TO

Professor A. H. LEFEBVRE Mr. E. R. NORSTER

-
BEST COPY

AVAILABLE

Variable print quality

 -_

IN THE NAUE OF GODTHISßEItEVOZE;

`tT AND
THE E0$r MERCIFUL
TUF COLLEGE OF AERONAUTICS

CRA IT PIRLD

T1FPARTMEl'IT OF AIRCRAFT PRnPti7I3InSi

DISCHARßF eOFFFIQIIII-IT AND JFT D'rrLýý+:

CTION STUDI,. 3

F0R
_..ý_ý_..
COMBU3TOR LINER AIR - NTRY HOLP::3`

BY: K. SH. KADDAR

Thesis submitted for the Diploma of Advanced Engineering

lot JUNE, 1964

 ACI4In'A'LM)c3NF,NT3

The author wishes to express his most sincere thanks

to Professor A. H. Lefebvre ' Head of the Department of

Aircraft Propulsion , and to Kr. E. R. Norater who

supervised this work for his helpful advice and assistance.

Thanks are extended also to the Propulsion staff members,

and to the College Authorities for providing all facilities

to carry out this work.

Thanks are also due to Workshop ataff, Mr. r. A. McNeill,

Photographic and Library staffs.

The author wishes also to express his thanks to Mrs. Archer

for the task of typing this thesis

 1.

Section I
S 11 14173 ARY

Theoritical and experimental investigations of discharge

coefficient ' and jet deflection 1avo been made for holes

having parallel to the plane of the hole

In the experimental investigation geometric configurations

studied were , circular holes , oval , rectangular , and

square alo-s . The main geometric and flow factors affecting

discharge coefficient and jet deflection were atudt1 .
Geor otrie factors ware : Ratio of hc1f: area to nprrouch
duct a, from O. 03'Y
'5 to O. F6C4C ; slot
area ranging aspect

ratio from 1,0 to 8.0 ; Slot major dimension along the flow

up to 2.0 inch with duct depth 1.25 inch ; and slot shape .
Flow factors were ; Static pressure ratio pi / pj up to
1,12 ; Bleed ratio of ; and approach velocity V1 up to1303f .
The results of these studies. have been correlated with
the dimensionless flow parameter Y. incorporating both the

geometric and flow factors

Data are presented in detailed form for direct use with

survey and-design applications to combustor liners
,
 2.

Section 2

corrcFtrr:3
vaLU1aF ý
3e tion Pa ?P
ýý
Title

Acknowledgments

1 SUMMARY 1I

2 CONTIMLIS 2

3 ?1rTAT In? T3 6

3.1 Liat of symbols 6

3.2 muff ix notation 8

4 IT7rRODLRTI(11 9

5 RY I*';; OF PREVIOUS WORK 11

5.1 Diacharge coefficient 11

5.2 Deflection angle 15

6 `r1 Eo ITICAL ITIVMTIGATIC"11 18

6.1 Analysis I 19

6.2 Analysis IL 23

7i PERIYF1TTAL ?'1VP:1TIGATIf" 27

7.1 Rig deoeription and levelopmer, to 27

7.2 Tent section dpa^ription and dnvnlcpm nts 27

7.3 Tont plates 29

7.4 Instruments and calibrations 32

7.4.1 measure measurements 32

7.4.1.1 Approach static rrennure 32

7.4.1.2 Jet static pressure 32

 3.

CONJ g (continued)

Section PAPM

7.4. i«3 Total and dynamic presaurea 33

7.4.2 Temperature meacuremento 33
7.4.3 Flog rate measurementa 34
7.4.4 Velocity measurements 35

7.5 Deflection angle measurement and

jet visulation 36
7.6 Preliminary work 37
7.6.1 Velocity profile 37

7.6.2 Deflection angle 38

7.6.3 precautions 39

7.7 Teat procedure 39

8 J Op R :3ULP 3 AND DIgcU3AUIo^t

PRESr'ETTTATIO? 42
8.1 Correlation of results 42
8.2 Discussion of theoritical analyses 42
8.2.1 Analysis I 42
8.2,2 Analysis U 44
8.2.2.1 Momentum loss factor 45
8.3 Discussion of experimental results 45
8.3.1 Disoh rge coefficient and jet deflection
for rcular holes 45
8,3.2 Discharge coefficient and jet deflection
for Oval alote 47
8.3.3 Discharge coefficient and jet deflection
for Rectangular slots Iý9
8.3.4 Discharge cocftioi, nt ani jet d#f3(-ct ion
for various configurations 51
 It*

COMMS (continued)

action Pn rrý..
n
.
8.34.1 Holes and sloto with equal areas
and nuMers 51
8.3.4.2 Slots with equal nreaa, nunber3
and aspect ration 52
8.3.5 Analysis of experimental results 53
8.4 Comparison of experimental results
with theory 54
8.4.1 Discharge coefficient 54
8.4.2 Jet deflection 55

8.5 Cociptiriaon of experimental results

with previous work 55

8.5.1 Discharge coefficient 55

8.5.2 Jet deflection 56

9 USE OF DATA 57

9.1 For survey applications 57

9.2 For design applications 58

10 CONNCLU3I02T3 e0

11 R?=MJF. IM. %TION"3 FOR y lT'2TI' ''? 'VORY 62

RF' ý;Tiý; `Ir,F3 63

APPF"IC F,:3

APPENDIX I low parameter K 65

APPENDIX 2 Reduction of experimental resulto 69

A2.1 Method used for calculation of

results 69
A2.2 Momentum Ions factor 71
 5.

COINT (continued)

Section Page

APPENDIX 3 Calibrations 74

A3.1 Orifice plate characteristics

and calibration 74

A3,2 Pitot tube characteristics and

calibration 77

APPENDIX tP Correlation of Gandamihardju 80

work (Ref. 7)

Appendix 5 Jet study 84

TABLES (30 Table) 87

V0LUMME 2
FIGURE3 ( 125 figure )

APPENDIOE1 FIGUR'R3 ( 26 figure )

PHOTOGRAPHIC PLATE3( 24 plate )

SEPARATE DÖQUMF.NT

DESIGN DRAWINGS ( '1:0 aheeta )

 6.
Section,, 3
tT 0'm

3.1 biet of eyi bola

A Area oq. ft.

at Area ration ratio of holen or clots to approacharea

a Rectanou1ar clot side normal to flow direction ft.

aIr Oval alot side normal to flow direction rt.

b Raotaneular slot side parallel to flow direction rt.
b Oval slot aide parallel to floe direction ft.

b,/a rectangular slot aspect ratio

1f Oval slot aspect ratio

C Dischar1e Coefficient

a ratio of measured to theoretical flow through hole or slot

CU Flow rate correction factor for temperature effect

CN Velocity correction factor for temperature effect

°k
CUUU/ib.
Cp Specific heat of air at constant pressure ."0.24
°k
C. Cpecifio heat of air at constant volume " 0.17 CRU/lb.

D Circular hole diameter it.

Gravitational 32.2 ft. /sec
g acceleration -

K Flow parameter

a Ratio of jet dynamic prossure to approach dynamic

Pressure . Pi "" P/Pl - pl

. Vj2/V12 for incomprecsiblo flow

L Air flow rate lb. /hr.

x Momentum lb. t

n Number of holes or elote

 7.
Nn"',,TI G (contimzed)

P Absolute total pressure lb. /eq. ft.

p Absolute static pressure lb. /sq. ft.

T Absolute temperature ok

V Air velocity ft. /bec.

Y Duct horizontal width normal to flow direction ft.

X Duct horizontal width parallel to flow direction ft.

Z Duct depth vertical to XY plane ft.

Y/Z Duct aspect ratio

of Bleed ratio " LA1

Ratio of actual air flow rate passing through

holes or slots to app caching flow rate

ii Ratio of specific heat at constant pressure to

specific heat at constant volume r 1.4

Momentumloss factor
Iý Ratio of Bleed to area ratio of /a*
"
i Density lb. /cu. ft.

Deflection angle, between jet axis and the

downstreamplane of the hole degrees

 8.

3.2 Suffix Lottian

Atx. Atmosphoric conditions

ß Blood orifice plate conditions

d Downstreamorifice plate conditions

j Jet conditions assumed at the yens contracts

0 No crone flow conditions

th Theoretical flow through holes or slots with C "" 1.0

Q upstrurs® orifice plate conditions

1 Steady flow conditions approaching holes or slots

1 ? lox conditions opposite to hole centre line

2 Steady flow conditions downstream the holes or slots

 9.
Sc ilon
ýiTfýOTSUCTIOK
. .. _.......... ý....

In the simple tubular combustor system for turbojets and ramjets,

figure 1, at the primary zone it is required to achieve a high heat

release and stable flame. At the dilution cone to have a good air

flow distribution (Ref. I ), and adequate penetration. In both zones

these requirements needs to be with the minimum total pressure loss

for high overall efficiency, short liner length, and light simple

construction (Ref. 2 ).

The air flow distribution between these two zones is still the

main problem of the design engineer. But the flow of annular air

to these zones is governed by the geometry of the eyetem, hole

coonetry and dicehargo coefficient of these holes. For light

construction and adequate penetration, the main governing factor is

the deflection angle of the liner jots.

The worts presented here is a theoretical and experimental study

of the main geometric and flow factors affecting discharge coefficient

and deflection angle for the commonly used hole shapes used in combustor

liners over a wide practical range of geometry. These are, plain

circular holes, flush oval and rectangular slots, geometric details

are shown at voction 7.3, slot major dimension with flow direction.

An attempt has been made to correlate discharge coefficient and

deflection ant,ls with flow parameter, which is the ratio of jet dynamic

proarüre to the approach dynamic prescure. This work can be applied

r
for both survey and denim applications up to static precruro ratio of
 10.

1.12 acrooa the holes. This is most suitable for low pressure lose

combustors and can allow up to pressure loss factor of IC%,,

Although these data corresponds to flow parallel to the plane

of the bole with zero internal flow case in the combustor, (Rete. 3 and 4)

showed that ouch data can be satisfactorily applied to both internal

and external flows, since the jet velocity is greater than the internal

flow velocity and the correct jet static pressure is used. Also

(fiefs. 5 and 6) showed that the effect of wall inclination and

curvature on discharge coefficient is negligible, and (Ref-7 )

showed the came negligible effects on deflection angle.

Theoretical analysis based on incompreesiblo flow gives good

agreement with experimental results within the range of the experiment.

Although some authors measured discharge coefficient with parallel

flow, i. e. Dittrich Graves (Refs. 5 6) for holes

and and circular

and rectangular slots; 1'ieldirig (Ref. 8 ) for circular holes over a

wide practical range; Callaghan (Rat. 3 ) for circular holes, square

and elliptical slots with normal flow into still and moving streams.

