aplicaii industriale frecvent de
25 Kw; laserii permit obinerea unui gradient mai mic de temperatur pe suprafa n comparaie
cu metoda prin impulsuri, necesitnd o mai mic densitate de energie i un timp mic de
expunere;
2. prin impulsuri, unde se folosesc laseri solizi pe baz de cristale de YAG + Nd;
lungimea de und activ este de 1,06 m, puterea medie a impulsului 1Kw, durata impulsului
(10-310-6) sec. i frecvena impulsului de ordinul zecimilor de secund; gradientul de
temperatur este mic iar densitatea de energie, la aplicaii industriale, de 109Kw/m2.
Prelucrarea termic cu laseri conduce n principiu la creterea durabilitii sculelor i a
organelor de maini prin urmtoarele grupe de tehnologii (tab.1):
1 - tehnologii fr topirea SS al materialului de prelucrat
a) -clirea superficial a oelurilor de scule pentru presare la rece, a oelurilor rapide, de
construcie i slab aliate, a oelurilor de rulmeni i a fontelor.
Parametrii de lucru sunt:
- densitatea de putere: 1025104Kw/m2;
- timpul de expunere: 1021 secunde; 0
- gradientul vitezei de nclzire: 104 Kw/s.
329
AFASES - 2008 -
Obinerea unui strat clit mai subire dect 0,12,5 mm i mai dur dect la clirea prin
inducie se poate face prin mrirea timpului de expunere de ordinul ctorva secunde.
Aplicarea LASERILOR la diverse operaii de prelucrare termic
Tipul operaiei
Clire superficial
Aliere
Fuziune superficial
5104106
(placare)
Fuziune superficial
105107
(placare sticloas)
Durificare
prin 1081010
explozie
milisecunde
microsecunde
nanosecunde
Tab.1
Tipul laserului
continuu
-continuu
-prin impulsuri
-continuu
-prin impulsuri
-continuu
-prin impulsuri
prin impulsuri
330
AFASES - 2008 -
331
AFASES - 2008 -
332
AFASES - 2008 -
Gheorghe Mare* , conf. univ. dr. ing., Universitatea Transilvania din Braov, B-dl Eroilor,
nr.9, Braov.
333
AFASES - 2008 -
334
AFASES - 2008 -
speed [m/s]
Starting
with
this
consideration, more complex and
better data processing can be
15
obtained using the MATLAB
programming environment. This
10
product offers a lot facilities (usual
5
known as toolboxes) that can
provide a lot of analysis.
0
0
1000 2000 3000 4000 5000 6000 7000 8000 9000 Moreover, the graphic interface is
very suggestive and user friendly.
no. of values
Fig. 2 Fire engine take-off
So, we will use this environment
from now on.
A lot of aspects concerning the general data processing will finally lead to one or more
dynamic serials. Among these aspects we can mention:
The existence and the level of the noise that occur into the dynamic serials of
an experimental data process;
Filtering the signal, if this thing hasnt been already realized during data
acquisition;
2xy () =
Gxy ( j)
(1)
Gxx ()G yy ()
where G () represents the unilateral power spectral densities, having asymmetrical limits:
[0, ) ; when bilateral spectral power densities are involved, the limits are symmetrical:
(, ) .
The coherence function between the two signals is a real amount, given by:
xy ( ) =
Gxy ( )
Gxx ( ) Gyy ( )
S xy ( )
S xx ( ) S yy ( )
Gxy ( ) Gxx ( ) G yy ( )
2
and
335
S xy ( ) S xx ( ) S yy ( ) then
2
(2)
AFASES - 2008 -
0 xy2 ( ) 1
Before computing
(3)
2
( ) we must eliminate the non-null averages from both signals,
xy
to avoid discontinuity in the self-origin of the Laplace transform [1, 5]. Moreover, we should
check whether Gxx ( ) and G yy ( ) have no zeros to avoid a division-by-zero error in (1).
The fundamental statistic equations for a linear SISO are taken into account:
S yy ( ) = W ( ) S xx ( )
(4)
S xy ( ) = W ( ) S xx ( )
(5)
where
W ( )
becomes:
( ) =
2
xy
W ( ) Gxx ( )
Gxx ( ) W ( ) Gxx ( )
2
=1
(6)
and, for a linear system, the coherence function equals 1 at all the frequencies.
On the other hand, if the signals were not correlated, meaning Gxy ( ) = 0 for all the
frequencies, then the coherence function is zero for all the frequencies. Hence, if the
coherence function is non-null and lower than 1, then two possibilities will occur:
S
2
=
1 2
Z
S + Z = y
(7)
S = 2 y ; Z = (1 2 ) y
(8)
A good example in using eq. (8) is the take-off situation of a fire engine. We
considered x(t ) as the fire engines speed and y (t ) the position of the throttle [4]. Applying
the above-mentioned algorithm we could get the graphics in fig. 3. So, anybody can notice
that the noise is really important in the signals.
Fig. 4 depicts the noise analysis for 18 non-filtered dynamic serials for different
take-offs of the same fire engine. The conclusion is the same: the noises are too high and
should be reduced.
336
AFASES - 2008 -
Total signal
Useful signal (S)
Noises (Z)
Values [%]
NOISES
Fig. 4. Maximal percentage values of the useful signal and noise for a group of experimental tests
(non-filtered signals of the fire engines speed)
337
AFASES - 2008 -
Most commonly met reasons to face outliers [3] are given by electromagnetic
interference, transducers malfunction, bad connections and so on.
The above mentioned software product helps us providing an example to locate and
eliminate the outliers, using its medoutlierfilt toolbox (Nov. 2006).
According to the classical theory, the outliers are those values that exceed 3 times the
intequartilic interval. This interval is defined as IQR = q3 q1 where q3 is the superior
quartile and q1 is the inferior one.
Fig. 6 depicts the result of a long (8192 events), non-filtered serial analysis in locating
and eliminating the outliers. We can notice that IQR=8019 gives no outliers.
Superior
quartile
Median
Trust interval
Inferior
quartile
All values
Cutting
interquartiles
Not valid
As can be noticed in the picture above, since the chosen IQR was pretty big, it seems
like there are no outliers. So, prior determining the outliers, a reasonable limit of IQR should
be considered. The IQRs value is most commonly determined on the researchers expertise in
this respect.
To prove the previous statement, the same dynamic serial has been re-processed for a
lower value of the IQR, that has been set to 3*IQR=2.75 that is much closer to a more
accurate analysis.
The result was remarkable. 51 values had to be rejected, which means about 2.5% of
the total amount of data. But, on the other hand, the time developing of the serial has also
been shrunk, from the real time of 13.6467 seconds to a final time of 13.4 seconds. Should be
also noticed that the general shape of the curve was not changed. But this is not a common
rule: if the rejection rate is too high, the shape of the resulting signal can be seriously altered.
338
AFASES - 2008 -
Valid = 1997
Not valid = 51
Values
Values = 2048
Cutting
interquartiles
Mitra filter
K=20
with outliers
without outliers
Fig. 6. Signal with and without outliers
IV. Conclusions
The new, powerful software allows the researcher to develop much more accurate
analysis and filtering of the experimental data. Therefore, we consider that, along with the
basic hardware filtering of the experimental signals, a software filtering should be performed.
Of course, this cant be stopped here and a much deeper research should be further
considered. In this respect, filters represent a powerful tool.
References
[1] Anderson, T.W. - The statistical analysis of time series, Wiley, New York, 2001
[2] Copae, I. .a. - Dinamica autovehiculelor, Editura Ericom, Bucureti, 2006;
[3] Elmqvist, H. .a. - Realtime simulation of detailed physically based automotive,
Proceedings of the 3rd International Modelica Conference, Linkping, 2003
[4] Lscteu, E. -Studii privind stabilirea strategiei de mentenan a automobilelor din dotarea
trupelor de pompieri, Tez de doctorat, Academia Tehnic Militar, Bucureti, 2006
[5] Olson, B. - Nonlinear dynamics of longitudinal ground vehicle traction, MS Thesis,
Michigan State University, 2001;
*
Marin MARINESCU - Assoc. Professor Eng., PhD, Military Technical Academy, Bd. George Cosbuc nr. 8183, sector 5, Bucharest, Phone +402 335 4660 ext. 0314, 0150, e-mail: marinescu@mta.ro
**
Longin IACOBESCU - Dipl. Eng., PhD Candidate, IMUT Moreni, Bd. 22 Decembrie 1989 nr. 60, Moreni,
DB, Phone +40245 666 525, e-mail: imut@minisat.ro, imut_moreni@adslexpresss.ro
***
Ioan Filip - Professor Eng., PhD, Military Technical Academy, Bd. George Cosbuc nr. 81-83, sector 5,
Bucharest, Phone +402 335 4660 ext. 0149
339
AFASES - 2008 -
Abstract:
This subject, propose to treat some theoretical aspects in obtaining multi-layer metallic
materials witch high properties and purposes for special industry, using the Finite Element
Method.
I.
Introducere
Modelul mecanic al procedeului de placare prin explozie utilizat n aceast lucrare este
conceput pe baza observaiilor de mai sus, ntr-un spaiu bidimensional (fig.1) i const din
urmtoarele elemente:
1- placa mobil;
2- placa de baz;
3- atenuator;
4- exploziv.
340
AFASES - 2008 -
v
p = p0 0
v
(1)
s-a adoptat o varia]ie liniar care pornete de la valoarea indicelui adiabatic de 1,4 la
presiunea mediului ambiant i ajunge la valoarea maxim pentru presiunea de detonaie.
341
AFASES - 2008 -
III.
342
AFASES - 2008 -
Simularea realizrii mbinrii sudate s-a fcut prin introducerea la interfa a unor
elemente speciale de captare, concepute de autor, care devin active n momentul n care un nod
de pe suprafaa inferioar a plcii mobile ia contact cu un element de pe suprafaa superioar a
plcii de baz i are viteza mai mare ca viteza minim de sudare. n caz contrar, cu toate c se
realizeaz contactul ntre cele dou plci, n punctul respectiv, nu se stabilete legtura ntre cele
dou materiale. Parcurgerea etapelor specificate mai sus a condus la realizarea modelului
bidimensional cu elemente finite al procedeului de placare prin explozie, reprezentat n fig.4.
Fig.4 Modelul 2D cu elemente finite pentru simularea procedeului de placare prin explozie
Calculul numeric concret a fost efectuat cu programul COSMOS/M, n modulul de
analiz dinamic neliniar. Durata analizei a fost stabilit, din condiia ca frontul undei de
detonaie s se propage n toat masa explozivului iar placa mobil s se deplaseze n ntregime la
contactul cu placa de baz, la o valoare de 100 s cu un increment de timp de 0,1 s. n
consecin, rularea programului s-a fcut n 1000 de pai de integrare.
IV.
343
AFASES - 2008 -
344
AFASES - 2008 -
Concluzii
Bibliografie
[1] NSTSESCU V.,
[2] MUSTA t. M.,
[3] FUIOREA I.,
[4] NSTSESCU V.,
MUSTA
tefan
Mircea,
345
AFASES - 2008 -
Abstract
The main objective of the study is to find a small wind turbine below 1 KW, of low cost
and with a very low cut in speed from 2-3 m/s. The blades proposed for this small wind
turbine are from textile materials, have different forms and profiles and must resist at big
wind, storms and degradation. This blades profile must be adapted to the climatic condition
from Romania, especially Brasov area. Because of the low wind speed from Brasov is
necessary to implement small wind turbines with a cut in speed from 1-2 m/s. The blades
forms are designed and tested with the software CATIA V5 R15 and with the method of Finite
Element Modeling. An optimize blade geometry is obtained.
1. Introduction
In designing wind blades is important the profile symbolization. The profiles have two
important coefficients of symmetry and width, and are depending on the blades top and axis.
For our blades we used three blades profile which are depending on the symmetry and
width coefficients. This coefficient are noted with: Cav, CAa and Cv, Ca, differ with the
blades profile and can have values of 1 and 0.
In the figures 2, 3 and 4 are presented the blades profiles choose for this design.
346
AFASES - 2008 -
The design of the blades profile is made with the software CATIA V5 R15 (figure 5).
Were designed three forms of small blades and were made tests for each of them to find the
proper blade for the small wind turbine. The three blades profiles have the same dimensions,
only the form is different.
a)
b)
Fig.5. Blades profile
c)
The analysis is made for all the blades from figure 5, with the software Catia V5 R15
and with the method of Finite Element Modeling. This analysis will be made at distortion and
at stress.
This blades are assembled in a rim of small dimentions. The rim is a part of the small
wind turbine how have the cut in speed from 1 2 m/s. This blades are assembled into the rim
at an angle of 45degree.
The small wind turbine is tested at different wind speeds to find the best wind blades
that can work. The blades are made from textile materials and because of that the small wind
turbine is tested at distortion ans stress.
The three type of bades assembled with the small wind turbine are presented in figure
6, figure 7 and figure 8.
347
AFASES - 2008 -
348
AFASES - 2008 -
= 0,15;
E = 400 10 6 N / m 2 ;
The Density:
= 140 kg/ m 3 .
In table 1 are presented the number of tests that are made for the blades at different wind
speed.
Table 1
Nr.crt
1
2
3
4
5
6
The analysis results for stress and distortion are presented in the figures 8 and 9, for
blades profile a).
Blades analysis at stress
0,4
0,35
20
Stress [MPa]
Distortion [mm]
0,3
0,25
0,2
Distortion
0,15
15
Stress
10
0,1
5
0,05
0
Tests
Tests
The analysis results for stress and distortion are presented in the figures 10 and 11, for
blades profile b).
Small wind turbine analysis at stress
30
20
25
15
Wind speed
10
[m/s],
Distortion [mm]
Wind speed
Distortion
20
Wind speed
15
[m/s],
Stress [Mpa]
10
Wind speed
Stress
0
Wind speed
Distortion
11
14
17
0
Wind speed
Stress
11
14
17
0,391
2,44
6,25
11,8
19,2
28,2
The analysis results for stress and distortion are presented in the figures 12 and 13, for
blades profile c).
349
AFASES - 2008 -
Wind speed
Distortion
0
Wind speed
Distortion
11
14
17
0,199
0,323
0,476
Wind speed
Stress
5
0
Wind speed
Stress
11
14
17
5,4
10,2
16,5
24,4
0,338 2,11
It was found that from the three form of blades the best to be used for the small wind
turbine is the blade with the profile a), (figure 5). This blade with profile a) at distortion and
stress have lower values (figure 8 and figure 9) and this type of blades can be used for a small
wind turbine.
4. Conclusions
The paper presents different blades profile from which is choose only one. The blades
are tested at different conditions and result small blades which resist at distortion and stress.
References
[1] Godfrey Boyle , Renewable Energy.Power for a sustainabel future,ISBN 0-19-926178-4, Oxford
University Press, Oxford 2004.
[2] Bent Sorensen, Renewable Energy, Third edition, Edited by Elsevier inc., UK, 2004, ISBN 0-12
656153-2ef
[3] P. Konor, Publisher: Universe, Renewable Energy Policy, 2004, ISBN: 0595312187
[4] Visa, I., Duta, A., Renewable Energy Systems, Basics, Ed. Univ. Transilvania, 2005, ISBN 973635-541-1, 351 pg.
[5] Alexandru, P., Visa I., Talaba, D., Alexandru, C., Antonya , Cs., The static-dynamic modelation
of the mechanisms of guiding for weheel of the cars. Ed. Lux Libris, Brasov, ISBN 9739458173,
2005.
[6] Visa, I., Duta, A., Renewable Energy Systems, Applications, Ed. Univ. Transilvania, 2006, ISBN
973-635-657-4
350
AFASES - 2008 -
351
AFASES - 2008 -
352
AFASES - 2008 -
montarea unei game largi de aparate de ochire, dar i a unor alte tipuri de accesorii, cum ar fi
arunctorul de grenade, la partea inferioar a armei (FN F2000).
Armele hibrid dintre o puc i un arunctor de grenade cu autoncrcare (cum sunt
XM 29 i PAPOP) produc ns numeroase probleme la asamblarea componentelor, inclusiv a
sistemului complicat de ochire.
4. Cadena i selecia tragerii
Provenind din arme neautomate, carabinele de asalt se disting prin posibilitatea
selectrii modului de tragere: n regim semiautomat (foc cu foc sau un singur cartu tras la o
apsare a trgaciului) sau complet automat (tragere continu pn n momentul eliberrii
trgaciului).
Avndu-se n vedere c, n cazul executrii unui foc continuu, doar primele gloane se
duc spre int, datorit reculului care modific poziia armei, tendina este de adoptare a unui
mecanism ce limiteaz tragerea la o serie scurt, de doar dou sau trei cartue, efectul find
optimizarea consumului de muniie i contracacrarea irosirii nejustificate a acesteia.
5. Posibilitatea utilizrii ambidextre
O alt tendin nregistrat n construcia de armament automat n general i n cea a
putilor de asalt n particular este aceea de proiectare a armamentului astfel nct el s poat fi
utilizat n acelai mod i cu aceeai uurin att de trgtorii dreptaci ct i de cei stngaci.
Acest lucru presupune, n primul rnd, posibilitatea selectrii prin intermediul unei comenzi
mecanice a prii spre care se face evacuarea tubului cartu, spre stnga sau spre dreapta.
Exist deja unii productori de puti de asalt care au adoptat acest principiu n
construcia lor.
Ceea ce este interesant de menionat este faptul c realizarea acestui deziderat este
mult mai dificil n cazul putilor de asalt cu construcie de tip bull pup dect n cazul celor
clasice, ridicnd o serie de probleme mai greu de rezolvat pentru proiectant.
6. mbuntirea performanelor balistice
Proiectarea unui tip de muniie pentru o puc de asalt prezint un set de probleme
dificil de rezolvat: realizarea unui recul suficient de redus astfel nct s permit controlul
armei n timpul executrii focului automat; reducerea dimensiunilor i greutii muniiei pe
ct de mult posibil pentru ca infanteristul s poat transporta un numr ct mai mare de
cartue.
Un ideal n construcia armamentului automat, deci i n aceea a putilor de asalt, l
reprezint optimizarea raportului dintre performanele balistice ale glonului i fora de recul a
armei.
Este cunoscut faptul c efectul omortor al glonului de 5,56 mm se bazeaz pe
fragmentarea acestuia la impact, dac viteza rmas este suficient de mare. La armamentul
clasic care utilizeaz aceast muniie, cu o lungime a evii de 51 cm, distana care asigur
acest efect este de 150-200 m. n cazul carabinelor de asalt ns, care au o eav mai scurt,
efectul de fragmentaie se produce la maxim 50-100 m. Tendina de adoptare a carabinei
automate M4 pe scar larg n armata american a redeschis controversele asupra eficienei
muniiei de cal. 5,56x45 mm NATO.
Expertul britanic n armament i muniii Anthony G. Williams trage dou concluzii
din aceast disput n articolul su [1]. Prima ar fi c ar puteatea exista multe beneficii
financiare i logistice prin adoptarea unui singur tip de muniie, att pentru putile de asalt, ct
i pentru mitralierele utilizate n cadrul NATO. A doua este c ar fi util adoptarea unei
muniii cu o putere mai mare la int dect cea a cartuelor cal. 5,56x45 mm. Ambele
concluzii duc ctre aceeai direcie, i anume ctre un cartu intermediar ca putere ntre cele
dou calibre standard NATO, 5,56x45 mm i 7,62x51 mm. Un astfel de cartu ideal ar trebui
s asigure o eficacitate la int comparabil cu cea a muniiei de 7,62 mm, combinat cu o
for de recul suficient de redus pentru a permite controlul armei la executarea focului
353
AFASES - 2008 -
automat. ntrebarea pe care i-o pune autorul este dac acest lucru este posibil, iar concluziile
pozitive pe care le trage se bazeaz pe cteva experiene reuite observate de-a lungul
timpului i pe calcule energetice.
Britanicii au avut rezultate foarte bune acum peste 50 de ani cu un cartu de calibrul
7x43 mm. Aceasta pare s fie limita superioar. Limita inferioar pentru un astfel de cartu o
putem fixa la 6 mm, ceva mai mic de att neputnd avea performane superioare actualului
cartu de 5,56 mm.Un element important n balistica glonului l reprezint raportul dintre
calibru i greutatea proiectilului. n calculele de balistic romneti se utilizeaz coeficientul
de greutate, exprimat ca raport dintre greutatea proiectilului q i ptratul calibrului acestuia:
q
(1)
cq = 3
d
Autorul britanic utilizeaz ca element de comparaie densitatea secional (SDR
sectional density ratio), definit ca raport ntre calibrul d, exprimat n inch i masa m:
d
(2)
SDR =
m
Cu ct densitatea secional este mai mare, cu att caracteristicile aerodinamice i
efectul la int ale glonului sunt mai bune. O densitate secional de 0.230 ar asigura unui
glon o vitez mai bun dect a unuia cu calibrul de 7,62 mm (care la o mas 9,33 grame un
SDR de 0.217), i deci rezultatele la int dorite la distane mari. Glonul clasic de 5,56 mm
are un SDR de 0.174, iar varianta recent introdus, Mk262, are un SDR de 0.220 la o mas de
5 g. Este de asemenea necesar o energie la gur de maxim 2500 J, astfel nct fora de recul
generat s nu fie peste limitele acceptabile pentru o puc de asalt.
n aceast arie a muniiilor cutat de autor se gsesc cteva variante constructive deja
existente:
glonul de calibru 7 mm, cu o mas de 8,4 g i o vite iniial de 770 m/s,
genernd o energie la gura evii de 2500 J;
glonul de 6,85 mm, cu masa de 7,9 g, viteza iniial de 784 m/s i energia
la gur de 2430 J;
glonul de 6,5 mm, cu masa de 6,9 g, viteza iniial de 820 m/s i energia
la gur de 2330 J;
glonul de 6,35 mm, cu masa de 6,5 g, viteza iniial de 834 m/s i enrgia
la gur de 2260 J;
glonul de 6 mm, cu masa de 5,9 g, viteza iniial de 854 m/s i energia la
gur de 2150 J.
Cteva cartue care ndeplinesc criteriile de mai sus au fost introduse nu de mult
vreme i n servciul militar, dnd dovada c n viitor aceasta ar putea fi soluia care va fi
adoptat.
Unul dintre ele este Remington SPC (Special Purpose Crtridge) de cal. 6,8x43 mm,
care trage un glon de 7,45 g cu o vitez iniial de 808 m/s i o energie la gur de 2430 J,
dintr-o eav cu lungimea de 16.5 inch. Acest glon dezvolt o energie la gur cu 55% mai
mare dect cea a glonului de 5,56 mm, la o distan de 550 m, energia lui fiind cu 84% mai
mare datorit coeficientului balistic superior (SDR=0.214). Traiectoria lui este aproape
identic cu aceea a glonului de 7,62 mm pn la distana de 500 m. Acest cartu este utilizat
de forele speciale americane.
Un altul este glonul de 6,5 mm Grendel (6,5x38 mm), ncapsulat ntr-un cartu
asemntor celui folosit de armele AKM (7,62x39 mm). Cu o greutate de 8 g i o vitez
iniial de 770 m/s, acesta genereaz o energie la gur de 2370 J, avnd o comportare
aerodinamic cu mult superioar gloanelor de cal. 5,56 Mk 262 sau 6,8 mm (SDR=0.252).
Utilizate cu o eav de puc de asalt ceva mai lung sau cu o mitralier, aceste gloane au
caracteristici mult mai bune dect cele de 7,62 mm calibru NATO.
354
AFASES - 2008 -
355
AFASES - 2008 -
356
AFASES - 2008 -
la ocularul aparatului de ochire al armei ctre ochiul trgtorului, acesta putnd cuta intele i
efectua trageri asupra lor, timp n care capul i zona superioar a toracelui sunt meninute n
spatele unei obstacol. Atunci cnd nu este nevoie de el, dispozitivul Oavd poate fi rotit n
afara axului lunetei standard a armei.
9. Aparatele de ochire
Probabil cea mai mare influen asupra preciziei tragerii, n afara antrenamentului
trgtorului, o au aparatele de ochire, iar acesta este un domeniu care a nregistrat schimbri
considerabile n ultimii ani.
Iniial, armele de asalt aveau un sistenm mecanic de ochire, format din nltor i
ctare. n ultimii ani, aparatele de ochire optice au devenit mai uzuale. Prima puc de asalt
conceput pentru a putea folosi n mod standard un aparat optic de ochire a fost EM-2. Acesta
era un aparat optic care nu mrea imaginea, sistemul de lentile avnd doar rolul de a
suprapune reticulul cu imaginea intei. Pentru a diminua minusul pe care l reprezint
poziionarea aparatului de ochire deasupra evii, firma productoare l-a combinat cu un mner
pentru transport.
Prima arm care a intrat n serviciu cu un aparat de ochire optic standard a fost puca
de asalt austriac Steyr AUG, urmat de arma britanic Enfield SA 80, cele dou aparate
avnd mrirea unghiular de 1,5x, respectiv 4x.
Se tie c, cu ct mrirea unghiular este mai mare, cu att precizia tragerilor la
distane mari este mai bun, dar scade cmpul vizual al aparatului, fcnd mai dificil
prinderea unei inte n obiectivul aparatului. Astfel a aprut un element alternativ la aparatele
oprice de ochire, i anume dispozitivele de marcare a intei cu punct rou. Acestea
utilizeaz ca surs de energie o baterie ce permitea folosirea lor att pe timp de zi ct i pe
timp de noapte. Avantajul lor l reprezint faptul c trgtorul poate inti cu ambii ochi
deschii, meninnd n acelai timp posibilitatea de a percepe ntregul cmp vizual. Ele sunt
foarte uor de utilizat, dar nu sunt eficiente la distane mari.
Puca de asalt german HK G36 ofer posibilitatea montrii att a unui aparat de
ochire telescopic, ct i a unuia cu marcarea intei, unul deasupra celuilalt.
De asemenea, aparatele optice pot fi suplimentate cu marcatoare laser, care
proiecteaz i ele un fascicul concentrat de lumin asupra intei.
n zilele noastre se utilizeaz i o serie de aparate de ochire pe timp de noapte, att cu
intensificator de imagine, ct i cu imagine termal, ns ele sunt n general destinate utilizrii
pentru scopuri speciale.
Toate indiciile duc ctre ideea c viitorul va aparine aparaturii de ochire video. O
camer video montat pe puca de asalt va transmite o imagine a intei pe un display ataat
ctii trgtorului (suprapus cu viziera). Acest lucru va avea avantajul de a permite
utilizatorului s trag cu precizie din locuri acoperite, expunnd focului advers numai arma i
camera video montat pe aceasta; de asemenea, va avea posibilitatea de a regla mrirea
imaginii n funcie de situaia dat i capacitatea de vedere pe timp de noapte.
10. Sporirea fiabilitii
Fiabilitatea este unul dintre cele mai importante elemente ale unei puti militare i s-a
dovedit, de-a lungul timpului, surprinztor de greu de atins. Condiiile de antrenament
efectuarea tragerilor de la o distan prestabilit, cu timp suficient la dispoziie pentru
curarea ulterioar a armei nu se ntlnes de cele mai multe ori i n lupta real, iar
experienele recente ale aciunilor militare n deert (mai ales n zonele cu un nisip foarte fin)
au artat ct de repede poate o arm s devin inoperativ dac nu este ngrijit cu atenie.
Posibile cauze ale fiabilitii sczute, cum ar fi muniia care d rateu sau ncrctoarele
defecte, nu sunt dependente de arma n sine, dar altele da.
Caracteristicile unei arme care s asigure o fiabilitate ct mai ridicat sunt: un
mecanism motor de putere mare, care s poat asigura funcionarea chiar i atunci cnd este
357
AFASES - 2008 -
influenat negativ de praf i mizerie; un nchiztor greu cu inerie mare, care ajut de
asemenea la nvingerea frecrii; tolerane mari ntre componentele dinamice ale
mecanismelor, pentru ca mizeria depus ntre ele s nu mreasc peste limite forele de
frecare; un dizain care s previn ptrunderea prafului i a altor impuriti n mecanismele
armei, pe ct de mult posibil, dar s faciliteze ndeprtarea acestuia, dac totu se depune; o
construcie solid/rigid care s permit aciuni mecanice dure fr afectarea funcionrii
armei.
Aceste caracteristici sunt deinute n mare msur de puca de asalt Kalanikov AK-47
i de derivatele ei, lucru care a dus la o reputaie legendar pentru meninerea capabilitii de
funcionare aproape indiferent de gradul de neglijen sau abuz al utilizatorului. Dezavantajul
este o greutate mai mare a armei i o precizie ceva mai redus dect aproape toate putile
vestice de calibru 5,56 mm. Trebuie avut ns n vedere c i o arm care ar putea fi, teoretic,
mai puin precis i mai grea, este mai bun dect o arm care nu funcioneaz deloc.
Atingerea unei fiabiliti ridicate rmne un deziderat pentru toi constructorii de
armament moderni.
11. Standardizarea
Cel puin n cadrul N.A.T.O., standardizarea reprezint un el, ncepnd cu concepia
luptei i terminnd cu sprijinul logistic.
Sistemele arm-muniie utilizate nu puteau face excepie. Domeniile de standardizare
n cazul acestora sunt n primul rnd muniia (calibru, tipul pulberii, greutatea, caracteristicile
balistice), dimensiunile camerei de ncrcare, aparatele de ochire, ncrctoarele, suportul de
prindere a unor dispozitive auxiliare.
Problemele nu sunt nc definitivate n acest domeniu, existnd mai multe grupuri de
lucru N.A.T.O., n cadrul crora experii ncearc s gseasc rspunsurile optime la
ntrebrile pe care standardizarea le ridic.
Concluzii
Evoluia putii de asalt s-a nregistrat de-a lungul timpului pe foarte multe direcii,
genernd multe soluii constructive pe ntreaga planet. Exist o permanent ncercare de
armonizare a inovaiilor cu costurile, a performanelor cu fiabilitatea. Cu toate noutile aduse
de la muniii la materiale constructive, de la aparatura de ochire la construcia modular,
principalele caracteristici ale unei arme de asalt performante rmn simplitatea constructiv,
fiabilitatea ct mai ridicat cu putin i, nu n ultimul rnd, performanele balistice. Dac
viitorul ne va rezerva nc pentru mult vreme arme bazate pe soluiile constructive de astzi
sau vom avea parte de o schimbare radical, asemntoare filmelor SF, rmne de vzut.
Bibliografie
[1] Anthony G. Williams, Assault Rifles and Their Ammunition: History and Prospects
[2] Maxim Popenker & Anthony G. Williams, Assault Rifle: The Development of The
Modern Military Rifle and its Ammunition
[3] Urban Warfare n Complete Guide by Armada, nr. 4/2006
[4] www.quarry.nildram.co.uk/AssaultRifle.htm
[5] world.guns.ru
[6] en.wikipedia.org
*
Lt. ing. drd. Piticari Ioan-Liviu, Academia Forelor Terestre Nicolae Blcescu Sibiu, e-mail:
pitymika@gmail.com
**
Cpt. ing. drd. Barbu Mdlin, Centrul 183 Mentenan Tehnic Special, Sibiu, e-mail:
madalinbarbu@yahoo.com
358
AFASES - 2008 -
Pentru realizarea pieselor turnate, aliajul lichid este introdus ntr-o cavitate care reproduce
forma i cu anumite condiii dimensiunile pieselor ce trebuiesc obinute.
Suprafaa exterioar a cavitii formei se realizeaz cu un element tehnologic numit
modelul piesei. Dei cu ajutorul modelului se obine configuraia i dimensiunile piesei
turnate, aceste modele au totui dimensiunile i uneori forme diferite fa de viitoarea pies
turnat.
1.1 Materiale utilizate la confecionarea modelelor:
Modele din lemn
Folosirea pe scar larg a lemnului ca material pentru modele se datoreaz faptului c:
se procur uor, este ieftin, are densitate mic i se prelucreaz uor manual sau mecanic.
Modele din ipsos i ciment
Ca material ipsosul i cimentul se ntlnesc mai rar n turntorie, fiind folosite n special
pentru turnarea modelelor sau cutiilor de miez din aliaje metalice sau rini sintetice.
359
AFASES - 2008 -
Modelele din ciment ca i cele din ipsos sunt mai groase, mai grele i mai putin rezistente n comparaie
cu modelele realizate din alte materiale i de aceea utilizarea lor este limitat numai la producia de serie
mic.
Modele metalice
Garniturile de model metalice sunt utilizate frecvent n cazul produciilor de serie mare
i de mas, cnd formarea este n mare parte mecanizat sau automatizat.
n acest caz complexitatea elementelor de baz ct i a celor auxiliare ale garniturii de
model este mult mai mare. Sunt utilizate : oelul, fonta, aliajele de cupru i mai ales
aliajele din aluminiu.
Avantajul principal este c au o rezisten la uzur mare dar ca i dezavantaj faptul c
sunt scumpe.
Modele din mase plastice (rini sintetice) simple sau armate
Pot nlocui aluminiul i alte metale care se prelucreaz greu
Prezint rezisten la uzur n anumite conditii mai mare dect modelele metalice
Modelele au o greutate redus n comparaie cu cele metalice
Modele pierdute
360
AFASES - 2008 -
Fixarea modelului prin intermediul unor bosaje lipite n interior figura 2.3
Lipirea modelului pe o plac din aceleai material (sau altul ce permite
lipirea) figura 2.4
361
AFASES - 2008 -
n cazul modelelor mari se pot aduga suplimentar nervuri de ntrire (confecionate din
Lipirea
modelului
pe placa
portmodel
acelai material ca i modelul)Fig.
fr2.4
a :afecta
greutatea
total
a modelului
in mod semnificativ
figura 2.5
Bibliografie
[1] Gh. Simionescu, C-tin Cernat, M. Pan : Ingineria Proceselor de Formare
Editura Elvarom Bucuresti, 2000
[2] tefnescu Florin, .a : Materiale compozite
Editura Didactic i Pedagogic Bucureti 1996
[3] Aurel Iacobescu : Materiale si tehnologii moderne folosite in constructia structurii de
rezistenta a motoarelor racheta cu combustibil solid
http://www.actrus.ro
[4] Pop M. A. : Metode noi de proiectare i execuie a garniturilor de model.
Materiale noi utilizabile. Contribuii originale
Universitatea Transilvania Braov
*
Pop, Mihai Alin, drd. ing, Facultatea: tiina i Ingineria Materialelor - Universitatea Transilvania Braov,
alinpma@yahoo.com
362
AFASES - 2008 -
363
AFASES - 2008 -
Nu s-au putut obine toate caracteristicile materialelor, acestea nefiind puse la dispoziie de
fabricant constituind secrete de serviciu, care d numai unele indicaii generale.
Pentru rinile poliesterice (nearmate) caracteristicile principale sunt date n tabelul 1.1
Tabelul 1.1
Tipul rinii
Proprietatea
Proprietti mecanice:
- greutate specific
- rezisten la compresiune
- alungire
- rezisten la traciune
- rezisten la oc
- indice de refracie
Proprieti electrice:
- rezistivitate volumic
- constan dielectric la:
- 60 Hz
- 103 Hz
- 106 Hz
U.M.
Rigid
Flexibil
MPa
%
MPa
J/m
1,10 - 1,46
90 - 200
<5
40 - 90
16 - 32
1,523 - 1,57
1,01 - 1,2
40-310
3,4 - 21
> 370
1,537 - 1,55
cm
1015
Caracteristici de rezisten:
- rezisten termic
C
- absorbia de ap (n 24 h ) %
la o grosime de 3,1 mm
- efectul luminii solare
3,0 - 4,36
2,8 - 5,2
2,8 - 4,1
4,4 - 8,1
4,5 - 7,1
4,1 - 5,9
121
0,15 - 0,6
121
0,5 - 2,5
nglbenire
nglbenire
S-au fcut ncercari prin variaia numrului de straturi de armatur i rin astfel :
- rin Norsodyne cu armtur tip Stratimat (3, 5, 7 straturi de armtur) figura 1.1
- rin Sirca cu armtur tip estur (3, 5, 7 straturi de armtur) figura 1.2
364
AFASES - 2008 -
UM
N/mm2
N/mm2
KCU
HB
3
238
157
17/2
41
5
293
163
18/2
45
Tabelul 1.2
7
215
159
17/2
44
Rm 50
3
Ri
KCU 0
0
HB 3
Rezistenta la tractiune
Rm [N/mm2]
Rezistenta la incovoiere
Ri [N/mm2]
Rezistenta la soc
[KCU]
Duritate [HB]
0
25
0
20
0
15
0
10
50
0
Numr straturi
365
AFASES - 2008 -
Aceste probe fizico-tehnice au fost efectuate pentru toate tipurile de composite realizate.
Toate ncercrile s-au fcut pe epruvete conform STAS-urilor aflate n vigoare.
Din acest material compozit s-au confecionat 2 (dou) modele cave care corespund
cerinelor formrii.
Bibliografie
Pop, Mihai Alin, drd. ing, Facultatea: tiina i Ingineria Materialelor - Universitatea Transilvania Braov,
alinpma@yahoo.com
366
AFASES - 2008 -
367
AFASES - 2008 -
(6)
J = J ( x, y )e z
(7)
unde ez semnific vectorul unitate pe direcia axei z. n consecin (2.5) devine>
(A) = J
(8)
1.2 Generatorul de cmp
Cu toate c este utilizat analiza bidimensional, s lum n considerare un caz
general. Densitatea de curent membrul drept al ecuaiei (5) poate fi determinat din ecuaia
materialului:
J = E
(9)
unde este conductivitatea. Combinnd (2.2) cu (2.4) se obine:
(10)
xA
t
care este satisfcut prin definirea densitii de curent astfel:
A
J =
(11)
t
unde este potenialul scalar electric.
Pentru conductori solizi pe direcia axial, de exemplu blocul rotor din carcasa mainii
de inducie sau a mainii sincrone, gradientul potenialului scalar electric n (11) poate fi
definit ca
u
u
= b e z = b e z
(12)
z
lb
Unde ub semnific tensiunea de-a lungul seciunii rotorului i lb este lungimea total a
blocului rotor. Conform acesteia, densitatea curentului n blocul rotor este:
u
A
(13)
+ b
J b =
t
lb
xE =
Pentru nfurrile de faz cu cteva bobine n serie, abordarea de tip curent turbionar
(11) ne-ar putea conduce spre modele extrem de inconveniente din cauza numrului mare de
conductori subiri din fiecare fant. n astfel de cazuri efectul de pelicul este exclus i
densitatea de curent este:
N i
(14)
JW = w w
Sw
unde NW este numrul de spire n bobin, iw este curentul n bobin i Sw este seciunea
transversal a suprafeei bobinei.
368
AFASES - 2008 -
care este o funcie nonlinear cu o singur valoare a densitii de flux B, ceea ce determin
excluderea histerezisului magnetic din analiz. ntruct curenii turbionari sunt redui foarte
mult de structura laminat, conductivitatea este stabilit la zero n miezul de fier laminat.
Axul i polii n form de pantof (shaft and pole shoes), care de obicei sunt fcute aliaj
de oel, sunt modelate ca o bucat de fier conductibil cu o curb de magnetizare nonlinear.
Pentru a modela curenii nvrtitori, densitatea de curent este dat de (2.11) unde gradientul
potenialului scalar electric este zero. Ca o consecin a analizei anterioare, cmpul magnetic
n diferite materiale poate fi prezentat n forma urmtoare
(15)
t
l
b
Sb
Sb
Cnd conductivitatea constant i aria seciunii uniform transversale Sb sunt considerate c
exist n seciune i inductana seciunii terminale Lbe este inclus, ecuaia de mai sus poate fi
prezentat astfel:
di
A
u b = lb dS + Rb ib + Lbe b
(17)
t
dt
Sb
unde Rb semnific rezistena seciunii incluse n zona terminal. Toate seciunile rotorului
sunt conectate printr-un inel n scurt circuit legat la ambele capete ale miezului rotoric.
Aceasta este luat n considerare prin definirea rezistenei la inelul de capt Rsc i a
inductanei la inelul de capt Lsc.
di
u SC = RSC i SC + LSC SC
(18)
dt
Unde usc and isc sunt vectori de tensiune i curent n inelului de capt care conecteaz
seciunile (barele) una cu alta.
369
AFASES - 2008 -
nfurrile de faz din rotor sau stator constau din mai multe bobine conectate n
serie i distribuite n cteva fante n miezul rotorului sau statorului. Cnd numrul prilor
bobinelor orientate pozitiv este Npos i numrul prilor bobinelor orientate negativ este Nneg,
integrarea densitii de curent prin toate poriunile bobinei ntr-o nfurare de faz genereaz
o ecuaie de tensiune:
Neg
Npoz N
N
di
A
A
(19)
u w = l w wn
dS wn
dS + RW iW + Lwe w
dt
n =1 S wn S wn t
n =1 S wn S wn t
Unde lw este lungimea bobinei n zona miezului, Nwn este numrul de spire ntr-o
poriune a bobinei n i Swn este zona seciunii transversale a unei pri a bobinei n. Tensiune
uw se aplic la ntreaga nfurare i curentul iw trece prin toate bobinele care aparin
nfurrilor de faz.
Rezistena Rw include toate bobinele i zonele terminale din afara miezului de fier.
Lwe este inductana din afara zonei miezului.
A = ja j
(20)
j =1
Unde Nn este numrul total de noduri libere n suprafaa elementului finit, j este
funcia de reprezentare a formei (shape function) asociat cu nodul j i aj este vectorul
potenial magnetic n nodul j. Pentru nodurile localizate pe suprafaa exterioar a statorului,
valoarea vectorului potenial este fixat la zero. De aceea aceste noduri nu sunt incluse n
calcule.
Funcia de reprezentare a formei (shape function) j are valoarea 1 n nodul j i zero
n toate celelalte noduri. ntre nodurile dispuse n elementele nconjurtoare, funcia de
reprezentare a formei este ajustat de ctre polinoamele de ordinul unu sau doi, depinznd de
ordinea elementelor.
n afara elementelor asociate, valoarea unei singure funcii de reprezentare a formei
este zero. Ecuaia cmpului numeric este derivat de ctre metoda lui Galerkin, unde (15) este
multiplicat de ctre funcia de reprezentare a formei i integrat n toate suprafeele elementului
finit. Pentru nodul i , obinem:
Nn
a
N i
i j a j + i j j i u b i w w d = 0
(21)
S w
t
j =1
Unde reprezint ntreaga arie a suprafeei elementului finit. Trebuie observat c,
oricum, ultimii trei termeni ai formulei (21) sunt prezeni numai n zonele unde , ub sau iw
este definit i are valori diferite de zero.
Aceeai aproximare (20) este aplicat ecuaiilor nfurrii (19) i (20). Ecuaiile
rezultate sunt:
370
AFASES - 2008 -
a j
Nn
di
u b = l b j
dS + Rb ib + Lbe b
dt
t
j =1
(22)
N neg
N n N pos
a j
a j
N
N
di
u w = l w wn
dS wn
dS + Rw i w + Lwe w
dt
j =1 n =1 S wn S wn t
n =1 S wn S wn t
(23)
(25)
rb = Fb a k + Bu bk Fb a k 1 + Bu bk 1 + Qb 0 ibk 1 = 0
(26)
T k
k
T k 1
k 1
rws = K ws Fws a k + K wsWs K ws
i ws + Qws ws
K ws Fws a k 1 + K wsWs 0 K ws
i ws + Qws ws
= 0 (27)
T k
k
T k 1
k 1
rwr = K wr Fwr a k + K wrWr K wr
iwr + Qwr wr
K wr Fwr a k 1 + K wrWr 0 K wr
i wr + Qwr wr
= 0 (28)
Unde a este un vector coninnd valorile nodale ale vectorului potenial magnetic, ub
un vector al tensiunilor electrice din blocul rotor, iwr un vector al curenilor de faz statorici i
iwr un vector al curenilor de faz rotorici. n general, indicele f se refer la cmpul magnetic,
b la blocul rotor, ws la nfurrile de faz statorice i wr la nfurrile de faz rotorice.
Exponenii k i k-1 se refer la curent i respective la treptele anterioare de timp. Matricele
coeficienilor S i So sunt funcii nonlineare ale vectorului potenial, definite astfel:
2
S ij = (a k )i j +
i j d
(29)
S 0ij = (a k 1 )i j
i j d
(30)
t
371
AFASES - 2008 -
unde reluctana (a) este determinat de vectorul potential separat pentru fiecare element,
depinznd de material i de densitatea fluxului. Conductivitatea are valoarea diferit de zero
numai in blocul rotor i n ax. Dimensiunile lui S i S0 sunt Nn X Nn, unde Nn este numrul
noduri libere n analiza elementului finit. Pentru blocul rotor, matricea Fb este definit ca:
1
Fb , ij = i d
(31)
lb
Unde nodul j aparine blocului rotor i. Altfel, intrrile Fb sunt zero. Dimensiunea lui
Fb
este nb x Nn, unde nb este numrul de bare rotorice n sectorul soluiilor. Matricele B i Qb0
sunt:
1
LSC
Lbe
t Rb
+ M b (32)
B=
1
R
2
1
R
2
M
M
+
+
+
+
be
SC
b
b
2lb Rb
2
t
t
Qb 0 =
L
t
t
1
RSC + 2 SC
2lb
2l b
t
L
1 + Rbe + 2 be
t
L
M b RSC 2 SC
t
L
1 + Rbe 2 be
t
M b
(33)
unde Rsc i Lsc sunt rezistena i inductana ntre barele rotorice adiacente din inelul n
scurtcircuit i, respectiv, Rbe i Lbe sunt rezistena i inductana captului de bar care se
gsete la exteriorul miezului de fier. Matricea unitate este reprezentat de 1 matricea de
conexiune Mb este:
1
0 ... 0 1
2
1
- 1 1... 0 0
0
-1
2... 0 0
.............................
Mb =
.............................
.............................
0
0
0 ...
2 -1
0
0 ...
-1 2
1
(34)
cu dimensiunea nbX nb. Intrrile n colul din dreapta sus i colul din stnga jos sunt 1 pentru
constrngerile de simetrie negative i -1 pentru constrngerile de simetrie pozitive n suprafaa
elementului finit. Curentul barei rotorice ik b este determinat dup soluia cmpului din:
1
L
L 1
L
L
i = RSC + 2 SC 1 + Rbe + 2 be M b M b ubk + ubk 1 + RSC 2 SC 1 + Rbe + 2 be M b ibk 1
t
t 2
t
t
k
b
(35)
Pentru nfurrile de faz n stator sau rotor Fw este definit ca:
Fw , ij = ji d
(36)
372
AFASES - 2008 -
1
0
0
Kw =
:
.
0 0...0 - 1
1 0... 0 - 1
0 1...0 - 1
: : : : :
. . . . .
0 0...1 - 1
(37)
i Mw este:
1 1 1... 1 0
0 1 1... 1 0
0 0 1... 1 0
Kw =
(38)
: : : : :
. . . . .
0 0 0... 1 0
cu dimensiunea nw -1 nw pentru amndou. n conexiunea delta, ambele Kw i Mw sunt
matrice unitate cu dimensiunea nw nw.
Matricile W, W0 i Qw sunt:
Rw,i t + 2 Lwe.i
(39)
Wii =
2l w
Rw,i t 2 Lwe.i
(40)
W0,ii =
2l w
t
Qw =
Mw
(41)
2l w
W i W0 sunt matricele diagonale cu dimensiunile nw nw.
Vectorul legturilor fluxului magnetic k n nfurrile de faz poate fi determinat din
soluiile de cmp prin:
wk = l w Fw a k
(42)
Trebuie notat c determinarea curenilor depinde de conexiune. n conexiunea tip stea,
vectorul curent iw conine nw-1 cureni de linie independeni i ultimul component este
determinat din legile lui Kirchhoff, deoarece suma curenilor trebuie s fie zero. n
conexiunea tip delta, elementele lui iw sunt curenii n nfurrile de faz, care sunt diferite
de curenii de linie. Elementele vectorului de tensiune electric vw sunt ntotdeauna tensiuni
electrice de tip linie -linie.
2.2 Soluii iterative nonliniare
Datorit caracteristicilor miezului de fier, ecuaiile (25)(28) formeaz un sistem
nonlinear de ecuaii care este rezolvat prin abordarea iterativ NewtonRaphson. Soluia este
bazat pe Jacobian i pe reziduuri, care sunt determinate de fiecare pas al iteraiei i folosite
pentru rezolvarea schimbrilor incrementale n variabile. Iteraia este ncheiat cnd creterile
n variabilele simulate la fiecare pas al iteraiei ncep s scad sub limita de convergen.
Rezultnd din etapele descrise anterior, sistemul final de ecuaii pentru simularea
nonlinear ealonat n timp a mainii electrice este:
373
AFASES - 2008 -
( )
J a k ,n
Fb
K F
ws ws
K wr Fwr
FbT
T
FwsT K ws
T
FwrT K wr
T
K wsWs K ws
0
0
0
T
K wrWr K wr
k ,n
r fn
a
n
k ,n
u b = rb
n
i ' k ,n
ws
rws
n
i ' k ,n
wr
rwr
(43)
unde exponentul n exprim pasul iteraiei. Jacobianul J(a) este definit la fiecare pas al
iteraiei prin:
( )
Nn
a k , n
J ij = Sij +
i k d
a
j
k =1
(44)
dm
= Te TL
dt
m =
(45)
d m
dt
(46)
unde J este momentul de inerie, m este viteza unghiular i m este poziia unghiular a
rotorului, Te este cuplul electromagnetic i TL este cuplul de sarcin.
Noua poziie a rotorului este determinat de nceputul fiecrei perioade de timp i
suprafaa nou este creat n golul de aer. (air gap).
Cuplul electromagnetic este determinat de principiul lucrului virtual:
Te =
H B dH d
m 0
(47)
unde zona de integrare acoper golul de aer. Implementarea analizei elementului finit
decurge din abordarea lui Coulomb (1983), n care micarea virtual este determinat cu
ajutorul unei matrice de transformare a coordonatelor fr modificarea suprafeei golului de
aer.
2.4 Starea iniial
Cmpul magnetic iniial pentru simularea distribuit n timp este determinat de o
analiz n timp armonic complex. Sursele i variabilele simulrii sunt determinate de
componentele armonice sinusoidale, n timp ce pentru variabile se folosete o notaie
complex cu amplitudine i faz. Soluia rezultat a cmpului de curent alternativ este
transformat n cmp de curent continuu prin luarea valorilor reale ale variabilelor i
nlocuirea reluctanei efective cu reluctana absolut. Cu toate acestea, soluia cmpului
374
AFASES - 2008 -
obinut nu corespunde exact cu cmpul real iniial, dar o situaie constant poate fi atins n
mod uzual dup simularea ctorva perioade al frecvenei sursei.
Toate informaiile despre suprafaa elementului finit, materialele i cmpul magnetic
sunt stocate ntr-un fiier text. Fiierul conine coordonatele nodurilor, asocierea dintre noduri
i elemente, caracteristicile materialului, valorile vectorului potenial i densitatea de curent,
dimensiunile i parametrii mainilor electrice i ali parametri cerui de calculaia MEF.
2.5 Concluzii
n aceast lucrare au fost prezentate metodele de calcul pentru modelarea i simularea
mainilor electrice i a circuitelor externe. Cmpul magnetic n maina electric este modelat
cu ajutorul MEF bidimensionale i rezolvate mpreun cu ecuaiile de circuit ale nfurrilor.
Modelul circuitului extern este cuplat cu modelul mainii electrice fie direct, cnd toate
ecuaiile sunt rezolvate simultan ntr-un sistem comun de ecuaii, fie indirect n SIMULINK,
folosind abordarea ieirii de curent sau abordarea parametrului circuitului.
n abordarea ieirii de curent, curenii de faz ai mainii electrice sunt rezolvai
mpreun cu cmpul magnetic, cnd sursa de tensiune este dat de intrri. n abordarea
parametrului circuitului, maina electric este caracterizat de fora electromotoare,
inductana dinamic, i rezisten, care sunt rezolvate de MEF la fiecare interval de timp
pentru o surs de tensiune dat. Ambele abordri sunt implementate ca funcii-S reprezentnd
blocuri funcionale n SIMULINK.
n urmtoarele capitole, metodele sunt aplicate la diferite studii de caz referitoare la mainile
electrice, circuite i sisteme de control. Rezultatele simulrii obinute prin diferite metode
sunt comparate unele cu altele pentru a trage concluziile referitoare la precizia i
aplicabilitatea lor n diferite cazuri.
BIBLIOGRAFIE
1. Abe, N.M. and Cardoso, J.R. 1998: Coupling Electric Circuit and 2D-FEM ModelWith
Dommels Approach for Transient Analysis, IEEE Transactions on Magnetics, Vol. 34, No.
5,
September 1998, pp. 34873490.
2. Ahn, S.C., Lee, J.H., and Hyun, D.S. 2000: Dynamic Characteristic Analysis of LIM
Using
Coupled FEM and Control Algorithm, IEEE Transactions on Magnetics, Vol. 36, No. 4, July
2000, pp. 18761880.
3. Albanese, R. and Rubinacci, G. 1992: Numerical Procedures for the Solution of Nonlinear
Electromagnetic Problems, IEEE Transactions on Magnetics, Vol. 28, No. 2, March 1992,
4. Ansoft 2005: http://www.ansoft.com/ (5.2.2005).
5. Ansys 2005: http://www.ansys.com/ (5.2.2005).
6. Arkkio, A. 1987: Analysis of Induction Motors Based on the Numerical Solution of
the Magnetic Field and circuit Equations, Electrical Engineering Series, No. 59, Acta
Polytechnica Scandinavica, Helsinki, 97 p., available: http://lib.hut.fi/Diss/198X/
isbn951226076X/ (5.2.2005).
7.Arkkio, A. 1990: Finite Element Analysis if Cage Induction Motors Fed by Static
Frequency
Converters, IEEE Transactions on Magnetics, Vol. 26, No. 2, March 1990, pp. 551554.
8. Bedrosian, G. 1993: A NewMethod for Coupling Finite Element Field Solutions with
External
375
AFASES - 2008 -
Circuits and Kinematics, IEEE Transactions on Magnetics, Vol. 29, No. 2, March 1993, pp.
16641668.
9. Cervera, M., Codina, R., and Galindo, M. 1996: On the Computational Efficiency and
Implementation of Block-Iterative Algorithms for Nonlinear Coupled Problems, Engineering
Computations, Vol. 13, No. 6, June 1996, pp. 430.
10.Charpentier, J.F., Lefvre, Y., and Piquet, H. 1998: An Original and Natural Method of
Coupling Electromagnetic Field Equations with Circuit Equations Put in a State Form, IEEE
Transactions on Magnetics, Vol. 34, No. 5, September 1998, pp. 24892492.
11.Comsol 2005: http://www.comsol.com/ (5.2.2005).
Costa, M.C., Nabeta, S.I., and Cardoso, J.R. 2000: Modified Nodal Analysis Applied to
Electric Circuits Coupled with FEM in the Simulation of a Universal Motor, IEEE
Transactions
on Magnetics, Vol. 36, No. 4, July 2000, pp. 14311434.
12.Coulomb, J.L. 1983: A Methodology for the Determination of Global Electromechanical
Quantities from a Finite Element Analysis and Its Application to the Evaluation of Magnetic
Forces, Torques and Stiffness, IEEE Transactions on Magnetics, Vol. MAG-19, No. 6,
November 1983, pp. 25142519.
13. De Gersem, H., Mertens, R., Lahaye, D., Vandewalle, S., and Hameyer, K. 2000:
Solution
Strategies for Transient, Field-Circuit Coupled Systems, IEEE Transactions on Magnetics,
Vol. 36, No. 4, July 2000, pp. 15311534.
14. Demenko, A. 1994: Time-Stepping FE Analysis of Electric Motor Drives with
Semiconductor
Converters, IEEE Transactions onMagnetics, Vol. 30, No. 5, September 1994, pp. 3264
3267.
Demerdash, N. and Nehl, T. 1999: Electric Machinery Parameters and Torques by Current
and Energy Perturbations from Field Computations Part I: Theory and Formulation, IEEE
Transactions on Energy Conversion, Vol. 14, No. 4, December 1999, pp. 15071513.
15. Depenbrock, M. 1988: Direct Self-Control (DSC) of Inverter-Fed Induction Machine,
IEEE
Transactions on Power Electronics, Vol. 3, No. 4, October 1988, pp. 420429.
16. Fujiwara, K., Nakata, T., Okamoto, N., and Muramatsu, K. 1993: Method for
Determining
Relaxation Factor for Modified Newton-Raphson Method, IEEE Transactions on Magnetics,
Vol. 29, No. 2, March 1993, pp. 19621965.
17. Gear, C.W. 1971: Numerical Initial Value Problems in Ordinary Differential Equations,
Prentice-Hall, Englewood Cliffs, New Jersey.
18. Kanerva, S. 2001: Implementation of Two-Dimensional Finite Element Method in
System
Simulator, Report 67, Helsinki University of Technology, Laboratory of Electromechanics,
19. Williamson, S., Lim, L.H., and Robinson, M.J. 1990: Finite-Element Models for Cage
Induction Motor Analysis, IEEE Transactions on Industy Applications, Vol. 26, No. 6,
November/
December 1990, pp. 10071017.
376
AFASES - 2008 -
377
AFASES - 2008 -
378
AFASES - 2008 -
379
AFASES - 2008 -
= 1
8 M (1 s )(1 t )(1 + r ) + N (1 + s )(1 t )(1 + r ) + O (1 + s )(1 + t )(1 + r ) + P (1 s )(1 + t )(1 + r )
(1)
1 1 1
f ( x, y, z )dxdydz = H k H j H i f (xi , y j , z k );
n
(2 )
k =1 j =1 i =1
Unde:
f(z y x ) funcia ce trebuie integrat;
Hk, Hj, Hi factor de greutate;
xi , y j , z k locaii pentru evaluarea funciei (aceste locaii sunt n mod uzual coordonatele s, t,
r );
n, m, l numerele punctelor de integrare (Fig. 6).
Ecuaiile matricei elementului finit pot fi derivate prin principii variaionale. Aceste
ecuaii exist pentru comportamente materiale liniare i nonliniare precum i pentru
rspunsul tranzitoriu i static. Bazndu-ne pe prezena materialelor nonliniare i liniare,
programul alege cea mai potrivit metod Newton- Raphson. Calculaia potenialului scalar
este restrictiv pentru analiza cmpului static cu permeabilitate ortotropic parial.
[K ] = [K ][K ] ;
[K ] = N [ ] N
M
vol
[K ] = H H
N
vol
(3)
d (vol ) ;
(4)
T
T
T d (vol )
N H N
;
H
(5)
380
AFASES - 2008 -
[J i ] = N
vol
[ ] H + H d (vol ) ;
g
c
(6)
Unde
N funciile de form ale elementului;
T
= N e potenialele scalare magnetice din nodurile elementului;
T operatorul gradient,
vol volumul elementului;
T preliminarul cmpului magnetic presupus;
H c vectorul for coercitiv,
[] matricea permeabilitii,
d h / d H derivata permeabilitii legat de magnitudinea intensitii cmpului magnetic.
3. Modelarea cmpului magnetic
Rezultatele cmpului electromagnetic de baz includ: intensitatea cmpului magnetic,
densitatea fluxului magnetic, forele magnetice i densitile de curent.
Prima derivat are ca rezultat intensitatea cmpului magnetic care se divide n dou
pri: un cmp generalizat H g i gradientul potenialului generalizat g .
Primul rezultat derivat (n cazul nostru H ) este evaluat la punctele de integrare (Fig.
6) folosind funcia de form ale elementului astfel:
T
H = N g ;
(7)
Unde:
T operatorul gradient ;
N funciile de form ale elementului;
g Vectorul nodal potenial generalizat.
Intensitatea cmpului magnetic este apoi:
H = H g + H
(8)
Valorile nodale ale intensitii cmpului i densitii fluxului sunt calculate din valoarea
punctelor integrate.
381
AFASES - 2008 -
Se pot prezenta mai multe exemple de cmp magnetic creat n bobin cu miez.
Modelul constructiv este format folosind un algoritm creat i fiiere de comand deja
create.(pre-made command files) .
Parametrii de construcie a modelului sunt artai n figura 7.
Numrul de spire al bobinei n=500, curentul prin spir i=6.
Valorile i distribuia intensitii cmpului magnetic calculat sunt artate n figura 8.
4. Concluzii
382
AFASES - 2008 -
383
AFASES - 2008 -
Abstract
The global approach to NOx emission controls was taken by the IMO whose Annex VI
(Regulations for the Prevention of Air Pollution from Ships) to Marpol 73/78 was adopted at
a diplomatic conference in 1997. Ships burning marine diesel oil and heavy fuel oil at that
time were reportedly responsible for around 7 per cent of global NOx emissions, around 4 per
cent of global sulphur dioxide emissions and 2 per cent of global carbon dioxide emissions
I. INTRODUCTION
Marine engine designers in recent years have had to address the challenge of
tightening controls on noxious exhaust gas emissions imposed by regional, national and
international authorities responding to concern over atmospheric pollution. Exhaust gas
emissions from marine diesel engines largely comprise nitrogen, oxygen, carbon dioxide and
water vapour, with smaller quantities of carbon monoxide, oxides of sulphur and nitrogen,
partially reacted and non-combusted hydrocarbons and particulate material (Figures 1 and 2).
Figure 1 Typical exhaust emissions from a modern low speed diesel engine
384
AFASES - 2008 -
385
AFASES - 2008 -
Figure 2 Typical composition of the exhaust gas products of a medium speed diesel engine
burning fuel with an average 3 per cent sulphur content. Some 6 per cent of the total
emission is carbon dioxide, with the real pollutants representing only a 0.3 per cent share
(MAN B&W Diesel)
Sulphur oxides (SOx)produced by oxidation of the sulphur in the fuelhave an
unpleasant odour, irritate the mucus membrane and are a major source of acid rain (reacting
with water to form sulphurous acid). Acid deposition is a trans-boundary pollution problem:
once emitted, SOx can be carried over hundreds of miles in the atmosphere before being
deposited in lakes and streams, reducing their alkalinity.
Sulphur deposition can also lead to increased sulphate levels in soils, fostering the
formation of insoluble aluminium phosphates which can cause a phosphorous deficiency.
Groundwater acidification has been observed in many areas of Europe; this can lead to
corrosion of drinking water supply systems and health hazards due to dissolved metals in
those systems. Forest soils can also become contaminated with higher than normal levels of
toxic metals, and historic buildings and monuments damaged. Hydrocarbons (HC)created
by the incomplete combustion of fuel and lube oil, and the evaporation of fuelhave an
unpleasant odour, are partially carcinogenic, smog forming and irritate the mucus membrane
(emissions, however, are typically low for modern diesel engines.) Carbon monoxide (CO)
resulting from incomplete combustion due to a local shortage of air and the dissociation of
carbon dioxide - is highly toxic but only in high concentrations. Particulate matter (PM) is a
complex mixture of inorganic and organic compounds resulting from incomplete combustion,
partly unburned lube oil, thermal splitting of HC from the fuel and lube oil, ash in the fuel and
lube oil, sulphates and water. More than half of the total particulate mass is soot (inorganic
carbonaceous particles), whose visible evidence is smoke. Soot particles (unburned elemental
carbon) are not themselves toxic but they can cause the build-up of aqueous hydrocarbons,
and some of them are believed to be carcinogens. Particulates constitute no more than around
0.003 per cent of the engine exhaust gases. Noxious emissions amount to 0.25-0.4 per cent by
volume of the exhaust gas, depending on the amount of sulphur in the fuel and its lower heat
386
AFASES - 2008 -
value, and the engine type, speed and efficiency. Some idea of the actual pollutants generated
is provided by MAN B&W Diesel, which cites an 18 V 48/60 medium speed engine in NOxoptimized form running at full load on a typical heavy fuel oil with 4 per cent sulphur content.
A total of approximately 460 kg of harmful compounds are emitted per hour out of around
136 tonnes of exhaust gas mass per hour. Of the 0.35 per cent of the exhaust gas formed by
pollutants, NOx contributes 0.17 per cent, sulphur dioxide 0.15 per cent, hydrocarbons 0.02
per cent, carbon monoxide 0.007 per cent and soot/ash 0.003 per cent. Carbon dioxide: some
6 per cent of the exhaust gas emissions from this engine is carbon dioxide. Although not itself
toxic, carbon dioxide contributes to the greenhouse effect (global warming) and hence to
changes in the Earths atmosphere. The gas is an inevitable product of combustion of all fossil
fuels, but emissions from diesel enginesthanks to their thermal efficiencyare the lowest
of all heat engines. A lower fuel consumption translates to reduced carbon dioxide emissions
since the amount produced is directly proportional to the volume of fuel used, and therefore to
the engine or plant efficiency. As a rough guide, burning one tonne of diesel fuel produces
approximately three tonnes of carbon dioxide. International concern over the atmospheric
effect of carbon dioxide has stimulated measures and plans to curb the growth of such
emissions, and the marine industry must be prepared for future legislation. (A switch from
other transport modesair, road and railto shipping would nevertheless yield a substantial
overall reduction in emissions of the greenhouse gas because of the higher efficiency of diesel
engines.)
The scope for improvement by raising the already high efficiency level of modern
diesel engines is limited and other routes have to be pursued: operating the engines at a fuelsaving service point; using marine diesel oil or gas oil instead of low sulphur heavy fuel oil;
adopting diesel-electric propulsion (the engines can be run continuously at the highest
efficiency); or exploiting a diesel combined cycle incorporating a steam turbine. The steaminjected diesel engine is also promising.
SOx emissions in diesel engine exhaust gaswhich mostly comprise sulphur dioxide
with a small amount of sulphur trioxide - are a function of the amount of sulphur in the fuel
and cannot be controlled by the combustion process. If the fuel contains 3 per cent sulphur,
for example, the volume of SOx generated is around 64 kg per tonne of fuel burned; if fuel
with a 1 per cent sulphur content is used SOx emissions amount to around 21 kg per tonne of
fuel burned. Chemical and washing/scrubbing desulphurization processes can remove SOx
from the exhaust gases but are complex, bulky and expensive for shipboard applications, and
increase overall maintenance costs. The most economical and simplest approach is thus to
burn bunkers with a low sulphur content. (If a selective catalytic reduction system is installed
to achieve the lowest NOx emission levels - see section below - then low sulphur fuels are
dictated anyway to avoid premature fouling of the systems catalyst package.) A global heavy
fuel oil sulphur content cap of 4.5 per cent and a 1.5 per cent fuel sulphur limit in certain
designated Sulphur Emissions Control Areas (SECAs)such as the Baltic Sea, North Sea and
English Channelis sought by the International Maritime Organisation (IMO) to reduce SOx
pollution at sea and in port. The European Union strategy for controlling air pollution calls for
all ships in EU ports to burn fuel with a maximum sulphur content of 0.2 per cent, which
would force uni-fuel ships to carry low sulphur fuel specifically for this purpose.
387
AFASES - 2008 -
Figure 3 Maximum allowable NOx emissions for marine diesel engines (IMO)
388
AFASES - 2008 -
Figure 4 NOx emission trends for typical two-stroke and four-stroke diesel engines
compared with IMO requirements (Wrtsil)
Dominating influences in the formation of NOx are temperature and oxygen
concentration: the higher the temperature and the higher the residence time at high
temperature in the cylinder, the greater the amount of thermal NOx that will be created. A
longer combustion timespan means that low speed two-stroke engines therefore generate
higher NOx emissions than medium and high speed four-stroke engines of equivalent output.
Apart from the use of alternative fuels, such as methanol, two main approaches can be
pursued in reducing NOx emission levels:
Primary (in-engine) measures aimed at reducing the amount of NOx formed during
combustion by optimizing engine parameters with respect to emissions (valve timing, fuel
injection and turbocharging). Emission levels can be reduced by 3060 per cent. Secondary
measures designed to remove NOx from the exhaust gas by downstream cleaning techniques.
Emission reductions of over 95 per cent can be achieved.
De-NOx technology options are summarized in Figure 5.
The primary NOx reduction measures can be categorised as follows: Water addition:
either by direct injection into the cylinder or by emulsified fuel. Altered fuel injection:
retarded injection; rate-modulated injection; and a NOx-optimized fuel spray pattern.
Combustion air treatment: Miller supercharging; turbocooling; intake air humidification;
exhaust gas recirculation; and selective non-catalytic reduction. Change of engine process:
compression ratio; and boost pressure. The basic aim of most of these measures is to lower
the maximum temperature in the cylinder since this result is inherently combined with a lower
NOx emission.
389
AFASES - 2008 -
Figure 5 Methods of reducing NOx emissions from marine diesel engines (Mitsubishi
Heavy Industries)
New generations of medium speed engines with longer strokes, higher compression
ratios and increased firing pressures addressed the NOx emission challenge (Figure 6). The
low NOx combustion system exploited by Wrtsil throughout its medium speed engine
programme, for example, is based on an optimized combination of compression ratio,
injection timing and injection rate. The engine parameters affecting the combustion process
are manipulated to secure a higher cylinder pressure by increasing the compression ratio. The
fuel injection equipment is optimized for late injection with a short and distinct injection
period. NOx reductions of up to 50 per cent are reported without compromising thermal
efficiency, achieving an NOx rate of 58 g/kWh compared with the 15 g/kWh of a typical
conventional engine with virtually no effect on fuel consumption.
Figure 6 Emission control measures and their potentials in reducing NOx output.
EGR = exhaust gas recirculation; SCR = selective catalytic reduction (Wrtsil Diesel)
390
AFASES - 2008 -
Figure 7 Wrtsils Direct Water Injection system features a combined injection valve and
nozzle for water and fuel oil
391
AFASES - 2008 -
Clean water is fed to the cylinder at a pressure of 210400 bar (depending on the
engine type) by a common rail system, the pressure generated in a high pressure pump
module; a low pressure pump is also necessary to ensure a sufficiently stable water flow to the
high pressure pump. The water is filtered before the low pressure pump to remove all solid
particles; the pumps and filters are built into modules to ease installation.
A flow fuse installed on the cylinder head side acts as a safety device, shutting off the
water flow into the cylinder if the water needle gets stuck; immediate water shut-off is
initiated in the event of excessive water flow or leakage. Water injection timing and duration
are electronically controlled by the control unit, which receives its input from the engine
output; timing and duration can be optimized conveniently via a keyboard for different
applications. Space requirements for the equipment are minimal, facilitating retrofit
applications.
Water and fuel are injected in a typical water-to-fuel ratio of 0.40.7:1, reducing NOx
emissions by 5060 per cent without adversely affecting power output or engine components.
NOx emissions are typically 46 g/kWh when the engine is running on marine diesel oil, and
57 g/kWh when heavy fuel oil is being burned. NOx reduction is most efficient from 40 per
cent load and higher of nominal engine output.
Numerous applications of the above DWI system have benefited Wrtsil 32 and 46
engines powering diverse ship types. Some W46 EnviroEngines (see section below)
combine electronically-controlled common rail fuel injection with DWI, the water then
injected into the combustion chamber separately from the side of the cylinder head through a
dedicated valve.
Mitsubishi Heavy Industries developed a stratified fuel-water injection (SFWI) system
using a common valve to inject slugs of fuel/water/fuel sequentially into the combustion
chamber. The system reportedly worked in a stable condition throughout extensive trials on a
ship powered by a Mitsubishi UEC 52/105D low speed engine, Nox emission reduction being
proportional to the amount of water injected.
IV. REFERENCES
1. Marian Ristea - Evoluia parametrilor geometrici i funcionali ai motoarelor cu ardere
intern n domeniul naval Raport de cercetare tiinific susinut n cadrul programului de
studii doctorale la Academia Tehnic Militar, Februarie 2008
2. *** - MAN B&W documentation
3. *** - DEUTX documentation
4. *** - PIELSTIK documentation
5. *** - Sulzer documentation
Ristea Marian, prep. univ. drd. ing, Academia Navala Mircea cel Batran, Constanta, Fulgerului, nr.1, e-mail:
ristea_k_marian@yahoo.com
**
Popa Ionel, sef lucr. univ. dr. ing, drd. ing, Academia Navala Mircea cel Batran, Constanta, Fulgerului, nr.1,
e-mail: popaionelac@yahoo.com
***
Popa Adrian, asist. univ. drd. ing drd. ing, Academia Navala Mircea cel Batran, Constanta, Fulgerului, nr.1,
392
AFASES - 2008 -
393
AFASES - 2008 -
394
AFASES - 2008 -
termic pentru mbuntirea rezistenei mecanice. Ele au o plasticitate foarte bun la orice tip
de deformare. [2, 5]
n funcie de compoziie i destinaie, aliajele titanului pot fi supuse diferitelor
tratamente termice i termochimice: recoaceri, cliri mbtrniri, cementri, nitrurri.
Utilizarea aliajelor de Ti este avantajoas ndeosebi la temperaturile de 400...600C, la
care aliajele de Al i Mg nu pot fi ntrebuinate datorit rezistenei lor sczute la fluaj, iar
oelurile aliate nu sunt indicate datorit greutii lor specifice mari. La utilizarea titanului i a
aliajelor pe baz de titan trebuie s se in seama i de unele inconveniente: prelucrabilitatea
prin achiere sczut, situat ntre prelucrabilitatea oelurilor pentru scule i aceea a
superaliajelor de nichel sau cobalt, tendina de deteriorare a pieselor supuse la frecare prin
alunecare (oboseal sau frecare), datorit coeficientului de frecare ridicat i conductivitii
termice sczute.
Se folosesc n construciile aeronautice (nveliul avioanelor, discurile, paletele i
bridele compresoarelor etc.), n tehnica rachetelor (corpurile motoarelor pentru a doua i a
treia treapt, butelii pentru gaze lichefiate i comprimate etc.), n industria chimic
(echipament pentru lucru n clor umed, n soluii apoase i acide de clor, schimbtoare de
cldur care lucreaz n acid sulfuric etc.), n construcii navale (elice propulsoare, nveli
pentru nave maritime, submarine i torpile), construcii energetice, fabricarea echipamentului
pentru tratamentul combustibilului nuclear etc. [6, 7] De asemenea, fiind foarte plastice la
temperaturi sczute, sunt utilizate n tehnica frigului.
Titanul i aliajele pe baz de titan constituie materiale ale viitorului, deoarece prezint
bune proprieti de turnare, bune proprieti plastice, bun rezisten chimic i bune
proprieti mecanice.
II. Aliaje grele
II.1. Aliaje cupru zinc (alame)
Alamele sunt aliaje ale cuprului cu Zn i conin cel puin 55% Cu. Ele pot fi obinuite,
adic simple aliaje binare cupru zinc, coninnd alte elemente numai sub form de
impuriti sau aliaje complexe, numite alame aliate care, alturi de cupru i zinc, conin i alte
elemente: Al, Sn, Mn, Fe, Ni, Pb etc.
Alamele cu un coninut pn la 39% Zn, la rcire lent sau n stare recoapt prezint
structur poliedric, iar dac rcirea este rapid, dup turnare, structura este dendritic.
Alamele monofazice sunt plastice, putnd fi deformate la rece. Dac procentul de Cu
depete 80% se numesc tombac; sunt foarte maleabile, ns se prelucreaz mai greu prin
achiere. Alamele cu un coninut mai mare de 39% Zn sunt dure i fragile i sunt prelucrabile
prin deformare plastic doar la cald.
n funcie de prelucrrile la care sunt supuse, att alamele obinuite, ct i cele aliate
se mpart n: deformabile i pentru turntorie.
II.2. Aliaje cupru staniu (bronzuri cu staniu)
Aliajele cuprului cu staniul se numesc bronzuri cu staniu sau bronzuri obinuite. Dac
n loc de Sn sau mpreun cu Sn mai particip i alte elemente (Al, Pb, Zn, Be, P, Ni), atunci
se formeaz bronzuri speciale sau aliate, care poart denumirea dup elementul de aliere
(bronz cu aluminiu, bronz cu plumb etc.).
Bronzurile cu Sn (binare i aliate), dup modul de prelucrare se mpart n deformabile
i de turntorie. [4, 7]
395
AFASES - 2008 -
396
AFASES - 2008 -
397
AFASES - 2008 -
Bibliografie
1. COLAN, H., .a. Studiul metalelor, Editura Didactic i Pedagogic, Bucureti, 1983.
2. DINESCU, I. Tehnologia materialelor, Procedee tehnologice, Editura Academiei Aviaiei
i Aprrii Antiaeriene, Braov, 1998.
3. DINESCU, I. Tehnologia materialelor, Materiale tehnologice, Editura Academiei Aviaiei
i Aprrii Antiaeriene, Braov, 1999.
4. DUMITRACHE, C. tiina materialelor, Editura Matrix Rom, Bucureti, 2002.
5. MITELEA, I., BUDU, V. Studiul metalelor ndreptar tehnic, Editura Facla, Timioara,
1987.
6. SMEAD, M., MANIU Al. Materiale utilizate n tehnica de aviaie, Sesiunea de
comunicri tiinifice cu participare internaional, Braov, 11-12 Mai 2007.
7. SMEAD, M. Factorii metalurgiei fizicii i chimiei care influeneaz obinerea
materialelor specifice aviaiei, Referat nr. 2 susinut n cadrul pregtirii prin doctorat, Catedra
Utilaj Tehnologic i tiina Materialelor, Facultatea tiina i Ingineria Materialelor,
Universitatea Transilvania din Braov, februarie 2008.
* Mihaela Smead, lector univ. drd. fiz., Academia Forelor Aeriene Henri Coand, Braov, Mihai Viteazu,
nr. 160, tel. 0268/423421, e-mail: mihaelasmeada@yahoo.com
** Ion Dinescu, profesor univ. dr. ing., Academia Forelor Aeriene Henri Coand, Braov, Mihai Viteazu, nr.
160, tel. 0268/423421, e-mail: dinescuion@yahoo.com
398
AFASES - 2008 -
Abstract: Pyrotechnic robots are service robots which are used to reduce the time for
intervention of pyrotechnic troops and to diminish the danger for the operators. Pyrotechnic
robots are used to inspect dangerous areas or/and to remove and to distroy explosive or
suspect devices/objects . These robots can be used to make some corridors through mined
battle fields, manipulation and neutralization of the intact ammunition, inspection of the
vehicles, trains, airplanes and buildings. For these robots a good functional activity is
determinated of dimensions of the work space , kinematics of the robotic arm and of
characteristics of the gripper. In the paper structural, kinematic, static synthesis and analysis,
design and functional simulation are shown of a robotic arm and the grippers attached from
a pyrotechnic robot designed by the authors.
1. Introduction
Pyrotechnic robots are service robots which are used to reduce the time for
intervention of pyrotechnic operators and to diminish the danger for them [5].
Pyrotechnic robots are used to inspect dangerous areas or/and to remove and to distroy
explosive devices. These robots can be used to make some corridors through mined battle
fields, manipulation and neutralization of the intact ammunition, inspection of the vehicles,
trains, airplanes and buildings. For these robots a good functional activity is determinated of
dimensions of the work space and kinematics of the robotic arm.
In the paper kinematic synthesis and analysis, one command schme and functional
simulation are shown regarding a robotic arm from a pyrotechnic robot designed by the
authors.
2. Pyrotechnic robots general aspects
A pyrotechnic robot has three main mechanical components: a mobile platform, a
robotic arm (the manipulator) and a gripper [2,3].