No one however has measured the discharge coefficient for oval slots,

Gandamihardja (Ref. 7 ) measured the deflection angle.

and only

Gandamihardja's work covers a small range of circular holes and have

been correlated into a more practical form Appendix 4.

For special applications such as liners designed with high

annular air velocity and low overall pressure loss, the work by

Dittrich and Graves (Ref. 6) covered discharge for most

coefficient

of these shapes such Be thumbnail scoops, step louvers and scoops.

over circular holos.

 11.

; 2.qjiop-j

R="VrEY 07 rcctl WORK

. _._.ý....
5.1 Di char, ýe Coeffieiez

Dittrich and Graves, (Ref. 5) measured discharge coefficient

for combustor linor air-entry holes having parallel flow unins

2.23 x 2.23 inch test section, with different duct aspect ratios and
test hole thickness from 0.04 to 0.5 inch, and at air temperature of
75 °r. They found that discharge decreases with
coefficient

increasing approach velocity up to V1 . 500 ft. /sec., and increases

with increasing static pressure ratio across the hole up to

PI/Pi so 1.32.

For no cross flow they showed that discharge I

coefficient

increases with a decrease of the ratio of hole area to duct area up

to a* = 0.6, and increase also with either the increase of wall

thickness at the hole or a decrease of hole diameter, so that the

ratio of hole diameter to wall thickness is less than 6.0. In their

study they showed that the effect of wall thickness and hole diameter

were found to be small, effects of duct height, boundary layer

thickness and static pressure level were negligible.

A good correlation had been made between discharge coefficient

WA flow paranoter, but at high pressure ratios a pressure ratio

,
correction factor wan introduced. This is the ratio of actual

discharge coefficient to discharge coefficient corrected for pressure

ratio effect. The introduction of this paraarater forms a common

curve for various approach velocities.

 12.

Provided the external flow pacaase walls are parallel with

flow direction, jst velocity greater than the internal parallel flow

velocity and jet static pressure aaaumed equal to static pressure

in the liner at the considered point. The corrected coefficient Cp

from figure 2 is multiplied by appropriate value of C/Cp from

Figure 3 to obtain the coefficient to be used.

The same authors, (Ref. 6 ) measured the discharge coefficient

for other various configurations of liner wall with approximately

the same conditions an used with (Fief. 5 ). For rectan, 6ular. flush

silots with major dimension parallel to flow direction, they found

that the discharge coefficient is slightly higher than those of

circular holes and square slots. And the increase in slot aspect

ratio increacea discharge coefficient slightly throughout the range

of flow paramoter. The rain factors affecting discharge coefficient

are otatie pressure ratio across the hole and duct stream velocity.
The addition of a scoop to a flush hole increases discharge

coefficient with flow range extension at low values of flow parameter

less than 1.0 (due to ram pressure effect). But it decreases discharge

coefficient at high values with scoop face areas less than 1.4 times

hole area, Iigsre 4. The effect of scoops or louvers is to increase

discharge coefficient and cake flow range wider at low values of flow

parameter depending on scoop height which makes them essential in

upstream region of combustor liners, having low pressure difference

across openings.

The authors also showed that the effect of boundary layer

 13.

displacement thickness is to reduce scoop face area. Thus to

correlate discharge coefficient in this case with flow parameter,

boundary layer correction factor was taken into consideration.

For calculations based on effective scoop area together with static

pressure ratio effect discussed before, the corrected discharge

coefficient for both boundary layer and static pressure ratio effects

can be read from curve, Figure 5. This being selected on the basis

of Geometric similarity (both shape and sise). This coefficient

must then be reduced to the corrected discharge coefficient for static

pressure ratio alone, by multiplying the ratio of effective scoop

area (estimated from pressure profile measured at scoop face) to

scoop face area as follows

e
Äff-
C CP CP. * (1
PÄ P

The desired discharCe coefficient can be calculated as for circular

holes using Figure 3.

C"" CP {C

P
Where
Af . face area of scoop ft?

Af* a effective scoop face azea tt2

b width of opening ft.

C" discharge coefficient

cp " discharge coefficient corrected for pressure

ratio effect
CpI discharge coefficient corrected for both static
cd
pressure ratio and boundary layer effects
eb*
0 boundary layer diaplacemont thickness.
 14.

In many combustor designs, the liner walls and outer shell are

not parallel. The same authors (Ref. 6) showed that for flush

holes and step louvers mounted in walls of convergent duct walls,

inclinations up to 20° had little effect on discharge coefficient,

with static pressure ratio (p1 + q1 sin e)/p4 instead of pl/pj for

parallel holes. 'hero () is the angle between direction of flow

and hole plans in degrees, and ql is the dynamic pressure of approach

stream, Figure 6.

Also they showed that multiple flush hole arrangements both in

the longitudenal direction range from 1.5 to 5 hole diameter (centre

to centre) and transverse direction from 2 to 4 hole diameter spacing

has no significant effect on discharge coefficient, Figures 7 and 0.

D. U. Fielding, (Ref. g) made a detailed investigation of

discharge coefficient for plain circular holes, O. 064 inch thickness,

using 1x5 inch test section. He studied effects of bleed, area

ratio up to a* .. 1.2, pressure loss coefficient, aspect ratio of

approach duct up to 1O/I, and jot Nach number up to 1.0, on discharge

coefficient. He found that for a given area ratio, discharge

coefficient increases as the bleed increases , till the flow in the

approach duct forma a vortex into the hole, causing decrease of

discharge coefficient. But in general it decreases with increase of

area ratio for any value of bleed. For the effect of approach duct

aspect ratio, he showed that for an aspect ratio 2.5/1, discharge

coefficient increases continuously with bleed for area ratio rang of

0.1 to 1.2. And with aspect ratio 5.0/1, discharge coefficient
 15.

behavoe as above case up to area ratio 0.14, above which, sudden

decrease in discharge coefficient occurs between bleed of 0.8 and-

0.9 due to vortex formation in the hole ! 'low, Figure 9. At aspect

ratio 10/1, sudden decrease in discharge coefficient occurs between

bleed values of 0.45 and 0.8 deeding on area ratio. -

R. Callaghan (Ref-3 ) measured discharge coefficient for circular

holes up to 0.625 inch diameter, squares arAL elliptic slate up to

es act ratio of 4/1, with equal areas and 0.064" thickness discharging

normally Into. still and moving air streams. -, Air jet total temperature

was 70 and 4AO0F. In, terms of pressure ratio up to 3.2 and jet

F.eynolds somber up to 64 x 104, ' he found. that up to chocking,

discharge coefficient ineieasea linearly with presaure ratio, with

transition region slightly higher than chocking. Then linearly again,

Figure 10, the rate of increase is high below chocking than above it.

At constant pressure ratio, diecih-arge coefficient decrease with

increasing; jet Reynolds number, but for the given pressure ratio and

jet total temperature, ellipses yields the hiChest discharge coefficient.

The effect of stream velocity on dicct. arge coefficient was neglected

provided that proper outlet static pressure was utilised.

5.2 Teflectton en le

Candamihardja (Ref. 7 1962) easured the deflection angle for

,

circular holes diameter range 0.25 - 1.0 inches with corrccpording

area ratio of 0.0344 - 0,314 respectively. He used the ewe rig

before. modificationa, Sections 7.1,7.2, with the jet issuing into

atmosphere. Vor deflection angle mrasuremsnt lie used carbon coated

 16.

plate, Section 7.5. Results of his work showed that with no cross

flow, deflection angle is function of area ratio a'F. As a*

increases decreases and the variation in within his range

was from. 83 degrees for largest holes to 88 degrees for the smallest.

With cross flow he found that 4) is (a* U*)

a function of bleed ratio

and a*. Where U* is the jet velocity component normal to the plane

of the holes. Also for a given value of a*, with increase in bleed

increases, and for a given bleed ratio as a* increases,

tP decreases. Results of his work are shown, Figure 11.

In a trial to correlate Candamihardja'o results Appendix 4 in

a more practical form. An empirical equation has been derived from

his work results giving Cp as a function of a*, a* U*-as follows

(i) With no cross C

ý *a 88 - 14.43 a"
(ii) with cýo,wFIllgx
.
[M. 42--ý 0575 (330)a* U*
88- a* 69 .
(a* tir).

These two equations give good agxeement with his work, Yiguro A4.69

and to the present work for circular holes Figures 11' 2 and119.

M. Cox (Ref .9) in his experiments on combustion chamber liner,

he estimated the value of Cos (P an the ratio of axial approach

velocity V to jet velocity, which is a good approximation for email

values of (P , and with increasing values of t(

, the axial approach

velocity decreases by a factor k obtained from previous tests

 I 7.

00 that
vi
Coo ý) .k .
Ili

whoro k m. (vz/`vl)

From examination of previous results Howell (Iief. 9) suggested

an empirical correlation for discharge coefficient as

C Co sin` q)

where C0 is the discharge coefficient for no cross flow assumed

of constant value.
 i8.

: section 6

TiMURMCAL INVF'.^,TIQAZ`IOtt

The analyses of this section deal with discharge coefficient Co

and deflection angle (p 0 in terms of the dimensionleos factors;

area ratio a*, bleed ratio o(, bleed area ratio , and flow

parameter K (Appendix 1).

Analysis I assuaoa ideal flow conditions with no change in the

axial direction of the approach momentum.

Analysis II deals with directional changes in approach

momentum, and noedn some exrerimental work to confirm the effects

of flow conditions on discharge coefficient and dofleotion angle.

This was attempted in Appendix 2.2.

The following assumptions have been used in the investigation

(1) Discharge coefficient is defined as the ratio of jet area at

vena contracts normal to jet axis and hole or slot area.

(2) Two dimensional in-compressible flow, since density ctangos

in the experiment within ate.

(3) Total pressure at the jet vena contracta is equal to the

total pressure of the approach stream.

(4) Static pressure opposite the test hole, is uniform and equal

to the approach atrea m static pressure.

(5) Jet static pressure at vans contracta is equal to the

surrounding pressure.
(6) Uniform flow conditions upstream, downstream jet
and at

versa contracts.
 19.

(7) Isentropic frictionleeas flow across the holes.

(ß) Effects of heat transfer and boundary layer thickness are

negligible.

Anay: ýis r,
,
In this investigation, the momentumequation is assumed to

be valid in x and Z directions with no change in the axial

direction of the approach momentum, i. e. with no axial momentum
lose (Figure I).

Applying continuity, Bernoulli, and momentumequations:

(I)
(2)
I ý
I
p E, --l - ---

V, E----ý VT
ý'_ _.
. __ap. -ºý ßz

FiG. i.

(1) Coninu. ltZ ecuntior}

52 A2 Y2
Lj ° Ah A Vh S1al v1 "

(1)
 2O .

(2) Fxvm &ornoulli'es catý

ýtion
V? Y12
F!..
P1 o pi f _f )
2
. pý t - ý P2 f'3Y'2 (2

(Vi 2-
or p1 '* Pj +2 Q12)

ýv322 i
P2 a p3 +2 -Y - Y2

p22
and P2 ý pl °2 (yl y2 ) (2i)
".

(3} r, e; ultion
_om_ertum
(i) Axial x direction

Ml - M2 - Mj cos 4) ° Al (P2 ý P1ý

SC Ah Yj 2 ý
5. Al (V V22) cos s Al (P2 ' P1) (3)
12 -
.F

Vertical Z direction
(ii) _
Ah (pl -pj)= Fis sin (p --- --
Ah (PI - pj) 5C Ah Vj2 sin
st `r

2
or P1 + Pj s. 1CY, Ein Cp (4)

Fa equation (1)
(vi - v2 )
Al
Vm
Ah Vi

Vi ' Y2
1
(5 )
a" vJ

ubetituting from equation 21 and 1, equation 3 becomes

Jn Yip
Coo (G)
2 Q,
 21.

but for incompressible flow using continuity equation

Li
= A131v1- A242v2
A2 52 V2
11.
or _.... m1- Vi
Al 31 vl A1-31

Since Al a A2

or
Y2 = Y1 (1 -o{) (7 )

by using equations and tAl-. 3) equation (5) becomes

(7)
(8)
c= fi
aý

equation (6) becomes

d-_ 2-Dl
4V ii IL/ - (9)
2 ýK

from equation (2), (4) becomes

substituting

1C
Sin
2 K1 (10)
from (9) and (10) by squaring and adding

2 ý2+ 2ýK x
2,2
a1
4K 4C
or
A K-1
(11)
JK [4K-(2- 61)Z]

for this to be true, equations (8) and (11) should be equal

or K-I
or
aý °K [4K
-(2-C4)]
or K2-K(2+4P2)+f``2(2-0()2+1 Q 0 (s2)
 22.

Equation (12) is a quadratic in X which by solving gives

2+ 2
Km 1+2%:
V aýa +

a
lk
(a oC- o<2) (13)

K1 an l+ 2ý2 -
V 4 f`4 + f`ý: (40(- 0e) {14)

Equation (14) represents the flow limit when c( . 0p this is

true when there is no flow through test holes i. e. icl - 1.0,

to equation (13) is the required solution.

A "- Catg with no croon flow

i. e. all the approaching flow passes through hole

i. e. 0ý . 1.0 0 /M ý"a (15)

using equations (8) and (9)

COw (16 )
aýý

Coo P (l'r)
° 21ý
ConbininQ (16) and (17)

a, o
cos
o" (is)
from equations (13) using t15)
2+
K an 1+21k 4fý+399

or 2-
2+ (1+- 2-) ä)
K¢ 1+ (1+

For . 054C a* .<,, 0.8, Table 1 gives the values of K, C, and 4,.

Various relations between a*, K, Coand () are represented

graphically Figures 12 to 15.

 23.

B- Case with cross flow

In this case the approach flow passes partially through holes

i. e. 0< co" c 1.0

using equations (8) and (9)

Al
Ca ý0-
8ý 1°

L z-ý
uOS(p -
zfx
Combining (8) and (9) togcther we get
Ca* 2-Oj (20)
Cos
2 0(
4 for bleed ratio
Tables 2 to 5 show values of K, C and range

2 < c<<0.8, and for values of a* within 05 < a* 0.8.

o. . 'K,

Various relations between fP, K, C and 4)

a', o(, are represented

graphically Figures 16 to 21, values of K from equation (13) as

K .. 1+ 2p2 +4ý+ J2 (4 b? -, tt2)

Analysis It

With this analysis a broader approach to the problem is

attempted,, from the physical behaviour of the approach stream

linos. As the approach stream approaches the holes, the stream

lines changes its direction from the axial direction of the

undisturbed steam to an inclined direction towards the hole as

 2L.

(I) (2)

ý ý
-; --,;.
n r-ý -
ý ý P1
ý
ý 1h. T-ý
" -ý-- -ý
.
_aj Nz
A, ýVA ýI

F-tC ii.

shown (Figure ii). This inclination depends on the flow rao e

through the hole, f. e. on the flow paramoter. For very rcall

values of K, the deflection angle approaches zoro and the approach

stream linos become axial. In this region results of this analysis

approaches that of Analysis I.

The effect of the chance in axial direction of the approach

momentum appears in this analysis in the momentum equation as a

factor ', is the momentum loss factor. This

which called

momentum factor affects the value of K for same Geometric and

flow factors compared with Analysis I, FiCure 22, as discussed later.

It is clear now the effect of this momentum factor on dischurie

coefficient and deflection angle, as K is the main factor Governing

them. Complete analysis of the momentum loss factor in discussed

in Appendix 2 . 2. Equation A2.2.3 shows that the simple form of

the momentum equation (4) is written in the forms

 25.

xJ sin (ýo I[ah (pl _p

j)]

_# r- I. _
0 (pl ) -0 ýj99 : in 4) 3 (21)
09 . - p, - L5`

usintf c.,.wtian (2)

" /yýý' ]. i: -1
ý ý (22)
1i ? Cý S,

fron (25)
(s) ruui

C (23)
ý4 id (2 ).
:"` - - o(

from equations (11) and (23)

K2 - K(2 * rº J1ý12ý`ý) ýýpl? ýý 12

-oý)ý t1]e0 ý'4J

Solution of equation (24) gives

Kd 1+ý2[2 ý aý4+ ý ]
(25)
äiao(-ý(2)
equation (25) gives the two values of K of which that of positive

value is used as discussed before thus

r2P2 ]
Ka1+ C'2 + R4+/2 (4 of-ot2) ('tý)
2

For no cross flow

[ 2+ )! 1 I
K 1+eb2 i«
(27)
ea
_A+
#
2
bjý2

It was found that the vaiub of depends on flow parameter.

For the confiturations studied was varying between 0.75 at high

values of flow parameter and 1.0 as tho flow parameter approaches

1.0, Figures A2-1 to A2-5.

 26.

These figures show that a suitable mean value of qa0.8

is covering most of the flow range for the configurations studied.

So substituting by this value of ýý. O.8 equations 26. and 27

becomes:

With No Cross Flowi

Pa
I-. [1 I I
K= 1 +. + 1+1.17a*2 (28)

With Cross Flows

C2 2+4j T+ V(40(-
Ka1+0.64 1.56 02)) (29)

Values of C and 0 can be derived with this new value of

using equations © and 9 with cross flow and 16 and 17 with

no cross flow.

F.esults of this analysis are calculated in Tables 6- 10,

represented graphically Figures 12 - 16 and 22 - 25.

 27.

Section

EaCPERIMFi1'PALIVESTIGATION

7.1 Rig Description and Developments

The rig used for the investigation, Figure 26, Plate 1, was

located at Allis Chalmers Compressor Test House, which was mainly

a 4.5 inch diameter steel pipe supplied with air from an Allis Chalmers

Rotory Vane Compressor. Compressor delivery 3120 cu. ft. /min. of air at

pressure 7 lbo per sq. in. (g) and running at 412 r. p. m. The test

section was inserted along the pipe 13 ft. from feed point

horizontally. Air control was made by 4 main valves, Figure 27.

(i) spill valve (s. v. ) at compressor outlet

(ii) upstream 4.5 inch butterfly valve (U. V. ) located

at rig inlet
(iii) Downstream 4 inch butterfly valve (D. V. ) located

8 fte downstream test section.

(iv) Bleed 2 inch gate valve (. B. V. ) located downstream..

test holes.

For rig details see (Ref. 7) Volume II0 and the author's design

drawings of rig scheme and components. Due to space limitations,

the rig was extended towards north direction, raised up to a level

of 445 fto from the floor.

7.2 Test Section Description and Develorments

The perspex section 1.25 inches high, 8 inches wide and 18 inches

long, used for jet deflection study (Rof. 7 ), was slightly

modified
 28.

for the present investigation as shown in Figure 28, Plate 3.

From static pressure survey along test section, Figure 29a to f

due to the large range of slot sizes, it was found that the position

of the constant approach static pressure value varies with slot size

used. So 12 static tappings 0.02 inch diameter were added to the

existing 5 tappings in the longitudinal direction with duct centre

line, and distributed from station 12 which was opposite to hole

centre line as shown in Figure 28b.

Plenum Chamber

For the necessity of accurate flow measurements through test

holes, especially with smaller hole sizes and very low pressure

ratios. The upstream and downstream orifice plates are not accurate

enough, and so the need of small size bleed orifice downstream test

holes gaire rise to the design of a perspex plenum chamber

7,25 x 7.25 inches cross sectional area and 12 inches deep, shown on

Plate 4. It w. s made from 0.25 inch thickness perspex sheet with

1 inch thickness flange 8.75 x 10 inches cross section, to fit tightly

in the test section instead of the perspex frame Plate 6, which was

used for high pressure ratio tests. Also it was used for visualisation

study of jet profiles for various hole configurations.

With both the plenum chamber and the frame, test plates were

flush with duct floor. Using enough studs and a suitable packing for

adequate sealing of the system. The plenum chamber 2 inch I. D.

perspex pipe was flexibly connected to the 2 inch diameter bleed

pipe ending with the bleed orifice plate and bleed valve exhausting
 29.

into atmosphere. The plemamichhmber connection with the rig is

shown in Figure 26, Plate 2.

7.3 Teat Plates

Test plates used for the investigation 1/16 inch

-were made of

thick aluminium plates (16 s. v. g. ), means of fitting teat plate

flush with duct floor were considered in plenum chamber design.

Sealing alongits edges was don* on plenum chamber side by sealing

strips of Bostick 5.

Teat holes were designed with an aspect ratio of the approach

stream (6.4/n) in the range of 2.5/1, to avoid vortex formation in

the flow through the holes, (Ref. 8 ). A uniform approach stream

was achieved by equi spacing each set of holes, which was taken an

odd number with central hole-centre line opposite to static tappings

row. Holes have square edges. :.wallest holes have a large number

for accurate measurements.

The details of the 20 configurations invedtigated are shown

below divided into three groups on basis of hole shape; Circular

holes, oval elöts and rectangular clots, covering area ratio range

from 0.0595 to 0.600 and slot aspect ratio from 1 to 8, Plates 7,

8 and 9.