The mobile platform can be equipped with wheels or crawlers, the robotic arm is a
manipulator with 5 or 6 axles and the gripper is usually a gripping mechanism with jaws or,
rarely, an anthropomorphic mechanical gripper.
After the comparative structural analysis for more types of pyrotechnic robots
(fig.2,a,b), for the designed robot we adopted the structural scheme of fig.2,c.
399
AFASES - 2008 -
This structure is formed of a platform with crawlers, a robotic arm and a linkage
gripping mechanism with two jaws.
The platform has degree of freedom M = 2, each crawler is powered by an electric
motor.
400
AFASES - 2008 -
After structural analysis of the robotic arm the following clonclusion can be
formulated: there is 6 outside connections (L = 6) and the degree of freedom is 5 (M = 5). The
gripper has M = 1.
c
Fig. 2. Structural schemes
3. Designed robot main characterisctics
3.1. Structural synthesis of a robotic arm
For the first time the kinematic synthesis is used to obtain the main dimensions (fig.3)
for an optimum work space.
401
AFASES - 2008 -
0
0
0
0
1
1
0
0
0 cos( y )
Ry :=
0
0 cos( x) sin( x)
0
0 sin( x) cos( x)
0 sin( y )
0
0
1
0 cos( z ) sin( z )
Rz :=
0 sin( z ) cos( z )
0
0
0
0
0
0
1
x
T :=
y
0
1
0
0
0
0
1
0
0
0
0 sin( y )
1
0
0 cos( y )
0
0
0
b 4
cos( fi54)
0
sin( fi54)
E 35 :=
0
sin( fi 43) sin( fi54) cos( fi 43) sin( fi 43) cos( fi54)
0 cos( fi 43) sin( fi54) sin( fi 43) cos( fi 43) cos( fi54)
402
AFASES - 2008 -
E25:=simplify(multiply(E23,E34,E45));
E15:=simplify(multiply(E12,E23,E34,E45));
This is verified with numerical values in the following form:
L1:=[fi10=Pi/3,fi21=Pi/6,fi32=Pi/3,fi43=Pi/4,fi54=Pi/12];
1
1
1
1
1
L2:=[b1=100,a2=600,b2=200,b3=500,b4=90];
L 2 := [b1 = 100, a 2 = 600, b 2 = 200, b3 = 500, b 4 = 90]
E05s:=evalm(Msubs(L1,L2,E05));
1
0
0
0
The kinematic dimensions and the forces between constructive elements were used for
the calculation of strength and then a 3D model was obtained with the CATIA soft [4]. This
model was used at a functional simulation for the robotic arm (fig.5).
The general scheme for command and control subsystem is shown in Figure 6.
403
AFASES - 2008 -
References
[1] Dutita, FL. s.a. Mecanisme articolate. Inventica-Cinematica,Ed. Tehnica,Bucuresti, 1989.
[2] Paun, A., Constantin, G. Tipuri de roboti utilizati la manipularea si determinarea munitiei
neexplodate. Revista de Robotica si Management, nr. 1/2002.
[3] Staretu, I., Daj, I.Roboti de interventie in locuri periculoasecaracteristici constructive si
functionale. Buletinul AGIR anul IX, nr. 1-2, ian iunie 2004, p. 55-58.
[4] *** Documentatie CATIA 5.
[5] *** Documentatie Internet.
_
Ionel Staretu, prof. univ. dr. ing., Transilvania University of Brasov, Bd. Eroilor nr. 29, Brasov, Romania, tel.
0268412921, e-mail: staretu@unitbv.ro.
**
Marius Ionescu, drd. ing., Transilvania University of Brasov, Bd. Eroilor nr. 29, Brasov, Romania, tel.
0268412921, marius.ionescu@unitbv.ro.
404
AFASES - 2008 -
Cap. I. Generaliti
Att la formarea garniturilor trenurilor ct i n timpul circulaiei lor apar ocuri
datorate existenei unor viteze relative ntre vagoanele nvecinate. n scopul diminurii
efectului acestor ocuri de traciune, sistemul de legare (traciune) ale vehiculelor este echipat
cu amortizori de traciune [1], [5]. Exist o gam larg de soluii constructive a acestor
elemente elastice (metalice, din cauciuc cu elastomeri, hidraulice, hidropneumatice sau
combinate). Elementele elastice metalice de tip RINGFEDER (figura 1) sunt o soluie
adoptat datorit avantajelor ce le ofer (gabarit potrivit, rezisten n exploatare, caracteristici
dinamice adecvate, revizie i ntreinere facil) [3], [4], [6].
405
AFASES - 2008 -
Figura 2 Amortizor cu 3 inele cu fant pentru aparat traciune pt. vagon AVA-200
Valorile caracteristicilor obinute pentru amortizor sunt urmtoarele:
- energie potenial de deformaie nmagazinat We = 18,525 KJ ;
- energie potenial de deformaie disipat Wa = 13,325 KJ;
- factor = Wa/We = 71,92 %;
- fora de precomprimare este de 20 KN;
- fora la cursa de 45 mm este de 380 KN;
- fora la cursa de 65 mm este de 570 KN;
- s-a utilizat unsoare U230 Ca 2 EPM.
Cap.III. ncercri de anduran
Amortizorul a fost supus la ncercarea de anduran cu urmtorul program:
- 200 cicluri cu o sarcina de 85 % din energia nmagazinat;
- 1200 cicluri cu o sarcin de 50 % din energia nmagazinat;
- 3000 cicluri cu o sarcin de 25 % din energia nmagazinat.
406
AFASES - 2008 -
Figura 3 Amortizor cu 3 inele cu fant pentru aparat traciune pt. vagon AVA-200 dup
primul program de solicitare
- We = 18,2 KJ;
- Wa = 13 KJ;
- = 71,42 %;
- fora de precomprimare este de 20 KN ;
- fora la cursa de 45 mm este de 380 KN;
- fora la cursa de 65 mm este de 560 KN;
- s-a utilizat unsoare U 230 Go 2 EPM.
- dup programul nr. 2 de solicitare figura 4
Figura 4 Amortizor cu 3 inele cu fant pentru aparat traciune pt. vagon AVA-200 dup al
doilea program de solicitare
407
AFASES - 2008 -
- We = 17,98 KJ ;
- Wa = 13,02 KJ;
- = 72,41 %;
- fora de precomprimare este de 20 KN;
- fora la cursa de 45 mm este de 380 KN;
- fora la cursa de 62 mm este de 580 KN;
- s-a utilizat unsoare Shell Alvania EP-1.
- dup programul nr. 3 de solicitare figura 5
Figura 5. Amortizor cu 3 inele cu fant pentru aparat traciune pt. vagon AVA-200 dup al
treilea program de solicitare
- We = 17,08 KJ;
- Wa = 12,51 KJ;
- = 73,24 %;
- fora de precomprimare este de 20 KN;
- fora la cursa de 45 mm este de 380 KN;
- fora la cursa de 61 mm este de 560 KN;
- s-a utilizat unsoare Shell Alvania EP-1.
n timpul ncercrii de anduran s-a urmrit ca temperatura maxim s nu fie mai mare de 60o C.
Temperatura de lucru a fost ntre 50o C i 55C. Frecvena de aplicare a solicitrii a fost de 1,8 cicluri pe
minut.
Pe tot parcursul ncercrilor de anduran efectuate cu acest program nu au fost constatate fisuri sau
alte defecte ale inelelor RINGFEDER, iar caracteristicile amortizorului dup ultimul ciclu de solicitare
difera cu maximum 8% fa de caracteristicile obinute dup proba de flexibilitate. n concluzie se poate
afirma c amortizorul a rspuns pozitiv la programul de ncercri la care a fost supus.
Cap. IV. ncercarea de anduran de 5000 cicluri
Programul de solicitare de 5000 de cicluri, conform fiei UIC 827.2 s-a efectuat
cu amortizorul care n prealabil a fost supus la ncercarea de anduran de 3 x 4400 cicluri.
Regimul de pulsare a fost cu 85 % din energia potenial de deformaie nmagazinat i
frecvena de 1.8 cicluri/minut. S-a folosit tipul de unsoare Shell Alvania EP-1, iar tremperatura de
408
AFASES - 2008 -
lucru a fost cuprins ntre 50 - 55C. Dup fiecare 500 cicluri aplicate amortizorul a fost uns cu unsoare
Shell Alvania EP-1, pstrndu-se temperatura de lucru cuprins ntre 50 - 55C.
n timpul programului de ncercri s-au constatat urmtoarele:
- la numrul 878 de cicluri s-a rupt inelul exterior nr. 3 fa de suprafaa de sprijin de jos, s-a
schimbat inelul rupt, s-a reuns amortizorul cu unsoare Shell Alvania EP-1 i s-a continuat proba de
anduran;
- la numrul de 1910 cicluri s-a rupt inelul exterior nr. 2 fa de suprafaa de sprijin jos. S-a
nlocuit inelul rupt, s-a reuns cu unsoare Shell Alvania EP-1 i s-au continuat ncercrile;
- la numrul 2040 de cicluri s-au rupt inelele exterioare 4 i 5 fa de suprafaa de sprijin jos, sau nlocuit inelele rupte i s-a continuat programul de ncercri pn la atingerea celor 5000 de cicluri
impuse.
Cap. V. Concluzii
Amortizorul testat a rspuns favorabil la ncercarea de flexibilitate i de anduran 3 x
4400 de cicluri. La ncercarea de anduran de 5000 de cicluri ntruct au aprut pe parcursul
ncercrilor 4 ruperi de inele se stabilete c cerinele impuse de normele internaionale n ce
privete fiabilitatea nu au fost respectate. n consecin s-a considerat c este necesar
urmrirea mai atent a respectrii tehnologiei de execuie, a verificrii riguroase
dimensionale i a aplicrii unor tratamente termice [2] ce s conduc alturi de un control
tehnic de calitate adecvat la creterea fiabilitii.
Bibliogafie
1.
2.
3.
Cioclov D.
4.
5.
6.
Aurelia Tnsoiu, .l. dr. ing., Universitatea Aurel Vlaicu Arad, tanasoiu.aurelia@rdslink.ro
409
AFASES - 2008 -
Studiul teoretic a evideniat calculul efectuat pentru arcurile elicoidale ale suspensiei
boghiului Y25Lsdi destinate echiprii vagoanelor de marf. A fost ales tipul de solicitare
axial de compresiune ntruct arcurile elicoidale avnd diametrul spirei de 31mm
respectiv 24,4mm sunt legate n paralel n cadrul suspensiei boghiului. Subansamblul
studiat a fost sprijinit pe seciunea transversal, pe captul leagnului boghiului, iar forele
de compresiune au fost aplicate n captul opus n acord cu situaia din exploatare.
1. Generaliti
Elementele elastice metalice, sub form de arcuri elicoidale, arcuri bar de torsiune i
arcuri lamelare sunt larg utilizate la vehiculele pe ine ca elemente de suspensie. Ca i
celelalte tipuri de elemente elastice folosite la suspensii, arcurile metalice sunt organe de
maini cu o mare elasticitate. Datorit acestui fapt ele nmagazineaz un lucru mecanic de
deformare pe care l redau, n cea mai mare parte, sistemului mecanic sub form de energie
cinetic, cnd aciunea sarcinii a ncetat. Folosirea acestei proprieti conduce, n funcie de
condiiile impuse suspensiei n plan vertical i orizontal de spa iul disponibil pentru
montarea arcului, de mrimea i felul solicitrii, la diferite tipuri de arcuri[1], [7], [3], [5], [4].
Arcurile elicoidale cai arcurile bar de torsiune, n comparaie cu arcurile lamelare,
au un grad ridicat de utilizare a materialului, tehnologia de fabricare i reparare este relativ
simpl, nu necesit ntreinere n exploatare i sunt sensibile la perturbaii mici.
2. Determinarea prin metoda elementului finit a strii de tensiune utiliznd
programul ALGOR
Prezentarea programului ALGOR
Mai multe tehnici analitice sunt disponible pentru a determina compatibilitatea
modelului analizat. Mediul Superview IV Results ofer acces la aceste tehnici oferind att
tensiunile brute n coordonate locale i multe valori derivate din acestea. Valorile derivate
410
AFASES - 2008 -
includ criteriile von Mises i Tresca, tensiunile principale minime i maxime i rezultate
specifice elementelor. Dat fiind c acurateea rezultatelor analitice depinde de modul de
alegere al discretizrii i aplicarea parametrior MEF, mediul Superview IV Results ofer o
estimare a preciziei la nodurile comune [6]. Aceast valoare a preciziei ajut la determinarea
compatibilitii modelului.
n continuare se definesc urmtoarele tensiuni oferite de program:
- Tensiune Von Mises
Afiarea rezultatelor este sub forma tensiunii echivalente Von Mises. Aceasta poate fi
afiat pentru elemente de suprafa (2-D, plci, etc.) i elemente de volum.
Ecuaia folosit este:
(1.1)
Unde x, y i z sunt tensiunile axiale n direciile globale, iar xy, yz, xz (figura 1.1)
sunt forele tietoare. Tensiunea von Mises este ntotdeauna pozitiv.
n funcie de tensiunile principale 1, 2 i 3 tensiunea echivalent von Mises este
dat de relaia:
e = 0,5 ( 1 2 )2 + ( 2 3 )2 + ( 3 1 )2
(1.2)
411
AFASES - 2008 -
412
AFASES - 2008 -
Fig. 2.1.
Fig. 2.2.
Fig. 2.3.
Fig. 2.4.
413
AFASES - 2008 -
Fig. 2.5.
Fig. 2.6.
Fig. 2.7.
Fig. 2.8.
414
AFASES - 2008 -
Tensiunile maxime de ntindere n toate situaiile tratate sunt cu cel puin 24% sub
valoarea limitei de elasticitate de 1511N/mm2, aa nct posibilitatea apariiei unei fisuri
datorat unor deformaii permanente este eliminat.
Rezultatele calculului sunt prezentate n tabelul 2.1 i n figurile 2.1 2.8.
Tensiuni calculate
Tens. von Mises
Tens. max. princ.
Tens.interm.princ.
Tens. min. princ.
Tensor z-z
Tensor y-y
Tensor x-x
Tensor z-x
Tensor y-z
Tensor x-y
Arc 31mm
max[N/mm2]
min[N/mm2]
1672
0,27
1150
-28,14
408
-37,17
156,40
-1128,58
937,51
-186,11
599,35
9,22
563,47
-842,01
658,88
-487,91
327,20
-740,45
638,18
-797,52
Tabel 2.1
Arc 24,4mm
max[N/mm2]
min[N/mm2]
1576
0,34
862
-41,50
356
-164,46
121,79
-1056,51
499,14
-755,64
258,14
-400,56
386,04
-644,13
215,91
-369,35
316,62
-604,97
575,34
-272,49
Arcuri de compresiune(interior) 24
Lungimea
Cursa
Cursa
sub sarcina
superioar
inferioar
static [mm]
[mm]
[mm]
L1 = 203
L0 = 217.8
Lu = 188.2
L1 = 203
L0 = 218.4
Lu = 187.6
L1 = 203
L0 = 219.5
Lu = 186.5
415
Cursa
[mm]
Numr de
cicluri
14.8
15.4
16.4
2.0 x 106
1.0 x 106
1.0 x 106
AFASES - 2008 -
Tabelul 3.2
Treapta de
sarcin
1
2
3
Cursa
[mm]
Numr de
cicluri
10.5
12.6
16.7
2.0 x 106
1.0 x 106
1.0 x 106
416
AFASES - 2008 -
Bibliografie
1.
2.
3.
4.
5.
6.
7.
8.
Aurelia Tnsoiu,
417
AFASES - 2008 -
Radu RULESCU *
Stelian RULESCU **
Rezumat
n aceast lucrare se prezint contribuii privind optimizarea configuariei senzorilor
unui robot mobil n vederea unei mai bune orientri n spaiul de lucru (mediu ambiant).
Pentru studiul modului de orientare (navigaie) s-a folosit un senzor ultrasonic analogic de
tip FESTO (aflat n laboratorul de senzoric al Catedrei de Mecanic Fin i Mecatronic a
Facultii de Inginerie Mecanic, Universitatea Transilvania Braov), cu un modul de
emitere i un modul receptor. S-a studiat modul de dispunere a senzorilor pe robotul mobil i
influna diferiilor factori perturbatori din mediul ambiant.
I. Introducere
Una dintere categoriile de senzori folosii la navigaia unui robot mobil o constituie
categoria senzorilor ultrasonici. ntlnii n literatura de specialitate i sub denumirea de sonar
(en. Sound Navigation and Ranging), senzorii ultrasonici folosesc un principiu oarecum
asemntor cu senzorii IR., dar n loc de a transmite fascicole luminoase, ei folosesc semnale
acustice (sunete). Un emitor sonar transmite un semnal acustic n mediu, urmnd apoi ca
reflecia acestuia s fie recepionat de componenta detector a senzorului. Timpul n care
semnalul este receptat napoi de senzor precum i atenuarea semnalului reprezint aspecte
exploatate de diferitele tipuri de senzori sonar.
Sunetele transmise de senzorii sonar sunt de regul n spectrul de sunete ultrasonice.
avnd o frecven foarte nalt pentru a nu putea fi detectate de urechea uman.
Principiul de funcionare al senzorilor sonar a fost mprumutat din lumea animal, att
liliecii ct i delfinii folosind aceleai tehnici pentru a localiza i identifica diferitele
caracteristici ale unui anumit obiect.
Dac semnalul acustic trimis de emitor se propag direct ctre un obiect, se reflect
de acesta i revine napoi la detector, distana pn la obiectul respectiv poate fi calculat
prin:
1
d = vt
(1)
2
418
AFASES - 2008 -
n formula de mai sus, t reprezint intervalul de timp n care semnalul acustic ajunge
de la emitor napoi la detector, iar v reprezint viteza sunetului emis n mediu, calculat cu
formula:
v = v 0 + 0,6 T [m/s]
(2)
Unde v0 reprezint viteza sunetului n aer la temperatura de 0 grade Celsius, iar T
temperatura n grade Celsius. Astfel, se poate calcula viteza semnalului acustic n aer la o
anumit temperatur.
Formula de mai sus este doar o aproximare. Viteza sunetului n practic mai este
afectat i de densitatea aerului (altitudine) precum i de ali factori precum umiditatea aerului
sau concentraia de praf n aer [1].
Senzorii ultrasonici au unghiuri de emisie i de recepie destul de mari (ntre 10 grade
i 120 grade), dar nu suficient de mari pentru a permite o investigaie panoramic (360 grade).
Aceasta poate fi obinut folosind mai multe sisteme emitor / receptor sau amplasnd senzorul
pe o platform rotativ [2].
II. Navigaia robotului dotat cu senzori ultrasonici
n figura 1 este prezentat un robot mobil aflat ntr-un mediu (spaiu de lucru) cu dou
obstacole, 1 i 2. Robotul trebuie s se mite, evitnd obstacolele, n vederea atingerii
punctului int. Prima etap este cea de detecie a obstacolelor [3]. n primul caz, figura 1.a),
robotul mobil are montat un sigur senzor acustic, amplasat frontal, avnd unghiul de emisie /
recepie de 15 grade. n acest sitiaie, robotul nu detecteaz nici unul din cele dou obstacole.
Acelai lucru se ntmpl i n cazul urmtor, cnd robotul are montai doi senzori acustici.
figura 1.b).
b)
a)
Fig. 1. a) Robot mobil cu un singur senzor acustic; b) Robot mobil cu doi senzori acustici,
amplasai la 7,5 grade fa de axa principal;
419
AFASES - 2008 -
c)
d)
Fig. 1. c) Robot mobil cu patru senzori acustici, amplasai la 7,5 grade fa de axa principal;
d) Robot mobil cu duzeci i patru de senzori acustici, amplasai la 7,5 grade fa de axa
principal.
n al trei-lea caz, figura 1.c), cnd robotul mobil are montai patru senzori acustici este
detectat obstacolul 1, ns obstacolul 2 rmne nedetectat. Pentru o investigaie complet,
robotul mobil trebuie s aib montai senzori acustici de jur mprejur, n numr de douzeci i
patru figura 1.d). Numrul de senzori este proporional cu dimensiunea unghiului emisie /
recepie. n acest caz ambele obstacole sunt detectate iar distana pn la ele poate fi calculat.
Aadar robotul poate demara etapele de planificare a drumului i de control a micrii.
O metod mai simpl este cea de montare a unui singur senzor acustic pe o platform
rotativ, figura 2. n acest mod se realizeaz scanarea complet a spaiului de lucru. Aceast
metod este folosit n mod uzual la roboii mobili echipai cu alte tipuri de senzori (vizuali,
laser), pentru confirmarea prezenei obstacolelor.
420
AFASES - 2008 -
n vederea orientrii robotului e folosit un senzor ultrasonic analogic de tip FESTO [4]
(aflat n laboratorul de senzoric al Catedrei de Mecanic Fin i Mecatronic a Facultii de
Inginerie Mecanic, Universitatea Transilvania Braov), cu un modul de emitere i un modul
receptor.
Oscilator
Amplificator
Unitate de evaluare
Condiionarea semnalului de ieire
Tensiune extern
Alimentare cu tensiune intern constant
Traductor ultrasonic
Ieire: Semnal curent
Tabelul 1. Date tehnice
421
AFASES - 2008 -
Tensiunea de operare
Consumul de energie electric
Rezistor de sarcin
Curentul la ieire
Distana de msurare
Temperatura de operare
Eroarea de liniaritate
Frecvena de ceas msurat
Unghiul de recepie a sunetului
Protecie la inversarea polaritii
Clasa de protecie
Materiale
Greutate
24 V DC
<35 mA
<400
6 20 mA
150 500 mm
-20 +75oC
0,2%
40 Hz
5o
da
IP 65
plastic
250 g
Puterea semnalului acustic care pleac de la o surs scade conform cu legea inversului
ptrat.
I=
I0
(3)
4R 2
Unde:
I = intensitatea (puterea per suprafa) la distana R
I0 = intensitatea maxim (iniial)
R = distana.
De asemenea exist i o pierdere exponenial asociat cu absoria molecular a energiei
sunetului de ctre mediul ambiant.
I = I 0 e 2R
(4)
= coefficient de atenuare pentru mediu.
Distana maxim de detecie pentru un senzor ultrasonic e dependent de puterea i de
frecvena de emisie a semnalului astfel: cu ct frecvena scade, cu att distana de detecie
crete. Atenuarea maxim teoretic pentru energia ultrasonic poate fi apreximat prin:
f
(5)
a max =
100
Unde:
amax = atenuarea maxim n dB/m
f = frecvena de lucru n KHz.
Pentru o transmisie cu frecvena 20-KHz, factorul tipic de absorie este aproximativ
0,065dB/m, n timp ce la 40-KHz pierderile sunt n jur de 0,2 i 0,3 dB/m.
I e 2R
(6)
I= 0
4R 2
422
AFASES - 2008 -
Avnd n vedere ecuaia de mai sus, se poate spune c intensiatea semnalului scade
direct proporional cu ptratul distanei.
b) Reflectivitatea obiectelor int
Ii Za + Z0
(8)
Unde:
Ir = intensitatea reflectat
Ii = intensitatea incident
Za = impedana acustic a aerului
Z0 = impedana acustic a obiectului int
Cu ct e mai mare dezacordul ntre impedanele din cele dou medii, cu att mai mult
energie va fi reflectat napoi la surs.
c) Influena temperaturii
Cu toate aceste probleme, senzorii ultrasonici sunt cu siguran cea mai des ntlnit
categorie de senzori de distan folosit la construcia unui robot mobil. Preul lor se
ncadreaz ntre preul foarte mic al senzorilor IR i preul mult mai mare al altor categorii de
senzori de distan mai precise.
423
AFASES - 2008 -
Bibliografie
Radu RULESCU, Asistent Univ. Ing, Universitatea Transilvania din Braov, Facultatea de Inginerie
Mecanic, Catedra de Termodinamica i Mecanica Fluidelor, Braov, 500036, Bdul Eroilor 29, Tel.:
0268412921 / int.193, Fax : +40 268 474761, E-mail radu.tarulescu@unitbv.ro
**
Stelian RULESCU Asistent Univ. Ing, Universitatea Transilvania din Braov, Facultatea de Inginerie
Mecanic, Catedra de Autovehicule i Motoare, Braov, 500036, Bdul Eroilor 29, Tel.: +40.68.413000 Int. 173,
E-mail s.tarulescu@unitbv.ro
424
AFASES - 2008 -
425
AFASES - 2008 -
conductor) poate servi ca suport geometric i pentru obinerea unor reprezentri animate ale
funcionrii.
1.1 Mecanismul cu biel
Fie schema cinematic a cunoscutului mecanism din figura 1. Poziiei curente i s-au
ataat vectorii corespunztori vitezelor punctelor p1 i p3 , care au fost astfel notai tocmai
pentru a urmri mai uor prezena coordonatelor acestor puncte n cadrul aplicaiei ce va fi
prezentat n continuare.
426
AFASES - 2008 -
!(distance p1 p6)
)
adic viteza momentan a elementului condus este la aceast poziie a elementului conductor
de 1088,31 uniti pe minut.
Fig. 2 Poziii succesive ale elementelor mecanismului care pot fi folosite pentru reprezentri
animate
Se cunoate c principiul animaiei a rmas acelai nc de la inventarea
cinematografului i anume se ataaz coordonatelor din plan (geometrice) una suplimentar:
timpul. Mediile specializate pentru animaie 2D nu au totdeauna i instrumente satisfctoare
pentru o modelare geometric precis, a unghiurilor i a distanelor. De aceea este necesar
colaborarea cu un program specializat i interactiv, aa cum este de exemplu AutoCAD, dar
nu numai acesta, care beneficiaz de astfel de instrumente. Poziiile din figura 2 pot fi
obinute toate odat, introducnd n cadrul aplicaiei o secven iterativ, sau succesiv, cte
una, aa cum realizeaz aplicaia urmtoare, numit abiela, care are predefinite valorile
razei de 100 uniti i lungimii bielei de 200 i cere doar introducerea unghiului curent u.
(DEFUN C:aabiela ( )
(SETQ o (list 400 600)
r 100
l 200
u (GETANGLE "Valoarea unghiului u:"))
(SETQ p1 (polar o u r)
a (* r (sin u))
b (sqrt (- (expt l 2) (expt a 2)))
c (* r (cos u))
p3 (polar o 0 (+ b c))
P4 (polar p3 (*(/ pi 180) 30) 20)
P5 (polar p3 (*(/ pi 180) 150) 20)
p6 (polar p3 (*(/ pi 180) 210) 20)
p7 (polar p3 (*(/ pi 180) 330) 20))
427
AFASES - 2008 -
428
AFASES - 2008 -
429
AFASES - 2008 -
p1 (polar o2 (+ pi u) d)
u1 (angle o1 p1)
p2 (polar o1 u1 400)
p3 (polar o1 u1 600)
p4 (polar p1 u 1000)
c1 (polar p4 (+ (*(/ pi 180) 30) u) 40);pistonulc1c2c3c4
c2 (polar p4 (+ (*(/ pi 180) 150) u) 40)
c3 (polar p4 (+ (*(/ pi 180) 210) u) 40)
c4 (polar p4 (+ (*(/ pi 180) 330) u) 40)
i1 (polar o2 (+ (/ pi 2) u) 40);cilindruli1i2i3i4
i2 (polar o2 (- u (/ pi 2)) 40)
i3 (polar i1 u 500)
i4 (polar i2 u 500)
d11 (distance p2 o3)
a11 (/ (- (+ (expt d11 2) (expt r2 2)) (expt r1 2)) (* 2 d11 r2 ));elem3si4
b11 (sqrt (- 1 (expt a11 2)))
c11 (/ b11 a11)
u11 (atan c11)
u12 (angle p2 o3)
u13 (+ u11 u12)
p11 (polar o3 (+ pi u13) r2)
)
(Command "Point" p1)
(Command "Point" o1)
(Command "Point" o2)
(Command "Point" o3)
(Command "Point" p2)
(Command "Point" p3)
(Command "Point" p4)
(Command "Point" p11)
(Command "circle" p3 40)
(Command "pline" o1 p3 "c")
(Command "pline" p1 p4 "c")
(Command "pline" p11 p2 "c")
(Command "pline" p11 o3 "c")
(Command "pline" c1 c2 c3 c4 "c")
(Command "pline" i1 i2 "c")
(Command "pline" i1 i3 "c")
(Command "pline" i2 i4 "c")
(Command "pline" i3 i4 "c")
)
n acest exemplu, pentru elementele constructive ale lanului cinematic s-au ales
urmtoarele valori: raza r (de la O1 la prima articulaie a lanului) de 200 uniti, razele
sistemului de biele r1 i r2 de cte 300 uniti, lungimea tijei pistonului motorului hidraulic de
1000 uniti (p4 (polar p1 u 1000)) i distana de la O1 la a doua articulaie de 600 uniti (p3
(polar o1 u1 600)). Toate acestea pot fi modificate cu uurin, dup cum este necesar n oricare
alt situaie.
Valoarea deplasrii pistonului este cerut de ctre aplicaie n mod repetat, prin funcia
de introducere:
(SETQ d (GETREAL "Valoarea cursei d:")
430
AFASES - 2008 -
n figura 5 este prezentat imaginea unei succesiuni de poziii ale elementelor, pentru
un pas al avansului tijei pistonului de 10 uniti. Aceast aplicaie este destinat obinerii unei
reprezentri animate a funcionrii, de exemplu cu ajutorul soft-ului Macromedia Flash
(exportul succesiv al modelelor key frame). Dar ea poate servi i ca punct de pornire pentru un
studiu cinematic al acestui mecanism, aa cum s-a artat la nceputul acestui articol n cazul
mecanismului biel-manivel. Adic ntreaga construcia a planului vitezelor i a planului
acceleraiilor poate fi automatizat, programul returnnd, pentru fiecare introducere a valorii
cursei pistonului, valorile instantanee ale vitezelor, respectiv, acceleraiilor elementului
condus.
.
Fig. 5 Succesiunea poziiilor elementelor mecanismului, cu un pas al deplasrii pistonului de
10 uniti, de la d = 500 la d = 550, obinut cu ajutorul aplicaiei aatren.
1.3 Probleme teoretice de GA i GD rezolvate cu mijloace CAD
Soluiile unor exerciii de geometrie analitic, referitoare la punct, dreapt i plan pot
fi calculate i vizualizate prin programare, folosind programul AutoCAD i interfaa lui
programabil, limbajul de nivel nalt, AutoLISP.
Coordonatele punctului de intersecie ale dreptei determinat de punctele M(xm, ym,
zm) i N(xn, yn, zn) cu planul punctelor A(x1, y1, z1), B(x2, y2, z2) i C(x3, y3, z3) sunt soluiile
urmtorului sistem liniar, alctuit cu ecuaiile (1) i (2):
x
y
z 1
x1 y1 z1 1
(1.4)
=0
x2 y2 z2 1
x3 y3 z3 1
x xm
y ym
z zm
=
=
xn xm yn ym zn zm
Pe de alt parte, pot fi alctuite aplicaii specifice (macrofuncii AutoCAD), cu
ajutorul crora pot fi vizualizate automat, n dubl proiecie ortogonal, att planul, prin
urmele sale ph, pv, ct i dreapta, prin proieciile ei d i d.
Din economie de spaiu, nu vom prezenta aceste aplicaii, ci doar reprezentrile
obinute cu ajutorul lor. Se cunosc operaiile grafice necesare aflrii dublei proiecii
ortogonale a punctului de intersecie a unei drepte, definit prin coordonatele unor puncte ale
ei , cu un plan, definit tot prin coordonatele celor trei puncte ale sale. Aceste operaii se
431
AFASES - 2008 -
repet, sunt algoritmice i ele se deosebesc doar prin valorile acelor coordonate. De aceea,
dup principiul artat anterior, se poate elabora un program care s reuneasc operaiile
grafice necesare. El a i fost deja scris i, cu ajutorul su, s-a obinut reprezentarea dublei
proiecii ortogonale a punctului I(i, i), din figura 6.
ABC(abc,abc) MN(mn,mn) = I(i,i)
(1.5)
i'
b'
1'
n'
c'
2'
m'
2c
a'
n
a
(1.6)
Tarziu Horia, Conf. dr. ing., Academia Fortelor Aeriene H. Coanda, Brasov, email: tarziuh@yahoo.com
Luculescu Doru, Conf. dr. ing., Academia Fortelor Aeriene H. Coanda, Brasov, email:
doruluculescu@yahoo.com
432
AFASES - 2008 -
433
AFASES - 2008 -
porturi. Pentru a realiza acest obiectiv s-a recurs la automatizarea operaiunilor de manipulare
n terminal, lucru care a condus la scderea considerabil a timpului de operare al navei.
n ceea ce privete navele portcontainer, i acestea au continuat s i mreasc
dimensiunile dar i puterea motoarelor. Impactul pe care l vor avea navele portcontainer
asupra porturilor dar i asupra terminalelor va fi unul sever att timp ct acestea vor crete att
ca mrime dar i n ceea ce privete capacitatea de transport.
2. Evoluia navelor portcontainer
Prima linie de portcontainere a nceput din portul Newark spre Huston, n 1956, cu o
nav tanc tramsformat, din al doilea Rzboi Mondial. Numele navei era Ideal X i putea
transporta 58 ( 35) de containere. Beneficiile timpului scurt de ncrcare i staionare n port
au fcut ca armatorii americani s transforme tot mai multe nave vechi pentru a transporta
mrfuri containerizate.
Prima nav specializat n transportul containerelor, dar fr mecanism de ncrcare, a
fost Sea Land Venture fiind introdus n operare la civa ani dup Ideal X .
Transportul containerizat s-a inventat n America i este de neles c dimensiunile
primelor nave portcontainer au urmat standardele americane. Mai trziu s-a introdus i
standardul pentru containere : lungime de 20 ft (6,085m), lime de 8 ft (2,438m) i nlime
de 8 ft (2,438m). Astfel containerul a devenit unitate de msur elementar folosit n ntreaga
lume, folosindu-se abrevierea TEU (Twenty Feet Equivalent Unit).
Succesul noului concept de transport a fost excepional iar navele purttoare de
containere au crescut rapid n mrime i capacitate.
La nceput navele au fost clasificate folosindu-se conceptul generaiei, astfel :
Prima generaie : 1000 TEU (au aprut n 1966)
A II a generaie : 2000 TEU (au aprut n 1969)
A III a generaie : 3000 TEU (au aprut n 1972)
Pentru navele care au urmat, datorit capacitii mari de transport, clasificarea pe
generaii nu a mai fost valabil. Cu toate acestea navele cresc dar n special geometria
ecluzelor canalului Panama limiteaz mrimea navelor (Fig.1) la dimensiunile urmtoare:
lungime de 294,1m, lime de 32,3m i pescaj de maxim 12m. Acestea erau denumite nave
Panamax avnd o capacitate de transport maxim de 4500 TEU (aceast limit a fost depit
n anul 1970)
n general limea navelor este dat de numrul de stive n direcie transversal. Astfel
limea navelor portcontainer a crescut treptat cu 8 ft, pn n prezent construindu-se nave cu
434
AFASES - 2008 -
435
AFASES - 2008 -
Portul
de export
Costurile
la teren
Costurile de
productie
Costurile de transport
interior la export
Portul
de import
Costurile
la dan
Costurile
portuare la export
Costurile de transport
pe mare
Costurile
la dan
Costurile
la teren
Costurile
portuare la import
Costurile de transport
interior la import
Costurile de distributie
si de vnzare
IMPORT
EXPORT
436
AFASES - 2008 -
Pe lng cele menionate mai sus unul din secretele obinerii unei performane ridicate, la
operarea navelor portcontainer, este folosirea utilajelor de manipulare, fie cele de pe cheu fie
din spaiul de depozitare a containerelor, la capacitatea maxim admis.
O alt cale pentru creterea producivitii n terminal este utilizarea prederelor care
lucreaz n tandem (Fig. 4). Aceste predere sunt capabile s ridice simultan dou containere
de 40 sau patru containere de 20. Legturile mecanice ntre cele dou predere asigur
reglarea nlimii celor dou containere.
437
AFASES - 2008 -
438
AFASES - 2008 -
BIBLIOGRAFIE
1. Review of maritim transport 2005 , Port development
2. Port Technology Internaional , The Ultra Large Container Ship (ULCS): presenting new problems
for ports?, 2005
3. 9th Symposium on Practical Design of Ships and Other Floating Structures, Germania , The
Challenge of Ultra Large Container Ship , A Hydrodynamic View, 2005
4. Port Technology International , Container Terminal Planning: improving system productivity to
service larger container vessels, 2007
Filis Tulei, preparator universitar, Academia Naval Mircea cel Btrn, Constana, Str. Fulgerului Nr. 1,
telefon 0723944993, e-mail: fixe2005@gmail.com
439
AFASES - 2008 -
440
AFASES - 2008 -
441
AFASES - 2008 -
Tip de
conexiune
Criteriu
Viteza de transmitere
Distana de tansmitere
Imunitatea electromagnetic
Izolaia n tensiune
Timpul de stabilire a
legturii
Conexiune
simpl
Conexiune
adresabil
Conexiune
radio
Conexiune prin
fibr optic
mic
medie
sczut
sute de V
mare
medie
medie
sute de V
mare
mare
medie
-
foarte mare
mare
foarte bun
mii de V
zeci de minute
minute
minute
minute
442
AFASES - 2008 -
443
AFASES - 2008 -
444
AFASES - 2008 -
445
AFASES - 2008 -
se conecteaz la linia de comunicaie prin intermediul a dou circuite asincrone de recepietransmitere de tipul Intel 8250. Circuitul Intel 8250 are proprietatea de a trece asincron din
modul de transmitere n cel de recepie n cazul n care sesizeaz informaie la intrare i de a
face conversia serial-paralel (sau paralel-serial, funcie de poziia sa n schem).
Comunicaia este asincron i compatibil cu comunicaia portului serial RS-232, fiind
asociat unui ciclu de ceas al crui semnal nu este trimis cu informaia, astfel c fiecare
cuvnt este sincronizat prin utilizarea bitului de start i a ceasului intern. Formatul utilizat
pentru transmitere este formatul obinuit 8N1 (informaie de 8 bii, fr bit de paritate i cu un
bit de stop). Realizarea unui flux bidirecional de date impune utilizarea cte unui circuit Intel
8250 att pentru transmiterea datelor ct i pentru recepia datelor, timpii alocai pentru
transmiterea datelor n ambele sensuri nesuprapunndu-se.
Pentru a putea realiza recepia fr erori, este necesar ca att transmitorul conectat la
interfaa calculatorului, ct i circuitul Intel 8250 s aib setat aceeai vitez de transmitere.
Dac pentru calculator setarea ratei de baud este uor de realizat, prin program, pentru partea
situat n instalaia de antene este necesar generarea unei rate de transmitere de 9600 Baud
att pentru recepie ct i pentru transmitere, ceea ce este uor de implementat cu ajutorul
generatorului ratei de baud construit n jurul circuitului integrat specializat de tipul 74HC4060
(circuit divizor de frecven) i a unui cuar extern. Valoarea cuarului extern este dictat de
valoarea impus pentru ieirea de baud i rata de divizare, astfel c, n cazul aplicaiei
descrise, pentru valoarea de 9600 Baud, se va alege un cuar de 2,4546 MHz.
5. Bilanul de putere al legturii de fibr optic
Un aspect foarte important al viabilitii legturii de fibr optic proiectate mai sus
este reprezentat de necesitatea furnizrii de ctre transmitoare a unei puteri optice suficiente,
care s asigure funcionarea corect a legturii. Mai mult de att totalul pierderilor pe legtura
optic i marginile de siguran optic trebuie s fie luate n considerare cnd se calculeaz
puterea cerut a transmitorului. Algoritmul de lucru pretabil a fi utilizat la calcularea
totalului pierderilor pe legtura de fibr optic, a bilanului de putere al legturii optice i a
puterii optice cerute a transmitorului, respect structura reprezentat n figura de mai jos:
446
AFASES - 2008 -
Cornel Ovidiu BADEA, drd. ing., Universitatea Politehnic, Bucureti, Splaiul Independenei nr. 313, tel.
0724.546.476, e-mail:cornelovidiu.badea@gmail.com
447
AFASES - 2008 -
I Modelul matematic
Posibilitile de modificare a turaiei motorului asincron decurg din analiza ecuaiei
[1]:
f1
(1 - s)
(1)
p
Din aceast expresie rezult c turaia unui motor asincron se poate regla prin
frecvena tensiunii de alimentare f1, numrul perechilor de poli din stator p i prin alunecarea
s a mainii.
Reglarea turaiei n limite largi se realizeaz prin modificarea frecvenei tensiunii de
alimentare f1 cu ajutorul convertoarelor statice de frecven. Aceste convertoare sunt
construite cu ajutorul tiristoarelor pentru mainile de putere mare sau cu tranzistoare pentru
n=
448
AFASES - 2008 -
maini de putere redus. Un astfel de convertor static indirect de frecven este prezentat n
figura 1, fiind construit dintr-un redresor comandat pentru modificarea tensiunii pe maina
asincron i un invertor pentru modificarea frecvenei.
(+ )
u1
u2
UD
u3
MA
3~
( )
Fig. 1: Convertor indirect de frecven.
Pentru a menine un flux constant n main la modificarea frecvenei tensiunii de
alimentare este necesar ca U/f = ct.
n realitate pentru ca maina s funcioneze la aceeai capacitate de suprasarcin,
randament i factor de putere este necesar ndeplinirea relaiei[4]:
U
f
M
(2)
=
UN
fN
MN
unde: UN, fN i MN, sunt mrimile nominale ale mainii.
Alimentarea mainilor electrice prin convertoare statice presupune analiza funcionrii
mainii n regim tranzitoriu, deoarece regimul permanent (staionar) este format dintr-un
lan de regimuri tranzitorii.
Studiul regimului tranzitoriu se realizeaz pe baza teoriei celor dou axe (modelul
ortogonal) la care ecuaiile mainii asincrone sunt scrise n raport cu timpul, parametrii
electrici fiind considerai constani.
n teoria celor dou axe, maina asincron trifazat este echivalent cu o main
asincron avnd nfurrile statorice i rotorice dispuse dup axele d i q n cvadratur
electric figura 2.
Maina echivalent are acelai cuplu, aceeai putere electromagnetic primit la borne
ca i maina real.
Ecuaiile pentru stator i rotor cu sistemul de referin dq fix fa de cmpul
magnetic din stator, sunt urmtoarele[3]:
d d
- 1 q
U d = R1 i d +
dt
d q
+ 1 d
U q = R1 i q +
dt
(3)
d dr
- (1 - ) qr
U dr = R 2 i dr +
dt
d qr
U qr = R 2 i qr +
+ (1 - ) dr
dt
unde fluxurile magnetice din stator i rotor sunt:
449
AFASES - 2008 -
(4)
id
ud
idr
udr
iq
iqr
u qr
uq
M m = M N
(7)
Cu ajutorul sistemului de ecuaii format din relaiile (3) i (5) se obin variaiile n
timp a id(t), iq(t), idr(t), iqr(t) i (t), de unde se poate determina curentul real din main[2]:
2
i=
[id cos( 1 t ) - i q sin( 1 t )]
(8)
3
450
AFASES - 2008 -
n cazul mainii asincrone cu rotorul n colivie avnd n vedere c barele din rotor
sunt scurtcircuitate cu ajutorul unor inele, tensiunile nfurrilor rotorice din axele d i q sunt
nule. (Udr = Uqr = 0).
n traciunea electric a troleibuzului, linia de contact fiind monofazat, redresorul ce
intr n componena convertorului de frecven este monofazat, iar invertorul este trifazat.
Unda de tensiune pentru prima faz la ieire din invertor este de forma prezentat n
figura 3.
u1
2
3
U D
3
2
U
3
1t
U2 =
U3 =
4 2
4 2
2
4 2
= 6 k 1
= 6 k 1
= 6 k 1
sin 1 t
sin 1 t
(9)
sin 1 t 2
Ud =
U cos 1 t - ( - 1)
3 =1
3
(10)
2 3
2
U sin 1 t - ( - 1)
Uq = 3 =1
3
451
AFASES - 2008 -
sin 1 t ( 1)
8
2
(
)
Ud =
U
t
cos
1
3
3 2
=1 = 6 k 1
(11)
sin 1 t ( 1) 3
3
8
sin t ( 1) 2
Uq =
U
3
3 2
=1 = 6 k 1
452
AFASES - 2008 -
Fig. 4a
Fig. 4b
Regimul tranzitoriu n procesul pornirii mainii asincrone
(PN =130 kW, n = 580 rot/min) pentru U = 380 V; f = 50 Hz i Mm =1000 Nm.
Fig. 5a
Fig. 5b
Regimul tranzitoriu n procesul pornirii mainii asincrone
(PN =130 kW, n = 580 rot/min) pentru U = 76 V; f = 10 Hz i Mm =1000 Nm.
453
AFASES - 2008 -
O influen mare asupra turaiei o are caracteristica cuplului rezistent care depinde de
ptratul vitezei unghiulare i armonicile de ordin superior din cmpul magnetic al mainii
asincrone. Prin urmare, pentru turaii mai mari dect turaia nominal este necesar
dimensionarea motorului asincron la putere superioar.
Fig. 6a
Fig. 6b
Regimul tranzitoriu n procesul pornirii mainii asincrone
(PN =130 kW) pentru U = 380 V; f = 65 Hz i Mm =1000 Nm
III Concluzii
454
AFASES - 2008 -
Bibliografie
[1] Dordea, T. Maini electrice, Editura ASAB Bucureti, 2002, pp. 124-139;
[2] Dordea, T. Maini electrice.Partea complementar, Editura Politehnica
Timioara, 2002, pp. 10-19;
[3] Cmpeanu, A. Introducere n dinamica mainilor electrice de curent alternativ,
Editura Academiei Romne, Bucureti, 1998, pp. 60-84;
[4] Mller, V. Maini electrice, Editura Politehnica Timioara 2005, pp. 5062.
Ioan Bele, prof.univ.dr.ing., Universitatea Aurel Vlaicu din Arad nr. 77, tel. 0257 250389, e-mail:
ioan_bele2002@yahoo.com
**
Liviu Sevastian Boci, prof.univ.dr.ing., Universitatea Aurel Vlaicu din Arad nr. 77, tel. 0257 250389, email: bociis@yahoo.com
***
Ioan Strinescu, conf.univ.dr.ing., S. C. ICPE SAERP S. A., Bucureti, Splaiul Unirii nr.313, cod 030138
sector 3, Bucureti, e-mail: saerp@saerp.ro
****
Valentin Dan Mller, conf.univ.dr.ing., Universitatea Aurel Vlaicu din Arad nr. 77, tel. 0257 250389, email: mullervalentin@yahoo.com
*****
Mihaela Popa, asist.univ.drd.ing., Universitatea Aurel Vlaicu din Arad nr. 77, tel. 0257 250389, e-mail:
mihaela2popa@yahoo.com
455
AFASES - 2008 -
456
AFASES - 2008 -
457
AFASES - 2008 -
R =
l
S
(1)
N2
L= n
;
lk
k =1 k s k
(2)
magnetic al bobinei.
C=
S
;
d
(3)
temperatur;
umiditate;
458
AFASES - 2008 -
concentraie de gaze;
vitez gaze (debit);
vacuum.
cmpul magnetic;
inducie.
intensitate luminoas;
flux luminos;
deplasri
(prin
modulaia
fluxului de radiaii).
for;
presiune.
deplasri liniare;
dimensiuni piese;
grosime;
nivel.
acceleraie;
vitez;
vibraii.
for;
presiune.
nivel;
grosime;
umiditate (solide).
459
AFASES - 2008 -
Bena marius Dan Drd. ing. Universitatea Transilvania din Braov mariusbenta@yahoo.com
Farkas Csilla Student Universitatea Transilvania din Braov
460
AFASES - 2008 -
461
AFASES - 2008 -
Detecie sensibil
Prghii flexibile
Vrfuri ascuite
Din fig. 1 se observa c ntreg ansamblu al instalaiei AFM este conectat la un calculator,
iar pentru validarea suprafeelor scanate se va utiluza mediul de programare LabView.
n continuare se prezinta schemele bloc al programului LabView pentru instalaia AFM
462
AFASES - 2008 -
463
AFASES - 2008 -
464
AFASES - 2008 -
Bena Marius Dan Drd. ing. Universitatea Transilvania din Braov, mariusbenta@yahoo.com
Ene Viorel Prof. Dr. ing. Universitatea Transilvania din Braov
465
AFASES - 2008 -
Abstract
The paper shows the possibility of application of adjoint transformation method for
the scheme achievement of current-mode Stephensen active network, using the voltage-mode
scheme. The both circuits have the same transfer function. The circuit parameters is
calculeted, and the response curves is presented.
I. Introducere
Modul de lucru n tensiune a fost i a rmas deocamdat dominant n circuistica
analogic de procesare a semnalelor. Acest lucru s-a datorat, n principal, faptului c acest
mod de lucru a dus la soluii atractive i elegante pentru realizarea circuitelor integrate, care
apoi s fie utilizate la construcia majoritii tipurilor de circuite analogice. Unul dintre aceste
circuite integrate, cu o rspndire extrem de larg, este amplificatorul operaional [1].
Odat cu apariia conveiorului de curent (1966) s-a deschis o nou posibilitate de
realizare a circuitelor electronice analogice. Lupta inegal pe care a trebuit s o duc de-a
lungul timpului cu amplificatorul operaional, principala component integrat cu mod de
lucru n tensiune, folosit la construcia circuitelor electronice analogice, a fcut ca aceste
circuite s rzbat greu, n pofida unor avantaje nete ca domeniu dinamic mai larg, band mai
larg, consum mai mic de putere, timpi de comutare mai mici, flexibilitate i versatilitate
sporite. Dintre toate tipurile de conveioare de curent s-a remarcat conveiorul de curent de
generaia a doua negativ, CCII- (Negativ Second Generation Current Conveyor), considerat
ca fiind un bloc constructiv potrivit pentru realizarea unor circuite analogice cu mod de lucru
n curent, cum ar fi: oscilatoarele RC sau filtrele active.
Un alt bloc constructiv cu mod de lucru n curent, dar aprut mult mai trziu, este
blocul nulor cu patru borne flotante, FTFN (Four Terminal Floating Nullor), mai flexibil i
mai versatil dect amplificatorul operaional sau dect conveiorul de curent, prin faptul c
realizeaz o combinaie ntre avantajele modului de lucru n tensiune i cele ale modului de
lucru n curent. Din acest motiv, multe circuite se recomand s fie realizate cu FTFN.
466
AFASES - 2008 -
(1)
467
AFASES - 2008 -
3.
Relaiile ntre mrimile la borne sunt sintetizate n ecuaia matricial (1).
i X 0 0 0 0 i W
i 0 0 0 0 v
Y=
Y .
v X 0 1 0 0 v W
i Z 1 0 0 0 v Z
Din ecuaia matricial se deduc urmtoarele relaii [1]:
i X = i Y = 0;
vX = vY ;
iZ = iW .
Impedanele de ieire vzute n bornele W i Z sunt arbitrare.
(4)
(5)
(6)
(7)
468
AFASES - 2008 -
V
Vii
VVo o
Vo
Vi
A
469
AFASES - 2008 -
Plecnd de la schema din figura 6, se pot pune n eviden trei nuloare, cu intrarea
dinspre dreapta. Utiliznd schema echivalent din figura 2, se pot nlocui nuloarele cu blocuri
CCII-, obinndu-se pentru reeaua activ schema din figura 7. Utiliznd schema echivalent
din figura 3,a, se pot nlocui nuloarele cu blocuri FTFN, obinndu-se pentru reeaua activ
schema din figura 8.
Ii
470
(9)
AFASES - 2008 -
(10)
SG0 = SG0 = SG0 = SG0 = SC 0 = SC 0 = .
2
3
4
6
1
2
2
Din relaia (10) se observ c senzitivitile filtrului n raport cu componentele pasive
care intr n alctuirea sa sunt mici.
5.2. Rezultatele simulrii
471
AFASES - 2008 -
472
AFASES - 2008 -
[1] Toumazou C., Lidgey F.J., Haigh D.G., Analogue IC design: the current mode
approach, Peter Peregrinus Ltd., p. 116, 1990
[2] Gnay E., Uzunhisarcikli E., Kili. Ali M, A Realization of SC-CNN-Based Using
FTFN, Turk J Elec Engin., vol 13, No. 1, p.39-50, 2005
[3] Sedra A.S., Roberts G.W., Gohn F., The current conveyor: history, progress and new
results, IEE Proceedings, vol. 137, Pt. G, No. 2, p. 80, 1990
[4] Mateescu A., Analiza si sinteza circuitelor electrice, Editura Didactic si Pedagogic,
p. 635, Bucuresti, 1975
[5] Uzunhisarcikli E., The Use of Adjoint Transformation Method to Tranformation
Current Mode from Voltage Mode in Universal Active Element Based Filters, G.U.
Jurnal of Science, Vol. 17, No. 3, p. 179-187, 2004
[6] Shah N. A., Malik M.A., Implementation of High Input Impedance Multifunction
Filter Using FTFNs, Jurnal of Active and Pasive Electronic Devices, Vol.1, p. 53-59,
2005
[7] Bogdan O., Considerations on accomplishment of active filters with current mode
circuits, Proceedings of the National Conferences with International Participation
ECO-EIC05, p.35-38, 23-24 septembrie 2005
[8] Prakobnoppakao S., Chipipop B., Surakampontorn W., Watanabe K., Design of
Current-Mode CCII-Based Filter from Immittance Function Simulator using
Commercial Available CCII (AD844), ITC-CSCC2002, p. 743-746, 16-19 iulie,
Phuket, Thailanda
[9] Liu S.I., Yang C.Y., High-input impedance filters using FTFNs, INTERNATIONAL
JOURNAL OF ELECTRONICS, Vol. 84, No. 6, p. 595-598, 1998
Bogdan Octavian-Ioan, profesor univ. dr. ing., Academia Forelor Terestre, Revoluiei nr. 3-5, tel. 0269
432990, e-mail: obogdan@armyacademy.ro
473
AFASES - 2008 -
474
AFASES - 2008 -
475
AFASES - 2008 -
Vectorii avalan reprezint rezultatul operaiei XOR ntre textul criptat rezultat folosind
cheia fix i textul clar aleator, respectiv textul criptat rezultat folosind cheia fix i textul clar
perturbat cu bitul de intrare de pe poziia i inversat, 1in. Pentru fiecare poziie n parte, se
nsumeaz n continuare ponderea vectorilor avalan corespunztori i n final se calculeaz
media. n modul ideal (e.g. calculat pentru ntreg spaiul valorilor posibile ale blocului de text
clar - n cazul acesta 2n), pentru o funcie care respect criteriul avalan media obinut ar
avea valoarea 0.5. ntruct se face o evaluare statistic, ne ateptm ca rezultatele obinute s
prezinte o abatere de la valoarea ideal. Prin urmare, se calculeaz maximul acestei erori
relative i, dac are o valoare acceptabil (sub 1%), se consider c testul a fost trecut cu
succes. Prin similitudine, inversnd rolurile cheii i al textului clar, se obine i testul
verificrii efectului avalan al cheii.
3.2 Criteriul SAC
Cazul SAC este similar celui avalan, cu deosebirea c se investigheaz efectul
inversrii bitului i de intrare asupra bitului j de ieire. Este vorba deci de o analiz mai
profund a efectului avalan i rezult astfel, n loc de un vector, o matrice bidimensional
nn de coeficieni, care se calculeaz tot din vectorii avalan.
n mod ideal, toi coeficienii matricii ar trebui s fie egali cu 0.5, ns ca i n cazul
efectului avalan, ne ateptm ca aceti coeficieni s aib nu valoarea exact de 0.5, ci valori
apropiate. Se calculeaz prin urmare eroarea maxim relativ i n funcie de valoarea
obinut se ia decizia corespunztoare. Valoarea prag aleas pentru acest test este de 5%.
Acest test l nglobeaz i pe cel mai simplu de completitudine. n cazul acestui test sar fi verificat doar dac exist vreun element nul n cadrul matricii bidimensionale. Pentru
acest motiv, nu s-a mai fcut pentru completitudine un test separat.
3.3 Criteriul BIC
Pentru msurarea conceptului de corelaie a biilor, se calculeaz coeficienii de
corelaie dintre vectorii avalan generai de fiecare poziie n parte. Este o analiz i mai
profund, care implic analiza unei matrice tridimensionale nnn de coeficieni. Se
calculeaz valoarea maxim a elementelor acestei matrici, denumit coeficientul de corelaie
BIC, cu valori ntre 0 i 1. n mod ideal, are valoarea 0.
Ca i cazul celorlate teste, se calculeaz eroarea relativ i dac este n limita acceptat
(5%), are loc validarea testului.
3.4 Criteriul AWD
n cadrul acestui test, se calculeaz mai nti ponderile vectorilor avalan i apoi se
calculeaz vectorul distribuiei acestor ponderi. Acest vector ar trebui s aib o distribuie
binomial. Acest lucru se poate verifica fie grafic (prin reprezentarea i vizualizarea
histogramei corespunztoare), fie computaional, calculnd parametrii statistici (media i
dispersia) ai abaterii fa de distribuia binomial. A fost aleas ultima variant. Valorile
statistice obinute ale mediei i dispersiei ar trebui s fie apropiate de valoarea 0 (cazul ideal).
IV. Rezultate experimentale
Pentru verificare, au fost alei algoritmi deja testai i cu proprieti criptografice
recunoscute i oarecum apropiate: ctigtorii concusurilor AES i NESSIE, Rjindael i
respectiv Camellia, cu cheie de 256 de bii. Testele au fost aplicate att pentru textul clar, ct
i pentru cheie. n cadrul testelor au fost generate seturi de cte 100.000 de valori aleatoare.
Rezultatele sunt prezentate n tabelul urmtor:
476
AFASES - 2008 -
Test
Avalan
SAC
BIC
AWD
media
dispersia
AES
Text clar
Cheie
0.081422 %
0.073422 %
1.384000 %
1.188000 %
1.493030 %
1.841811 %
0.006517
0.006601
0.000001
0.000001
CAMELLIA
Text clar
Cheie
0.097328 %
0.108266 %
1.272000 %
1.310000 %
1.542629 %
1.557944 %
0.006620
0.006559
0.000001
0.000001
Gigi-Daniel BUDARIU, Agenia de Cercetare pentru Tehnic i Tehn. Mil., ACTTM, gbudariu@acttm.ro
Constantin GROZEA, Agenia de Cercetare pentru Tehnic i Tehn. Mil., ACTTM, cgrozea@acttm.ro
***
Mihai MARIN, Agenia de Cercetare pentru Tehnic i Tehn. Mil., ACTTM, marin_mihai2001@yahoo.com
**
477
AFASES - 2008 -
478
AFASES - 2008 -
esenial ca instalaia s urmeze intrarea chiar i n cazul n care dinamica instalaiei trebuie
modificat puin pe timpul lucrului. Un sistem a crei mrime de ieire, y(t), urmrete
ndeaproape mrimea de intrare, r(t), incluznd cazul n care r = 0, este definit ca fiind un
sistem de reglaj. Un sistem de reglaj performant este caracterizat de :
Rejecia bun - un bun reglaj n prezena semnalelor perturbatoare;
Sensibilitate sczut la parametrii sistemului - un bun reglaj bun n prezena
schimbrilor n parametrii instalaiei;
Robustee o bun rejecie a perturbaiilor i o sensibilitate sczut.
479
AFASES - 2008 -
Pupitrul de comand
Instalaia de antene
480
AFASES - 2008 -
481
AFASES - 2008 -
482
AFASES - 2008 -
483
AFASES - 2008 -
484
AFASES - 2008 -
Aurelian CRCIUNESCU, prof. univ. dr. ing., Universitatea Politehnic, Bucureti, Splaiul Independenei nr.
313, e-mail:aurelian_craciunescu@yahoo.com;
Valentin NVRPESCU, prof. univ. dr. ing., Universitatea Politehnic, Bucureti, Splaiul Independenei nr.
313, e-mail:valentin_ navrapescu@gmail.com;
Cornel Ovidiu BADEA, drd. ing., Universitatea Politehnic, Bucureti, Splaiul Independenei nr. 313, tel.
0724.546.476, e-mail:cornelovidiu.badea@gmail.com.
485
AFASES - 2008 -
Abstract. This work presents the problems related to power quality in the aircraft electric
systems, the problems related to internal electromagnetic compatibility (wire-to-wire
coupling) and lightning indirect effects on the quality of the electricity supply in aircraft
electric system. The disturbance's generation and propagation is described for to guarantee
reliability and safety in aircraft performances.
I. Introduction
The use of new materials and the enlargement of electromagnetic environment
underlined the need of paying greater attention to the electromagnetic interference in the
design of modern aircrafts. For an aircraft to fulfill its tasks reliably and satisfactorily, the
following problems, must be solved [3]: Internal electromagnetic compatibility (the avionic
equipment must not interfere with the electric and electronic systems on aircraft); External
electromagnetic compatibility (the satisfactory performance of the electric and electronic
systems must be ensured even when exposed to external radiations); Lightning protection;
Nuclear electromagnetic pulse (NEMP) [4].
The aircraft structure may be built with the following materials: Pure metal
structure: aluminum; Mixed structure: aluminum structure with avionic; access doors or
panels made of carbon fiber; composite (Carbon Fiber Composite, CFC). The materials used in
modern aircrafts are more sensible to electromagnetic environment compared to the ones
used in the past.
The problems related to electromagnetic effects increase very much as time goes on
and technology advances. The problems concerning lightning protection become relevant,
while the ones relating EMC (electromagnetic compatibility) appear to increase less
strongly, the aircrafts are more dependent on the electronics.
486
AFASES - 2008 -
The currents flowing along the wires of a circuit create magnetic fields that couple to
another circuit inducing signals in both of them. The unintended interaction between two or
more circuits via electromagnetic fields can cause interference problems. This phenomenon is
called crosstalk [1]. The task in modeling crosstalk is the prediction of these signals and
whether they will cause the respective loads to malfunction in order to reduce this
phenomenon.
The analytical procedure to predict the induced terminal voltages of the receptor
circuit is reported in [1], [2], [3], [4] with reference to the Multiconductor Transmission Line
(MTL) theory. The induced terminal phasor voltages of the receptor circuit are the sum of a
component due to the mutual inductance between the two circuits (inductive coupling) and the
component due to the mutual capacitance between the two circuits (capacitive coupling).
Generally one component dominates the other.
The inductive coupling dominates capacitive coupling for low impedance loads and
the capacitive loads for high impedance loads. The sensitivity of the coupling depends on
the variations in relative wire position in cable bundles [4], [5].
487
AFASES - 2008 -
When the lightning stepped-leader attaches to one of the streamers emanating from the
aircraft extremities, and enters the aircraft skin, the currents are about a thousand amperes, but
when the lightning channel approaches the ground they may reach 30000 A (return-stroke).
The Lightning effects [6] on aircraft on classified as: Direct Effects, which include the
physical damage at the point of arc attachment, and Indirect Effects, which are
electromagnetically induced by field coupling to wires or avionic equipment. The indirect
effects could also be produced by lightning, which did not directly contact the aircraft, but
these would not be significant.
The Lightning currents induce voltages in aircraft electrical circuits, thus inducing
strong magnetic fields surrounding the conductive aircraft. Some of this magnetic flux may
leak inside the aircraft through apertures such as windows and joints. These internal fields
pass through aircraft electrical circuits and induce voltages between both wires of a two-wire
circuit (differential mode voltages) or between both wire and the airframe (common mode
voltages).
As lightning current flows through the aircraft resistive voltage drops may appear
between circuit wires and the airframe. For metallic airframe made of highly conductive
aluminum these voltages may be significant when the lightning current must flow through
resistive joints. The resistance of titanium and composite materials is ten times and hundred
times that of aluminum, so resistive voltages in future aircraft employing these materials may
be significantly higher. The newer materials do not provide the degree of electromagnetic
shielding in the 10 kHz to 1 MHz frequency band as has been afforded by conventional allaluminum aircraft.
Increasing use of plastic of solid state electronics, further miniaturization of solid state
electronics, and greater dependence of electronics to perform flight critical functions, greater
congestion in terminal airways requiring more frequent flight through adverse weather
conditions at altitudes where lightning strikes frequently occur (6000 m), could increase the
problems due to indirect effects.
Another phenomenon that can affect an aircraft is the NEMP, which produces an
electromagnetic field fifty times greater than the one produced by lightning and a thousand
times faster to reach its peak value. Protection against its effects isn't much different from
that used against lightning effects.
488
AFASES - 2008 -
Every electric and electronic system essential for the aircraft safety must guarantee
their performances even during adverse conditions such as lightning. Protection against
lightning indirect effects can be applied either by designing the aircraft structures to minimize
wiring susceptibility to the lightning induced effects, or by applying protective methods to the
avionic equipment.
The lightning shield can usually consist of an overall braided shield on a group of
conductors, being grounded to the aircraft structure at least at the ends.
489
AFASES - 2008 -
Concerning the induced voltage on the ground conductor, that reaches the value of 400
V, it must be considered its level of disturbance associated to equipment and electronic
controlled devices.
This kind of interference is more relevant compared to the lightning effects because is
a permanent phenomenon while the second is transient and does not often occur.
490
AFASES - 2008 -
V. Conclusions
This work presents the problems related to internal electromagnetic compatibility
(wire-to-wire coupling) and lightning indirect effects on the quality of the electricity supply in
aircraft power systems.
The disturbance generation and propagation are briefly recalled and a protection
system design is described in order to guarantee reliability and safety in aircraft performances.
References
[1] Clayton, R.P., Cables and crosstalk, International Conference on Electric power quality
on aircraft: analysis and propagation of disturbances, Proceedings of IEEE (Milano)
2001, Vol. 1, Page(s): 222 - 227.
[2] Middlebrock, R.D., Input Filter Consideration in Applications of Switching Regulators,
Proceedings of the IEEE Industry Applications Society, 2000, pages 91-97.
[3] Braidotti, D., Zanini A., Power Quality studies on aircraft electric systems, International
Conference on Harmonics and Power Quality Orlando (Florida, USA), 2000.
[4] Braidotti, D., Cables and electromagnetic compatibility into aircraft, International
Conference on Insulated Power Cable, Versailles, France, 1999.
[5] Albert, E.H., Stability and Dynamics of Power Systems with Regulated Converter,
Proceedings of the IEEE Symposium on Circuits and Systems, Society, 1998, pages 143145.
[6] Jaeger, D., Increasing significance of electromagnetic effects in modern aircraft
development, International Aeronautical Conference on Research and Technology,
Melbourne, 1998.
[7] Airbus A320, Volare Airlines, Flight Operation Manual.
*Eleonora Darie, Conf. univ. dr. ing. Univ. Tehnic de Construcii Bucureti, Catedra de
Electrotehnic, Bucureti, Pache Protopopescu nr. 66, e-mail: eleonora_darie@yahoo.com
**Emanuel Darie, Conf. univ. dr. ing. Lt. Col., Academia de Poliie Alexandru Ioan Cuza,
Bucureti, Privighetorilor nr.1, tel. 021 2304833, e-mail: edarie@yahoo.com
***Garibald Popescu, Lector univ. dr. ing. Lt. Col., Academia de Poliie Alexandru Ioan
Cuza, Bucureti, Privighetorilor nr.1, tel. 021 2304833.
491
AFASES - 2008 -
492
AFASES - 2008 -
493
AFASES - 2008 -
abur la temperatur ridicat. Cel mai important aspect este acela c presiunea maxim n eava
de cldur trebuie limitat prin restricionarea atent a masei totale de lichid de lucru.
Controlul activ a fluxului de cldur poate fi eficientizat prin adugarea unui rezervor
de lichid cu volum variabil la seciunea de evaporare. evile de cldur cu conductibilitate
variabil utilizeaz un rezervor imens de gaz inert imiscibil ataat la seciunea de condensare.
Varierea presiunii rezervorului de gaz schimb volumul gazului ncrcat n condensator care,
n schimb, limiteaz zona disponibil pentru condensarea aburului. Astfel, o mai mare palet
de fluxuri de cldur i variaii de temperatur poate fi ajustat printr-un singur proiect.
O eav de cldur cu rezervor care nu are conexiuni capilare la eava de cldur de la
captul evaporatorului poate fi utilizat, de asemenea, ca diod termic. Aceast eav de
cldur va transfera cldur ntr-o singur direcie, acionnd ca un izolator pentru ceilali.
evile de cldur plane au aceleai componente de baz precum evile de cldur
tubulare. Aceste componente sunt un recipient gol sigilat ermetic, un fluid de lucru i un
sistem de recirculare capilar cu circuit nchis.
n comparaie cu o eav de cldur tubular de o singur dimensiune, lrgimea unei
evi de cldur de dou dimensiuni permite o secionare corect pentru fluxul de cldur chiar
i cu un dispozitiv foarte subire. Aceste evi de cldur subiri i plane se gsesc n cadrul
calculatoarelor. Se pot produce evi de cldur subiri de 0.5 mm.
evile de cldur utilizeaz rcire de evaporare pentru a transfera energia termic de
la un punct la altul prin evaporarea i condensarea unui lichid de lucru sau de rcire. evile de
cldur se bazeaz pe diferena de temperatur dintre capetele evii i nu pot scdea
temperaturile la niciunul din capete mai mult dect temperatura ambiental (prin urmare,
acestea tind s egalizeze temperatura din eav).
Atunci cnd unul din capetele evii este nclzit, lichidul de lucru din interiorul evii de
la acel capt se evapor i mrete presiunea aburului din interiorul cavitii evii de cldur.
Cldura latent de evaporare absorbit de vaporizarea fluidului de lucru reduce temperatura la
captul fierbinte al evii.
Presiunea aburului asupra fluidului lichid fierbinte la captul fierbinte al evii este mai
mare dect presiunea aburului de echilibru asupra fluidului de lucru de condensare la captul
mai rece al evii, iar aceast diferen de presiune conduce la un transfer rapid de mas ctre
captul de condensare unde aburul n exces se condenseaz, i elibereaz cldura latent i
nclzete captul rece al evii. Gazele necondensatoare (de exemplu, cele cauzate de
contaminare) din abur ntrzie fluxul de gaz i reduc eficiena evii de cldur, n special la
tempearturi sczute, unde presiunile aburilor sunt sczute. Viteza moleculelor dintr-un gaz
este aproximativ egal cu viteza sunetului, iar n absena gazelor necondensatoare, aceasta
este cea mai ridicat vitez cu care acestea ar putea trece prin eava de cldur. n practic,
viteza aburului prin eava de cldur depinde de rata de condensare la captul rcit.
Ulterior, fluidul de lucru condensat curge napoi nspre captul fierbinte al evii. n
cazul evilor de cldur orientate vertical, fluidul poate fi mutat de fora de gravitate. n cazul
evilor de cldur care conin fitile, fluidul este ntors de aciunea capilar.
Atunci cnd de fabric evile de cldur, nu este necesar cearea unui vid n eav.
Trebuie doar fiert fluidul de lucru n eava de cldur pn ce aburul rezultat epureaz gazele
necondensatoare din eav i apoi sigileaz captul.
O trstur interesant a evilor de cldur este temperatura peste care acestea sunt
eficiente. La prima vedere, putem suspecta c o eav de cldur umplut cu ap ar funciona
doar atunci cnd captul fierbinte atinge100 C i apa fierbe, declannd transferul de mas
care reprezint secretul evii de cldur. Totui, punctul de fierbere al apei depinde de
presiunea la care este inut. ntr-o eav evacuat, apa va fierbe chiar la 0 C. De aceea,
transferul de cldur va ncepe atunci cnd captul fierbinte este mai cldu dect captul rece.
494
AFASES - 2008 -
495
AFASES - 2008 -
evile de cldur sunt folosite pentru a disipa cldur n Sistemul de Conducte TransAlaska. Cldura produs prin friciune i turbulen n uleiul aflat n micare ar conduce n jos
picioarele de susinere a evii i ar topi permafrostul care le fixeaz. evile de cldur cu
radiatori la vrf sunt utilizate pe fiecare picior pentru a le menine reci astfel nct s nu
topeasc permafrostul i s lase conducta s cad.
n aplicaiile de nclzire solar termic a apei, un tub colector golit poate avea pn la
40% mai mult eficien n comparaie cu tradiionalii nclzitori solari de ap "plani i
drepi". Colectorii de tuburi evacuai elimin nevoia de a aduga aditivi anti-nghe pe msur
ce vacuumul ajut la prevenirea pierderilor de cldur. Aceste tipuri de nclzitori solari
termici de ap sunt protejai mpotriva ngheului pn la mai mult de -3 C i sunt utilizai n
Antarctica pentru nclzirea apei.
4. Limitri
evile de cldur trebuie adaptate la anumite condiii de rcire. Alegerea materialului,
a dimensiunii i a rcitorului evii influneaz mpreun temperaturile optime la cre
funcioneaz evile de cldur.
Atunci cnd este nclzit peste o anumit temperatur, ntregul fluid de lucru din eava
de cldur se va evapora, iar procesul de condensare nu mai are loc. n asemenea condiii,
conductibilitatea termic a evii de cldur este eficient redus la proprietile de transmitere
de cldur ale metalului su solid care se autoizoleaz. Dat fiind c majoritatea evilor de
cldur sunt fabricate din cupru (metal cu conductibilitate ridicat de cldur), o eav de
cldur supranclzit va continua, n general, s conduc cldura la aproximativ 1/80 din
conductibilitatea original.
n plus, sub o anumit temperatur, fluidul de lucru nu se mai evapora deloc, iar
conductibilitatea termic va fi redus la nivelul celei de izolare a metalului solid. Unul din
criteriile eseniale pentru selectarea fluidului de lucru este gradul de temperatur operaional
dorit al aplicaiei. Limita mai sczut de temperatur apare de obicei la cteva grade deasupra
punctului de nghe al fluidului de lucru.
5. Rcirea componentelor din cadrul notebook-urilor folosind evile de cldur
Vremurile cnd evile de cldur erau folosite majoritar pentru egalizarea temperaturii
n tehnologia spaial i satelii au luat sfrit. n zilele noastre, evile de cldur se utilizeaz
cel mai frecvent n tehnologia calculatoarelor, pentru console de jocuri i chiar integrate n
rcitoarele normale pentru procesoare.
Un sistem termic de evacuare a caldurii slab proiectat va declana temperaturi de
operare ridicate pentru echipamentele electronice, ducnd la numeroase poteniale probleme
496
AFASES - 2008 -
sau la eecuri premature. De aceea, este necesar intensificarea termic adiional pentru
disiparea cldurii.
Cel mai cunoscut modul de rcire pentru un procesor este format din radiator i un set
de ventilatoare extinse cu rol de disipare a cldurii.
Cnd e vorba de rcirea unui PC, "eav de cldur" este un cuvnt cu rezonan. O
eav de cldur este un dispozitiv cu conductibilitate termic foarte ridicat, utilizat pentru a
transporta cldur. Pe msur ce lichidul se evapor, energia sub form de cldur trebuie
luat din mediu. De aceea, un lichid care se evapor va rci zona ce l nconjoar. Acesta este
modul n care o eav de cldur rcete efectiv sursa de cldur. Totui, acesta nu scap de
cldur ntruct cldura se deplaseaz cu aburul. Ca int pentru transportul de cldur, eava
de cldur trebuie s fie rcit, folosind de exemplu o chiuvet de cldur. n acest caz, are
loc efectul invers: lichidul se condenseaz, i deci, emite cldur. Folosind acest efecte, este
posibil s construim evi de cldur cu conductibilitate termic de mii de ori mai mare dect o
pies din cupru de aceeai dimensiune.
497
AFASES - 2008 -
structur de fitil pe interiorul evii de cldur. eava de cldur are avantajul de a avea
rezisten termic foarte sczut, iar performanele ei sunt insensibile la lungimea acesteia. n
consecin, poate fi aranjat n multe forme pentru a primi diferite configuraii. A fost
implementat cu succes n aplicaiile pe laptop. Mai jos se poate vedea o eav de cldur
conectat la un sistem de rcire de laptop.
Dat fiind c nu exist pri care se mic, nu e necesar service-ul i nu este nimic de
spart. Unii sunt ngrijorai de posibilitatea ca fluidul s se scurg din eava de cldur n
componentele electronice. Cantitatea de fluid dintr-o eav de cldur de acest diametru este
foarte mic. ntr-o eav de cldur corect proiectat, apa este reinut total n structura
capilar de fitil i exist mai puin de 1 atmosfer de presiune. Dac integritatea vasului unei
evi de cldur ar fi vreodat compromis, aerul s-ar prelinge n eava de cldur, n loc ca
apa s se scurg afar. Apoi fluidul s-ar evapora ncet pe msur ce atinge punctul de fierbere
atmosferic.
evile de cldur au demonstrat c reprezint un mijloc excepional prin care se
furnizeaz controlul termic al desktop-lui i al sistemelor laptopului. evile de cldur pot
mica i disipa cldura generat de CPU n mod selectiv prin sistem, fr a afecta
componentele sensibile la temperatur. evile de cldur cu puini wai (sub 20 wai) au
plcue de alimentare standard la eava de cldur. Conectarea la schimbtorul de cldur prin
eava de cldur poate avea orice numr de configuraii pentru a nlesni plasarea
componentelor, nivele multiple de putere i opiuni de ventilator.
498
AFASES - 2008 -
[2] Gaugler, Richard (1944), Heat Transfer Devices, Dayton, Ohio: U.S. Patent Office, pp. 4,
2350348.
[3] Jim Danneskiold, Los Alamos-developed heat pipes ease space flight. Los Alamos News
Release, April 26, 2000.
[4] Grover, G.M., T. P. Cotter, and G. F. Erickson (1964). "Structures of Very High Thermal
Conductance". Journal of Applied Physics 35 (6): 1990-1991
[5] Heat Pipe research at LANL
Petre-Daniel Dinescu, Mr.drd.ing., Agentia pentru Sisteme si Servicii Informatice Militare, Bucureti, B-dul
Drumul Taberei nr.7-9, sector 6, tel. 021 319 59 59, 0745 590 865, e-mail: ddp_ro@yahoo.com
Dinu Comnescu, confereniar univ. dr. ing., Universitatea POLITEHNICA Bucureti, Catedra de Mecanic
Fin, Bucureti, Splaiul Independenei nr.313, tel. 021 316 95 84, 0744 580 541, e-mail:
dinucomanescu@yahoo.com
Adriana Comnescu, profesor univ. dr. ing., Universitatea POLITEHNICA Bucureti, Catedra de Teoria
Mecanismelor i a Roboilor, Bucureti, Splaiul Independenei nr.313, tel.021 402 9 632, 021 316 95 84, e-mail:
adrianacomanescu@yahoo.com
499
AFASES - 2008 -
500
AFASES - 2008 -
Fig. 1 Ilustrare schematic a efectului piezoelectric: (a) n stare iniial; (b) generarea
curentului electric prin comprimare (efect piezoelectric direct); (c) dilatare la aplicarea
curentului electric (efect piezoelectric invers) .