(i) Circular Boles t Code C- Plato ?

Five hole configurations

Hole coder Cl C2 0 C4 CS

No, of Holest 77553

Hole diameter incht . 329 . 472 . OCIT . 965 1.59

Arca ratio a*t . 0595 . 1223 . 2542 3610 5950

. .
 30.

(ii) Oval elotst Code 0- Plate ß

Aspect ratio

1+ bý8

Seven slot configurations divided to 2 subgroups:

(a) five configurations with the same slot aspect ratio of

2.0, with similar number of slots and slot area to circular

holes W.

Slot code: 01 02 03 04 05

No* of e16to: 77553

aI m b' inch: . 218 312 531 656 1.062

. . .
Area ratio as . 0595 . 1217 . 252 384 604
. .

(b) For aspect ratio ranco from 1 to 8, with same area ratio

of 0.252 and same number of slots 5 each. Slots code 03t 03

covers aspect ratio of 1 and 2 respectively. Slots Code 06

and 07 to cover aspect ratio of 4 and 8 respectively.

 31.

Blot aspect ratio 1248

Plate code: C3 03 06 07

No. of slots: 5555

a inch: 807 dia. 531 375 25

. . . .
b inch: 807 dia. 531 1.122 1.85
. .
Area ratio a*: . 2542 252 252 252
. . .

(iii) I'ectanoý;
ulni sli Code R- Plate 9

b
ei
slot aspect ratio

u b/a

Eight slot configuration divided to 2 sub-group©s

(a) Five configurations of aspect ratio 2.0, and similar in

number and slot areas to circular holes and group (ii - a) oval

slots

Slot code: RI R2 R3 N4 R5

No. of slots: 77553

a inch: 206 295 500 62 1.0

. . . .
b inch: 412 591 1.00 1.24 2.0
. .
Area ratio aes . 0595 . 122 250 384 600
. . .
 32.

(b) For aspect ratio range from 1 to 0, with same area ratio

of 0.252 and same number of slots, 5 each. Slots code R6, R7

and RS together with slot code R3 for aspect ratio of 2.0.

These are similar to oval slots (ii b)

-

Slot aspect ratio: 124a

Plate code: RG R3 R7 IRS

No* of slots: 555.5

a inch: . 7CS 25
.5 "355 .
b inch: .? C6 100 1.42 2.0

Area ratio: . 252 . 250 . 252 250

.

7.4 Inez ents and Calibrations

._
7.4.1 Pressure measurements

7.4.1.1 krnroaoh static nressuret A complete study of static

pressure distribution along test section Figure 29a to f and

opposite to hole centre line was measured by 17 static tappings.

Figure 28. We to frictional losses and change in velocity profile,

the approach static pressure should be measured as near to test hole

as possible. Since the static pressure at the wall opposite to the

centre of the hole, measured at station 12 varies with flow through

the hole. This approach static pressure was found to be steady at

station 2,7.875 inches upstream hole centre line.

7.4.1.2 Jet tatio rressuret Due to practical difficulties of

measuring this pressure at the jet vend contracta, it was assumed

 33.

that this pressure is equal to the surrounding pressure measured

by static probes made of 1.6 mm hypodermic steel tubes with static

tappings of 0.0135 inch diameter located tangential with jet profile

plates 4 and 6. Tap location is not critical in this case due to

the absence of parallel flow on the downstream side of the plate.

Secondary flow effect inside taps was neglected due to the very

small sitze of these taps.

7.4.1.3 Total and dyn-imic pressures: These were measured by

ordinary pitot tube with square cylindrical nose . 039 inch O/D and

. 028 inch I/D. The tubes arranged in two sets of 7 probes each,

1 inch cipart across the test section and located at station 3;

6.875 inches upstream hole centre and station 15; 2.625 inch

downstream hole centre. These were designed to move vertically

across the duct to allow velocity survey across the duct, Plate 3.

Manometers: Three types of double tube j inch I/D, Plate 1.

(a) Inclined water manometers for accurate measurements of small

and negative pressures up to 10 inches of water.

(b) Vertical water manometers for measurements of intermediate

pressures 55 inch range.

(c) Vertical mercury manometers for measurements of highor

pressures than 55 inch water.

Errors with water manometers lies within ;t0.1 inch water.

7.4.2 Temperature measurements

Aix temperature was varying with delivery pressure and running

time of the compressor. Temperature was changing within a small

 34.

range 25 - 50°C. All temperatures were measured by mercury

expansion thermometers 550C range.

7.4.3 Flow rate measurements

Air flow rates were measured using sharp edged orifice plate

with D and D/2 tappings according to B. 8.1042 (Rof. 10 ). Three

orifico sizes had been used, 3 inch, 2.25 inch for upstream and

downstream ends. For accurate measurements with call test hole

sizes, 1.25 inch bleed orifice plate was used on the bleed pipe.

Orifice arrangement is shown Figure 27. Flow characteristics of

these plates are detailed in Appendix 3.1. The characteristic

formulae used are suzarieod below.

3-inch orifice plato dia., m- . 445

L.. OfL
C 100 Zý hVP3 LB/Hp

2.25 inch orifico plate dia. # ma0.25

LoC fL C 56.3 ZAVp LB/Hr

1.25 inch orifice plate dia.. m "" 0.39

Li3/H r
LvCtL
[18-3 xZIYI

wh®res
L. Air flow rate LB/JZ'

CfL a Flow rate temperature correction factor

ZW Combined multiplier

hW Differential pressure across the plate inches of water

p. Absolute pressure at the high pressure tapping inches

of mercury.
 35.

Orifice plate calibrations:

A separate calibration for each arranjement used in the

experiment has been done. The calibration curves are shown

Appendix 3.1. Errors lie within ± 22.

7.4.4 Velocity Measurements

Approach velocity was calculated by the measurement of approach

dynamic pressure as the difference between total and static pressures

at station 3.
[ii
Vý. cfv 369 p Pl) ft. /see.
1

Jet velocity was calculated by the measurement of jet dynamic

prescuro as the difference between total approach pressure and jet

static pressure
(P1 p)
Vý a 18.29 g--J) ft. /sec.
3

where Cfv Temperature correction factor

Pl + Pl = Approach dynamic pressure inch H2O

Pl Pi Jet dynamic pressure inch H20

a Approach static pressure inch 8

Qpi Jet density 1b. /ft3

Pitot tube calibrations

Details of pitot tube characteristics and calibration together

with the derivation of above equation are shown in Appendix 3.2.

 36.

7,5 Deflection angle measurement and Jet Visualisation

The method used in (Refs. 4 and 7) was continued in this

work at the high pressure ratios with test section arrangement,

Plate 6. The carbon powder was replaced by "saturn yellow

sunbonded fluorescent pigment series A", giving clearer picture

and cleaner use. The method is simply by mixing suitable portions

of the above powder and paraffin, then coating a thin aluminium

plate (20 s. w. g. ) with the mixture, evenly distributed using a

special roller.

With the wet coating, the plate was inserted below the test

plate normal to its surface, i. e. in the XZ plane with one of its

straight sides bisecting th@ jet. Care should be taken for that

side not to touch the test plate otherwise the effective hole area

will decrease. A gap k inch is sufficient. Also the plate was

inserted and removed while the compressor is not running, with

sufficient time allowed, for few seconds after the compressor gets

its running speed. Thus a clear jet pattern was formed on the

plate surface as shown on Plates ._ 12 to 2L1..

Deflection angle (j) was measured directly using an accurate

bevel protractor, with its edge running in line with jet core axis

as shown on Plate 10.

At low pressure ratios, a simple mechanism was used on the

plenum chamber; by trying to put a rotating deflector across the

jet pattern within its core. Then the angle can be measured

directly on a transparent protractor stuck on the inner surface of

 37.

the plenum chamber,Plate 11.

The deflector was made of a steel shim 42 s. w. g. (. 004") thick,

1 inch long and 'inch wide. It was soldered at right angle,

through a thin groove at the centre of a hypodermic tube 1.6 mm

diameter, forming one plane with deflector surface. The tube was

supported * inch below and parallel to the test hole surface, and

free to rotate arround its axis with a pointer at its and opposite

to the degrees scale.

To avoid gravitrltion effect on the measured angle, due to shim

weight, it, was balanced by two tin spots on the other side of the

axis opposite to the shim.

The errors in both. systems depends on the reading accuracy and

was within ;t1 degree.

7.6 Preliminnry work

7.6.1 velocity profile

Some preliminary work had been attempted for the study of flow

inside teetieection. From velocity survey of the approach stream,

with the use of round 7/32 inch I/D, 1 inch long honeycombs located

22 inch upstream hole centre line, the resulting velocity profile

was fluctuating within q4 of the mean velocity 100 ft. /sec. measured

at duct centre line. These fluctuations will cause uneven pressure

distribution across test holes.

Since mesh gauze screens reduce these fluctuations more

effectively duo to large number of very small jets produced and

rapidly blend, producing relatively uniform stream although pressure

 38.

drop is high. Several trials had been attempted using several

mesh screens at different positions upstream (see author's lab. book),

the best arrangement was found with 16 mesh screen located 4 inch

upstream the honeycombs together with 40 mesh screen located 11 inch

upstream halo centre, this arrangement is shown Figure 30. The

resulting fluctuations lies within 4.7,' of the mean velocity

130 ft. /sec., Figure 31. The velocity profile across test section

Z direction, is shown Figure 32. The resulting static pressure

distribution is shown Figure 33. Jet static pressure measured all

over the sot of holes was varying within j 0.5% Figure 34. Jet

dynamic pressure was also measured across each hole and showed

negligible variation ± 0.5%.

Preliminary work for static pressure survey with the 5 stations,

arises the addition of more 12 stations for pressure measurement

test as discussed (Section 7.2).

along section

7.6.2 Deflection angle

For deflection angle measurement over a set of holes, preliminary

work showed that the angle variation from one hole to another was

negligible and within ±1 degree. Also the angles measured by the

mechanism, Plate 11, were checked using fine tuft pieces hung from

duct floor through the holes with short length to lie within the

potential core. Wall effect on side holes was considered in hole

arrangements. Several tests with different deflector lengths using

different hole configuration sizes gives a good deflector length"'of

1 inch with minimum pointer vibrations even with smallest holes.

 39.

7. G.3 Prey

According to the type of compressor used, lubricating oil

was necessary for cooling and lubrication of its moving parts.