Prin aplicarea tensiunii mecanice, se produce o separare a centrelor de greutate a
sarcinilor electrice, negative i pozitive, ceea ce d natere unui dipol electric, caracterizat
printr-un moment electric dipolar.
n efectul piezoelectric direct aplicarea unei tensiuni mecanice conduce la
redistribuirea sarcinilor electrice n volum, rezultnd o polarizare electric volumic i
implicit o sarcin electric indus pe suprafa.
Efectul piezoelectric invers este produs prin aplicarea unui cmp electric cristalului,
avnd ca rezultat deformarea cristalului sau apariia unei fore (cauza este electric, efectul
este mecanic).
Dintre cele 32 de clase de cristale existente numai 20 pot prezenta efect piezoelectric.
Majoritatea materialelor piezoelectrice importante sunt i feroelectrice. Materialele
feroelectrice prezint o faz de nalt simetrie cristalin care apare deasupra unei temperaturi
critice (temperatur Curie = TC) i care nu se poate polariza spontan. Faza de la temperatur
sczut prezint n general efect piezoelectric. Dup descoperirea piezoelectricitii, s-a
considerat mult vreme c efectul piezoelectric este limitat doar la monocristale, deoarece
materialele policristaline au gruni orientai n mod ntmpltor, astfel nct efectele lor se
anuleaz reciproc, rezultnd un efect global nul. n deceniul al-V-lea al secolului XX s-au
descoperit materialele piezoceramice, cu constante dielectrice, K, foarte ridicate. Constanta
dielectric este definit drept raportul dintre permitivitatea electric a dielectricului respectiv
() i cea a vidului (0):
(1)
(2)
k'=
n care:
501
AFASES - 2008 -
direcia lui X i Y. Modul de comportare contrar apare cnd materialul este solicitat n planul
X-Y.
502
AFASES - 2008 -
(3)
(4)
(5)
(6)
(7)
undeY este modulul de elasticitate al lui Young, iar mrimile electrice satisfac relaia:
D = E = o rE = oE + P
(8)
503
AFASES - 2008 -
Nanometrologie, msurarea
sistemelor, neutralizarea vibraiilor;
dimensiunilor
critice,
microlitografie,
verificarea
504
AFASES - 2008 -
surselor acustice aeriene (sisteme de nalt frecven pn la 106 Hz). Datorit impedanei
mecanice foarte mare, pentru anumite cmpuri de deplasri senzitivitatea lor este redus.
Bibliografie:
1] BARLEA Nicolae-Marius, Fizica senzorilor; Cluj-Napoca: Editura Albastr, 2000;
[2] VASIU Dan, ROCA Ileana, OLTEANU Ciprian, Senzori i traductoare : Vol. I:
Traductoare giroscopice; Braov: Reprografia Universitii Transilvania, 1994;
[3] BUCAN Constantin, DEMIAN Traian, DUMITRU Adrian, Senzori i traductoare pentru
roboi; Medro, 1996;
[4] PANTELIMON Brndua, ILIESCU Constantin, ANGHEL Florin; Senzori i traductoare:
Experimentri/ Bucureti: Tritonic, 1995;
[5] AGOSTON Katalin, Senzori i traductoare: ndrumar de laborator, Targu-Mures:
Universitatea Petru Maior, 2005;
[6] AGOSTON Katalin, Senzori i traductoare: Note de curs, Targu-Mure: Universitatea
Petru Maior, 2001;
[7] STROE Ioan, Senzori i traductoare: Braov: Universitatea "Transilvania" din Braov,
2000;
[8] TODOS Petru, GOLOVANOV Carmen; Senzori i traductoare: Chiinau: Editura
Tehnica-Info, 1998.
505
AFASES - 2008 -
506
AFASES - 2008 -
Fig. 2. AFM-urile pot s msoare elasticitatea prin apasarea cu vrful pe eantion i msurarea
parghiei optice.
507
AFASES - 2008 -
(a)
(b)
Fig. 5 imagini AFM pentru un strat Si/SiO2/Py(5 nm); (a) suprafaa probei; (b) aspectul
tridimensional al suprafeei probei [9]
Suprafaa scanat 2065x2043 nm2
508
AFASES - 2008 -
(a)
(b)
Fig. 6 imagini AFM pentru un strat Si/SiO2/Py(100 nm); (a) suprafaa probei; (b) aspectul
tridimensional al suprafeei probei [9]
Suprafaa scanat 17974x17996 nm2
Valoarea maxim a rugozitii: 133,59 nm
Valoarea ptratic medie a denivelrilor (rms): 15,25 nm
Valoarea medie a rugozitii: 12,84 nm
Fig. 7 imagine AFM-3D a suprafeei unui ML Si/SiO2/[Py(10 nm)/Cu(4 nm)]9- Py(10 nm)
depus prin pulverizare magnetron r.f. [9]
Valoarea medie a rugozitii suprafeei: 6,43 nm
Mixturi colisionale - regiuni mixte Py-Cu la interfee
- grosimea efectiv a stratului magnetic micorat
fluid
vacuum
Special gas
AFM
TEM
OPTICAL
SEM
509
AFASES - 2008 -
optica
SEM
0 20
100
200
400
300
nm
500
optica
SEM
0 20
100
200
300
400
500
nm
510
AFASES - 2008 -
7. Meyer, G. and Amer, N.M. (1988) Novel optical approach to atomic force microscopy.
Appl. Phys. Lett. 53(12), 1045-1047
8. Binnig, G., Quate, C.F., and Gerber, Ch. (1986) Atomic force microscope. Phys. Rev. Lett.
56(9), 930-933
9. www. agilent/find/afm
511
AFASES - 2008 -
I. Introduction
This paper presents a cryptographically secure identity certification system. This is a
modified ID which contain the photo of the pretended person, an arbitrary textual message, and a
2D color bar-code which encodes an RSA signature of the message hash and the compressed
representation of the face encompassed by the photo.
512
AFASES - 2008 -
The signature is created using the private key of the party issuing the ID. Verification is
performed by a simple, intelligent, and off-line scanning device that contains the public key of
the issuer. Unlike typical identification methods, which require a person to have something--an
identification card, a personal identification number (PIN), or password--biometric information is
part of a person. Since biometrics are so closely bound to a person, they are thought to be more
reliable and not easily forgotten, lost, stolen, falsified, or guessed. This is because a biometric
identifier relies on unique biological information about a person.
Since 2001, there has been a great deal of interest in using biometrics for verification of
identity. The interest is particularly acute in the areas of visa and immigration documentation and
government-issued identification card programs.
Biometric passports must be interoperable, which means that data must be written on the
smart card (chip) in such a way that they can be read by the border authority of any foreign
country, regardless of what hardware and software is used at their border crossing points. In order
to ensure interoperability, the contents of the data on the chip are strictly defined. Only the data
that are also visibly printed in the passport (personal name, nationality, birth date, gender,
country of issue, passport number and expiry date) are stored on the chip, as well as the
photograph that is in the passport. An european biometric pasport example is presented in the
Figure 2.
open biometric
passport with the
data page with
invisibly
stored
contactless chip.
OCR-B
fonts
enable access to
data stored on the
chip
513
AFASES - 2008 -
There are a set of procedures to assure the authenticity of such documents such is
holograms, micro-printing, special print paper, but modern printing technologies have made highquality printing devices quite inexpensive. The availability of such printers has rendered forging
most personal ID documents a relatively simple task with results often perceptually comparable
to the originals.
On the other hand, authenticating the stored photograph and the textual message can be
made highly reliable using the combined technology of face recognition and public cryptography.
Typically, the stored photograph and the textual message are hashed and then signed using the
private key of the user at enrolment.
Authentication is performed using the public key of the user by a verification device,
which also must display the signed data.
514
AFASES - 2008 -
detail to which an adversary, whose photo has not been taken for the ID, must resemble the facial
features of the person on the authentic face.
Consider f a vector of weights that describe the projection of the original face into the
face space. Messages f and z are merged into a message m using a reversible operator such is
stream encryption, where message m is resulted encrypted using z as a key for f.
Note that the purpose of using an encryption function is not related to security, it is rather
a way to make the final message m dependant upon f.
Then, m is divided into 1023-bit parts, where each of them is signed with the private
1024-bit RSA-key. The signature is encoded and printed as a 2D bar-code onto the ID.
At verification the verifyer scans all three printed components: the photo, the text, and the
bar-code. The bar-code is decoded into the originally printed signature.
The scanned textual data is also converted into a text-string using reliable optical
character recognition. Next, after encrypting the signature with the public RSA key of the issuer,
the verifier obtains the signed message m.
After the verifier hashes the text to obtain z, it decript m using z as key. Then, the verifier
decompresses f into a subimage of the original photo that contains the facial features.
Finally, the verifier quantifies the level of similarity between the decompressed and
scanned face. If the two images are similar within the maximum tolerable compression-printscan
noise, only then the verifier reports an authentic ID.
The security of system stems from the fact that changing a single bit of the textual
message or altering the photo beyond the compression-print-scan noise causes a global change in
the bar-code that appears to be random without the knowledge of the issuers private key.
References
[1] Cryptographically secure identity certificates, Darko Kirovski and Nebojsa Jojic, 2004
[2] Eigenfaces for recognition, Matthew Alan Turk and Alex Pentland, 2004
[3] R.L. Rivest, et al. A method for obtaining digital signatures and public-key cryptosystems.
CACM, vol.21, no.2, pp.1206, 1978.
[4] A.J. Menezes, et al. Applied Cryptography. CRC Press, 1996.
[5] P. Viola et al. A unified framework for face datection and recogntion. Learning workshop,
Snowbird, 2002.
[6] L. OGorman, I. Rabinovich. Secure identication documents via pattern recognition and
public-key crypto. PAMI, pp.1097102, 1998.
**
515
AFASES - 2008 -
Abstract: This paper describes a device capable of throwing metal rings. The
apparatus was invented by the American engineer and inventor Elihu Thompson (18531937)
to demonstrate his pioneering research in alternating current and high frequency. It is
possible to transform this apparatus into powerful electromagnetic gun different by railgun
and coilgun.
Este posibil obinerea levitaiei magnetice cu ajutorul cmpului magnetic oscilant.
Acest fenomen se bazeaz pe inducerea curenilor turbionari ntr-un conductor aflat ntr-un
cmp magnetic oscilant. Cmpul magnetic oscilant poate fi obinut prin dou metode. Prima
metod const n deplasarea unui conductor ntr-un cmp magnetic constant n timp. Aceast
metod este utilizat practic n contoarele de energie electric unde un disc de aluminiu este
rotit ntr-un cmp magnetic constant generat de un magnet permanent. Curenii turbionari
indui n discul de aluminiu frneaz micarea de rotaie a acestuia. Cu ct viteza de rotaie a
acestuia este mai mare, curenii turbionari sunt mai mari iar forele care se opun micrii sunt
mai mari.
A doua metod de obinere a cmpurilor magnetice oscilatorii const n utilizarea unei bobine
alimentat n curent alternativ. Prin plasarea unui conductor n cmpul magnetic generat de
bobin, n acesta se induc cureni turbionari. Acetia creeaz un cmp magnetic care se opune
cmpului magnetic generat de bobin.
Pentru demonstrarea acestui fenomen, inginerul i inventatorul american Elihu Thompson
(18531937) a realizat un dispozitiv n anul 1887. Acesta era compus dintr-o bobin
alimentat n curent alternativ, un miez feromagnetic precum i un conductor sub form de
inel dispus deasupra bobinei ca n figura 1.
516
AFASES - 2008 -
517
AFASES - 2008 -
5
Curentul I
0
0
-5
-10
518
10
t.e.m.
AFASES - 2008 -
Fig. 3.4. Variaia tensiunii electromotoare din inel n funcie de curentul I de alimentare a
solenoidului
Se observ din figura 4 c tensiunea electromotoare indus n inelul conductor se
opune variaiei cmpului magnetic al solenoidului. Dar deoarece rezistena electric a inelului
este mic, pe circumferina acestuia ia natere un curent. Acesta ar trebui s fie n opoziie de
faz cu tensiunea electromotoare indus ceea ce ar face ca jumtate de timp inelul s fie atras
de bobin iar jumtate de timp acesta s fie respins de bobin. Aici intervine inductana L a
inelului care ntrzie n timp curentul prin inel.
Pentru calculul defazajului n timp a curentului indus se utilizeaz relaia:
L
R
Unde:
-
= tg
5
Curentul I
0
t.e.m .
Curentul indus
-5
-10
519
AFASES - 2008 -
520
AFASES - 2008 -
521
AFASES - 2008 -
522
AFASES - 2008 -
Laurian Gherman, Mr. ing. drd., Academia Forelor Aeriene, Mihai Viteazu nr. 160, tel 0268 423421,
e-mail: laurian.gherman@afahc.ro
523
AFASES - 2008 -
Abstract: The use of electromagnetic energy to launch and kinetic energy to achieve
lethality affords the opportunity to entirely remove both propellants and explosives from the
weapon system, significantly increasing system and survivability. An electromagnetic rail gun
uses a large Lorentz force to fire a projectile. The classic configuration uses two conducting
rails with armature that fits between and closes the circuit between the two rails. A large
current is sent through the rails from one side creating a magnetic field that is perpendicular
to the current in the armature and forces the armature to move down the rails away from
whichever end the current is being applied. Since the force pushing apart the rails is large as
well, they must be built so as to withstand the force developed during firing.
524
AFASES - 2008 -
cretem raza de aciune sau viteza proiectilului la gura evii, vom constata c ajungem la un
nivel peste care nu putem trece.
Armele bazate pe energia chimic nu pot accelera un proiectil astfel nct viteza la
gura evii s fie mai mare de 1,5 km/s iar distana de tragere s fie mai mare de 50 km. Pentru
mrirea distanei de tragere s-a gsit soluia realizrii unui proiectil propulsat de un motor de
rachet. Aceast soluie mrete distana de tragere pn la 100 km. Peste aceast distan
costurile cresc peste costurile utilizrii unei rachete. Astfel pentru distane mai mari de 100 de
km este mult mai eficient utilizarea rachetelor. Dar rachetele sunt foarte scumpe. Avantajul
tunurilor const n faptul c pot trage proiectile mai mici ca dimensiuni i mai ieftine ca i
costuri n comparaie cu rachetele.
Pentru a putea egala performanele rachetelor la un cost asemntor tunurilor este
necesar abordarea problemei dintr-o nou perspectiv. Tunul electromagnetic ofer o nou
direcie de dezvoltare a armelor.
Dac lum n considerare faptul c cea mai rapid i mai puin costisitoare cale pentru
mrirea distanei de tragere este creterea vitezei proiectilului la gura evii, atunci tunul
electromagnetic este cea mai bun alegere. Deoarece tunul electromagnetic utilizeaz energia
electric pentru propulsarea proiectilului se elimin pericolul exploziei pulberii care
propulseaz proiectilele convenionale. Acest lucru prezint un alt mare avantaj, deoarece
sistemul de propulsie se afl pe tun i nu la fiecare proiectil, se pot nmagazina mai multe
lovituri pentru acelai volum al magaziei. Dar odat cu creterea vitezei proiectilului se obin
o serie de alte avantaje cum ar fi reducerea timpului de zbor pn la int dar i creterea
energiei de impact ceea ce face inutil prezena unei ncrcturi explozive n proiectil. Se
poate elimina astfel utilizarea materialelor explozive ceea ce determin eliminarea msurilor
de siguran ce se aplic n lucrul cu astfel de materiale.
Dac comparm performanele sistemelor convenionale cu cele ale tunului
electromagnetic se observ foarte clar avantajele acestuia.
Cele mai puternice tunuri navale aflate n exploatare pot trage proiectile a cror energie la
gura evii ajunge pn la 10 MJ (megajouli). Pentru mrirea distanei de tragere s-a adoptat
soluia utilizrii unor motoare de rachet cu combustibil solid. Aceste proiectile sunt lansate
cu ajutorul unor evi de tun convenionale i sunt ghidate la int cu ajutorul sistemului global
de poziionare (GPS).
Crme
Motor de rachet
ncrctur exploziv
Focos
Baterie
Dispozitiv de ghidare
525
AFASES - 2008 -
526
AFASES - 2008 -
La ora actual cele mai avansate cercetri pentru dezvoltarea acestei tehnologii a
tunului electromagnetic sunt realizate avndu-se la baz conceptul denumit railgun sau n
traducere tun cu ine. n urmtorul subcapitol voi face o scurt prezentare a acestui dispozitiv.
2. Railgun
Dintre toate sistemele de accelerare a proiectilelor utiliznd fore electromagnetice,
sistemul denumit railgun este cel mai avansat la ora actual.
Denumirea de railgun provine din faptul c sistemul este alctuit din dou conductoare
electrice care au i rol de ghidare. Aceste dou ine conductoare sunt fixate rigid la o distan
constant una fa de cealalt. Circuitul electric format de aceste dou ine se nchide printrun conductor mobil care se poate deplasa n lungul acestora. Acest conductor poate fi
proiectilul nsui sau un sabot care mpinge proiectilul. De asemenea conductorul poate fi un
material conductor solid sau plasm. Plasma poate fi obinut prin utilizarea unui conductor
subire de metal care se nclzete rapid la trecerea curentului electric. Bineneles c n acest
caz trebuie luate o serie de msuri care s previn distrugerea proiectilului. Sistemul de
rigidizare a celor dou ine trebuie s formeze o eav cu diametrul interior egal cu diametrul
proiectilului.
Pentru obinerea performanelor prezentate n primul capitol circuitul format de cele
dou ine fixe i conductorul mobil trebuie alimentat cu un curent de intensitate foarte mare
pentru o perioad scurt de timp. Acest impuls de curent continuu poate fi asigurat de ctre
baterii pe baz de acid, care pot furniza cteva mii de amperi pentru scurt timp. Aceast
soluie nu poate fi utilizat pentru alimentarea unui tun, deoarece sunt necesare un numr
mare de astfel de baterii. O alt soluie ar fi utilizarea unor condensatoare care pot stoca o
cantitate foarte mare de energie. Utilizarea lor n scopuri militare este limitat deoarece
utilizeaz ntreaga energie nmagazinat doar pentru tragerea unui singur proiectil. Soluia
utilizrii mai multor baterii de condensatoare pentru a putea trage mai multe proiectile nu este
viabil datorit volumului mare ocupat precum i a timpului mare de ncrcare.
O soluie aflat nc n studiu este utilizarea generatoarelor de impulsuri de curent
alternativ. Acestea nmagazineaz energia inerial prin rotirea unui rotor la viteze mari cu
ajutorul unui motor electric. Aceast energie poate fi eliberat sub form de impulsuri de
curent alternativ. Deoarece generatorul de impulsuri produce curent alternativ este necesar
utilizarea unui convertor de curent care s asigure curentul continuu necesar alimentrii
tunului electromagnetic. n funcie de mrimea i viteza de rotaie a rotorului, se poate
nmagazina suficient energie pentru a trage mai multe proiectile.
527
AFASES - 2008 -
r
r
r
F = q E + qv B
Unde:
-
(1)
r
F este fora care acioneaz asupra sarcinii q [N/C];
r
E este intensitatea cmpului electric [N/m];
r
B este inducia magnetic [T];
v este viteza particulei [m/s];
q sarcina particulei [C];
n cazul prezentat q reprezint curentul electric din sabotul ce face legtura dintre cele
dou ine. Fora necesar pentru accelerarea sabotului este dat de al doilea termen al ecuaiei
1.
tiind c:
m
C
q[C ] v[ ] = i[ ] l[m] (2)
s
s
Rezult c:
r
r
r
F = q v B = i l B
(3)
528
AFASES - 2008 -
i2
i2 r + d
1
1
F=
(
)d =
ln(
)
+
r (4)
4 (r + d ) (r + )
2
0
d
Unde:
-
Dei aceast ecuaie are limitrile ei datorit unor aproximri, totui ea se dovedete
util deoarece se poate calcula curentul necesar pentru producerea unei fore F date. Aceast
for poate fi mrit dac se utilizeaz materiale cu permeabilitatea magnetic ridicat, dac
distana d dintre ine este maximizat i nu n ultimul rnd dac raza inelor r este minim.
Pentru calculul induciei magnetice B plecm de la ecuaia 3 unde nlocuim l cu d,
distana dintre ine sau mrimea sabotului.
r
i
F
r+d
=
B=
ln(
)
i d 2 d
r
(5)
529
AFASES - 2008 -
530
AFASES - 2008 -
Laurian Gherman, Mr. ing. drd., Academia Forelor Aeriene, Mihai Viteazu nr. 160, tel 0268 423421,
e-mail: laurian.gherman@afahc.ro
531
AFASES - 2008 -
532
AFASES - 2008 -
protocoale complexe i cu consum mare de putere [3]. O imagine mai detaliat asupra acestor
standarde este oferit n Tabelul 1.
TECHNOLOGII
ZigBee
Wi-Fi
Bluetooth
10 100
1-100
1-10
100 1000
0.5 5
17
Distana de
transmisie (metrii)
Durata bateriei
(zile)
Dimensiunea
reelei (noduri)
65000
32
Aplicaii
Monitorizare
i Control
Web, Email,
Video
nlocuirea
cablajului
4 32
1000
250
20 250
11000
54000
720 1000
2400/ 5000
2400
Dimensiunea stivei
(KB)
Viteza datelor (kb/
s)
Spectrul de
frecven (MHz)
Costuri
868/ 915/
2400
MICI
MEDII
MICI
Puterea consumata
MIC
MEDIE
MIC
Caracteristici de
succes
Fiabilitate,
Putere, Cost
efectiv
Viteza de
date,
Flexibilitate
Cost,
Comoditate
533
AFASES - 2008 -
534
AFASES - 2008 -
535
AFASES - 2008 -
536
AFASES - 2008 -
transmisii eficiente n reeaua wireless din punct de vedere al influenei nefaste a factorilor de
degradare.
Odat ce nodurile sunt amplasate corespunztor, reeaua ZigBee se poate implementa,
prezentnd caracteristici de fiabilitate i flexibilitate caracteristice unei reele de tip plas, n
care datele trimise n reea cunosc mai multe salturi pn s ajung la destinaie i o pot lua pe
mai multe ci n funcie de algoritmul de rutare [8]. n cazul n care un nod iese din reea,
reeaua trebuie s se adapteze la aceast situaie, i alte ci pentru transmisia de date sunt
alese, efectele cauzate de absena acelui nod fiind repede compensate. Figura 4 prezint un
scenariu de optimizare a reelei pentru o locaie n care deja au fost efectuate msurtori.
537
AFASES - 2008 -
538
AFASES - 2008 -
Concluzii
Aceast lucrare a prezentat aspectele teoretice i practice legate de propagarea
eficient a datelor n reele wireless din cadrul incintelor i din afara lor folosind tehnologia
ZigBee. S-a analizat calitatea legturii n funcie de trei metrici: RSSI, LQI i PRR. LQI se
dovedete a fi cel mai precis n estimarea performanei i reprezint punctul de pornire n
vederea optimizrii reelei fr fir innd seama de efectele fenomenelor de degradare care
rezult din msurtorile experimentale efectuate.
Dei viteza de transfer a datelor este mic (maxim 250 kbps), totui posibilitatea de a
forma reele fr fir cu senzori rmne valid, deoarece datele de la senzori nu necesit rate
mari de transfer. Securizarea nodurilor centrale se face prin actualizarea n timp real a poziiei
nodului coordonator este posibil prin transmiterea de cadre de semnalizare de la nodurile
nvecinate. Prin corelaia LQI cu distana i prin implementarea modelului matematic descris
de metoda LCP, se estimeaz poziia nodului n cauz ct mai fidel cu realitatea.
Bibliografie
[1]
T. Zahariadis, Evolution of Wireless LAN and PAN Standards, Computer
Standards and Interfaces, vol. 26, pag. 175185, Mai 2004.
[2]
H.S. Kim, J.H. Song, i S. Lee, Energy-efficient traffic scheduling in IEEE 802.15.4
home automation networks, IEEE Transactions on Consumer Electronics, vol. 53, numrul
2, pag. 369374, Mai 2007
[3]
S. Krishnamurthy, Platform-independent software enables true interoperability in
Zigbee Networks, Portable Design., vol.13, numrul 8, pag. 2630, August 2007.
[4]
S. Rao, Estimating ZigBee Transmission in ISM Band, EDN Europe, vol. 52, pag.
4043, Iulie 2007.
[5]
Strassberg, D.: Simple Networks. In: EDN Europe, vol.51, numrul 5, pag.3542, Mai
2006.
[6]
J. Polastre, R. Szewczyk i D. Culler, Enabling Ultra-Low Power Wireless
Research, n Proceedings of the Fourth International Conference on Information Processing
in Sensor Networks: Special track on Platform Tools and Design Methods for Network
Embedded Sensors (IPSN/SPOTS), pag.364369, Aprilie 2005.
[7]
R. Grossmann, Localization in Zigbee-based wireless sensor networks, Technical
Report (German), University of Rostock, Institute MD, Aprilie 2007.
[8]
G. Ferrari, P. Medagliani, S. Di Piazza, i M. Martalo, Wireless Sensor Netowrks:
performance analysis in indoor scenarios, Eurasip Journal on Wireless Communications and
Networking, Ianuarie 2007.
539
AFASES - 2008 -
Abstract:
The paper presents some vulnerability that can appear in the improper usage of public key
cryptographic algorithm RSA (Rivest, Shamir, Adellman). Are presented some bad
implementations of the encryption method, which can conduct to derive the private key from the
public key and some tricks, used in generation of a public/private key, to enable a smart brute
force attack. This paper is intend to be a cryptanalyst point of view in the implementation of
RSA.
I. RSA Algorithm
RSA method is based on unanimously accepted hypotheses, that in some conditions raising
at a power fixed modulo m is a trapped unique function.
Setting up a RSA system follow this steps:
- Choose two different large prime numbers p and q;
- Compute modulus.
n = p q.
(1.1)
(n ) = (p 1) (q 1).
540
(1.2)
AFASES - 2008 -
(1.3)
(1.4)
(1.5)
c = m e mod n.
(1.6)
(1.7)
(1.8)
(1.9)
(1.10)
541
AFASES - 2008 -
(2.1)
p + q = n + 1 - (n).
(2.2)
p q = (p + q) 2 4 n.
(2.3)
From equation (2.3) and (2.4) we can obtain very easy p and q.
2.2 Attack upon RSA in case of small difference between prime numbers p and q
A condition in generating primes number p and q is that, the difference between them, p-q,
to be large, because if a cryptanalist can guess for instance:
n = t 2 s2.
(2.4)
t = ( p + q ) / 2.
(2.5)
s = ( p q ) / 2.
(2.6)
with
by checking all likely values, it follows from that also can be determined. From this
relations p and q can be found, which implies that the system has been broken. Assuming that n
< t2 we can obtain relation t n .
We start with t = n + 1 and increment t by 1 until the value of t2 n is a perfect square s2.
From relations (2.5),(2.8) we can observe that the next value of t2 n is obtain by adding 2t+1.
(t + 1) 2 - n = (t 2 - n) + 2t + 1.
(2.7)
542
(2.8)
AFASES - 2008 -
(2.9)
c 2 = m e 2 mod n.
(2.10)
x 1 e1 + x 2 e 2 = 1,
(2.11)
y = (c 2 - 1) mod n,
(2.12)
m = c1 x 1 y - x 2 ,
(2.13)
- We conclude that the same plaintext should not be crypted twice with the RSA algorithm
543
(2.14)
AFASES - 2008 -
c (k +1) = E ( c (k) ).
(2.15)
(2.16)
3 d < n 14 .
(2.17)
From the theory of continued fractions, it is known that any approximation of a/b e must be
one of the convergents of the continued fraction expansion of a/b.
544
AFASES - 2008 -
This expansion can be obtained from the Euclidean Algorithm, as we describe now. A
(finite) continued fraction is an m-tuple of non-negative integers, say [q1,,qm], which is
shorthand for the following expression:
a
= q1 +
b
q2 +
1
1
q3 + ... +
1
qm
(2.18)
Now, for 1 < j < m, define Cj = [q1, q2, ..., qj]. Cj is said to be the jth convergent of [q1, q2,
..., qm].
Each Cj can be written as a rational number cj /dj, where the Cj 's and dj 's satisfy the
following recurrences:
1
c j = q 1
q c + c
j-2
j j-1
if j = 0
if j = 1
(2.19)
if j 2
0
if j = 0
d j = 1
if j = 1
q d + d
j-2 if j 2
j j-1
(2.20)
References
[1] Friedrick L. Bauer - Decrypted Secrets - Methods and Maxims of Cryptology - 4th Edition,
Springer 2007.
[2] Schneier B., Applied Cryptography, John Wiley & Sons, 1996.
[3] Douglas R. Stinson - Cryptography, Theory and Practice, Third Edition, Chapman & Hall
CRC, 2003
[4] M. J. Wiener, Cryptanalysis of short RSA secret exponents. EUROCRYPT '89 Proceedings.
Lecture Notes
Mihai MARIN, Scientifical Researcher 2nd degree, Military Equipment and Technologies Research Agency,
ROMANIA,marin_mihai2001@yahoo.com
**
Florin RSTOCEANU, Scientifical Researcher, Military Equipment and Technologies Research Agency,
ROMANIA, rastoceanu_florin@yahoo.com
545
AFASES - 2008 -
546
AFASES - 2008 -
0
0
(3)
Pentru aer sau vid valoarea impedanei de und n cmp ndeprtat are valoarea Z0 =
120 377 . Umiditatea influeneaz puternic impedana aerului prin valoarea permitivitii
dielectrice 0.
Diferena de faza ntre componenta electric i cea magnetic a cmpului generat de o
surs este semnificativ doar n zona apropiat, cei doi vectori tinznd s fie n faz odat cu
creterea distanei fa de surs.
Pentru domeniul de frecven vizat, 80 MHz 3 GHz, distana minim fa de un
dipol electric, conform relaiei (1), este cuprins n domeniul (1,6 60) m. Distana mai mare
trebuie asigurat pentru sursele de frecven mai mic. De asemenea, elementele definitorii
ale zonelor caracteristice antenei dipol (scurt) se pstreaz din punct de vedere calitativ pentru
orice alt sistem radiant.
In prezent exist mai multe modaliti de a caracteriza mediul electromagnetic
intenionat: prin descrierea caracteristicilor surselor, prin tipurile de efecte pe care le pot
produce, prin descrierea naturii cmpurilor electromagnetice aplicate sau prin determinarea
intensitii cmpului electric n jurul surselor de perturbaii electromagnetice [9]. In lucrare se
utilizeaz metoda de caracterizare a cmpului electromagnetic prin msurarea componentei
electrice a cmpului electromagnetic.
III. Determinri experimentale
Pentru stabilirea nivelurilor de perturbaii s-a uitlizat un sistem portabil de msurare a
emisiilor electromagnetice, tip Rohde&Schwarz TS-EMF, format din analizor de spectru i
anten izotrop, cu urmtoarele caracteristici:
Domeniu de frecven: 80 MHz 3 GHz;
Sensibilitate: 1 mV/m;
Gam dinamic de msurare: 1 mV/m - 100 V/m;
Detectare valori de vrf i medii.
Veridicitatea datelor obinute este asigurat prin: efectuarea msurtorilor n spaii
deschise pentru a se evita reflexiile de semnal, observarea temperaturii, umiditii i presiunii
atmosferice.
O variaie semnificativ a acestor condiii duce la obinerea de valori eronate, de aceea
calibrarea sistemului n funcie de nivelul de zgomot i verificarea reproductibilitii
msurtorilor este esenial.
547
AFASES - 2008 -
a)
b)
548
AFASES - 2008 -
549
AFASES - 2008 -
550
AFASES - 2008 -
200-230 MHz
550-580 MHz
920-970 MHz
1800-1840 MHz
2110-2180 MHz
Zona urban
supravegheat
Zona A
Zona B
Zona C
Zona D
Zona A
Zona B
Zona C
Zona D
Zona A
Zona B
Zona C
Zona D
Zona A
Zona B
Zona C
Zona D
Zona A
Zona B
Zona C
Zona D
Zona A
Zona B
Zona C
Zona D
Ora
9,00
14.44
44.5
19.6
76.1
18.2
96.7
8.2
29.7
12.2
163
3.95
12.7
276
61.9
81.4
228
43.4
35.8
162
17.8
17.1
14.7
20.8
12.2
Ev,max.
[mV/m]
Ora
12,00
15.6
50.1
22.1
42.2
17
96.2
11.4
30.1
14.6
183
5.59
74.2
238
43.2
261
189
71.7
30.1
177
115
24.6
18.8
18.5
27.4
Ora
16,00
7.68
17
22.5
32.1
10.2
8.93
19.1
30.9
17.2
36.8
43.6
77.4
84.2
70.3
252
107
152
27.8
226
106
62
19
49.9
22.8
Ev, med
[mV/m]
A [%]
Media
Abaterea
12.6
37.2
21.4
50.1
15.1
67.3
12.9
30.2
14.7
127.6
17.7
54.8
199.4
58.5
198.1
174.7
89.0
31.2
188.3
79.6
34.6
17.5
29.7
20.8
38.92
54.30
8.41
51.80
32.60
86.73
48.06
2.21
17.27
71.16
146.14
76.81
57.77
26.11
58.92
38.74
70.72
14.62
20.00
77.64
79.36
16.00
67.83
41.35
551
AFASES - 2008 -
552
AFASES - 2008 -
Bibliografie
[1] * * *: More on EMC Terminology, IEEE Transactions on Electromagnetic Compatibility,
Vol. EMC-29, No. 3, August 1987, p. 202-205.
[2] Adcliei A., Ball R., Creu M., .a., Electromagnetic Compatibility Testing and
measurement. Parctical Manual, 2002, Warwick University.
[3] Galvao, B.S.M.C., Santos, G., Onusic, H., Sant'Anna, L.F.P.: Electromagnetic
environmental measurements in specific populated areas of Brazil, IEEE International
Symposium on Electromagnetic Compatibility, 2001, Volume 1, 13-17 Aug. 2001 pp.
106 110.
[4] Bdic M., Pslaru-Danescu L., Stefan M.: Bazele Ecranrii Electromagnetice, Vol. I,
Editura Electra, 2007, Bucureti, ISBN: 978-973-7728-93-7, 978-973-7728-97-5.
[5] Poljak, D.: Advanced Modeling in Computational Electromagnetic Compatibility, Editura
John Wiley & Sons Inc., 2007, New Jersey, USA, ISBN: 978-0-470-03665-5.
[6] Marian Costea: Metode i Mijloace de Asigurare a Imunitii Electromagnetice, Editura
AGIR, 2006, Bucureti, ISBN: 973-720-041-1.
[7] Perambur S. Neelakanta, Rajeswari Chatterjee: Antennas for Information Super Skyways:
An Exposition on Outdoor and Indoor Wireless Antennas, Research Studies Press Ltd.,
2003, Baldock, England, ISBN: 0-86380-267-2.
[8] Schmitt, Ron: Electromagnetics Explained, Newnes publishing house, 2002, ISBN 07506-7403-2.
[9] D. V. Giri, F. M. Tesche: Classification of Intentional Electromagnetic Environments
(IEME IEEE International Symposium on Electromagnetic Compatibility, Vol. 46, No.
3, 2004, Pagina: 322-328.
[10] Dipak L. Sengupta, Valdis V. Liepa: Applied Electromagnetics and Electromagnetic
Compatibility, Editura John Wiley & Sons Inc., 2006, New Jersey, USA.
[11] * * *: Guidelines for Limiting Exposure to Time-Varying Electric, Magnetic, and
Electromagnetic Fields (up to 300 GHz), International Commission on Non-Ionizing
Radiation Protection, 1998, Health Physics Society.
[12] * * *: Ministerul Sntii Publice: ORDIN nr. 1.193 din 29 septembrie 2006, pentru
aprobarea Normelor privind limitarea expunerii populaiei generale la cmpuri
electromagnetice de la 0 Hz la 300 GHz, publicat n Monitorul Oficial nr. 895 din 3
noiembrie 2006.
(*) Adrian Marian Maoi, drd.ing., Universitatea Transilvania din Braov, Facultatea de Inginerie Electric i
Stiina Calculatoarelor, Str. Politehnicii nr. 1, tel. 0268474718, e-mail: matoi@unitbv.ro
(**) Elena Helerea, prof. univ. dr.ing., Universitatea Transilvania din Braov, Facultatea de Inginerie Electric i
Stiina Calculatoarelor, Str. Politehnicii nr. 1, tel. 0268474718, e-mail: helerea@unitbv.ro
553
AFASES - 2008 -
554
AFASES - 2008 -
555
AFASES - 2008 -
2003
100
2004
80
60
2005
40
20
2007
2006
0
1
10
556
AFASES - 2008 -
100
90
80
Nr de defecte
70
60
2003
50
2004
40
2005
30
20
10
0
1
Tipul de instalatie
Nr.de defefect
50
40
HR
BV
30
CV
20
10
0
Anul 2003 - 2007
Nr. de defecte
SB
AB
TG
557
AFASES - 2008 -
.
R (t )
R (t ) d t
t
z (t ) d t =
0
dR (t )
R (t )
0
=ln R (t ) 0 = ln R (t ) ln R 0 = ln R (t ) ln 1 = ln R (t ).
t
558
AFASES - 2008 -
(8)
m = t f (t )dt
0
m = t e t d t
0
(10)
0
= .
t e d t = e t + e d t;
t
1
e 0
0
0
0
Pentru nlturarea nedeterminrii se aplic regula lui l Hspital:
lim
t e t
= lim
= 0.
e t
e t 0
(11)
= 0.
t e
dt =
=
0
(0 1) =
2
. m=
(13)
m=
(12)
2
(t m ) f (t ) dt;
D=
(t m )2
= u;
2(t m)dt = du
e t d t = d v ;
1 t
1
1
e
= v ; t = u; dt = du; e t d t = d v ; e t = v ;
D=
(15)
= D=
=m
Relaiile (13), (15) i (16) definesc indicatorii de fiabilitate n cazul unei repartiii
exponeniale a timpilor de bun funcionare.
559
(16)
AFASES - 2008 -
22
f = exp[a + bD ]
(17)
m[ f ] =
(19)
iar dispersia:
2[ f ]=
(20)
exp( bD )
b
exp(a )
b
(21)
exp(a )
. Atunci valoarea medie a unei variabile aleatoare care are densitatea de
b
exp(a )
repartiie dat de relaia (21) mprit la
este:
b
560
AFASES - 2008 -
f = exp[a bD ]
b
= b exp[ bD ]
exp(a )
(22)
valoarea
s 2 = 0,016206
iar
abaterea
medie
ptratic
de
selecie
este
s = s 2 = 0,12730 .
V. Concluzii
Drd.ing., SC FDEE Electrica Distributie Transilvania Sud S.A., BRAOV, Str. Pictor Luchian nr. 25, Tel: 4
0268 305 314, Fax:4 0268 305 004, email: catalin.mihai@electricats.ro.
Prof. univ. dr.ing., Universitatea Transilvania Braov, Facultatea de Inginerie Electric i tiina
Calculatoarelor, Str. Politehnicii nr. 1-3, Braov, Tel/Fax: 0040.268 47 47 18, e-mail: helerea@unitbv.ro.
561
AFASES - 2008 -
I.
MATHEMATICAL DETERMINATION
A basic buck circuit is shown in Fig.1. For simplifying mathematical analysis of the
BUCK circuit operation, the components are assumed to be ideals. The circuit of Fig.1 has,
basically, two switches which determine its mode of operation:
Mode 1: T - on, D - off;
Mode 2: T - off, D - on.
Laplace analysis of the equivalent circuits establishes the expressions for currents and
voltages in each operation mode.
tbl
T1
tcon
Ui
Command
circuit
D1
U
-
562
AFASES - 2008 -
U i = rL i L + L dt + u
du
iC = C C
dt
(1)
u
u
rC iC
=
C
u = R i
i = i L iC
The system results are presented in the next expressions, obtained with MathCAD routine:
last( S )
Re
k= 0
UC rL, rC, L, C, R, I , U , Sk
1 rL, rC, L, C, R, Sk
.e
S .( t
k
t )
Mode 2
Like in mode 1, the system in mode 2 is presented in the next. Its results are made with
MathCAD functions.
di L
rL i L + L dt + u = 0
du
C C = iC
dt
u C u + rC iC = 0
u R i = 0
i L iC i = 0
Expressions for the current and voltage in this function mode are:
563
(2)
AFASES - 2008 -
II.
Using the previous equations, the MathCAD program will simulate the operation of
the circuit. The calculated values of current and voltage will be used to determine when the
circuit moves from one sequence to the next. The final values of currents and voltages are
used as the initial values for the next sequence.
The MathCAD routine determines the values of components converter: inductances [H] and
capacitors [F].
The Buck simulation is realized with the equations, which characterized the steady state. For
this it is necessary to determine the initial values of current and voltage (I, I and U, U).
Define the following:
D = duty cycle (don= tonfc) matrix;
= accuracy.
The initial values of steady state are calculated with a function MathCAD, CondInit(D,).
This function determines the currents and voltages initial values at turn-on moment (T) I , U and at turn-off moment (T) - I , U for all duty cycle. Values matrix is shown in table
1.
Table 1. Initials values at turn-on and at turn-off moment
The matrix columns correspond to all duty factors. The matrix rows represent:
Row 1: the values of I;
Row 2: the values of U
Row 3: the values of I;
Row 4: the values of U;
The waveforms for the model of buck converter with Lm = 0,7mH, rL = 0,01, C = 10F, rC =
0,1, R = 5, Ui = 25V, in case of different duty factor values (1 d = 0,1; 2 d = 0,3; 3
d = 0,5; 4 d = 0,9), are shown in fig.3.
3
2
1
t [s]
t [s]
a.
564
b.
AFASES - 2008 -
2
1
t [s]
c.
2
1
t [s]
d.
EXPERIMENTAL RESULTS
The correctness of the MathCAD modelling is verified with the experimental montage
of the buck converter. The waveforms of the load voltage and different currents of the circuit,
from fig.3, have small variations confronted by the continuous current waveform. These
variations are noticed by the calculus routine and their values and forms are correct drawn on
the diagram? is a problem of the MathCAD modelling.
In fig. 4 are presented an analogy between the MathCAD modelling waveforms and
experimental waveforms
Time [s]
a.
b.
Time [s]
a.
b.
565
AFASES - 2008 -
In fig. 5 the load voltage waveform is presented. The peak-to-peak value of the output
(load) voltage with MathCAD modelling is 0.041V, and experimental value is 0.040V. The
experimental measured output voltage is 11.96V.
In the next tables are presented the values of the converter inductive current and the load
voltage at the beginning (I, U) and at the end (I, U) of the period. The presented values are
those that are obtained with modelling (Ic, Uc, Ic, Uc) and experimental (Im, Um, Im,
Um)-in open loop.
Table 2.
The current through L (IL)
Ic
(A)
Im
(A)
Ic
(A)
Im
(A)
Uc
(V)
Um
(V)
Uc
(V)
Um
(V)
0.35
0.5
0.65
0.8
0
0
0
0
0
0
0
0
1.1313
1.31368
1.1742
1.017
1.102
1.305
1.125
0.94
11.9644
15.5538
18.2121
19.9349
11.93
15.45
18
19.82
11.9844
15.5757
18.2267
19.9378
11.94
15.48
18.1
19.85
The maximum error of the MathCAD modelling is 4.35% for the inductive current iL
and 1.16% for the load voltage.
IV.
CONCLUSION
The analysis of the performance of a buck converter becomes easy by breaking the
cycle of operation into its logical modes. It is simple to generate the time functions for all the
currents and voltages, to read various important values and to incorporate these equations into
a MathCAD program to produce an accurate, detailed and fast running model of the system.
REFERENCES
566
AFASES - 2008 -
d
id
ud
idr
iq
iqr
uq
567
AFASES - 2008 -
dt
(1)
+ B = q
A
didr
idr 0
iqr 0
didt
qr
dt
unde:
0 L1h 0
L1
0
1 L1
1 L1h
R1
L1
0 L1h
0
R1
1 L1
1 L1h
0
A=
B=
0
R2
s1 L1h
s1 L2
0
L1h 0 L2
s1 L1h
s1 L2
R2
0
0 L1h 0 L2
Relaia (1) trebuie completat cu ecuaia micrii:
J ds
M Mm =
p1 dt
(2)
(3)
unde:
L1h
0
0
0
L1
0
L1h
0
L1h
L2
0
0
L1h
0
L2
0
0
0
0
0
0
J
p1
(4)
sau:
568
AFASES - 2008 -
J
( L1L2 L12h ) 2 0
p1
(5)
2
n mod uzual L1 L2 > L1h
i deci determinantul este diferit de zero. Pentru mainile
asincrone ce au ntrefier mic, cuplajul magnetic dintre stator i rotor este foarte bun,
L1 L2 L12h i n cazul acesta determinantul matricei tinde ctre zero.
Considernd determinantul matricei diferit de zero, sistemul exprimat n form
normal este [2]:
did
dt = g1 (t , id , iq , idr , iqr , s )
di
q = g 2 (t , id , iq , idr , iqr , s )
dt
didr
= g 3 (t , id , iq , idr , iqr , s )
dt
diqr = g (t , i , i , i , i , s )
4
d q dr qr
dt
ds
= g 5 (t , id , iq , idr , iqr , s )
dt
(6)
(7)
g 1
iq
g 2
id
M
g 5
id
g 1
idr
g 2
id
M
g 5
id
g 1
iqr
g 2
id
M
g 5
id
g 1
s
g 2
id
M
g 5
id
(8)
Daca aceasta matrice jacobiana n punctul de echilibru are toate valorile proprii cu
partea real negativ, atunci acesta este punctul de echilibru asimptotic stabil. Daca cel putin
una dintre valorile proprii are partea real strict pozitiv atunci punctul de echilibru este
instabil.
569
AFASES - 2008 -
Pentru un motor de inducie trifazat cu rotorul n colivie de putere PN= 4 kW, UN= 380
V si nN= 1440 rot/min se dau urmtorii parametrii electrici [3] : R1= 1,694 , R2= 1,124 ,
L1= 0,196 H, L2= 0,197 H, p1= 2, J= 0,024 kgm2 .
Determinatul matricei (4) are valoarea 1,06510-7 0 i prin urmare sistemul poate fi
exprimat n forma normal :
did
dt = 25140,361 112,073id ( 4082,463 + 3768,263s )iq + 71,324idr
dt
+ 71,324iqr
didr
= 24113,312 + 107,494id (3915,684 3915,684 s )iq 74,114idr
dt
(3768,263 4082,463s )iqr
ds
3
dt = 31,5iqidr + 31,5idiqr + 2,167 10
3
112,073 3,775 10 3
71,324
3,935 10
3
G ' (a ) = 107,494
3,762 10
74,114
3,608 10 3
3
74,114
107,494
3,608 10 3
3,762 10
48,73
372,614
230,895
364,928
570
2,154 10 4
2,818 10 3
2,239 10 4
2,928 10 3
-56,167+298,140i.
AFASES - 2008 -
112,073 669,317
1,025 10 3 1,835 10 3
71,324
3
3,361 10 3
71,324
669,317 112,073 1,025 10
G ' (b) = 107,494 1,022 10 3 74,114
1,906 10 3
1,38 10 3
3
74,114
107,494 1,38 10 3
3,493 10 3
1,022 10
1,871 10 3 1023 10 3 1,966 10 3 1,038 10 3
Bibliografie
[1] Dordea, T. Maini electrice. Partea Complementar, Editura Orizonturi
Universitare, 2002
[2] Babescu, M. , Punescu, D. Maini electrice. Analiza matematic a regimurilor
tranzitorii, Editura Politehnica Timioara, 2001
[3] Mller, V. Maini electrice, Editura Politehnica Timioara, 2005
Valentin Dan Mller, conf.univ.dr.ing., Universitatea Aurel Vlaicu din Arad nr. 77, tel. 0257 250389, e-mail:
mullervalentin@yahoo.com
571
AFASES - 2008 -
I. Introducere
n ultimii ani, odat cu tendinele de reducere a costurilor, a greutii i a siguranei n
funcionarea autovehiculelor, a crescut i interesul pentru asigurarea unui confort optim n
autovehicule. Construciile de autovehicule au evoluat de la construcii simpliste spre maini
moderne cu tehnologii de ultima or, organizate pe criterii funcionale i estetice, care s-i
asigure omului confort, ergonomie i siguran.
Factorii care determin confortul ambiental n autovehicule sunt [1]:
factori de construcie: formele i dimensiunile interioarelor autovehiculului, ansamblul
instalaiilor i al echipamentelor interioare;
factori fizici: microclimatul (compoziia aerului, umiditatea, temperatura), iluminarea,
culoarea, zgomotul, estetica i igiena interioar.
Confortul termic [2] n autovehicule reprezint senzaia subiectiv de echilibru termic
ce apare n corpul uman atunci cnd parametrii mediului din autovehicul - temperatura
aerului, umiditatea aerului, temperatura radiant, viteza aerului, nivelul activitii umane i
nivelul de izolare termic a mbrcmintei se situeaz ntr-un interval de valori bine definit.
572
AFASES - 2008 -
573
AFASES - 2008 -
baza mai multor standarde de confort termic, cele mai semnificative fiind standardele ISO
7730 (1994), ISO 9920 (1995) [12]-[15] i ASHARE 55 (1992), ASHARE (2001) [16], [17].
Modelul lui Fanger calculeaz indicele de confort termic, PMV, pentru orice mediu
particular. Acest lucru poate fi fcut dac se tiu valorile celor ase parametri de confort
(temperatura aerului din interiorul autovehiculului, umiditatea aerului, temperatura radiant,
viteza aerului, nivelul activitii umane i nivelul de izolare termic a mbrcmintei).
n vederea evalurii senzaiei de confort termic se utilizeaz scara subiectiv de
confort termic ASHARE (American Society of Heating, Refrigerating and Air Conditioning
Engineers) cu apte niveluri [18]. Pe aceast scar, valoarea indicelui PMV este cuprins ntre
-3 i +3, corespunznd senzaiei de la foarte rece la foarte cald, respectiv, valoarea de 0 n
care indicele PMV este neutru (Fig.1).
+3 Foarte cald
temperatura aerului din
interiorul autovehiculului;
umiditatea aerului;
temperatura radiant;
viteza aerului;
nivelul activitii umane;
nivelul de izolare termic a
mbrcminii;
+2 Cald
PMV
calculat cu
modelul Fanger
+1 Cldu
0 Neutru
-1 Rcoros
-2 Rece
-3 Foarte Rece
(1)
unde:
M - cldura dezvoltat prin metabolism n corpul uman [Wm-2]; L - lucrul mecanic dezvoltat
de om [Wm-2]; Qsk - cldura pierdut prin piele [Wm-2]; Qres - cldur pierdut prin respiraie
[Wm-2]; C - cldura cedat de corpul uman prin convecie [Wm-2]; R - cldura cedat de
corpul uman prin radiaie [Wm-2]; Esk - cldura pierdut prin evaporare prin piele [Wm-2];
Cres - cldur sensibil (de convecie) pierdut prin respiraie [Wm-2]; Eres - cldur latent de
vaporizare pierdut prin respiraie [Wm-2].
este:
574
AFASES - 2008 -
(2)
35.7 0.032 M (1 ) t h
=
0.155 RCL
(3)
M
-0.042
AD
PMV = 0.352 e
M
(1 ) 0.35 43 - 0.061
(1 ) p a
AD
AD
M
(1 ) 50 0.0023
(44 - p a )
0.42
AD
AD
+ 0.032
M
8
0.0014
(34 - t a ) 3.4 10 f CL
AD
4
4
(t CL + 273) (t r + 273) f CL h C (t CL t a )
t CL
M
(1 - ) - 0.18R CL
= 35.7 0.032
AD
2.05 (t CL t a )0.25
hC =
10.4 v
(4)
(5)
f h (t t )
a
CL C CL
pentru 2.38 (t CL t a )
pentru 2.38 (t CL
575
> 10.4 v
0.25
t a ) < 10.4 v
0.25
(6)
AFASES - 2008 -
kcal
;
M - coeficientul metabolic (activitatea uman),
2
h m
AD - suprafaa total a corpului uman (valorile medii pentru aduli sunt cuprinse ntre 1.65 i 2
m2);
- randamentul activitii umane;
RCL - rezistena termic a mbrcmintei (capacitatea de a reduce transferul de cldur),
unitatea folosit este clo,
C m2 h
[1 clo] = 0.180
;
cal
+ 0.2179PMV 2
(7)
Teoretic, indicele de confort termic PMV are valoarea optim egal cu zero, dar n
reglementri (exemplu, standardul ISO-77300 se consider c domeniul de confort termic
optimal corespunde intervalului de valori cuprinse ntre -0.5 i +0.5.
Relaia (7) permite aprecierea indicelui PPD (fig. 2). Cnd PMV este neutru, PPD are
valoarea 5. Asta valoare semnific faptul c pentru orice mediu confortabil, aproximativ 5%
din populaie l va considera neconfortabil.
576
AFASES - 2008 -
Corelaia ntre indicii PMV i PPD are forma unei curbe n forma de U. Se observ
c n situaia n care PMV are valorile -2 sau +2, indicele PPD are valoarea 80, ceea ce
semnific faptul c 80% din populaie va fi nesatisfcut, i aa mai departe. Se observ c
valoarea optim pentru PPD este de 5 % i se poate obine doar cu ajutorul instalaiilor de
climatizare cu grad nalt de automatizare.
Relaiile (1) (7) permit calculul indicilor PMV i PPD. In ttabelul 1 este prezentat
rezultatul calcululului indicilor PMV i PPD pentru ansamblul de parametri:
temperatura mediului ambiant: ta = 24 C;
temperatura radiant medie: tr = 24 C;
presiunea vaporilor de ap: pa = 103 Pa;
viteza relativ a aerului: v = 0,15 m/s.
Tabelul 1: Valori obinute cu relaiile (1) (7) pentru indicii de confort termic n autovehicul.
M
RCL
PMV
PPD
-2
2
-1
[Wm ]
[m C W ]
[%]
50
0,130
-1,0
27,7
58
0,155
0,0
5,0
66
0,180
0,4
8,8
85
0,130
0,5
10,5
100
0,155
0,9
22,6
115
0,180
1,2
36,4
Rezultatele din Tabelul 1 indic o dependen semnificativ a indicilor de confort
termic din autovehicule de caracteristicile metabolice ale pasagerilor.
III. Cerine i limite privind utilizarea modelul fizic cu manechine termice
577
AFASES - 2008 -
genera o pelicul de apa la suprafaa pielii pentru simularea ratei de evaporare n mediu sau de
absorie n mbrcminte.
Cercetrile n domeniu arat c se dezvolt manechine termice care s posede un grad
mare al rezoluiei, un rspuns termic local, un sistem de traspiraie, un timp scurt de rspuns i
o reacie continu i adaptat mediului termic. Un manechin termic avansat, cu aceste
capaciti, ar sprijini dezvoltarea de sisteme de control climatic mai performante i mai
eficiente n mediul de transport dar i acolo unde exista medii termice neuniforme i
tranzitorii.
Dezvoltarea i mbuntirea manechinelor termice ca un sistem de evaluare al
mbrcminii este important din punct de vedere al sntii. Un manechin termic trebuie s
aibe urmtoarele proprieti pentru a putea simula coprul uman: (1) mrime i form corect;
(2) un control al emisiei de cldur; (3) un control al distribuiei cldurii la suprafaa pielii; (4)
un control al evaporrii la nivelul pielii; (5) un control al distribuiei transpiraiei la suprafaa
pielii; (6) controlul micrii; (7) un control al simulrii rspunsurilor fiziologice ale corpului
uman. Un manechin care s aibe toate aceste proprieti nu s-a dezvoltat nc, totui
manechine care controleaz transpiraia la nivelul pielii i viteza de micare sunt n cercetare
i dezvoltare n unele laboratoare.
Manechinele termice sunt limitate la o distribuie uniform a termperaturii la suprafaa
pielii, chiar dac la nivelul extremitilor corpului uman exist pierderi mari de cldur. Acest
lucru duce inevitabil la o supraestimare a pierderii de cldur la nivelul extremitilor corpului
uman. Msurrile climatului local cu un manechin termic sunt corelate cu senzaiile termice
experimentate pe subieci supui acelorai condiii. Criteriul pentru condiii climatice
acceptabile poate fi definit n termeni cantitativi msurat utiliznd manechinul.
IV. Concluzii
578
AFASES - 2008 -
Bibliografie
[1] Brujan, E.,A., Ventilaia i condiionarea aerului.
[2] Srbu, I., Modele de evaluare a confortului termic, Consideraii teoretice, nr.2 (43), 2007,
Universitatea Politehnic Timioara.
[3] S. Atthajariyakul, T. Leephakpreeda, Neural computing thermal comfort index for HVAC
systems, School of Manufacturing Systems and Mechanical Engineering, Sirindhorn
International Institute of Technology, Thammasat University, Patumthami, Thailand.
[4] John P. Rugh, Desikan Bharathan, Predicting human thermal comfort in automobiles,
National Renewable Energy Laboratory, NRELl/CP-540-37784, presented at the Vehicle
Thermal Management Systems Conference and Exhibition, May 2005, Toronto, Canada.
[5] Itsuhei Kohri1, Tohru Mochida, Evaluation Method of Thermal Comfort in a Vehicle with
a Dispersed Two-Node Mode Part 1 Development of Dispersed Two-Node Model,
Journal of the Human-Environmental System, vol.6, no.1, 2002.
[6] McGuffin, R., Burke, R., Huizenga, C., Hui, Z., Vlahinos, A., Fu, G., Human Thermal
Comfort Model and Manikin, National Renewable Energy Laboratory, Society of
Automotive Engineers, Inc, 2002, Colorado, USA.
[7] * * *, Thermal Confort in Automobiles, LumaSense Technologie.
[8] ISO 7726 (1998). Ergonomics of the thermal environmentinstruments for measuring
Physical quantities, Geneva: International Standards Organization.
[9] Fanger, P., O., Thermal Comfort: Analysis and Applications in Environmental
Engineering, McGraw-Hill Book Company, 1970, USA.
[10] Fanger, P., O., Thermal Comfort. McGraw-Hill Book Co., 1973, New York, USA.
[11] Joel M. Devonshire, James R. Sayer, The effects of infrared-reflective and antireflective
glazing on thermal comfort and visual performance: a literature revie, University of
Michigan, Transportation Research Inst., Report no.UMTRI-2002-4 March 2002, U.S.A.
[12] ISO 7730. (1994) Moderate thermal environmentsdetermination of the PMV and PPD
indices and specification of the conditions for thermal comfort. Geneva.
[13] ISO 9920 (1995). Ergonomics of the Thermal Environment: Estimation of the Thermal
Insulation and Evaporative Resistance of a Clothing Ensemble. Geneva.
[14] Tri Harso Karyono, The applicability of the ISO 7730 (Fanger's comfort model) and the
adaptive model of thermal comfort in Jakarta, Indonesia, Agency for the Assessment and
Application of Technology, Republic of Indonesia.
[15] Parsons,K.,C., Introduction to thermal comfort standards, Loughborough University, UK
[16] American Society of Heating Refrigerating and Air Conditioning Engineers, Thermal
Environmental Conditions for Human Occupancy (ASHRAE Standard 55),1992, Atlanta.
[17] American Society of Heating Refrigerating and Air Conditioning Engineers (ASHRAE),
Thermal comfort. In ASHRAE Handbook: Fundamentals, 2001, Atlanta, GA.
[18] Charles,K.E. Fangers Thermal Comfort and Draught Models, Institute for Research in
Construction, Report RR-162, October, 2003, Canada, Ottawa.
[19] Joel M. Devonshire, James R. Sayer, Radiant heat and thermal comfort in vehicles, The
University of Michigan, Transportation Research Institute, Report No. UMTRI-2003-32,
November 2003, Michigan, U.S.A.
[20] Huizenga, C. Zhang, H., Mattelaer, P., Tiefeng, Yu, Arens, E., Window performance for
human thermal confort- final report, The National Fenestration Rating Council, february
2006, Center For the Built Environment, University of California, Berkeley.
[21] Parsons, Ken, Human thermal environments second edition, Taylor & Francis Group,
London, ISBN 0-415-23793-9.
**
Radu Musat, inginer, Renault Technologie Roumanie, tel. 0723 151844, e-mail: r_musat@yahoo.com.
Elena Helerea, prof. dr. ing.,Universitatea Transilvania Brasov, tel. 0744624682, e-mail: helerea@unitbv.ro.
579
AFASES - 2008 -
I. INTRODUCTION
High energy irradiation (gamma or accelerated electrons) is a well-known tool used for
the modification of polymers [1, 2]. The radiation modified polymer materials reveal the potential
applications on various areas of work. One of the most important usages of polymers, especially
polyethylene, is the electrical insulation because of its good mechanical and chemical resistances
under the strong and long-term action of degrading agents.
580
AFASES - 2008 -
Several papers have been reported on the evaluation of the operational features and durability of
polymer cable materials due to the necessity of proper finding of suitable compounds for the
manufacture of electrical insulations. Sun at al [3] and Chou et al [4] investigated the contribution
of carbon black on the thermal and electrical properties of polyolefins: low density polyethylene
and ethylene-propylene elastomer, respectively. Electrical conduction of different insulation
compositions was assessed [5 15], who have emphasized the influences of structural
parameters, the presence of stabilizers or the interaction between blending components. The
output information is useful either for the quality control of materials or for the development of
new technologies in the branch of dielectric materials.
The environmental conditions play an important role in the degradation of polymers. The
actions of oxygen, heat, ionizing radiation, UV light cause the advanced modifications in the
molecular structures of macromolecular matrix, which may change the functional features of
exposed materials [16]. Also, the presence of any stabilizer brings about the improvement in the
lifetime of items in direct connection with its antioxidant efficiency [17]
In the recent years, a main attention is focused on the durability of electrical materials in
order to validate the material availability for the special applications including military purposes
(insertion of proper insulation that can resist on the wartime or on terrorist attacks). The problem
of material durability must be solved for all areas of usage, because it determines the warranty of
products where they are one of components.
The energy transfer onto polymer materials causes damages by bond scissions. The
primary processes occurred in high energy irradiation were previously presented by Stephens et
al. [18]. Free radicals that are formed react with diffused molecular oxygen and hydroperoxides
appear (Figure 1). The further processes involve these intermediates that generate final and stable
oxygenated products. The consequence of this degradation process is the accumulation of
compounds with oxygenated functions like ketones, aldehydes, acids, hydroperoxides and other
related structures, and the depletion of double bonds that are the reactive sites in relation with
free radicals and O2. The oxidative degradation is a chain process, which is promoted by RO2.
radicals. The mechanism of thermal and radiochemical degradation is described by the Bolland
and Gees scheme [19] (Figure 1).
581
AFASES - 2008 -
II. EXPERIMENTAL
Low density polyethylene (LDPE) was provided by ARPECHIM Piteti (Romania) as
A23 BP085 type. Neat material was inserted in an electrically heated plate press and it was
maintained at 1500C and 180 bars for 10 minutes. Sheets of 150 m thickness were obtained.
They were subjected to the action of -rays provided by 137Cs source incorporated in
GAMMATOR M-38-2 installation (USA) at a dose rate of 0.4 KGy/h. Irradiation/measurement
cycles were applied because of the accumulation of dose.
Spectral measurements were carried out with JASCO 4200A (Japan) FTIR
spectrophotometer and electrical investigations were performed with Keithley 6517A (USA)
electrometer. Oxygen uptake determinations (testing temperature: 1800C, oxidize environment:
air at normal pressure). Isothermal and isobaric experiments were done in an original equipment
built up in our laboratory [20].
III. Results and discussion
The degradation of low density polyethylene was assesses by the accumulation of
oxygenated products in accordance with the mechanism presented in figure 1. Figure 2 shows the
FTIR spectra recorded on the neat polyethylene and on 100 kGy air irradiated polyethylene.
Transmision (%)
60
50
40
30
20
10
0
4000
LDPE
0 kGy
100 kGy
3500
1500
-1
Wave number (cm )
1000
582
AFASES - 2008 -
The changes in the prominent peaks of recorded spectra in the region of 3450 cm-1 and
1720 cm-1 illustrate the increase in the amounts of hydroperoxides and carbonyl units,
respectively. The initiation of oxidation is ensured by peroxyl radicals, which produce various
oxygenated products [21]. They are dipoles, which are involved in the variation of electrical
behavior. The progress in the main oxygenated functions is presented in figure 3. It may be
noticed that the ROOH content (expressed as absorbance at 3350 cm-1, A3350) exhibits a slower
increase that happens with carbonyl units expressed as absorbance at 1720 cm-1, A1720). The
conversion of hydroperoxides (alcoholic function) into carbonyl structures takes place
simultaneously with its accumulation from the reactions of free radicals with molecular oxygen
that penetrates and diffuses in polyethylene.
Fig. 3. Changes in the absorbance of bonded hydroxyl (left side) and carbonyl (right side) during
the -irradiation of LDPE.
(red line: exposure in water; blue line: exposure in air)
The oxidative process occurs with different rates, when polyethylene samples are
radiation-processed in water or in air [22]. The contributions of radiolysis intermediates that exist
as radicals react with the first layer of LDPE and transfer the radical position onto
macromolecules. The new intermediates act simultaneously with high energy radiation
themselves for the formation of degradation precursors, peroxyl radicals (figure 1).
The oxidation states of LDPE specimens are characterized by two main kinetic
parameters: oxidation induction time (the period when material preserves its initial structure; it is
expressed in minutes) and rate of oxidation (the quantity of oxygen that reacts with material; it is
expressed in mmol O2/polymer massxmin). These parameters were calculated from the
oxydograms recorded for each dose and the type of irradiation environment. The induction time
was determined as the time value obtained by the intersection of Ox axis with the tangent on the
oxidation curve on the propagation step; the oxidation rate was computed by the division of a
certain oxygen amount to the corresponding time of reaction. In figure 4 the oxydograms drawn
for air-irradiated low density polyethylene are presented. The effect of radiation exposure may be
583
AFASES - 2008 -
remarked by the shortening of oxidation induction times and the increase in the oxidation rates of
processed polyethylene sheets.
The full description on the radiochemical oxidation of studied LDPE samples is exhibited
in table 1.
air
Induction
time
(min)
195
141
73
24
12
water
Oxidation rate
(mmol O2/g.min).10-3
2.42
3.55
5.21
7.03
9.86
Induction
time
(min)
195
122
52
15
6
Oxidation rate
(mmol O2/g.min) .10-3
2.42
4.02
6.34
9.50
11.79
The oxidation of cables takes place in the presence of oxygen and water vapors. It means
that the synergistic effect of these stressors would be expected. These data regards the polymer
degradation at 1800C, the high temperature for usual regime of operations, except the cases of
584
AFASES - 2008 -
6
4
2
0
20
40
60
Dose (kGy)
80
100
-13
-13
accidents. By the evaluation of oxidation rates and oxidation induction time under the accelerated
testing, a clear consequence of cable-insulation resistance in outdoor positions can be revealed.
However, the oxidative degradation occurs, even the temperature is not so high. The
accumulation of degradation product brings about a depreciation of material quality in direct
relation with the working temperature, the mechanical charge, the electrical voltage, the duration
of operation [23].
The conduction of irradiated LDPE samples is modified by the increase in the irradiation
dose. Because a very small amount of double bonds is formed during - exposure and due to the
appearance of oxygenated products that plays the role of electron scavengers (the
electronegativity of oxygen), a fraction of -electrons are trapped on these structural positions. In
addition, the electrons which escape from bond-orbital level contribute to the measure current
intensity. The resistivity measured for radiation processed LDPE mitigates and the quality of
electrical insulation is diminished. In figure 5 the values of measured resitivities are presented in
comparison with the unirradiated material.
6
5
4
3
2
1
0
20
40
60
Dose (kGy)
80
100
585
AFASES - 2008 -
has to be taken into consideration the synergistic effects of these factors, which diminish the
lifetime of materials.
IV. CONCLUSION
The study on the different aspects of polymer degradation involves the evaluation of
chemical modifications occurred during this process and the direct consequences on the
functional characteristics of material. An advanced state of oxidation determines the most
probable breakdown of material. The accelerated testing of electrical insulations, like low density
polyethylene, is a good tool for the evaluation of material quality and its limits of operation. The
characteristics of electrical insulators must be judged by the correlation between chemical
strength under hard conditions and the answer of material, which determines the area of
applications.
References
[1] Zaharescu, T., Podin, C., Radiochemistry of Polymers, Bucharest University Printing House,
2003.
[2] Gillen, K. T. Clough, R. L., Ageing Predictions of Nuclear Power Plants: Crosslinked
Polyolefins and EPR Cable Insulation Materials, Report SAND-91-0822, 1991.
[3] Sun, Y. Y., Fan, L. H., Watkins, K., Peak, J., Wong, C. P., J. Appl. Polym. Sci., 93, 509
(2004).
[4] Chou, H. W., Huang, J. S., Lin, S. T., J. Appl. Polym. Sci., 103, 1244 (2007).
[5] Acosta, J. L., Gonzles, L., del Rio, C., Ojeda, C., Rodriguez, A., J. Appl. Polym. Sci., 79,
2136 (2001).
[6] Lee, M. G., Nho, Y. C., J. Appl. Polym. Sci., 83, 2440 (2002).
[7] Soljovruji, E., Kostoski, D., Dojilovo, J., Polym. Degrad. Stabil., 74, 167 (2001).
[8] Gillen, K. T., Bernstein, R., Clough, R. L., Celina, M., Polym. Degrad. Stabil., 91, 2146
(2006).
[9] Gasa, J. V., Liu, Z., Shaw, M. T., Polym. Degrad. Stabil., 87, 77 (2005).
[10] John, H., Josef, R., Mathew, K. T., J. Appl. Polym. Sci., 103, 2682 (2007).
[11] Tjing, S. C., Liang, G. D., e-Polymer, no. 037 (2007).
[12] Zamora, F., Gonzles, C., Carballo, T., Zamora, N., Lezama, L., e-Polymer, no. 032 (2008).
[13] Souza F. G jr, Michel, R. C., Soares, B. G., Polym. Testing, 24, 998 (2005).
[14] Fouracre, R. A., MacGregor, S. J., Judd, M., Banford, H. M., Radiat. Phys. Chem., 54, 209
(1999).
[15] Zhou, X. Z., Zhou, Y., Zhou, Y., Tang, Q., Radiat. Phys. Chem., 63, 267 (2002).
[16] Ranby B., Rabek J. F., ESR Spectroscopy in Polymer Research, Springer Verlag, Berlin,
1977.
[17] Gorghiu, L. M., Dumitrescu, C., Jipa, S., Olteanu, R., Zaharescu, T., Gigante, B., Santos, C.,
Silva, A. M., Rev. Chim. (Bucharest), 55, 777 (2004).
[18] Stephens, C. P., Benson, R. S., Ling, X., Song, H., Ham, H.-J., Buchanan, R.A., Chipar,
M.; e-Polymer, no. 042 (2008).
[19] Bolland J. L., Gee, G., Trans. Faraday Soc., 42, 236 (1946).
[20] Zaharescu, T., Jipa, S., Setnescu, R., Rev. Chim. (Bucharest), 55, 320 (2004).
586
AFASES - 2008 -
Emil Dumitru Popescu, PhD candidate, Military Technical Academy, 81-83 George Cosbuc
Av., Bucharest 050141, tel: 0723556562, e-mail: edp171@yahoo.com
**
Florin Bue, PhD candidate, Military Technical Academy, 81-83 George Cosbuc Av.,
Bucharest 050141, tel: 0723556562, e-mail: buseflorin@yahoo.com
***
Gheorghe Gavril, Prof. PhD, Military Technical Academy, 81-83 George Cosbuc Av.,
Bucharest 050141, tel. 021 3354660, e-mail: atm@mta.ro
****
Traian Zaharescu, Senior Researcher, Assoc. Prof., PhD, National Institute for Electrical
Engineering (INCDIE ICPE CA), 313 Splaiul Unirii, Bucharest 030138, tel. 0726636222, e-mail:
zaharescu@icpe-ca.ro; traian_zaharescu@gw-chimie.math.unibuc.ro
587
AFASES - 2008 -
588
AFASES - 2008 -
OP2177
Amplificatorul de instrumentaie
589
AFASES - 2008 -
Filtrul trece-jos
590
AFASES - 2008 -
2.0mV
1.0mV
0V
-1.0mV
0s
0.2s
0.4s
0.6s
0.8s
1.0s
1.2s
1.4s
1.6s
1.8s
2.0s
-V(2)
Time
Alimentarea circuitului
2.6V
2.4V
2.2V
0s
0.1s
0.2s
0.3s
0.4s
0.5s
0.6s
0.7s
0.8s
0.9s
V(5)
Time
591
1.0s
AFASES - 2008 -
2.
Conform teoriei filtrelor [3], filtrul cu panta cea mai abrupt este filtrul Cebsev, dar
rspunsul lui n band conine supracreteri (pulsaii), n funcie de numrul de poli. Urmeaz
n ordine filtrele Butterworth i Bessel. n lucrare, se studiaz comportarea filtrelor
Butterworth i Bessel de ordinul 5, filtrul cu rspuns Cebsev alternd acurateea sistemului.
S-au analizat caracteristicile filtrelor trece jos de tip Butterworth i Bessel n funcie de
caracteristicile amplitudine-frecven.
Pentru ca rspunsurile filtrelor Butterworth i Bessel s poat fi comparate, s-au
considerat aceleai frecvene de frngere ale caracteristicilor i aceeai amplificare.
Prin suprapunerea caracteristicilor amplitudine-frecven ale celor dou filtre se
observ o asemnare aproape perfect rezultnd concluzia c ar putea fi utilizat oricare dintre
cele dou tipuri analizate (fig. 7).
100
-0
-100
-200
-300
1.0Hz
3.0Hz
DB(V(27))- DB(V(15))
10Hz
DB(V(270))- DB(V(15))
30Hz
100Hz
300Hz
1.0KHz
3.0KHz
10KHz
30KHz
100KHz
Frequency
-100
-200
DB(V(27)) - DB(V(15))
0d
-100d
-200d
SEL>>
-300d
1.0Hz
3.0Hz
P(V(27)) - P(V(15))
10Hz
30Hz
100Hz
300Hz
1.0KHz
3.0KHz
10KHz
30KHz
100KHz
Frequency
592
AFASES - 2008 -
100
-100
-200
DB(V(270)) - DB(V(15))
0d
-400d
SEL>>
-800d
1.0Hz
3.0Hz
P(V(270)) - P(V(15))
10Hz
30Hz
100Hz
300Hz
1.0KHz
3.0KHz
10KHz
30KHz
100KHz
Frequency
2.5V
2.4V
SEL>>
2.3V
V(270)
5.0V
2.5V
0V
0s
0.2s
0.4s
0.6s
0.8s
1.0s
1.2s
1.4s
1.6s
1.8s
2.0s
V(27)
Time
Fig. 10: Rspunsurile n timp ale filtrelor analizate (Bessel sus, Butterworth jos)
IV. Concluzii
Simularea Spice a unui sistem electronic de prelucrare a semnalelor
electrocardiografice conduce la urmtoarele concluzii:
Pentru acurateea procesrii analogice a semnalelor, amplificatorul de instrumentaie
trebuie alimentat cu tensiune dubl;
Filtrarea semnalului se face cel mai bine cu un filtru trece-jos de tip Bessel.
Bibliografia
[1]
[2]
[3]
[4]
Ana-Maria Puca, student, Anul V, EA, Universitatea Transilvania, Braov, Str. Politehnicii, Nr. 1-3, Tel.
0268-478705, e-mail: ana-maria.puscas@yahoo.com
Gheorghe Pan, conf. univ. dr. ing., Universitatea Transilvania, Braov, Str. Politehnicii, Nr. 1-3, Tel. 0268478705, e-mail: pana@vega.unitbv.ro
593
AFASES - 2008 -
Abstract:
There is a tendency for miniaturization in the electrical equipments industry. Due to the
miniaturization, electrical stress is increasing, so that defection may appear. Defection
processes modeling needs complete information. In order to change a determinist design to a
operational one, that will take into account the reliability, one needs to know each parameter
variation low. In this paper one make a forecast estimation of the helicoidally cylindrical
springs characteristics, by the reliability indicators and by the performance degradation
coefficient of the spring metal b, that is experimentally measured using some original stand
in heating conditions.
I. REZULTATE EXPERIMENTALE PENTRU COEFICIENTUL DE
DEGRADARE A PERFORMANELOR SISTEMELOR ELASTICE CILINDRICE
LA TEMPERATURI INALTE
Concepia din fiabilitate i mentenabilitate a echipamentelor electrice, presupune o
strategie diferit de cea tradiional.
Algoritmii de dimensionare preliminar se bazeaz pe considerarea defectului, utilizarea
modelelor stochastice i a coeficienilor operaionali de siguran. Proiectarea calitii
echipamentelor electrice necesit organizarea unui sistem informaional total, care s pun la
dispoziia concepiei datele i informaiile necesare (bnci de date). Cunoaterea
mecanismelor fizice de oboseal este esenial n considerarea fiabilitii n faza de concepie.
n structura aparatajului electric de joas i nalt tensiune o importan deosebit o
reprezint elementele i subsistemele cu acumulare de energie, prevzute n general cu
elemente elastice, care asigur fie nchiderea, fie deschiderea cilor de curent.
Pentru a gasi o relaie ntre fora elementului elastic dup N cicluri fa de fora
elementului elastic la zero cicluri cnd sgeata resortului este dat, innd seama de
temperatura de lucru a resortului, s-a construit un stand original de ncercari
A fost stabilit o temperatur de 80oC apoi de 60oC. Schema electric a standului de
termostatare este dat n lucrarea [1]. El a fost montat pe standul de ncercri de fiabilitate
pentru temperatura mediului ambiant dat n figura 1.
594
AFASES - 2008 -
Pentru 80oC. msurtorile au fost fcute pe patru tipuri de arcuri, reprezentate fiecare
de 20 de arcuri. Modul de msurare a fost similar cu cel pentru temperatura mediului ambiant.
Msurtorile sunt exemplificate n tabelul 1.
Pentru 60oC. msurtorile au fost fcute pe 3 tipuri de arcuri, reprezentate fiecare de 20
de arcuri. Au fost ntocmite tabele de msurtori. La fiecare tip de resort a fost verificat legea
de repartiie a valorilor msurtorilor reale cu legea de repartiie normal utiliznd testul de
concordan Massey. Legile au corespuns.
Valorile msurate au fost trecute n tabele iar funciile
P N 2 = f(N) i ln P N 2 = f(N) au fost reprezentate grafic. Pentru 80oC. exist tabelul 2 iar
pentru 60oC. exist tabelul 4.
Algoritmul de calcul pentru coeficinii de degradare b1t, b2t, b3t, b4t este similar cu cei ai
coeficienilor b1, b2, b3, b4 corespunztori mediului ambiant cf.[2]. A fost ntocmit tabelul 3
pentru 80oC. i tabelul 5 pentru 60oC.
Stand ncercri fiabilitate
220Vca
N
N
PE
F1
S K B
F2
1
F3
F4
220V
T
24V
F5
KR
NI
C
3
S1
5
2
2
K
4
S2
A
KR B
M
N
M
PE
Sistem
experimentari
resoarte
K
Fig.1. Schema electric
595
AFASES - 2008 -
596
AFASES - 2008 -
N (cicli)
P 2 (N)
ln P 2
Diagrama
Resort cilindric de compresiune A.C. 11, i = 80oC
0
106
5105
2106
2,5106
62,02
60,21
58,74
57,07
55,93
310
54,91
4,127
4,006
4,098
4,073
4,044
4,024
tg = b 1t =
tg = b 2t =
P 02 = P N 2 e b 2t N
b 2t = 3,73.10 8 ;
pt. N 10 6
597
cicli ;
Tabel 2
6
AFASES - 2008 -
tg = b4t =
P 02 = P N 2 + b4tN;
Nr
crt
tip
AC 11
AC 12
AC 19
AC 20
Nc (cici)
Nr
crt
1
2
3
Compararea coeficienilor b
Scal semi logaritmic
Scal normal
Arc
1
2
3
4
b4t =2,1210-6;
b1ti
5,4010-8
5,1010-8
5,3010-8
5,1010-8
P 2 (N)
0
62,02
ln P 2
4,127
b2ti
3,7310-8
3,6010-8
4,0610-8
3,8010-8
b3ti
3,2810-7
0,8210-6
1,6910-6
1,0510-6
b4ti
2,1210-7
0,5310-6
1,2210-6
0,7310-6
= 80 C
= 80 C
= 80 C
= 80 C
Diagram
Resort cilindric de compresiune A.C. 11, i = 60oC Tabel 4
106
5105
2106
2,5106
3106
58,40
56,98
55,36
54,25
53,26
4,067
4,028
4,014
3,994
Compararea coeficienilor b
scal semilogaritmic
scal normal
b1ti
b2ti
b3ti
b4ti
-8
-8
-7
5,2410
3,6210
3,1410
2,0610-7
4,9510-8
3,4810-8
0,7910-6
0,5110-6
4,9410-8
3,6510-8
1,0110-6
0,7010-6
Tip
arc
AC 11
AC 12
AC 20
Tabel 3
Obs.
3,975
Tabel 5
Obs.
= 60C
= 60C
= 60C
OBSERVAII
a) Experimentrile s-au efectuat exclusiv asupra resoartelor elicoidale cilindrice de
compresiune din aparatele electrice de comutaie preluate din producia de serie a uzinei
Electroaparataj Bucureti, executate din srm RR, avnd diametrul maxim al srmei de 2
mm i diametrul minim al srmei de 0,55 mm. Fora msurat maxim a fost de 131 N i fora
msurat minim de 1,18 N.
b) S-au efectuat msurtori a resoartelor n instalaia de termostatare numai pentru
temperaturile de 80oC. i 60oC
c) Coeficienii b1ti, b2ti au valori apropiate ntre ele la fiecare tip de coeficient b, deci se
poate folosi o valoare medie a fiecrui tip de coeficient (b1t, b2t).
-Pentru = 80 C:
P N 2 = P 02 e b1t N ;
4
b1t =
b1ti
i =1
= 5,22 10 8
pt. N 10 6 cicli.
598
AFASES - 2008 -
P N 2 = P 02 e
b 2 t N
; b 2t =
b 2ti
i =1
- Pentru = 60 C:
= 3,80 10 8
P N 2 = P 02 e b1t N ;
b1t =
b
i =1
1ti
= 5,04 10 8
pt. N 10 6 cicli.
b2t =
2 ti
i =1
= 3,58 10 8
d) Coeficienii b3ti, b4ti au valori ndeprtate, deci nu se poate folosi o valoare medie a
acestora.
e) Analiznd funciile PN2 = f(N) se constat c variaia funciei poate fi considerat
liniar. innd seama de experimentrile specificate se poate accepta ipoteza degradrii
liniare a performanelor arcurilor
f) Coeficienii b1ti, b2ti au valori apropiate ntre ele la fiecare tip de coeficient b, att
pentru = 80 C. ct i pentru = 60. C, deci se poate folosi o valoare medie a fiecrui tip
de coeficient (b1t, b2t) 5.135 10 8 pentru N < 10 6 cicli i 3.69 10 8 pentru N > 10 6 cicli.
II. CONCLUZII
Cu ajutorul unui stand de ncercri i a unei instalaii de termostatare s-au efectuat
cercetri asupra resoartelor pentru a determina influena unor factori (oboseala prin ciclu
oscilant i temperatur) asupra performanelor lor.
Relaia care aproximeaz cel mai bine scderea forei n funcie de numrul de cicli este
de tip exponenial att la temperatura mediului ambiant ct i la temperatura de 80oC. i 60oC.
innd seama de aceast variaie a forei i impunnd o anume fiabilitate aparatelor i
elementelor componente se poate determina o for de calcul, astfel nct aparatul s
funcioneze la N cicluri i o temperatur dat.
n acest fel, se poate nlocui o proiectare determinist cu o proiectare operaional care
s in seama de indicatorii de fiabilitate.
Bibliografie
[1].Ratiu Gheorghe, Experimental results on the degradation coefficient of the elastic
tronconical systems performances for electric equipments in high temperature, articol la
Academia Forelor Terestre, Sibiu, la a 12-a sesiune de comunicri tiinifice cu participare
internaional din.11-14.06.2007
[2].Raiu Gheorghe, Asigurarea calitii i fiabilitii elementelor active din
componena echipamentelor electrice, Tez doctorat, Universitea Politehnic, Bucureti,
2001.
Gheorghe Ratiu, conf.univ.dr.ing., Academia Forelor Terestre Nicolae Blcescu Sibiu, str. Revoluiei
nr. 3-5, tel. 0269 432990/ 1332, e-mail:ghratiu@armyacademy.ro
599
AFASES - 2008 -
600
AFASES - 2008 -
Msurtorile au fost fcute pentru 3 respectiv 2 tipuri de arcuri, fiecare tip fiind reprezentat de
20 de arcuri pentru fiecare temperatur(80oC i 60oC). Msurtorile sunt similare cu cele
fcute la temperatura mediului ambient i sunt date n tabelul 1.
La fiecare tip de resort a fost verificat legea de repartiie a valorilor msurtorilor reale
cu legea de repartiie normal utiliznd testul de concordan Massey. Legile au corespuns.
Valorile msurate au fost trecute n tabele iar funciile
P N 2 = f(N) i ln P N 2 = f(N) au fost reprezentate grafic. Pentru 80oC. exist tabelul 2 iar pentru
60oC. exist tabelul 4.
Algoritmul de calcul pentru coeficinii de degradare b1t, b2t, b3t, b4t este similar cu cei ai
coeficienilor b1, b2, b3, b4 corespunztori mediului ambiant cf.[2]. A fost ntocmit tabelul
final 3 pentru 3 tipuri de arcuri i pentru 80oC. i a fost ntocmit tabelul final 5 pentru 2 tipuri
de arcuri i pentru 60oC.
Instalaia de termostatare:
1
220Vca
X
X
S K B
F2
N
PE
F1
F3
F4
Sistem de
experimentare
arcuri
220V
T
24V
1
F5
2
Tc
RE
0
a
1
K
RE
0
601
AFASES - 2008 -
4 buc.;
1 buc.;
1 buc.;
1 buc.;
1 buc.;
1 buc.;
1 buc.;
1 buc.;
602
AFASES - 2008 -
Diagram
Arc conic de compresiune tip AC 25 i = 80oC
Tabel 2
No.
(cicli)
5105
106
1,5106
2106
2,5106
3106
P 2 (N)
ln P 2
1,639
0,494
1,593
0,466
1,557
0,443
1,532
0,427
1,506
0,409
1,485
0,395
1,452
0,373
P 02 P N 2
N
P 02 = P N 2 + b3tN;
b3t = 0,8210-7; pt. N 106 cicluri
Pentru curba 4 scal normal
603
AFASES - 2008 -
tg = b4t =
P 02 = P N 2 + b4tN;
Nr
crt
arc
tip
1
2
3
AC 21
AC 22
AC 25
scal semilogaritmic
b2ti
b1ti
-8
3,7.10-8
5,3010
3,8510-8
5,1510-8
-8
3,6010-8
5,1010
scal normal
b3ti
b4ti
-6
2,1510-6
3,310
1,0510-6
0,7510-6
-7
0,8210
0,5310-7
Diagram
Arc conic de compresiune tip AC 25
Tabel 3
Obs.
= 80 C
= 80 C
= 80 C
i = 60oC
Tabel 4
No.
(cycl)
5105
106
1,5106
2106
2,5106
3106
P 2 (N)
ln P 2
1,639
0,494
1,545
0,435
1,487
0,397
1,486
0,396
1,460
0,378
1,445
0,365
1,408
0,342
Nr
crt
1
2
arc
tip
AC 21
AC 25
Tabel 5
Obs.
= 60 C
= 60 C
OBSERVAII
P N2 = P 02 e
b1t N
; b1t =
b1ti
i =1
= 5,18 10 8
pt. N 10 6 cicli.
P N2 = P 02 e
b 2t N
; b 2t =
-Pentru = 60 C:
b 2ti
i =1
= 3,72 10 8
604
AFASES - 2008 -
P N2 = P 02 e
b1t N
; b1t =
b1ti
i =1
= 5,03 10 8
pt. N 10 6 cicli.
P N2 = P 02 e
b 2t N
; b 2t =
b 2ti
i =1
= 3,6 10 8
d) Coeficienii b3ti, b4ti au valori ndeprtate, deci nu se poate folosi o valoare medie a
acestora.
e) Analiznd funciile PN2 = f(N) se constat c variaia funciei poate fi considerat
liniar. innd seama de experimentrile specificate se poate accepta ipoteza degradrii
liniare a performanelor arcurilor
f) Coeficienii b1ti, b2ti au valori apropiate ntre ele la fiecare tip de coeficient b, att
pentru = 80 C. ct i pentru = 60 C, deci se poate folosi o valoare medie a fiecrui tip de
coeficient (b1t, b2t)
5.105 10 8 , pentru N < 10 6 cicli i 3.66 10 8 pentru N > 10 6 cicli.
II. CONCLUZII
Gheorghe Raiu, conf.univ.dr.ing., Academia Forelor Terestre Nicolae Blcescu Sibiu, str. Revoluiei
nr. 3-5, tel. 0269 432990/ 1332, e-mail:ghratiu@armyacademy.ro
605
AFASES - 2008 -
1. Introducere
Realizarea unor centrale pentru energia valurilor constituie o preocupare deosebit
pentru rile care beneficiaz de rmuri cu valuri puternice, problema prezint totui interes i
pentru rile care dispun de valuri de mic amploare. Astfel, avnd n vedere necesitatea
construirii unor diguri de protecie costier se pune n prezent problema recuperrii energiei
valurilor care ajung la aceste diguri. Dei energia valurilor Mrii Negre este relativ redus i
nu se poate pune n prezent problema livrrii energiei captate n reea, exist nenumrate
aplicaii n care aceast energie este foarte util. [1]
Avnd n vedere cercetrile efectuate la nivel mondial i experimentrile efectuate n
ara noastr se poate concluziona c energia valurilor Mrii Negre poate fi captat i utilizat
pentru aplicaii locale n condiii eficiente cu ajutorul instalaiilor hidropneumatice cu coloan
oscilant.[1,3,7]
606
AFASES - 2008 -
II. Instalaii de captare a energiei valurilor pentru zonele din apropierea rmului i
pentru zonele de larg
Instalaii cu coloan oscilant - OWC (oscillating water column). Constau n
principal dintr-o incint n care valurile care ptrund formeaz o coloan oscilant. Aceast
coloan acioneaz un volum de aer care trece printr-o turbin de aer unisens;
Instalaii tip TAPCHAN. Constau dintr-un bazin n care valurile care vin printr-un
canal special amenajat fac ca apa s se acumuleze la un nivel superior nivelului mrii.
Diferena de nivel obinut permite alimentarea unor turbine;
Instalaii pendulare cu panou articulat. ntr-un spaiu special amenajat un panou
oscileaz datorit valurilor care se propag pe orizontal i acioneaz o pomp hidraulic.
Pompa alimenteaz o turbin hidraulic.
Instalaia danez cu pomp i flotor. Este o instalaie n care un flotor acioneaz
o pomp ancorat de fundul mrii care acioneaz turbine hidraulice;
Instalaia suedez HOSEPUMP. Se bazeaz pe un cilindru realizat din elastomeri
care antrenat de un flotor permite expulzarea apei din interior, alimentarea unei pompe
hidraulice i acionarea unei turbine;
Instalaia McCABE WAVE PUMP . Const din nite pontoane care se mic fa de
un ponton central, acionnd pompe hidraulice care acioneaz turbine hidraulice.
Preluarea energiei valurilor se realizeaz prin: coloan oscilant format de apa
mrii; flotoare de diferite forme i dimensiuni; panouri oscilante; acumularea volumului de
ap ridicat de valuri ntr-un bazin[2,3,4,5].
Acionarea direct a generatoarelor electrice se realizeaz cu turbine de aer i turbine
hidraulice.
Agentul de acionare a turbinelor poate fi: aerul; apa de mare; lichidul acionat n circuit
nchis.
Din analiza acestor elemente de caracterizare se constat c nu apar elemente de
acionare mecanic a generatoarelor cum sunt: multiplicatoarele de vitez, bielele, manivelele,
frei-urile i volanii. Acionarea generatoarelor centralelor de valuri actuale se realizeaz
direct de turbine sau cu ajutorul pompelor care acioneaz turbine cuplate cu generatoare.
Aceasta se explic, prin discrepana care exist ntre viteza necesar generatoarelor electice i
viteza de deplasare a flotoarelor. Datorit acestui fapt se evit utilizarea dispozitivelor
mecanice de multiplicare a vitezei flotoarelor care conduc la un randament global al
instalaiilor redus i se utilizeaz turbine pentru aer sau lichide.
Dintre toate tipurile de instalaii pentru captarea i conversia energiei valurilor se
constat c cele care prezint cea mai mare atenie sunt instalaiile hidropneumatice cu
coloan oscilant, cunoscute sub denumirea de instalaii OWC (oscillating water column).
Printre avantajele instalaiilor hidropneumatice se pot enumera:
- schema cinematic foarte simpl;
- lipsa sistemelor de multiplicare a vitezei, turbina fiind cuplat direct cu
generatorul;
- nu folosesc agent lichid pentru transmitera energiei, evitnd utilizarea
etanrilor;
- nu au structuri cinematice n contact cu apa mrii;
Subansamblul cel mai important al acestor instalaii este turbina unisens.
607
AFASES - 2008 -
608
AFASES - 2008 -
motivul c este ntubat, turbina unisens propus are o eficien net superior turbinelor
Wells.
Microcentrala este alctuit dintr-un cheson prevzut cu o deschidere spre largul mrii i cu o
deschidere circular n plafon peste care se asambleaz o coloan cilindric. n aceast
coloan este montat un modul generator alctuit dintr-o turbin unisens i un generator trifazat
cu puterea nominal de 5 kW
Coloana cilindric, montat vertical deasupra chesonului este prevzut cu o deschidere n
partea superioar pentru vehicularea aerului acionat de valuri. Protecia modulului generator
este asigurat de coloan care este etan si deschis la o nlime la care nu ajung valurile de
mare amplitudine.
Deschiderea chesonului spre larg este amenajat pentru captarea cu eficien att a valurilor
mari ct i a valurilor mici.
Instalaia este susinut de 6 piloni ncastrai ntr-o plac de beton armat care permite
aezarea intalaiei pe un pat amenajat pe fundul mrii i lestarea ansamblului.
Pentru obinerea unui randament superior modulul generator al microcentralei este prevzut
cu un generator trifazat cu excitaie realizat din magnei permaneni din pmnturi rare.
Funcionarea instalaiei propuse pentru cercetare se bazeaz pe urmtoarele principii:
Spre deosebire de instalaiile hidropneumatice cu coloan oscilant cunoscute, n
instalaia propus valurile ptrund pe direcia lor de propagare, evitnd schimbarea direciei
fluxului energetic al valurilor i permind captarea att a valurilor mari ct i a valurilor mici;
609
AFASES - 2008 -
610
AFASES - 2008 -
Bibliografie
[1] Iulian C., Lazr P.D, Energia valurilor. Ed. tiinific i enciclopedic, Bucureti, 1982
[2] Tnsescu F.T. Conversia energiei - Tehnici neconvenionale. Ed. Tehnic, 1986
[3] Olaru Gh., Lazr P.D. Rotor de turbin pentru conversia energiei valurilor.
Brevet Ro. 104576/1989
[4] Iulian C. Utilizarea energiei valurilor. Ed. Tehnic, Bucureti. 1990
[5] Olaru Gh., Lazr P.D. Contribuii privind implementarea digurilor energetice n largul
Mrii Negre. n: Energetica, Mart- Apr. 1990
[6] *** M.E.E - ISPH . Lucrarea nr.3509 Contract 2539/1984, Studiu privind realizarea
unei centrale electrice marine Sintez
[7] *** Wavegen a world leader in marine renewable energy. httm://wwww. wavegen.co.uk
Gheorghe Samoilescu, profesor univ.dr.ing. Academia Naval " Mircea cel Btrn , Constana, email
samoilescugheorghe@yahoo.com
** *
Barbu, Alina, assist.univ.Dr. Academia Naval " Mircea cel Btrn , Constana, email
alucu013@yahoo.com
611
AFASES - 2008 -
Capitolul I
I. Etapele modelrii
Se consider un ecran n care cele dou plci metalice plan paralele au grosimea d,
conductivitatea , permeabilitatea , fiind separate prin distana 2x0.
Modelarea cmpului de reacie indus ntr-un perete metalic al navei, utiliznd metodologia de
calcul a unui ecran plan-paralel, are trei faze, i anume: 1. calculul cmpului n interiorul
ecranului; 2. calculului cmpului n exteriorul ecranului; 3. calculul cmpului n peretele
ecranului. [1,3,4,6]
612
AFASES - 2008 -
2
d 2 Hyi
2
i = j 0 = j
=
Hy
i
i
( )
0 ,
dx 2
; Hy i = Hy i x ;
Hyi ( x ) = C1 e j0 x + C 2 e j0 x ; x x0
Hyi ( x ) = Hyi ( x ) ; C 1 = C 2 = C
(1)
(2)
(3)
Hyi ( x) = 2C cos(0 x ) ;
(4)
dEz i = j 0 Hyi dx ; Ez i = Ez i ( x) ;
(5)
Aceast relaie este valabil n oricare din cele trei regiuni-exterior, interior, perete ecran.
Dac s-ar neglija curentul de deplasare :
Ez i ( x ) = j 0 Hyi x + C
(6)
( )
iar dac acest curent nu se neglijeaz, Ezi x va rezulta prin integrarea ecuaiei (5).
Ecuaiile de propagare fiind liniare, iar excitaia (unda incident la ecran) fiind
considerat armonic, cmpul interior va fi armonic.:
Ez i ( x ) = j 0 Hyi ( x )dx + C ; C = 0
(7)
respectiv:
(8)
(9)
Ez i ( x ) = j
0
Hi0 sin ( 0 x ) + C
0
,
613
(10)
AFASES - 2008 -
respectiv
Ez i (x ) = j
0
Hi0 sin ( 0 x )
0
; C = 0,
(11)
=
=
=
c0 T = 0 c 0 =
2
2
2f
0 2
0 0
0 Z 0
=
,
0
(12)
(13)
Z 0 = 120
unde :
se obine:
Ez i ( x ) = j
Z0
(14)
(16)
(17)
i:
Hy e ( x) = C 2 = He
(18)
unde He este cmpul emis de surs, la distan de ecran, avnd o valoare constant i
cunoscut, putnd fi msurat. Cmpul magnetic exterior devine:
Hy e ( x ) = He e _ j0 x x x0 + d
;
(19)
Cmpul electric din exteriorul ecranului se obine din ecuaia:
dEze ( x) = j0 Hye (x )dx ;
(20)
(21)
;He = ct; x x0 + d
614
AFASES - 2008 -
;
Cmpul electric este de forma:
Ez p ( x ) =
p
x
-
Ae - Be
px
) = Zs(Ae
px
(22)
(23)
- Be
- p x
(24)
+
Ht ( x0 ) = Ht ( x0 )
Ht ( x0 + d ) = Ht ( x0 + d )
Et (x0 ) = Et ( x0 )
unde:
Ht ( x0 )
(25)
+
Ht ( x0 + d ) - cmpul Hy (x), pentru x=x +d; Ht (x0 + d )+ - cmpul Hy (x), pentru x=x +d;
p
0
e
0
Et (x0 )+
Et ( x0 ) - cmpul Ez (x), pentru x=x ;
- cmpul Ezp (x), pentru x=x0.
i
0
innd cont de aceste condiii, se obin relaiile:
Hi0 cos( 0 x0 ) = Ae
Ae
p ( x0 + d )
+ Be
p x0
+ Be
p ( x0 + d )
p x0
= He e j0 ( x0 + d )
jZ 0 Hi0 sin( 0 x0 ) = Zs Ae
p x0
Be
p x0
(26)
se obine:
Ae
Ae
p x0
p ( x0 + d )
Ae
p x0
+ Be
+ Be
p x0
= Hi0 cos( 0 x0 )
p ( x0 + d )
Be
p x0
= He e
= j
j0 ( x0 + d )
(27)
Z0
Hi0 sin( 0 x0 )
Zs
(28)
(29)
615
AFASES - 2008 -
Se calculeaz:
Ez ( x )
rot E = j
x
Se scrie legea induciei electromagnetice:
B
rot E =
t ,
care devine n complex:
rot E = jBy j = jHy i j
(30)
(31)
.
Prin compararea relaiilor (30) i (31) se obine:
dEz ( x ) = jHy i ( x )dx .
(32)
(33)
(34)
(35)
Avnd ecuaiile de propagare liniare i excitaie armonic, cmpul interior fiind
armonic, deci C=0; ca urmare:
Ez i ( x ) = j 0 Hi0 cos( 0 x0 )x .
(36)
Dac n relaia (33) se nlocuiete Hy(x)cu relaia sa din perete, (23), se obine:
x
x
dEz p (x) = j Ae p + Be p dx
,
(37)
2
j
= p / rezult:
prin integrare i notnd
Ez p (x) =
p
x
x
Ae Be
.
(38)
p
Ez p (x) =
Ae
x p0
Be
p x0
(39)
616
AFASES - 2008 -
Ae
p x0
unde:
K=
Be
- p x0
= KHi0 cos ( 0 x0 )
(41)
0
p x0
; p= perete.
Hi0
[cos( 0 x0 ) + K 0 sin( 0 x0 )]
2
;
Hi0
p ( x0 + d ) j0 ( x0 + d )
B = He e
e
(42)
Din relaiile (42), se calculeaz cmpul Hi0, n centrul ecranului, precum i constantele A i
B, n funcie de He. nlocuind pe A i B n relaia (23) i punnd x=x0+d, se obine cmpul
magnetic de reacie la suprafaa peretelui exterior:
(x +d )
( x +d )
Hyp e = Ae p 0 + Be p 0
(43)
V. Concluzii
Bibliografie
[1] Sotir, A., Samoilescu, Ghe, Compatibilitate electromagnetic. Baze teoretice. Ecranarea
antiperturbativ, Ed. Academiei Navale " Mircea cel Btrn , Constana, 2001
[2] Mocanu, C. I., Teoria cmpului electromagnetic, EDP, Bucureti, 1981
[3] Hadrian, W., Measuring tests for shielding effectiveness of an enclosure of energy
distribution devices used in building complexes, EMC 80
[4] Nicolau, E., Radiaia i propagarea undelor electromagnetice, Ed. Academiei RSR,
Bucureti, 1989
[5] Schwab, J.A., Compatibilitatea Electromagnetic, Ed. Tehnic, Bucureti, 1996
[6] Oltu, O, Elemente de teoria ecranrii antiperturbative, Ed. Universitatea Politehnica
Bucureti,1994
617
AFASES - 2008 -
Gheorghe Samoilescu, profesor univ.dr.ing. Academia Naval " Mircea cel Btrn , Constana, email
samoilescugheorghe@yahoo.com
** *
Barbu, Alina, assist.univ.Dr. Academia Naval " Mircea cel Btrn , Constana, email
alucu013@yahoo.com
618
AFASES - 2008 -
Resume
This paperwork presents an analysis of the functioning in commutation of a
single-phased transformer with charge pump, made in a Kaap scheme. The commutation
elementh is a diode. The commutation study is done for the two working regimes: periodic
and aperiodical.
I. INTRODUCTION
The simplified equivalent scheme of the transformer in commutation with a diode in
the secondary circuit, with charge pump is presented in figure 1.
Fig. 1. The diode commutation of the (Kaap) transformer with RC charge in parallel
In diode conductions lenght of time, the sistem of equations associated to the circuit is the
following:
d i S'
=
=
+
+ R i S'
u
U
sin(
t
)
R
i
L
1
M
d
t
(1)
'
i = i ' + i ' ; i ' = R ' C ' d i S ; u ' = R i '
S
C
C
S
S
dt
619
AFASES - 2008 -
In diode blockages lenght of time, the sistem of equations associated to the circuit is the
following:
i = 0 ; i 'C = i S' ; u S' = R ' i S'
(2)
The recurrence relations for the circuits currents expressions are the following:
The n-st commutation
Diode D conducts t pn + 2 (n 1) , n + 2 (n 1)
[ t 2 ( n 1 ) p n ]
Hn
'
i S = I S sin (t ) +
cos ( t c n ) e
cos (c n )
'
[ t 2 ( n 1) p n ]
Hn
(
)
(
)
i
I
tg
cos
t
sin
t
b
c
e
=
C S
n
cos b cos (c n )
(4)
I
i = S sin (t + ) +
cos
Hn
tg
[t 2( n 1) p n ]
+ I
cos( t c n )
sin (t + b c n ) e
S cos (c n )
cos b
where it is noted:
u n = 2 (n 1) + p n ; p n = cos u n ; q n = sin u n
p + qn
q pn
(5)
mn = n
;nn = n
R 1 tg tg min
An = t
sin P n sin p n
R
cos
R 1 tg tg min
Bn = t
sin P n + cos p n
R
cos tg
A m Bn p n
H n = B n q n A n n n ; tg c n = n n
Bn q n A n n n
From the condition i [t = 2 ( n 1 ) + n ] = 0 results the expression of the blockage angle n :
[
[
]
]
sin ( n + )
Hn
+
cos (2 ( n 1) + n ) c n
cos
cos(c n )
tg
( n p n )
sin ( 2 ( n 1) + n ) + b c n e
=0
cos b
620
(6)
AFASES - 2008 -
From the initial condition the value of the current trough the
diode enters in its blockage lenght of time.
Hk
I S n = I S sin ( n ) +
cos ( 2( n 1) + n ) c n
cos(c n )
the
condition
( n p n )
e
(7)
(8)
R I S n
2 + p ( n + 1 ) n ctg
(9)
UM
The value of the current trough the charge when the diode enters in its conduction lenght of
time is:
R
U
1 tg tg min
(10)
I S p ( n +1) = M sin p ( n +1) = t I S
sin p ( n +1)
R
R
cos
The expressions of the relative values are the following:
We refer the value of the current trough the charge (R) to the value of the current when the
charge is pure resistive:
i'
i 'r =
(11)
I SM
Diode D conducts (the n-st time) t 2 (n 1) + p n , 2 (n 1 ) + n
[ t 2 ( n 1) p n ]
Hn
i S' , r = d sin (t ) +
cos ( t c n ) e
cos (c n )
'
[ t 2 ( n 1) p n ]
Hn
sin (t + b c n ) e
i C ,r = d tg cos (t )
cos b cos (c n )
d
sin (t + ) +
i r =
cos
[t 2 ( n 1) p n ]
+ d H n cos( t c ) tg sin (t + b c ) e
n
n
cos b
cos (c n )
The relative value of the current trough the charge when the diode enters in its
lenght of time.
Hk
( n p n )
I S n ,r = d sin ( n ) +
cos ( 2( n 1) + n ) c n e
cos(c n )
621
(12)
blockage
(13)
(14)
AFASES - 2008 -
R
(2 + p ( n + 1 ) n ) ctg
(15)
I S n , r cos min e
Rt
The relative value of the current trough the charge when the diode enters in its conduction
lenght of time is:
R
R
1 tg tg min
1
I S p ( n + 1 ) , r = t d
sin p ( n +1) = t
sin p ( n +1)
(16)
R
cos
R cos min
sin p ( n +1) =
III. THE COMMUTATION STUDY WHEN THE FREE REGIME SOLUTION HAS
A DEEMED APERIODICAL CHARACTER
The recurrence relations established for the circuit currents expressions are the following:
R
U
1 tg tg min
(17)
I Sp 2 = M sin p 2 = t I S
sin p 2
R
R
cos
The n-st commutation
Diode D conducts t pn + 2 (n 1) , n + 2 (n 1)
The initial conditions are the following:
i S' t = p n + 2 (n 1) = I Sp n
(18)
'
i C t = p n + 2 (n 1) = I Sp n
The expressions of the currents are:
[ t 2 ( n 1) p n ]
'
i 'C = I S tg [ cos (t ) +
[ t 2 ( n 1) p n ]
(
)
(
)
[
]
(19)
+
H
tg
c
tg
b
ch
t
1
tg
b
tg
c
sh
t
e
n
n
n
i = I S sin (t + ) + I H + tg (tg c tg b ) ch t +
S
n
n
cos
[ t 2 ( n 1) p n ]
+
tgc
tg
(
1
tgb
tgc
)
sh
e
+
n
n
where it has been noted:
q pn
p qn
u n = 2 (n 1) + p n ; p n = ch u n ; q n = sh u n ; m n = n
;nn = n
[
[
]
]
R t 1 tg tg min
sin P n sin p n
R
cos
R 1 tg tg min
Bn = t
sin P n + cos p n
R
cos tg
A m Bn p n
H n = B n q n A n n n ; tg c n = n n
Bn q n A n n n
An =
622
)
)
(20)
AFASES - 2008 From the condition i [t = 2 ( n 1 ) + n ] = 0 we obtain the expression for the blockage
angle n :
sin ( n + )
+ Hn
cos
+ tg ( tgc n tgb) ch (2 (n 1) + n ) +
( n p n )
+ tgc n + tg (1 tgb tgc n )sh (2 (n 1) + n ) e
=0
(21)
From the initial condition we determine the value of the current trough the charge when the
diode enters in its blockage lenght of time:
I S n = I S sin ( n ) +
( n p n )
H n ch (2 ( n 1) + n ) + tgc n sh ( 2 (n 1) + n ) e
The expression of the current trough the charge:
i S' = I S n e [t 2 ( n 1) n ] ctg
+ IS
u S'
sin p ( n +1) =
R I S n
(22)
(23)
2 + p ( n + 1 ) n ctg
(24)
UM
we obtained the expression of the conduction entrance angle:
The value of the current trough the charge when the diode enters in its conduction lenght of
time is:
R
U
1 tg tg min
I p ( n +1) = M sin p ( n +1) = t I S
sin p ( n +1)
(25)
R
R
cos
The expressions of the relative values are the following:
Using relation (12) results that:
Diode D conducts (the n-st time) t 2 (n 1) + p n , 2 (n 1 ) + n
'
[ t 2 ( n 1) p n ]
'
i C , r = d tg [ cos (t ) +
[ t 2 ( n 1) p n ]
(26)
+ H n [ (tg c n tg b ) ch t + (1 tg b tg c n )sh t ]e
i = d sin (t + ) + d H + tg (tg c tg b ) ch t +
n
n
r cos
[ t 2 ( n 1) p n ]
The relative value of the current trough the charge when the diode enters in its blockage
lenght of time.
623
( n p n )
(
)
(
)
ch
2
(
n
1
)
tgc
sh
2
(
n
1
)
+
n
n
n
st
Diode D is blocked (the n- time) t 2 (n 1) + n , 2 n + p ( n +1)
+ d Hn
(27)
(28)
I p ( n + 1 ) , r =
RI S n
2 + p ( n +1) n ctg
R t 1 tg tg min
R
1
d
sin p ( n +1) = t
sin p ( n +1)
R
cos
R cos min
(30)
Bibliography:
[1] Matlac, I.: Convertoare electroenergetice, Editura Facla, Timioara, 1987.
[2] Matlac, I., Cernat, M., Szekely, I.: Convertoare electrice, Universitatea Braov, 1975.
[3] Matlac, I.,Marinescu, C., Helera, E.: Convertoare electrice, vol I, Universitatea
Transilvania Braov, 1994.
[4] Nicolaide, A.; Maini Electrice, Editura Scrisul Romnesc, Craiova, 1975.
[5] Strmbu, C.: Comutaia periodic a circuitului RC paralel cu un tiristor. Sesiunea
Naional de Comunicri tiinifice, Academia Aviaiei i Aprrii Antiaeriene Henri
Coand i Centrul Regional pentru Managementul Resurselor Umane, Braov, 12
noiembrie 2002, pp 247- 256, ISSN 1453-0139.
[6] Strmbu, C.: Consideraii asupra funcionrii redresoarelor de tensiune monofazate,
Sesiunea Naional de Comunicri tiinifice, Academia Forelor Aeriene Henri
Coand, Braov , 26 27 noiembrie 2004, pag 196-201, ISSN 1453-0139.
[7] Strmbu, C.; Cozlov, L.; Chita, E. M.: Regimul Tranzitoriu al Comutaiei Periodice n
c.a. cu Diod a unui Circuit RC Paralel, International Session of Scientific Papers
Education and Scientific Research at European Standards, Henri Coand Air Force
Academy, Braov 11-12 may 2007, Romania, vol II, pp 686-691, ISBN 978-973-841545-4;
[8] Strmbu, C.; Cozlov, L.; Chita, E. M.: Regimul Tranzitoriu al Comutaiei Periodice n
ca cu Tiristor a unui Circuit RC Paralel, International Session of Scientific Papers
Education and Scientific Research at European Standards, Henri Coand Air Force
Academy, Braov 11-12 may 2007, Romania, vol II, pp 662-700, ISBN 978-973-841545-4;
Constantin Strmbu ef lucrri dr.ing., Academia Forelor Aeriene Braov, Mihai Viteazu nr. 160, e-mail
strimbu_c@yahoo.com
*
Laura Cozlov masterand la Universitatea Bucureti, Facultatea de Limbi Strine, secia Interpretare de
Conferin, Sublocotenent, Baza 90 Transport Aerian Otopeni, e-mail laura_cozlov_2003@yahoo.com
624
AFASES - 2008 -
Resume
This paperwork presents an analysis of the functioning of the single-phased
transformer with charge pump, modelled by a Kaap scheme. The theoretical values that the
capacitor can have are determined in the first part of this article, using the phaser diagrams
asociated to the scheme and the condition that the charge must behave like a capacitor. The
mathematical expressions of the current flowing through the charge are presented next in the
situation in witch the transformer is functioning in the two work regimes: periodical and
aperiodical.
I. THE PHASE DIAGRAMS ASOCIATED TO THE CIRCUIT
The simplified equivalent scheme of the charge pump transformer is presented in
figure 1.
625
AFASES - 2008 -
k 2U
The phasors diagram associated to the charge pump transformer in Kaap scheme is
presented in figure 2.
R ' = k 2U R ; C ' =
Fig. 2 The phasors diagram of the charge pump transformer in Kaap scheme
The phase differences between the values are noted like this:
tg = R ' C ' 0 , phase difference between I and U S' ;
r
r
R ' C ' =
tg > 0 , phase difference between U '2 and U S' ;
tg =
r+R
r+R
2
X 1 + tg R ' tg
, phase difference between U1 and I ;
tg 1 =
R t + R tg 2
tg 2 =
R ' tg
(2)
(3)
(4)
(5)
R + r 1 + tg
R tg + X
(6)
tg =
> 0 , phase difference between U1 and U S' ;
R t X tg
Because the charge is behaving like a capacitor it is necessary that tg 1 0 , from were it
results that for a given transformer (the equivalent scheme parameters being known) the
capacitors capacity must fulfill the following condition:
'
2
2
'
' 2
' 2
R R 4 (X ) R + R 4 (X )
'
(7)
C
,
2 R ' X
2 R ' X
'
( )
( )
Fig. 3. The phasors diagram of the Kaap scheme transformer with resistiv charge
626
AFASES - 2008 -
Fig. 4. The phasors diagram of the Kaap scheme transformer, with pure capacitive charge
In this case you can observe that:
X
tg 1C = tg 1 ; tg 1C = lim tg 1
C
R
where tg 1C is the phase difference between U1 and I ;
(10)
d i S'
u
U
sin(
t
)
R
i
L
=
=
+
+ R i S'
1
M
d
t
i = i ' + i '
S
C
'
i ' = R ' C ' d i S
C
dt
u S' = R i S'
The differential equation, non homogeneous of second degree is:
d 2 iS 1
R di
U
1 R
+ ' ' + S + ' ' t i S = ' M'
sin t
2
dt
R C L
R C L
R C L dt
Noting:
R
1
1 R
2 = ' ' + ; 02 = ' ' t ; = 2 02
R C L
R C L
you observe that:
2
1 tg 1C + tg 02
1
=
; 2=
; tg = 2
2 tg 1C tg
tg tg min
0 2
from (13) and (14) the differential equation becomes:
d 2 i S'
d i S'
U M 2
2 '
2
+
i
=
sin t
0 S
dt
X tg
dt 2
of witch generals solution is:
i S' = i S' , l + i S' , f
627
(11)
(12)
(13)
(14)
(15)
AFASES - 2008 -
Determining the particular solution of the non homogeneous equation (the forced
component of the current) i S' , f :
i S' , f = I S sin (t )
(16)
where:
U M cos
sin
1
= UM
=
2
X + R tg
X tg
1
(17)
U
cos
cos
= UM
= M
R t X tg R t 1 tg tg min
you can see that the maximum value of the current through the charge it is obtained when the
value of the C capacitor is very low (resistive charge).
U
(18)
I SM = M cos min
Rt
UM
cos
I
R 1 tg tg min
cos
1
We note: d = S = t
=
(19)
UM
1 tg tg min cos min
I SM
cos min
Rt
The determination of the general solution of the differential homogeneous equation
(the free component of the current) i S' , l .
IS =
L
C ' ' 2 R ' + R t 2 R ' R t , ' 2 R ' + R t + 2 R ' R t
R (R )
R (R )
[(
[(
(20)
(21)
The constants K1 respectively K2, are determined from the initial conditions.
The free regime solution has a deemed non periodic character 2 > 02
From (13) results the condition:
L
L
C ' 0 , ' 2 R ' + R t 2 R ' R t ' 2 R ' + R t + 2 R ' R t , (22)
R (R )
R (R )
[(
[(
628
(23)
AFASES - 2008 -
Bibliography:
[1] Matlac, I.: Convertoare electroenergetice, Editura Facla, Timioara, 1987.
[2] Matlac, I., Cernat, M., Szekely, I.: Convertoare electrice, Universitatea Braov, 1975.
[3] Matlac, I.,Marinescu, C., Helera, E.: Convertoare electrice, vol I, Universitatea
Transilvania Braov, 1994.
[4] Nicolaide, A.; Maini Electrice, Editura Scrisul Romnesc, Craiova, 1975.
[5] Strmbu, C.: Comutaia periodic a circuitului RC paralel cu un tiristor. Sesiunea
Naional de Comunicri tiinifice, Academia Aviaiei i Aprrii Antiaeriene Henri
Coand i Centrul Regional pentru Managementul Resurselor Umane, Braov, 12
noiembrie 2002, pp 247- 256, ISSN 1453-0139.
[6] Strmbu, C.: Consideraii asupra funcionrii redresoarelor de tensiune monofazate,
Sesiunea Naional de Comunicri tiinifice, Academia Forelor Aeriene Henri
Coand, Braov , 26 27 noiembrie 2004, pag 196-201, ISSN 1453-0139.
[7] Strmbu, C.; Cozlov, L.; Chita, E. M.: Regimul Tranzitoriu al Comutaiei Periodice n
c.a. cu Diod a unui Circuit RC Paralel, International Session of Scientific Papers
Education and Scientific Research at European Standards, Henri Coand Air Force
Academy, Braov 11-12 may 2007, Romania, vol II, pp 686-691, ISBN 978-973-841545-4;
[8] Strmbu, C.; Cozlov, L.; Chita, E. M.: Regimul Tranzitoriu al Comutaiei Periodice n
ca cu Tiristor a unui Circuit RC Paralel, International Session of Scientific Papers
Education and Scientific Research at European Standards, Henri Coand Air Force
Academy, Braov 11-12 may 2007, Romania, vol II, pp 662-700, ISBN 978-973-841545-4;
Constantin Strmbu ef lucrri dr.ing., Academia Forelor Aeriene Braov, Mihai Viteazu nr. 160, e-mail
strimbu_c@yahoo.com
*
Laura Cozlov masterand la Universitatea Bucureti, Facultatea de Limbi Strine, secia Interpretare de
Conferin, Sublocotenent, Baza 90 Transport Aerian Otopeni, e-mail laura_cozlov_2003@yahoo.com
629
AFASES - 2008 -
The space mapping techniques supplied designers with strong and reliable tools necessary for the
complex and demanding needs of modern circuit design. They utilize the circuit responses and possibly
derivative information in the optimization loop. Recently, commercial software packages have been developed
that solve Maxwells equations for circuits of arbitrary geometrical shapes. Such simulators are denoted as
Electromagnetic (EM) simulators. They utilize different methods of the analysis of microwave circuits such as
the Finite Element Method (FEM), the Method of Moments (MoM),etc. These simulators are accurate but they
require intensive CPU time. The optimization algorithms employ the coarse model in the search for the fine
model minimizer. This is done through a parameter mapping, the so called space mapping, which in effect makes
the coarse model behave as the fine model. We call this combination of the space mapping and the coarse model,
the mapped coarse model. Hence, in the space mapping technique, applied on microwave RLC circuits, mapped
coarse model is to take the place of the fine model in search for a minimizer model. Beginning with space
mapping conditions we can present some software algorithms for optimization algorithms applied on microwave
RLC circuits models.
I. Introduction
The initial use of these EM simulators was limited to validating designs obtained
through traditional optimization of empirical/analytical models. Over the years, empirical and
circuit theoretic models of many microwave circuits have been developed and accumulated.
The empirical and circuit-theoretic models are denoted as coarse models. Advances in the
technology of workstations and PCs enabled traditional EM optimization of simple structures.
However, the increasing complexity of microwave circuits still makes traditional EM
optimization a formidable task. A mathematical link (mapping) is established between the
spaces of the parameters of the empirical and EM models. This approach directs the bulk of
the required CPU time to the fast model while preserving the accuracy and confidence
supplied by few EM analyses. The target of circuit optimization is to determine a set of values
for the circuit parameters such that certain design specifications are satisfied. These
specifications represent constraints on the circuit responses. Usually, a model of the physical
circuit is utilized in simulating and thus optimizing the circuit. Traditional optimization
techniques utilize the simulated circuit responses directly and possibly available derivatives.
Engineering models used in simulating the circuit responses vary in accuracy and speed. In
some engineering problems, applying traditional optimization using the accurate models
directly may be prohibitively impractical. On the other hand, applying optimization using the
less accurate models may indicate feasibility of the design but could lead to unreliable results.
630
AFASES - 2008 -
These results must be validated using the accurate models or even using
measurements. It follows that alternative optimization approaches must be utilized.
The problems to be solved by the optimization algorithms in this presentation have two
general models available: first model denoted the fine model, being the model of primary
interest, and the second denoted the coarse model. We are proposing for our analyze four
models covering the general models algoritms. It is expected that the coarse model somehow
resembles the behavior of the fine model. Further, it is expected that the coarse model is
cheaper to evaluate than the fine model, and therefore it is most likely less accurate than the
fine model. The optimization algorithms employ the coarse model in the search for the fine
model minimizer. We call this combination of the space mapping and the coarse model, the
mapped coarse model.
Freq
FINE
MODEL
Rf
xf
COARSE
MODEL
Rf
w
FIG.1 ALG1
631
AFASES - 2008 -
ALG2: Hybrid Space Mapping Model. The function ALG2 implements the hybrid space
mapping algorithm, with a gradual switching between the mapped coarse model and the linear
Taylor model of the fine model.
TAYLOR LINIAR
Freq
Rf
FINE
MODEL
xf
OPTIMIZ
COARSE
MODEL
xc
Rc Rf
w
FIG.2 ALG2
ALG3: Hybrid Space Mapping Model with Orthogonal Updates. The function Alg3
implements a hybrid space mapping algorithm with orthogonal updating steps of the space
mapping approximation. If the space mapping fails within the first n iterations the algorithm
evaluates the fine model at a step in a direction orthogonal to previous steps, this is in order to
improve the quality of the space mapping secant approximation. Which of the orthogonal
directions that is chosen and the length of the step in that direction can be controlled by the
options ortho met, ortho scale type and ortho scale. If a single orthogonal step is not
sufficient, further steps are taken, until the fine model has been evaluated at most n times.
Thereafter the algorithm switches to a linear Taylor model of the fine model.
Freq
FINE
MODEL
Rf
xf
COARSE
MODEL
ORTHOGONAL
Rf
w
FIG3. ALG3
ALG4: Hybrid Space Mapping Model with Response Correction. The function ALG4
implements a hybrid space mapping algorithm with response correction of the mapped coarse
model.
632
AFASES - 2008 -
Freq
Rf
FINE
MODEL
xf
COARSE
MODEL
xc
CORRECTION
Rc Rf
w
FIG.4 ALG4
The RLC microwave models implementation for all four algorithms was run together
with the convergence tests on all SM models and the results for all problems are presented in
[10]. The exact non-linear mapping (RLCB) results are represented in Table 1.
Alg.1
Alg.2
Alg.3
Alg.4
100
1
1
1
1
10-1
65
2
2
2
-2
10
13
24
23
-3
10
59
67
10-4
83
-5
10
93
10-6
123
-8
10
139
10-10
148
-12
10
152
-14
10
154
STOP
VE
VE
VE
VK
Table 1: Exact non-linear mapping (RLCB)
VK : The algorithm stop because the fine model evaluation reached the condition (3)
VE : The algorithm stop because the time of the last tentative step was to small
III. Conclusions
The original space mapping algorithms are in general not preferable over any other
algorithms we have studied and presented in this paper, exept special cases presented in [10].
It is most likely a better choice to initialize a direct, classical method with Jacobian
approximation obtained from the course model, than to use the original space mapping
algorithms. The Hybrid Space Mapping algorithms showed good results for initial
convergence and for microwave RLC circuits the Hybrid Space Mapping with Response
Correction are recommended because of the good final convergence better to that of the
direct, classical method started in the course model minimizer.
633
AFASES - 2008 -
IV. Bibliography
[1] Paulo j. Rodrigues Computer-aided analysis of nonlinear microwave circuits Artech
house inc. Boston,1998 / 2003
[2] Stephen Mass Nonlinear microwave circuits , Artech House inc. boston,1998
[3] Microwave Office 3.2 User Manual, 2004
[4] Alexandru Stefan, Constantin Strambu Simularea circuitelor de microunde, Editura
albastra, Cluj-Napoca, 2000
[5] Alexandru Stefan, Indrumar de laborator pentru studiul dispozitivelor de microunde,
Academia Aviatiei, Brasov, 1997
[6] Alexandru Stefan Semnale, Circuite, Sisteme , vol.1 Academia Aviatiei, Brasov, 2001
[7] Constantin Strimbu, Alexandru Stefan Analiza semnalelor cu pspice Editura albastra,
Cluj-Napoca, 2000
[8] Alexandu Stefan Extragerea parametrilor circuitelor de microunde, Simpozion de
referate, Institutul de cercetari militare, Bucuresti, 2002
[9] Alexandru Stefan Optimizarea circuitelor de microunde folosind metoda suprapunerii
spaiale Academia Tehnica Militara, 2002
[10] Alexandru Stefan Research Contract 1013 01/03/03: Microwave Circuits Models
Optimization, V Entr., NJ, USA 2003
[11] J.W.Bandler, M.A.Ismail, J.E.Rayas-Sanchez, and Q.J.Zhang, Neuromodeling of
Microwave Circuits Exploiting Space-Mapping Technology, IEEE Trans. Microwave
Theory Tech., vol. 47, no. 12, pp. 24172427,Dec. 1999.
[12] J.W.Bandler, R.M.Biernacki, S.H.Chen, P.A.Grobelny, and R.H.Hemmers, Space
Mapping Technique for Electromagnetic Optimization, IEEE Trans. Microwave Theory
Tech., vol. 42, no. 12, pp. 25362544, Dec. 1994.
Lt.col(r) Dr. ing. Alexandru STEFAN, Air Force Academy, Brasov ROMANIA
634
AFASES - 2008 -
635
AFASES - 2008 -
x& = Ax + Bu, y = Cx + Du ,
(3.1)
where A represents a state matrix, B is input matrix, C is output matrix, and finally, D is a
direct feedforward matrix, respectively, and x is a state vector, and u represents input vector.
Block diagram of the closed loop control system built by equations (3.1) can be seen in
Figure 3.1., when feedforward matrix is zero, i.e. D=0.
Optimal control law can be determined evaluating the following integral performance criteria
[10, 11, 12]:
J=
1 T
T
x Qx + u Ru dt Min .
2 0
(3.2)
In this cost function main design parameters are weights on state vector Q0, i.e. it is a
positive semi-definite matrix, and weights on control vector R>0, i.e. it is a positive definite
weighting matrix. If Q is very large relative to R one can get a closed loop system response
with large overshoots. If R is chosen to be very large relative to Q control system has smaller
636
AFASES - 2008 -
actuators, electric motors, amplifier gains and other devices. During controller synthesis
weighting matrices can be derived using the so-called inverse square rule.
The LQ optimal control problem can be solved using wide variety of techniques. Let us
consider method of Euler-Lagrange equations, Hamilton-Jacobi-Bellman theory and
Pontriagin's minimum principle. Firstly let us define the so-called Hamiltonian matrix:
H ( x, , t ) =
1 T T
x Qx + u T RuT + T (Ax + Bu ) ,
2
(3.3)
H
H
H
= x& ;
= & ;
=0.
x
u
(3.4)
Using rules for differentiation of matrices and vectors equations of (3.4) can be rewritten in
the following manner
x& = Ax + Bu, x(0) = x o ,
(3.5)
& = Qx + A T , (T ) = 0 ,
(3.6)
u o = R 1B T .
(3.7)
Formula (3.7) defines the optimal control law of the closed loop control system. The coupled
equations (3.5), (3.6) and (3.7) are main equations for the so-called two point boundary value
problem (TPBWP). Substituting equation of control law (3.7) into state equation (3.5) results
in following formula:
x& A BR 1 B T x
x
=
H
& =
.
A T
Q
(3.8)
=Px,
(3.9)
where P is the so-called cost matrix. Differentiating both sides of equation (3.9) with respect
to time and considering equations (3.5) and (3.7) the following equation can be derived:
d dP
dx dP
=
x+P
=
x + PAx PBR 1B T Px = Qx A T Px .
dt
dt
dt dt
(3.10)
The sufficient condition for optimal control is that P must satisfy the following Ricatti
differential equation:
dP
= A T P + PA + Q PBR 1B T P, P (T) = 0 .
dt
(3.11)
Solution of the optimal controller synthesis problem using Ricatti equation in control theory is
regarded as the finite time problem. This solution results in the linear time varying static
controller of the feedback, i.e.:
u o (t ) = K (t )x(t ), K (t ) = R 1 B T P(t ) .
(3.12)
637
AFASES - 2008 -
Equation (3.11) is a nonlinear first order matrix differential equation, which has to be solved
backwards in time. During solution of the infinite time LQR problem it is considered that
T . It is obvious that under mild conditions cost matrix P can be considered as constant
and solution of Ricatti equation results in the asymptotically stable closed loop control
system. In this particular case equation (3.11) can be rewritten as:
A T P + PA + Q PBR 1 B T P = 0 ,
(3.13)
(3.14)
In modern control theory equation (3.13) is known as algebraic Ricatti equation (ARE).
Conditions defined by equations (3.13) and (3.14) are necessary and sufficient for existence
of the optimal controller, which will asymptotically stabilize the closed loop control system.
Thus, procedure of optimal control law synthesis consists of the following two steps:
1, solution of the ARE equation (3.13) in order to find constant cost matrix P,
2, substitution cost matrix P into equation (3.14). The resulting feedback gain matrix K is an
optimal for the chosen weighting matrices Q and R.
This chapter deals with methodology of the closed loop automatic flight control system
preliminary design, and analysis, thus, gives main guideline for control engineers. The closed
loop automatic flight control system will be synthesized using optimal Linear Quadratic
Regulator (LQR) design method outlined in deep details in Chapter III.
Let us consider automatic flight control system used for height control system, which is given
in Figure 4.1. The block diagram is general, and suggested to use it for height control of the
UAV first time by the authors.
Block diagram given in Figure 4.1 represents the full state feedback control system.
Dynamics of the UAV is represented by the following transfer function:
Y (s) =
v ac ( s)
A
5
=
=
,
E ( s) 1 + sT 1 + 2 s
(4.1)
638
AFASES - 2008 -
theoretically, however transfer function must represent a slow dynamics with pure dynamic
characteristics. Airspeed sensor is represented by transfer function of Ys1 ( s) , and its transfer
function is one of the design parameters. Barometric altimeter is given with its transfer
function of Ys2 ( s) , which will be supposed to have unity gain. Note that this model is
simplified to that of the possible simplest one. The series compensator is represented with
transfer function of Yc (s ) and, it is the second design parameter to be found.
4.1. Analysis of the Uncontrolled UAV Dynamics
The open loop UAV was tested both in time and in frequency domain. Result of the computer
simulation can be seen in Figure 4.2.
Figure 4.2-1.
Figure 4.2-1.
Figure 4.2-3.
Figure 4.2-4.
Figure 4.2-1 shows impulse response, Figure 4.2-2 shows step response of the hypothetical
UAV dynamics. From these figures it is easily can be seen that UAV dynamics has first order
term feature. Response functions are exponential ones, the UAV responses to inputs very
slowly. Regarding dynamic performances, flying and handling qualities given in [8, 9, 14, 19]
it can be said that dynamic performances shown by UAV dynamics out of the required ranges.
Figure 4.2-3 represents transient response of the open loop system to square signal having
period of time of T = 10 sec , or frequency of f = 0,1 Hz . As this figure shows, having slow
dynamics UAV is unable to follow the reference input of the square signal. Thus, UAV needs
flight control system making it able to achieve fast responses, in other words,
fast maneuvers. Analysis in the frequency domain shows that UAV has good dynamic
performances, i.e. gain margin is Gm = , and phase margin is Pm = 102 0 . From this argue
comes out the evidence of the need of the automatic flight control system to control the height
639
AFASES - 2008 -
of the UAV. There are many classical and modern methods, which can be used for control
system preliminary design. Regarding problems of UAV systems design, for example design
of the energy system, the more feasible methods are limited to optimal design methods.
4.2. Gain Selection Using LQR Design Method
Before start design of the height control system shown in Figure 41. let us define dynamic
performances of the closed, and open loop dynamic characteristics and requirements to be as
follows in Table 1. [14, 19]:
Table 1. Requirements of the closed loop control system.
Gain margin: Gm 10 dB
Before using LQR design procedure to find system parameters, closed loop control system
given in Figure 4.1. in complex frequency domain, it must transferred to the following one,
given in the time domain. Block diagram of the time domain closed loop control system given
in Figure 4.3.
Figure 4.3. State Space Model of the Closed Loop Control System.
Using basic equations of the inverse Laplacetransformation, matrices, and vectors used in
Figure 4.3. easily can be found to be as follows:
1
x& = Ax + Bu = T
1
A
0,5 0 v ac 2,5
0 v ac
+
E
H + T E = 1
0 H 0
0 0
.
1
0
v
0
ac
y = Cx + Du =
+ E
0 1 H 0
(4.2)
K = Ys1
Yc .
(4.3)
During controller synthesis for closed loop control system given in Figure 4.3. first step must
be done is selection of the weights of the integral criteria defined by equation of (3.2).
There are many well-known methods for selection of the weights, such as inverse square rule,
unity weighting, and, finally, heuristic set of weights. Having no information about limitations
640
AFASES - 2008 -
of the automatic flight control system we will choose for the first step unity weighting for
matrices Q, and R, respectively, i.e.:
1 0
Q1 =
; R1 = r1 = 1 .
0 1
(4.4)
For the unity weighting matrices defined by equation of (4.4) the static feedback gain matrix
is as follows [10, 11, 12, 17, 18]:
(4.5)
The closed lop control system for given static controller was tested for dynamic
performances.
Table 2. Dynamic performances of the closed loop control system.
Eigenvalues
Damping ratio,
1,08
1,08
2,31
2,31
Step response of the control system can be seen in Figure 4.4. From this figure it is evident
that steady-state settling time is tss 3,3 sec , and, damping ratio has unity value. The response
of the closed loop control system to the reference unity gain signal is exponential one.
At the next stage of the control system design find heuristically the appropriate set of
weighting matrices Q, and R providing for the closed loop control system, in this particular
case it is a height control system of the UAV, dynamic performances defined before. After
several attempts of the control system synthesis and analysis carried out by the authors
following set of weighting matrices shall be used for the controller synthesis:
0,1 0
Q2 =
; R 2 = r2 = 15 .
0 15
(4.6)
641
AFASES - 2008 -
For this set of matrices static state feedback gain is as follows [17, 18]:
K 2 = [0,7201 1] ; Ys1 = 0,7201 ; Yc = 1 .
(4.7)
For this particular case dynamic performances were found and put into Table 3.
Table 3. Dynamic performances of the closed loop control system.
Eigenvalues
Damping ratio,
1,15 + 1.08 i
0,727
1,58
1,15 1.08 i
0,727
1,58
The closed loop step response was analyzed. Result of the computer simulation can be seen in
Figure 4.5.
From Figure 4.5. it is easily can be seen that t ss 1,8 sec , and damping ratio is = 0,727 .
Thus, time domain dynamic performances of the closed loop control system are in ranges
with those of defined in Table 1.
Comparison of the two systems designed using special sets of weighting matrices can be
performed using Figure 4.6. Right side shows transient responses of the inner loop, which is
loop for airspeed of the aircraft. It is evident from this figure that change of elements of
weighting matrices to those of given in equation (4.6) result in system, which behaves more
oscillatory than previous one. However, this allows to closed loop control system to react
faster to reference unity input measured in the height of the flight.
642
AFASES - 2008 -
Figure 4.6-1.
Figure 4.6-2.
Finally, let us find Bode diagram of the control system opening outer loop of the height
control system of the UAV given in Figure 4.1. Results of the computer simulation can be
seen in Figure 4.7.
From Figure 4.7. it is evident that frequency domain dynamic performances are as follows:
Gm = , Pm = 66,6 0 .
(4.8)
It is easy to determine that dynamic performances fit frequency domain requirements given in
Table 1. At this point we can finish our preliminary design task.
4.3. Analysis of the Disturbance Rejection Ability
Preliminary design of the automatic flight control system must be followed by analysis of the
disturbance rejection ability of the control system. Figure 4.1 shows vertical gust wg (t )
acting as environmental disturbance affecting motion of the UAV. It is evident that
disturbance wg (t ) depends on many circumstances, and conditions. In this paper not going
into deep detail let us suppose to have following initial data and conditions [20]:
H = 100 m 328,084 feet; U 0 = 25 m / s = 90 km / h ,
0,45 m / s w 1,8 m / s , Lw = 580 m ,
(4.9)
(4.10)
643
AFASES - 2008 -
where H is height of the flight [m], U 0 is airspeed of the UAV [m/s], w r.m.s. is gust
velocity [m/s], and finally, Lw is the integral scale of the atmospheric turbulence [m].
The minimum value of w represents NASA parameter for the vertical gust speed at light
wind, while maximum value of w is defined for moderate the vertical component of the
wind [20]. Time histories of the vertical speed component for minimum, and maximum of w
can be seen in Figure 4.8.
Controller synthesis using LQR method is performed in Section 4.2. System parameters for
given heuristically set weighting parameters were derived to be as follows:
K 2 = [0,7201 1] ; Ys1 = 0,7201 ; Yc = 1 .
(4.11)
Disturbance rejection ability of the closed loop automatic flight control system can be
analyzed using block diagram given in Figure 4.1. Suppose straight equilibrium flight of the
UAV when random turbulence acts on it, while reference signal of the closed loop control
system is zero, H r (t ) = 0 . In this particular case closed loop transfer function of the
automatic flight control system is as follows:
Wd ( s ) =
H ( s)
wg ( s )
=
H r ( s ) =0
2 s + 4,6005
.
2 s + 4,6005s + 5
(4.12)
From equation (4.12) change of the height of the flight easily can be derived as:
H ( s ) = Wd ( s) wg ( s ) .
(4.13)
Using equation (4.13) closed loop control system was analyzed for disturbance rejection.
Results of the computer simulation for light and moderate wind weather conditions can be
seen in Figure 4.9.
644
AFASES - 2008 -
Figure 4.9 shows that closed loop automatic flight control system is unable to maintain
altitude constant, i.e. random turbulent air moves UAV from its equilibrium height for given
static error (see Figure 4.9). For the moderate wind static error can be found as:
H (t ) = H r (t ) H (t ) = H (t ) .
(4.14)
(4.15)
which is 0,7 % of the equilibrium height of H 0 = 100 m . It is easily can be derived that this
inaccuracy means high quality although automatic flight control system uses simple series
Pcontroller.
However, from automatic control theory it is known that unwanted effects from external
disturbances affecting motion of the UAV can be reduced, or removed totally using
PIcontroller in the feedforward path of the automatic flight control system. Let transfer
function of the series compensator is as follows:
1
(4.16)
Yc = 1 + .
s
Using transfer function of equation (4.16) transfer function of the closed loop control system
can be derived as:
Wd ( s ) =
H (s)
wg ( s )
=
H r ( s ) =0
2 s 2 + 4,6005s
.
2 s 3 + 4,6005s 2 + 5s + 5
(4.17)
Applying series controller defined by equation (4.16) in the feedforward path of the closed
loop automatic flight control system given in Figure 4.1 the closed loop control system was
tested for disturbance rejection ability. Results of the computer simulation for light and
moderate wind weather conditions can be seen in Figure 4.10.
645
AFASES - 2008 -
From Figure 4.10 it is evident that application of the PI-controller is minimizing static steady
sate error from the constant component of the vertical random turbulence closed loop control
system, i.e. using final value theorem of the Laplace-transformation yields to:
2s 2 + 4,6005s
=0.
H (t ) = lim sH ( s ) = lim sWd ( s) wg ( s ) = 3
s 0
s 0
2 s + 4,6005s 2 + 5s + 5
t
(4.18)
Figure 4.10 shows that weather conditions affect basically the response from external
disturbance of the vertical wind gust. Figure 4.11. demonstrates responses of the closed loop
control system.
Figure 4.11-1
Figure 4.11-2
Figure 4.11-1 shows responses for light wind weather conditions, and Figure 4.11-2 shows
response from the automatic flight control system for moderate wind meteorological
conditions. Dealing with problems of the analysis of the disturbance rejection ability it is
worth to mention that we have to re-arrange analysis of the reference signal tracking ability of
the closed loop height control system having PI- controller. We have just kept in mind
disadvantages of the integral compensation, regarding automatic control theory. This theory
deals in deep details with filtering problems of the closed loop control systems. Final outcome
from here is to apply PID-controller eliminating partly, of fully disadvantages of the integral
compensation.
646
AFASES - 2008 -
Figure 4.12-1
Figure 4.12-2
Figure 4.12-1 demonstrates behavior of the closed loop control system in the time domain.
Step responses show that applying PI-controller dynamic performances provided with the
simple P-controller are going worse, i.e. steady-state response time increased to that of
t ss 8,7 sec achieving large value of the overshoot. In frequency domain it can be stated that
phase margin goes worse having value of Pm 22,8 deg . Evaluating these results both form
reference tracking and disturbance rejection it is easy can be determined that control engineers
must agree in compromises allowing satisfaction of all pre-defined dynamic performances.
Disadvantages of application of the PI-controllers can be omitted using PID-controllers.
In our example we have considered following transfer function found heuristically of the
series controller:
Yc = 1 +
1
+ 0,25s .
s
(4.19)
Using transfer function defined by equation (4.19) height control system was tested. Results
of the computer simulation are summarized in Figure 4.13.
Figure 4.13-1
Figure 4.13-2
Figure 4.13-1 demonstrates responses from all closed loop control systems having P-, PI, and
PID-controllers investigated before in previous sections. Results of the time domain analysis
647
AFASES - 2008 -
of the height control system show that PID controller improves dynamic performances both
in time and in frequency domains, i.e. steady-state response time is decreased, and phase
margin is increased (see Figure 4.13-2). From results of the computer simulation it is obvious
that further investigation of the PID-controller tuning is required.
[2]
[3]
[4]
[5]
[6]
[7]
[8] MIL-F-9490D, Notice 1, Flight Control Systems Design, Installation, and Test of
Piloted Aircraft, General Specification, 1992.
[9] MIL-C-18244A, Amendment 1, Control and Stabilization System: Automatic, Piloted
Aircraft, General Specification, 1993.
648
AFASES - 2008 -
* Dr. habil., Rbert SZABOLCSI *Mikls Zrnyi National Defense University, Associate Professor, Director H1581 Budapest, POB. 15, Hungary Email: szabolcsi.robert@zmne.hu
**
Gyrgy MSZROS QualiProx Ltd., Managing Director ,H-8630 Balatonboglr, Mricz Zsigmond Str. 2/a.,
Hungary, Email: qprox@externet.hu
649
AFASES - 2008 -
650
AFASES - 2008 -
R=
AB AP
AB
unde AB AP este aria paralelogramului cu laturile AB i AP.
(3)
N y = ( y yi ) (z j zi ) (x j xi ) ( z zi )
(4)
R12 = ( x xi ) + ( y yi ) + (z zi )
(5)
N z = ( z zi ) (z j zi ) ( x xi ) ( y j yi )
Avnd n vedere i expresiile
2
2
2
AB 2 = (xi x j ) + ( y i y j ) + (z i z j )
2
R22 = (x x j ) + ( y y j ) + (z z j )
2
rezult:
R =
2
N x2 + N y2 + N z2
;
AB 2
R 2 + AB 2 R22
;
cos 1 = 1
2 R1 AB
(6)
R22 + AB 2 R12
2 R2 AB
Cosinuii directori ai induciei magnetice sunt deci:
N
N
N
cos x = x ; cos y = y ; cos z = z
N
N
N
cos 2 =
(7)
unde N = N x2 + N y2 + N z2
Fig.1
n situaia c toate spirele sunt parcurse de acelai curent I/NS, componentele cmpului
magnetic rezultant n punctul P se obin prin efectuarea sumelor
651
AFASES - 2008 -
Fig.2
NS
Nl
B xt =
S =1 l =1
NS
Nl
B yt =
S =1 l =1
1 cos 1 + cos 2
cos x 10 7 I = Fxt 10 7 I
NS
R
1 cos 1 + cos 2
cos y 10 7 I = Fyt 10 7 I
NS
R
(8)
1 cos 1 + cos 2
cos z 10 7 I = Fzt 10 7 I
N
R
S =1 l =1
S
iar cmpul rezultant este:
NS
Nl
B zt =
(9)
2. Programul de calcul
652
AFASES - 2008 -
START
NL, NS, NP, XPk, YPk
zPk, k=1, NP
FXTk=0
FYTk=0, k=1, NP
FZTk=0
IS=1
NU
DA
k=1, NP
Fig.3
653
IS=NS+1
AFASES - 2008 -
CALL LOOP
k=1
FXk=0
FYk=0
FZk=0
I=1
J=I+1
Nx=
N=
DA
Ny= Nz=
AB2= R2=
R>0,2gx
NU
cos 2 =
F = (cos 1 + cos 2 ) / R
FXk=FXk + F . Nx/NS
FYk=FYk + F . Ny/NS
FZk=FZk + F . Nz/NS
DA
I=NL
DA
NU
j=1
k=k+1
NU
k=NP
i=i+1
RETURN
DA
J=NL
NU
Fig.4
2.1. Subprogramul de calcul al cmpului magnetic produs de o spir
1
2
2
3
3 . i .
4 i+1
654
NL-1
NL
Tabelul 1
NL
1
AFASES - 2008 -
Bibliografie
Dan George Ton, prof. univ. dr. ing., Universitatea din Oradea, Facultatea de Inginerie electric i Tehnologia
Informaiei, e-mail: dtont@uoradea.ro,
**
Gabriela Ton, conf. dr. ing., Universitatea din Oradea, Facultatea de Inginerie electric i Tehnologia
Informaiei, e-mail: gtont@uoradea.ro,
655
AFASES - 2008 -
Fig.1
656
AFASES - 2008 -
Fig.2
Fig.3
Numrul de spire al unei bobine este NS=M . N unde
M=(d2 - d1)/ d - numr de spire pe strat
N=(a2 - a1)/ a
- numr de straturi.
Soluia exact a cmpului ntr-un punct P (o, o, z) de pe axa oz, rezult trecnd la
limit pentru a i d foarte mici. Efectund calculele se obine:
Bzex = Fgex . 10-7 I
(2)
2
S
[(d 2 + z )ln a 2 +
a1 +
(d1 + z )ln
a 22 + (d 2 + z )
a12 + (d 2 + z )
a2 + a22 + (d1 + z )
a1 + a12 + (d1 + z )
+ (d 2 z )ln
(d1 z )ln
a2 +
a 22 + (d 2 z )
a1 +
a12 + (d 2 z )
2
a2 + a22 + (d1 z )
2
2
a1 + a1 + (d1 z )
(3)
Bxw = Fxw 10 7 I
Bzw = Fzw 10 7 I
(4)
unde
Fxw =
) ]
xz
0,00719 31 2 36 2 + 0,51792 4 z 2 3x 2 + ...
5
R
657
(5)
AFASES - 2008 -
4 0,71553
0,008
31 2 36 2 2 z 2 x 2 .
Fzw =
+
1
4
4
60
R
R
R
)(
0,144 4
8 z 24 z 2 x 2 + 3 x 4 + ...].
R4
(5/)
a1 + a2
, = a2 a1 , = d 2 d1
2
Considernd dimensiunile R=0,5m, d=0,25m, =0,04m i =0,02m, s-a calculat
cmpul magnetic pentru diferite variante ale numerelor Nl i NS rezultatele fiind date
parial n tabelul 1.
Din rezultatele calculului se desprind urmtoarele concluzii:
- n puncte situate pe axa de simetrie a bobinelor, la toate variantele considerate,
F Fgex
eroarea
nu
depete
valoarea
de
0,02%,
1 = z
100
Fgex
iar R =
( Fz =
cos 1 + cos 2
cos z ).
R
- Relaiile (5) sunt suficient de exacte numai pentru puncte situate n jurul originii.
F Fz
Astfel eroarea 2 = zw
100 rmne sub valoarea 0,2% pentru x 0,2 i z 0,1
Fz
x
z
( 0,4; 0,4
R
d
Pentru puncte situate la distane apropiate de R, eroarea 2 este foarte mare, fiind necesar ca
n rel.5 s se ia n considerare mai muli termeni ai seriei.
X
[m]
0
0
0
0
0
0
0,1
0,2
0,2
0,4
Z
[m]
0
0,1
0,2
0,3
0,4
0,5
0,1
0,2
0,4
0,4
Ns
N
Fz
17,9834
17,9517
17,5407
16,2051
13,8908
11,1365
18,0355
18,7052
14,3162
13,2688
1
24
17,9816
17,9514
17,5430
16,2079
13,8916
11,1352
18,0348
18,7113
14,3159
13,2333
8
24
Fzx
17,9817
17,9514
17,5430
17,2079
13,8916
11,1352
18,0348
18,7113
14,3160
13,2335
8
48
17,9815
17,9513
17,5429
16,2078
13,8915
11,1352
18,0347
18,7111
14,3159
13,2335
32
24
17,9816
17,9514
17,5429
16,2078
13,8915
11,1352
18,0347
18,7111
14,3159
13,2337
32
48
17,9813
17,9498
17,4574
15,3110
9,5216
-2,6948
18,036
18,8563
15,6837
31,7832
Tabelul 1
Fg ex
17,9828
17,9518
17,5424
16,2078
13,8891
11,1367
-
658
AFASES - 2008 -
superpoziie, rezultatele obinute aici cu privire la erori se pot extrapola i pentru programul
general de calcul.
Fig.4
Bara cu seciune dreptunghiular din fig.4 se descompune ntr-un numr NS de bare
elementare cu seciunea gx . gy astfel nct S = a b = N S g x g y .
Inducia magnetic n punctul P(x, y) dat de bara curent (xi, yi) este:
0 I
B=
2R N S
unde I este curentul total din bar. Componentele cmpului dup axele x i y sunt:
yi y
cos 1 + cos 2
2
10 7 I =
cos x 10 7 I
Bx =
2
2
R
N S ( x xi ) + ( y y i )
(6)
x xi
cos 1 + cos 2
2
10 7 I =
cos y 10 7 I
(7)
2
2
R
N S ( x xi ) + ( y y i )
nsumnd pentru cele N bare se obine cmpul total:
NS
cos 1 + cos 2
B xt =
cos x 10 7 I = Fxt 10 7 I
R
1
NS
cos 1 + cos 2
B yt =
cos y 10 7 I = Fyt 10 7 I
(8)
R
1
Expresia exact a cmpului se obine trecnd la limit pentru gx i gy foarte mici i
By =
rezult
1
x ( x, y ) 10 7 I = Fx ex 10 7 I
a b
1
=
y ( x, y ) 10 7 I = Fy ex 10 7 I
a b
B x ex =
B y ex
unde:
x (x, y ) = ( x a ) ln
+ 2 y arctg
( x a )2 + y 2
( x a )2 + ( y b ) 2
(9)
x 2 + ( y b)
+
x2 + y2
2
+ x ln
ay
a( y b )
+ 2( y b )arctg
+
y + x( x a )
( y b )2 + x ( x a )
2
+ 2( y b )arctg
a( y b )
( y b )2 + x ( x a )
659
(10)
AFASES - 2008 -
(x a )
y ( x, y ) = ( y b ) ln
2
+ ( y b)
+ y ln
x2 + y2
( x a )2 + y 2
bx
b( x a )
+ 2 x arctg 2
+ 2( x a )arctg
x + y( y b)
( x a )2 + y ( y b )
x + ( y b)
(10/)
n tabelul 2 se dau parial rezultatele calcului aproximativ pentru mai multe variante de
mprire a seciunii transversale a unei bare ptrate (0,4 x 0,4m2) i ale calculului exact n
cteva puncte din exteriorul barei parcurs de curent, puncte situate pe axa y=b/2=0,2m.
Tabelul 2
y [m]
0,2
0,2
0,2
0,2
0,2
0,2
0,2
0,2
0,2
0,2
0,2
0,2
x[m]
0,4
0,401
0,402
0,403
0,405
0,41
0,415
0,425
0,44
0,48
0,55
0,70
Ns
Fy
10,000
9,9502
9,9009
9,8522
9,7561
9,5238
9,3023
8,8888
8,3333
7,1428
5,7142
4,0000
1
8,3336
8,3250
8,3155
8,3051
8,2816
8,2105
8,1254
7,9290
8,4970
8,4879
8,4769
8,4641
8,4342
8,3404
8,2305
7,9921
8,5296
8,5203
8,5085
8,4947
8,4621
8,3602
8,5514
8,5418
8,5293
8,5146
8,479
8,5785
8,5684
8,5547
8,5383
8,499
16
64
100
144
256
Fy ex
8,6600
8,6324
8,6049
8,5774
8,5229
8,3882
8,2561
7,9995
7,6332
6,7634
5,5710
3,9735
660
AFASES - 2008 -
de bobin de form rectilinie, caz n care scade timpul de calcul (Nl rezult foarte mic), iar
soluia aproximativ poate deveni chiar exact.
Programul nu poate fi folosit dac nu se cunoate repartiia de curent pe seciunea
transversal a bobinei, ca de exemplu n regim variabil cu luarea n considerare a efectului
pelicular.
Dac se pune problema calculului cmpului magnetic n interiorul laturii de bobin
exist dou posibiliti de descompunere n spire elementare i anume:
- punctul n care se calculeaz cmpul se situeaz ntre dou laturi de spir;
- acest punct s fie pe axa seciunii transversale a laturii de spir.
n ultimul caz latura respectiv trebuie eliminat din calcul (deoarece R=0, iar cmpul
corespunztor ei este nul n axa laturii din spir). Programul de calcul se poate aplica n
ambele situaii dar se recomand a doua variant, caz n care numrul de spire NS rezult mai
mic, avnd n vedere d p g / 2 , unde g este cel mai mic dintre gx i gy.
4. Bibliografie
Dan George Ton, prof. univ. dr. ing., Universitatea din Oradea, Facultatea de Inginerie electric i Tehnologia
Informaiei, e-mail: dtont@uoradea.ro,
**
Gabriela Ton, conf. dr. ing., Universitatea din Oradea, Facultatea de Inginerie electric i Tehnologia
Informaiei, e-mail: gtont@uoradea.ro,
661
AFASES - 2008 -
Abstract: The most obvious rapprochement between the areas of signal processing
and that of dynamical systems is their very fundamental interest: the analysis, prediction and
control of real world physical systems, either natural or man-made. This occurred at the level
of analysis and modeling tools, either research area borrowing with great ease methods and
concepts from another. This paper reviews several such convincing examples. The question of
the chaotic systems is also addressed. While already used in large bandwidth
communications, their influence is shown to be extendable even further. On this line, their use
in sea electromagnetic echo is presented.
Chapter I - Introduction
A relatively large number of physical systems, either natural or man-made, are studied
under two paradigms: either in the traditional and reductionist way, either from a holistic,
systemic perspective. Usually, the classical signal processing (SP) techniques, aimed at
extracting pieces of information about the analyzed system (or signal) falls in the first
category. This is not to say that the two domains are dichotomic. At the contrary, they
strongly interfere, by the very nature of their object of study.
Many often encountered systems which may be analyzed and modeled under a
systemic paradigm are nonlinear and a part of them are even chaotic ones. One could
enumerate: the meteorological phenomena, the eco-systems, the social and economical
systems, the financial markets, several electronic circuits (like oscillators), chemical reactions,
and communications and urban/highway traffic etc.
Many of these may be very easily modeled by using a dynamical system (DS)
perspective. On the other hand, their analysis using the classical reductionist analysis tools
may become overwhelming.
662
AFASES - 2008 -
Another common point of dynamical systems and signal processing techniques lies in
the study of nonlinearity. The latter benefits of an increasing interest from researchers in
signal processing area, allegedly being the buzzword of the day. It is its considerable
modeling power, which may be applicable to even very complex systems, which justifies this
research trend, despite the underlying difficulties. On the other hand, the dynamical systems
had assumed, for many years now, hypothesis of nonlinear behaviors in their studies. As such,
is very appealing and it should be easy for signal processing researchers to benefit of this
existing expertise.
One such example, where nonlinearity is not only assumed and intensely studied, but
is also a prerequisite, is the study of chaotic systems. The chaotic systems are either discrete
or continuous time dynamical (whose state is time-varying) systems which exhibits apparent
disorder, similar to randomness, while still remaining completely deterministic. In fact, the
proper sense of the term chaotic should be read as extreme dependence of initial
conditions. The figures representing such systems often have very interesting and
complicated shapes, widely known as strange attractors, most of them being, in fact,
fractals.