This oil mixes with the air inside the compressor and then separated

by an oil separator at compressor delivery. But oil vapour still

harried by the air which forms oil droplets after some time of

running. These droplets forms on the tips of pitot probes and

static tappings, so care should be taken to blow those droplets

from time to time especially when significance pressure readings

are notices on the manometer panel. Also to clean the inner surface

of test section when needed.

Accurate measurement of deflection angle needs to check the

free rotation of the pointer about the axis, and the zero zvading

before start. Also the right setting of the plate and the bevel

protractor in the correct position.

The rig should be leak proof, so leakage was checked at

different connections, especially the test section connection with

the plenum chamber using soap solution.

7.7 Test procedure

For each test plate arrangement, two tests werecarried out as

follows, roter to the rig flow diagram Figure 27.

(a) With no cross flow (D. Q. fully closed)t

With this arrangement all the approach flow pass through the

hole i. e. o< .. 1.0. The approach duct flow was gradually

increased and consequently increasing pressure ratio across the hole

 40.

and the flow Va;,. The duct approach static pressure was

kept constant at a value of 33" Ng absolute.

With each setting, the following observations were made:

(i) Upstream pressure, temperature and pressure drop across

orifice plates U. O., D. O. and 8.0.

(ii) Constant value of approach static pressure from the

static pressure survey along the duct.

(iii) Jet static pressure.

(iv) Dynamic pressure drop across the test hole measured at

station 3,6.875 inch upstream hole centre.

(v) Static pressure drop across test hole.

(vi) Dynamic pressure of the approach stream measured at

station 3.

(vii) Deflectionangle as shownsection 7.5.

(b) Wjtticross flow:

With this arrangement, a fraction o[ of the approach flow

passed through the holes so that 0< 1.0, at two constant

of 80 and 130 ft. /sec., and constant approach

approach velocities

static pressure of 33" Hg abs. The approach mass flow was adjusted

to give the required value of V1 by means of the spill valve to keep

the same approach boundary layer conditions. The value of « and

hence the flow parameter was controlled by means of valves D.Y. and

B. Y.

Some observations were made as above with case (a). At high

 41.

values of pressure 'ratios control was madeby means of valves

s. v* and D.v.

The method used for reducing experimental data in terms of

bleed ratio o( , bleed to area ratios flow parameter g

,
'r-
and discharge coefficient C, is shown in Appendix 2.1.

Cý
MISSING

PRINT
 42.

Section 6

"NTATION OF RF.StJLTS AND DISCUSSION

PRESP.

8.1 Correlation of results

Discharge coefficient and deflection angle were correlated to

the dimensionless flow parameter K, which is the ratio of bet dynamic

pressure to approach dynataio pressure. The flow parameter is an

indication to the flow range of the hole with minimum limit of 1.0.

At this limiting va1t& the pressure drop across the hole is zero

and pl/Pj = 1.0. K increases with increasing pl/psr and with

decreasing V1 for a given pressure ratio. Appendix 1. Also K

increases with increase in of and decrease in at.

The combination o& is the bleed area ratio as of /a*.

of , a*

So J''k increases with increase in of and with decrease in a*, i. e.

increases with increase in Figure 22.

2 Discussion of theoretical analyses

. 2.1 Ant1 aim

v
This analysis assumes no momentumloss in tho axial direction.

Although the solution is somewhat far from practical results as

hown in the behaviour of K, Figures 17 and 22. It gives a good

nd to the variation of C and () with flow and geometric factors.

(a) With No Crose Flow

This is the cace of final entry holes of the liner where all

the approach flow passes through the holes. Within the theoretical

range investigated, . 05 ,< a* ' . $. The main factor governing

 43.

C and () is the area ratio. Covaries from . 44 with a* a .8 to

.5
with a* 4 . 05, Figure 12. () varies from 79o with a* a to 890
.8
with a* n . 05, Figure 14. Figure 15 shows the variation of 4o and

K with no cross flow.

(b) With CrossFlows

Figure 16 shows the variation of C with K. As K tends to

1.0 0 approaches zero at high rate of change. At high values of K.

C increases slowly with increase in K. The limiting value of C®0.5.

Figure 17 shows the variation of r and K. Figure 18 shows that

for a given configuration, C increases with increase in of and hence

pl/pi. At low values of o? the rate of change is very high with

smallest hole and small with the largest hole. At high values of oC ,

the rate change is small, and C approaches Cc as two to 1.0.

of

Figure 19 shows the high rate of change of 0 high

with of with

values of a*. With a* a . 8,0 changes from 33o to 80o as q

changes from to 1.0., while with small value of a* w . 05 0 changes

.2
from 83 to 89 as of changes from .2 to 1.0. Figure 20 shows the

() K for
complete variation of with all values of o( and a*.

ll't low flow ranges, with the care value of K, the largest holes

give top values of 4) . This is because the value of for both

holes is the me Figure 22. Bence as o( =/k a* and o{ is larger

with larger holes, equation 9 gives the value of 1P as

Q7 j2- ýl
. coä
2V'K
4) increases
i. e. at same value of K, with increase in of and
 44.

hence with increase in a*.

Figure 21 shows the direct relation between of

, a* and !

which gives great help in the application of data.

Equation 20 shows the relation between C and as

("")
C ý c0 8 `i. ' äý
2 -oý

This relation shows that for a given hole configurations

with increasing os ,C increase due to the increase in C and

hence in jet area.

8.2.2 AMI- sis 11

With this analysis a broader approach to the practical case

is achieved by considering the axial momentum loss. Figure 22

shows the complete agreement between theory and experiment-with

this analysis. Figures 23-25 show the variations of a d?

and

with b(, a* and K. The trend is the same as with analysis It

but the values are higher for C and lower for 4) the given
with

flow and geometric conditions Figures 12-16.

0
 45.

8.2.2.1 Momentumloss factor

As discussed in appendix 2-2, the effect of axial momentum

loss with analysis II is indicated by lo" .c varies within

. 75 with high momentumloss at high values of * K# and 1,0 with

low momentumloss at low values of K (Figures A2-1 to 5).

C
Although was considered as 0.8 in analysis II9 this was

the mean value covering most of flow parameter range, and was

used for comparison. For more accurate solution, exact values

of should be used within the flow range investigated.

ca
The effect of is significant on the value of K as it

appears as a second order factor in equations 26 and 27.

More work can be extended for further study of this

momentumeffect in both X and Z directions.

8.3 Discussion of e2Merimental results

8.3.1 Discharge coefficient and et deflection for circular holes

(a) With no cross flow:

Figures 35-41 show the variation of PI/Pi with C and C'ý) for

five circular configurations with . 059a'. "595. Figure 40 shows

that smallest hole with at _ . 0595 have the largest value of C0 up

to 0.7 at p1.12 while the largest hole with at _ . 595 give the

smallest value of Co = . 58 at samepressure ratio.

Figure 41, shows the same effect on 4) while the smallest hole
 46.

have a value of 0p
# sbo at pl/pj - 1.12, the largest has a value

of Cpo s 770 at same pressure ratio.

The effect of pressure ratio is small especially with large holes,

but both C0 and 4)0 increase slightly with increasing pressure drop

a. T033 the hole.

The effect of a* can be explained as follows. With decreasing

hole diameter for the same hole thickness, the flow passage inside the

hole approaches flow inside a short pipe with pipe length equal to

hole thickness and with diameter equal to hole diameter. This

configuration gives the jet a guidance to flow along hole axis.

Thus <ýp increase and C0 increase due to the increase in jet area

at versa contracts following s The range investigated covers hole

o.
diameter to wall thickness ratio, from 5.25 - 25.4.

(b) Withcror flow

Figures 35-39 show the variation of PI/Pi with C and for

two approach velocities of 80 and 130 ft. /sec. For a given hole

configuration at a constant value of V1, the rate of chance of C and

4) is small at high values of pZ/pj. As Pi/Pi approaches unity

the rate of chance becores large and at the limit at pl/pj = 1.0

C and 0 have a nui values. q is

both The effect of bleed ratio

the same as that of precuiro ratio Figures 111 and 120 respectively.

For a given configuration with constant value of pl/pj C and 0

increase with decrease in V1. The chance is small ei high pressure

ratios and increases as p2/p3 decreases and dimirishea as pl/pj

approaches 1.0.
 47.

Figures 42-46 show the variation of. C and ( with flow

parameter.

Figure 47 shows that wallest holes have- highest value of

dicchario cooz"ficient over the whole range of flow parameter.

Figure 413shows that cmall holes have top values of d) with

no cross flow as the top ends of the curges indicate. At. low

values of K, the largest holes given the top values of (p

. The

reason is discussed at Section 8.2.1.

8.3.2 Dirch {e coefficient and jet deflection for oral slots

(a) with n cross flow

Figures 49-55 show the variation C 4) for

of pl/pj with and

five configurations with same aspect ratio of 2.0 and with

. 059< a. <, . 604.

Figure 54 shows that the smallest slot with a* 0595 given
.
higher value of Co up to 64 at pl/pj W 1.12. The lar est with
.

a* s . 604 give a low value of CO ... 6 at came pressure ratio.

Figure 55 shows same effect on 4) while the eallest clot

.
has a value of (;; Poa 84° at pl/pj = 1.12 the largest slot has a value

of CPO m ?8° at the same pressure ratio.

The effect of pressure ratio is gall, but both CO and

increase with increasing pl/Pj

As ratio effect t
,t
Figures 51 and 56-59 show the variation of C and with pl/p,

for 3 configurations of the same a* = . 252 and a mw numbor of slots.

The aspect ratios are 2,4 and 8 respectivoly.

 48.

3'igure 58 shows that at a given value of pl/p,, Cc increase

with increasing slot acpoct ratio. But the slot of aspect ratio 4

gives the m3cimun values of Co over the pressure ratio range. With

geometric study of these configurations the major dimensions are

respectively 1.06,1.497 and 2.1 inches. The increase in value of

C0 is clear with the first two slots, since the depth of the approach

stream is in the range of duct depth. With the last slot, the

approach stream has the whole depth of the duct, so it is more

disturbed by boundary layer thickness before passing through the slot.

Figure 59 shown tra same effect on P but the last slot

c,
with aspect ratio 8 gives highest value at higher pressure ratios.
(b) With croso flow

Figures 49-53 and 56-57 show the variation of C and with pl/pj

for the same approach velocities of 80 and 130 ft. /see. The betr}viour

in the same as with circular holes.

Figures 60-66 show the variation of C and ( with flow parameter.

Figure 65 shows that smallest slots have high values of discharge

coefficient at high values of K. With largest slot, C has the top

values over low flow parameter range. This is due to the increased

slot major dimension in the direction of flow.

Figure 66 shows that small slots have highest values of (P with

no cross flow. But at low values of X the largest slots have top

values of 4) as with circular holes,

Asioctlatio effect

Figures 51 and 56-57 show the variation of C and C1)with pl/p3

 49.

for the slots of different aspect ratios.

Figures 62,67-70 chow the variation of C and C7 with flow

parameter.

Figure 69 shows that slots with large aspect ratio have high

values of C all cvcr the range of flow parameter. With slot major

dimension within duct depth. As slot major dimension exceeds duct

depth, C starts decreasing all over the flow range.

Figure 70 shows that slots with large aspect ratio have top

values of all over low ranges of flow parameter. : lots with

low aspect ratio have high values of Ct) all over high ranges of

flow parameter.

8.3.3 Dischar
r e_Coefficient . ir andr jet 4' riection for Slots
.r..rý
. . ýýrrrrReetaWzlar

(a) With no crosflow

s
Figures 71-77 show the variation of C an 0 with pl/p,

for five configurations with same aopoot ratio of 2.0, and with

0595 -a* .. 600.

.
Figure 76, chow that small slots have high CO values.

C0m . 69 with a* a . 0595 at Pi/Pi = 1.12. Largo slots have low

values of Co " . 59 with a* ß . 600 at sa:Ae pre cure ratio.

Figure 77 shows the camp effect on L'hile the smallest slot

has a value of (Po = 8t3°g the larg^ost have 'Po a 7e at SFr prof sure

ratio of 1.12.

increases both Co and q)o increase with

Also as pressure ratio

a small rate for a given configuration.

Aspect ratio effect

Figures 73,0-82 show the variation of C and with

 50.

pl/pi for four configurations with same a* R . 25P and same number of

slots, 5 each, and four aspect ratios 2, It 4 and 8 respectively.

Figure 81 shows that C increases with increase in aspect ratio

over the pressure ratio Square slots (aspect ratio

all range. * 1.0)

have the lowest value of at low values of pressure ratio.

Figure 82 shows the same effect on () as with oval slots.

Square slots have the lowest 0o

value of all over pressure ratio

range. The effect of slot major dimension is clear, as the slot

(o
with b/a .4 has the top value of as discussed with oval slots.
(b) With cross flow

Figures 71-75 and 78-80 show the variation of C and 4) with

pl/pj at the same approach velocities of 80 and 130 ft. /sec. The

behaviour is the same as oval slots.

Figures 83-407 show the variation of C and4. )with flow parameter.

Figure 88 shows that at high flow range, small slots have top

value of C all over this range. At low flow range large slots have

top values of C all over this range except with the smallest slot.

95 shows the variation 0

Figure of with K. Like oval slots,

small slots have top values of (p with no cross flow. But with cross

flow largest a lots have high values of ( at low flow ranges all

over this flow range.

sr)ect ratio effect

Figures 73,78-80 show the variation in C and 1P with pl/pj for

slots with aspect ratios 2, It 4 and 8 respectively.

Figures 85,90-94 show the variation of C and with flow parameter.

 51.

Figure 93 shows an increase in C with increasing aspect ratio

all over the range of flow parameter. The square have the lowest

value of C all over the range of K.

Figure 94 shows that at high flow range ýP increases with

increasing aspect ratio. Slot major dimension is still affecting (

At low flow range 4) increase with decreasing slot aspect ratio,

although the slot with aspect ratio of 8.0 still has top values of

() over most of this range.

8.3.4 DisWArgw coefficient gE4 et de ion for vari a nfi ratio

In this section each figuration is compared with other similar

configurations of different shapes.

8.3.4.1 Holes and equal areas and mmben

-slots
Circular holes are compared with oval and rectangular slots

having aspect ratio of 2.0. The comparison is based on equal hole

area and number of holes.

Figure 951 7 holes and a* "" . 0595 show that ovals have higher

values of C over high flow range, while circulars have top values

over low flow range. fiectangulars lie in between circular and oval.

The effect on CP is the same but rectangulars have top values

of ( at high flow range.

Figure 961 7 holes and a* a, . 122 show that circulars have top

values of C and 4) over most of the flow range. Reotangulars have

lowest values. Ovals lie in between.

Figure 971 5 holes and a* ".. 252 show that circulars still have

top values of C all over flow range. Ovals have lowest values of C
 52.

at high flow rare, rectangulars in between.

The effect on () show at high flow ranges ovals hive highest

have the lowest and rectangulars in between.

values of , circulars
At low flow range circulars have top values of 4) , rectangulars have

the lowest and ovals in between.

Figure 93; 5 holes, a* b . 334 show that ovals have top value of

C all over flow range. Circulars have lowest values and rectangulare

in between.
4,
The effect on () show that circulars have top values of

roctangulars have lowest and ovals in between all over the flow range.

Figuro 99; 3 holes, a* ... 600, show that ovals have top values

of C all over flow rangge, circulars have lowest values and rectangulars

in between.,

The effect on show that circulars still have top values of

especially at low flow range. Co significant change for ovals and

rectangulars all over flow range.

8.3.4.2 ; lots with equal-areas, numbers and aspect ratios

Figure 100; 5 holes a* " . 252, aspect ratio " 1.0, comparison

between circulars and squares. This shows that circulars have top

values of C all over flow range.

Effect on 4) show that circulars have top values of (P over low

.
flow range and squares have top values over high flow range.
Figure 101; 5 slots, a* n . 252, aspect ratio . 2.0 show that
.
have top values of C over lox flow range. Reotangulara have
ovals

top values over high flow range.

 53.

Effect pn0 show that ovals have top values of 4) than

rectangulars all over flow range.

Figure 102; 5 slots, a* a . 252, aspect ratio v 4, show that

ovals have top values of both C and than rectangulars all over

flow ranvge.

Figure 103; 5 slots, a* a . 252, aspect ratio 8, show that

a

ovals have top values of C than rectangulars over high flow ranges,

the change is small.

Effect on A) show that rectangulara have top values of P than

roctangulars over high flow range. Ovals have top values over low

flow rams.

8.3.5 Analysis of experimental results

For all configurations studied, with a given hole geometry, both

C and Q) increase with decreasing V1 at constant pressure ratio.
C and q) reach the top values as V1 approaches zero i. e. for small

configurations with no cross flow. With cross flow for a given approach

velocity C and 4) increase increase. The effect

with pressure ratio

of pressure ratio is small with no cross flow especially for large

configurations. But with cross flow pl/pJ effect is large at very

low pressure ratios and small at higher pressure ratios. Bleed ratio

effect is the same as pressure ratio effect. As p1/pp approaches 1.0,

or of approaches zero, both C and 4) With

approaches a nul values.

a given flow factors, C and () generally increase with decreasing

hole geometry, i. e. with decrease in a+.

The combined effect of flow and geometric factors IS the flow

 54.

parameter 8. For a given hole configuration C and C increase with

increasing flow parameter. At low flow ranges large configurations

have higher values of C and 4) for circular holes and oval,

rectangular slots with aspect ratio o. 2.0. Other geometric factors

affecting C and 4) for different configurations ares

Ratio of hole diameter to hole thicknesss For circular holes

both C and 0 increase with decrease in this ratio.

Slot major dimensions With narrow slots (high aspect ratios),

as slot major dimension exceeds approach duct depth both C and

decrease.

Aspect ratio effects With slot major dimension within approach

duct depth an increase in slot aspect ratio discharge coefficient all

over flow range, and increase 4) at high flow ranges.

Effect of-hole zeometry (share)s At low flow large

ranges with

configurations, circular holes have higher values of C 0 than

and

ovals (with aspect ratio Rectangular slots have the lowest

."2.0).

values. square clots have lowest values of C and C) than circulars

over most of the flow range.

Generally oval slots have higher values of C and (P than

rectangular slots over most of the flow range.

8.4 Com aon14t exnerinental results with theo; Z

8.4.1 Dischame Coefficient

Figures 104,105 show a good agreement between theory (Analysis It)

and experiment for circular holes, especially at low values of a+ with

no cross flow and at low values of K with cross flow. At high values
 55.

of 8 because the momentumloss factor is rather than 0.8 as shown

in Appendix 2.2.

Figures 106-110 show theoretical results quite close to

experimental results for all the configurations studied. The slight

difference is due to the effects of q and the aspect ratio effect

which have not been studied theoretically.

8.4.2 Jet Deflection

Figures 112,113, show quite agroement between theory (Analysis II)

and experiment for circular holes, especially at low values of a* with

no cross flow and low values of K with cross flow, as discussed with

discharge coefficient.

Figures 114-116 show the theoretical results within the

experimental region for all configurations studied. The reason for

the slight difference is the same as above.

8.5 Camri$on of experj ental results with previous work

.
8.5.1 Diec coefficient

Figures 104-106 show good agreement between experimental results

and those by Dittrich and Graves for circular holes (Ref. 5) all

over the flow range. Figure 105 show comparison with Cox results

(Ref. 9 ).

Figures 109,110 show results by Dittrich and Graves for

rectangular slots (Ref. 6) lying within the experimental region

for the investigated configurations, all over flow range.

Comparison with Fielding's results for circular holes, (Ref. 8)

gives quite agreement Figure 111.

 56.

The slight variations between author'a work and previous work

is due to different conditions of experiments, and different

apparatus geometry.

6.5.2 Jet deflectio_n

Comparing with Candam hardja'a work for circular holes (Rot 7 ).

Figures 112,113 chow quite good agreement with no cross flow.

Figures 119,120 show Gandamihardja's remit quite close to author's

results. Figure 120 show Gandamihardja's results are slightly higher

especially at high flow parameter range i. e. with smaller holes. The

reason for this may be due to the alight error he made in measuring

the angle due to the partial bloc' go of the test hole with the

measuring plate touching hole surface. This effect is more

significant for ¬rnaller holes.

Figure 113 shows a comparison with coicresults (Ref. 9) for

1
estimating 4) as CO3 V1/Vi. These results are somewhat far

from both theoretical and experimental investigations.

 57"

Section

USF CF DATA

Although these data are based on sere liner flow, it can be

applied satisfactorily to both internal and external flows as the

case of actual combustor, provided that jet velocity is greater than

internal flow velocity and correct jet static pressure is used.

Data are most suitable for low pressure loss combustors, and allows

up to pressure loss factor of lQT across the hole.

9.1 For survey applications

The given data are; Area ratio a*, duct geometry, and approach

flow rate. It is required to check combustor behaviour, i. e. to

find C, and of . To find these parameters, the knowledge of

pressure ratio Pi/Pi and flow parameter K is required. Either pl/pj

and K can be calculated from combustor flow rate, air density, duct

geometry, or from direct pressure measurements. Choose the suitable

similar configuration using (Section 7.3).

geometric
(i) with the value of pl/pj, and calculated approach velocity V1

(for holes rather than final liner holes), C and 4> can be

read directly with suitable a* from

Figures 35-41 for circular holes

Figures 49-59 for oval slots

Figures 71-82 for rectangular slots.