The sensitivity to the initial conditions translates in increasing divergence in system
evolution. After n steps, this divergence is assumed to bind to an exponential model:
f n ( x + ) f n ( x) = exp(n )
(1)
where is the Lyapunov exponent. This quantity is positive for chaotic systems.
Some chaotic models, as will be shown, already made their way into traditional signal
processing areas, such as in radar noise modeling. However, they remain a relatively new
concept for signal processing researchers. Good introductory works are, by example, [1] and
[2].
Chapter II Applying DS knowledge to SP
One of the first breakthroughs of the DS theory into a domain traditionally belonging
to the signal processing area is the proposed use of chaotic models for the high-resolution,
low grazing angle sea clutter returns [3], in the case of coastal maritime surveillance radars.
While still subject to discussions and not fully accepted by the scientific community, the
subject is of great importance nonetheless.
The traditional ways of doing with sea clutter and the degradation of target detection
are by use of Neyman-Pearson or Bayesian hypothesis testing criteria. They require the
probability distribution of noise amplitude (and of the target echo) to be properly modeled.
While for target echo a number of models have been defined, notably the Swerling models
[4], the noise was generally accepted as white and Gaussian.
The problem is that, for large bandwidth radars and for low grazing angles, the
Gaussian hypothesis is far from true. This misassumption greatly penalized the detection
performance of the first large bandwidth coastal radars, despite the higher resolution leading
to a better SNR.
Different others probability density functions have been proposed for the sea clutter
amplitude, serious candidates being Weibull [5], lognormal [6] or K-compound [7].
However, all of them failed to achieve general consensus. More, the estimation of their
underlying parameters is often tedious.
In this context, supported by results obtained while analyzing the published IPIXDartmouth [8] radar dataset, professor Haykin made the hypothesis of an underlying purely
deterministic, albeit chaotic, model of the sea clutter. His first results [3] were encouragingly,
showing sea clutter exhibiting some well-known characteristics of dynamical systems, such as
positive Lyapunov exponent and a fractal dimension.
663
AFASES - 2008 -
However, the presented results were questioned, when a number of publications [9]
shown their ambiguity. The main source of confusion lied in the very similar behavior that
some particular purely random (i.e. stochastic, not chaotic) processes exhibit. It is the case of
1/ f processes, whose spectral density is:
s ( f ) = f (2 H +1)
(2)
where H (0,1) is the Hurst parameter.
While the outcome is still debatable in terms of theoretical soundness, other scientists
[10] developed Haykins original idea, by proposing a new kind of chaotic model:
1
x(t )
= 1 + (1 q ) q t 1 q
(t ) = lim
(3)
x (0) 0 x (0)
which is a generalization of (1).
The latter paper shown surprisingly good results in terms of naval target detection,
based on the different values that = (1 q) 1 in the presence or absence of a target (see
Figure 1). Values between 0.8 and 1 indicate that a target is present in the analyzed
resolution cell.
Another interesting and very useful application of methods originating in the DS field
is presented in [11]. This time, one of the most central concepts of the signal processing is put
under the loupe: the stationarity.
This concept, while mathematically formalized in a rigorous way, is largely
interpretable in the framework of a particular application.
Figure 1: Values of for 14 range resolution cells; the higher values denotes presence of target
By example, when random processes are considered, stationarity stands for temporal
invariance of signals statistics. On the other hand, when deterministic signals are studied,
signals stationarity means the invariance of its spectral properties. Finally, claiming that a
signal is stationary or not is only pertinent with respect to a particular interval of time or
frequencies, the pertinence of which depends on the problem at hand. This localization
clearly contradicts with the permanence notion associated with the stationarity concept. The
same signal may appear as stationary or not, function of the length of analyzing interval
which, basically, defines the scale at which the signal is studied.
In an intuitive way, for a given average spectrum, a stationary signal should present no
temporal structure, i.e. no predictability. This should manifest for stationarry signals only, in
the phase of their spectra.
Based on this observation, a mathematical, albeit practical, definition of stationarity
may be given. It uses the notion of surrogates, which are signals having the same power
spectrum as the original signal, but are stationnarized. This is achieved by replacing the phase
664
AFASES - 2008 -
of original signals spectrum by a random phase, completely destroying its temporal structure
(see Figure 2). A time-frequency representation shows a completely dispersed spectral content
(see Figure 3).
The concept of surrogates originates in the field of DS, being a tool for testing the
nonlinearity [12] and the chaoticity [13].
As acknowledged, the surrogates are stationary signals, having the same spectrum
amplitude as the original signal, but their spectrum phase is noise. The original signal is then
compared against its surrogates. To compare, several tools may be used: the Kullback-Leiber
and the log-spectral divergences [11], neural networks, SVM etc.
The results of this procedure are shown in ... (for amplitude-modulated noise and for a
frequency sinusoidal modulated deterministic signal) on three scales: macro (upper), medium
(middle) and micro (lower). The first and the third are identified as stationary, wile the middle
one is not.
Figure 2: A signal with sinusoidal frequency modulation (left) and one of its substitutes
Figure 3: Spectrograms of the original signal, a surrogate, and the mean over multiple surrogates and the
spectrum
Figure 4: Samples and spectrograms of amplitude modulated noise (left) and frequency sinusoidal
modulated signal
Most of the usual signal processing techniques quickly find their application in the DS
domain, as they are valuable information extraction and characterization tools. At the very
basic level, there are the purely mathematic ones, such as ARMA modeling, PCA, ICA etc.
On a higher level, more specific techniques are used, such as the spectral and
spectrogram representation methods. An illustrative application is presented in [14], showing
the analysis of a Duffing oscillators impulse response. The author prefers these tools instead
665
AFASES - 2008 -
of the more traditional bone curves because they clearly give the information concerning
the evolution of the system in time.
A remarkable contribution of time-frequency (and, generally speaking, signal
processing) methods is in distinguishing between a chaotic and a random system. In fact, this
remains an open issue for the scientific community and there is not any absolute reliable
criterion.
A possible use of time-frequency representation has been presented recently by [15],
exploiting one fundamental characteristic of deterministic chaotic systems: the approximate
(but not identical) recurrence of their behavior and sequence of states.
While the recurrences and their occurrences remain unpredictable, it is possible to
average their time-frequency representations, thus filtering out the measurement noise and the
intrinsic instability of the system, while enhancing the common, deterministic part.
The first step is to identify these recurrences. A method specific to the DS, the time
delay embedding [15] is used for this purpose. Its central idea is that the evolution of all
parameters of a dynamical system may be approximated by a temporal embedding of just only
one known parameter. This translates in constructing the following vectors:
T
z n = zn , zn + , zn + 2 ,..., zn + ( d 1)
(4)
where d is the embedding dimension and is the delay value, and z0 , z1 ,... are the values of
the known parameter.
The neighborhood of radius of z n is defined as:
N n = def {z k : z k z n < ,1 k L (d 1) }
(5)
Figure 5 shows an example for the state x of the chaotic Lorenz attractor of
parameters (10, 28,8 / 3) . The black points are the samples between the extreme elements of
vector (4), which starts, each time, in the reference point or in one of its neighbours.
Figure 5: The reference point (black disk) and two of its neighbours (white disks), along with the associated
embedded vectors (black points)
s n = def zn , zn +1 , zn + 2 ,..., zn + ( d 1)
(6)
Intuitively, the Figure 5 shows that the waveforms s n occur irregularly, but they are
somehow similar. For each neighborhood N n , the matrix Dn = x k1 , x k2 ,..., x kN where
x ki = s ki s n is defined. Then, the covariance matrix is computed: Cn = N 1Dn DTn whose
eigenvalues are i and corresponding eigenvectors are u i . The Neighbourhood-based
Spectral Estimator is then computed as:
1 Lw
PNSE ( ) =
i | Vi ( ) |2
Lw i =1
666
(7)
AFASES - 2008 -
Figure 6: NSE time-frequency representations for a chaotic (left) and random (right) process, along with the
mode frequencies
Acknowledgments
This work was supported by the French Ministry of Foreign Affairs/EGIDE, under the
grant ECO-NET 2007, by Office of Naval Research Global under grant number N00014-081-1074, and by the North Atlantic Treaty Organization, under the grant EAP.RIG.982933.
667
AFASES - 2008 -
References
[1] D. K. Campbell, Nonlinear science from paradigms to practicalities, Los Alamos (Special
Issue), pp. 218-262, 1987.
[2] J. Palis, A view on chaotic dynamical systems, in Brazilian Journal of Physics, vol 24, no.
4, December 1994.
[3] S. Haykin, S. Puthusserypady, Chaotic dynamics of sea clutter, in Chaos, no. 7, pp. 777802, 1997.
[4] P. Swerling, Detection of fluctuating pulsed signals in the presence of noise, IRE Trans.
on Inform. Theory, IT-3, pp. 175-178, Sept. 1957.
[5] F. A. Fay, J. Clarke, R. S. Peters, Weibull distribution applied to sea-clutter, in Proc. IEE
Conf. Radar77, pp. 101103, London, UK, 1977.
[6] H. C. Chan, Radar sea-clutter at low grazing angles, in Proc. Inst. Elect. Eng. F137, pp.
102112, 1990.
[7] E. Jakeman, P. N. Pusey, A model for non-Rayleigh sea echo, in IEEE Trans. Antennas
Propagation 24, pp. 806814, 1976.
[8] http://soma.crl.mcmaster.ca/ipix/
[9] C. P. Unsworth, M. R. Cowper, S. McLaughlin, B. Mulgrew, False detection of chaotic
behavior in the stochastic compound k-distribution model of radar sea clutter, Proc.
10th IEEE Workshop on SSAP, pp. 296-300, August 2000.
[10] J. Hu, J. Gao, K. Yao, Power-law sensitivity to initial conditions in sea clutter, in Radar
Conference, 2005 IEEE International Volume, Issue 9-12, pp. 956 961, doi:
10.1109/RADAR.2005.1435965, May 2005.
[11] J. Xiao, P. Borgnat, P. Flandrin, Testing stationarity with time-frequency surrogates, in
Proceedings of EUSIPCO 2007, pp. 2020-2024, Poznan (Poland), 2007.
[12] T. Schreiber, A. Schmitz, Improved surrogate data for non-linarity tests, in Physics
Revue Letters, vol. 77, no. 4, pp. 635-638, 1996.
[13] M. Small, R. G. Harrison, C. K. Tse, A surrogate test for pseudo-periodic time-series
data, in American Institute of Physics Proceedings of Experimental Chaos: 6th
Experimental Chaos Conference, vol. 622, pp. 190-198, 2002.
[14] S. Braun, M. Feldman, Time-frequency characteristics of non-linear systems, in
Mechanical Systems and Signal Processing, nr. 11, vol. 4, pp. 611-620, 1997.
[15] J. Sun, Y. Zhao, T. Nakamura, M. Small, From phase space to frequency domain: A
time-frequency analysis for chaotic time series, in Physical Review E 76 (article
016220), doi : 10.1103/PhysRevE.76.016220, 2007.
Felix Totir, CS III dr.ing., Military Equipment and Technologies Research Agency, Bucharest, 16,
Aeroportului street, 077060 Clinceni, Romania, phone: 0214233058, email: ftotir@acttm.ro
Bebe Ionacu, CS ing., Military Equipment and Technologies Research Agency, Bucharest, 16, Aeroportului
street, 077060 Clinceni, Romania, phone: 0214233058, email: bionascu@acttm.ro
tefan Popa, CS ing., Military Equipment and Technologies Research Agency, Bucharest, 16, Aeroportului
street, 077060 Clinceni, Romania, phone: 0214233058, email: spopa@acttm.ro
668
AFASES - 2008 -
Abstract:
The aim of this note is to show that the following difference equation:
x
xn +1 = + k 1 n k , n=0,1,, has solutions which monotonically converge to zero. This
ci xni
i =0
result shows the existence of such solutions which were not shown in the recent papers.
Key words: Convergence to zero, Difference equations, Positive nonoscillatory solutions.
2000 Mathematics Subject Classifiction: 39A10.
1.
INTRODUCTION
In [4] the author investigates the behavior of solutions of the difference equation
x
xn +1 = + n 1 ; n = 0,1,,
(1)
xn
where < 1 . The case > 0 was considered in [1] and a generalization of Eq. (1) was
considered in [6]. Asymptotics for solutions of difference equations have been investigated by
L. Berg and S. Stevic (see for example [3], [6]).
The problem is solved by constructing two appropriate sequences yn and zn with
yn xn0 zn
(2)
for suficiently large n. Let us describe briefly a Berg-Stevic method which can be used in
proving that a difference equation has monotonic solutions. For a difference equation with the
equilibrium x we consider its linearized equation about the equilibrium. If the characteristic
equation of the linearized equation has a zero, (0,1) , we can assume that the difference
equation has solutions with the following asymptotics
x + n + o( 2n ) .
(3)
At this point we require a theorem which will guaranty the existence of solutions that have
the needed asymptotics. A quite general result will be proved in Section 2.
Consider now the equation
x n +1 = +
xn k
, n=0,1,,
k 1
c x
i =0
(4)
n i
669
AFASES - 2008 -
n-1
i=0
Note that the linearized equation for Eq. (4) about the equilibrium x = + 1 is
k 1
(5)
i =0
For the case > 1 the characteristic polynomial associated with Eq. (5) i.e.
k-1
(6)
i=0
has a positive root t0 belonging to the interval (0,1). To see this, note that P(0) = 1 and
P(1) = + 1 > 0 . This fact made us believe that in this case there are solutions of Eq. (4)
which have the following asymptotics
xn = +1 + a.t0n + 0(t0n ) ,
(7)
( xn +1 ) ci xn i xn k = 0 .
(9)
i =0
ci xn i xn k = 0 .
(10)
i =0
P1 (t ) = ci t k i + 1 = 0
i =0
satisfies P1 (0) = 1 and P1 (1) = 1 + < 0 and that P1 is decreasing on (0,1). Hence P1 has a
unique characteristic zero t = t1, which belongs to the interval (0,1) if < 1 . For such
chosen t we expect that Eq. (4) has solutions which have the first two members in their
asymptotics as in (8). The following result plays a crucial part in proving the main result.
k +2
Theorem 1. Let f : I I be a continuous and non-decreasing function in each
argument on the interval I R ; and let ( yn ) and ( zn ) be sequences with yn < zn for
n n0 and such that
670
(11)
AFASES - 2008 -
(12)
N = n0 + k
(n0 ,..., n0+ k ) A ; then the corresponding solutions of (12) satisfy (2) for all n n0 , as desired.
2. THE MAIN RESULT
In this section we prove the main result in this note:
Theorem 2. For each < 1 there is a nonoscillatory solution of Eq. (4) converging to zero
as n .
Proof. First note that Eq. (4) can be written in the following equivalent form
def
F ( xn k , xn k +1 ,..., xn , xn +1 ) = ( x n+1 )
k 1
c x
i =0
i n i
xn k = 0 .
(13)
We expect that solutions of Eq.(4) converging to zero have the asymptotic approximation
(8). Consider F = F (n k , n k +1 ,..., n , n +1 ) . Then, we have
k 1
k 1
F = (at n +1 + bt 2 n + 2 ) a ci t n i + b ci t 2 n 2i (at n k + bt 2 n 2 k )
i =0
i =0
k 1
k 1
k 1
= at n k ci t n i + 1 + t 2 n a 2 ci t1i ab ci t 2i bt 2 k + o(t 2 n )
i =0
i =0
i =0
nk
2n
2n
= at P1 (t ) + t A(a, b, t ) + o(t )
(14)
where
k 1
k 1
i =0
i =0
1i
2 k
(15)
k 1
P1 (t ) = cit k i + 1 = 0 .
i =0
For such chosen t, choose b R and a 0 such that A(a, b, t0) =0.
Since
k 1
k 1
i =0
i =0
671
(16)
AFASES - 2008 -
k 1
a 2 ci t01i
b=
> 0.
2( k i )
t ci t0
+ 1
i =0
n
2n
n = at0 + qt0 , we obtain
2 k
0
If
i =0
k 1
(17)
zn = n + o(t02 n ) , for n n0 ,
where q = b ln n ; in particular, the solution converges monotonically to zero for n n0 .
Hence, the solution xn + n0 + k is also such a solution when n k .
Remark 1. Since a > 0 is an arbitrary parameter, by Theorem 2 we find a set of
solutions of Eq. (4) converging nonincreasingly to zero.
Remark 2. If we take a < 0 in the proof of Theorem 2 we obtain the set of solutions
of Eq. (4) converging nondecreasingly to zero.
3.
REFERENCES
[1]
[2]
[3]
[4]
[5]
J.
[6]
[7]
672
AFASES - 2008 -
Gheorghe Anton, teacher, Trades and Services High School,Buzau, 14 bis, Bazalt treet,g10.anton@gmail.com
Gheorghe Radu, lecturer, Henri Coanda Air Force Academy, 500183 Brasov, #160, Mihai Viteazul Street,
phone: +40 723 604084, e-mail: gh.radu@gmail.com
673
AFASES - 2008 -
674
AFASES - 2008 -
675
AFASES - 2008 -
676
AFASES - 2008 -
There are very few methods of image segmentation that addressed partitioning and
obtaining content description of segments simultaneously:
In [3], the problem is posed as texture segmentation where the textures are modeled by
Gauss-Markov random fields.
In [22], the problem is posed as joint estimation of the partition and class parameter
variables, and a Bayesian estimation multi-scale image segmentation method is
proposed. The method recognizes the variability of content description depending on
the complexity of the image regions and effectively addresses it. The notion of a class is
introduced as that which gives rise to different segments with same content description.
In particular, the method partitions the data as well as obtains descriptions of classes for
a large family of parametric models. These parametric models are used to describe the
content of the classes. Central to the method is the formulation of a cost functional
defined on the space of image partitions and the class description parameters that can be
minimized in a simple manner.
1.3 Object feature (descriptor) extraction
The result of the segmentation step is a number of regions corresponding to objects in the
analyzed images. The next step is to find properties of these regions : object features that can help
in their identification. These features can refer to the boundary of a region or to its content.
Commonly used boundary features belong to one of the following categories:
o Bounding box;
o Extremity points (top-left, bottom-right);
o Chain codes;
o Simple geometric shape descriptors such as length or curvature;
o Fourier descriptors of the boundary;
o Features derived from the B-spline representation of a contour;
o Fractal dimension;
o Shape invariants such as number of corners/vertices and shape number.
Examples of classes of content (regional) features are:
o Simple scalar region descriptors such as area, compactness, elongatedness;
o Moments;
o Convex hull;
o Graph representation based on region skeleton.
Furthermore, there are descriptors that refer to the relation between regions: a region
is inside another one or one is above another. Examples of techniques to derive such
descriptors are:
o Region decomposition;
o Region neighborhood graphs.
1.4 Object classification
After their descriptors were extracted, objects are classified based on these. It is necessary
that the system has a description of the objects of interest (reference objects), i.e. it has their
descriptors. Having this information, the system compares the descriptors of the objects found in
an image with the descriptors of the reference objects. Based on the result of the comparison, a
decision is taken whether the found object is of a certain type.
677
AFASES - 2008 -
In the ideal case, the system has descriptions of all the objects that can be encountered in
the analyzed images and, consequently, can identify all objects in an image. Practically, a system
has descriptors of a certain set of objects of interest and it is able to identify whether any of those
objects are present in an image.
2. Audio Analysis and Automatic Indexing Techniques
There are some main directions in audio analysis: segmentation, classification and
similarity and retrieval (by content) [12]:
o Segmentation deals with options to split an audio signal into intervals, minding the
semantics or the sound source composition. Audio scenes is an important concept here.
o Classification aims at categorizing segments of audio according to predefined (mid-level)
semantic classes, such as speech, music, silences, background noise, etc. This also
includes possible further differentiation within a class (e.g. according to genre or mood in
the case of music, or speaker identity in the case of speech), or detection of concurrency
of specific classes (e.g. speech and background noise). Furthermore, classification may
also extend towards identifying the genre of entire clips and programs.
o Retrieval by content is concerned with similarity measures, enabling to e.g. retrieve from
an audio database item(s) that perceptually or semantically match the query.
It should be noted that the segmentation and classification are typically interdependent.
Segmentation of an audio signal into shorter frames if often the first step towards classification;
alternatively, pre-classified segments can be fused using different strategies to obtain
segmentation at a higher semantic level.
2.1 General Audio Classification Techniques
Concern the algorithms developed for separating predefined general audio classes, such as
speech, music, silences, background noise, etc. Some of them also extend towards discriminating
sub-classes of a given class, e.g. genres in music or different types of noise and environmental
sound. All the approaches can be characterized as a multi-stage processing, where a feature
extraction.
[18] describe a classification in five basic groups: classical music, popular music, speech,
noise and crowd noise (applauding and cheering), with the popular music containing samples
from seven genres: Jazz, Folk, Electronic, R&B, Rock, Reggae and Vocal. The feature sets
include: (1) low-level signal properties, (2) mel-frequency spectral coefficients, (3) psychoacoustical features including roughness, loudness and sharpness, and (4) an auditory model
representation of temporal envelope fluctuations.
A framework targeted to real-time segmentation and classification of audio from TV
broadcast streams is described in [26]. It adopts common low-level features, such as signal
energy, average zero-crossings rate, fundamental frequency and spectral peak track, and the
usual milliseconds feature extraction rate. A heuristic rule-based procedure is employed for
segmentation and classification, being based on morphological and statistical analysis of the
time-varying functions of the extracted features. The classes considered include: speech, music,
song, environmental sound, speech with music background and silence.
In [17], the classification is performed using support vector machines. The audio clip is
first divided into non-overlapping sub-clips, having the duration of one second. Each of the subclips is classified in two initial categories: speech and non-speech.
678
AFASES - 2008 -
In [7], yet different classes are considered, including laughter, applause and cheer.
Mostly the conventional low-level features are used, including the more perceptual brightness
and MFCC. HMM are used to model the sound effects, and a log-likelihood based scores to make
the final decision. [10] uses the HMMs and GMMs in order to model gunshots, explosions,
engines, helicopter flying and cheers. A GMM, event scores and a sliding window concept are
further used to combine multiple events, as means to model the context of different movie scenes.
2.2 Silence Detection (Speech/silence discrimination)
Belongs to the general audio classification methods.The main approach is in fact the
endpoint detection, which determines the boundaries of words and sentences based on a global
signal energy/zero-crossings threshold scheme [20].
An alternative approach is to seek for exclusion of all silences, rather than to isolate all
utterances. This is expected to improve the silence detection in the presence of (background)
noise [16].
3. Multimedia (Images, Audio and Video) Semantic Content Management
Image processings purpose: to find the segments or regions that are homogeneous
according to some mathematical or statistical criteria (considered a low-level operation).
Computer vision or artificial intelligence (AI) methods are concerned with high level
operations dealing with the content of the image, i.e., understanding the meaning behind the
segments or knowledge representation.
However, recently a combined approach based on content based coding of images and
video is emerging which tries to represent image or video data using the segments within them,
where the content of each segment is characterized mathematically:
Content based segmentation methods (for example, accessing images depicting a sunset):
the initial raw data, both images and video, has to be partitioned such that the description of each
segment is related to its content.
The semantic contents of multimedia data can be managed by two approaches.
1. The first one is to annotate an image, a video, or an audio data with text. Impractical
especially in large multimedia databases: the raw data accompanied by annotations is
assumed to be updated frequently. Another problem: it is very difficult to keep the
consistency of the annotations.
2. The other approach is to provide the system with a rule base or a knowledge base where
knowledge or rules are used to extract features from the raw data, to match content, to
analyze queries, and so forth.
[2], [15], [19], [25]: the knowledge could be associated directly (as fundamental
property) with the content feature values and/or the spatial relationships.
[24]: the knowledge is defined for the subject of interest indirectly
[15]: there is developed a hierarchy called Type Abstraction Hierarchy (TAH) to
define a general level of concepts to detailed level with sets of attribute values. Knowledge is
referred to the evaluation of shapes and spatial relationships of the objects (e.g., a tumor). TAHs
conceptualize the objects and their semantics and incorporate the domain expert knowledge in
order to improve search efficiency in radiological databases.
Other approaches make use of knowledge that interprets semantic contents in terms of
image representations, establishing image features specific to the content. In [19], the meaning of
the image is represented by a keyword that is defined with a description of image features such as
679
AFASES - 2008 -
regions of colors and their locations. The semantic contents are represented by the spatial
composition of the color regions. The hierarchical relation between the primitive color regions
and the semantic contents is captured in a state transition model. In the content-based retrieval
system exposed in [25], the authors define a pseudo attribute which associates a query condition
with a domain knowledge describing the contents to be retrieved. The method for extracting
semantic features from multimedia data, the rules to transform query conditions into an internal
representation whose type is the same as the extracted semantic features, or the rules to transform
a certain operator into content-dependent calculus is called a domain knowledge.
Some issues of this approach: how to interpret a query condition into semantically
equivalent expressions and evaluate the expressions with features extracted from the raw data?
Next, how to keep the knowledge base semantically consistent with the database schema?
A possible solution: to make the knowledge base and the database schema semantically
dependent on each other by integrating them together with rules that prescribe semantic
association of one with the other [23].
Generally speaking, the second approach is more practical than the first one even in large
multimedia databases.
Conclusions
The automatic indexing of images and audio documents is not a trivial problem: in
addition to the complexity of the involved algorithms, the result is often influenced by the large
variability in the appearance of represented objects, depending for example on illumination
conditions, distance to the object, parameters of the photo equipment, etc.
Together with the most relevant and popular indexing techniques, the article presented
also the approaches in multimedia semantic content management. A very promising direction is
to combine the knowledge management techniques with those promoted by the semantic Web
community.
Bibliography
[1] Acuna, C.O., Texture modeling using Gibbs distributions. Computer Vision, Graphics, and
Image Processing: Graphical Models and Image Processing, 1992. 54(3): p. 210-222
[2] J.R. Bach, S. Paul, and R. Jain, A Visual Information Management System for the
Interactive Retrieval of Faces, IEEE Trans. Knowledge and Data Engineering, vol. 5, no.4, 1993
[3] Bouman, C.A., Liu, B., Multiple Resolution Segmentation of Textured Images, IEEE
Trans. Pattern Analysis and Machine Intelligence, vol. 13, no. 2, pp. 99-113, February 1991
[4] Bovik, A.C., M. Clark, and W.S. Geisler, Multichannel texture analysis using localized spatial
filters. IEEE Trans. Pattern Analysis and Machine Intelligence, 1990. 12: p. 55-73
[5] Castleman, K.R., Digital Image Processing. 1996: Prentice Hall
[6] Clark, M. and A.C. Bovik, Experiments in Segmenting Texton Patterns Using Localized
Spatial Filters. Pattern Recognition, 1989. 22: p. 707-717
[7] Rui Cai, Lie Lu, Hong-Jiang Zhang, Lian-Hong Cai, Highlight Sound effects detection in
audio stream, Proceedings of the International Conference on Multimedia and Expo, vol. 3, pp.
37-40, 6-9 July, 2003
[8] Chatterjee, S., Markov Random Fields - Theory and Application, R. Chellappa and A. Jain,
Editors. 1993, AP. p. 159-178
680
AFASES - 2008 -
[9] Chen, Shu-Ching, Kashyap, R. L., Ghafoor, Arif, Semantic Models for Multimedia Database
Searching and Browsing, Kluwer Academic Publishers, NY, 2002, ISBN 0-792-37888-1
[10] Wen-Huang Cheng, Wei-Ta Chu, Ja-Ling Wu, Semantic Context Detection based on
Hierarchical Audio Models, Proceedings of the 5th ACM SIGMM international workshop on
Multimedia information retrieval, Berkeley, California, USA, 2003, pp. 109-115
[11] Daugman, J.G., Complete discrete 2-d Gabor transforms by neural networks for image
analysis and compression. IEEE Trans. Acoustics, Speech, and Signal Processing, vol.,36, 1988
[12] Jonathan Foote, An overview of audio information retrieval, in Multimedia Systems, vol.
7 no. 1, pp. 2-11, ACM Press/Springer-Verlag, January 1999
[13] Gonzales, R.C. and R.E. Woods, Digital Image Processing. Second ed. 2002, Prentice Hall
[14] Haralick, R.M., K. Shanmugam, and I. Dinstein, Textural features for image classification.
IEEE Trans. Systems, Man and Cybernetics, 1973. 3: p. 610-621
[15] C.C. Hsu, W.W. Chu, and R.K. Taira, A Knowledge-Based Approach for Retrieving
Images by Content, IEEE Trans. Knowledge and Data Engineering, vol. 8, no. 4, 1993
[16] Ying Li, C.-C. Jay Kuo, Video content analysis using multimodal information, Kluwer
Academic Publishers, 2003
[17] L. Lu, S. Li, H.-J. Zhang, Content-based audio segmentation using support vector
machines, IEEE International Conference on Multimedia and Expo, 2001 (ICME 2001), pp. 749
- 752, August 2001
[18] Martin F. McKinney, Jeroen Breebaart, Features for Audio and Music Classification,
Proceeding of the International Symposium on Music Information Retrieval (ISMIR 2003), pp.
151-158, Baltimore, Maryland, USA, 2003
[19] A. Ono, M. Amano, M. Hakaridani, T. Satou, and M. Sakauchi, A Flexible Content-Based
Image Retrieval System with Combined Scene Description Keyword, Proc. Intl Conf.
Multimedia Computing and Systems, pp. 201-208, 1996
[20] L. Rabiner and R. Schafer, Digital processing of speech signals, Prentice- Hall Inc., 1978
[21] S. Sista. Image And Video Segmentation Using Unsupervised Classification in a Bayesian
Set up. Ph.D. Thesis, Purdue University, May 1999
[22] S. Sista and R.L. Kashyap, Bayesian Estimation for Multiscale Image Segmentation, IEEE
International Conference on Acoustics, Speech, and Signal Processing, March 1999
[23] A. Yoshitaka and T. Ichikawa, A Survey on Content-Based Retrieval for Multimedia
Databases, IEEE Transactions on Knowledge and Data Engineering, vol. 11, no. 1, pp. 81-93,
January/February 1999.
[24] A. Yoshitaka, T. Ishii, M. Hirakawa, and T. Ichikawa, Content- Based Retrieval of Video
Data by the Grammar of the Film, Proc. Intl Symp. Visual Languages, pp. 314-321, Sept. 1997
[25] A. Yoshitaka, S. Kishida, M. Hirakawa, and T. Ichikawa, Knowledge-Assisted ContentBased Retrieval for Multimedia Databases, IEEE Multimedia, vol. 1, no. 4, pp. 12-21, 1994
[26] Thong Zhang, C.-C. Jay Kuo, Audio Content Analysis for Online Audiovisual Data
Segmentation and Classification, IEEE Transactions on Speech and Audio Processing, vol. 9,
issue 4, pp. 441-457, May 2001
Mihaela Brut, lecturer, Faculty of Computer Science, University Alexandru Ioan Cuza Iai, Berthelot nr. 16,
phone 0232 201544, e-mail: mihaela@infoiasi.ro
681
AFASES - 2008 -
Capitolul I
1.1 Sisteme multi-agent. Noiuni de baz.
Instrumentul folosit pentru realizarea aplicaiei propuse este Zeus Agent Toolkit.
Acesta a fost dezvoltat de ctre British Telecom i mai precis de ctre grupul Sisteme
inteligente i are marele avantaj de a fi gratuit pentru uz academic. Acesta furnizeaz un
mediu grafic care permite dezvoltarea sistemelor agent distribuite, iar limbajul de programare
folosit este Java. Furnizeaz o bibliotec de componente software i instrumente care
faciliteaz dezvoltarea rapid a aplicaiilor agent colaborative i poate rula pe orice platform
care are instalat o main virtual JDK 1.2. Componentele scrise n limbajul de programare
Java, sunt grupate n trei biblioteci funcionale, i anume: biblioteca componentelor tip agent
(Agent Component Library), un instrument de construire a agenilor (Agent Building
Software) i un set de ageni utilitari (Utility Agents) acetia fiind nameserver-ul, facilitatorul
i visualizer-ul[1].
Aplicaiile care se pot rezolva folosind acest instrument, sunt cele orientate pe task-uri,
ca de exemplu: furnizare de servicii, managementul lanului de fabricaie, managementul
682
AFASES - 2008 -
resurselor, precum i managementul proceselor, din aceast ultim categorie fcnd parte i
sistemul multi-agent propus. Lucrul cu instrumentul Zeus Agent Toolkit presupune
cunoaterea i implementarea urmtoarelor noiuni de baz: sistem multi-agent (SMA), agent
inteligent, arhitectur agent, cooperare, comunicare, ontologie, coordonare, tipuri de medii de
lucru, task, descriere PASM (Percepii, Aciuni, Scopuri i Mediu), scop i fapt.
Agenii din cadrul unui sistem multi-agent au asociate task-uri specifice. Pentru a
putea realiza aceste task-uri, agenii trebuie s comunice ntre ei, s existe un anumit grad de
coordonare, cooperare i o anumit ontologie definit anterior.
Limbajele de comunicare folosite pentru ageni, sunt aa numitele ACL-uri (Agent
Comunication Languages), acestea fiind FIPA ACL, KQML i KIF. Coordonarea este acel
proces n care agenii unui SMA se angajeaz s asigure coerena sistemului, fiind necesar
pentru alocarea resurselor i eliminarea conflictelor. n ceea ce privete cooperarea, aceasta
este necesar pentru a obine un comportament global robust al sistemului multi-agent i se
realizeaz prin comunicaia ntre componentele software ale sistemului[2].
Poate cel mai important aspect al unui SMA este acela al definirii ontologiei, ce const
n specificarea unui vocabular cu termenii recunoscui de agenii din sistemul respectiv,
termeni ce apar n mesajele acestora.
n momentul n care se dorete rezolvarea unei aplicaii cu ajutorul Zeus Agent
Toolkit, trebuie parcurse mai multe etape, acestea fiind[3]:
1. Definirea aplicaiei;
2. Analiza aplicaiei, etap ce presupune urmtoarele: specificarea rolurilor agenilor,
ntocmirea listei de responsabiliti pentru fiecare agent, realizarea diagramei fluxului
de informaie, descrierii PASM i a diagramei de interaciune;
3. Proiectarea aplicaiei, etap ce const n modelarea cunoaterii i stabilirea
ontologiei;
4. Realizarea aplicaiei ce presupune: crearea ontologiei, crearea agenilor, stabilirea
task-urilor pentru fiecare agent, configurarea agenilor utilitari i a celor task,
implementarea i rularea aplicaiei.
n cadrul acestei lucrri se va prezenta un mod de implementare a noiunilor de baz i
a etapelor de mai sus, pentru aplicaia propus spre rezolvare.
Sistemele multi-agent se bazeaz pe ideea c ntr-un mediu de lucru cooperativ, care
cuprinde componente software ce au capaciti de rezolvare a problemelor, se pot rezolva
probleme dificil de rezolvat altfel, cu metode tradiionale, aceste sisteme constituind poate
singura soluie pentru rezolvarea anumitor tipuri de probleme, n special a celor complexe.
Capitolul II
2.1 Construirea SMA-ului
Exemplul de sistem multi-agent propus n cadrul acestei lucrri, const n modelarea
activitii didactice depus de ctre un cadru didactic, respectiv de un asistent, abordnd
fiecare aspect al acesteia. Aceast activitate nu se rezum numai la aciunile posibile ale
asistentului i ale studenilor n cadrul unui laborator de informatic, de exemplu, la
interaciunile dintre acetia, ci i la interaciunile care au loc ntre titularul cursului i asistent,
la interaciunile asistentului, respectiv studenilor cu mediul de lucru, acesta fiind, n cazul
aplicaiei date, laboratorul.
n cadrul acestei aplicaii sunt construite patru entiti, fiecare dintre ele fiind capabile
s interacioneze una cu alta, n scopul obinerii ct mai multor informaii, lmuririi aspectelor
683
AFASES - 2008 -
684
AFASES - 2008 -
Fig. 3:
Ontologia
aplicaiei
685
AFASES - 2008 -
686
AFASES - 2008 -
687
AFASES - 2008 -
688
AFASES - 2008 -
Bibliografie
[1] Collins, J., Ndumu, D., The Application Realisation Guide, Inteligent Systems Research
Group, BT Labs, 1999
[2] Collins, J., Ndumu, D., ZEUS Technical Manual, Inteligent Systems Research Group, BT
Labs, 1999
[3] Collins, J., Ndumu, D.,The Role Modelling Guide, , Inteligent Systems Research Group,
BT Labs, 1999
Mdlina Crbureanu, Prep. mat., Univ. Petrol-Gaze, Ploieti, B-dul Bucureti, nr. 39, tel 0244 573171, e-mail:
mcarbureanu04@yahoo.com
**
Irina Tudor, As. ing., Univ. Petrol-Gaze, Ploieti, B-dul Bucureti, nr. 39, tel 0244 573171, e-mail:
trinelle@yahoo.com
689
AFASES - 2008 -
Corina CHIRIAC *
Abstract.We present an iterative algorithm for approximate solutions of mixed quasivariational inequalities in the setting of reflexive Banach spaces.We use Fan-KKM lemma
and concept of -cocoercivity of a composition mapping to prove the existence and
convergence of approximate solutions to the exact solution of quasi- variational inequalities.
INTRODUCTION
Variational inequality theory has stimulated new results dealing with partial differential
equations, but also has been applied to a large variety of problems arising in mechanics,
optimization and control problems, management science, operations research, and so
forth.One of the most important and interesting problems in the theory of variational
inequality is the development of an efficient and implementable algorithm for solving
variational inequality and its generalizations.These methods include projection method and its
variant forms, linear approximation, descent and Newtons methods, and method based on
auxiliary principle technique .
Let H be a real Hilbert space whose inner product and norm are denoted by
, and
,respectively. Let D be a nonempty convex subset of H.Let T, A:D H, and :D D H be
mappings, and let f:D R be a real-valued function. Recently Ansari and Yao [1] considered
and studied the mixed variational inequality problem (MVLIP), which is to find u* D such
that
+f
-f
690
(1)
AFASES - 2008 -
When A
, this problem reduces to the following problem considered by Dien [2], Noor [4]
and Siddiqi [5]: find
such that
+f
-f
(2)
+(u, u),
v D
(3)
In this paper, we propose a new iterative algorithm for computing the approximate
solution of MQVLIP (3).Our proposed algorithm is an extension and generalization of Ansari
and Yaos [1] iterative algorithm.
PRELIMINARIES
In this section we will recall some definitions and known results.
Definition 1.Let D be a nonempty subset of a Banach space V with the dual space V*.Let
T:D V* and :D D V be two mappings. Then
(i) T is called -monotone, if
0,
(4)
(ii) T is called -strongly monotone, if there exist a constant >0, such that
2
(5)
(6)
(iv) T is called Lipschitz continuous, if there exists a constant L>0, such that
,
(v) is called Lipschitz continuous, if there exists a constant >0, such that
691
(7)
AFASES - 2008 -
(8)
Lemma 2. Let (v,.):D V and K be a sequentially continuous from the weak topology to the
weak topology and from the weak topology to the strong topology, respectively, where v is
any fixed point in D. Then the function g:D R, defined as g(u)=<K(u),(v,u)> for each fixed
v D, is weakly sequentially continuous on D.
Lemma 3 (Fan-KKM). Let D be an arbitrary nonempty subset in a topological vector space E,
and let G:D 2E be a KKM mapping. If G(u) is closed for all u D and is compact for at least
one u D, then
Proposition 5.Let K be a differentiable -strongly convex functional with constant >0 on a
convex subset D of V, and let :D D V be a mapping such that (u,v)+(v,u)=0, for all
u,v D.Then K is -strongly monotone with constant >0.
Definition 6. The bifunction :V V R
is said to be skew-symmetric if
V.
ITERATIVEALGORITHM
In this section, following the approach of Ansari and Yao [1], we will apply the auxiliary
variational inequality technique to suggest a general algorithm for finding approximate
solutions of MQVLIP (3) with skew-symmetric term (,). Moreover, we will also study
convergence analysis of proposed algorithm.
Algorithm.LetK:DR be a given Frchet dierentiable functional, let
be a sequence
of positive parameters, and let u0 be any initial guess in D. For each given iterate un D,
D, such that
(9)
D.
Theorem 7.Let D be a nonempty close convex subset of a reflexive Banach space V with the
dual space V*.Let T, A:D V*,P:V* V* V*, and :D D V be mappings. Let w* V* and
692
AFASES - 2008 -
:V V R
be skew-symmetric and weakly continuous, such that for each v V,
int(dom(.,v)) D
and (.,v) is proper convex. Suppose that K:D R is -strongly convex
with constant >0, and K is sequentially continuous from the weak topology to the strong
topology. Suppose also that
(i) the mapping u P(Tu, Au) is -cocoercive with constant v>0;
(ii) is Lipschitz continuous with constant >0, such that
(a) (u, v)=(u, z)+(z, v) for each u, v, z D
(b) (.,.) is affine in the first variable,
(c) for each fixed v D, u (v, u) is sequentially continuous from the weak topology to
the weak topology;
(iii) for each fixed n 0 and z D,
{u D:
}.
Note that, since v G(v), G(v) is nonempty for each v D. Now we claim that G is a KKM
mapping. Indeed suppose that there exists a finite subset {v1, v2,,vk} of D and that i 0, for
all i=1,2,,k, with
=1 such that =
is not in G(vi), for all i=1,2,,k, that is
},
i=1,2,
,k.
Since (.,v) is proper convex for each v V, by virtue of assumptions (a) and (b) in (ii), we
have
0=
, )=
=
<0,
Which yields a contradiction. Therefore G is a KKM mapping.
Since K is sequentially continuous from the weak topology to the strong topology,
and (.,.) is weakly continuous on V V, so it follows from condition (ii)(c) and Lemma 2 that
693
AFASES - 2008 -
G(v) is a weakly closed convex subset of D for each v D. Moreover from condition (iii) we
know that G(v) is weakly compact for at least one point v D. Hence by using Lemma 3 we
have
, which clearly implies that there exists at least one solution to problem
(9).
Uniqueness of solutions of problem (9).Let u1 and u2 be two solutions of problem (9).Then,
for all v D.
Taking v=u2 in the first relation and v=u2 in the second relation, we obtain
Since K is -strongly monotone with constant >0 by Proposition 4, from relation above we
have
REFERENCES
[1] Q. H. Ansari and J, -C. Yao, Iterative schemes for solving mixed variational-like
inequalities,
Journal of Optimization Theory and Applications 108 (2001), no. 3, 527-541.
[2] N. H. Dien, Some remarks on variational-like and quasivariational-like inequalities,
Bulletin of the Australian Mathematical Society 46 (1992), no. 2, 335-342.
[3] S. Karamardian, The nonlinear complementarity problem with applications. II, Journal of
Optimization Theory and Applications 4 (1969), no. 3, 167-181.
[4] M. A. Noor, Variational-like inequalities, Optimization 30 (1994), no. 4, 323-330.
[5] A. H. Siddiqi, Q. H. Ansari and R. Ahmad, Some remarks on variational-like inequalities,
Mathematics and Its Applications in Industry and Business (A. H. Siddiqi and K. Ahmad,
eds.), Narosa, New Delhi, 2000, 101-108.
[6] P. Tseng, Further applications of a splitting algorithm to decomposition in variational
inequalities and convex programming, Mathematical Programming, Series B 48 (1990), NO.
2, 246-263.
*
Corina CHIRIAC, lector univ. dr. ,Universitatea George Baritiu, Brasov, Str. Lunii nr. 6, tel. 0238 711125, email: corinalchiriac@yahoo.com
694
AFASES - 2008 -
It
Pr (F
i =1
AP
= Fti ) = 1
E (P
i =1
AP
ti
(1)
695
AFASES - 2008 -
E AP (Pti Qti )
It
E (P
i =1
AP
ti
Pr AP (Ft = Fti )
(2)
Pr (F
i =1
cos t
= Fti ) = 1
Because the denominator is a sum for all index i, we will consider all quantities, not
just a particulary quantity qti , so the repartition became
E AP (Pti qti )
PrUC (Ft = Fti ) = I t
PrAP (Ft = Fti )
(3)
E AP (Pti Qti ) PrAP (Ft = Fti )
i =1
In the periode Z 0 we have a mean value of the costs E AP ( P0i Q0i ) . So we define the
cost level as
Pti Qti
(4)
Yti =
E AP ( P0i Q0i )
With the density function
f AP ( y ti ) = E AP (P0i Q0i ) f AP ,cos t [E AP (P0i Q0i ) y ti ]
(5)
The mean value of the cost level is
E AP (Pti Qti )
Pti Qti
E AP (Yti ) = E AP
(6)
=
E AP (P0i Q0i ) E AP (P0i Q0i )
For a repartition of the cost level on elements quantity units we consider the
random variable
qti Pti qti
Y(UC )ti =
(7)
EUC ( P0i Q0i )
The resulted density function is
f UC ( y ti ) = EUC (P0i Q0i ) f UC ,cos t [EUC (P0i Q0i ) y ti ]
(8)
For the density function we can also use the following formula
E AP (qti Pti qti )
f AP ,cos t [EUC (P0i Q0i ) y ti ]
(9)
f UC ( y ti ) = EUC (P0i Q0i )
E AP (Pti qti )
qti Pti qti = y (UC )ti EUC (P0i Q0i )
(7)
696
AFASES - 2008 -
E AP (Pti qti )
f AP ,cos t [EUC (P0i Q0i ) y (UC )ti ]
(10)
II. Common Probabilistic Repartition for Production Factor and Cost of Production
Level per elements act of production
In the model there was presented two probability prepatition wich we will now contain
in a common repartition.
Both the variable cost level and the caracteristic factor per acts of production
are dimensionless so the repartition can be presented on the same dial.
The random variable cost level Yti sis formed under the existing condition of a
production factor Fti in the act of production i. We use this propriety to write the random
variable cost level as a conditioned variable
Yti = Yt (Ft = Fti )
i=1,2,..., I t
(12)
accordingly the density function of cost level per elements act of production can be
written as
i=1,2,..., I t
(13)
f AP ( yti ) = f AP ( yt (Ft = Fti ))
For the factor Fti used in the i acts of production the probability repartition per act of
production is given as
Pr AP (Ft = Fti )
i=1,2,..., I t
(14)
Now we can form easily a common repartition of the cost level and factor per acts of
production.
f AP ( y ti ; Ft = Fti ) = f AP ( y t (Ft = Fti )) Pr AP (Ft = Fti )
(15)
where
+ I t
f (y
i =1
AP
ti
; Ft = Fti )dy t = 1
(16)
III. Common Probabilistic Repartition for Production Factor and Cost of Production
Level per elements quantity units
We consider the density function of the common repartition of cost level and factor
per acts of production per elements quantity units
fUC ( yti ; Ft = Fti ) = fUC ( yt (Ft = Fti )) PrUC (Ft = Fti )
From the relation (10) we have:
f UC ( y ti ) = f UC ( y t Ft = Fti ) = EUC (P0i Q0i )
then
fUC ( yti ; Ft = Fti ) = EUC(P0i Q0i )
EAP(Pti qti )
697
AFASES - 2008 -
It
f (y
i =1
UC
ti
; Ft = Fti ) =
E (P
E (P
i =1
It
i =1
AP
AP
ti
ti
f AP ( y t Ft = Fti )dy t =
f AP ( yt Ft = Fti )dyt = 1
It has been considered that because the denominator is a sum for all index i, we will
consider all quantities, not just a particulary quantity qti , so qti Pti qti = Pti Qti .
1. References
1. Lange, K.: Eine Theorie der Preisstatistik: Preis,Preisrelation,Preisindex, Editura
Daniela Damian, Dept. of Economic Science, George Bariiu University of Braov, e-mail:dana.damian@yahoo.com
698
AFASES - 2008 -
699
AFASES - 2008 -
700
AFASES - 2008 -
701
AFASES - 2008 -
Stratul de date
Stratul de mijloc
Stratul de comunicare
Stratul de interfa
(GUI)
Utilizator
H ( X ) = p( x) log p( x) .
(1)
702
AFASES - 2008 -
Demonstraie:
Funciile de variabile aleatoare independente sunt tot variabile aleatoare independente.
Rezult c deoarece X i sunt i.i.d., atunci i log p ( X i ) sunt independente. Aadar, din legea
slab a numerelor mari, rezult c
1
1
ate
log p( X 1 , X 2 ,..., X n ) = log p ( X i ) inprobabilit
M log p( X i ) = H ( X )
n
n i
iar teorema este demonstrat.
(3)
Dac FX (x) este continu, atunci variabila aleatoare este continu. Fie f ( x) = FX ( x) dac
f ( x) = 1 ,
(4)
n literatur exist numeroase studii despre modelarea unei aplicaii web ca un lan
Markov finit i ergodic [1]. n acest model, fiecare sesiune de navigare a utilizatorilor este
interpretat ca o secven din lanul Markov ergodic. Se presupune c se cunoate o distribuie
iniial empiric. Se consider c distribuia empiric a probabilitilor lanului Markov este
staionar pe o perioad de timp n concordan cu probabilitile de tranziie actuale.
Entropia lanului Markov este esenial n aceast abordare. Distribuia empiric fiind
staionar n timp, rezult c entropia unei secvene tipice este apropiat de entropia lanului
Markov. Aceast remarc este consecina proprietii asimptotice de echipartiie (PAE).
703
AFASES - 2008 -
iniiale.
Un lan Markov ergodic are o distribuie staionar notat cu care satisface
egalitatea:
P =
(6)
n
Distribuia staionar a lanului Markov poate fi calculat din fiierele LOG care
sunt nregistrate pe server [4]. Aceast distribuie indic paginile preferate de utilizatori n
timpul navigrii n aplicaia web. Aceasta este o form de page ranking, care poate fi folosit
pentru a mbunti relevana rspunsurilor ctre utilizatori n cazul motoarelor de cautare.
n consecin, estimnd entropia unei secvene tipice din lanul Markov i folosind
proprietatea PAE, se poate estima entropia ntregului lan.
Bibliografie
[1.] David N., On Some Probability Distributions in Web Surfing, New Challenges in the
Field of Military Sciences 2007, Zrnyi Mikls National Defence University, Budapest,
Hungary, 2007.
[2.] Ratiu I.G., David N.,Working with Databases in Java, Buletinul Stiintific, Academia
Fortelor Terestre Sibiu, anul XII, nr 2(24)/2007, pag. 232-237, ISSN 1224-5178.
[3.] Janosek M., Modelling and simulation of fighter aviation at averflow air targets, The 8th
International Carpathian Control Conference ICCC 2007, p. 248-251, ISBN: 978-80-8073805-1.
[4.] Levene M., Loizou G., Computing the Entropy of User Navigation in the Web,
International Journal of Information Technology and Decision Making 2(3): 459-476, 2003.
[5.] Mezey G., Mapping to storage of a data model nework structure. 4th European Meeting
of the Psychometric Society and the Classification Societies, Cambridge UK, 1985.
[6.] Damian D., Major Types of Multiple Regression, 19th Scientific Session on Mathematics and its
Applications, Brasov, 2005, ISBN 973-635-605-1.
Nicoleta David, lect. univ. drd., Univ. George Bariiu din Braov, email: nicoletadavid@yahoo.com
704
AFASES - 2008 -
I. Introduction
Today, the network technologies are available to a large number of students many of
them having at least a medium level in computers use skills. This is providing since the
beginning of the educational process a valuable resource what can be used. An effective
handle of this resource capitalized together with the potential of the networked technologies
and the instructors skills can be the problem-solving approach that defines the solution of an
educational process developed in an e-learning environment. In the past few years the use of
the World Wide Web (WWW) features in the educational process has changed the way how
the instructors approach the process (i.e.: teaching practices, curriculum evolving etc.) and the
way how students understand to approach the learning in a more effective manner. From this
point of view several features of the WWW have been exploited considerably as follows:
9 Communication features These mechanisms, such as: e-mail, Web conferencing, newsgroups, chat rooms etc. have the capability to facilitate the
exchange of information and ideas between instructors and students and to
provide instructional feedback in a short time and a friendly manner. The
communication mechanisms above mentioned shown that it is possible to
increase the amount of time available to each student compared to traditional
class room communication mechanisms. By increasing this amount of time than
a major improvement of the student cognitive skills should appear shortly.
705
AFASES - 2008 -
706
The research studies in the psycho-pedagogical domain shown that human beings are
able to retain 10% from what we read, 20% from what we hear, 30% from what we see, 70%
from what we discuss with others, 80% from what we experience and 95% from what we
teach the others [2].
Specialists in e-learning domain consider that this educational process is carried out on
the basis of the e-knowledge management [3], through several phases such as, e-knowledge
supply, e-knowledge transfer, e-knowledge acquisition and e-knowledge control. In figure 1 it
is presented a basic concept of the e-knowledge management phases.
2.1
process. Also, the students participate in the improvement process providing their feed-back
in regards of the courses effectiveness. The system proposed by the author tries to
interconnect several sets of learning elements such as, the learning objectives, the curriculum,
the assessment procedures etc.
Further, the system uses as implementation solution the implementation by subsystems: the learning subsystem and the assessment subsystem. These subsystems are
perceived as independent ones, autonomous entities that interact with each other through
interfaces. In the figure 3 is presented the proposed integrated system.
Fig. 3 : The flow chart of the experimental integrated learning assessment system
The proposed model has been designed as an automated system for learning and
assessment covering all phases of knowledge management process.
System inputs were defined to meet the individual and group needs and demands on
basis of the knowledge and skills necessary to be transferred / acquired and developed.
The logistics aspects concerning the technical resources availability (hardware and
software) is also one important system input as well as the timing schedules to accomplish the
teaching process and assessment process.
AFASES - 2008 -
System inputs were defined to meet the individual and group needs and demands on
basis of the knowledge and skills necessary to be transferred / acquired and developed.
The logistics aspects concerning the technical resources availability (hardware and
software) is also one important system input as well as the timing schedules to accomplish the
teaching process and assessment process.
Learning subsystem is deemed to cover the knowledge supply / transfer phase. The
curriculum is transferred in modules containing the learning materials. The students are
supplied by the instructors through the web with the necessary information. Each learning
module has attached an application module deemed to help students in acquiring the
knowledge by practicing.
The Evaluation subsystem is dedicated to the knowledge control phase. Being
designed as a cyclic system, the students will go through the assessment process once they are
confidents in their capability to pass that learning module. The system is developing
automatically the tests once is receiving the request from the student side. Each student
receive a different test like content the system being design to avoid the repetition of the same
questions in different tests. Of course, the main requirement in this case is to have a large data
base with questions from where the program can generate the tests. However, the assessment
process is not referring only to the students but also to their instructors and to the system
itself. The idea is not to restrict the use of this integrated system only to the classic
educational purposes but also to integrate this one into the quality management systems by
providing the necessary support in measuring and the performances, at individual and overall
levels. Therefore, a part of the normal tests dedicated to each module the students have the
opportunity to evaluate themselves the instructors and the system capabilities by using
specific appreciation questionnaires similar with the customer satisfaction questionnaires. At
the end of the day, the students can be considered the internal clients of the educational
process.
System outputs are deemed to be used for continual improvement of the process. These
outputs could be used for further discussions or to draw some conclusions about success and
failure at individual and group level. The system is capable to build various histograms in
order to have a clear picture of the situation at any moment than the conclusions can be easily
withdrawn and necessary measures applied.
The system is working in two modes, the administrator mode and the user mode.
Obviously, the administrator mode is dedicated to instructors and the user mode to the
students. The administrator of the system could be an individual person or could be a team.
However, the team should have a leader with full rights in operating and modifying the
system in same time the team members being dedicated to certain sequential activities to
support the learning and assessment subsystems.
An administrator will be responsible for defining the system implementation strategy,
this implementation process consisting from specific activities what can be carried out once or
in repeatable way depending on the requirements at individual level and group level.
Generally, the activities could be as follows:
9 Analysis of the actual situation in the light of the educational process
requirements and goals
9 Learning and assessment processes scenario definition
9 E-learning platform selection
9 Authoring tool selection
9 Network capabilities analysis
9 Web-based learning material and assessment modules design & development
9 Provide learning material and assessment modules to the users
9 Tutoring organization by defining the requirements for the logistic needs
709
AFASES - 2008 -
The author believes that at this stage and taking into consideration the actual situation
when most of the higher education institutions are in the phase to revamp their strategies
regarding the educational process a good point for discussion can be the cooperation between
these institutions. Seeking for such opportunities, exchanging experience, sharing knowledge
etc. can be only beneficial for all parties involved and here the author would like to mention
the end user of the educational process, which is the modern society represented by the social,
political and economical environment. Without a full support coming from these
environments the higher education institutions will not be able to move forward in developing
new opportunities for knowledge resources development.
III. Conclusions
Today requirements in higher education teaching involve re-ordering the leap
magnitude in the instructors ability to create, acquire, assimilate and share the knowledge to
their students. The available information and communication technologies re-shape on daily
basis the educational environment. In the next decade the knowledge sharing methods and
techniques will be re-invented significantly. Clearly, at present all these transformation are
underway and in the author opinion there is no other option than to follow this mega-trend
and to try to contribute as much as possible to the future changes in the benefit of all parties
involved in the educational process. Across the globe, we see examples of the e-knowledge
revolution [5].
The author believes that integrating learning and assessment by sharing the
responsibility of the educational process goals achievement is the proper solution for these
incoming transformations. Without considering the proposed experiment as the unique
solution the author consider that a web-based integrated system for learning and assessment
can support both, undergraduate and postgraduate students as well as the instructors needs and
demands with a very high rate of effectiveness.
Development of the knowledge base at individual level, for each student,
independently but in same time without moving outside from the modern society
requirements is the approaching challenge for the higher education. Today the knowledge is
one of the key resources for the human society and the higher education institutions are
playing a critical role in developing this valuable asset.
Bibliografie
[1] Anil K. Aggarwal, Web-Based Education: Learning from Experience, IRM Press (an
imprint of Idea Group Inc.), U.S.A., 2003.
[2] Maria-Ana Lazr, nvarea prin e-learning: azi o provocare, mine o realitate, Sesiunea
Judeean de Comunicri tiinifice a Cadrelor Didactice, Deva, 2005.
[3] M. Klein, D. Sommer and W. Stucky, WeBCEIS A Scenario for Integrating Web-Based
Education into Classical Education, University of Karlsruhe, Germany, 2003.
[4] Iuliana Dobre, Aspects of Collaboration Tools in E-Learning, Buletinul Universitii
Petrol-Gaze Ploieti, Romnia, 2006.
[5] D. Norris, J. Mason and P. Lefrere, Transforming e-Knowledge A Revolution in the
Sharing of Knowledge, Society for College and University Planning, U.S.A., 2003.
*
710
AFASES - 2008 -
711
AFASES - 2008 -
712
AFASES - 2008 -
713
AFASES - 2008 -
714
AFASES - 2008 -
The only thing left to configure is the interval between running the cron.php script
and the path where the script is located. This feature is very reliable and indifferent to the user
logged in the system and its requirements to the processor are minimal.
Users accounts
Since Moodle is currently only in the testing period when we are getting acquainted
with it, the option of manual creation of users accounts was selected. Presently we are
considering only minimal operation of this application and only trial courses for a limited
number of students and teachers of University of defence. The possibilities of this application
are, however, much wider. For verification and logging of users almost any database or
commonly used and also more elaborate systems like LDAP servers can be used. User
accounts can also be created using emails and POP3 or IMAP protocols, including their safer
variants based on SSL/TSL. All these methods are more practical for a bigger group of users.
For safetys reasons we chose an option where even teachers cant create users
accounts for their students, only request their set up. The creation is then realised by the
Moodle administrator. This part of Moodle configuration lasts about 2-3 hrs.
Real life with Moodle
So far Moodle has been successfully running at the University scale for a limited
group of students and teachers. Some basic types of courses are being offered. Also the
teachers themselves are only getting acquainted with the application and with what this
software pack offers and what are the possibilities of a real life with Moodle. Our future plan
is to enlarge moodle as standard learning support and using of Moodle as support for NEC.
Summary
Moodle is very dynamic and powerful learning environment based on internet
standards. Installation of Moodle can be adjusted to meet concrete education requirements.
Community of moodle is permanently growing and moodle is used by large and recognized
institutions like universities and high schools all over the world. As moodle is open source
software it is often sought by small communities and organizations. There are also many
discussion forums on the internet solving possible problems with running this system. In this
case, support of moodle is on a very quality and professional level. This is also one of the
reasons why moodle is the most popular learning system. The native web page of Moodle
project is http://www.moodle.org or you can test our installation at our department pages
which is placed at http://k209moodle.unob.cz .
Lt. Ing. Robert Drmola, University of defence Faculty of military technologies Communications and
information systems department
715
AFASES - 2008 -
ABSTRACT
The shortest spanning tree (SST) of a graph has applications in cases where roads
(gas pipelines, electric power lines, etc.) are to be used to connect n points together in such a
way so as to minimize the total length of road that has to be constructed. If the n points to be
connected are on a Euclidian plane, they can be represented as vertices of a complete graph
G with arc costs being the straight line distances between the corresponding end points. The
SST of G is then (provided no road junctions outside the given n points are allowed) the
required minimum-cost road network. If junctions outside the given n points are allowed ,
then an even shorter road network may be possible, and finding it is a problem known as
Steiners problem.
Keywords: spanning tree, graph, Kruskal, Prim.
I. INTRODUCTION
One of the most important concepts of a graph theory, and one which appears often in
areas superficially unrelated to graphs, is that of a tree [2].
Definitions. The following definitions of a non-directed tree can be quite easily shown to be
all equivalent to each other.
A non-directed tree is:
(1) A connected graph of n vertices and (n-1) links,
or (2) A connected graph without circuit,
or (3) A graph in which every pair of vertices is connected with one and only one elementary
path.
If G = (X,A) is a non-directed graph of n vertices, then a spanning tree of G is
defined as a partial graph of G which forms a tree according to the above definition. Thus, if
716
AFASES - 2008 -
G is the graph of fig. 1.1(a), then the graph of fig. 1.1(b) is a spanning tree of G, and so is the
graph of fig. 1.1(c). From the above definition it can be seen that a spanning tree of G could
also be considered as a minimal connected partial graph of G (where minimal is used in the
sense of no subset of the links of the tree exist which forms a connected partial graph of G).
x2
x2
x2
x3
x1
x4
x3
x5
x1
x4
x6
x3
x5
x1
x4
x6
x7
x6
x7
Fig. 1.1.(a)
Graph G
x5
x7
Fig. 1.1.(b)
A spanning tree of G
Fig. 1.1.(c)
Another spanning of G
The concept of the tree as a mathematical entity was first proposed by Kirchhoff [7] in
connection with the definition of fundamentals circuits used in the analysis of electrical
networks. About 10 years later Cayley [1] rediscovered trees independently and produced
most of the early works in this area.
A directed tree (also called an arborescence), is defined in a similar way as follows.
Definition. A directed tree is a directed graph without a circuit, for which the in-degree of
every vertex, except one (say vertex x1), is unity: the in-degree of x1 (called the root of the
tree) being zero.
Figure 1.2 shows a graph which is a directed tree with vertex x1 as the root. The above
definition immediately implies that a directed tree of n vertices contains (n-1) arcs and is
connected. It is also quite obvious that not every directed graph has a spanning directed tree,
as can, for example, be seen from the graph of fig. 1.3.
Root: x1
x4
x3
x5
x9
x6
x10
x14
x2
x1
x7
x21
x15
x3
x8
x12
x11
x16
x18
x19
x17
x2
x13
x20
x5
x6
x4
It should be noted that a non-directed tree can be converted into a directed one by
arbitrarily picking any vertex for the root and choosing directions for the links so that there is
a path between the root and every other vertex. Conversely, if
717
5
3
x4
5 5
3
x5
x1
3
x3
5
x2
x4
x1
x5
x2
x4
x3
x5
x2
x3
Fig. 2.1 (a) Graph G Fig. 2.1 (b) Shortest path tree from x1 Fig. 2.1 (c) Shortest spanning tree
The calculation of the shortest spanning tree (SST) of a graph is one of the few
problems in graph theory which can be considered completely solved. Thus, let Ti and Tj
be any two sub-trees produced by the addition of links during the constructions of the SST.
If Ti is also used to represent the set of vertices of the sub - tree, then
ij can be
defined to be the shortest distance from a vertex in Ti to a vertex in Tj .
ij
ij
(2.1)
It can then be quite easily shown that a repeated application of the following operation will
produce the SST of a graph.
Operation I: For some sub-tree Ts, find that sub-tree Tj* for which sj* = min [ sj], and
____
Tj
let (xs,xj*) be the link whose cost produced
in
equation
(2.1).
sj*
718
719
A function is symmetric in variables x1, x2 , , xm if the exchange of any two variables in the function leaves
its value unchanged. Both the monotonicity and symmetry restrictions are forced by the condition that C < C
should imply a decrease in the cost of the tree.
720
the spanning tree of G which minimizes (Ci1 + Ci2 + + Cin-1 ) or (Ci1 x Ci2 x x
Cin-1) where Ci1 , Ci2 , , Cin-1 are the costs of any n-1 links forming a spanning tree of
G is the same as the SST of G. In addition, since for the first of the above mentioned
functions the power of 3 could be replaced by any other power p>0, and since as p :
n-1
[ Ci l ] { max [ Ci1] }p ,
l=1
l=1,...,n-1
the spanning tree which minimizes the cost of the most costly link contained, is also the same
as the SST of G.
2.4 EXAMPLE [ 11]
In the graph G shown in Fig. 2.4.1 each vertex represents a person and a link ( xi,xj)
implies that xi could communicate with xj and vice versa. It is required that a confidential
message be circulated amongst the 12 people in such a way as to minimize the probability of
it becoming known outsider. For each transmission of the message from xi to xj there is a
probability ij that the massage may be intercepted by an outside person and these
probabilities are given as percentages in fig. 2.4.1.
15
x2
x3
6
8
x1
5
x5
18
11
6
x8
11
x4
9
14
7
10
x7
17 12
15
x6
18
13
5
7
x11
x12
x9
19
2
x10
721
x3
(x1, 15)
x8
(x7, 9)
x4
(x7, 7)
T1
x5
x7
x6
(x5, 15)
x12
(0, )
x11
(x7, 12)
x9
(x7, 4*)
x10
(x7, 14)
Fig. 2.4.2 (a). Partially formed tree T1 with labels on vertices not in T1
_ _ _ _ _ _ Next link to be added
722
x2
a2
x1
a1
x3
x4
a3
x5
x6
a6
a7
x8
a4
x7
a11
x11
a10
a5
x12
a9
a8
x9
x10
REFERENCES
[1] CAYLEY, A., Collected papers, Quart Jl. Of Mathematics,13, Cambridge, 1897.
[2] CHRISTOFIDES, N., Graph Theory An Algorithmic Approach, Imperial College,
London, 1975.
[3] DIJKSTRA, E. W., A note on two problems in connection with graphs, Numerische
Mathematik, 1, 1959.
[4] GOWER, J. C. and ROSS, G. J. S., Minimum spanning trees and single linkage cluster
analysis, Applied Statistics, 18, 1969.
[5] KERSHENBAUM, A., and Van SLYKE, R., Computing minimum spanning trees
efficiently, Proc. of the Ann. Conf. of ACM, Boston, 1972.
[6] KEVIN, V. and WHITNEY, M., Algorithm 422 Minimal spanning tree, Comm. Of.
ACM, 15, 1972.
[7] KIRCHHOFF, G. , Annalen der Physik and Chemie, 1847.
[8] KRUSKAL, J.B.Jr., On the shortest spanning sub-tree of a graph and the traveling
salesman problem, Proc. American Mathematical Soc., 7, 1956.
[9] LOBERMAN, H. and WEINBERGER, A. , Formal procedures for connecting terminals
with a minimum total wire length, Jl. Of ACM, 4, 1957.
[10] OBRUCA, A. Algorithm 1 MINTREE, Computer Bulletin, 1964.
[11] PRIM, R.C., Shortest connection networks and some generalizations, Bell Syst. Tech.
Jl.,36, 1957.
Gabriela Simona Dumitrescu, ing. profesor, Liceul teoretic Traian Constanta, Traian n.r.55, tel. 0241672295, e-mail: profdumi@yahoo.com.
723
AFASES - 2008 -
ABSTRACT
In this article will be dealing with two questions [5]:
Problem 1. Given a general graph G, find a Hamiltonian circuit of G, (or alternatively
all circuits), if one or more such circuits exist.
Problem 2. Given a complete graph G whose arcs have arbitrary costs C = [cij]
associated with them, find that Hamiltonian circuit, (or path), with the least total cost. The
problem of finding the least cost Hamiltonian circuit is known in the literature as the
traveling salesman problem [1, 2, 7]. It should be noted that if G is not complete, it can be
considered as a complete graph with infinities inserted as the costs of the non-existent arcs.
Solutions to the travelling salesman problem and its variants have a large number of
practical applications in many diverse fields. For example, consider the problem where a
vehicle leaves a central depot to deliver goods to a given number of customers and return
back to the depot. The cost of the trip is proportional to the total distance traveled by the
vehicle so that, given the distance matrix between customers, the least cost trip is the solution
to the corresponding travelling salesman problem. Similar types of problem occur in
collecting mail from mail boxes, scheduling school buses through a number of stops, etc. The
problem generalizes quite readily to one where more than one vehicle performs the deliveries,
(or collections), although this problem can also be reformulated into a larger size single
travelling salesman problem [6]. Other applications include the scheduling of operations on
machines [4], the design of electricity supply networks [3], the operation of sequential
machines [8] etc.
Keywords: Traveling, salesman, graph, Hamiltonian, circuit, cost.
INTRODUCTION
In a number of industries, the following basic scheduling problem arises: A number (say
n) of items is to be manufactured using a single processing facility or reaction vessel. The
facility (vessel), may or may not have to be reset (cleaned), after item pi has been
724
AFASES - 2008 -
manufactured, (but before production of pi is started); depending on the item combination (pi,
pj). The cost of resetting the facility is constant regardless of the item pi that has been
produced or the item pj that is to follow; and there is no cost incurred if no resetting of the
facility is required. 1 Suppose that these n items are to be manufactured in a continuous cyclic
manner, so that after the production of the last of the n items the manufacture of the first item
in the fixed cycle is started again.
The problem then arises as to whether a cyclic production sequence for pi (i = 1, 2, ,
n) can be found which requires no resetting of the facility. The answer to this question
depends on whether a graph G whose vertices represent the items and where the existence of
an arc (xi, xj) implies that pj can follow the production of item pi on the facility without
resetting, possesses a Hamiltonian circuit or not. (A Hamiltonian circuit is a circuit passing
once and only once through each vertex of the graph).
For any given graph G, random finite costs can be allocated to the arcs of G to form a
traveling salesman problem. If the solution to this problem the shortest Hamiltonian circuit
has a finite value, then this solution is indeed a Hamiltonian circuit of G.
If the solution has infinite value, then G does not possess a Hamiltonian circuits. An
alternative interpretation can be placed on problem 1 as follows. Consider once more a
complete graph G1 with a general arc cost matrix [cij], and consider the problem of finding
that Hamiltonian circuit of G1 whose longest arc is a minimum. This problem could be called
the mini-max traveling salesman problem because of the mini-max nature of its objective, as
compared with the classical traveling salesman problem which could in he present terms also
be called the mini-sum problem. I will show that problem 1 above is in fact equivalent to the
mini-max traveling salesman problem.
In the complete graph G1 mentioned above, we can certainly find a Hamiltonian
circuit. Let this circuit be 1 and the cost of the longest arc in 1 be 1. Remove from G1 any
arc whose cost is greater than or equal to 1 to obtain the graph G2. Find a Hamiltonian circuit
say 2 in G2, and let the cost of the longest arc in 2 be 2. Again remove from G2 any arc
with cost greater or equal to 2 to form G3 and continue in the same way until a graph Gm+1 is
found which contains no Hamiltonian circuit. The Hamiltonian circuit m in Gm (with cost m,
say) is then, by definition, the solution to the mini-max traveling salesman problem, since the
lack of a Hamiltonian circuit in Gm+1 implies that no Hamiltonian circuit exist in G1 which
does not use at least one arc with cost greater than or equal to m. Thus, an algorithm for
finding a Hamiltonian circuit in a graph also solves the mini-max travelling salesman
problem. Conversely, if we possess an algorithm for solving this latter problem, a
Hamiltonian circuit in an arbitrary graph G can be found by constructing a complete graph G1
on the same set of vertices as G, placing unit costs on the arcs that correspond to arcs in G and
infinite costs on all other arcs. If the solution to the mini-max traveling salesman problem for
G1 has a finite cost (in fact unity cost), then a corresponding Hamiltonian circuit in G has
been found; if the solution has infinite cost then no Hamiltonian circuit in G exists.
I. THE TRAVELLING SALESMAN PROBLEM
1.1 A lower bound from the assignment problem
The linear assignment problem for a graph with a general cost matrix C = [cij] can be
stated as follows.
Let ij be an n x m matrix of 0 1 valued variables so that ij =1 if vertex xi is
assigned to vertex xj and ij = 0 if xi is not assigned to xj. In the traveling salesman problem
1
The above problem can arise in two ways; either the resetting costs are in reality independent of the items, or, alternatively,
detailed cost data is not available and an average constant value is taken as an approximation.
725
AFASES - 2008 -
we could use a similar scheme where ij = 1 would mean that the salesman travels from xi to
xj directly and ij = 0 would mean that he does not. For this last problem we can start by
setting all cii (i=1, , n) to thus eliminating non-sensical solution with ij = 1.
The assignment problem now becomes:
Find 0 1 variables ij so as to minimize:
n
z = cij ij
(1.1)
j =1 i =1
subject to
ij
= ij = 1
(1.2)
(1.3)
Equations (1.2) simply insure that the solution is cyclic one arc enters and one leaves
every vertex.
Equations (1.1) (1.3) together with the additional constraints that the solution must
form a single (Hamiltonian) circuit and not just a number of disjoint circuits, can also be used
to represent a formulation of the traveling salesman problem. Since the addition of any
constraint to the assignment problem can only increase or leave unchanged the minimum
value of z as calculated from equations (1.1) (1.3), this value of z is a valid lower bound to
the cost of the solution to the traveling salesman problem for a graph with a cost matrix [cij].
1.2. A lower bound from the shortest spanning tree
726
(1.4)
AFASES - 2008 -
Let us define G(TSP) to be the partial graph of a non-directed graph G, formed by the
vertices and those links of G that are used by the optimal travelling salesman circuit. Similarly
let us define graphs G(AP) and G(SST) formed by the same vertices but having links which
appear in the optimal solutions of the assignment and shortest spanning tree problems
respectively.
1
5
5
>
<
4
3
(a) G (TSP)
(b) G (AP)
Fig. 1.1 Graphs of traveling salesman, assignment, and shortest spanning tree problem
The graph G(TSP), which in the case of a 5-vertex graph is shown in Fig.1.1 (a), has the
following properties:
(i)
The graph is connected; every vertex can reach every other vertex via a path
using the links;
and
(ii)
The degree of every vertex is 2; there are two links incident at each vertex.
The graph G (AP) does not necessarily possess property (i) above, as can it seen from
the example of Fig. 1.1 (b) but does, by definition, have property (ii). If, however, it so
happens that the solution to the assignment problem does have property (i) as well, then this is
also the solution to the traveling salesman problem.
The graph G (SST) has property (i), by definition, but does not have property (ii). If it
so happens that the shortest spanning tree does have property (ii) except for two end
vertices (x1 and x2 say) which by necessity must have a degree of unity then the shortest
spanning tree is also the shortest path passing through all the n vertices. If, moreover, link
(x1, x2) is in the optimal travelling salesman circuit, then the links of the shortest spanning tree
plus link (x1, x2) will be the solution to the travelling salesman problem. If it is not certain that
(x1, x2) is in the optimal travelling salesman circuit then some small modification is needed.
Thus, the solution of the assignment and shortest spanning tree problems are dual in
the sense that they possess properties which are complementary with respect to the properties
of the travelling salesman problem. Two possible avenues of investigation now reveal
themselves which can lead to solutions for this latter problem.
(a) Use the solution of the assignment problem which possesses property (ii) and force
this solution to conform with property (i);
or (b) Use the solution of the shortest spanning tree which possesses property (i) and force it
to conform with property (ii).
The assignment problem is defined for graphs with any arbitrary cost structure, the
shortest spanning tree is only defined for non-directed graphs graphs with symmetrical, (cij
= cji) cost matrices. What is being said about the relationship between the travelling salesman
problem and the shortest spanning tree for non-directed graphs, has an exact equivalent
applying to the relationship between the travelling salesman and the shortest arborescence for
the case of directed graphs.
727
AFASES - 2008 -
The assignment problem defined by equation (1.1), (1.2) and (1.3) may have solutions
composed of a number of disjoint circuits, and one method of solving the traveling salesman
problem would be to impose restrictions until the solution becomes a single (Hamiltonian)
circuit. In the present section I will investigate procedures of imposing these restrictions
within the framework of a decision-tree search algorithm, for the decision-tree search based
on the shortest spanning tree.
2.1 A decision tree search algorithm
Let the solution of the assignment problem with a cost matrix [cij] (and cji = ), be
composed of a number of disjoint circuits. For example if the solution of an 8-vertex problem
is given by 1,2 = 2,6 = 3,5 = 4,7 = 5,4 = 6,1 = 7,8 = 8,3 = 1 and all other ij = 0, then the
solution corresponds to the two circuits shown in Fig. 2.1(a). What is now needed is to
eliminate this solution together with as many other solutions as possible without eliminating
the solution to the travelling salesman problem under the same cost matrix. Since the
travelling salesman solution is a Hamiltonian circuit we will attempt to eliminate any
solutions which correspond to more than one circuit.
1
3
5
8
2
7
(a) Solution to problem P0
3
5
6
8
4
7
b) Solution to problem P1
Fig. 2.1 Solution to assignment problems
728
AFASES - 2008 -
xk, x1) is a solution to at least one of the problems P1, , Pk and hence the optimal travelling
salesman solution is the solution to one or more of these sub-problems.
For the example of Fig. 2.1(a), choosing to eliminate the circuit of cardinality 3 results
in the decision-tree of Fig. 2.2 in which the problems P1, P2, and P3 are represented by nodes
of the tree derived from the initial problem P0. Let the problem P1, P2, and P3 be solved as
assignment problems and let the corresponding costs be C1, C2 and C3. Since C1 is a lower
bound on the traveling salesman solution to problem P1 and similarly for P2 and P3 the
number L = min [C1, C2, C3] is a lower bound to the value of the solution of the initial
travelling salesman problem.
P0
c12 =
c18 =
P4
c26 =
P1
P5
c61 =
P2
c26 =
c82 =
P6
P3
c61 =
P7
729
AFASES - 2008 -
C61=
C12= -M
C26= -M
C26=
P2 C12= -M P3
REFERENCES
[1] BELLMORE, M. and NEMHAUSER, G. L., The travelling salesman problem a survey,
Ops. Res., 16, 1968.
[2] BELLMORE, M. and MALONE, J. C., Pathology o traveling salesman sub-tourelimination algorithms, Ops. Res. ,19, 1971.
[3] BURSTALL, R. M., Tree-searching methods with an application to a network design
problem, In: Machine Intelligence, Vol. I, Collins and Michie, London, 1967.
[4] CHARLTON, J. M. and DEATH, C.C, A method of solution for general machinescheduling problems, Ops. Res, 18, 1970.
730
AFASES - 2008 -
[5] CHRISTOFIDES, N., Graph theory An algorithmic approach, Computer Science and
Applied Mathematics - A series of monographs and textbooks, University Maryland,
1975.
[6] CHRISTOFIDES, N. and EILON, S., An algorithm for the vechicle dispatching problem,
Opl. Res., 20, 1969.
[7] EILON, S., WATSON-GANDY, C.D.T. and CHRISTOFIDES, N., Distribution
Management: Mathematical modelling and practical analysis, Griffin, London, 1971.
[8] HARING, D.R., Sequential-circuit synthesis, MIT Press, Research Monograph 31,
Cambridge, Massachusetts, 1966.
Gabrila Simona Dumitrescu, ing. profesor , Liceul teoretic Traian Constanta, str. Traian nr. 55, tel.. 0241
672295, e-mail: profdumi@yahoo.com.
731
AFASES - 2008 -
Where: Qtp is the theoretical volumetric debit of the pump; Qlp is the volumetric debit of the
pumps flow; Qtm is the theoretical volumetric debit of the engine; Qlm is the volumetric debit of
the engines flow; Qs is the volumetric debit evacuated trough the valve orifice; Qcs is the
command volumetric debit of the valve; e is the equivalent elasticity module of the liquid from
the volume Vt .
The evolution of the pressure in the command chamber of the valves poppet results from the
continuity equation attached to the chamber:
732
AFASES - 2008 -
Vc
p& cs
(2)
dx s
is the speed of the poppet; Vc is the
dt
average volume of liquid located in the command chamber of the valve.
The distance between the poppet and the drawer x s (the valve opening) results from the
movement equation attached to the poppet:
ms &x&s = Fcs + Fhs + Fht + Fe
(3)
Where: ms is the equivalent mass of the poppet and the spring; &x&s is the valve acceleration; Fcs is
the pressure force which assures the command of the valve; Fhs is the hydrodynamic force
corresponding to the fluid speed variation with respect to the space inside the valve (stationary
component); Fht is the pressure force corresponding to the variation of the fluid speed with respect
to the time inside the valve (transitory component); Fe is the elastic force generated inside the
helicoidally spring
Its admitted that both the volumetric machines have variable capacities (the general
case). Their theoretical debits are of the following form: Qtp = n pV p ; Qtm = nmVm
Where: Acs is the surface of the command piston and
Where: n p is the pump revolution (constant); V p is the pump capacity (variable); nm is the engine
revolution (variable); Vm is the engine capacity (variable).
It is admitted that the dominant flows of both machines are produces in a laminar regime
therefore they are direct proportional with the pressure differences that they create:
Qlp = K lp ( ps pt )
(4)
Qlm = K lm ( ps pt )
(5)
Is admitted that the exit joints of the valve and the drain joints of the machines are connected
straight to the tank, which is at the atmospheric pressure pt 0 .
The debit evacuated trough the valve derives from the Bernoulli equation and the equation of
continuity for the current tube:
Qs = K s xs p s
(6)
Where K s is the valve constant. For a valve with a conic poppet, at small openings
K s = d s cds
sin
(7)
Where: d s is the diameter of the valve; cds is the debit coefficient of the valve orifice; the halfangle of the poppet cone; the liquid density
The debit coefficient depends on the angle . The command debit for the valve is generated by
the pressure difference ps pcs . It is considered a damper of a ring type slot, having the medium
l
diameter d s , the length Ls and the width l s . The usual value of the report s = 1
, and the
ds
100
command debit is relatively small. Therefore, it can be admitter that the movement trough the slot
is a plane movement of Hagen Poiseuille type. If the damper is not centered (the usual case)
733
AFASES - 2008 -
Qcs = 2,5
d s l s3
( p s pcs ) .
12Ls
(8)
Where is the dynamic viscosity of the liquid. This relation can be written under the following
form:
Qcs = K cs ( p s pcs )
(9)
Where K cs = 2,5
d s l s3
12Ls
is the constant of the damping butterfly valve. The equivalent mass of the poppet and the resort
is:
ms = mv + 0.33me
(10)
Where mv is the poppet mass and me the spring mass.
The poppet is accelerated by the force:
d 2
Fcs = Acs pcs = s pcs
(11)
4
The hydrodynamic force is:
Fhs = K hs x s p s
(12)
Where K hs is the constant of the hydrodynamic force in stationary regime.
The spring of the valve generates the force:
Fe = K es ( x s + xe )
(13)
Where K es is the static rigidity and xe is the pre-compression.
For the case of an uncompressed poppet, the movement equation for the poppet becomes:
Fcs Fhs + Fe = 0
(14)
Or Acs p s K hs x s p s = K es ( x s + xe ) .
From this equation one can calculate the valve opening based on the pressure from the valves
upstream:
A p K e xe
xs = cs s
.
(15)
K es + K hs p s
The valve begins to open at the pressure:
K x
(16)
ps 0 = e e
Acs
A ( p ps 0 )
So, xs = cs s
. The valve working process in stationary regime is described by the
K es + K hs p s
equation:
ps ps 0
.
(17)
Qs = K s Acs p s
K es + K hs p s
Establishing the transfer function for the valve
The complete mathematic model of the valve is composed of the following equations [1], [4]:
ms &x&s = Acs pcs K hs x s p s K e xe Fe
(18)
734
AFASES - 2008 -
Vc
p& cs
(19)
n pV p nmVm K lm p s K cs ( p s pcs ) K s x s p s =
Vt
p& s .
(20)
(21)
This measure represents the available theoretical debit in the pump engine valve node. The
flows of the two machines can be represented focused on the coefficient:
(22)
K lpm = K lp + K lm
The equation (22) becomes:
V
Qt K lpm p s K cs ( p s pcs ) K s x s p s = t p& s
(23)
If the compressibility of the liquid from the command chamber is neglected the equation (21) is
essentially simplified:
K cs ( p s pcs ) = Acs x& s
(24)
The system consisting in the equations (18), (19) si (23) contains 5 known measures:
xs , pcs , p s , Fe , Qt . If the elastic force is maintained constant and the debit Qt is defined variable,
the transfer function of the valve can be defined with respect to the available debit in the node
between the pump, engine and valve:
p (s)
(25)
H Q (s) = s
Qt ( s )
If Qt is constant and the pre-compression of the spring is changing, the transfer function of the
valve with respect to the elastic force:
p (s)
(26)
H F ( s) = s
Fe ( s )
To establish the two transfer functions we introduce in these three equations the variation of the
involved measures with respect to the values for a point of working within the stationary regime.
The valve characteristic can be described as:
Q
Q
Qs = Qs 0 + s dx s + s dp s + K
(27)
x s 0
p s 0
In which:
Qs
= K s p s 0 = K Qx
xs 0
(28)
Qs
p s
(29)
=
0
K s xs 0
2 ps 0
= K QP
Results:
Qs = Qs Qs 0 = K Qx x s + K QP p s
(30)
735
AFASES - 2008 -
d 2 (xs )
= Acs pcs K hs xs 0 ps K hs p s 0 xs K e xs Fe
dt 2
(31)
d (xs )
K cs (p s pcs ) = Acs
dt
d (p s )
Qt p s (K1 + K cs + K QP ) + K cs pcs K Qx x s = Et
dt
In the last equation,
V
Et = t
ms
(32)
(33)
(34)
represents the volume elasticity for the liquid between the pump, engine and valve.
We apply the Laplace transformation to the equations with the finite differences, resulting:
(35)
s 2 ms xs = Acs pcs K hs xs 0 ps K hs ps 0 xs K e xs Fe
K cs (ps pcs ) = sAcs xs
(36)
Qt p s (K1 + K cs + K QP ) + K cs pcs K Qx xs = sEt ps
(37)
To establish the first transfer function for the valve, we consider Fe = 0 . This has the expression:
x s
Acs K hs x s 0
(38)
=
p s
A 2 cs
2
s ms + s
+ K et
K cs
The first transfer function of the valve has the expression:
p s
s 2 b2 + sb1 + b0
= 3
(39)
Qt s a3 + s 2 a2 + sa1 + a0
Where:
Acs2
(40)
; b0 = K et ;
b2 = m; b1 =
K cs
a3 = m s E t ; a 2 = m s K p + E t
Acs2
A2
; a1 = Et K et + K p cs + Acs ( Acs K hs xs 0 );
K cs
K cs
a0 = K p K et + K Qx ( Acs K hs xs 0 )
The measure K p = K1 + K QP can be named total coefficient of command for the pressure p s .
In the command systems built with valves that are normally closed, the pumps debit is
maintained constant and the pre-compression of the valves spring is changed. It is therefore
useful to determine the transfer function of the valve with respect to the elastic force of reference.
To determine the second function we consider: Qt = 0 . The transfer function will then have the
expression:
p s
sc1 + c0
= 3
(41)
Fe s a3 + s 2 a2 + sa1 + a0
Where: c1 = K Qx ; c0 = Acs .
736
AFASES - 2008 -
For the particular case of the non-compensated system the time constant can be precisely
determined trough simplifying the transfer function. If we neglect the second and the third term
results:
p s
1
1
(42)
=
Qt K QP sTs + 1
The time constant of the valve is:
A2
Et K et + K p cs + Acs ( Acs K hs xs 0 )
K cs
(43)
Ts =
K p K et + K Qx ( Acs K hs xs 0 )
From the analysis of the measure order of the time constant results that it can be written under the
following form:
E
(44)
Ts = t
Kp
It can be so noticed that the time constant is influenced by the volume of liquid submitted to the
pressure variations between the pump, engine and valve, on the slope of the static characteristic
of the valve and by the drain coefficient of the pump and engine.
Stability analysis trough the algebraic criterion
(45)
(46)
(47)
Considering the expression for the coefficient K cs one can determine the maximum value of the
dampers maximum clearance:
L A K
(48)
j s max = 3 1.257 s cs et Et
d s K Qx
The maximum clearance of the damper is twice bigger than the optimum one. It can be notice
that the resistance required by the amortization butterfly valve depends directly of the volume of
the liquid located between the pump, engine and valve. This relation explains the high intensity
of the pressure oscillation among the weekly amortized dumping connections of the pumps
equipped with valves for pressure restriction. Placing an elastic joint between the pump and the
engine can essentially contribute to the reduction of the pressure pulsation when the engine is
blocked by an excessive overload. The parameters which influence the mathematic model are
take over from the specific literature and are presented in the following table [1]:
Acs = 2.8274e-005; ms = 0.0497; ds = 0.006; cds =0.8; cvs = 0.98; beta =
; js = 6*10^(-5);
12
= 22.9; ls = 0.009; = 900; Khs =0.0074; Kes =100000; xe =0.0085; Ee =7*10^7;
737
AFASES - 2008 -
Fig. 1. The response of an non compensated valve = to a step signal of low amplitude
738
AFASES - 2008 -
Bibliography:
[1]. Vasiliu, N.: Actionari Hidraulice si Pneumatice, Vol. I, Ed. Tehnica, 2005
[2]. Halanay, A., s.a.: Periodic motions for loaded two control edges hydraulic copying
systems, Comput. Meth. Appl. Eng. 158 (1998) 367-374
[3]. Halanay, A., s.a: Stability of equilibrium in a four dimensional nonlinear model of a
hydraulic servomechanism, J. of. Eng. Math. 49: 391-406, 2004
[4]. Florea O.: Mathematical Models And The Dynamics Optimization For Hidraulic
Servomechanisms Installed In Ideal Conditions, Proceeding NAV-MAR-EDU 2007,
337-348
[5]. Florea, O.: Mathematical Modeling of a Hydraulic System Equipped with a Pipe Rupture
Valve, Bull. Transilvania Brasov, Vol 13(48)2006, 153-157
Olivia Florea, Assist. Univ. drd., Transilvania University of Brasov, olivia.florea@unitbv.ro, oaflorea@gmail.com
739
AFASES - 2008 -
740
AFASES - 2008 -
Fig.1: Procesul de prelucrare a unei imagini privit sub aspectul principalilor factori ce l
caracterizeaz .
Avnd n vedere etapele de baz pe care le parcurge o imagine, conceptul de calitate a
unei imagini poate mbraca mai multe aspecte. Astfel, alturi de calitatea prelucrrii, se mai
pot enumera urmtoarele componente ale calitii unei imagini: calitatea achiziiei imaginii,
calitatea afirii i, respectiv calitatea observrii/analizei imaginii. S-a realizat astfel o
transpunere a calitii imaginii n planul calitii proceselor pe care imaginea respectiv le
parcurge, n vederea analizei acesteia de ctre om/main.
II. Prezentare sintetic a metodelor de evaluare a calitii imaginilor
2.1 Evaluarea subiectiv
Una dintre cele mai sigure i comune aprecieri a calitii imaginii o reprezint
evaluarea subiectiv, realizat de ctre observatori umani. Pentru asemenea evaluri sunt
utilizai observatori neantrenai (non-experi), astfel ca aprecierea asupra calitii unei
imaginii s corespund percepiei unui observator mediu. Testele sunt conduse de observatori
experi antrenai, care au dobndit abilitatea de a observa degradrile unei imagini la scar
mic i, cruia unui non-expert i-ar scpa. Sunt dou grupe de modaliti subiective de
evaluare: absolute i respectiv comparative. n cazul unei evaluri absolute, observatorului i
se cere s aprecieze calitatea unei imagini, innd seama de o scal de evaluare predefinit.
Pentru o calibrare subiectiv a aprecierii calitii, observatorului i se poate furniza un set de
imagini standard de referin, n timp ce, n alte situaii experimentale, observatorul poate fi
forat s realizeze o apreciere bazat numai pe o experien anterioar. Evaluarea
741
AFASES - 2008 -
comparativ implic observatorul la niruirea unui set de imagini, de la cea mai bun la cea
mai slab calitativ, ntr-un grup particular de imagini.
n principiu, cel puin 20 de subieci sunt considerai necesari pentru a asigura o
certitudine statistic n analiza subiectiv a calitii imaginii. Rezultatele evalurii subiective
sunt prezentate n mod normal, ca un scor de opinie mediu definit astfel:
K
C=
n C
k =1
K
n
k =1
care indic suma erorilor componentelor spectrale corespunztoare pixelului de poziie (i,j). n
mod similar, se poate dezvolta expresia erorii k (de component spectral k) astfel:
N
742
AFASES - 2008 -
Indicatorii de tip bivariabili mai des folosii n analiza asistat de calculator a calitii
imaginilor sunt:
raportul semnal/zgomot SNR (Signal to Noise Ratio);
raportul de vrf semnal/zgomot PSNR (Peak Signal to Noise Ratio);
eroarea ptratic medie MSE (Mean Squared Error);
eroarea medie absolut MAE (Mean Absolute Error).
Practic, aceti indicatori (msuri obiective ale calitii imaginii) reprezint extensii
bidimensionale ale msurilor de calitate folosite pentru caracterizarea semnalelor
unidimensionale. Pentru implementarea funciilor de analiz sunt folosite ca date de intrare
(argumente ale funciilor) imaginea original i imaginea deteriorat (a crei calitate trebuie
determinat) care se presupune c a fost afectat ca urmare a aciunii unor factori
perturbatori.
SNR = 10 log
f 2 (m, n)
m =1 n =1
M
[dB]
m =1 n =1
n general, se consider c o imagine de calitate are indicatorul SNR > 25 dB, iar o
imagine cu diferene imperceptibile are SNR > 30 dB (C. Vertan -[5]).
743
AFASES - 2008 -
Fig.3: Prezentare a unor imagini degradate, cu diferite mrimi ale degradrilor evaluate prin
indicatorul SNR.
MN [max f (m, n)]2
m ,n
PSNR = 10 log
[dB]
m =1 n =1
MSE =
MAE =
1
MN
1
MN
g (m, n) f (m, n)
m =1 n =1
n vederea analizei modului de evoluie a calitii unei imagini supuse procesrii se pot
folosi i msuri (indicatori) de tip monovariabile. Astfel de msuri mai des folosite n
evaluarea calitatii unei imagini sunt:
Nivelul mediu al intensitii imaginii ca msur a strlucirii;
Gradientul mediu caracterizeaz informaia de detaliu coninut ntr-o imagine
f(x,y) i se calculeaz cu relaia:
2
M 1N 1
1
f ( x, y ) f ( x, y )
g = G[ f ] =
( M 1) ( N 1) x =0 y =0 x y
def
Valoarea acestui indicator este o expresie a claritii imaginii pentru care a fost calculat.
Indicatorul notat cu Hsd (Histograms standard deviation) care reflect gradul de
abatere al histogramei (aferente imaginii) de la una ideal. Pentru evaluarea indicatorului
respectiv se folosete relaia:
Hsd =
1 L1
h( x ) h
L x =0
744
AFASES - 2008 -
a (i , j )
i = [ p + 1, p + M ]; j = [q + 1, q + M ]
1
1 p+M q+ N
std2 =
a(i, j ) a
M 1 N 1 i = p +1 j =q +1
745
AFASES - 2008 -
pentru evaluarea abaterii standard ntr-o regiune selectat din imagine, se poate
apela funcia MATLAB std2.
V. Concluzii
9 Evaluarea calitii unei imagini este n principiu asociat cu aprecierea mrimii
fidelitii imaginii respective, pentru care se pot stabili msuri adecvate care s reflecte gradul
de ndeprtare a imaginii evaluate fa de o anumit imagine standard (de referin).
9 Msura calitii unei imagini poate fi dat de calitatea proceselor pe care aceasta le
parcurge. Astfel, alturi de calitatea prelucrrii, se mai pot enumera urmtoarele componente
ale calitii unei imagini: calitatea achiziiei imaginii, calitatea afirii i, respectiv calitatea
observrii/analizei imaginii.
9 Evaluarea subiectiv este una dintre cele mai sigure i comune aprecieri a calitii
imaginii, fiind realizat de ctre observatori umani. Rezultatele evalurii subiective sunt
prezentate n mod normal ca un scor de opinie mediu.
9 Principalii indicatori de tip bivariabili folosii n analiza asistat de calculator a
calitii imaginilor sunt:
eroarea ptratic medie MSE (Mean Squared Error);
raportul semnal/zgomot SNR (Signal to Noise Ratio);
raportul de vrf semnal/zgomot PSNR (Peak Signal to Noise Ratio);
eroarea medie absolut MAE (Mean Absolute Error).
9 Printre msurile de tip monovariabile care se pot folosi n analiza evoluiei calitii
unui imagini supuse procesrii, se pot aminti: gradientul mediu, nivelul mediu al intensitii,
indicatorul Hsd (Histograms standard deviation).
9 MATLAB constituie un instrument informatic util care se preteaz a fi folosit n
aplicaiile de procesare a imaginilor i n analiza asistat de calculator a calitii imaginilor.
Bibliografie
[1] Ismail Avcibas, Bulent Sankur, K. Sayood: Statistical evaluation of image quality
measures, Journal of Electronic Imaging 11(2), (April 2002).
[2] Marin Ghinea, Virgiliu Fireeanu: MATLAB -calcul numeric i grafic, aplicaii, Editura
Teora, Bucureti, 2001.
[3] William K. Pratt: Digital Image Processing, cap.7.:-Sampled image quality measures.
A. Wiley Interscience Publication, New York, 1991.
[4] Perry Sprawls, Ph.D.: Image Characteristics and Quality.
http://www.sprawls.org/ppmi2/IMGCHAR/
Tamarina Florea, profesor, Colegiul Virgil Madgearu, Iai, Calea Plieilor nr. 9, tel. 0232.237.539, e-mail:
ionelcfr@yahoo.com
746
AFASES - 2008 -
TILING OF SURFACES
Monica FODOR *
Monica TEFU **
Lorentz JANTSCHI ***
Mircea DIUDEA ****
Tiling a planar surface by various polygonal or curved regions was an ancient human
activity. Covering is nowadays a mathematically founded science. A computer program, CVNET,
useful in generating coverings with application in chemistry was made. The program has options
for map operations, and was made for generating molecular graphs representing nanostructures.
In this paper are presented mainly the algorithms that analyze the original graphs and then
generate the new ones.
Key words: algorithms, map operations, molecular graphs, nanostructures
2000 MSC: 51H05, 90C35, 90-04
747
AFASES - 2008 -
studied by theoretical and experimental methods as well and display unusual and distinctive
electrical, thermal, and mechanical properties.
A classical fullerene called also buckyball has a pseudo-spherical shape made up entirely
of pentagons and hexagons. As the surface is curved with non-zero Gaussian curvature and
according to Eulers formula [4,5], these fullerenes have 12 pentagons and v/2-10 hexagons,
where v is the number of atoms [6]. As theoretical models, fullerenes with various numbers of
pentagons or with larger face have been proposed.
Cylindrical fullerenes called carbon nanotubes, can have only a few nano-meters in
diameter, yet up to millimeters length. They are constructed mostly from hexagonal arrangements
of carbon atoms [7], but they often contain defects as well, as non hexagonal polygons. In the
domain of molecular modeling, derived forms of nanotubes have been proposed, as nanotubes
whit various covering, the faces having between 4 and 8 vertices [8], nanotori and branched
nanotubes.
C60 classical fullerene (a)
748
AFASES - 2008 -
II Map Operations
The Map Operations can be classified in simple, composite, generalized and others
operations [8,14]. Similar with the graph notations, we denote in a map: v the number of
vertices, e the number of edges, f the number of faces and d the vertex degree. A subscript
index 0 will mark the corresponding parameters in the parent map.
Dualization Du of a map is built as follows: locate a point in the centre of each face and
join two such points if their corresponding faces share a common edge. The new map is called
the dual Du(M) (Poincar). The following relations exist in the resulting map [10]:
v = f 0 ; e = e0 ; f = v0
Du(M):
Dual of the dual, recovers the original map: Du(Du(M)) = M.
Medial Me [15] puts the new vertices as the midpoints of the original edges and join two
vertices if and only if the original edges span an angle on a face in the original map. Only the new
vertices are retained. The resulting parameters are:
v = e0 ; e = 2e0 ; f = f 0 + v0
Me(M):
Points in the Medial represent original edges, thus this operation can be used for
topological analysis of edges in the parent polyhedron. Similarly, the points in dual give
information on the topology of parent faces.
All the Platonic, Archimedean and Catalan objects can be obtained, by simple map
operations, from smaller objects. Figure 2 illustrates Dual and Medial operations on the Cube.
Cube
Dual(Cube) = Octahedron
Cubeoctahedron = Me(Cube)
(1)
which predicts the multiplication factor m = v/v0 in a 3-valent map transformed by a composed or
generalized operation.
749
AFASES - 2008 -
Icosahedron
Truncation operation
Le( M ) = Du ( St ( M )) = Tr ( Du ( M ))
In trivalent maps it has the multiplication factor 3. Figure 3 illustrates the Le operation on
the Dodecahedron.
Dodecahedron (C20)
Icosahedron = Du(Dodecahedron)
C60 = Le(Dodecahedron)
750
AFASES - 2008 -
Q(Cube) (a)
Ca (Cube) (b)
Truncated Hypercube
Leapfrog(Hypercube)
751
AFASES - 2008 -
To find the faces in a map, the program uses a recursive algorithm to detect all the cycles
of the graph with length less or equal to a input parameter named max_c. In this respect, the
program uses the table of links between vertices from the input file. Then the program selects the
cycles as faces, in ascending order of their size, under the condition that any edge shares
maximum 2 faces. We will detail below this algorithm.
Lets assume that the variables ci and c are three-dimensional arrays, where ci contains all
the cycles or rings of size less or equal to max_c and in the array c will obtain the faces of the
map. In the array ci, ci[i,0] is the number of faces contained in the vector ci[i], which begins at
the vertex i; nvi = ci[i,j,0] is the number of vertices of the face ci[i,j], while ci[i,j,1] ... ci[i,j,nvi] is
the list of vertices of the face ci[i,j]. The array c will have a similar structure. The variable n
stores the number of vertices in the input map. The algorithm works as follows:
Int i,j,k,size_c, next_v, faces_no, add;
for size_c := 1 to max_c
// the size of the cycles goes from 3 to the input parameter max_c.
for i := 1 to n // crosses the array ci
for j := 1 to ci[i,0]
if (size_c == ci[i,j,0]) and (verify_faces(ci[i,j])){
// if the cycle ci[i,j] has the size size_c and if ci[i,j] doesnt contains other faces
for k := 1 to ci[i,j,0]{ // verify the edges of ci[i,j]
add=1;
if k< ci[i,j] then next_v := ci[i,j,k+1];
else next_v := ci[i,j,1];
verify_edge(ci[i,j,k], next_v,faces_no);
//verify if on the edge (ci[i,j,k], next_v) is more than 1 face in the array c. The procedure
//returns the parameter faces_no. If there is an edge of the cycle ci[i,j] on which are already 2
//faces, the cycle will not be added in the faces array c.
if (faces_no == 2) then add=0;
} // end for
if (add == 1) then
add_cycle(ci[i,j],c); //add the cycle in the faces array c
} // end if (size_c == ci[i,j,0])
Next, the algorithm of map operations works in the following steps, for every face or
cycle:
1. Put the vertices of the resulting map at their coordinates, calculated function of the
coordinates of the vertices from the initial map, for every old face.
2. Link the vertices located inside of an old face.
3. Link the vertices located in different old faces. In case of chiral resulting objects, the old
faces are visited by adjacency, so that the proper links are accomplished (Figures 7 and 8).
752
AFASES - 2008 -
Figure 8. The order of visiting the faces in case of a chiral resulting object: if the algorithm
starts with the green face, the next will be the red, then the yellow and the blue faces.
The maximum size of the input structure is about 3.000 vertices. The program has some
constants for allocating variables for the number of vertices, valence, number of faces, faces
shared by an edge, etc. They was modified when the option for all the cycles was introduced.
REFERENCES
[1] Grnbaum B., Shephard G. C., Tilings and Patterns, Freeman, New York, 1985.
[2] Stefu M., Diudea M. V., CageVersatile (.Net), Babes-Bolyai University, Cluj, 2005.
[3] http://en.wikipedia.org/wiki/Fullerene, http://en.wikipedia.org/wiki/Fullerene#Variations
[4] L. Euler, Comment. Acad. Sci. I. Petropolitanae 1736, 8,128-140.
[5] L. Euler, Novi Comment. Acad. Sci. I. Petropolitanae, 1758, 4, 109-160.
[6] Diudea, M. V. Chemistry of Fullerenes, Master Course, Fullerene1.ppt file, Faculty of
Chemistry and Chemical Engineering, 2006.
753
AFASES - 2008 -
[7] Diudea M. V., and P. E. John, Covering polyhedral tori. Commun. Math. Comput. Chem.
(MATCH), 2001, 44 103-116.
[8] Diudea M. V., Covering Nanostructures, In: M. V. Diudea, Ed., Nanostructures-Novel
Architecture, NOVA, New York, 2005, 203-242.
[9] Harary, F., Graph Theory, Addison - Wesley, Reading, M.A., 1969.
[10] Pisanski T. and Randi M., in Geometry at Work, M. A. A. Notes, 2000, 53, 174-194.
[11] Klein D. J., Zhu H., in: A. T. Balaban, (Ed.), From Chemical Topology to Three Dimensional Geometry, Plenum Press, New York, 1997, pp. 297-341.
[12] La Vaissire B. de, Fowler P. W., Deza M., J. Chem. Inf. Comput. Sci., 2001, 41, 376-386.
[13] Deza M., Fowler P. W., Shtorgin M., Vietze K., J. Chem. Inf. Comput. Sci., 2000, 40, 13251332.
[14] Diudea M. V., Covering Forms in Nanostructures, Forma (Tokyo), 2004, 19 (3), 131-163.
[15] Fowler P. W. and Pisanski T., J. Chem. Soc. Faraday Trans. 1994, 90, 2865-2871.
[16]Goldberg M., Tohoku Math. J. 1937, 43, 104-108.
[17] Fowler W., Phys. Lett., 1986, 131, 444.
Fowler P. W., Steer J. I., J. Chem. Soc., Chem. Commun., 1987, 1403-1405.
Fowler P. W., Rogers K.M., J. Chem. Soc., Faraday Trans., 1998, 94, 1019-1027.
Fowler P. W. , Rogers K.M., J. Chem. Soc., Faraday Trans., 1998, 94, 2509-2514.
[18] Brinkmann G., Fowler P. W., and M. Yoshida, MATCH Commun. Math. Comput. Chem.,
1998, 38, 7-17.
[19] Stefu M., Butyka D., Diudea M. V., Jantschi L., and Parv B., Algorithms for basic operations
on maps, In: M. V. Diudea, Ed., Nanostructures-Novel Architecture, NOVA, New York,
2005, 243-267.
Monica Fodor, Professor at National College Mihai Viteazu, Bariiu Str. nr. 82, Turda, PhD student at Technical
University Cluj-Napoca, B-dul Muncii no. 103-105 Cluj-Napoca, e-mail: monicai.fodor@yahoo.com
**
Monica tefu, PhD. student at Babes-Bolyai University, Mihail Koglniceanu Str., no. 1, Cluj-Napoca, tel.
0720.568.971 e-mail: mstefu02@yahoo.com
***
Lorentz Jantschi, Sef L. Dr., Technical University Cluj-Napoca, B-dul Muncii no. 103-105 Cluj-Napoca, tel.
0766.239997, e-mail: lori@academicdirect.org
****
Mircea Diudea, Prof. Dr., Babes-Bolyai University, Mihail Koglniceanu Str. no. 1, Cluj-Napoca, e-mail:
diudea@chem.ubbcluj.ro
754
AFASES - 2008 -
Abstract: The dissipative power in an electric arc to the REZ process assures the
necessary quantity of heat in normal conditions. We offer the premises to a mathematical
analysis of the thermal transfer in the electric arc. Using the thermal equilibrium equation, an
initial time of the initiation of the electric arc, an integral transformation technique, we will
formulate a boundary value problem of elliptic type. Some techniques of bifurcation theory
can be applied to reveal that a ramification of solution is achieved from the first eigen-value
of the linear part and a continuation is given by means of implicit function theorems.
AMS Subject Classification: 70E20, 74B20, 74A45, 74R10
Key words: electric arc, heat extraction control, quasi-linear boundary value problem
I. Introduction
From the technological point of view, the implemented power in electric arc to a REZ
process (160kVA order) on a pilot installation for ingots of half and low weight, assures the
necessary quantity of heat for another melt of ingot-electrode in normal conditions of work
(on standard installation parameters). With an energetic balance on the electric arc level
developed between ingot-electrode and metallic bath, we create premises to
thermodynamically property analysis of the arc. Difficulties in mathematical modeling of
energetically equilibrium of the arc are risen up by the specification of power dissipation
function through column and implicitly of heat extraction function on the electrode foot level.
We put the problem of perturbations analysis which appears on changing power steps. The
calculation of gyved up heat could be done at electric current parameters supplied by
transformer: 2000 A, 50-80 V ([5], page 45). In case of soliciting installation electric energy
variations, we impose the preliminary calculus requirement of extract heat on the ingotelectrode foot level, for REZ process behavior characterization. We can specify the power
dissipation function at the column arc level and of the heat extraction function on the
electrode foot level. Taking account a linear perturbation, depending o a rate factor of power
in changing the power step, we study the bifurcation of solution [4]. We will see how results
of bifurcation theory assure the ramification of solution u of (3.1) problem from a solution u0
of an associated eigenvalue problem.
755
AFASES - 2008 -
II. Some technological features of the electric arc and a modeling problem.
We adopt for heat transmitted in metallic mass the thermal diffusion equation. In
analyzing heat source context, we require heat quantity determination through an
energetically result:
(2.1)
dQ = (U I )dt ,
where U is electrical tension, I electrical current intensity from electrode, a dissipation
function of electrical energy in arc, dQ elementary variation of extraction heat. For
determining the extraction heat dissipation function trough thermal conduction is necessary to
determine the distribution of temperature in arc column. This is the equation of heat diffusion
in liquid metal mass:
T
(2.2)
aT = 0,
t
T is thermo dynamical temperature, a thermal spreading. We accept the cylindrical form of
electric arc developed between electrode and metallic bath. We consider T=T(t,r,,z), r,,z
are the cylindrical coordinates of the system of reference. From grounds of symmetry, it is
sufficient to have T=T(t,r,z) and take the laplacian in cylindrical coordinates:
1 2
=
(2.3)
.
r +
r r r z 2
We have 0 (r , z ) = T (0, r , z ) the temperature at the moment of the initiation of electric arc
between ingot and the main board of crystal-maker, covered with electric-conducting slag.
We impose as a boundary condition the proportionality of heat flux q with P, distribution of
power on the surface level of metallic bath:
P
r, r b
T
(t , r ,0) = b 2
(2.4)
q (r , t ,0) =
z
0, r > b,
where b is the thickness of electric arc. On this condition it is considered that power transmits
itself like a step function on the contact surface electrode-melt steel. It is used the expression
in cylindrical coordinates of diffusion equation (2.2), with expression (2.3) at laplacian. We
introduce the laplacian transformer:
(2.5)
(r , z , p ) = e tpT (t , r , z )dt ,
0
1 2
a
r
+ 2 2 + p =T 0 (r , z ), (r , z ) ,
r r r z
(2.7)
P 1 1
b 2 p 2 p 0 (r ,0), r b
0 (r , p ) =
1 (r ,0 ), r > b.
p 0
756
AFASES - 2008 -
We consider that at the initial moment the temperature 0 is uniformly distributed on the
contact surface electrode-slag. We introduce the variable:
(2.8)
u (r , z , p ) = (r , z , p ) 0 (r , p ),
1
(r u ) + 2 2 + p u = f (r , z, p ), u = 0, for z=0,
a
z
r r
a1
P 1
+ 0 (r ,0 ), for r b ,
r 0 (r ,0) + 2
p r r r
b p
a1
f (r , z , p ) = 0 (r , z ) +
r 0 (r ,0) + 0 (r ,0), for r > b .
p r r r
and it is independent from u, it is accepted f L2 ( ) and we could see (2.9) like a boundary
value problem on a restricted domain, associated with a quasi-linear operator.
III. Some preliminaries results.
Lu u + f (u ) = 0 in , Bu = 0 on ,
aij
u + a.u ,
t
xi x j
aij C 1+ ( ) (0,1), Bu = u , it is accepted the operational form of equation (3.2):
(3.3)
F (u , ) = K (u f ) = u ,
757
AFASES - 2008 -
(3.4)
The eigenvalue 1 of the operator L from (3.2) are characteristic value for application K from
(3.3). It is considered the eigenvalue quasi-linear problem:
K (m.n ) = u , m C ( ), m( x ) > 0 a. p.t. x ,
(3.5)
( )
( )
( )
( )
v C , v 0 , so that G 'u (u, )v = 0 . Then, equation F 'u (u, )v = v, has the solution v 0 ,
so the eigenvalue problem K ( f ' (u ))v = v has v as a eigenfunction, for a particular
eigenvalue k ( f ' (u )) = 1, with k > 1 specified. On the other hand, u verifies
f (u )
problem:
(3.6)
f (u )
K
u = v.
u
f (u )
has an eigenvalue j
= 1, with u eigenfunction, positive or negative, which is the
u
f (t )
solution to equation F (u , ) = u. We observe that function f satisfies
< f ' (t ), t 0.
t
Indeed, the conditions f (0 ) = f ' (0 ) = 0, f convex, assure f as a local increasing function for
758
AFASES - 2008 -
f (t )
is inferior to graphic tangent :
t
f ' (t ) . We have
f (t )
> f ' (t ) = > 0, from where, using monotony of eigenvalues, results
t
f (u )
f (u )
1 ( f ' (u )) 1
= 1, that is to say 1
= 1 . But, for (2.6)
u
u
f (u )
1 = 1
> 1 ( f ' (u )) > k ( f ' (u )) = 1, is contradiction (because k > 1, k 0 ).
u
1 + k +
Proof: We see that application F (, ) = K ( f ()) does not vary the positive cone
( )
PC
([1])
and
is
an
asymptotic
Having
application,
having
F ' + (0, )v =
k+
v > v , v P .
1
now
linear
(1 , 1 + k + ) ,
then
k+
> 1,
< 1,
1
1
so
1
>
= 1 (Lemma 3.1). Because f ' (u ) > k + , we have 1 = 1
u
u
The results from Lemma 3.1 and 3.2 assure irreversibility of application G (.,.) , so it is
a local isomorphism for 1 . The following result shows us existence of two no vanish
branches of solutions, making a bifurcation from solution (0, 1 ) at the problem (3.3),
parametrically appointed through:
( )
( )
+ = (u, ) C R | u = u+ ( ), 1 1 + k+ ,
(4.1)
= (u, ) C R | u = u ( ), 1 1 + k ,
Theorem 4.1: Let f C 1 (R ) be a function that satisfies the imposed conditions; then the
multitude no vanish solutions of equation (3.3), making a bifurcation in point (0, 1 ) , form
759
1 + +
AFASES - 2008 -
(3.3) has only vanish solution for (0, 1 ) and neither positive nor negative solution for
2 .
Proof: We prove in the first place that F (u , ) = u has solution for 1 < < 2 , with
f (u )
f (u )
< < 2 the
3.1.). Using continuation property of no vanish solution u, we have for
u
f (u )
f (u )
f (u )
f (u )
negative.
If u = u+ ( ) is the positive solution for F (u, ) = u, conforming Lemma 3.2 there exist for
1 < < < 1 + k+ the solution branch + . Analog, a no vanish negative solution branch
exists. Because for < 1 , the only solution is vanish, the solutions branches are bifurcated in
point (0, 1 ) .
Corollary 4.1: Having c1 = min{1 + k + , 1 + k } and c2 = max{1 + k+ , 1 + k } , then the
boundary value problem (2.2) has two no vanish solutions branches for (1 , min (c1 ,c2 )) , a
no vanish solutions branch, if c1 < min (c2 , 2 ) and [c1 , min (c2 , 2 )] , none of the no
vanish solutions if c2 2 and [c2 , 2 ] .
For many other considerations see [2], [3]. A numerical analysis of this solution and other
technological considerations will be presented in a forthcoming paper.
References
[1] H. Amann, Nonlinear operators in ordered Banach spaces and some applications to
nonlinear boundary-value problems. Lect. Notes Math. 543, Springer-Verlag, (1976), p.
1-55.
[2] C. Ghi, Some considerations about monotone iterative schemes to nonlinear eigenvalue
problems, Lucrri tiinifice. Matematic, Fizic, Proc. Summer school Constana, 1981.
[3] C. Ghi, Metalurgie Matematic. Analiza nestandard a proceselor metalurgice, Ed.
Academiei Romne, Bucureti 1995.
[4] D. H. Sattinger, Topics in stability and bifurcations theory, Lect. Notes Math., 309,
Springer-Verlag, 1973.
[5] I. Tripa & C. Pumnea, Retopirea i rafinarea oelurilor, Ed. Tehnic, Bucureti, 1981.
760
AFASES - 2008 -
761
AFASES - 2008 -
care se ocup de aceast problem; una din acele tehnologii utile este cea care folosete logica
de tip fuzzy [2] pentru a ajusta adaptiv, on-line, covariana zgomotului de msur.
Dei manevra cu scurte acceleraii i diversitatea de zgomote de msur au fost
discutate, s-a lucrat destul de puin la a rezolva cele dou probleme n aceeai arhitectur.
Voi prezenta aici o viziune de arhitectur multisenzor distribuit de urmrire ce
integreaz reea neuronal i logic fuzzy. n acest sistem de fuziune, reeaua neuronal ia
datele de la senzori ca intrri, apoi export estimata acceleraiei ctre filtrul Kalman i, n
acelai timp, la sistemul de logic fuzzy, care ajusteaz covariana zgomotului de msur al
senzorilor i ofer un grad de ncredere pentru a fuziona rezultatele date de filtrele Kalman.
Capitolul II: DESCRIERE A NOII ARHITECTURI DE FUZIUNE DATE
2.1 Filtrul Kalman
Filtrul Kalman este un algoritm recursiv de procesare a datelor. Ecuaiile sale
ncorporeaz o nou msurtoare n estimata a priori pentru a obine o estimat a posteriori
mbuntit. Dac notm cu x k estimata vectorului de stare xk, iar Pk este matricea de
covarian corespunztoare erorii de estimare a strii,
Pk = E{( xk x k )( xk x k )T }
(1)
Termenul H k x k este ieirea zk prezis la momentul anterior, iar ( z k Hx k ) este
eroarea de predicie de la momentul anterior, cunoscut i sub denumirea de secvena de
inovaie sau rezidualul:
rk = ( z k H k x k )
(2)
Inovaia reprezint informaia adiional disponibil filtrului, ca o consecin a noii
observaii zk. Inovaia ponderat, K k [ z k H k x k ] , acioneaz ca o corecie la estimata
predicionat x k pentru a forma estimata x k ; matricea de ponderare Kk este referit ca fiind
ctigul filtrului sau matricea de ctig Kalman.
Algoritmul filtrului Kalman pornete de la condiiile iniiale la k = 0, acestea fiind x 0
i P0 . Cu trecerea timpului, pe msur ce noile msurtori zk devin disponibile, ciclul
corecie-estimare al strilor i covariana corespunztoare a erorii vor continua recursiv la
infinit.
2.2 Reeaua neuronal
Studiul asupra reelelor neuronale ne dezvluie c orice transformare neliniar
continu poate fi aproximat ct de mult dorim cu o reea neuronal multistrat, cu o topologie
de propagare nainte i funcii neliniare sigmoidale (de forma lui S).
Elementul de procesare de baz (neuronul) din aceste reele de aproximare a funciei
are o caracteristic de intrare-ieire care se obine prin formarea unei sume ponderate ale
intrrilor recepionate i producerea unei ieiri care este o funcie neliniar de aceast sum
ponderat, conform relaiei:
m
y (t ) = g[ wi ui (t )]
i =1
762
(3)
AFASES - 2008 -
unde ui(.): RR, i = 1,2,,m sunt intrrile; y(.): RR este ieirea; wiR, i = 1,2,,m sunt
ponderile. Aici g(.): RR este o funcie de activare neliniar selectat n mod adecvat, care
satisface urmtoarele condiii: xg(x) > 0 pentru toi x R ; lim|x|g(x) = ksgn(x), k > 0;
g(x1)/x1 g(x2)/x2 pentru toi |x1| |x2|.
Cele mai des ntlnite funcii de activare sunt caracteristicile sigmoide; spre exemplu,
g(x) = tanh(x) sau g(x) = (1+e-x)-1. Arhitectura de procesare a unei reele multistrat cu
propagare nainte, cu un strat de intrare, un strat de ieire i cteva straturi ascunse este
prezentat n figura 1. n aceast arhitectur, stratul de intrare are 4 noduri, care doar
distribuie intrrile la nodurile din urmtorul nivel (primul strat ascuns), iar stratul de ieire are
dou noduri, cu rolul doar de a combina ieirile nodurilor din stratul anterior (ultimul strat
ascuns). Straturile ascunse au un numr arbitrar de noduri care realizeaz procesarea neliniar,
conform regulii stabilite.
De o importan fundamental pentru antrenarea n bune condiii a reelei neuronale
este selectarea unui set adecvat de caracteristici de intrare. Utiliznd un numr suficient de
caracteristici n procesul de antrenare, este posibil s obinem proprieti de foarte bun
toleran la erori (robustee) pentru procesorul cu reea neuronal. Mai mult, ntruct
arhitectura de reea se poate alege astfel nct s accepte ca intrri date colectate de la mai
muli senzori (la fel sau diferii), fuziunea datelor se poate face implicit. Uneori este necesar
o preprocesare a datelor pentru a putea fi utilizate avantajos ca intrri n reea.
La selecia caracteristicilor adecvate ne putem ghida dup observaia c n general
exist trei entiti de baz care ajut la obinerea unei bune estimate a manevrei intei. Acestea
sunt: (1) intensitatea acceleraiei, (2) direcia vitezei tangeniale i (3) viteza iniial la
momentul acceleraiei.
Sunt disponibile dou intrri pentru reeaua neuronal, extrase din datele brute de la
senzor:
1: intensitatea acceleraiei, definit astfel:
ry2 (k )
rx2 (k )
1 (k ) =
+
(4)
S xx (k ) S yy (k )
unde r(k) = [rx ry] este inovaia (rezidualul), iar Sxx(k) i Syy(k)sunt elementele diagonale ale
matricei covarian:
S (k ) = HP(k | k 1) H T + R
(5)
2: modificarea de cap-compas, definit astfel:
2(k) = LT(k) - LT(k - 1)
(6)
unde LT(k) i LT(k - 1) sunt estimatele capetelor compas, calculate innd seama de ultimele
poziii (N este 3) la momentele k i respectiv k-1, adic:
N
i =1
i =1
LT (k ) = [ ( yi y ) 2 / ( xi x ) 2 ]1 / 2
(7)
unde
N
= sgn ( yi y )( xi x )
(8)
i =1
x = (1 / N ) xi
(9)
y = (1 / N ) yi
(10)
i =1
N
i =1
763
AFASES - 2008 -
din valorile anterioare ale inovaiei) este egal cu zero; dar atunci cnd exist o manevr, ea
nu va mai fi zero.
2.3 Sistemul de logic fuzzy
Principalele avantaje ale utilizrii unui sistem cu logic fuzzy sunt simplitatea
abordrii i capacitatea de a lucra cu informaie imprecis i de asemenea posibilitatea de a
include cunotine euristice despre un fenomen luat n considerare.
Capacitatea unui sistem cu logic fuzzy de a ajusta n mod adaptiv covariana
zgomotului de msur a fot demonstrat n [2]. Fundamentul acestei aplicaii este numit
algoritm de estimare adaptiv bazat pe inovaie [3]. Ideea de baz din spatele acestui algoritm
este de a face valoarea actual a covarianei rezidualului compatibil cu valoarea sa teoretic.
Secvena de inovaie r(k) are o covarian teoretic S(k) definit prin ecuaia (5) de mai
nainte. Avnd secvena de inovaie r(k), covariana sa actual C rk este aproximat prin
covariana de la pai diferii prin mediere n cadrul unei ferestre de estimare glisante de
dimensiune N.
1 N T
C rk =
(11)
ri ri
N i = i0
unde i0 = k N + 1 este primul eantion n cadrul ferestrei de estimare. Dimensiunea ferestrei
N este aleas empiric pentru a da o anumit mediere statistic.
Astfel, dac se observ c actuala covarian a lui r(k) este diferit de valoarea sa
teoretic, atunci trebuie s facem ajustri pentru R n sensul corectrii acestei nepotriviri. Este
definit o nou variabil numit Gradul de Potrivire (GP) pentru a detecta discrepana dintre
S(k) i valoarea sa actual C rk ,
GP = S C
(12)
k
rk
764
AFASES - 2008 -
Sistemele FIS i FLO (de inferen i observatorul) se pot integra ntr-un singur sistem
fuzzy, care va lua GP ca intrri i va scoate R ajustat ctre filtrul Kalman i w (gradul de
ncredere) ctre modulul de fuziune.
2.4 O nou arhitectur de fuziune date
Aceast nou arhitectur de fuziune date am gndit-o n spiritul ideii mai largi de a
nva reeaua neuronal cu parametrul acceleraie al manevrei, n mod special pentru
acceleraiile scurte i rapide; apoi de a ajusta on-line covariana zgomotului de msurare cu
ajutorul sistemului cu logic fuzzy. Sunt trei algoritmi utilizai n aceast arhitectur: reea
neuronal, filtru Kalman i logica fuzzy. Relaiile dintre ei sunt ilustrate n figura 1. Mai nti,
toate msurtorile noi ale senzorilor de la momentul k vor fi transmise la reeaua neuronal,
sistemul logic fuzzy dar i la filtrele Kalman. Sistemele cu logic fuzzy vor regla parametrii R
(covariana zgomotului de msur) ai filtrelor Kalman i n acelai timp vor furniza gradul
corespunztor de ncredere n informaiile senzorilor ctre modulul de fuziune. Modulul de
reea neuronal va exporta parametrul acceleraie la filtrele Kalman i la sistemele fuzzy.
Modulul de fuziune va fuziona ieirile filtrelor Kalman, corespunztor gradului de ncredere.
n figur avem: 1: intensitatea acceleraiei definit anterior, 2: modificarea de cap-compas
definit anterior, 3 - n: alte intrri pentru reeaua neuronal care sunt utilizate n estimarea
parametrului acceleraie, formele lor detaliate depinznd de proprietile senzorului, Z(k):
msurtoarea senzorului la momentul de eantionare k, S(k): covariana secvenei de inovaie,
definit anterior, a: parametrul acceleraie estimat la momentele de eantionare k, X(k - 1):
valoare estimat al momentele de eantionare k 1, GP: Gradul de Potrivire, intrare a
sistemului logic fuzzy, R(k): covariana zgomotului de msurtori la momentele de
eantionare k, una din ieirile sistemului logic fuzzy, w: gradul de ncredere pentru un senzor,
una din cele dou ieiri ale sistemului logic fuzzy.
n aceast arhitectur, mai trebuie introduse alte dou module: modulul de operaii
(prelucrri) matematice i modulul de fuziune. Modulul de operaii matematice este utilizat
pentru a executa operaii matematice simple. Modulul de fuziune are ca scop fuzionarea
ieirilor filtrelor Kalman ctre o valoare care este mai apropiat de valoarea real bazat pe
algoritmul de fuziune, astfel:
m
wi xi (k )
~
i =1
X (k ) =
m
wi
(14)
i =1
unde m este numrul de senzori, wi este gradul de ncredere al celui de-al i-lea senzor,
xi (k ) este estimarea valorii reale la momentele de eantionare k date de al i-lea filtru Kalman,
~
X (k ) este rezultatul fuziunii la momentele de eantionare k.
Capitolul III: SIMULRI I EVALUAREA PERFORMANEI
Pentru a verifica eficiena unei asemenea arhitecturi, n aceast seciune voi prezenta o
simulare. Vom considera doi senzori n cadrul sistemului nostru de urmrire, ambele
furniznd date de msurtori n dou dimensiuni. Cei doi senzori au ns condiii diferite de
lucru i sensibiliti diferite la situaiile meteo nefavorabile. Modelul de urmrire ales este
unul liniar:
765
AFASES - 2008 -
1 2
0
T
wx
(
)
(
1
)
x
k
x
k
1 0 T 0
2
w
y (k )
y ( k 1)
u
0
1
0
1
T
y
x
2
+ 0
T +
wv x
v x ( k ) 0 0 1 0 v x ( k 1)
2 u y
T 0
(
)
(
1
)
v
k
v
k
0
0
0
1
w
y
vy
y
0 T
(15)
x(k )
y (k )
z x (k )
x
(16)
= [1 1 0 0] v (k ) +
y
x
z y (k )
v y (k )
unde T este perioada de eantionare.
Reeaua neuronal are patru intrri i dou ieiri. Numrul de neuroni ales pentru
stratul ascuns se poate decide doar dup antrenare. Pentru a avea o bun acuratee la estimarea
acceleraiei, am creat 2000 de seturi de date de antrenare dup urmtoarea schem:
- Acceleraie: ntre 0 la 20, mprit n 20 de nivele;
- Viteza iniial: ntre 100 i 200, mprit n 10 nivele;
- Covariana zgomotului de msur: ntre 0,1 i 1, mprit n 10 nivele.
Rezultatele antrenrii au artat c 35 de neuroni sunt suficieni pentru stratul ascuns.
Se ridic dou ntrebri n legtur cu arhitectura de fa. Mai nti, dac aceast
arhitectur lucreaz eficient atunci cnd avem n mod sistematic manevre de scurt durat i
diverse zgomote de msur; apoi, pentru situaia cnd covariana zgomotului de msur al
celor doi senzori se modific n acelai timp.
766
AFASES - 2008 -
767
AFASES - 2008 -
Scopul acestei lucrri a fost de a prezenta o nou arhitectur de fuziune ce s-ar putea
utiliza ntr-un sistem de urmrire a intelor i care poate rezolva problema manevrelor cu
scurte acceleraii i a ajustrii adaptive a covarianei zgomotului de msur al senzorilor.
Rezultatele simulrii au artat c aceast arhitectur poate rezolva cele dou probleme
anterioare dar are totui dezavantajul c senzorii utilizai n sistemul de urmrire nu trebuie s
aib acelai zgomot de msurtori.
Bibliografie
[1] M.W. Owen and A.R. Stubberud, NEKF IMM Tracking Algorithm, Signal and Data
Processing of Small Targets, 5204:223-233, 2003.
[2] P.J. Escamilla-Ambrosio, N. Mort, A hybrid Kalman filter-fuzzy logic architecture for
multisensor data fusion, International Symposiumon Intelligent Control, pg. 364-369, 2001.
[3] A. H. Mohamed and K. P. Shwarz, Adaptive Kalman filter for INS/GPS, Automatic
Control, pg.193-203, 1999.
Grni Marian-Silviu, doctorand, Academia Tehnic Militar, Bucureti, George Cosbuc nr. 81-83, tel.
0213354660, e-mail: silviu_girnita@yahoo.com
768
AFASES - 2008 -
Abstract: RSS stands for Rich Site Summary and is a type of XML document used to share
news headlines and other types of web content. RSS has been adopted by news syndication
services, weblogs, and other online information services. RSS and RSS feeds are relatively
new to internet users, being the most important technology to arrive on the internet since
blogging. Because this technology provides us with a method of getting relevant and up to
date information sent to us and for us to read in our own time, it saves a lot of time and helps
getting the information we want quickly after it was published. The content comes to us
instead of going to get it! The strength of RSS is its simplicity. It is exceptionally easy to
syndicate website content using RSS.
2000 MSC code: 68U35
Key words: web syndication, RSS feeds.
I. Introduction
RSS, or Really Simple Syndication as it is commonly known, is a technology that
gives webmasters the ability to easily distribute and publish syndicated content on the
Internet. But a lot of webmasters have absolutely no idea how to create an RSS feed or how
they can incorporate an RSS feed on their website. A tool is not very useful if they do not
know how to use it. It is with this motivation that this article is written, to give a basic
understanding of how to create an RSS feed, validate/incorporate an RSS feed into the
website, promote the feed, and finally how to subscribe a feed.
II. What is RSS?
If we think about the types of websites that offer RSS feeds we see: first, there is the
news and article related websites. These make up the majority of the websites who use RSS.
There are also forums, web portals, search engines, and news aggregators, to name a few. The
one thing all these types of websites have in common is that they are all filled with a lot of
769
AFASES - 2008 -
information that changes. Organizing this information is the difficult part, and organizing it in a
way that others can syndicate and customize the format is even more difficult.
RSS organizes information within tags or labels and places this information into
what could be considered an outline format. All information can be organized into separate
parts. As an example, an article website is made up of articles. Each article can be considered
its own part of the site. Within each article there are parts as well, such as the title of the article,
a description of the article, the date the article was published, who wrote the article, and so on.
What RSS does is to present these parts in a uniform, organized format.
RSS organizes information the same way every time. An RSS feed can be broken
down into a few parts. First RSS presents the header information such as the XML version
and various comments. This is more for the computers than it is for the readers. Next RSS
presents information about the website. The information presented here can vary, but typically
there will be the name of the site, a link to the site, the webmasters e-mail address, and
maybe the last time the feed was updated. The next part to an RSS feed is the actual content
of the feed.
So, to understand about how RSS organizes information, it would be useful to see how
an RSS feed is published.
III. Publishing an RSS Feed
Because RSS is focused on organizing content, creating an RSS feed is fairly easy to
do. The below example is extremely simplified. RSS has quite a bit more flexibility than is
demonstrated here, but for most webmasters a basic RSS feed is all that is needed.
An RSS feed can be broken up into a few simple parts. Similar to regular HTML, the
first part of an RSS feed is the header information. A sample RSS header is located below:
[code]
<?xml version="1.0" encoding="ISO-8859-1" ?>
<!-- Publishing tool used -->
<rss version="0.91">
[/code]
The header is normally the same from feed to feed. After the header comes the actual
content. All of the content is surrounded by channel tags. Below is a snippet of an RSS feed
that contains just two articles:
[code]
<channel>
<title>Website Title, or Title of Section on the Site</title>
<link>http://www.thesite.com/</link>
<description>A short description of what the website is about</description>
<webmaster>Who is the webmaster ofthe site?</webmaster>
<language>en-us</language>
<item>
<title>Title of the first article focusing on</title>
<link>http://www.thesite.com/directlinktoarticle.html</link>
<author>Authors Name</author>
<date>The date the article was published</date>
<description>This is the description of this article</description>
770
AFASES - 2008 -
</item>
<item>
<title>The title of the second article</title>
<link>http://www.thesite.com/directlinktootherarticle.htm</link>
<author>Authors Name</author>
<data>Date Published</data>
<description>This description will describe the second article.</description>
</item>
</channel>
[/code]
The channel tags surround everything. Within the channel tags lies the content being
syndicated. There is the title of the site, the site description, a place for a webmasters e-mail
address, and the language that the site is in. It can be added more tags if neded. For example,
if we want to include information on when the feed was last updated, we might include a tag
called lastUpdate. Whatever information we want to give those who are syndicating our
content, we can give them.
After this information are the item tags. The item tags allow separating the content
being syndicated. In this example, the item tags separate each article. Each item tag separates
items.
Within the item tags is the RSS feeds content itself. In the above example, each item
tag separates an article. Within the item tag, we have more tags which identify the title of the
article, a link to the full article, the articles author, the date published, and a description of the
article. Again, we could include more information or less information depending on what our
goals are.
The final step to create the RSS feed consists in closing all tags. RSS is very strictly
about open tags. So, when creating the RSS feed, the RSS tag that we opened back in the
header must be closed:
[code]
</rss>
[/code]
This information is saved as an XML document and the result will be a valid RSS
feed. The final sample RSS feed looks like the following:
[code]
<?xml version="1.0" encoding="ISO-8859-1" ?>
<!-- Publishing tool used -->
<rss version="0.91">
<channel>
<title>Website Title, or Title of Section on the Site</title>
<link>http://www.thesite.com/</link>
<description>A short description of what the website is about</description>
<webmaster>Who is the webmaster of the site?</webmaster>
<language>en-us</language>
<item>
<title>Title of the first article focusing on</title>
<link>http://www.thesite.com/directlinktoarticle.html</link>
<author>Authors Name</author>
771
AFASES - 2008 -
772
AFASES - 2008 -
So, in order to promote the RSS feed, it is necessary to submit it to as many directories
as possible, such as: Best Blog Directory And RSS Submission Sites - Robin Good,
FeedSubmitter, RSSFeedPromoter.
Finally, it's recommended to "ping" one of the major services that track when web logs
and RSS content changes. By doing this, we ensure that other sites that monitor these know to
check back at our site for more content. Weblogs.com is one of these major sites. By entering the
site's name and the URL of the feed into the manual Ping-Site Form, it will know we have
updated our feed. The Specs page explains how to set up automatic notification.
blo.gs is another major change monitoring site. It is supposed to receive any changes
that come from Weblogs.com, so we shouldn't need to notify it separately. The ping form
explains how to set up automatic pinging, as well.
VI. Subscribing and reading RSS feeds
There are a number of news aggregators or RSS readers allowing to subscribe to and
read RSS feeds. While the in-browser feed readers are convenient they are still quite
rudimentary.
We should consider switching to a full-fledged feed aggregator for more features and
increased usability. There are many feed readers to choose from, each have different levels of
complexity and features. There are essentially two types: web-based and application-based.
With a web-based aggregator we can check our feeds from any computer but this comes at the
cost of limited functionality and speed. With application-based feed readers we get excellent
features and speed but lack mobility.
The best and most popular aggregator (online feed reader), in my opinion, is
Bloglines. Signing up for an account and clicking My Feeds on the top left we can test out
Bloglines by adding a few feeds. If we found the feed on the website we can paste that in the
subscribe page now. However, Bloglines has a feature where we can just type in the URL of
the website and it will search for feeds: it may find several feeds and other times it will not
find any. The figure bellow display a RSS feed in Bloglines aggregators window:
773
AFASES - 2008 -
VII. Conclusions
There are a lot of good resources available and the challenge is actually seeing the
information. RSS is still not a very familiar technology but it's already very obvious what role
RSS will be playing in the future of the Internet. Because the web has become a sprawling
network of different services, RSS is helping the web connect these services in an organized
fashion - the new Web 2.0. In the coming few years, RSS will definitely be seen on every
website and blog. Any service without RSS would lag behind.
RSS is a great supplemental communication method that does not burden the publisher
with maintaining lists or following strict privacy guidelines. RSS feeds are compiled according to
the user's choices, removing any burden that is placed on publishers of email newsletters.
Publishers no longer need to be concerned with spam, privacy policies, and age
guidelines. Publishers using RSS as a communication vehicle are able to create keyword-rich,
themed content, establishing trust, reputation, and ongoing communication with current and
prospective customers.
Rarely in the last 10 years has there been a technology with so much potential to explode
online exposure. A part of Web 2.0 has been the increasing popularity of RSS feeds and
aggregation-syndication that changes the way content is distributed, valued and consumed.
References
[1]. RSS Specifications, http://www.rss-specifications.com/
[2]. http://dictionary.reference.com/wordoftheday/wotds.rss
[3]. http://www.rssguide4u.com/Web_2.0.html
[4]. http://www.syndic8.com/documents.php?Section=HowTo
Luminita Giurgiu, Associate professor, PhD, Nicolae Balcescu Land Forces Academy, Sibiu, Revolutiei
Street, 3-5 No., tel. 0269432990, e-mail: lumigee@armyacademy.ro
**
Ghita Barsan, Professor, PhD, Nicolae Balcescu Land Forces Academy, Sibiu, Revolutiei Street, 3-5 No.,
tel. 0269432990, e-mail: gbarsan@armyacademy.ro
774
AFASES - 2008 -
775
AFASES - 2008 -
The humanitarian subjects are taught with respect to their military applications. Nevertheless,
the graduate study programmes are versatile and offer broad lifelong use in military
professional as well as civilian life.
The UoD provides education in accredited study programmes and in branches which are
structured to ensure academic education in accordance with the needs of the CAF
requirements as well as civilian institutions, research and industry. The university offers
accredited education in bachelor, master programmes, and doctoral studies. It should be
emphasized that, compared to civilian schools, military education at the UoD has a few
important differences. First of all, the UoD is relatively small. The number of its graduates in
every specialization is lower than their number at civilian schools. This situation is caused by
customers demands i. e. by the MoD which sticks to the CAF figures planned. Bearing in
mind future graduates positions, the subjects are partially tailored for the future commitment.
The Department of Communication and Information Systems (CIS) was established at the
Faculty of Military Technology on September 1st, 2005 and has included two former
departments Department of Special Communication Systems and Automated Command
Systems and Informatics Department. The Department of CIS is divided into five groups of
specialists. These groups are oriented on radiocommunication systems, telecommunication
systems, information systems and programming, computer networks and operation
systems and security of information. Department experts are members of NATO working
groups in the field of information systems, distance education and computer modelling. The
Department of CIS is also involved in the educational project Cisco Networking Academy
(CNA) for network specialists. Establishing a Cisco regional training centre at the UoD,
Department of CIS has enabled to furnish department laboratories with needful technical
equipment. The project facilitates data communication topics to be incorporated into regular
education. This education has a practical orientation and, therefore, it enhances theoretical
knowledge.
The CIS Department takes care of the continuous development in communication technology
(CT) and information technology (IT) for all students with regard to the requests of Network
Enabled Capability (NEC) which includes an integrated network interface and security of
troops on a battlefield.
Graduates of CIS Department are posted both in the field of CIS administration of the MoD
and permanent networks on the countrys territory as well as in foreign missions. It is of vital
importance to prepare them so that they are able to tackle challenging situations
independently and improvise if necessary.
776
AFASES - 2008 -
the course is possible, too. Selected methods and tools mentioned later, could help to
modernize contemporary teaching and learning processes at the military university
environment.
2.1 The individual study agreement
The author of this contribution met the individual study agreement in the Net-Trainers
course and found it to be a very good tool for the motivation support. This tool can
especially be used in lifelong learning processes. There is a necessity to pay a good
deal of attention to the appropriate content of the individual study agreement between
every individual student and his/her teacher (tutor). The individual study agreement
should be formulated with respect to the special profile of the student and based on the
information gained from the questionnaire survey before starting of the teaching and
learning processes. This document should set the rights and duties of all education
participants. The strong stress should be put on communication rules. When the
education is provided in the distance form the introductory face to face session is the
best opportunity for the creation of the individual study agreement. It is possible to say
that in the Net-Trainers course the individual study agreement helped many
participants to pass the course successfully. The full utilization of this tool potencial
depends on the social skills of the tutor.
2.2 Working group creation and usage
Nowadays meeting the demands of military profession depends on close cooperation
of individuals, on their capabilities to work together, and their ability to create useful
working groups. The Net-Trainers course participants were also divided into working
groups. There were five working groups per 7 people created since the beginning of
the course. The most of participants were from the university environment but some
members from the state administration and economic sector were also included. The
age distribution was from the age of 24 to the age of 65. There were various kinds of
university degrees from professors to Ph.D. students of the Czech universities.
The Net-Trainers course used the concept of the large diversity of personalities in each
working group. This methodology was evaluated by the course participants as the best
and it can be very inspiring especially for the area of lifelong learning. The working
groups of participants of various gender, age, qualifications, skills, and interests were
very innovative during tasks solution and this groups structure brought the best results.
It is obvious that team-work skills should be developed, it has to be an important goal
of education. Besides fixed working groups there can also exist temporary working
groups according to the needs of study activities. Temporary working groups can be
limited by time or by concrete tasks fulfilment.
2.3 On-line study materials
The technology base of the Net-Trainers course was the Learning Management
System (LMS) AllWebSolution. Every course participant had online study materials
continually at his/her disposition. Study materials were in the Czech language but the
course participants could also use alternative study materials in all other languages of
777
AFASES - 2008 -
the course. The course consisted of five modules and there was given an opportunity
for limited number of the Czech participants to study the third module in other course
language in a foreign study group.
From the authors point of view the Net-Trainers course study materials prepared in a
few alternative languages can serve as a pattern for the other education materials
production. In this case a student can learn the content and a foreign language
together. This solution has an importance for the Europe integration and the student
can compare various national approaches to the same topic.
The Net-Trainers course showed convincingly that online study materials should be
brief, apposite, links to other sources have to be included. Other sources of
information should be divided in compulsory and optional (recommended sources).
The very important fact is that online study materials can be updated easily, colours
and multimedia can help their understanding.
2.4 Study activities planning and fulfilment
Every Net-Trainers course participant had his/her own study plan for each module of
the course. Every study activity had its date of fulfilment, its deadline. The beginning
and the end of the module was set by the tutor but timeline of study activities of every
student depended on his/her real self-evaluation, his/her contemporary level of
knowledge, time possibilities and study activities demandingness. Students were
forced to overcome difficulties in team work planning and team work outputs
production.
This concept of study activities planning seems to be the useful tool for the adult
education. It is possible to expect the higher implementation of online studies in the
future. Especially group study activities fulfilment can form special and social
competencies and skills of students in formal and also in informal manner. It was
shown that individual and team study activities should be implemented in every
education smartly.
2.5 On-line study tutoring
One of the qualities of the Net-Trainers course was the high level of its tutoring. This
was very well done by the tutor from the University of West Bohemia in Pilsen. The
course participant had seven months for the appreciation of this fact and during this
rather long time period they could get a general idea of the demandingness of the good
tutor work. Time demands, special and social skills requirements cannot be met by
everyone. The tutor can be a decisive factor of motivation, he/she has to be capable to
solve possible study problems of students sensitively and he/she has a dominant role
in ensuring fruitfulness of online education.
2.6 Tools for communication support during studies
There were implemented three hierarchical levels of the asynchronous communication
in the Net-Trainers course. They were: the international discussion forum, the national
Czech discussion forum and the special Czech discussion forum for every Czech
778
AFASES - 2008 -
working group. The course participants could also use a synchronous communication
tool. The LMS AllWebSolution provided the chat. All communication among the
course participant was controlled by the means of server located in Greece.
Three levels of asynchronous communication seems to be an optimal solution for
multinational online education. It would be useful to implement: international
discussion forum for common topics discussion in English or other suitable world
language, national forums for communication in national languages of every
participating state and working groups discussion forums for every national working
group in national language. Chat at international level can sometimes bring some
problems because of various students different language competencies. It needs some
experience to use chat in foreign language and write briefly and clear under time
pressure.
E-mail communication need not be implemented in LMS, every student can use e-mail
by his/her option. E-mail communication is the optimal tool for sending the results of
study activities to tutor and e-mail communication is fully acceptable for all
contemporary students.
A very important communication channel is a voice communication. Students can use
phone, mobile phone or computer network. It is possible to recommend the use of the
Skype software, it is free of charge. In the case of the appropriate equipment
ownership, videoconference can also be a suitable solution.
2.7 The portfolio as a study activities fulfilment review
Every module of the Net-Trainers course required a students portfolio creation at its
end. After the last study activity of the concrete module every student had to create a
review of his/her work. Portfolio contains the results of all study activities of the
student. Some of them were individual activities, some of them were group activities.
Student had to select his/her own part of outputs of the group activity and to add it to
his/her portfolio. The final file was sent to tutor.
In the university environment the portfolio can be used as a useful tool. It can serve in
the process of assessment of students work and especially in the case of great number
of students it can help the teachers in their decision process about credits and exams
evaluation.
III. Conclusion
The Net-Trainers course belongs to a very good European educational technology.
The text of the paper describes seven of the most interesting methods and tools of the course.
From the authors point of view they can be very usefully implemented at the military
university environment.
779
AFASES - 2008 -
References
[1] Web pages of the University of Defence in Brno. [online]. [cit. 2. 4. 2008]. URL:
http://www.unob.cz/en/.
[2] HRUB, Miroslav, COUFALKOV, Aneta, SMEJKALOV MAZLKOV,
Markta. European On-line Courses - a Step to Closer Co-operation in Education. In
Proceedings 7th International Conference Virtual University VU06. Bratislava : E-Academia
Slovaca, 2006. p. 109-111. ISBN 80-227-2542-0.
[3] VEJVODOV, Jana. Developing European Online Courses. In Online Educa Berlin
2007. 13th International Conference on Technology Supported Learning &Training. Berlin :
ICWE, 2007. s. 214-217. ISBN 3-9810562-7-2.
[4] Web pages of European NetTrainers Association (ENTA) about the Net-Trainers course.
[online]. [cit. 2. 4. 2008]. URL: http://www.nettrainers.org/en/index.asp?p=1-1.
[5] Web pages of Lifelong learning Institution in Pilsen about Net-Trainers course. [online].
[cit. 2. 4. 2008]. URL: http://www.ucv.zcu.cz/kurz.php?num=130&kurz=1.
Ing. Miroslav Hrub, CSc., University of Defence, Faculty of Military Technology, Department of
Communication and Information Systems (K-209), Kounicova 65, 61200 Brno, the Czech Republic, tel. 00420
973 443 646, e-mail: miroslav.hruby@unob.cz
780
AFASES - 2008 -
781
AFASES - 2008 -
782
AFASES - 2008 -
la clasa a IV-a. Aceleai elemente grafice de interfa pot fi folosite n diferite momente ale
leciei, n funcie de obiectivul urmrit.
Astfel, elementul grafic din Fig. 2 poate fi folosit la momentul
comunicrii noilor cunotine, cnd elevul este pus n situaia de
nvare n care i se sugereaz s apropie pointer-ul mouse-ului de
imagine. Pointer-ul se transform din forma normal n forma
link, iar sub pointer-ul mouse-ului se deschide o fereastr de afiare
n care apare cuvntul Rabbit. Noua form a pointer-ului mouseului sugereaz elevului s dea click pe imagine i acesta aude n
Fig. 2. Rabbit
difuzor pronunia cuvntului. Obiectivul operaional vizat este cel al
nvrii pronuniei i a transcrierii fonetice a cuvntului. Se cuvine a
fi remarcat faptul c n momentul n care pointer-ul mouse-ului se transform n forma link,
nu mai este necesar ca elevului s i se precizeze sarcina de a da click pe imagine, aceasta este
sugerat de forma pointer-ului i indus de obinuin.
Acelai element grafic poate fi folosit n momentul verificrii
gradului de asimilare a noilor cunotine (fig. 3), cnd elevul este pus
n situaia de nvare n care i se sugereaz s dea click pe imagine.
Acesta aude n difuzor pronunia cuvntului Rabbit i sub imagine
se deschide o fereastr de dialog n care i se cere s introduc
transcrierea fonetic a cuvntului respectiv. Obiectivul operaional
vizat este cel al realizrii feed-back-ului.
Combinarea elementelor de grafic i sunet pot facilita
nvarea n situaii specifice atingerii unor obiective cum ar fi
formarea abilitilor de parcurgere a unui text ntr-o limb strin sau
chiar n limba romn, de pronunie sau retroversiune. Astfel,
Fig. 3. Rabbit
sincronizarea redrii nregistrrii sonore a unui text cu parcurgerea
acestuia n maniera karaoke conduce la formarea deprinderilor de citire, pronunie sau
traducere a unui text.
Datorit faptului c elevii din clasele a III-a pn la a VI-a, pot ntmpina dificulti la
citire, aceast tehnic este recomandat n proiectarea programelor de nvare pentru
discipline din planul de nvmnt, la care prezentarea coninutului tiinific se face n mare
parte sub form de text i n special la care apar numeroi termeni noi necunoscui elevului,
dificil de pronunat[6].
Tehnic privind, critica cea mai vehement se ndreapt asupra calitii de
interactivitate a programelor de nvare. Interactivitatea vizeaz n mod direct calitatea feedback-ului secvenial, ameliorativ, asigurarea reteniei, asigurarea transferului, concretizarea
particularitilor populaiei int, realizarea evalurii continue, formative i a evalurii
sumative.
Soft-urile interactive de comunicare de noi cunotine ofer o gam larg de tehnici i
procedee didactice de realizare a feed-back-ului secvenial, pornind de la utilizarea unor
elemente de legtur (link) sau declanare, marcate ca text personalizat, astfel nct, un dublu
click cu mouse-ul pe aceste cuvinte s produc deschiderea unei ferestre de dialog i
continund cu elementele de interogare care implementeaz instrumentele de evaluare
alternative. Utilizarea acestora se face ntr-un scop bine conturat, precis; cel al ameliorrii
calitii procesului instructiv. Evaluarea formativ joac aici cel mai concret rol de mijloc de
nvare, recurgnd la o serie de alte mijloace didactice care urmresc ntrirea reteniei,
consolidarea cunotinelor i asigurarea transferului.
De o deosebit importan este modalitatea de reprezentare a itemilor de evaluare
utilizai n realizarea feed-back-ului sau n constituirea testelor docimologice pentru evaluarea
formativ. Acetia trebuie realizai utilizndu-se elementele grafice de interfa ale sistemului
783
AFASES - 2008 -
784
AFASES - 2008 -
elevul i organizeaz ideile pentru a i le exprima coerent, este destul de dificil de elaborat un
numr suficient de mare de distractori (rspunsuri aparent corecte) plauzibili.
Situaii n care pot fi utilizai permit msurarea rezultatelor nvrii, a
cunotinelor asimilate, cum ar fi: verificarea cunoaterii terminologiei, a unor principii,
cunoaterea unor metode, proceduri, cunoaterea elementelor specifice (cine, ce, cum, cnd?);
dar i msurarea rezultatelor nvrii ca reflectare a nivelului comprehensiv i aplicativ:
capacitatea de a identifica aplicaii ale unor rezultate teoretice, capacitatea de a interpreta
relaia cauz-efect, capacitatea de a argumenta metode i proceduri.
- itemi cu rspuns scurt (de completare) (fig.
7), care solicit elaborarea unui rspuns n scris, prin
actualizarea integral a rezultatului de nvare
urmrit de profesor. Pot fi avute n vedere
urmtoarele avantaje: se construiesc uor, sunt uor
de cuantificat, premisa poate utiliza diferite forme de
comunicare (text, scheme, diagrame, desene,
grafice), se poate elabora un model complet al
rspunsului corect, permite abordarea, ntr-un
interval scurt de timp, a unui numr mare de
obiective. Ca dezavantaje remarcm: capacitate
redus de abordare a unor nivele superioare ale
domeniului cognitiv, formularea rspunsului corect
poate fi afectat de capaciti i cunotine care nu
sunt direct implicate n obiectivele evaluate
(capacitatea de exprimare n scris - ortografie,
Fig. 7. Rspuns scurt
capacitatea de utilizare a unor simboluri), furnizarea
de rspunsuri cu grade diferite de corectitudine.
Principalele situaii de utilizare sunt: cunoaterea terminologiei, a unor fapte
specifice, a unor principii, a unor metode sau proceduri, interpretarea unor date, aplicarea
unor cunotine, capacitatea de a opera cu simboluri, formule i convenii standardizate[5].
- ntrebrile structurate, sunt un set de ntrebri care se succed logic i se leag ntre
ele i care solicit elaborarea unui rspuns n legtur cu un coninut tiinific prezentat. Pot fi
avute n vedere urmtoarele avantaje: testeaz cunotine, priceperi, capaciti la nivele
superioare ale domeniului cognitiv, permit utilizarea unor materiale auxiliare, transform un
item subiectiv, de tip eseu, ntr-o suit de itemi obiectivi sau semiobiectivi sau minieseuri,
mai uor de notat. Ca dezavantaje: materialele auxiliare sunt dificil i laborios de proiectat,
exist riscul condiionrii unei ntrebri de rspunsul celei sau celor anterioare. Situaii de
utilizare recomandate: verificarea cunoaterii unor definiii, proprieti, principii, evaluarea
capacitii de a observa fenomene, a utiliza cunotine, a efectua experimente, a manevra
informaii; evaluarea capacitii de a raiona i a relaiona elemente[4].
Exemplu.
Obiectiv. Msurarea capacitii de a opera cu variabile de tip ntreg.
Enun. Citete cu atenie ntrebrile care urmeaz. Nu te grbi cu rspunsurile! Folosete
tabele de verificare pentru analiza secvenelor de program. Fi atent la iniializarea
variabilelor!
Premis.
1. Care sunt tipurile de date ntregi folosite n limbajul Pascal?
2. Precizai numrul de octei de memorie ocupat i determinai intervalele de valori pentru
fiecare tip de date ntreg.
3. Cum se reprezint n memorie date de tip ntreg?
4. Cum se determin ctul i restul mpririi ntregi a dou numere naturale?
785
AFASES - 2008 -
Bibliografie
[1] V. Ciubotaru - Ce este un scenariu, www.poezie.ro/index.php/article/
[2] Giordan, A., Apprendre, Paris, Ed. Belin, 1998
[3] I. Maxim, nstruire asistat de calculator, Suceava, Universitatea tefan cel Mare, Note
de curs, 2007, http://dppd.usv.ro/cursuri/cursuri.html
[4] I. Maxim, C. Moroanu, Informatic Didactica specialitii, Iai, Editura Universitii
Al. I. Cuza, 2007
[5] I. Maxim, S. Buduanu, V.V. Maxim, Educaie tehnologic, Cluj-Napoca, Editura
InfoData, 2007
[6] C. Masalagiu, I. Maxim, I. Asiminoaie, Metodica predrii informaticii, Bucureti, Editura
MatrixRom, 2001
[7] Siebert, H., Pedagogie constructivist, Iai, Institutul European, 2001,
[8] Sorel, M., Pratiques nouvelles en ducation et en formation. Lducabilit cognitive,
Paris, LHarmattan, 1994
[9] Stanciu M., Didactica postmodern, Editura Universitii Suceava, 2003
Lector univ. dr. Departamentul pentru Pregtirea Personalului Didactic, Universitatea tefan cel Mare
Suceava, Universitii, 13, 720225, e-mail: maximioan@yahoo.com
Lector univ. drd. Facultatea de tiine Economice i Administraie Public, Universitatea tefan cel Mare,
Suceava, Universitii, 13, 720225, e-mail: socaciu@seap.usv.ro
786
AFASES - 2008 -
Abstract:
The goal of this paper is to show possibilities of on-line course at Communication and
Information Systems Department.