(U) For higher V1 than 130 ft. /sec. use the

reasonable values of

value of K and by interpolation read C and 4)

calculated

directly from
 58.

Figures 42 - 48 for circular holes

Figures 60 - 70 for oval slots

Figures 83 - 94 for rectangular slots.

To find d%; with the value of K calculated, from Figure 22

read bleed area ratio From Figure 21, read of at the given

value of a*.

At values of a* > .6 which is the top limit of this experimental

investigation. Refer to the theoretical investigation Section 6,

with the values of K, o? calculated from above. Equations 8 and 9

give the required values of C andIP .

9.2 For deem applications

The given data are

a- Approach flow rate from compressor delivery and engine

type.

b- Air/Fuel ratio and hence of .

.. p
c- Pressure loss factor across the hole p% and
2
hence PI/Pi

d- Combustor geometry according to compressor and turbine

designs, and combustion efficiency.

It is required to find hole size and C) s The hole shape should be

decided from point of view of mixing, penetration and liner size.

To find these parameters, K must be calculated from air flow rate,

air density and duct geometry or from the ratio of V/V1 using

Figure Al-l. From Figure 22, read the value of P at corresponding

 59.

value of K. From Figure 21, read value of a* at the corresponding

valued of )'- and of

Hole area Ah = a* x Approach area.

With circular holes Ah - . 7854 n D2

With oval clots and suitable value of slot major dimension b' +ä
(according to duct depth).
[slot
Ah an a'2 aspect ratio - . 2146 J
With Reotangular slots; find b as for oval slots

Ah on a2 C slot aspect ratio

Suitable configuration choice can be referred to (Section 7.3)

according to requirements of design with reference to(Seetion 8.3) .

To find tp the value and V1
, with of a* calculated, p1/pj,
(for holes than final
rather liner holes), read (P directly from

Figures as with 9.1 (i).

With the value of S calculated or with higher reasonable values

V1 > 130 ft. /sec. read 4) directly from Figures 9.1 (ii),
of as with

by interpolation.
 60.

Seqtior-10
,
COTIC_, O:3S

The following results were obtained from theoretical a nd

experimental studies of discharge coefficient C, and deflection

angle IX> for various configurations. These results were correlated

with the dimensionless flow parameter K incorporating geometric and

flow factors affecting C and (D.

1. The flow parameter indicate the flow range through the hole

limiting value of K for flush holes is unity. Flow parameter K is

function of area ratio a", bleed ratio oý , both combined as bleed

area ratio-. Also K is a function of pressure ratio pl/ps, ratio

of jet to approach velocities.

2. For a given hole configuration C and 4) increase with increasing

With no cross flow the rate of change is small especially

with larL^e holes. With cross flow C and () Increase with increasing

pl/p3 or c4. Rate of chance is high at very low values of pi/pi

With no cross flow C and ) increase

and oý , with decreasing a*,

and for circular holes C and Cpincrease with decrease in the ratio of

hole diameter to hole thickness.

Axial momentumloss have great effect on both C and all over

the flow range. This effect vanishes as flow parameter approach unity.

4. At low flow ranges, largo configurations either circular or oval

and rectangular with aspect ratio 2.0 have higher values of C and

than small configurations.

5. Discharge coefficient and usually deflection angle increase with

 61.

increasing slot aspect ratio, since slot major dimension is within

the depth of the approach duct.

6. At low flow ranges with large configurations, circular holes

have higher values of C and 0 than ovals (aspect ratio 2.0).

Rectaagulars have the lowest values.

?. Square slots have low values of C and () than circular slots.

Q. Generally oval slots have higher values of C and d) thane

rectangular slots over mast of the flow range.

9. Good agreement have been achieved between experimental results

and both theory and previous work.

 62.

oni
Sect. 11

E6'COMMA'II ,t;3 FM FURT'dFR ýiORK

.,._.. . .... f . ý.. ý. _..

(i) York can be extended for studying the behaviour of other hole

configurations such as triangular slots.

(2) Although geometric and flow conditions studied were not too high,

to cover all types of combustors, this was the maximum limit allowed

by the rig to avoid compressor overloading. Extension of rig

rapacity can be made to allow for higher conditions. Compressibility

effect should be taken into consideration.

(3) More work can be extended for the study of the momentumeffect

in both directions.
 63.

97ECES

'1. Grobman, J. Z., Dittrich, R. T.: "Pressure drop and airflow

distribution in Gas turbine combustors" Trana. ASP4E.Vol. 79,

No. 7. October 19579 pp* 1601 - 1G079

8nvthorn, W. R., Olson, W. T, 8 "High speed aerodynamics and

Jet propulsion". Prinaton aeries Vol. Il.

I'd Callaghan, E. E., Bowden, D. To: "Investigation of flow coefficient

of circular, square and elliptical orifices at high preccure

ratios". N. A. C. A. T J. 19470 1949.

Dewey, P. E.; "A preliminary investigation of aerodynamic

characteristics of =all inclined air outlet at transonic Mach

mucbere". N. L. C. T. N. 3442,1955. (S3upersedoe U. A. C.A. We 153610).

5: Dittrich, R. T., Gravos, C. C. * "Discharge coefficients for

combustor-liner air-entr7 holes. Is Circular holes with

parallel flow". N. A. C. A. T. N. 3663,1956"

6. " Dittrich, R. T., Graves, Co Cot "Discharge coofficientn for

combustor-liner air entry holes. lit Flush rectangular holes,

stop louvers and scoops".

I, Gandamihardja, Os Jet deflection. Thesis work submitted at

College of Aeronautics, Cranfiold. June 1962.

B. Fielding, 2. ß. t "Data on coefficient of discharge of a plain

circular hole ina wall of a duct". Iucaa report Ito. B45,893.

May 1960.
 6L..

RF.:F`E (CO21t11n1@d)

Co:, H. i "Air flow and momentumexchange through the dilution

holes of a combustion chsamber". N. G.T. E. Note No, T. N. 281,1957.

10. "Float measurement" - B. S. 1042,1943.

11. Gracey, W., Lotko, W., Rassel, W. E.: "Wind tunnel investigation

of a number of total pressure tubes at high angles cf attack".

N.A. C.A. T. Pt.2331.1951. (5upexDedes N. A. C.A. R.M. 150 G19).

/ddi ions. tieferences t

12. Anon: "Fluid moters, their theory and application". Fourth edn.

A. S. M.E. Res. pub. A. B. M.E. (N. i. ) 1937.

13. Lefebvre, A. U. a Supplement lectaraa on combustion at

College of Aeronautics, Cranfield.

14. Lord Rayleigh: "Theory of Sound". Vol. II. }aciillan & Co.

15. Pankhurat and Holdere Wind Tunnel Technique. Pitman.

`ý 16. Roshko, Anatols "Some measurements of flow in a rectafsgular

cut out". N. A. C. A. T. N. 3488,1955.

17. Stokes, G. M., Davies , D. D., Sellers, T. B.: "An experimental

study of porosity aharaoteristioo of perforated materials

in nonanl and parallel flow". Z. A. C.A. T. W. 3085.1954.

(Supersedes N. A. C.A. 1t.H. 1531W?).

11. Aerodynamic measurements. K. I. T. Press.

f
APP'T, 2rDICE- 3
 65.

APP.
-MIX

ýaý
Frý,
vv ýWMER-K
"r a'-

It was found that the ratio of jet velocity to the approach

velocity is the main factor governing discharge coefficient and

deflection angle, Figure Al - L. For incompressible flow, this

ratio is simply the. square root of the ratio of jet dynamic pressure

qj The approach dynamic pressure ql, thus

Z
'V
lý qJ/ql
tl'. - ýv ' A1.1
i
The flow parameter was defined as the ratio of jet dynamic pressure

to; approach, dynamic pressurep and for incompressible flow

g3/qj Al. z
.ý 2
or v2 Al A1.3
j
I--" L:ý
With an error less than within the experiment conditions,

equation A1.3 is represented graphically Figure Al-i. When Vj/vd,

i. e. with noflow through the holes, K-1.0 as discussed before

(Section 6)*-- This shows that K is an indication to the flow

through the holes, which has a mini-um value 1.0 for flush holes.

This flow range can be ex ended at low values of K leco than 1.0

by adding scoops to flush holes (Ret. 6 ).

From A1.3, using continuity equation

ý.ý:ý, ':. ,
ý .42
X=2 G2
C a'ý
A1.4
or g 1'`2/C
 66.

A1.4 shows K as a function of flow and geomctrio factors o{

a* which both affects the discharge coefficient. With assumption 3,

Section 6), using equation A1.2.

pl` pi
K
+1 .
ql

!,
Pi -1
1 -V ---- f1
Y12
2

For a given approach density and with astuaption5 (Section 6),

thuß

K"m -1- (p1 +1 A1.5

pj -1)
Y12

where m is a constant depending on and pj.

Equation A1.5 is represented graphically Figure Al-2 with atmospheric

jet pressure and a mean working conditions with velocity ranging from

50 to 200 ft. /sec. Figure Al-2, shows that for any value of V19
pl/pj
when m It in this case K-1.0 and no flow passes through

the hole.

With no cross flow, Yi is a function of a*, pl/pj and C.

From continuity equation,

V1-V2 - Ca*Yj

using equation (7) Section 6

v2/ 1. (1 0
-Q1

i. e. V1 "C a" Vi A1.6

 67.

P ; Pj
But 2g
j

°ýL ?,ýr pl +ý pý
26

ý26
a rMS - 1)

using 11.6

v1 .Cým(-- 1) A1.7

whore m is the constant disaussad before in A1.5.

With no cross flow i. e. a 1.0

V1 . CQa* m (pj , 1) A]..8

Equation Ai. ß is represented graphically Figure Al-3, using a mean

value of Cc a 0.65 for a* range of . 0y a* 0.6. Using this

value of Y1 for a given value of pl/pj, S for no cross flow can

be calculated from A1.5.

A complete study of flow parameter, the factors affecting it

and its relations with discharge coefficient and deflection angle

have been discussed at Section 6, which shove that from equation 13

ZtV4 (4 ý( 2)
Km 1+2r ý'"4 + pa - of

for a given valuo of a*

K- f(0) A1.9

and from equation Al-5

ý)
I{ a (ý ºi Al. lo
 68.