E-learning is a new and very popular form of education at universities. It is very
suitable for modules, where are rapid progressions and we need to change studying materials
very often. At the beginning, there is a big problem with knowledge from e-learning, tutoring
and on-line education field. We need someone, who is able to create and tutor on-line course.
It is not only about creation and tutor the course, it is necessary to find the right form
of e-learning education.
How to get knowledge about e-learning as a support of education? I passed
the European Net-Trainers [1] on-line course in 2006. This course gave me a lot
of information about e-learning and how to prepare and tutor an e-learning course by the
right form.
On-line course Moodlinka was the second experiences how to prepare on-line
course and lots of aspects of education by this form of education. Moodlinka (Moodle-link)
is LMS (Learning Management System) at Masaryk University in Brno and the tried
to explained principles of using this LMS and its advantages. Moodle is the LMS which was
installed for support the education at our department.
This paper deals with e-learning utilization for education supporting
at communication and information systems department. It involves experiences and effects
of pilot course Net-Trainers and few words about preparing course Optical Communication
Systems and Support of Software for Communication Technologies. There are mentioned
the differences between classic type of education and e-learning. It is concentrated
on positives and negatives in e-learning utilization as a support of education and e-learning
utilization at University of Defence. The knowledge about how to create and how to tutor
on-line course went from Net-Trainers on-line course.
787
AFASES - 2008 -
788
AFASES - 2008 -
789
AFASES - 2008 -
(4.1)
(4.2)
Now consider generalization to the case of a discrete function,
by letting
, where
, with
, ...,
. Writing this out gives the discrete Fourier
transform
as
(4.3)
is then
(4.4)
Discrete Fourier transforms are extremely useful because they reveal periodicities
in input data as well as the relative strengths of any periodic components. There are a few
subtleties in the interpretation of discrete Fourier transforms, however. In general, the discrete
Fourier transform of a real sequence of numbers will be a sequence of complex numbers
of the same length. In particular, if are real, then
and are related by
(4.5)
for
, 1, ...,
, where denotes the complex conjugate. This means that the component
is always real for real data.
4
3
2
1
0
1.5
2.5
3.5
4.5
5.5
1.5
2.5
3.5
4.5
5.5
4
3
2
1
0
790
AFASES - 2008 -
Fig. 1 : Discrete Fourier transforms using the function of Fast Fourier transform and
matrix transform
This picture is only an example of using figures for demonstration of equation
application. The software MATLAB was used for creating signal process of Discrete Fourier
transform. I use this software for demonstrate time behaviours of the signals. For this reason
I would like to prepare the module Software Support for teaching how helpful this software is.
Chapter V Conclusion
I have passed the on-line course Net-Trainers and it was my firs meeting with on-line
education. It seems to be very useful for our department. The Net-Trainers course can serve
as a sample of modern educational method, shows European educational technology
utilisation in the Czech environment. The project about e-learning module creation has started
in special research. There has been an idea creation e-learning courses. It has been because
we have needed to create new modules from our department.
I would like to finish all materials which are necessary for completion e-learning
courses and I would like to try it next year.
bibliography reference:
[1] Hortov, N., Vov, A., a kol.: Pilotn kurzy, Masarykova univerzita; Brno, 2007
[2] Kopeck, K.: Modern trendy v elektronick komunikaci; Olomouc, 2007
[3] Kopeck, K.: E-learning (nejen) pro pedagogy; Olomouc, 2007
Reference
[1] Net-trainers Course [online]. [cit. 2006-05-13]. URL: http://www.nettrainerstraining.org/en/extranet.asp?p=1-1
[2] Hrub, M.: From the Communication and Information Technology Modules Specification
to the Distance Learning. Simulation and Distance Learning Brno 2007, Univerzita
Obrany.
[3] Smejkalov Mazlkov, M.: Education in Optical communication systems by e-learning.
Virtual University Bratislava 2006.
[4] Jechyle, T.; Vejvodov, J.: European Net-Trainers: A Translational Course Framework.
Online Educa Berlin 2006.
[5] Vejvodov, J.: Metodick pruka pro autory on-line kurz. Zpadoesk univerzita v
Plzni, Plze, 2005.
[6] Vejvodov, J.: Metodick pruka pro tutory on-line kurz. Zpadoesk univerzita v
Plzni, Plze, 2005.
[7] Smejkalov Mazlkov, M.: The analysis and optimization of premises of laser intersatellite communication system, WSEAS Transaction on Communications, 2007.
[8] Wikipedia, [cit. 2008-03-30]. URL: http://en.wikipedia.org/wiki/Fourier_transform
[9] [cit. 2008-03-30]. URL: http://mathworld.wolfram.com/DiscreteFourierTransform.html
*
Marketa Mazalkova, Ing, University of Defence, Communication and information systems department,
marketa.mazalkova@unob.cz
791
AFASES - 2008 -
792
AFASES - 2008 -
from each base, is in the same state of superposition. Conventionally, the binary value 0
corresponds to the states and and the value 1 corresponds to the other two states,
which are called qubits states.
( P2 = b( l 1) / 2+1 b( l 1) / 2+ 2 ... b( l 1) ) . They will repeat the division of the groups each time
when they realize that the parity is different, and each time they will erase the last bit from the
groups whose parity is announced publicly. This way, they will avoid a possible intruder
(Eve), and in the end Alice and Bob will share the same row of bits.
793
AFASES - 2008 -
How to obtain the private key is the most important step of the procedure. The second
step of the procedure is the entanglement of the qubits of the message with the ones of the
key, and the transportation of the message to the receiver. In fact, the qubits of the private key
will have a very important part in this process, because with their help the transportation of
the qubits of the message will be realized from the sender (Alice) to the receiver (Bob). The
second step is as follows: after obtaining the key, the sender (Alice) will divide the message in
sub-messages whose dimensions are equal to that of the private key (number of qubits from
the key = number of the qubits from the sub-message). When the number of the qubits of the
last sub-message is lower than the one of the private key, the qubits of the key will be send to
the receiver un-entangled.
(1.1)
will be obtained.
The next step is to apply the C-NOT gate, and the state of the qubit from the key will change
as follows:
C NOT : 1 / 2 [( 0 1 ) + ( 1 + 0 )] 0 == 1 / 2 ( 00 11 ) + ( 11 + 00 ) =
= 1 / 2 [ ( 00 + 11 ) + ( 00 11 )]
(1.2)
If we consider that:
794
AFASES - 2008 -
1 / 2 ( 00 + 11 ) = 00
1 / 2 ( 00 11 ) = 10
1 / 2 ( 01 + 10 ) = 01
1 / 2 ( 01 10 ) = 11
Bob will receive the qubits and he will "read" them using the Bell bases. Knowing the
private key, he will be able to extract the message with the help of the quantum gates used for
the entanglement of the states (the quantum gates are reversible). If during the reading of the
message Bob will notice the presence of the qubits which cannot be "read" using the Bell
bases, he will think that Alice's message ended, and these qubits are in fact the ones left unentangled of the private key. The communication is considered to be realized through an
isolated quantum channel, and in the absence of an intruder (ideal conditions). When the
communication is realized through a noisy channel, Bob will take into consideration the errors
which can appear during the transmission (decoherence and redundancy).
Bibliography
[1] David P. DiVincenzo, Quantum Gates and Circuits, Proceedings of the Royal Society,
London, 1997.
[2] G. Johansson, Quantum Algorithms-Lectures in Quantum Informatics, applied Quantum
Physics, MC2, Chalmers, S-412 96 Gteburg, Sweden, 2005.
[3] Michael E. Nielsen, Isaac L. Chuang, Quantum Computation and Quantum Information,
ISBN 0-521-63503-9, 2000.
[4] Nikolas P. Papadakos, Quantum Information Theory and Applications to Quantum
Cryptography, University of Athens, 2001.
Gabriela Mogos, Alexandru Ioan Cuza University, Computer Science Department, Iasi, Bd. Carol I no.11,
Tel:0232-201091, gabi.mogos@gmail.com
795
AFASES - 2008 -
QUANTUM ORACLES
Gabriela MOGOS *
Abstract:
The 20th century has taught us that the physical world is characterized by fundamental
quantum processes. Computers are physical objects that have not evolved beyond their ability
to explore classical physical phenomena (even when they do take into account certain
quantum phenomena, these latter's quantum essence is not being explored). Research in
quantum information technology consists of exploring quantum physics' potential, laws and
limits in order to illustrate, analyze and communicate information. Attempted first in
laboratories, quantum IT research has its origin in questions explored by scientists like
Richard P. Feynman, Nobel Prize winner for physics in the early 80s. It was not until the
second half of the 90s that several IT researchers, mostly involved in theoretical IT work
(information theory, complexity, cryptography), reconsidered their research field deciding to
approach the quantum information field. This paper proposes an overview of quantum
oracles.
Key words: Quantum bits, Quantum gates, Quantum algorithms.
2000 MSC: 81-08.
I. Introduction
Quantum algorithms are also called oracles. Oracles are instruments employed to
answer questions by a simple 'yes' or 'no.' Oracles' architecture is very appropriate for
quantum computers in as much as quantum systems are based on probabilities. Therefore, if
one asks a question, the answer will take the shape of a wave function. This study will present
four algorithms (oracles).
796
AFASES - 2008 -
the answer can be found with the help of a single evaluation which means that, from a
quantum perspective, 0 and 1 can be grouped together simultaneously in one x.
Deutsch-Joszas Oracle. This oracle represents a generalization of the Deutsch algorithm,
where the function is f : {0,1}n {0,1} . The question is identical to the previous one: is the f
function constant or balanced? In this case, for the function to be balanced, its value must be
0 in half of the arguments, and 1 in the other half. Of course, in this case, the function cannot
be either constant or balanced. Therefore, the oracle does not apply because it can only
answer with 'yes' or 'no' which renders the answer meaningless.
Bernstein -Vazirani's Oracle. Let there be a function f : {0,1}n {0,1} where f ( x) = a x ,
where a is a vector constant with 0 and 1, and . is the scalar product. How many
measurements are there necessary in order to find a? Classically, we should apply the formula
to all possible arguments and solve the linear equation system for a. From a quantum
perspective, a can be found in one step.
Simon's Oracle. Assuming a function f : {0,1}n {0,1}n . The function is presumed to be 2to-1, that is for any value of f there is always x1 and x 2 such that f ( x1 ) = f ( x 2 ) . The function
is presumed to be periodic, meaning that for a binary vector s we have f (x s) = f (x) . The
algorithm returns the period s in O(n) measurements.
2.1 Deutschs Oracle
(2.1)
H marks the Hadamard gate and U f is a control gate defined by the relation:
U f x y = x y f (x)
797
(2.2)
AFASES - 2008 -
where f : {0,1} {0,1} is either constant in which case f (0) = f (1) or balanced where
f (0) = 1, f (1) = 0 f (0) .
Next we will analyze the impact of the different gates of this circuit on the input state.
The first Hadamard operator on the top entry affects the 0 state and leads to:
H0 =
1 1 1 1
1 1
1 1 0
1
=
=
(0 + 1)
+ =
2 1 1 0
2 1
2 0 1
2
(2.3)
which means:
H
0
(0 + 1)
(2.4)
2
It can be noted that the newly obtained state is a combination of 0 and 1 . Analogously, for
input state 1 we will get:
H1 =
1 1 1 0
1 1
1 1 0
1
=
=
(0 1)
=
2 1 1 1
2 1
2 0 1
2
(2.5)
which means:
H
1
(0 1)
(2.6)
2
According to these results, the first pair of H gates (top entry and bottom left) change the
initial input state 0 1 to state:
1
( 0 + 1 )( 0 1 )
2
Then we apply the controlled U f gate to this state. We will follow the impact of the U f gate
on each of the terms in the equation above.
U f x ( 0 1 ) = (1) f ( x ) x ( 0 1 )
(2.7)
(2.8)
(2.9)
(2.10)
798
AFASES - 2008 -
1
1
1
[(1) f ( 0 ) H 0 + (1) f (1) H 1 ] ( 0 1 ) = [(1) f ( 0)
( 0 + 1 )+
2
2
2
1
1
+ (1) f (1)
( 0 1 ) ( 0 1 ) = [((1) f ( 0 ) + (1) f (1) ) 0 +
2
2
1
+ ((1) f ( 0) (1) f (1) 1 ]
(0 + 1)
2
Based on this equation, the following conclusions can be drawn:
1. If the f (x) function is constant, then we have:
(1) f ( 0 ) (1) f (1) = 0
the term between square parentheses in equation (2.12) becomes:
1
[(1) f ( 0 ) + (1) f (1) ] 0 = 0
2
and (2.13) expresses the final state on the top entry in Figure 1.
(2.12)
(2.13)
(2.15)
where f (x) is the evaluated function of a single qubit. Taking into account the fact that there
are n -qubits entry data, the global problem is if the f (x) function is constant or balanced.
799
AFASES - 2008 -
0 +1
0 +1
0 1
)=
0 1
(2.17)
)
2
2
2
2
2n
x{0 ,1}n
According to the similar step in Deutsch's algorithm, we must evaluate the effect of the U f
)...(
)(
operator:
2 =
(1) f ( x ) x
0 1
(2.18)
)
2
2
Finally, we must evaluate the effect of the last set of Hadamard gates. For a single
qubit, the Hadamard effect can be written as:
(1) xz z
H x =
(2.19)
2
x{0 ,1}
which, by extension to the whole register, becomes:
x1 ,..., x n =
( 1)
x1 z1 +...+ xn z
z1 ,..., z n
(2.20)
2n
z1 ,..., z n
(1) xz + f ( x ) z
2
0 1
2
(2.21)
register.
If the f ( x) function is constant, then the sum above x is independent of x, therefore we will
obtain 2 n , identical simple terms by placing the sign 1 in front of the 0
state.
state, probability 1.
n
positive terms and 2
negative
2
2
terms which cancel each other. So, for a balanced f ( x) function, the studied probability has a
probability zero in all zero states. Thus, it is necessary to measure the register and, if one or
more of the qubits are in state 1, the f ( x ) function is balanced, otherwise the f ( x) function
is constant.
800
AFASES - 2008 -
As a consequence, with the help of the quantum algorithm, we can answer these global
questions (register work) by a single U f , routine run as long as a classical algorithm will
2n
require a number of
+ 1 rounds, since Deutsch-Josza's algorithm is an algorithm of
2
exponential computing.
(1)
a x
(2.22)
(1) x y y
x =0
(0 1)= a
2
and, in this case, the control line measurement returns a.
y
1
2
(0 1)
Let there be the f (x) function f : {0,1}n {0,1}n . There is a sequence s so that
f ( x) = f ( y ) y = x s , for any x, y {0,1} . The problem is finding this s.
In 1993, Simon proposed an algorithm which starting from a x , can calculate x f ( x) in an
n number of steps.
801
AFASES - 2008 -
x{0 ,1}n
f ( x)
x{0 ,1}n
( x + x s ) f ( x)
n/2
1
2
((1)
x y
+ (1) ( x s ) y ) y
y{0 ,1}n
where x y = ( xi y i ) mod 2 .
i
III. Conclusions
The theory of quantum computation allows us to look at the question of consciousness
from a new perspective, since every physical object can be regarded as a quantum computer,
any physical process can be considered a computation and Deutsch's universality result means
it must be possible to simulate conscious rational thought using a quantum computer.
Bibliography
[1] Zdzislaw Meglicki, Three Lectures on Quantum Computing, 2002.
[2] David P. DiVincenzo, Quantum Gates and Circuits, Proceedings of the Royal Society,
London, 1997.
802
AFASES - 2008 -
Mogos, Gabriela, Al.I.Cuza University, Computer Science Department, Iasi, Bd. Carol I no.11, Tel.: 0232210910, gabi.mogos@gmail.com
803
AFASES - 2008 -
Z+T
f (t) g(t)
dt + c(b) K
sup |F (x) G(x)| b
T
t
x
(1)
Aceasta teorema joaca un rol important n studiul ratei de convergenta n teoremele limita.
In lucrarea de fata se regaseste o aplicatie a teoremei lui Esseen pentru cazul fiabilist.
Fie v(0) = 0 si 0,n intervalul de timp de la starea 0 la starea n, n 1, adica
0,n = inf {t : v(t) = n | v(0) = 0} .
(2)
k k
k=0
(3)
n1
, n 1. Relatia (3) a fost demonstrata n lucrarea [8].
unde 0 = 1 si n = 0 1 1 ...
2 ...n
Sunt prezentate n continuare cateva rezultate intermediare foarte importante n demersul
ce va urma.
1
Transilvania University of Brasov, Departament of Mathematical Analysis and Probability, e-mail:
b.munteanu@unitbv.ro
804
AFASES - 2008 -
1
,
n (z)
(4)
(5)
(6)
Remarca 2. ([5])
n,2 =
n1 Pk
X
k=1
E[0,s ]s
.
k k
s=1
(7)
In lucrarea [9] s-a tratat exemplul ce va urma, unde s-a urmarit calculul fiabilitatii unui
sistem cu doua componente (stari). Acum pentru acelasi sistem, cu aceleasi caracteristici,
vom studia abaterea distributiei timpului de viata al acestuia fata de distributia exponentiala.
Sistemul are doua unitati operationale cu ratele (1) = (2) = 1, o unitate n standby-ul rece
(3) = 0.5 si o componenta n stanby-ul cald (4) = 0.2. Exista un singur canal de reparatie
cu rata de reparatie 1 = 10. Sistemul se degradeaza daca numarul total de componente defecte
este doi. In aceste conditii, ratele ascendente de tranzitie ale sistemului sunt
0 = (1) + (2) + (3) + (4) = 2.7 ,
1 = (1) + (2) + (3) = 2.5 ,
iar conform relatiei (7),
2,1 = E[0,2 ] =
2,2 =
cu 0 = 1 1 =
0
1
0 + 1
0
= 2.112
+
1 1
0 0
1
1 E[0,1 ]
= 0.148
=
0 1
1 1
= 0.1.
805
AFASES - 2008 -
astfel nc
at supx G (x) K. Atunci pentru orice T > 0 si b > 1/(2) avem:
sup
0<x<ln
unde I =
1
K
(8)
+T
R
0.148t+1.112i
si c(b) este o constant
a pozitiv
a ce depinde de b.
(1it)(0.148t
2 +2.112it1) dt
+T
R
0.148t+2.112i
brul drept al inegalitatii (1) este I =
(1it)(0.148t2 +1.112i1) dt.
T
Rezolvarea integralei se bazeaza pe tehnicile bine cunoscute de integrare a unei functii irationale
prin procedeul de descompunere a unei fractii n fractii simple.
806
AFASES - 2008 -
Referinta
[1] DEKKING, F.M., A Modern Introduction to Probability and Statistics, Springer, 2005
[2] ESSEEN, C.G., Fourier analysis of distribution functions. A mathematical study of the
Laplace-Gaussian law, Acta. Math., 77, 1-125, 1945.
[3] ESSEEN, C.G., On the Liapunoff limit of error in the theory of probability, Arkiv Mat., och
Fysik, 28A(2), 1-19, 1942.
[4] FELLER, W., On the Berry-Esseen theorem, Z.Wahrsch. Verw. Geb.,10,261-8, 1976.
[5] GNEDENKO, B.V. and SOLOVYEV, A.D., A general model for standby with renewal,
Engineering Cybernetics, 12(6)(1974), 82-86.
[6] KATZ, M., A note on the Berry-Esseen theorem, Ann. Math. Statist., 34, 1007-8, 1963.
[7] MUNTEANU, B. GH., On a Generalization of Esseens inequality, Italian Journal of Pure
and Applied, 27(2007), in press.
[8] MUNTEANU, B. GH., Deviation from exponentiality in birth and death process of ergodic
exit and sojourn times, A 10-A Conferinta Anuala a Societatii de Stiinte Matematice din
Romania, 2006, in press.
[9] MUNTEANU, B. GH., On numerical example of a reliability of system with renewable components and fast repair, Sesiunea de Comunicari Stiintifice cu Participare Internationala
Educatie si Cercetare Stiintifica la standarde comunitare, Academia Fortelor Aeriene
Henri Coanda Brasov, Proceedings Education and Scientific Research European Standards, vol.2(2007), p.482-491, (pe CD, ISBN: 978-973-8415-45-4).
[10] PAULAUSKAS, V.I., On a smoothing inequality, Litovsk. Nat. Sb., 11, 861-6, 1971.
[11] PETROV, V.V., Limit Theorems of Probability Theory :Sequence of Independent Random
Variables, Oxford Science Publication, Oxford, 1995.
807
AFASES - 2008 -
(1)
=
+ D k + k + R k
q k
dt q& k q k
unde T =
1 n n
a kjq& k q& j este energia cinetica, U(q) este energia potenial corespunzatoare
2 k =1j =1
U
( dac exista fore Fe cu rotFe = 0 Fe = gradU ), Dk
q k
sunt fore disipative corespunzatoare forelor de frecare cu lucru mecanic negativ, k sunt
fore generalizate giroscopice cu lucru mecanic nul si R k sunt fore neconservative(
provenite din forte Fk cu rotFk 0 ) depinznd de coordonatele poziionale.
Ecuaiile (1) se scriu matricial:
A&q& + Cq& + Kq + Gq& + Pq = Q(q, q& )
(2)
unde A este matricea inerial, C, matricea de amortizare( corespunztoare forelor disipative,
cu componente simetrice c ij = c ji ), K, matricea forelor elastice(poteniale), G, matricea
forelor generalizate Q k =
808
AFASES - 2008 -
( )
~
q=~
q t, t 0 , ~
q0 , ~
q& 0 , ~
q& = ~
q& t , t 0 , ~
q0 , ~
q& 0 . Dac regimul perturbat, ~
q, ~
q& , ramne in vecintatea
regimului de control, (q c , q& c ) , spunem, dup Liapunov, c regimul de control este simplu
(q c , q& c )
spunem ca avem stabilitate asimptotic, (S.A). Se observ c un system (S.A) este (S.S).
Pentru sistemele liniare sunt cunoscute criteriile de stabilitate Hurwitz-Routh iar
pentru cele neliniare criterii in prima aproximaie utiliznd stabilitatea sistemului liniarizat si
deasemeni metoda funciei lui Liapunov.
Considerm sistemul multidimensional (2) cu membrul stng liniar i matricea A
nesingular . Sistemul liniar i omogen adus la forma normal va fi:
&q& + Cq& + Kq + Gq& + Pq = 0
cu polinomul caracteristic corespunztor
(4)
(5)
Condiia de stabilitate asimptotic pentru soluia nul q=0 a sistemului (4) este ca
rdcinile polinomului characteristic s aib partea real negativ. Acest lucru are loc dac
minorii diagonali ai determinantului Hurwitz corespunztor sunt pozitivi,
A1 1 0 0..............0
A 3 A 2 A1 1 0.........0
A
H = A 5 A 4 A 3 A 2 A11.....0 , H1 = A1 , H 2 = 1
A3
..................................
1
....., H 2s = H
A2
0 0...................0 A 2s
(6)
r
Forele care acioneaz in sistemul (6) sunt fora de amortizare, C(c1x& , c 2 y& ) , fora
r
r
elastic conservativ, K (k1x, k 2 y ) ,fora giroscopic, G (g1 y& , g 2 x& ) i fora neconservativ
r
P(p1 y, p 2 x ) .
Polinomul caracteristic va fi:
809
AFASES - 2008 -
P( ) =
2 + c1 + k1 g1 + p1
2 + c 2 + k 2
g 2 + p2
(7 )
4 + 2c3 + c 2 + g 2 + 2k 2 + 2(ck + gp ) + k 2 + p 2 = 0
(8)
)(
p p
k + g 0 . Inecuaia este satisfcut dac (p c )1 (p c ) (p c )2 , unde (p c )1 , (p c )2
c c
referin triortogonal cele dou rdcini definesc dou suprafee, domeniul cuprins intre ele
fiind domeniu de stabilitate.
Pentru sistemul (6) vom enuna o serie de teoreme Thomson-Theta-Cetaev privind
r r r r
structura forelor C, K, G, P ce vor influena stabilitatea soluiei nule.
Teorema1.
a) In cazul sistemului conservativ, dac energia potenial are un minim izolat, atunci
sistemul este S.S in jurul punctului de minim.
b) Dac sistemul conservativ are in jurul punctului de minim izolat regim S.S i dac
exist coordonate ciclice atunci stabilitatea simpl se conserv in raport cu
coordonatele poziionale i micarea va fi uniform in raport cu coordonatele ciclice.
Teorema 2.
Dac avem un regim izolat potenial S.S atunci prin ataarea la sistem a forelor
r
r
r
C (disipativ) i G se pstreaz stabilitatea simpl; dac disiparea este total ( C ) atunci
r r r
sistemul K, C, G devine S.A(asimptotic stabil).
Teorema 3.
r
Intr-un regim potenial K , instabil in vecintatea unui punct de maxim pentru energia
potenial i unde energia potenial are i valori negative, dac se aplic forele disipative
r r
C, G instabilitatea se pstreaz.
Teorema 4.
r
Sistemul in care avem fore giroscopice G , intotdeauna avem stabilitatea vitezei (S.S).
r
Dac adugm i forele K sistemul devine asimptotic stabil in raport cu viteza i simplu
stabil in raport cu coordonatele.
Teorema 5.
r
Un sistem in care avem numai forele nepoteniale P este intotdeauna instabil. Un
astfel de sistem este reprezentat de ecuaiile:
&x& + py = 0
&y& px = 0
810
AFASES - 2008 -
k1 + k 2 f 0, k1k 2 + p 2 f 0, k1 k 2 f 2p . Intr-un
C2
C1 C2=-p2
O
C1
C1= - C2
C1
C1 - C2=2p
a.)
b.)
C2, k2
k1= k2=k
a.) p= 0
c1 c2=p2/k
C1, k1
c.)
Fig.1 Domenii de stabilitate
b.) p 0 , k1=k2=0
811
c.) k1=k2=k
AFASES - 2008 -
r
K
stabil i coeficienii
r
k1 = k 2 = L = k s = 2 atunci, prin adugarea forelor nepoteniale P , stabilitatea va fi
perturbat indiferent de termenii neliniari (X,Y).
Exemplificm pentru s=2. Sistemul de ecuaii corespunztor este:
&x& + 2 x + py = X
consider
sistemul
conservativ
&y& + 2 y px = Y
2
Ecuaia caracteristic este 4 + 2 2 2 + 4 + p 2 = 2 + 2 + p 2 = 0 . Soluiile vor
fi:
p 2 + 4 + 2
p 2 + 4 2
=
i
2
2
812
AFASES - 2008 -
fig.2
R+C
dS
R
x
O1
ptratul vitezei relative, (R v )2 i densitatea fluidului. Aceast for este tangent la rotor.
Expresia ei este:
r
C
d
dS = (R v )2 Rd = R 3 2 1
(9)
2(C + x cos )
Proiectm (9) pe axele x i y i integrm intre 0, 2 . Dezvoltm in serie expresia din
parantez:
2
1
1
x
x
1
= 1 + cos cos 2 + ...
2(1 + x C cos ) 2 C
C
813
AFASES - 2008 -
x
R 3 2 2
R 3 2
Sy =
1
cos
...
cos
d
+
x + ...
4
C
2C
0
In mod asemntor se obine componenta dup axa x:
(10)
R 3 2
y + ...
(11)
2C
r
Asupra rotorului considerm c acioneaz i fora elastic F cu componentele Fx=-cx
i Fy=-cy, unde c este coeficientul de rigiditate la incovoiere.
Ecuaiile de echilibru sunt ecuaiile lui Newton (ecuaiile de micare ale centrului
maselor):
M&x& = Fx + S y + X
(12)
M&y& = Fy + S y + Y
Sx =
unde M este masa rotorului, X i Y sunt componentele de ordin superior neglijate. In (12)
imprim cu M i notm:
3 2
c
2 R
=k ,
=p
M
2CM
Ecuaiile de micare devin:
&x& + k 2 x py = X
&y& + k 2 y + px = Y
Sistemul obinut este asemntor cu cel de la teorema 8 din paragraful precedent i
putem aplica concluzia respectiv, sistemul este instabil. Pentru a evidenia natura
instabilitii considerm sistemul omogen i introducem variabila complex z = x + iy . Se
obine:
&z& + k 2 + ip z = 0
(13)
locul
(15)
pt
2
Al doilea termen din (15) tinde la zero cnd t tinde la infinit i z = C1e k e ikt . Deci,
pt
geometric descris de punctul O1 este o spiral logaritmic cu r = C1e 2k i = kt .
814
AFASES - 2008 -
Bibliografie
1. S.Chiriacescu: Sisteme mecanice liniare. Editura Academiei Romane, Bucuresti, 2007
2. C.I.Ghiorghiu: Metode numerice pentru sisteme dinamice. Casa Crii de Stiin, Cluj
Napoca, 2006
3. D.R.Merkin : Introducere n teoria stabilitii micrii. Moscova, Nauka, 1987
4. M.Lupu, O.Florea: Studiul transferului de mas n cazul micrii staionare a fluidelor
viscoase intre cilindrii cu axe paralele. Proceedings of International Conference,
Academia Forelor Aeriene Henri Coand, Braov, CDROM, 2007
5. S.Nedelcu, A.Postelnicu: Rezolvarea problemei neliniare a unui lagr circular hibrid.
International Journal of Applied Mechanics and Engineering,University Press Zielona
Gora, Poland,vol.7,nr.2, 2002
**
815
AFASES - 2008 -
AN INTRODUCTION IN THE ASYMPTOTIC STUDY OF NONLINEAR STOCHASTIC DIFFERENTIAL EQUATIONS IN THE FIELD
OF OSCILLATIONS
Gabriel NEPOTU*
Abstract: Starting from the uni-dimensional It formula, certain aspects of a second order
equation including a random phenomenon called white noise are presented. The calculus
will make use of the amplitude and phase of the oscillations, as well as a small parameter.
Key words: It formula, white noise, oscillations, amplitude, phase.
Consider the equation:
d 2x
2
+
x
=
f
x
x
+
f
x
x
(1)
(
,
)
(
,
)Wt
1
2
dt 2
where f1 and f2 are non-linear functions, is a small positive parameter, and Wt is a random
process of the white noise type.
Equation (1) can be written as a stochastic system,
dx
xdt
=
(2)
d x = x + f1 ( x, x) dt + f 2 ( x, x)dBt
x = A sin , where = t +
x = Acos ,
Observation: A is the oscillation amplitude and its phase.
Considering x= Acos(t+), its derivative is
(3)
x = A cos(t+)-Asin(t+)
Due to the fact that the oscillations are small, A may be considered 0.
Equations (3) results in
2
x
A= x + 2
= arctg
x
Here follows the calculation procedure used in the paper.
2
816
(4)
AFASES - 2008 -
d = K 2 ( A, )dt + M 2 ( A, )dBt
Using the previous theorem, as well as the following calculations, the expressions of the
functions Ki and Mi , i=1,2 are found.
2
A = x2 +
dA
dx
2 2 x2 + x
x
d
1
= arctg
=
x
dx
2 x
1
2
1+
1 A sin
sin
2x2
1 2x2
1
cos
2 2 =
A cos =
2
x A
A
A
d2A
2
dx
d 2
2
dx
2 x2 x
= A
= A
( 2 x 2 + x ) 2
2x
= x
( 2 x 2 + x ) 2
f 2 ( A cos , A sin )
2A
1
M 1 ( A, ) = f 2 ( A cos , A sin ) sin
K 1 ( A, ) =
1
M 2 ( A, ) =
f 2 ( A cos , A sin ) cos
A
In system (5), the change of variable is performed:
A = A + u ( A, )
817
(6)
AFASES - 2008 -
= + v( A, )
where u ( A, ), v( A, ) are represented using the small parameter :
u ( A, ) = u1 ( A, ) + u 2 ( A, ) + 2 u 3 ( A, ) + ...
(7)
v( A, ) = v1 ( A, ) + v 2 ( A, ) + 2 v3 ( A, ) + ...
Functions u and v are so chosen that the deterministic parts of the equation system (5)
in the new variables A, should not contain vibrating terms. Therefore, the new variables
after the change (6) have to verify the new equations:
d A = C ( A)dt + B( A, )dBt
(8)
d = D( A)dt + E ( A, )dBt
where C, B, D, E are to be found as series such as in (9):
C ( A) = C1 ( A) + C 2 ( A) + 2 C 3 ( A) + ...
(9)
D( A) = D1 ( A) + D2 ( A) + 2 D3 ( A) + ...
B ( A, ) = B1 ( A, ) + B2 ( A, ) + 2 B3 ( A, ) + ...
E ( A, ) = E1 ( A, ) + E 2 ( A, ) + 2 E3 ( A, ) + ...
The
problem
is
solved
when
the
coefficients
u n ( A, ), v n ( A, ), C n ( A), Dn ( A), Bn ( A, ), E n ( A, ) where n=1,2,... , are known.
An example of equation (1) is the equation describing the motion of a pendulum with small
random disturbances in its environment.
x (t ) + (1 + Wt ) x = 0 , with x(0) and x(0) given, R+ .
The stochastic system is
dx = xdt
d x = ( x)dt xdBt
x(t) solves the stochastic Volterra equation of the form:
t
x(t ) = x (0) + x (0) t + a (t , s )x( s )ds + b(t , s ) x( s )dB s where a(t,s)= s t , b(t,s)= ( s t )
References
1. R.Voinea, F. Stroie, Introducere n teoria sistemelor dinamice, Ed. Academiei, Bucure ti,
2000.
2. E. Deciu, M. Bugaru, C. Dragomirescu, Vibra ii neliniare cu aplica ii n ingineria
mecanic, Ed. Academiei, Bucure ti, 2002.
818
AFASES - 2008 -
OPTIMIZING QUERIES
Maria-Carmen PNGLEANU *
Abstract:
The performance of applications depends on the database's design and the queries
written for it. This paper describes goals for tuning, how to identify high-resource SQL
statements, explains what should be collected, and provides tuning suggestions.
The tuning of SQL statements is very important for the database system performance
tuning. SQL tuning involves three basic steps:
-Identifying high load or top SQL statements that are responsible for a large share of
the application workload and system resources, by reviewing past SQL execution history
available in the system.
-Verifying that the execution plans produced by the query optimizer for these
statements perform reasonably.
-Implementing corrective actions to generate better execution plans for poorly
performing SQL statements.
These three steps are repeated until the system performance reaches a satisfactory
level or no more statements can be tuned.
This paper is a study case which provides information for understanding and
managing SQL statements to optimal performance and contents examples with performing
SQL statements.
819
AFASES - 2008 -
820
AFASES - 2008 -
To improve SQL efficiency, use equijoins whenever possible. Statements that perform
equijoins on untransformed column values are the easiest to tune.
Avoid Transformed Columns in the WHERE Clause
Use untransformed column values. For example, use:
WHERE a.order_no = b.order_no
Rather than:
WHERE TO_NUMBER (SUBSTR (a.order_no, INSTR (b.order_no, '.') - 1))
= TO_NUMBER (SUBSTR (a.order_no, INSTR (b.order_no, '.') - 1))
Do not use SQL functions in predicate clauses or WHERE clauses. Any expression
using a column, such as a function having the column as its argument, causes the optimizer to
ignore the possibility of using an index on that column, even a unique index, unless there is a
function-based index defined that can be used. Avoid mixed-mode expressions, and beware of
implicit type conversions.
Avoid the following kinds of complex expressions:
- col1 = NVL (:b1,col1)
- NVL (col1,-999) = .
- TO_DATE(), TO_NUMBER(), and so on
These expressions prevent the optimizer from assigning valid cardinality or selectivity
estimates and can in turn affect the overall plan and the join method. Add the predicate versus
using NVL() technique. The SQL functions doesnt used on the columns on which have an
index. For example:
SELECT employee_num, full_name Name, employee_id
FROM mtl_employees_current_view
WHERE(employee_num=NVL(:b1,employee_num))AND (organization_id=:1)
ORDER BY employee_num;
Also:
SELECT employee_num, full_name Name, employee_id
FROM mtl_employees_current_view
WHERE (employee_num = :b1) AND (organization_id=:1)
ORDER BY employee_num;
Written as shown, the database cannot use an index on the somecolumn column, because the
expression involving that column uses the same column on both sides of the BETWEEN.
To use the index whenever real values are given for :loval and, then can rewrite the
example in the following logically equivalent form:
SELECT /* change this half of UNION ALL if other half changes */ info
FROM tables WHERE ...
AND somecolumn BETWEEN :loval AND :hival
AND (:hival != 'ALL' AND :loval != 'ALL')
UNION ALL
SELECT /* Change this half of UNION ALL if other half changes*/ info
FROM tables WHERE ...
821
AFASES - 2008 -
If you run EXPLAIN PLAN on the new query, then you seem to get both a desirable
and an undesirable execution plan. However, the first condition the database evaluates for
either half of the UNION ALL is the combined condition on whether: hival and: loval are
ALL. The database evaluates this condition before actually getting any rows from the
execution plan for that part of the query. When the condition comes back false for one part of
the UNION ALL query, that part is not evaluated further. Only the part of the execution plan
that is optimum for the values provided is actually carried out. Because the final conditions on
:hival and :loval are guaranteed to be mutually exclusive, only one half of the UNION ALL
actually returns rows.
Use of EXISTS versus IN for Subqueries
In certain circumstances, it is better to use IN rather than EXISTS. In general, if the
selective predicate is in the subquery, then use IN. If the selective predicate is in the parent
query, then use EXISTS.
Sometimes, can rewrite a subquery when used with an IN clause to take advantage of
selectivity specified in the subquery. This is most beneficial when the most selective filter
appears in the subquery and there are indexes on the join columns. Conversely, using EXISTS
is beneficial when the most selective filter is in the parent query. This allows the selective
predicates in the parent query to be applied before filtering the rows against the EXISTS
criteria.
The next examples are two examples that demonstrate the benefits of IN and EXISTS.
Both examples use the same schema with the following characteristics: there is a unique index
on the employees.employee_id field, on the orders.customer_id field, on the
employees.department_id field and the employees table has 27,000 rows.
Example 1: Using IN - Selective Filters in the Subquery This example demonstrates how
rewriting a query to use IN can improve performance. This query identifies all employees
who have placed orders on behalf of customer 144.
SELECT /* EXISTS example */
e.employee_id, e.first_name, e.last_name, e.salary
FROM employees e
WHERE EXISTS (SELECT 1 FROM orders o /* Note 1 */
WHERE e.employee_id = o.sales_rep_id /* Note 2 */
AND o.customer_id = 144); /* Note 3 */
The following plan output is the execution plan (from V$SQL_PLAN) for the
preceding statement. The plan requires a full table scan of the employees table, returning
many rows. Each of these rows is then filtered against the orders table (through an index).
ID OPERATION OPTIONS OBJECT_NAME OPT COST
---- -------------------- --------------- ---------------------- --0 SELECT STATEMENT s1
1 FILTER
2 TABLE ACCESS FULL EMPLOYEES JOHN 155
3 TABLE ACCESS BY INDEX ROWID ORDERS JOHN 3
4 INDEX RANGE SCAN ORD_CUSTOMER_IX JOHN 1
The following plan output is the execution plan (from V$SQL_PLAN) for the preceding
statement. The optimizer rewrites the subquery into a view, which is then joined through a
unique index to the employees table. This results in a significantly better plan, because the
view (that is, subquery) has a selective predicate, thus returning only a few employee_ids.
822
AFASES - 2008 -
Example 2: Using EXISTS - Selective Predicate in the Parent This example demonstrates
how rewriting a query to use EXISTS can improve performance. This query identifies all
employees from department 80 who are sales reps who have placed orders.
The following SQL statement uses IN:
SELECT e.employee_id, e.first_name, e.last_name, e.department_id,
e.salary FROM employees e
WHERE e.department_id = 80 /* Note 5 */
AND e.job_id = 'SA_REP' /* Note 6 */
AND e.employee_id IN (SELECT o.sales_rep_id FROM orders o)/*Note 4 */
The following plan output is the execution plan (from V$SQL_PLAN) for the preceding
statement. The SQL statement was rewritten by the optimizer to use a view on the orders
table, which requires sorting the data to return all unique employee_ids existing in the orders
table. Because there is no predicate, many employee_ids are returned. The large list of
resulting employee_ids are then used to access the employees table through the unique index.
ID OPERATION OPTIONS OBJECT_NAME OPT COST
---- -------------------- --------------- ---------------------- -0 SELECT STATEMENT S3
1 NESTED LOOPS 125
2 VIEW 116
3 SORT UNIQUE 116
4 TABLE ACCESS FULL ORDERS ANA 40
5 TABLE ACCESS BY INDEX ROWID EMPLOYEES ANA 1
6 INDEX UNIQUE SCAN EMP_EMP_ID_PK ANA
The following plan output is the execution plan (from V$SQL_PLAN) for the
preceding statement. The cost of the plan is reduced by rewriting the SQL statement to use an
EXISTS. This plan is more effective, because two indexes are used to satisfy the predicates in
the parent query, thus returning only a few employee_ids
ID OPERATION OPTIONS OBJECT_NAME OPT COST
---- -------------------- --------------- --------------------0 SELECT STATEMENT S4
1 FILTER
2 TABLE ACCESS BY INDEX ROWID EMPLOYEES ANA 98
3 AND-EQUAL
4 INDEX RANGE SCAN EMP_JOB_IX ANA
5 INDEX RANGE SCAN EMP_DEPARTMENT_IX ANA
6 INDEX RANGE SCAN ORD_SALES_REP_IX ANA 8
823
AFASES - 2008 -
the optimizer, can choose a more effective way to execute a SQL statement. You can use hints
in SQL statements to specify how the statement should be executed.
Hints, such as /*+FULL */ control access paths. For example:
SELECT /*+ FULL(e) */ e.last_name FROM employees e
WHERE e.job_id = 'CLERK';
Join order can have a significant effect on performance. The main objective of SQL
tuning is to avoid performing unnecessary work to access rows that do not affect the result.
This leads to three general rules:
- Avoid a full-table scan if it is more efficient to get the required rows through an index.
- Avoid using an index that fetches 10,000 rows from the driving table if you could instead
use another index that fetches 100 rows.
- Choose the join order so as to join fewer rows to tables later in the join order.
The following example shows how to tune join order effectively:
SELECT info FROM taba a,
WHERE a.acol BETWEEN 100
AND b.bcol BETWEEN 10000
AND c.ccol BETWEEN 10000
AND a.key1 = b.key1
AND a.key2 = c.key2;
tabb b, tabc c
AND 200
AND 20000
AND 20000
1. Choose the driving table and the driving index (if any).
The first three conditions in the previous example are filter conditions applying to
only a single table each. The last two conditions are join conditions. Filter conditions
dominate the choice of driving table and index. In general, the driving table is the one
containing the filter condition that eliminates the highest percentage of the table. Thus,
because the range of 100 to 200 is narrow compared with the range of acol, but the ranges of
10000 and 20000 are relatively large, taba is the driving table, all else being equal. With
nested loop joins, the joins all happen through the join indexes, the indexes on the primary or
foreign keys used to connect that table to an earlier table in the join tree. Rarely do you use
the indexes on the nonjoin conditions, except for the driving table. Thus, after taba is chosen
as the driving table, use the indexes on b.key1 and c.key2 to drive into tabb and tabc,
respectively.
2. Choose the best join order, driving to the best unused filters earliest.
The work of the following join can be reduced by first joining to the table with the
best still-unused filter. Thus, if "bcol BETWEEN ..." is more restrictive (rejects a higher
percentage of the rows seen) than "ccol BETWEEN ...", the last join can be made easier (with
fewer rows) if tabb is joined before tabc.
3. You can use the ORDERED or STAR hint to force the join order.
Use Caution When Managing Views Be careful when joining views, when performing outer
joins to views, and when reusing an existing view for a new purpose.
Use Caution When Joining Complex Views Joins to complex views are not recommended,
particularly joins from one complex view to another. Often this result in the entire view being
instantiated, and then the query is run against the view data.
Do Not Recycle Views Beware of writing a view for one purpose and then using it for other
purposes to which it might be ill-suited. Querying from a view requires all tables from the
view to be accessed for the data to be returned. Before reusing a view, determine whether all
tables in the view need to be accessed to return the data. If not, then do not use the view.
Instead, use the base table(s), or if necessary, define a new view. The goal is to refer to the
minimum number of tables and views necessary to return the required data.
Use Caution When Unnesting Subqueries Subquery unnesting merges the body of the
subquery into the body of the statement that contains it, allowing the optimizer to consider
them together when evaluating access paths and joins.
824
AFASES - 2008 -
Use Caution When Performing Outer Joins to Views In the case of an outer join to a
multitable view, the query optimizer can drive from an outer join column, if an equality
predicate is defined on it.An outer join within a view is problematic because the performance
implications of the outer join are not visible.
Store Intermediate Results Intermediate, or staging, tables are quite common in relational
database systems, because they temporarily store some intermediate results. In many
applications they are useful, but Oracle requires additional resources to create them. Always
consider whether the benefit they could bring is more than the cost to create them. Avoid
staging tables when the information is not reused multiple times.
Restructuring the Indexes
Often, there is a beneficial impact on performance by restructuring indexes. This can
involve the following:
- Remove nonselective indexes to speed the DML.
- Index performance-critical access paths.
- Consider reordering columns in existing concatenated indexes.
- Add columns to the index to improve selectivity.
Application developers sometimes think that performance will improve if they create
more indexes. If a single programmer creates an appropriate index, then this might indeed
improve the application's performance.
Modifying or Disabling Triggers and Constraints
Using triggers consumes system resources. If you use too many triggers, then you can
find that performance is adversely affected and you might need to modify or disable them.
Restructuring the Data
After restructuring the indexes and the statement, you can consider restructuring the data.
- Introduce derived values. Avoid GROUP BY in response-critical code.
- Review data design. Change the design of system if it can improve performance.
- Consider partitioning, if appropriate.
Maintaining Execution Plans Over Time
It can maintain the existing execution plan of SQL statements over time either using
stored statistics or stored SQL execution plans. Storing optimizer statistics for tables will
apply to all SQL statements that refer to those tables. Storing an execution plan (that is, plan
stability) maintains the plan for a single SQL statement. If both statistics and a stored plan are
available for a SQL statement, then the optimizer uses the stored plan.
Visiting Data as Few Times as Possible
Applications should try to access each row only once. This reduces network traffic and
reduces database load. Consider doing the following: combine multiples scans with CASE
statements, use DML with RETURNING Clause and modify all the data needed in One
Statement.
Combine Multiples Scans with CASE Statements
Often, it is necessary to calculate different aggregates on various sets of tables. Usually, this is
done with multiple scans on the table, but it is easy to calculate all the aggregates with one
single scan. Eliminating n-1 scans can greatly improve performance. Combining multiple
scans into one scan can be done by moving the WHERE condition of each scan into a CASE
statement, which filters the data for the aggregation. For each aggregation, there could be
another column that retrieves the data.
The following example asks for the count of all employees who earn less then 2000, between
2000 and 4000, and more than 4000 each month. This can be done with three separate
queries:
SELECT COUNT (*)
FROM employees
WHERE salary < 2000;
825
AFASES - 2008 -
However, it is more efficient to run the entire query in a single statement. Each number is
calculated as one column. The count uses a filter with the CASE statement to count only the
rows where the condition is valid. For example:
SELECT COUNT (CASE WHEN salary < 2000
THEN 1 ELSE null END) count1,
COUNT (CASE WHEN salary BETWEEN 2001 AND 4000
THEN 1 ELSE null END) count2,
COUNT (CASE WHEN salary > 4000
THEN 1 ELSE null END) count3
FROM employees;
3. Conclusions
These operations and information are tested in an application system based on
Oracle Database which offers many tools and facilities which provides the management and
optimization of SQL statements. Today, the performance for an integrated applications system
is a priority.
Bibliography
1. www.metalink.oracle.com
2. Rebecca M. Riordan , Designing Effective Database Systems, 2005
3. Oracle Database , Performance Tuning Guide
PhD
Student,
826
Academy
of
Economic
Studies
Bucharest,
AFASES - 2008 -
827
AFASES - 2008 -
828
AFASES - 2008 -
CLIENTII
PORTULUI
AUTORITTI
Declaratii vamale
Vmuire, control
Consultare
AGENTI
MARITIMI
Anunt de sosire
VAMA
Pregtirea escalei
POLITIA DE
FRONTIER
VTMIS
CPITNIA
PILOTAJ
ADMINISTRATIA
PORTULUI
REMORCAJ
LEGAREDEZLEGARE
Manevre
(sosiri, plecri,
n port)
EXPLOATARE
Alocare echipe, utilaje
Operare nave
Depozitare mrfuri
OPERATORI
PORTUARI
NTRETINERE
ntretinere preventiv
Reparatii
TRANSPORTATORI
CLIENTI
Documente transport
Expedieri / primiri mrfuri
829
AFASES - 2008 -
830
AFASES - 2008 -
1. Subsistemul NAVA
3. Subsistemul
MRFURI
Gestiunea tarifelor
Gestiunea
mrfurilor de export
Gestiunea
mrfurilor de import
Urmrirea
afluirii/expedierii
mrfurilor
Facturarea prestaiilor
efectuate mrfurilor
4. Subsistemul
RESURSE UMANE I
UTILAJE
Gestiunea echipelor
Gestiunea utilajelor
Planificarea echipelor i
utilajelor la operare
Urmrirea folosirii
echipelor i utilajelor
831
AFASES - 2008 -
Sistemele informatice
ale
transportatorilor
terestri
Sistem
integrator
(Vama)
AIS
Sistemele informatice
ale
furnizorilor si
beneficiarilor
de mrfuri
Sistemul
informatic portuar
+ VTMIS
Sistemele informatice ale
operatorilor economici portuari
si ale autorittilor portuare
832
AFASES - 2008 -
LOCATIA
CENTRAL
BAZ
DE DATE
BAZ
DE DATE
Client
PORTAL
Protectie
(firewall)
Servere de aplicatie
Server WEB
Modem
INTERNET
Concentrator de date
Concentrator de date
HUB
HUB
EXTRANET
Client
INTRANET
Client
Client
Client
Client
Client
de client.
n ceea ce privete arhitectura tehnic, soluia actual adoptat de sistemele mari care
includ un numr important de calculatoare legate n reea (de regul peste 50) se bazeaz pe
arhitecturi tehnice (hardware i software) de tipul client/server, respectiv:
a) un server activ funcionnd sub un sistem de operare specific (de ex.: UNIX,
AIX etc) care asigur stocarea datelor ntr-o baz de date relaional de tip SQL i gestioneaz
cererile venite din partea clienilor pentru nregistrarea i livrarea informaiilor solicitate i un
server de rezerv care s preia sarcinile serverului de baz n caz de incident hardware sau
indisponibilitii acestuia din diferite motive (revizii hard/soft, up-grade etc.);
b) staii de lucru (clieni) reprezentate de calculatoare de tip PC (Personal
Computer) pe care se execut aplicaiile specifice iniiate de utilizatori i prin care se solicit
serverului accesul la baza de date pentru nregistrarea sau citirea informaiilor (datelor) care
sunt transpuse pe ecranul calculatorului ntr-o form accesibil i convenabil utilizatorului.
Clienii pot fi legai ntr-o reea intern organizaiei (Intranet), la reeaua
internaional de schimb de date (Internet) sau la o facilitate de comunicaii exterioar de la
care s aib acces direct la serverul de date (Extranet). Bazele de date sunt de tip relaional
bazate pe standardul SQL utiliznd unul din sistemele consacrate de baze de date (Oracle,
Sybase, Informix etc.).
Arhitectura unui asemenea sistem este asemntoare cu ce a unui sistem dezvoltat
pentru marile organizaii (fig. 5)
4. Tendine n dezvoltarea sistemelor informatizate portuare
Tendinele de dezvoltare a sistemelor portuare informatizate trebuie s se nscrie n
direcia celor dou componente majore care alctuiesc sistemul:
Activitatea portuar care reprezint obiectul sistemului informatizat trebuie s se
nscrie n tendina porturilor de a deveni porturi de generaia a III-a care implic portul
ca un actor principal n lanul de transport prin:
833
AFASES - 2008 -
834
AFASES - 2008 -
O alt diferen ntre EDI i XML este c acesta din urm suport i o serie de alte
funciuni de manipulare a datelor cum ar fi gestiunea coninutului i un sistem foarte dezvoltat
de cutare web. Pe de alt parte EDI a fost proiectat ca un sistem pentru reele private n timp
ce XML este dezvoltat pentru a utiliza n mod ct mai complet avantajele Internetului.
Unul din riscurile XML-ului este c definirea datelor i structura acestora pot fi alese
liber de utilizator ceea ce face ca, n cazul n care o organizaie i dezvolt propriul su
sistem utiliznd definiii i structuri unice, nestandard, aceasta poate face extrem de dificil
schimbul de date fiind necesare instruciuni auxiliare pentru a translata definirea i structura
datelor de la un sistem la altul.
Pentru a elimina acest neajuns UNCEFACT (United Nations Centre for Trade
Facilitation and Electronic Business) a lansat mpreun cu OASIS (Organisation for
Advancement of Structured Information Standard) iniiativa ebXML (electronic business
XML) care i-a propus furnizarea unei infrastructuri bazat pe XML care s poat asigura o
folosire global a informaiilor comerciale electronice ntr-o manier interoperabil, sigur
i consistent pentru toate prile. Scopul ebXML este de a facilita comerul deschis ntre
organizaiile comerciale de orice mrime prin folosirea XML ntr-o manier consistent
pentru schimbul electronic de date comerciale.
5. Concluzii
n concluzie este evident c situaia actual care exist, la nivelul transportului naval,
o penurie de schimburi de informaii pertinente ntre participanii la activitile portuare, poate
fi substanial mbuntit prin dezvoltarea unor sisteme informatice integrate care, cel puin
din punct de vedere tehnic, este posibil n condiiile actuale de dezvoltare tehnologic.
Oarecum ignorat n prezent, cel puin n porturile maritime romneti, tocmai din
cauza penuriei de informaii, planificarea activitilor portuare, poate deveni o component
major a utilizrii corespunztoare a tuturor disponibilitilor i resurselor pe care le au
porturile i poate contribui substanial la creterea performanelor acestora.
Bibliografie:
1. Robert Schware, Paul Kimberey - Information technology and national
trade
facilitation, World Bank technical paper number 316
2. Siveco S.A. materiale de prezentare a sistemelor informatice de ntreprindere - EAS
3. dbh Datenbank Bremische Hfen GmbH, Bremen Harbour Telematics, version 2.1.,
march 1995
4. ICI Bucureti, EDIPA Electronic data interchange between port administrations,
2000
5. ICI Bucureti, Sistem de monitorizare avansat a mobilitii mijloacelor de transport
i mrfurilor multimodal suport pentru realizarea societii naionale
informaionale mobile proiect, contract Amtrans 3C03/12.08.2002
6. *** Port technology international, anii 2003 2007
7. IAPH, Ports and harbours, anii 1999 2007
8. www.ebxml.org/specs
i
Pascale Dinu, Asistent universitar, APMC, incint port, tel. 0724219122, e-mail: dpascale@constantzaport.ro
**
Tulei Filis, Preparator universitar, Academia Naval Mircea cel Btrn, tel. 0730293633,
e-mail: fixe2005@gmail.com
835
AFASES - 2008 -
Introduction
1.1
The problem of assuring and maintaining reliability has many facets, including original equipment design, control of quality during production, acceptance inspection, field
trials, life testing, and design modifications. To complicate matters further, reliability
competes directly or indirectly with a host of other engineering considerations, chiefly
cost, complexity, size and weight and maintainability.
We define the reliability of a product as the probability that it will function within
specified limits for at least a specified period of time under specified environmental conditions.
Many systems can be considered to be series or parallel systems, or a combination of
both. A series system is one in which all components are interrelated in such a way, that
the entire system will fail if any one of its components fails; a parallel system is one that
will fail only if all of its components fail.
n
Q
For a system of n components connected in series we have Rs = Ri ,where Ri is the
i=1
reliability of the ith component and Rs is the reliability of the series system.
For a system of n independent components connected in parallel, if Fi = 1 Ri
is the unreliability of the ith component, we can again apply the special rule of
n
n
Q
Q
multiplication for probabilities to obtain Fp = Fi and Rp = 1
(1 Ri ) ,where Fp
i=1
i=1
is the unreliability of the parallel system, and Rp = 1 Fp is the reliability of the parallel
system.
836
AFASES - 2008 -
Rt
The reliability function for the exponential model is R (t) = 1 ex dx = et and
n
P
n
Y
t
i
i t
Rs (t) = 1
e
= e i=1 ,
i=1
n
Q
a
parallel
system
Fp (t)
=
1
(1 ei t )
and
i=1
Rp (t) = n1 et n2 e2t + ... + (1)n1 ent .
The reliability function associated with the Weibull failure-time distribution is given by
For
1.2
Theorem 1 (i) (Extremal Types for Maxima). Let Mn = max (i ) , where i are inde1in
pendent identically distributed (i.i.d.) random variables. If for some constants n > 0, n ,
w
we have P {n (mn n ) x} G (x) for some nondegenerate G, then G is one of the
three extremal types for maxima:
Type 1: G1 (x) =
exp (ex ) , < x <
0, x 0
Type 2: G2 (x) =
bn (Mn an )
(1)
as n and let the distribution function G (x) = P ( < x) of the random variable
be nondegenerate.
If relation (1) holds we say that the distribution function F belongs to the domain of
attraction of the law G and we write F D (G) .
As noted the minimum mn = min i is simply given as mn = max (i ) , such that
1in
1in
limiting results for minima can be obtained from those for maxima.
Theorem 2 (i) (Extremal Types for Minima). Let mn = min (i ) , where i are i.i.d.
1in
1 exp (x ) , > 0 , x 0.
(ii) The min - stable distributions are those given in (i) above [2], [3].
837
AFASES - 2008 -
Classical Extreme Values Theory - the asymptotic distributional theory for maxima
of independent, identically distributed random variables - may be regarded as half a
century old, even though its roots reach further back in to mathematical antiquity. Recent
researches made by G.S.Watson, S.M. Berman, R.M. Loynes, H. Cramer, L. Haan, S.
Resnick, M.R. Leadbetter, G. Lindgren, H. Rootzen, who were interested in extending the
theory to include, first, dependent sequences and then continuous parameter stationary
processes, and I. Matsak (who extended the theory for sequence of independent normal
random functions belong to extending intervals, for Banach lattices).
The literature abounds with applications of the classical theory, and many of these
are really dependent cases although assumed independent. More importantly, however,
we feel that each potential application should be understood as well as possible in terms
of its underlying physical principles so that extremal theory may be thoughtfully applied
in the light of such principles, rather than by routine trial and error fitting.
Extreme value distributions have found widespread use for the description of strength
of materials and mechanical structures, often in combination with stochastic models for
the loads and forces acting on the material. Thus it is often assumed that the maximum of
several loads follows one of the extreme value distributions for maxima. More important,
and also less obvious, is that the strength of piece of material, such as a strip of paper or
glass fibre, is sometimes determined by the strength of its weakest part, and then perhaps
it follows one of the extreme value distributions for minima.
2.1
Suppose that a piece of material, such as a glass fibre or an iron bar, is subject to
tension, and that it breaks if the tension exceeds the inherent strength of the material.
Let L be the random strength of a piece of material L of length l, and suppose that the
material can be divided, at least hypothetically, into smaller pieces, L1 , ..., Ln of arbitrary
lengths l1 , ..., ln , and with (random) strength L1 , L2 , ..., Ln , respectively. We say that
the material is stochastically:
(i) brittle if L = min (L1 , ..., Ln ) ;
(ii) homogeneous if the marginal distributions of L1 , ..., Ln depend only on l1 , ..., ln ;
(iii) disconnected if L1 , ..., Ln are independent for all disjoint subdivisions L1 , ..., Ln
of L.
Suppose a material satisfies (i)(iii) and let Fl (x) be the (non-degenerate) d.f.(distribution
function) of the strength of a piece of length l. Then
1 Fl (x) = 1 Fl/n (x)
n
(2)
Now, any d.f. Fl (x) satisfies (2) for some Fl/n (x) , e.g. Fl/n (x) = 1 (1 Fl (x))1/n ,
such that in order to obtain a simple structure, we have to introduce one extra restriction
on the material.
A material is called stochastically
(iv) size-stable if the distribution of L is of the same type regardless of length l, i.e.
there are constants al > 0, bl and d.f. Fl (x) = F (al (x bl )) .
For a size-stable material, write n = al/n a1
l , n = bl n bl/n , Fl/n (x) = Fl (n x + n ) ,
so that, by (2),
838
AFASES - 2008 -
1 Fl (x) = (1 Fl (n x + n ))n .
(3)
Thus, Fl , is min-stable [2], and hence, by Theorem 2, F can be taken to have one of the
three minimum extremal types.
Thus strength distributions for materials satisfyng (i) (iv) are one of the three
minimum extremal types. If, furthermore, the scale of measurement is such that the
measured strngth L is bounded from below, L x0 , for some x0 (, ) , then the
only possibility is for Type III (or Weibull) distribution, with location parameter x0 , and
general scale parameter, i.e.
F (x) = 1 exp { (a (x x0 )) } , x x0 ,
for some constant > 0. Often x0 = 0 is a natural choise, expressing the fact strength
can never be negative. However, even in cases when the strength must be positive, one
may prefer to use one of the other types, the double exponential,
F (x) = 1 exp ea(xb) , < x < ,
since if the location parameter b is large enough, the probability of a value below the a
priori lower bound is negligible.
2.2
A strip of paper will burst when it is subject to a tension which exceeds the strength
at its weakest point. Here we discuss in the mean strength of paper strips (of constant
width) of length varying from 8 cm up to 10 cm.
The next table gives location and scale parameters and the mean and standard deviation
R 1ofxstrength as functions of the tested length l (expressed in terms of () =
x e dx). For completeness, all three types of distributions are included in the
0
table, even though the Type 2 d.f. is not commonly used as strength distribution.
Type 1 (double exponential, F (x) = 1
exp (ex ))
1
F (x) = F (a (x b )) = 1 exp ea(xb+a log l)
l
al = a
bl = b a1 log l
ml = b a1
( + log l)
1
l = a / 6
Type 2 F (x) = 1 exp (x) , x < 0
Fl (x) = F (al (x bl )) = 1 exp(l a (x x0 ) , x < x0
al = al1/
bl = b = x0
ml = x0 l1/ a1 (1 1/) (if > 1)
1/2
l = l1/ a1 { (1 2/) 2 (1 1/)}
(if > 2)
839
AFASES - 2008 -
Two qualities of paper were tested and for one of the qualities three series of experiments were carried out, but no decisive conclusion can be drawn from these data regarding
the best fitting distribution [2]
In a stochastically inhomogeneous material the strength distribution for a small piece
of the material varies with its locations. However, if the conditions (i) , (iii) , and (iv) are
satisfied, one can still obtain one of the extreme value distributions under mild conditions
on the inhomogeneity. One simple kind of inhomogeneity is obtained by measuring the
size of the material not as its physical dimension but in terms of an integral of a positive
local size function, as described below.
Let (x) , x L be a non-negative, integrable function, and define, for Li L,
Z
li = (x) dx.
(4)
Li
n
S
i=1
Then it is readily seen that (2) and (3) stil hold, so that Fl (x) is one of the three extremal
types for minima and it is min-stable. Using a, b as the parameters in Fl (x) , we have
in the Weibull and
exponential case Fl (x) = 1 exp { (a (x x0 )) } , x > x0
double
Fl (x) = 1 exp ea(xb) , respectively.
Starting with the local size function (x), one can simply derive closed-form expressions for theR location and scale parameters in the strength distribution for a piece
L of size l = (x) dx, which is F (L) (x) = Fl (x) = F (al (x bl )), with al and bl
L
given in the previous table. In order to obtain these expressions, define the functions
a (x) , b (x) as a (x) = a1/ (x) (Weibull), b (x) = b a1 log (x) (double exponential)
1/
R
R
and A (L) =
a (x) dx
, B (L) = a1 log eab(x) dx.
L
respectively. These formulae motivate the use of a (x)1 and b (x) as local scale and
location functions in standard extremal models.
In both cases the average strength is given by
x0 + A (L)1 (1 + 1/) ,
E (L ) =
a1 + B (L) .
840
AFASES - 2008 -
References
[1] R. A. Johnson, Probability and statistics for engineers, Miller & Freunds, sixth edition, 2000, 546 - 554.
[2] M.R. Leadbetter, G. Lindgren, H. Rootzen, Extremes and related properties of random
sequences and processes, Springer, New York , 1983, 272 - 277.
[3] A. M. Proca, Applications to reliability, 18th Scientific Session on Mathematics and
its Applications, Brasov, April 2004, 139 - 142.
[4] A. M. Proca, Extremal properties of random sequences, Buletinul Universitatii Transilvania, 2005, 69 - 74.
[5] A. M. Proca, Application of extremal value theory to strength of materials, Scientific
Session on Mathematics and its Applications, Brasov, May 2006, 109-114.
841
AFASES - 2008 -
842
AFASES - 2008 -
843
AFASES - 2008 -
Fig. 4: TGPF
844
AFASES - 2008 -
Modul de funcionare
n rndurile urmtoare voi oferi o privire general asupra modului de funcionare i
implementrii programului.
Pentru a putea genera terenul folosind algoritmul fractalilor este necesar crearea unei
grile de puncte asupra creia va fi aplicat ulterior alogritmul. Aspectul
grilei poate fi modificat din mai multe puncte de vedere. n primul rnd
se poate ajusta numrul de linii i coloane ale acesteia apelnd meniul :
Fractal Parameters / Detail & Aspect. Va aparea o fereastr n care
putem modifica numrul de linii i coloane ale grilei ce urmeaz a fi
generat (prin modificarea parametrilor Number of Rows, Number of
Fig. 5
Columns se ajusteaz numrul de linii, respectiv coloane ale grilei). n
figura 5 este desenat o gril generat cu 3 linii si 4 coloane. Punctele roii reprezint
nodurile grilei (dimensiunile minime sunt 1x1 reprezentnd un ptrat).
n acest moment putem genera grila apsnd butonul Generate New Terrain Grid. Se
observ c unele dintre butoanele care erau inactive iniial pot fi acum accesate. Apsnd
butonul Adjust Terrain Grid este lansat engine-ul ATG care permite modificarea altitudinii
diferitelor noduri ale grilei. Rezoluia se poate modifica apelnd meniul Rendering / Video
Options n care vom avea disponibile 3 rezoluii (640x480, 800x600 i 1024x768), precum i
2 adncimi de culoare (16 bii i 32 bii). Pentru a modifica nlimea anumitor puncte : se
selecteaz punctele dorite (cu butonul stnga al mouse-ului, innd eventual apsate tastele
Ins/Del pentru selectare/deselectare de puncte) i se modific altitudinea innd butonul al
treilea al mouse-ului apsat. De asemenea putem roti camera innd apsat butonul dreapta al
mouse-ului. Pentru a executa zoom-in se folosete tasta A, i Z pentru zoom-out.
Dup ce suntem mulumii cu aspectul grilei, putem s ajustm parametrii cu care va fi
generat terenul. Acetia sunt:
Seed Properties : se acceseaz din meniul Fractal Parameters / Seed Properties.
Este folosit pentru a iniializa modul de generare a valorilor aleatoare. Putem
specifica un numar (Seed Number) sau putem spune calculatorului s-l iniializeze n
funcie de ceasul intern al acestuia (bifnd check-box-ul Time Seed). Aici avem dou
alternative: s initializm generatorul de valori aleatoare doar naintea execuiei
algoritmului, sau naintea fiecrei iteraii a acestuia.
Initial Roughness Value : se acceseaz din meniul Fractal Parameters / Roughness;
aceasta este valoarea care influeneaz duritatea, asperitatea terenului generat (de
fapt este amplitudinea maxim cu care poate fi modificat altitudinea unui nod).
Roughness Factor : se acceseaz din meniul Fractal Parameters / Roughness;
parametrul specific valoarea la care este mprit duritatea, procedura care se
execut la fiecare iteraie a algoritmului, din cauz c dimensiunea ptratelor se
micoreaz de 4 ori la fiecare iteraie.
Number of Iterations : se acceseaz din meniul Fractal Parameters / Detail &
Aspect; indic numrul de iteraii ( la fiecare iteraie grila se fragmenteaz, devenind
din ce n ce mai fin).
Pentru a alege metoda de generare, acionm butonul Terrain Genesis (fig. 6).
845
AFASES - 2008 -
Fig. 7
Fig. 6
n acest moment putem apsa butonul Generate Terrain. Vei observa c n seciunea
Processing a ferestrei principale este afiat task-ul curent executat de calculator, precum i
stadiul n care acesta se afl.
Dup generarea terenului vor deveni disponibile urmtoarele butoane : Terain
Modification (fig.7), Terain Combination (fig.8), Adjust Camera, Smooth, Raise/Lower
Terrain, Render.
Adjust Camera : lanseaz engine-ul
AC, care permite ajustarea unghiului
i poziiei camerei. Fa de engine-ul
ATG se mai pot folosi i tastele S i
X pentru a modifica altitudinea
camerei. Nu mai este posibil
modificarea grid-ului.
Smooth : aplic un efect de smooth
terenului, n cazul n care acesta este
prea aspru.
Raise/Lower Terrain : modific
altitudinea tuturor punctelor gridului
relativ la poziia curent a acestora.
RENDER : randeaz imaginea la o
calitate ridicat (utiliznd engine-ul
FR), iar fa de engine-ul AC, are
Fig. 7
implementate i efecte speciale :
reflexii, n cazul n care este randat
i apa, precum i efectul de cea (fog).
n cazul realizrii de peisaje, se apeleaz la meniul Landscape (fig.8) i Landscape
Setings(fig.9).
846
AFASES - 2008 -
Fig. 9
Fig. 8
847
AFASES - 2008 -
Fig. 11
Fig. 12
O alt aplicaie de acest gen, dar mult mai simpl n utilizare, accesibil unei categorii
largi de utilizatori care doresc obinerea de simulri rapide de peisaje fr pretenii deosebite o
constituie TerrTexGen (fig. 13 i fig.14)
Fig. 13
Fig. 14
Bibliografie
1. Ackermann,F., 1999, Airborne laser scanning- present status and future
expectations, ISPRS Journal of Photogrammetry & Remote Sensing nr.54, 1999.
2. Radu,A.G.,Trziu,H. Anatomy of a Povray file for using Geomorph . Conferina
Internaional Nav-Mar-Edu 2007, Constana 2007
3. Radu,A.G.,Harabagiu,A. About the Apollonian Gasket or Curviliniar Sierpinsky
Gasket. Conferina Internaional Nav-Mar-Edu 2007, Constana 2007
848
AFASES - 2008 -
4.
5.
6.
7.
Radu, A.G. Modelare i simulare n 3D. Universitatea Petrol i Gaze, Ploieti, 2006
http://www.planetside.co.uk./terragen/
http://en.wikipedia.org/wiki/fractal/
***OEEPE-Workshop on Application of Digital Photogrammetric Workstations,
Lausane 4-6 Martie, 1996
Adrian-Gelu Radu, prof. inf.-fiz., Grup colar Meserii i Servicii Buzu, tel. 0722453007,
e-mail: agerbit@yahoo.com
**
Anca Harabagiu, prof. drd. inf.-fiz., Inspectoratul colar Judeean, Buzu, tel. 0721126577,
e-mail: anca_hara@yahoo.com
849
AFASES - 2008 -
1. Julia Set
Julia Set poate fi descris ca un set de numere care niciodat nu crete la infinit cnd se
itereaz funcia:
f (z ) = z 2 +
(1)
850
AFASES - 2008 -
unde variabilele z i sunt n mod uzual numere complexe, dar formula poate fi deasemenea
aplicat quaternionilor. Alegnd diferite valori pentru constanta , pot fi generate diferite
seturi.
Cel mai comun set Julia, cu numere complexe, poate fi ataat diagramelor
bidimensionale impunnd partea real n lungul axei x i partea imaginar n lungul axei y.
Forma ce apare este o imagine binecunoscut (fig.1)
n special pentru tipul de fractali Mandelbrot.
Utiliznd quaternieni n loc de numere
complexe, se poate obine un set de numere ce pot fi
ataate n 4-dimensiuni. A.Norton a fost primul care
a implementat aceasta, creind imagini prin reducerea
unei dimensiuni, folosind astfel doar cele trei
dimensiuni rmase. El a creat o metod de a expune
acest set prin calcularea numrului de puncte ale
suprafeei setului i apoi expunndu-le pe un
monitor.
Adoptnd metode de ray-tracing la
renderizarea setului Julia, dezvoltatori ai metodei ca
Hast, Sandin, Kaufman [4], prin tragerea razelor n
set pentru crearea camerei virtuale, au reuit
Fig. 1 Standard 2D Julia care corespunde crearea mai rapid a imaginilor dect prin metoda
Norton.
quaternionului Julia ( = 0,7 0,2i )
Se poate ncerca crearea unei camere virtuale
n spaiul 4D astfel nct s poat fi mutat i rotit liber n orice direcie dintre cele patru,
ntruct se sper s se captureze ct mai mult din setul Julia n 4D. Valoarea tiinific a
acestor imagini poate fi ndoielnic, dar acestea fascineaz atenia. Oricum, dezvoltarea
tehnicilor vizuale ajut cercetarea funciilor dinamice i iterative.
2. Quaternionii
Quaternionii au fost descrii pentru prima dat de W.Hamilton ca o extensie a
numerelor complexe mperecheate cte dou n perecheri cte patru. Acetia au forma :
q = q1 + q 2 i + q3 j + q 4 k
(2)
unde i,j,k sunt uniti imaginare cu proprietile speciale:
i 2 = j 2 = k 2 = 1; ij = ji = k ; ki = ik = j; jk = kj = i
(3)
Important de notat aici, faptul c multiplicarea quaternionilor nu este comutativ.
Aplicnd acest set de reguli, formula Julia q 1 = q 2 + poate fi calculat astfel :
q1' = q12 q22 q32 q42 + 1; q2' = 2q1q2 + 2 ; q3' = 2q1q3 + 3 ; q4' = 2q1q4 + 4 ; (4)
Imaginile renderizate ce vor fi discutate n continuare sunt derivate din funcia
zn +1 = f ( z ) = zn2 + unde z este variabila de intrare iar o constant. Rezultatul a n iteraii
851
AFASES - 2008 -
3. Ray-tracing
Ray-tracingul este o metod foarte realistic de renderizare a formelor geometrice.
Efecte grafice ca umbre, reflexii i refracii pot fi uor implementate. Cum tehnica este foarte
lent i aproape imposibil de a genera renderizri n timp real cu computerele actuale, de
multe ori se renun la o parte din rezoluie.
Avnd poziia camerei, se poate calcula proiecia virtual n plan a punctelor
intersectate de raze. n ray-tracingul normal cu forme Euclidiene, se poate rezolva o ecuaie
pentru fiecare raz i obiect. Dup rezolvarea ecuaiilor se determin punctele unde raza intr
prin suprafaa obiectului. La aceste puncte se pot calcula suprafeele normale care sunt
folosite pentru iluminare i reflexii. Cnd se aplic ray-tracingul fractalilor Julia nu sunt
metode finite care s determine exact punctele suprafa unde razele intersecteaz arbitrar
setul. n timp ce se itereaz de un numr finit de ori, se obine de fapt o suprafa
aproximativ.
Gsirea punctului de intersecie prin scanarea spaiului 4D, testnd fiecare punct de-a
lungul razei reprezint o caracteristic a setului Julia. Acest proces este mare consumator de
timp-computer, deoarece fiecare punct se bazeaz pe un numr larg de funcii de iteraie.
Scanarea poate fi fcut la diferite rezoluii (z-rezolution) cea mai ridicat rezoluie fiind a
suprafeei finale, dar aceasta necesit un timp-computer apreciabil. Suprafeele setului Julia
difer foarte mult n funcie de constanta i de adncimea de iterare n.
4. Camera 4D
Pentru a fi posibil randarea setului Julia pentru diferite unghiuri i poziii, trebuie
definit modul de determinare a direciilor razelor. Soluia general pentru aceasta este crearea
unei camere virtuale. Aceasta este descris de doi vectori 4D, unul descriind punctul de
origine i cellalt descriind punctul de sosire. Referirea la al patrulea element al vectorilor,
este notat elementul w. Cnd razele create sunt ndreptate dinspre camer, proiecia plan
reprezint imaginea produs n faa camerei. Distana de la camer la planul de proiecie
determin lrgimea cmpului de vedere.
Metodele folosite n mod normal n 3D-ray-tracing care conin vectori 3D, pot fi
aplicate direct cnd sunt folosii vectori 4D. ncercnd imaginarea a patru vectori toi
perpendiculari unul pe cellalt nu este uor, dar matematic este posibil:
4 Dcross :4 4 4 4 unde q = 4 Dcross(a, b, c ) conduce la:
q1 = a2 (b4 c3 b3c4 ) + b2 (a3c4 a4 c3 ) + c2 (a4b3 a3b4 )
(6)
852
AFASES - 2008 -
Medierea efectului de plan ndeprtat face ca acesta s fie difuz. Distana trebuie
aleas astfel nct razele s se opreasc n interiorul setului Julia, nct acestea s par
retezate. Cum n mod obinuit razele nu pornesc imediat din camer, se poate implementa un
plan apropiat astfel nct s nceap testarea cu un bit n faa camerei. Alegnd planuri bune
de apropiere i deprtate, putem fi siguri c punctele testate au ansa de a fi n setul Julia. O
cale de a alege aceste planuri este ca la prima renderizare rezoluia imaginii setului s fie
sczut i nregistrnd unde se opresc i unde continu interseciile. Se pot seta planurile
apropiate i ndeprtate la aceste valori, cu o eroare mic.
5. Umbrirea
Umbrirea este un proces extrem de important n redarea
realistic i plcut estetic a imaginilor. Cnd se umbrete o
imagine, fiecare pixel este iluminat n funcie de normala sa la
suprafa. Unghiul dintre direcia luminii i suprafaa luminat
determin strlucirea pixelilor:
b = normal lightdirection
(7)
Culoarea suprafeei este apoi multiplicat cu factorul de
strlucire b. Pentru a performa acest calcul, trebuiesc alese normala
Fig. 2
la suprafa pentru fiecare pixel i direcia luminii (fig.2). Acestea
sunt metode mai realistice dect (7), de exemplu algoritmii unde
distana de la sursa de lumin produce efecte de luminozitate.
6. Animaia
Imaginile ray-tracing produse prin metodele amintite, atunci cnd setul Julia este
animat prin atribuirea de diferite valori constantei produc scene fascinante cu modificri i
rsuciri dinamice. Mutnd poziia camerei n timpul animrii creaz deasemenea imagini
dinamice interesante. Dac se mut camera n primele trei dimensiuni, se obine o imagine
dinamic a setului Julia ce poate fi vizualizat din diferite unghiuri. Mutndu-se poziia
camerei i vectorii int n lungul axei w, se poate randa Julia la diferite 4D-unghiuri.
Cu ajutorul animrii camerei se poate spera la captura unor proprieti ce nu sunt nc
cunoscute.
7. Post-stepping
Pentru a crete viteza de strbatere transversal a razei, s-a dezvoltat o metod care
este denumit post-stepping. Algoritmul este foarte simplu dar poate fi efectiv complet,
depinznd de civa parametri.
Raza transversal n mod obinuit este dirijat dinspre ochi spre scen n pai
incrementai. Lungimea acestor pai depinde de rezoluia z aleas.
853
AFASES - 2008 -
Fig. 4
Fig. 3
Cnd raza atinge un punct ce aparine setului Julia (un hit) procesul post-step se
declaneaz. Procesul post-step este un proces interactiv care poate gsi cu mult acuratee
poziii pe o suprafa. n fiecare pas, se mut cu jumtate distana de de vizionare, fiecare
punct avanseaz sau se retrage n funcie de parametrii sortai (fig. 3). Acesta poate porni ca o
d
funcie inductiv. Distana de salt n pasul n este definit ca: d n = n 1 .
2
Poziia p a suprafeei aproximate dup n iteraii:
pn 1 J : pn 1 d n
(8)
pn
pn 1 J : pn 1 + d n
unde d i p sunt vectori 4D iar J este setul Julia.
Procesul post-step crete acurateea suprafeei aproximative de puncte. Aceasta va
conduce la obinerea de imagini netede i cu detalii, chiar la rezoluii z sczute. Cnd se
utilizeaz o iteraie post-step, adncimea lui n este comparabil cu rezoluia z:
Re solutionz' = Re solutionz 2n
(9)
Cnd metoda este utilizat cu rezoluie z sczut, pasul maxim poate duce la depirea
suprafeei, nct raza dispare. O raz cu pas nalt poate depi uor marginile i astfel poziiile
sunt rrite n set, ceace duce la margini zimate i pierderi de arii ale suprafeei (fig. 4). Este
imperativ s se aleag corect rezoluia z pentru Julia specific ce urmeaz s fie renderizat.
8. Rezultate
Am renderizat un mare numr de imagini n scopul evalurii algoritmului post-step.
Imaginile au fost renderizate cu un 3,2GHz Pentium 4 PC cu 512MB RAM sub Windows
XP.
Aa cum se vede din fig. 5,
procedura post-step crete detaliile n mod
considerabil. Fiecare iteraie a fost
renderizat cu rezoluia z = 250 .
Imaginea din stnga, fr post-step
cu 10 pai are defecte cauzate de
neacurateea normalelor, ducnd la pai largi
Fig. 5
ai razelor transversale. n imaginea din
dreapta fractalul este renderizat cu
= 0,803762 0,40615i , n = 8
Cnd se ncearc captura proprietilor
4D ale quaternionilor Julia, se renderizeaz
854
Fig 6.
AFASES - 2008 -
imagini pentru diferite puncte n spaiul 4D. n fig. 6 se poate vedea acelai set Julia cu
diferite poziii w ale camerei i = 0,7323 0,2179i . Imaginea din stnga are camera la
( 2,1,5,0) n timp ce imaginea din dreapta are camera la ( 2,1,5,8) .
n fig. 7 apar dou imagini ale setului Julia
renderizate cu diferite adncimi de iteraie. Cea din
stnga s-a oprit la 8 iar cea dindreapta a mers pn
la 12 pai oferind mai multe detalii. = 0,4 + 0,5i
Figura 8 prezint dou imagini care arat
cum o rezoluie z sczut poate cauza rateuri ale
razelor n setul Julia. n imaginea din stnga
rezoluia z este 2000, pe cnd n cea din dreapta
doar 35. Datorit post-stepping suprafeele n
Fig. 9
ambele figuri sunt netede.
= 0,8 + 0,4i .
Fig 8.
Deasemenea, prezint un exemplu de cod pentru renderizare n POV-Ray a
quaternionilor i interfaa cu imaginea renderizat (fig. 9).
// Persistence Of
Vision raytracer
version 3.5
// -w320 -h320
// -w800 -h800
+a0.3
#include
"golds.inc"
#include
"metals.inc"
#include
"finish.inc"
#declare VP = <2,1.5,1.5>;
#declare VU =
<0,1,0>;
#declare VD =
vnormalize(<0,0,0
> - VP);
#declare VR =
vcross(VU,VD);
#declare ConstC =
<-0.08,0.0,0.83,-0.025>;
#declare
SLICEDIST = 0.1;
camera
Fig. 9
855
AFASES - 2008 -
{ right x
location <-1, 2.8, -3.5>*1.2 look_at y*.5
focal_point 0 blur_samples 40 aperture .35 }
light_source
{ <23,15,-20>, 1
area_light x*5,y*5,12,12 circular orient jitter adaptive 1 }
light_source
{ <23,15,20>, <.3,.4,.5>
area_light x*5,y*5,12,12 circular orient jitter adaptive 1 }
light_source
{ <-15,45,35>, <.8,.6,.3>
area_light x*15,y*15,12,12 circular orient jitter adaptive 1 }
julia_fractal
{ <-.4, .2, .3, -.2>
quaternion sqr
max_iteration 8
precision 50
pigment { rgb .5 }
finish
{ specular 2 roughness .02 metallic
reflection { .2, .8 }}
rotate z*-20
translate y*.9 }
#declare noise3d = function { internal(76) }
#declare Plank=
isosurface
{ function { pow(x, 10)+pow(y*5, 10)-(1-.5*noise3d(x*20,y*10,z)) }
contained_by { box { -<1,1/5,50>, <1,1/5,50> } }
max_gradient 50
}
union
{ #declare Seed = seed(0);
#declare Ind = 0;
#while(Ind < 15)
object { Plank translate<-8+2*Ind, 0, -10*rand(Seed)> }
#declare Ind = Ind+1;
#end
texture
{ pigment
{ wood color_map
{ [0 rgb <.8,.5,.3>][.5 rgb <.7,.45,.2>][1 rgb <.65,.4,.2>]
}
turbulence .5
scale <1,1,10>
}
finish { specular .8 reflection { .1, .3 } }
normal
{ average normal_map
{ [1 bumps .4 scale <.1, .1, 1>]
[1 dents 1 scale .1]}}
}
texture
{ pigment
{ bozo color_map
{ [0 rgb 0 transmit .5]
[.5 rgbt 1]
}
scale 5
warp { turbulence 2 }
scale 1/5}}
}
856
AFASES - 2008 -
Bibliografie
1. Hart J.C., Sandin D.J., Kauffman L.H. Ray Tracing Deterministic 3-D Fractals,
Computer Graphics, Vol 23, 1989.
2. degaard T., Wennergen J. Raytracing 4D Fractals, Institute of Computer
Science, Sweden 2005
3. Radu A.G., Harabagiu A. About the Apollonian Gasket or Curviliniar Sierpinsky
Gasket. Conferina Internaional Nav-Mar-Edu 2007, Constana 2007
4. Radu A.G. Modelare i simulare n 3D. Universitatea Petrol i Gaze, Ploieti, 2006
5. *** http://www.physcip.uni-stuttgart.de/phy11733/index_e.html
6. *** http://www.povray.org
Adrian-Gelu Radu, prof. inf.-fiz., Grup colar Meserii i Servicii Buzu, tel.
0722453007,
e-mail: agerbit@yahoo.com
**
Anca Harabagiu, prof. drd. inf.-fiz., Inspectoratul colar Judeean, Buzu, tel.
0721126577,
e-mail: anca_hara@yahoo.com
857
AFASES - 2008 -
858
AFASES - 2008 -
Abstract:
In this paper it is proposed an operational method for solving difference equations. In
the first part an operational representation of the linear difference equation of the first and
second order is obtained, using lag operator L and, in the second part it is applied to the
solving of linear first order vector difference equation.
Key words: Difference equation.
2000 Mathematics Subject Classification: 39A05, 39A10.
1.
INTRODUCTION
859
AFASES - 2008 -
Therefore, as long as t = t1 , ~
y t is also a solution. Note that t = t1 == t 0 , for
arbitrary initial value 0 = c. Hence, the general solution is
bxt
a
(4)
yt =
+
+ t c ,
1 L 1
where c is an arbitrary constant.
Lets suppose that {xt} is a bounded string and we are interested in finding the bounded
solution {yt}.
a. Then, when || < 1, the bounded string {yt} from (4) can be obtained by following
backward representation with c = 0:
a
a
y t = b(1 + L + (L) 2 + ...) xt +
.
(5)
= b j xt j +
1
1
j =0
b. If || > 1, we need to use forward representation in order to get the bounded string {yt}
as follows:
a
a
(L) 1
yt =
bx
b
j xt j +
+
=
;
(6)
t
1
1
1
1 (L )
j =0
again, setting c = 0.
c. Note that if we dont set c to 0, the string {yt} will diverge, that is, either
lim y t = or lim y t = .
t
bxt
a
+
+ c1t + c 2 tt .
(9)
2
(1 L)
(1 L) 2
Again, if we are interested in a bounded string of {yt} mapped from a bounded string {xt},
then we need to set both cs to zero, and focus on a particular solution.
yt =
.
=
(1 1 L)(1 2 L) 1 2 1 1 L 1 2 L
a1. If |1| < 1 and |2| < 1, we can solve (10) backward so that
860
(10)
AFASES - 2008 -
1
1
1 (1L) j 2 (2 L) j = 1
=
(1 1L)(1 2 L) 1 2 j =0
j =0
1 2
(
j =0
j +1
j +1
)Lj
(11)
yt =
b
1 2
(
j =0
j +1
2 j +1 )Lj xt j +
a
+ c11t + c22t .
(1 1 )(1 2 )
a2. If, without loss of generality, |1| < 1 and |2| > 1, then
1
1 1
L1
=
+
1
(1 1 L)(1 2 L) 1 2 1 1 L 1 ( 2 L)
1 2
(1L) j +
j =0
1 2
( L )
j =1
(12)
b2 j +1
b
a
j +1
yt =
x
+
( xt + j +
+ c11t + c2 2t .
1
t j
1 2 j =0
1 2 j =1
(1 1 )(1 2 )
a3. Last, if |1| > 1 and |2| > 1, then
1
1 L1
L1
=
=
+
1
1
(1 1 L)(1 2 L) 1 2 1 (1 L)
1 (2 L)
1
(13)
( ( j +1) 2( j +1) ) L( j + 2) .
1 2 j =0 1
Plugging these solutions to (8) for each case, we get the solutions for
b
a
yt =
(1 ( j +1) 2 ( j +1) ) xt + j + 2 +
+ c11t + c2 2t .
1 2 j =0
(1 1 )(1 2 )
1
1
=
=
( j + 1)(L) j
(14)
2
(1 1 L)(1 2 L) (1 L)
j =0
and general solution to (7) is
a
yt = b ( j + 1) j xt j +
+ c1 t + c2t t .
2
L
(1
)
j =0
b2. While if || > 1,
1
1
=
=
(15)
( j + 1)(L) ( j + 2) .
(1 1 L)(1 2 L) (1 L) 2 j =0
Plugging these solutions to (9) for each case, we get the solutions for
a
yt = b ( j + 1) ( j + 2) xt + j + 2 +
+ c1 t + c2t t .
2
(1 L)
j =0
y t:
yt :
c. Complex Eigenvalues
It should be noted that if one eigenvalue turns out to be a complex number, then the other
eigenvalue is the complex conjugate of it, that is, 1 = + i and 2 = i , where
861
AFASES - 2008 -
(
j =0
j +1
j +1
)Lj =
1 j
r sin( ( j + 1)) L j
sin j =0
sin j =0
(1 1 )(1 2 )
c2. If || > 1, from (13),
1 j
1
1
j +1
j +1
r sin( ( j + 1)) L j .
=
(1 2 )Lj =
sin j =0
(1 1 L)(1 2 L) 1 2 j =0
(16)
(17)
Plugging these solutions to (8) for each case, we obtain the solutions for yt.
b ( j + 2)
a
yt =
r
sin( ( j + 1)) xt + j + 2 +
+ c11t + c2 2t .
sin j =0
(1 1 )(1 2 )
2.
T
T21 T22
0 2
z 2t
T
11
where y1t, a1 and z1t are (k l) 1 vectors, F11, T11, 1, T are (k l) (k l) matrices, and
B1 is (k l) m. Then, the first (k l) equations can be written as
z1t = 1 z1,t 1 + (T 11 B1 + T 12 B2 ) xt + (T 11 a1 + T 12 a 2 )
(21)
and the remaining l equations are
z 2t = 2 z 2,t 1 + (T 21 B1 + T 22 B2 ) xt + (T 21a1 + T 22 a2 ) .
862
(22)
AFASES - 2008 -
Since first set of the equations in (21) has the roots less than the unity in norm, we can
solve it backward, whereas equations in (22) can be solved forward yielding following
particular solutions:
z 1t =
j =0
j
1
(T 11 B1 + T 12 B 2 ) x t j + ( I 1 ) 1 (T 11 a1 + T 12 a 2 ) ,
z 2t = 2( j +1) (T 21 B1 + T 22 B 2 ) x t +1+ j ( I 21 ) 1 21 (T 21 a1 + T 22 a 2 ) .
(23)
(24)
j =0
For general solutions, we need to add ci ti , i=1,2,.k, for each equation, where cis are
any constants. However, since we are interested only in bounded strings, we will ignore these
terms by setting ci = 0, i. Once we obtain solutions by (23) and (24), the solutions for the
original variables in yt in (18) can be recovered by yt = Tzt. That is
y1t T11 z1,t + T12 z 2,t
.
=
(25)
0 J 2 ... 0 ,
(26)
J =
M M O
0 0 ... J
s
where s < k is the number of distinct eigenvalues and Ji is a (ni ni) matrix with following
form where ni is the number of times that the eigenvalue i is repeated
i 1 L 0
0 i L 0
,
(27)
Ji =
M M O 0
0 0 L
i
suppose n1 = 3, n2 =5, , n s = k ni . Note that z3t , the last row variable in J1, satisfies
1
i =1
863
AFASES - 2008 -
yt = Mzt .
(29)
(30)
Such a scalar difference equation can be represented by a similar vector difference equation
as in (18),
yt= Fyt1+bxt+a, t,
(31)
where xt is a scalar, and
yt
[a1 a 2 L a p ]
y t 1
;
yt =
F =
; ( pxp
)
I
M
0
M
( px1)
(p -1)(p -1)
(p-1)(p-1)
y
t p +1
y t 1
b
a
yt 2
0
0
=
=
y t 1 =
b
a
;
;
(32)
M
M .
( px1)
( px1)
M
( px1)
0
0
y
t p
In order to solve this system of equations by methods described before, we need to obtain the
eigenvalues of the matrix F. Note that the eigenvalues are the solutions from |FIp| = 0(pp).
It is easy to show that the eigenvalues s of F satisfy the following characteristic equation
p a1p1 a2p2 ap1 ap = 0 .
(33)
th
We can solve following equivalent p order polynomial equation
1 a1z a2z2 apzp = 0,
(34)
where the solutions to (34) zs are the inverses of s. Once we obtain the eigenvalues, it is
straightforward but tedious to solve the system by aforementioned methods.
3.
REFERENCES
[1]
Gh. Anton and Gh. Radu, Ecuaii ordinare cu diferene i aplicaii, Editura Albastr,
Cluj-Napoca, pag.36-75, 1998
Gh. Anton, On Operational Method for the Solving of Difference equations, Sesiunea
tiinifica anual ,,Mathematics And Its Applications dedicat Centenarului Grigore
C. Moisil, Universitatea Transilvania, Braov, 12 13 Mai 2006
Gh. Anton, Gh. Radu, T. Socaciu, Ecuaii cu diferene i scheme de calcul, Editura
Albastr, Cluj-Napoca, 2008
Gh. Radu, Gh. Anton, Asupra unor ecuaii cu diferene de ordinul doi, Proceedings
of International Conference on Naval and Marine Education Romanian Naval
Academys 130th Anniversary, Section V: Mathematics and Computer Science, pp.
31-38, Constana, November 14-16, 2002
Gh. Radu, Gh. Anton, On Operational Method for the Solving of Difference
Equations, International Multidisciplinary Scientific Symposium UNIVERSITARIA
SIMPRO, Petroani, 13-14 octombrie 2006, vol. Matematic-Informatic i Fizic,
pp. 7-10, 2006
[2]
[3]
[4]
[5]
Gheorghe Radu, lecturer, Henri Coanda Air Force Academy, 500183 Brasov, #160, Mihai Viteazul Street,
phone: +40 723 604084, e-mail: gh.radu@gmail.com
Gheorghe Anton, teacher, Trades and Services High School, Buzau, 14 bis, Bazalt Street,
g10.anton@gmail.com
864
AFASES - 2008 -
I. INTRODUCERE
Sistemele de baze de date sunt o component esenial a vieii de zi cu zi n societatea
modern. n cursul unei zile, majoritatea persoanelor desfoar activiti care implic
interaciunea cu o baz de date: depunerea sau extragerea unor sume de bani din banc,
rezervarea biletelor de tren sau avion, cutarea unei referine ntr-o bibliotec computerizat,
cumprarea unor produse etc.
Bazele de date pot avea dimensiuni (numr de nregistrri) extrem de variate, de la
cteva zeci de nregistrri (de exemplu, baza de date pentru o agend cu numere de telefon)
sau pot ajunge la zeci de milioane de nregistrri (de exemplu, baza de date pentru plata
taxelor i a impozitelor).
Utilizatorii unei baze de date au posibilitatea s efectueze mai multe categorii de
operaii asupra datelor memorate:
introducerea de noi date (insert);
tergerea unora din datele existente (delete);
actualizarea datelor memorate (update);
interogarea bazei de date (query) pentru a regsi anumite informaii, selectate dup
un criteriu ales.
n sensul cel mai larg, o baz de date (database) este o colecie de date corelate din
punct de vedere logic, care reflect un anumit aspect al lumii reale i este destinat unui
865
AFASES - 2008 -
anumit grup de utilizatori. n acest sens, bazele de date pot fi create i meninute
manual (de exemplu, fiele de eviden a crilor dintr-o bibliotec, aa cum erau folosite cu
ani n urm) sau computerizat, aa cum este majoritatea bazelor de date folosite astzi.
O baz de date (database) este o colecie de date creat i meninut computerizat,
care permite operaii de introducere, tergere, actualizare i interogare a datelor.
Simple colecii de fie (documente pe hrtie) sau fiiere de date, care conin
nregistrri de date, dar nu permit operaii de interogare, nu sunt considerate baze de date.
II. COMPONENTELE UNUI SISTEM DE BAZE DE DATE
Un sistem de baze de date (Database System) este un sistem computerizat de
meninere a evidenei unei anumite activiti, folosind baze de date. Componentele unui
sistem de baze de date sunt: hardware, software, utilizatori, date persistente.
Hardware. Sistemele de baze de date sunt instalate, de regul, pe calculatoare de uz
general, de la calculatoare PC standard, pn la staii multiprocesor puternice.Performanele
generale de operare ale calculatorului (numrul i viteza procesoarelor, dimensiunea i viteza
de operare a memoriei principale etc.) influeneaz n mod corespunztor performanele
sistemului de baze de date. Ceea ce intereseaz n mod deosebit n utilizarea unui calculator
pentru un sistem de baze de date, este volumul (capacitatea) memoriei secundare, utilizat
pentru memorarea coleciei de date persistente ale bazei de date.
Software. ntre baza de date (colecia de date memorate fizic n fiiere pe hard-discuri)
i utilizatorii sistemului exist un nivel software, numit Sistem de Gestiune a Bazei de Date
(SGBD) - (Database Management System -DBMS) - (fig. 1.1).
Sistemul de gestiune a bazei de date - SGBD - (Database Management System DBMS) recepioneaz cererile utilizatorilor de acces la baza de date (pentru operaii de
introducere, tergere, modificare sau interogare), le interpreteaz, execut operaiile
corespunztoare i returneaz rezultatul ctre utilizatori.
Sistemul SGBD ofer utilizatorilor o viziune (vedere - view) a bazei de date la un nivel
nalt i i elibereaz de necesitatea de a cunoate organizarea particular a sistemului
(driverele de disc, structura nregistrrilor de date etc.). Mai mult, sistemul de gestiune asigur
protecia datelor fa de accese neautorizate sau defecte de funcionare, asigurnd integritatea
bazei de date. Pe lng SGBD, care este cea mai important component software a unui
sistem de baze de date, exist i alte componente: sistemul de operare, care asigur controlul
execuiei programelor, biblioteci i instrumente software (toolset-uri) pentru proiectarea,
dezvoltarea sau exploatarea sistemelor de baze de date i a aplicaiilor de baze de date.
O aplicaie de baze de date (Database Application) este un program care ofer o
anumit utilizare unei baze de date. Exemplu, programul care permite monitorizarea
activitii angajailor unei firme (ncadrare, calificare, salarizare etc.), folosind informaiile
despre angajai memorate ntr-o baz de date, reprezint o aplicaie de baze de date.
Utilizatorii unui sistem de baze de date se pot mpri n cteva categorii:
programatorii de aplicaii, utilizatorii finali i administratorul bazei de date.
866
AFASES - 2008 -
867
AFASES - 2008 -
868
AFASES - 2008 -
Sistemele de baze de date distribuite pot fi reprezentate ntr-un mod asemntor din
perspectiva structurrii client-server (fig. 1.4).
O baz de date distribuit este o colecie de date care aparin din punct de vedere logic
aceluiai sistem, dar care pot s fie din punct de vedere fizic, memorate n mai multe staii de
calcul (locaii - sites) conectate printr-o reea de comunicaie. Sistemul software care
869
AFASES - 2008 -
870
AFASES - 2008 -
relativ la domeniul de interes pentru care se definete modelul respectiv, astfel nct s asigure
diferenierea acelei entiti fa de restul universului.
Toate entitile similare, care pot fi descrise prin aceleai atribute, aparin unui acelai
tip de entitate (entity type), iar colecia tuturor entitilor de acelai tip dintr-o baz de date
constituie o mulime de entiti (entities set). n general, n modelul E-A se folosete aceeai
denumire att pentru un tip de entitate ct i pentru mulimea entitilor de acel tip.
Prin analogie cu modelul obiect, se poate spune c un tip de entitate corespunde unei
clase, o entitate este o instan a unui tip de entitate i corespunde unui obiect, iar mulimea
entitilor de un tip dat corespunde mulimii obiectelor (instanelor) unei clase.
O asociere (relationship) este o coresponden ntre entiti din dou sau mai multe
mulimi de entiti.
Gradul unei asocieri este dat de numrul de mulimi de entiti asociate. Asocierile
pot fi binare (de gradul 2, ntre 2 mulimi de entiti) sau multiple (ntre k mulimi de entiti,
k>2).
Asocierile binare sunt, la rndul lor, de trei categorii, dup numrul elementelor din
fiecare dintre cele dou mulimi puse n coresponden de asocierea respectiv (fig. 1.5). Fiind
date dou mulimi de entiti, E1 i E2, se definesc urmtoarele categorii de asocieri binare:
Asocierea unul-la-unul (one-to-one) este asocierea prin care unui element
(entitate), din mulimea E1 i corespunde un singur element din mulimea E2 i reciproc; se
noteaz cu 1:1.
Asocierea unul-la-multe (one-to-many) este asocierea prin care unui element din
mulimea E1 i corespund unul sau mai multe elemente din mulimea E2, dar unui element din
E2 i corespunde un singur element n mulimea E1; se noteaz cu 1:N.
Asocierea multe-la-multe (many-to-many) este asocierea prin care unui element
din mulimea E1 i corespund unul sau mai multe elemente din mulimea E2 i reciproc; se
noteaz cu M:N. Cardinalitatea (multiplicitatea) unei asocieri fa de o mulime de entiti
(cardinality, multiplicity), este numrul maxim de elemente din acea mulime care pot fi
asociate cu un element din alt mulime a asocierii.
Fig. 1.5. Categorii de asocieri ntre dou mulimi de entiti: a - asociere 1:1; b - asociere 1:N; c- asociere M:N.
871
AFASES - 2008 -
872
AFASES - 2008 -
Este evident asemnarea dintre ierarhiile de tipuri de entiti din modelul E-A extins
i ierarhiile de clase din modelul obiect-orientat, dar modelul E-A extins este un model de date
mult mai general (de nivel nalt), care poate fi transpus n diferite modele de date specializate,
inclusiv modelul obiect-orientat.
4.3. Modelul de date ierarhic
n modelul ierarhic (Hierarchical Model) o baz de date se reprezint printr-o
structur ierarhic de nregistrri de date (records) conectate prin legturi (links).
Modelul de date ierarhic a fost primul model folosit pentru dezvoltarea bazelor de
date. Cea mai cunoscut realizare de SGBD ierarhic este sistemul IMS (Information
Management System) dezvoltat de firma IBM n cadrul programului de cercetri Apollo, n
perioada anilor 1960.
O nregistrare de date n modelul ierarhic este o instan a unui tip de nregistrare
(record type) i const dintr-o colecie de cmpuri (fields), fiecare cmp coninnd valoarea
unui atribut. Un tip de nregistrare corespunde unui tip de entitate, iar o nregistrare
corespunde unei entiti din modelul E-A.
Un tip de legtur n modelul ierarhic este un tip de asociere cu raportul de
cardinalitate 1:N (de tip printe-fiu) ntre dou tipuri de nregistrri. Tipul de nregistrare din
partea cu multiplicitatea 1 a asocierii este numit tip de nregistrare printe, iar tipul din partea
cu multiplicitatea N a asocierii este numit tip de nregistrare fiu.
Schema conceptual a unei baze de date n modelul ierarhic se reprezint printr-un
numr oarecare de scheme ierarhice (fig. 1.10). O schem ierarhic este un arbore direcionat,
reprezentat pe mai multe niveluri, n care nodurile sunt tipurile de nregistrri, iar arcele sunt
tipurile de legturi. Fiecare nod (cu excepia nodului rdcin), are o singur legtur ctre un
nod de pe un nivel superior (nodul printe) i fiecare nod (cu excepia nodurilor frunz) are
una sau mai multe legturi ctre noduri de pe nivelul imediat inferior (noduri fii).
Se poate stabili o coresponden ntre o schem conceptual ierarhic i o diagram EA: tipurile de nregistrri corespund tipurilor de entiti, iar tipurile de legturi corespund
tipurilor de asocieri. (fig. 1.10, a i b).
n modelul ierarhic nu sunt admise dect legturi de tipul printe-fiu, care corespund
asocierilor 1:1 i asocierilor 1:N din modelul E-A. Asocierile M:N din modelul E-A nu se pot
reprezenta n mod direct n modelul ierarhic, ci numai prin multiplicarea nregistrrilor de tip
fiu, atunci cnd sunt referite de mai multe nregistrri de tip printe. Acest lucru conduce la o
mare redundan a datelor. Corespunztor schemei ierarhice a unei baze de date se pot
reprezenta mai muli arbori de instaniere a datelor, care sunt, de asemenea, arbori direcionai
873
AFASES - 2008 -
(fig. 1.10, c). Fiecare arbore de instaniere conine ierarhii pe mai multe niveluri de
nregistrri ntre care exist legturi de tipul printe-fiu.
Fig. 1.10. Baz de date ierarhic: a - diagrama E-A; b - schema conceptual a bazei de date ierarhice; c - arbori
de instaniere a datelor.
Ioan-Gheorghe Raiu, lector universitar, dr. ing., Universitatea George Bariiu, Braov, Str. Lunii Nr.6, tel.
0740-080-463, e-mail: ratiu_2000@yahoo.com
874
AFASES - 2008 -
Abstract. This paper contains a description of K-NN (K Nearest Neighbor) based method - that can be used for univariate time series prediction. The given
method is inspired by machine learning techniques. Experimental results and comparisons with the predicted values obtained by ARIMA show a good behavior of
this approach.
Key words: ARIMA, K Nearest Neighbor, Prediction.
2000 MSC : 91B70, 91B84, 60G25.
Introducere
In acest articol voi prezenta o alta metoda de predictie pentru serii de timp bazata
pe metoda celor mai apropiati k vecini, cunoscuta n literatura de specialitate ca metoda
k-nearest neighbor (k-NN). Este o metoda simpla din punct de vedere conceptual, dar
care si-a dovedit utilitatea n rezolvarea problemelor de clasificare si de regresie. Desi
conceputa pentru clasificare, ea poate fi extinsa pentru efectuarea de predictii de serii de
timp. Astfel de variatii sunt descrise de exemplu n [2], [3], [8], etc.
875
AFASES - 2008 -
ca n (1) si (2):
v
uX
u p
dE (x, xi ) = t (xij xj )2
(1)
j=1
si respectiv
dL1 (x, xi ) =
p
X
|xij xj |.
(2)
j=1
(3)
Rn
876
AFASES - 2008 -
Daca notam cu P valoarea minima posibila pentru Pn (e), atunci n [1] se arata ca:
m
P .
(4)
P P = lim Pn (e) P 2
n
m1
Proprietati similare sunt enuntate pentru k > 1.
k-NN poate fi folosit pentru regresie, n cazul n care vectorilor xi le sunt asociate
valori reale. Valoarea asociata vectorului x se calculeaza ca medie a valorilor asociate
celor mai apropiati k vecini:
k
X
yo(i)
.
(5)
y=
k
i=1
Predictia are acuratete mai mare daca se foloseste ponderarea valorilor vecinilor astfel
ncat formula (5) devine:
k
P
wo(i) yo(i)
i=1
(6)
y=
k
P
wo(i)
i=1
unde wo(i) este o functie nedescrescatoare n raport cu distanta dintre x si xo(i) . Ponderarea arata ca vecinii cu apropiere mare de vectorul x au un impact mai mare n calcularea
valorii de iesire asociate lui x.
In practica s-au dovedit eficiente folosirea ponderilor invers proportionale cu distanta
la vecini:
1
, i {1, 2, . . . , k}.
(7)
wo(i) =
d(x, xo(i) )
Pentru d(x, xo(1) ) = . . . = d(x, xo(k) ) = constant formula (6) devine (5).
Pentru folosirea lui k-NN ntr-o problema de predictie pe serii de timp univariate,
trebuie sa avem n vedere urmatoarele:
1. definirea strategiei de cautare a vecinilor cei mai apropiati;
2. construirea valorilor prezise pentru seria de timp.
Rezultate experimentale
Aceasta sectiune contine niste studii de caz pe serii de timp monovariabile care provin
din situatii reale si reprezinta evolutia unor procese din domeniul economic. Un rol
important n stabilirea modelului final al seriei de timp analizate l are masura de acuratete
MAPE definita prin ecuatia (8).
n
1 X |Yt Yt |
100, (Yt 6= 0),
MAP E =
n t=1 |Yt|
(8)
877
AFASES - 2008 -
Seria
ARIMA k-NN
MAPE MAPE
S
omaj
8.60
6.28
V
anzare NSA
1.65
1.58
V
anzare SA
1.32
0.66
Constructie
5.02
4.10
Educatie
0.62
0.46
Export
8.16
6.99
Gaz
19.22
10.9
Tabela 1: Rezultatele obtinute cu ARIMA, respectiv cu k-NN
Datele seriilor de timp analizate au fost luate de pe site-urile www.economagic.com si
http://research.stlouisfed.org/fred2/series si fac parte din domeniul economic.
Din experimentele realizate si prezentate n tabela 1 putem extrage urmatoarea concluzie: de fiecare data, k-NN duce la valori MAPE mai bune decat modelarea prin ARIMA.
Aceste valori ale lui MAPE sunt cu 4 50% mai mici decat valorile lui MAPE pentru
ARIMA.
Bibliografie
[1] Duda, O., R., Hart, E., P., Stork, G., D., Pattern Clasification, Second Edition,
John Wiley & Sons, Inc., USA, 654p., 2001
[2] Oswald, R.K., Scherer, W.T., Smith, B.L., Traffic Flow Forecasting Using Approximate Nearest Neighbor Nonparametric Regression, A Research Project Report No.
UVACTS-15-13-7 For the National ITS Implementation Research Center, December
2001
[3] Plummer, E.A., Time Series Forecasting with Feed-Forward Neural Networks guidelines and limitations, Master Thesis, 2000
[4] Sasu, A., An Application of ARIMA Models, Bulletin of the Transilvania University
of Brasov 11(46), Series B1. Transilvania Univ. Press-Brasov, 61-68, ISSN 1223964X, 2005
[5] Sasu, A., Trends in Economic Time Series, 7-th European Conference E-COMMLINE 2006, Bucuresti, C7-60-06, ISBN 973-88046-0-4, ISBN 978-973-88046-0-9,
2006
[6] Sasu, A., Time series forecasting methods, Sesiunea de comunicari stiintifice cu
participare internationala, Academia Fortelor Aeriene Henri Coanda, BrasovRomania, ISBN 978-973-8415-45-4, 2007
[7] Sasu, A., Time Series Forecasting using Arima Models, 8-th European Conference
E-COMM-LINE 2007, 20-22 Sept., Bucuresti, ISBN-10: 973-88046-6-3, ISBN-13:
978-973-88046-6-1, 67e,4 pg., 2007
[8] Yakowitz, S., Nearest-neighbor methods for time-series analysis. Journal of Time
Series Analysis, Vol. 8, No. 2, 235-247, 1987
878
AFASES - 2008 -
879
AFASES - 2008 -
Cele dou constrngeri sunt impuse de utilizarea intensitii luminoase nenegative. Intervalul
G=Bmax-Bmin este cunoscut sub denumirea de gam dinamic. Potrivit constrngerii 2, fie
suportul de dimensiune N M . El va fi eantionat cu pasul l0 , uniform pe ambele axe. Fie
N, M numere naturale, astfel nct N l0N i Ml0M. Semnalul imagine numeric este
descris de funcia
u: D u G , ( i, j) u( i, j), i = 1K N , j = 1K M
(4)
iar NM este domeniul 2-conex, ce constituie suportul ei.
2. Metode de calcul a funciei de intercorelaie bidimensional
xy [i, j] =
x[ n, m] y[ n + i, m + j]
(5)
n = m=
xy [i, j] =
N 1 N 1
x[ n, m] y[ n + i, m + j]
(6)
n = 0 m= 0
Domeniul de valori al funciei xy[.] este 2N 2N. Calculul direct presupune efectuarea a N
multiplicri i N-1 adunri de N ori i N-1 adunri. Operaia se repet de 4N2 ori,
corespunztor tuturor perechilor (i,j) de pe suportul funciei, iar timpul de calcul total va fi:
t cd = 4 N 2 N 2 t m + ( N 1) 2 t a = 4 N 4 t m + [2 N ( N 1)] t a
(7)
unde tm i ta corespund timpului n care este efectuat o multiplicare, respectiv o adunare. Se
cunoate i o cale indirect de calcul.
Expresia transformatei Fourier bidimensionale directe a semnalului u[.] calculat dup un
algoritm rapid este dat de [B1]:
U[ k, l ] =
S1 S1
u[ n, m] w kn wlm
n = 0 m= 0
880
(8)
AFASES - 2008 -
2
2r
unde w = e
pentru r = log 2 N
(9)
iar k,l sunt frecvene spaiale. Expresia transformatei inverse este:
1 S1 S1
u[ n, m] = 2 U[ k, l ] w kn w lm
(10)
S k=0 l=0
Aplicnd transformata Fourier bidimensional direct asupra relaiei, schimbnd ordinea de
sumare, grupnd termenii i apoi prin intermediul unei schimbri de variabil, se poate pune
n eviden relaia:
xy [ k, l ] = X*[ k, l ] Y[ k, l ]
(11)
Ca urmare, funcia de intercorelaie va avea expresia:
1 2 N 1 2 N 1 *
xy [i, j] =
X [ k, l ] Y[ k, l ] w ki w lj (12)
2
4N k=0 l=0
Calculul transformatei Fourier bidimensionale directe a unui semnal M M presupune
calculul a M transformate Fourier rapide (FFT 2 ) (n timpul tFFT n cazul algoritmului
monodimensional), adic:
t TFBD = 2 M t FFT
(13)
Multiplicarea punctual a dou matrici complexe de dimensiune M M presupune un
consum de timp de
(14)
t pm = 4 M 2t m + 2 M 2t a
Considernd durata de execuie a mpririi egal cu cea a nmulirii, timpul de calcul al
transformatei Fourier bidimensionale inverse va conine i termenul corespunztor normrii.
Astfel, pentru semnalul definit pe 2N 2N, calculul prin metoda indirect se realizeaz n:
t ci = 2t TFBD + t pm + tTFBI =
8 Nt FFT + 16 N 2 t m + 8 N 2 t a + 4 Nt FFT + 4t m =
(15)
= 12 Nt FFT + 4 4 N 2 + 1 t m + 8 N 2 t a
Lund n considerare timpul necesar executrii transformatei Fourier rapide a unei secvene
de lungime 2N, i anume N(log2N-1)tm i [N(log2N-1)-1]ta, se obine:
tci = 24 N 2 log 2 N + 16 N 2 + 4 t m + 24 N 2 log 2 N 16 N 2 t a (16)
Metoda direct se dovedete prohibitiv din punct de vedere al costului, fiind practicabil
numai pn la dimensiuni de cel mult 64 64 .
Volumul de calcul este ridicat n cazul n care unul din semnale este un detaliu mic din
cellalt, ntruct zona valorilor nule obinute prin proiecia pe suportul 2N este foarte ntins.
Presupunem n continuare c semnalul x are eantioane nenule, ncepnd de la
eantionul d, astfel nct + d<N. n acelai mod, semnalul y are eantioane nenule,
<<. Funcia xy[.] va avea 2(+-1) 2(+-1) eantioane nenule, ncepnd de la
eantionul (d+d). Prin limitarea suportului, semnalele x i y au devenit periodice spaial:
suportul finit acioneaz ca o fereastr bidimensional aplicat semnalului imagine infinit.
Dac (+-1) N, nu vor apare intervale de interferen. xy[.] poate fi obinut n ntregime
sau parial extrgnd perioada N N pornind din punctul (d,d).
Fie =N i d =0. Evident, (+-1) =L N. Rezultatul neafectat de erori se va regsi n
intervalele 2N-L, ncepnd din punctul (L-N,L-N). Dac d =L-N (d + << N) rezultatul este
corect pe ntreaga perioad 2N. mprim suportului N n K intervale de lungime P, ultimul
fiind Pr<P astfel nct N=KP+Pr . Produsul de intercorelaie determinat pentru primul
881
AFASES - 2008 -
0
w1
w2
L w N 1
w
2 N 1)
(18)
W = w 0
w2
w4
L w (
M
M
O
M
M
0
2( N 1)
N 1)2
N 1
(
L w
w
w
w
Semnalele x[.] i y[.] se proiecteaz pe un suport 2N 2N prin completarea cu zerouri.
Calculul (5) prin metoda indirect, scris sub form matricial este:
XY = X*Y
( ) = W 1( X*Y)(W 1)
H
*
= W 1 ( WxW H ) ( WyW H )( W 1 )
xy = W 1 XY W 1
= (19)
Imaginea este o funcie real cu valori strict pozitive i se poate realiza o cretere a vitezei de
calcul a funciei de intercorelaie pentru metoda indirect (pe baza linearitii transformatei
Fourier).
Fiecare rnd de ordin impar se scrie ca parte imaginar a rndului anterior (primul este 0),
calculndu-se apoi FFT pentru fiecare semnal obinut. Rezultatul este disponibil dup un
timp de NtFFT.
Se construiete matricea spectrului spaial al liniilor utiliznd relaii similare celor prezentate
n [B2]:
882
AFASES - 2008 -
(20)
Re[ X l +1 [k , p ]] = {Im[ X [k , p ]] + Im[ X [2 N k , p ]]} 2
Im[ X l +1 [k , p ]] = {Re[ X [k , p ]] Re[ X [2 N k , p ]]} 2
l = 2 p, p = 0,K, N 1
unde p este indicele de ordine al transformatei (perechii de linii par-impar). n continuare,
aceast matrice va fi transpus i se vor calcula transformatele monodimensionale pentru
fiecare linie. Rezultatul conine numai numere complexe, nefiind aplicabil o nou reducere
de calcul dup transpunere. Al doilea semnal este transformat ntr-un mod similar. Timpul
consumat pn la aceast etap corespunde valorii de 2(NtFFT+2NtFFT). Se execut
multiplicarea punct cu punct a celor dou semnale n domeniul de frecven spaial, operaie
ce presupune costul dat de (7), dup care se revine n domeniul spaial prin aplicarea unei
transformate Fourier bidimensionale inverse calculat dup un algoritm rapid. n final, costul
total de calcul va fi:
t ci = 2( Nt FFT + 2 Nt FFT ) + t pm + t TFBI =
(21)
= 6 Nt FFT + 16 N 2 t mc + 8 N 2 t ac + 4 Nt FFT + 4t mc =
= 10 Nt FFT + 4 4 N 2 + 1 t mc + 8 N 2 t ac
+ 4 4 N 2 + 1 tm + 8N 2ta =
(22)
= 20 N 2 log 2 N + 16 N 2 + 4 t m + 20 N 2 log 2 N 12 N 2 t a
Potrivit valorilor din tabelul 1, timpul de calcul prin metoda direct este adus la o valoare
comparabil cu cea realizat prin metoda indirect.
3. Concluzii
883
AFASES - 2008 -
884
AFASES - 2008 -
p(m, 0) = 0
p(0, n) = for n > 0, and 0 for n = 0
p(m, n + 1) = min {1 + 2 * p(m, k) + p(m - 1, n - k) | k = 0, n}
885
AFASES - 2008 -
Aceasta formula pare mult mai atractiva deoarece complexitatea unui algoritm de programare
dinamica bazat pe aceasta idee este mai de complexitat mai scazuta decat cele bazate pe
vechea recurenta (vezi [3]).
Pentru a evalua valorile lui p() cu
(m,n) [0 .. Maxm] x [0..Maxn]
putem folosi urmatorul algorithm (vezi [1]):
Algorithm A2
For m = 0 to Maxm do
For n = 0 to Maxn do
If n = 0 then
P[m,n] := 0
Elsif m = 0 then
P[m,n] := infinite
Else
X := infinite
For k := 0 to n-1 do
Y := 1 + 2 * p[m,k] +
p[m-1,n-k-1]
If Y < X then
X := Y
End If
End For
P[m,n] := X
End If
End For
End For
End Algorithm
Teorema: Complexitatea timpului de executie a lui A2 este (Maxn2 x Maxm).
Demonstratie: vezi [2]
IV. Al treilea algoritm (bazat pe o conjectura)
Teorema: Daca notam
a) A(i, j) = (i+j-2)! / ((i-1)! (j-1)! )
b) s(m,n) acel numar pentru care
A(s(m, n), n) <= m < A(s(m, n) + 1, n)
atunci:
a) s(m,n) este binedefinit
b) A(m,n-1) + A(m-1, n) = A(m,n)
Demonstratie: vezi [6].
886
AFASES - 2008 -
887
AFASES - 2008 -
Teorema: Complexitatea din punct de vedere al timpului de executie a lui A3 este O(Maxn x
Maxm2).
Demonstratie: vezi [2]
V. Concluzie
Tipul
algoritmului
Serial
Notatia
Landau
utilizata la
specificarea
complexitatii
A1
A2
Serial
Maxn2 x Maxm
A3
Serial
Maxm2 x Maxn
Bibliografie
[1] Tiberiu Socaciu, A Conjecture of Generalized Hanois Problem, in New Chalenges in the
Field of Military Sciences, Proceedings IV: Informatics, Communication and Applied
Mathematics, Budapest, Hungary, 13-14 november 2007, pp. 103-106 (ISBN 978-963-877062-2).
[2] Tiberiu Socaciu, Some Algorithms for Generalized Hanoiss Problem, submitted to
WSEAS Transactions on Computers.
[3] Tiberiu Socaciu, A Conjecture of Generalized Hanois Problem, at New Challenges in the
Field of Military Sciences 2007, International Conference, 13 - 14 November 2007, Budapest,
Hungary.
[4] Tiberiu Socaciu, Ioan Maxim, Antoanela Naaji, Claudiu Persa, Conjectures about
Generalized Hanois Problem, Conferina Naional de Matematic i Informatic, CNMI
2007, Bacu, 16-17 noiembrie 2007
[5] Tiberiu Socaciu, O extindere a problemei Turnurilor din Hanoi, in Gazeta de Informatica,
nr. 11-12/1994, pp. 19-27. (An extension of Hanois Problem)
[6] Tiberiu Socaciu, Problema Generalizata a Turnurilor din Hanoi, in GInfo, nr. 17/3, pp. 2629. (Generalized Hanois Problem)
[7] Tiberiu Socaciu, Ioan Maxim, Antoanela Naaji, Claudiu Persa, On Prooving a Conjecture
of Generalized Hanois Problem with Parallel Architectures, in The Knowledge Based
Organization Conference Procedings 11, Sibiu, 22-25 november 2007, pp. 133-136 (ISSN
1843-6722).
[8] Tiberiu Socaciu, Invatati Backtracking. Exemple in Pascal si C, Editura InfoData Cluj
2006, ISBN 973-87774-1-0, 160 pagini (Learning Backtracking. Examples in Pascal and C).
888
AFASES - 2008 -
Tiberiu Socaciu, lector universitar, Universitatea Stefan cel Mare Suceava, FSEAP, socaciu@seap.usv.ro
Ioan Maxim, lector universitar dr, Universitatea Stefan cel Mare Suceava, DPPD, maximioan@yahoo.com
**
889
AFASES - 2008 -
Alexandru STEFAN *
ABSTRACT
e-Learning continues to grow at a tremendous rate. E-learning strategists redict that
by the year 2008, more than half of all training may be online.
E-learning companies are springing up everywhere. It seems as though you cant pick
up a business or training magazine without seeing articles about the benefits or the
problems that are a result of e-learning. The field is growing at an amazing rate and its
standards have yet to be developed or even agreed upon. So how in the world does a
training department go about implementing an e-learning program in an organization? One
way is to develop a strategy for creating e-learning courses that can serve as a guide or
road map as you are working your way through the chaos. It is essential to link e-learning
goals to business goals to ensure the ultimate success of the entire e-learning program.
Introduction
The increasing availability of the Internet and computer technologies we are now able
to change what and how we deliver training and instruction to learners separated by time or
space. Thus eLearning has begun to evolve.
One of the most important questions to ask when considering eLearning is "What
does eLearning change?"; in Horton's book he suggests that most importantly it removes
barriers of time and space, changes the economics of delivering training, and also changes
how we can deliver training.
The field of eLearning is growing at an amazing rate and its standards have yet to be
developed or even agreed upon.
So, how does a training department go about implementing an eLearning program in an
organization?
One way is to develop a strategy for creating eLearning courses that can serve as a
guide or road map as you are working your way through the chaos.
To create your e-learning strategy, you need to:
Link eLearning goals with business goals
Ensure support from top management
890
AFASES - 2008 -
Desig
n
Test
Analyze
Re-Build
Re-Design
Re-Analyze
891
AFASES - 2008 -
892
AFASES - 2008 -
NO
DO YOU HAVE IT?
YES
ACHIEVE
PRESERVE
eLearning training packages with three levels Creativity,
experience
and
of tuition:
enthusiasm of the current staff
No support
Opportunities to inform and
Tutoring: tuition+10%
involved customers
in
Tutoring & Labs + 20%
course selection
Monthly/ 6 month/annually fee to access Improving
online response
training programs ( networking, management, time and effectiveness
web design )
Positive
image
of
the
Develop short term discounts for training organization
packages: sale events
Diversification of services:
Customize web base training package for online business activities
corporate;
Flexibility to respond to
Design new e-learning courses
different needs
Custom Online Course Authoring
Updating course tuition with
Customized student page
regarding the market
Create marketing plans for each tool that
Define target customers
clarified target customers and the specific,
Develop international market by
quantifiable value proposition for the elocalizing the business
Commerce tool;
projects
and
Business version for our online International
affiliation under international
training site
universities
Attract revenue by adds
YES
YES
ELIMINATE
Lower operating costs
Barriers to progress
The barriers to degree programs
NO
NO
YES
893
AFASES - 2008 -
Bibliography
1. The DNA of eLearning, excerpt from Beyond eLearning, Internet Time Group,
www.internettime.com
2. Pam Pervenanze, LearnSource- Connecting Companies with Quality Learning
Solutions and Pinneast, CeleXxs Performance Media Group
3. Building a Business Case for eLearning, www.geolearning.com
4. ITPS-Online Solution, http://online.bloomfield.edu, USA
5. Aligning eLearning with Business Goals & Measuring Results, Spring 2008, MBA
Program, Poly University
894
AFASES - 2008 -
895
AFASES - 2008 -
( e , )
( e , )
. n aceste condiii introducnd variabila de ecruisare = ( , ) din cele
e
de mai sus obinem c variabila de ecruisare izotrop q i tensiunea anterioar q sunt:
1
q = 2 K + H = K
(2)
q = 1 K + H = H
i =
q = H
p
= C ( )
(3)
f ( , ) =
K ( ) cu n =
(4)
3
sn +1 q n +1 qn +1
j =1 j
(5)
n cazul ecruisrii neliniare expresiile (3) devin:
p
sn +1 = 2 (en +1 en+
1)
(6)
q n +1 = H n +1
3
qn +1 = K ' ( n +1 )
f
f
f
2
= n,
= n,
=
Pe de alt parte relaia (5) innd cont c
poate fi scris sub
s
q
3
q
urmtoarea form:
896
AFASES - 2008 -
s = sT 2 n
n +1
n +1 n +1
T
2
q n +1 = q n +! + H nn +1
3
2
T
n +1 = n +1 +
3
(7)
unde:
1
T
p
p
p
sn +1 = 2[en +1 en + 3 ( s 1)(en en 1 )]
2
1
T
q n +1 = H [ n + ( s 1)( n n 1 )] ,
3
3
1
T
n +1 = n + 3 ( s 1)( n n 1 )
s = 1, 2
(8)
Multiplicatorul din cadrul relaiilor de mai sus se determin n urma rezolvrii ecuaiei
neliniare:
2
2
2
[ Y + K ( nT+1 +
)] = 0
g ( ) = nT+1 (2 + H )
(9)
3
3
3
prin metoda Newton, unde funcia g ( ) este presupus a fi o funcie convex.
n aceste condiii algoritmul metodei radial regresive poate fi sintetizat astfel:
1. Calculm urma tensiunii elastice pentru en +1 = dev( n +1 ) :
T
2
q n +1 = q n +1 + 3 H nn +1
T
sn +1 = sn +1 2 nn +1
e p = e 1 s
n +1 n +1 2 n +1
2
T
n +1 = n +1 +
3
2 ' T
f T ( n +1 , n +1 ) = nT+1 Y +
K ( n +1 )
3
cu nT+1 = snT+1 q n +1 .
Dac f nT+1 0 atunci se trece la urmtoarea etap:
3. Lum nn +1 =
nT+1
i l vom calcula pe pe baza metodei Newton.
nT+1
897
(10)
(11)
AFASES - 2008 -
T
2
=
+
q
q
H nn +1
1
1
+
+
n
n
T
sn +1 = sn +1 2 nn +1
e p = e 1 s
n +1 n +1 2 n +1
2
T
n +1 = n +1 +
3
(12)
n cadrul acestui paragraf vom implementa metoda radial regresiv n Matlab, pentru
cazul ecruisrii izotrope neliniare ntr-un singur pas. n acest caz modulul de ecruisare
cinematic este H = 0 , prin urmare i q = 0 .
n urma particulazrii metodei radial regresive pentru cazul ecruisrii mixte, pentru
ecruisarea izotrop obinem:
sn +1 = snT+1 2 nn +1
(13)
2
T
=
+
n +1
n +1
3
unde
snT+1 = 2 (en +1 enp )
(14)
T
n +1 = n
Criteriul von Mises pentru ecruisare izotrop este:
2
f ( s, ) = s
K ( )
(15)
3
unde
K ( ) pentru
cazul
ecruisrii
K A
e 1 n care K i A sunt parametrii materiarului.
A
n aceste condiii algoritmul metodei radial regresive n Matlab, pentru cazul ecruisrii
neliniare este:
izotrope neliniare este K ( ) = y
898
AFASES - 2008 -
899
AFASES - 2008 -
end
if LOP(5)==1
EPS13=APPLY(5); SIG13=2*C44*APPLY_E(5);
else
EPS13=APPLY_E(5)/(2*C44)+E13P; SIG13=APPLY(5);
end
if LOP(6)==1
EPS12=APPLY(6); SIG12=2*C44*APPLY_E(6);
else
EPS12=APPLY_E(6)/(2*C44)+ E12P; SIG12=APPLY(6);
end
SEFF_LOAD=sqrt((SIG11-SIG22)^2+(SIG22-SIG33)^2+(SIG11-SIG33)^2+
+6*(SIG23^2+SIG13^2+SIG12^2))/sqrt(2);
EPSMEAN=(EPS11 + EPS22 + EPS33)/3;ED11 = EPS11 - EPSMEAN; ED22 = EPS22 - EPSMEAN; ED33
= EPS33 - EPSMEAN; ED23 = EPS23; ED13 = EPS13; ED12 = EPS12;PRESS = BULK*EPSMEAN*3; ST11
= 2*C44*(ED11 - E11P); ST22 = 2*C44*(ED22 - E22P); ST33 = 2*C44*(ED33 - E33P); ST23 =
2*C44*(ED23 - E23P); ST13 = 2*C44*(ED13 - E13P); ST12 = 2*C44*(ED12 - E12P);
XNORM=sqrt(ST11^2+ST22^2+ST33^2 + 2*(ST23)^2 + 2*(ST13)^2 + 2*(ST12)^2);
if (XNORM*SQ3/SQ2)<YIELD
SIG11C = ST11 + PRESS; SIG22C = ST22 + PRESS; SIG33C = ST33 + PRESS; SIG23C = ST23;
SIG13C
= ST13; SIG12C = ST12;
end
XN11 = ST11/XNORM; XN22 = ST22/XNORM; XN33 = ST33/XNORM; XN23 = ST23/XNORM; XN13 =
ST13/XNORM; XN12 = ST12/XNORM; INEWT=0; GAMMADT=0; EPQ1 =0;
XKAPPA=YIELD;
DXKAPPA = HARD; G = -(SQ2/SQ3)*XKAPPA + XNORM;
while abs(G)>TOL_NEWTON
EPQ1 = EPQ + (SQ2/SQ3)*GAMMADT; XKAPPA = YIELD - (HARD/A)*(exp(-A*EPQ1) - 1);
DXKAPPA=HARD*exp(-A*EPQ1);G=-(SQ2/SQ3)*XKAPPA-2*GAMMADT*DXKAPPA/3+ XNORM
2*C44*GAMMADT;
DG = -2*DXKAPPA/3- 2*C44; GAMMADT = GAMMADT - G/DG; INEWT = INEWT + 1;
end
EPQ = EPQ + (SQ2/SQ3)*GAMMADT; E11P = E11P + GAMMADT*XN11;E22P = E22P +
GAMMADT*XN22; E33P = E33P + GAMMADT*XN33; E23P = E23P + GAMMADT*XN23; E13P = E13P
+ GAMMADT*XN13; E12P = E12P + GAMMADT*XN12;XKAPPA = YIELD - (HARD/A)*(exp(-A*EPQ) 1);
SEFF=2*C44*(SQ3/SQ2)*sqrt((ED11-E11P)^2+(ED22-E22P)^2+(ED33-E33P)^2+2*(ED23-E23P)^2+
+2*(ED13-E13P)^2+2*(ED12-E12P)^2);
CONVERGED =1;
if abs((SEFF - XKAPPA)/SEFF)>TOL_GLOBAL
CONVERGED =0;
end
if abs((SEFF_LOAD - XKAPPA)/SEFF_LOAD)>TOL_GLOBAL
CONVERGED =0;
end
end
S11=(SQ2/SQ3)*XKAPPA*XN11;S22=(SQ2/SQ3)*XKAPPA*XN22;S33=SQ2/SQ3)*XKAPPA*XN33;
SIG23=(SQ2/SQ3)*XKAPPA*XN23; SIG13 =(SQ2/SQ3)*XKAPPA*XN13;
SIG12= (SQ2/SQ3)*XKAPPA*XN12; SIG11=S11+PRESS;SIG22 =S22+PRESS; SIG33=S33+PRESS;
SIG11C = 2*C44*(ED11 - E11P) + PRESS; SIG22C = 2*C44*(ED22 - E22P) + PRESS;
SIG33C = 2*C44*(ED33 - E33P) + PRESS; SIG23C = 2*C44*(ED23 - E23P); SIG13C = 2*C44*(ED13 E13P); SIG12C = 2*C44*(ED12 - E12P);
SIG11a(INC)=SIG11; EPS11a(INC)=EPS11;
EPS22a(INC)=EPS22;
End
900
AFASES - 2008 -
A = 100
11 i 22 pentru A = 1000
Soluiile prezentate mai sus sau obinut n unsprezece iteraii ale metodei.
4. Bibliografie
901
AFASES - 2008 -
8. J.C. Simo Numerical analysis and simulation of plasticity, In P.G. Ciarlet and
J.L.
Lions,
Eds. Handbook of Numerical Analysis, volume VI, Elesevier, Amsterdam, 1998;
9. P. P. Teodorescu, V. Ille Teoria elasticitii i introducere n mecanica solidelor
deformabile,
Editura Dacia, Cluj-Napoca, 1976
10. W. Soboyejo Mechanical Properties of Engeneered Materials, New York, 2002
902
AFASES - 2008 -
903
AFASES - 2008 -
proactive testing approach ensures that network devices will excel under real-word conditions
when they are installed in private network.
There are three possible methods for realism-testing of firewalls with using Spitents
Avalanche/Reflector platforms. These methods enable to reflect the reality of a private
network by adding more parameters to firewall tests and include additional devices in series
with the firewall. This evaluates a firewall in a context similar to a private network, which
reduces deployment risk.
The set of tests are:
Firewall Load measures how well a firewall maintains availability under load.
Determine firewall operating limits during multiple IP protocol-based traffic and
Distributed Denial of Service (DDoS) attacks.
The test of validation of maximum new TCP connection per second. There was
searched to find the maximum TCP Connection Rate of the firewall and there was
observed the degradation of its performance. It was possible to evaluate the
performance degradation by increasing the object size requested the client.
The object size defines the number of bytes contained in body, excluding any bytes
associated with the HTTP header. The performance is evaluated by downloading the
following web page body size from 512B to 100KB.
The simulated client uses HTTP/1.1 without keep alive. Each client opens one TCP
connection sending one HTTP request and then closes the TCP connection. Using this
configuration, the simulated client doesnt try to keep any TCP connection opened
after every each HTTP transaction. The TCP connection will just be closed. This
means that any concurrent TCP connections are not coming from the client side.
The server uses HTTP/1.1 without keep alive, and the server will close any TCP
connection after one HTTP request. Using this configuration, the server doesnt try to
keep alive any TCP connection opened after each HTTP transaction. The TCP
connection will just be closed. This means that any concurrent TCP connections are
not coming from the server side.
Performance level ranking insures quality of service on Layer 4 (OSI). There is
possible to check the maximum TCP connection rate but also the quality of service on
Layer 7.
The test of validation of maximum concurrent TCP connections. There was searched
the maximum concurrent TCP connections of firewall and there was observed the
degradation of its performance. It was possible to evaluate the performance
degradation by increasing the object size requested by the client.
904
AFASES - 2008 -
The simulated clients use HTTP/1.1. The clients open one TCP connection sending ten
HTTP requests and then close the TCP connection. Thirty second think time between
each HTTP transaction per simulated client is used in order to keep the TCP
connection opened.
The server uses HTTP/1.1, and the server will close any TCP connection after ten
HTTP request.
Performance level ranking insures quality of service on Layer 4. There is possible to
check the maximum concurrent TCP connections but also the quality of service on
Layer 7.
The test of validation of maximum new HTTP transactions per second. There was
searched to find the maximum new HTTP Transactions Rate of the firewall and there
was observed the degradation of its performance. It was possible to evaluate the
performance degradation by increasing the object size requested the client.
The simulated clients use HTTP/1.1. The clients open one TCP connection sending ten
HTTP requests and then close TCP connection.
The server uses HTTP/1.1, and the server will close any TCP connection after ten
HTTP request.
Performance level ranking insures quality of service on Layer 4. There is possible to
check the maximum HTTP transaction rate but also the quality of service on Layer 7.
Current response time for HTTP transactions exceeds the quantification time. In this
case, data delivery to the client is taking to much time, explaining the higher response
time.
905
AFASES - 2008 -
The security policy of the firewall was defined as full inspection of HTTP packets according
to the RFC 2616 with other elements such as: prevention of VBscript, components of
ActiveX, rejection of not allowed symbols in header of HTTP packet (e.g. *, +, x).
The result of measurement of the stress-test shows the relationship between attempted and
successful transactions (vide graph 1.). Generated requirements are attended with success
until 104th second. There started to display a work-load of the firewall and every transaction
over number 2500 took time-delay. The highest time-delay, what I measured, was 1.8 second.
With decreasing a number of transactions, the time-delay was reduced and then it started to
increase even the number of successful finished requirements again. At the end of the graph,
there is seen a definite number of unsuccessful transactions. It is evoked by determined length
of the test.
906
AFASES - 2008 -
30
25
Average Response
Time Per URL
20
Attempted Transaction
15
Successful
Transactions
Unsuccessful
Transactions
10
124
112
100
88
76
64
52
40
28
16
0
Time (s)
100
90
70
60
Goodputs (http)
50
40
30
20
10
88
10
0
11
2
12
4
76
64
52
40
28
16
0
4
80
Time (s)
907
AFASES - 2008 -
Graph 2 measured in the same test as graph 1, watches the bandwidth of useful load.
Goodputs is presented like the number of transferred bites for one time unit. In this case, one
million bites per second. The maximum transmission of a hundred bites line is 12.5 MBps.
The results from graph 2 show that measured values in the test (11.3 MBps) are very close to
the maximum transmission of a hundred bites line.
Conclusion
It is impossible to standardize something so flexible. Firewalls right now are a highly
competitive market place. We can constantly asked vendors to differentiate their products
against those of their competition. The end result is that things which are very similar are
touted as very different, and some types of technologies are cast as inadequate while others
are positioned as superior. It's difficult to sort out the hype from the substance.
The real problem is that a "good" firewall for any given purpose depends a lot on the
customer's needs, security requirements, and budget. For an authority to specify evaluation
criteria for good firewalls, we'd all have to be using them for the same purpose, and they'd all
have to have roughly the same properties and it is impossible. For this case, we would be able
to simulate transmission in a private network through security devices and measure the
requirements for a performance.
Lt. Ing. Miroslav tolpa, Univerity of defence, Faculty of military technologies Communications and
information systems department
908
AFASES - 2008 -
carteziene definit de reperul ales care const n bijecia ntre P i R 2 prin care punctul M P
i se pune n coresponden perechea ( x, y ) a abscisei respectiv ordonatei punctului M .
Bijecia invers ataeaz perechii arbitrare ( x, y ) din R 2 punctul din P notat M ( x, y ) avnd
exact coordonatele x i respectiv y.
Dac f : I R , I R , este o funcie arbitrar atunci i punem n coresponden
submulimea din planul P numit epigraful sau subgraficul lui f format din toate punctele
M ( x, y ) pentru care x I i y f (x) ; aceast submulime se noteaz epi( f ) .
Deci
epi( f ) = {M ( x, y ) : x I , y f ( x)}
(1)
Funcia f prin definiie se va numi convex dac epi( f ) este o mulime convex.
909
AFASES - 2008 -
Capitolul II
i.
ii.
f ( x1 ) f ( x0 )
f ( x 2 ) f ( x0 )
f ' ( x0 ) m f ' + ( x 0 )
x1 x0
x 2 x0
i x1 < x0 < x2 , rezult f ( xi ) f ( x0 ) + m( xi x0 ) , i = 1,2,... deci graficul funciei f se afl
910
AFASES - 2008 -
i.
ii.
TEOREMA 4
Urmtoarele dou afirmaii sunt echivalente
f este o funcie convex
f = sup g ( x) : g F f n fiecare punct x I .
Cu alte cuvinte o funcie (pe un interval deschis) este convex dac este nfurtoarea
superioar a minoranilor ei liniari.
Demonstraie:
i ii rezult din teorema 3
Pentru x0 I fixat g ( x) = ax b , a [ f ' ( x0 ), f ' + ( x0 )] i b = ax0 f ( x0 ) este un
minorant i avem f (x0 ) = g ( x0 ) deci n (ii) n loc de supremum avem chiar maximum.
ii i f satisface (ii) atunci epi( f ) = epi( g ) : g F f
Deoarece funciile g sunt convexe (de exemplu g ' ' = 0 ) mulimile epi( g ) sunt
convexe (mai exact epi( g ) sunt semiplane nchise) deci i intersecia lor, epi( f ) , este
convex ceea ce nseamn tocmai convexitatea lui f.
Capitolul IV
Definim f * n a I * prin
f * (a) = sup{ax f ( x) : x I }
(2)
Din echivalenele: M (a, b) M ax f ( x) b pe I sup{ax f ( x) b}
M (a, b) epi( f ) .
911
AFASES - 2008 -
definiia (2), iar f * determin f prin faptul c f * determin mulimea M = epi ( f * ) , iar M
determin familia F f i n final, n baza teoremei 4 F f determin f.
Tot din definiia (2) rezult:
f * (a ) + f ( x) ax , x I , a I *
(3)
numit inegalitatea YOUNG-FENCHEL.
n final indicm o procedur simpl pentru determinarea conjugatei f * n cazul n
care f ' este strict cresctoare (deci f convex).
n aceast situaie I * va fi imaginea lui I prin f ' (se aplic faptul c funcia derivata
are proprietatea lui DARBOUX). Pentru a I * = f ' ( I ) se alege punctul unic x a I cu
f ' ( x a ) = a i din f ' ' > 0 se conchide c x a este un punct de maxim pentru ax f ( x) , deci
f * (a) = ax a f ( x a ) .
Aplicaie
Fie q > 1 i f ( x) =
xq
definit pe (0,+) .
q
a = f ' ( x)
admite
1
f (a ) = a x a f ( x a ) = 1
q
singura
a
rdcin
xa = a
1
q 1
consecin
q
q 1
1 1
q
ap
rezult c p > 1, + = 1 i f * (a ) =
. n final
q 1
p q
q
YOUNG-FENCHEL ne conduce la baza inegalitii lui HOLDER:
Introducnd parametrul p =
inegalitatea
a p xq
+
ax .
p
q
Bibliografie
[1] Benko, Iosif: Elemente de Analiz Matematic, Universitatea Transilvania Braov, 1992
[2] Moise, E.E.: Geometrie elementar dintr+un punct de vedere superior, Editura Didactic i
Pedagogic, Bucureti, 1980.
912
AFASES - 2008 -
913
AFASES - 2008 -
START
Profilul
client
Profilul utilizator
Profilul funcional:
Numrul de funcii;
Lista iniial de funcii;
Explicit/implicit
Variabile de mediu;
Lista final de funcii;
Probabiliti de apariie.
Profilul operaional:
mprire execuie n rulri (seturi);
Identificare spaiu de intrare.
Selectare teste
STOP
914
AFASES - 2008 -
Probabilitate de apariie
0,45
0,35
0,20
915
AFASES - 2008 -
Subprofilul D
Valoarea variabilei
Probabilitatea de
eseniale de intrare y
apariie
y1
0,7
y2
0,2
y3
0,1
Probabilitate de
apariie
0,42
0,21
0,12
0,07
0,06
0,06
0,03
0,02
0,01
Urmtorul pas este acela de a determina variabilele de mediu i domeniul lor de valori.
Variabilele de mediu caracterizeaz condiiile ce influeneaz cile de execuie ale
programului, ns nu se suprapun direct capabilitilor. Exemple de variabile sunt configuraia
hardware i ncrcarea cu date.
nainte de crearea listei finale de funcii (tabelul 4), variabilele de mediu i
caracteristicile trebuie verificate sub raportul dependenelor. Lista trebuie s fie ct mai
ortogonal posibil, aa nct dac o variabil depinde semnificativ de alta, ea poate fi
eliminat.
Numrul final de funcii este calculat ca fiind produsul dintre numrul de funcii din
lista iniial i numrul de niveluri ale variabilelor de mediu, minus combinaiile de funcii
iniiale i valori ale variabilelor de mediu care nu apar.
Ultimul pas este asignarea probabilitilor de apariie. Acestea pot proveni la modul
ideal din msurtorile fcute ultimei versiuni sau unui sistem similar nregistrate n log-urile
sistem. Valorile probabilitilor de apariie obinute pe baza datelor istorice trebuie aduse la zi
pentru a considera noi funcii, medii de operare sau utilizatori. Dialogul cu utilizatorii poate
indica relativa importan acordat de acetia unor funcii, fapt ce poate conduce la reducerea
numrului funciilor i implicit la creterea fiabilitii.
916
AFASES - 2008 -
Variabil de mediu
x
y
x
y
x
y
x
y
Deviaia standard
Corelaia
Variaia
Regresia
1.5. Profilul operaional
Tip de execuie
(stare de intrare)
Operaie
(domeniu)
Funcie
917
AFASES - 2008 -
918
AFASES - 2008 -
Cezar Vasilescu, lt. col. lector univ. dr. ing., Departamentul Regional de Studii pentru Managementul
Resurselor de Aprare, Braov, Mihai Viteazul, nr. 160, tel. 0268/401800, email: caesarv@crmra.ro
919
AFASES - 2008 -
APLICAREA METODOLOGIEI
DE ESTIMARE I PREDICIE A FIABILITII
PE BAZA PROFILELOR OPERAIONALE
LA SISTEMELE C4ISR
Cezar VASILESCU *
Abstract: The current article focuses on the command and control systems within Air Force.
Its purpose is to describe one of the solutions that exists and that addresses the critical
problem of reliability. Thus, it will detail on how the methodology of hardware/software
systems operational profile development can be applied to a case study that is representative
of the C4ISR systems. In order to identify the operational profiles of a missile system within
the USA Air Force, which is approached as a C4ISR execution subsystem, the topics
discussed cover the content and the means of establishing customer profile, user profile,
system mode profile, as well as system operational profile developed in accordance with its
functioning principles. In the end, based on this methodology employed to estimate and
predict systems reliability on the basis of operational profiles the necessary conclusions will
be drawn.
Keywords: reliability, hardware/software systems, operational profiles;
2000 MSC: 68M15
I. PREZENTAREA SISTEMULUI SELECTAT
Exemplul urmtor trateaz determinarea profilelor operaionale ale unui sistem de
rachete (subsistem de execuie C4ISR) aflat n dotarea forelor aeriene i navale ale SUA.
Datele reale au fost alterate pentru a pstra caracterul general al analizei i a servi scopurilor
de exemplificare [4] [7].
Cerina de fiabilitate pentru sistem este aceea c impactul rachetei trebuie s fie sub o
raz de 10 metri de int n cel puin 95% dintre cazuri. n plus, fiabilitatea dispozitivului de
transport i lansare (definit ca probabilitatea ca proiectilul aerian s treac un test IBIT
Initiated Built-in-test, iniiat ntre comanda de lansare i lansarea propriu-zis) trebuie s fie
de cel puin 0,98. Aceste cerine sistem au fost alocate elementelor configuraiei hardware i
software-ului operaional de zbor.
Software-ul operaional de zbor a primit o alocare de fiabilitate de 97% (pentru zbor
liber), adic eecul rachetei datorat unei erori software nu trebuie s apar mai mult de 3 ori la
100 de trageri. Cerina de fiabilitate pentru IBIT a fost de 0,995.
920
AFASES - 2008 -
Probabilitate de apariie
0,73
0,27
Probabilitate de apariie
0,38
0,22
0,21
0,13
0,06
921
AFASES - 2008 -
apare o singur dat pe durata misiunii. Profilul mod sistem determinat este prezentat n
tabelul 3.
Tabelul 3. Profilul mod sistem rachet.
Mod
Zbor liber
Test iniial (IBIT)
Probabilitate de apariie
0,1818 (90/110)
0,182 (20/110)
IBIT
Controler IMU
Controler GPS
Controler LP
Secveniator de misiune
Tratare mesaje
De baz / alternativ
Controler TAS
Controler JPF
Starea vehiculului
Controlere de comunicaie
Controler interfa GEM
Utiliti
922
AFASES - 2008 -
Factori IBIT
Timpul JPF rezultat
Rezultatul BIT ATE
Timpul ATE rezultat
Rezultate discrete
Urmtorul pas a fost acela de a stabili limite pentru fiecare dintre parametri i numrul
de nivele n care fiecare variabil ar trebui partiionat. Am decis limitarea la numai dou
nivele. Parametrii cu limitele variabilelor asociate i probabilitile de apariie estimate pentru
IBIT sunt prezentate n tabelul 6, iar pentru modul zbor liber n tabelul 7.
Tabelul 6. Profilul operaional IBIT pentru rachet.
Parametri
Rezultatul BIT IMU
Rezultatul BIT GPS
Rezultatul BIT TAS
Rezultatul BIT JPF
Rezultatul BIT ATE
Rezultate discrete
Timpul IMU rezultat
Timpul GPS rezultat
Timpul TAS rezultat
Timpul JPF rezultat
Timpul ATE rezultat
Valori
(Pass, Fail)
(Pass, Fail)
(Pass, Fail)
(Pass, Fail)
(Pass, Fail)
(Pass, Fail)
(2-10, 10-20)
(2-10, 10-20)
(2-10, 10-20)
(2-10, 10-20)
(2-10, 10-20)
Probabiliti IBIT
(0,99, 0,01)
(0,99, 0,01)
(0,99, 0,01)
(0,99, 0,01)
(0,99, 0,01)
(0,99, 0,01)
(0,05, 0,95)
(0,05, 0,95)
(0,10, 0,90)
(0,15, 0,85)
(0,50, 0,50)
Valori
(Da, Nu)
(Orizontal, Vertical)
(On, Off)
(normal, Loft)
(mic, rapid)
(lung, scurt)
(lung, scurt)
(lung, scurt)
(40K, 18K)
(mic, mare)
(lung, scurt)
Probabiliti IBIT
(0,75, 0,25)
(0,90, 0,10)
(0,70, 0,30)
(0,80, 0,20)
(0,60, 0,40)
(0,95, 0,05)
(0,75, 0,25)
(0,85, 0,15)
(0,90, 0,10)
(0,65, 0,35)
(0,50, 0,50)
Fiecare dintre aceste profile este implicit. O stare specific de intrare este generat
prin selectarea aleatoare a unei valori pentru fiecare dintre aceti parametri. n primul rnd va
fi selectat un interval de valori, n raport cu probabilitatea de apariie. De exemplu, exist 60%
anse s fie selectat o vitez mic. Aceasta se poate ncadra n intervalul 0,4 la 0,9 Mach.
Apoi va fi aleas aleator o valoare din acest interval, toate valorile avnd aceeai probabilitate
de a fi selectate.
923
AFASES - 2008 -
Aceste profile operaionale au fost utilizate pentru a genera cazuri de test pentru
software pe parcursul testului sistem. Efectuarea unui numr substanial de teste aleatoare
permite atingerea unui nivel bun de acoperire a spaiului intrrilor n sistem.
n acest exemplu nu a fost dezvoltat un profil funcional deoarece nu exist o
interaciune uman cu sistemul pe durata modurilor sistem considerate.
IV. OBSERVAII ASUPRA ESTIMRII I PREDICIEI FIABILITII PE BAZA
PROFILELOR OPERAIONALE
Metodologia de estimare a fiabilitii folosind profile operaionale prezentat anterior
impune cteva observaii:
Ofer o procedur clar i util pentru construirea unui profil operaional pe baza
cruia pot fi obinute cazuri de test adecvate; detaliaz, dei nu n totalitate, cei
cinci pai care se impun a fi urmai i dezvoltai;
Exemplific fiecare subprofil obinut, astfel nct s dea relevan maxim
entitilor obiect implicate;
Permite apropierea testrii software de cea hardware, tabelele 6 i 7 realiznd
conversia variabilelor de intrare n variabile binare (cu dou niveluri);
Metodologia nu ofer garanii cu privire la detectarea oricrui defect.
Astfel, testarea aleatoare a profilului operaional din tabelul 6 cu un anumit prag de
ncredere, necesit un numr de experiene [1]
p (1 p )
N=
,
2 (1 )
unde:
p - probabilitatea de eec; - eroarea de aproximare; - pragul de ncredere.
Pentru p=0,01, =0,001 i =0,9 rezult N 10 5 experiene. Cu alte cuvinte, nu
este garantat dect acoperirea a 90% din cazuri, cu o eroare de 10-3, chiar dac se
fac 10 5 experiene.
Pe de alt parte, testarea exhaustiv a profilului operaional din tabelul 6 necesit
un numr de experiene N=211=2048, care acoper sigur toate situaiile posibile;
Testarea bazat pe profilurile operaionale din tabelele 6 i 7 ignor structura
funcional a OFS (Operational Flight Software), precum i faptul c prin testarea
unui regim se obin informaii despre comportamente care apar i n cellalt regim;
Metodologia de testare nu precizeaz dac modulele din tabelul 4 sunt testate
individual i simultan (caz n care ar trebui s existe un regim special de testat) sau
nu;
Metodologia de testare ignor fiabilitatea modulelor (numrul sau procentul lor de
erori reziduale), iar durata testrii lor este considerat implicit (fiecare modul
funcioneaz o singur dat pe durata unui test); acest fapt conduce la o eficien
sczut a procesului de testare;
Metodologia de testare propus nu ofer detalii cu privire la ieirile testate i la
evaluarea acestora.
Pe baza observaiilor enumerate anterior, concluzia care se impune este aceea c ea nu
ofer garanii cu privire la detectarea oricrui defect, are o eficien sczut i este
perfectibil. Modul n care pot fi eliminate n parte deficienele acestei metodologii, mpreun
cu alte aspecte perfectibile ale metodei poate fi consultat n [7].
924
AFASES - 2008 -
Unul dintre beneficiile majore ale metodologiei este acela referitor la creterea
fiabilitii i teste de demonstrare, bazate pe profile operaionale. Meritul principal al acesteia
este acela c ea ofer o abordare orientat spre utilizator a fiabilitii sistemului, care prezint
numeroase oportuniti pentru modelarea i creterea fiabilitii sistemului.
n concluzie, putem aprecia c metodologia prezint att avantaje, ct i limitri, iar
dezvoltrile oferite de [1], [2], [3], [5], [6] i [7] ofer o baz mai bun pentru alocarea i
predicia fiabilitii.
IV. BIBLIOGRAFIE
[1]
[2]
[3]
[4]
[5]
[6]
[7]
Cezar Vasilescu, lt. col. lector univ. dr. ing., Departamentul Regional de Studii pentru Managementul
Resurselor de Aprare, Braov, Mihai Viteazul, nr. 160, tel. 0268/401800, email: caesarv@crmra.ro
925