From A1.9 and £1.10, it is clear that for a given area ratio ae,

the blood of is'a function of V1 and Pl/Pj

Pl/pý
ise" " t (ý )
OC 1

and for a given value of approach velocity Yl

'.
0 O4
" Al, u
So the floe paramoter K which in a function of of , a* is

also a function of Yl, pl/pj and a*, i. e. the value of K

reprecenta completely the flow across the hole.

Effect of momentum loss on flow parameter have been

discussed in se ction 6 (Analysis it)

0
 69.

APº-F:MIX 2

Reduction of N. erjj ental Results

A2-1 rý.Mothod
r. r used for calculrttion of results
Discharge coefficient was calculated as the ratio of sweat d

txtsa flow to theoretical mass flow through the holes or slots

C« Lj/Lth. Theoretical mass flow Lth was calculated as the

product of jet density jot velocity V, and hole or slot area Ah.
Yj
was calculated assuming isentropic expansion through the

holes
holen

Pi/Yý pi/ p
11 "

or (pj/filýý lb. /cu. ft.

j1

where lb. /cu. ft. A2. l. 1

1"i
1

Y, was calculated using Appendix 3-2 as

511,
Y, "" 18.2g it. /aoo. A2.1.2

wher© h is the jet dynamic pressure, inches water.

Ah t Hole or clot area cq. inch

nt Amber of holes or slots

Thus Lth " 25 43 Vn lb. /hr. A2.1.3

Ah

Lj t Jot flow ruts Measured as the difference between up'

and downstream orifice plate measurements, with suitable plate

sizes* Using Appendix 3-1 thin value of Lj was checked by bleed

orifice plate coasurementn, thus:

 70.
L
ý..
C. ý. ........ ý. A2. t. 4
25ýj Vj aAb

The floe parameter was calculated as'the ratio of the

calculated jet dynaxto pranure to the approach dynamic pressuro.

Due to the negligible temperature effect, IC was calculated as

follows
ýY-
1%

B1 Y1

Bleed ratio of is the ratio of the measured flow rate through

the holes to the approach flow rate, i. e.

L
oC "" Ll

Dlead/arca ratio
a
):oDentum lose factor q (: 6),
using equation(21)1, roction

ýg " Al.,(pl - pj)/X, ýin 0

Since the moulting force on the hole is due to the pressure

difference aorooa it. For more accurate values of '5 . it is

rccaonded to calculate this force by static pressure curvey

ovor the hole, and integrating over hole croza section. In this

work pl was taken as the static pressure acting on hole centre

ee pý co
?j)/Hj sin
(pl )
a+ - Pj
-x 4.15 104
Lj P, Siti4)
 71.

where

(pi pj) = static pressure drop across the hole inch water

L, jet flow rato lb. /Ur.

VJ jet velocity ft. /coc.

Experimental calculations are shown in Tables 11 to 30

and represented graphically Figures 35 to 120

A2-2 FlorreptuM Ions factor

Refer to (ioction 6), Analysis 11

(') ('3 c2ý

.ý
-1----ý
ý ý
P, r' _ -
i ýj pz
-9
r- ý
i
VI ýI Vz
0
r --ý ß,
"I V'
Z

F'IG. ii' ý
Atȼo; plierc.
From the physical observation of the approach stream lines

behaviour, using fine tuft tachniqu© visualication in the approach

duct. Tho directional behaviour of the approach stream is rather

than axial as ascumed in Appendix 1. As the stream lines approach

the test hole, it starts changing its direction and becomes

inclined towards the holes. This inclination varies with flow

range through the holes, i. e. with flow parcmeter. The effect of

 72.

this directional chan; e is in fact an indication to the momentum

loan in the axial direction, and the amount of axial momentum

exchangod to the atmosphere and hence to the vertical direction.

At low values of K, i. eo with low values of 4) , the axial

momentumloan, is small, and the amount of momentumexchanged is

also small. As the flow range reduces, i. e. K terdo to 1.0,

4) tends to zero and both the momentumloan and amount of momentum

exchan,; od tends to zero. With high values of K, i. e. largo

values of (), the amounts of axial momentumlose and momentum

exchanged becomes large. In the limit at very high values of K,

tends to9 O the axial momentumis completely lost and the amount

exchanged is maximum.

Considering momentum equation in the axial direction the

directional effect of approach momentum is small, because we are

considering the undisturbed uniform upstream conditions. In the

vertical direction the approach momentum behaviour has a

significant effect as discussed before.

Applying momentumequation in the vertical. Z direction

Figure (d)

Sin (Ml - Mj) a Ah (p, - pl) A2.2.1

'.
. SýTIl> (m3- L1V1) m Ah (pl - p3)

writinQ L1, V1 in tarag L3, V3 using Al and A2-1

lr )
n3 äý- ALýP1 " p,

ý [Ah (pi A2.2,2

or M3 ..
cin nÄ.: - pj
d- i
 73 .

Cozpan d with that of &m;lysia rl oztion 4 as

(pl
A, sia
0 a. 16h - p?

! obis another factor arises in the R. 11.3. of equation A2.2.2.

This factor is a unction of the unknown value of IC. Sinco the

solution of the problem in to find the Value of K. thus we rued

to find the value of this factor. Analytical solution of finding

it is outside the ccopp of this work: #

truation &2.2.2 was written in another fora

93 Cin 4) c[, nn(pl - p3)]

.. ßa2.2.3
In equation A2.2.3 thin factor as a function of oý is noted
iI
as which is called the momentumlose factor. Values of 1p for

various eonfigurationa have been calculated using erporimontal work

as shown 1'icures A2-1 to A2-3.

These figures show that the 476

value of varies with the flow

parameter for all configurations studied. The rate of shame in

am-al at high flow parameter values and high at low valuon of K.

0.75 and 1.0. The value of ý-0.8

Also' wan varying within

as considered in Analysis II covers most of the flow Farmmeter

'range. Unless at low values of flow parameter, is rather than

. a. but at this range the effect of cbj becomes ezall as Iä tends

towards 1.0 as discussed before.

 7t.

AP r,.'DTJC 3

CALII3RArIOM

Aj. 1 Orifice Elato charaCteriBtiCs n.M calibration

,
(i) Orifice plate chiricteriatics

According to B. S. 1042,1943 (Ref-. 10) with D and D/2 tapping

AVhW -
L ,. 359.1 x CZ A3-1.1

Whore

L flow rate lb. /hr.

Ca orifice discharge coefficient

Za combined multiplier for effects of Rd" E and D.

RdvE and D are Reynolds number, expansion factor and

pipe diameter respectively

da orifice diameter inches

---ý-
--- velocity approach factor
--M2
1

be differential inches 60 OF
pressure of water at
(undor air)

wo. fluid dcnsity lb. per cu. ft" at working

conditions at high pressure tapping

where
p "" absolute pressure lbo por sq. ft. at high

pressure tapping
OX for air
R" gas constant - 96 lb. ft. per The
0K
T. absolute temperature at high pressure tapping
 75. '

Due to the different flow ranges of the experiment with

bleed range from 0 to 1.0 and different hole areas, accurate

measurement of differential pressure h was attained by using

three different orifice plate sizes arranged as shown Figure 27.

(a) 3 inch orifice with a. 0,445 t0.0) mainly used

at high flow conditions upstream to avoid compressor

overloading.

(b) 2.25 inch orifice with m (D, O)

.n0.25
designed for low flow conditions mainly used at

downstret: n and and at upstream with small hole areas*

to) 1.25 inch orifice with aa0.39 (H, 0)

used at the downstream end of the plenum chamber for

accurate bleed measurement at very low values of

static pressure ratio across the holes, and with

=all hole areas.

For these orifice plates, equation A3.1.1 van arranged in terma

of measured parametero, p in inches of mercury (absolute) and he

0X
Considering constant air to raturn of Sot) and introducing a

flog temperature correction factor C.

CCL Ascuainß constant

absolute viscosity of air 3.88 x1 77 slug per ft. seo. due to the

call change in temperature.

F--
CfL 'SOOT A3.1.2

The resulting flow equations in order are

(U, 0)a L° CrL

[ ita x Zý hYp3 a3.1.3
 76.

(D, 0)b rh ,
. CtL
[56.3 xZ IrT. A3.1.4

[207#8 ; -H ]
.. CtL xZ Y ,. A3.1.41

(B, 0)Q CfL 118.3

= xZFhrPJ A3.1.5

E67.4
- CZL xZY5J A3.1.5
where H= Differential pressure, inch of mercury
These equations are Mreseated graphically Figures A3-1 to

A3-3 and the value of Cfh from Figure A3-4 ucinn equation A. 1.2.

(ii) To1e noes

These calculated values of flow rate allows for an overall

tolerance limit of ± 1.5 .

(iii) Orifice p- ste o.4librationa

For all differentaorifice arrangemonts used in the experiment

tabulated below, calibrations were down as shown Figures A3-5 to A3-7.

Flow Upstream Downstream Maximum

Arrangement %NP
Bleed Orifice
Condition Orifice Orifice
error
I 3 inch dia. 2.25 inch dia. ý 2
No Bleed
2 2.25 inch dia. 2.25 inch dia. r 1.32

NO 1.25 inch dia. 0.628

3 2.25 inch dies.
Cross Blow

4 With 3 inch dia. 2.25 inch dia. 1.25 inch dia. 3.5
5 Cross Flow 2.25 inch dia. 2.25 inch dia. 1.25 inch dia. 4
 77.

A3.2 Pit2t tube characteristics and calibration

(1) Pitot tube characteristics for velocity measurements
compzeasible
Applying Bernoulli equation for flow

P
Ya 29 A3.2.1
ý .

A3.1.2 can be written in terms of measured parametors, at given

station as follows
V. 18.29 h/5 ",A.3.2.2
FT b/p
. 21.3 A3.2.3
Q
cfV [ 369 ý/p 3 A3.2.4

where at the Given station,

V velocity ft. /sec.

CfV "" temperature correction factor for volocity. Since

T is considered constant value of 300°K.

Cfv T1300 13.2.5

T- temperature OX
absolute air
h dynamic pre©sure inches of water mP-p

p absolute air presnure inches of mercury

g- air density lb. per cu: ft.

Equation A3.2.4 is represented graphically Figure A3-8 with the

value of CfV from FiCure A3-9 uainC equation A4.2.5.

(ii) Pitot tube calibration

The meaeure4 velocity represents the local velocity at stagnation

point. So the use of pitot tube for measuring the mean stream

velocity, is done either by velocity measurement survey across the

 . 78.

the section, or by flattening of velocity profile at given section

and calibrating the measured value to the calculated mean velocity

Figure A3-10.

(iii) Yaw effect on ritot tube reading

Calculated velocity from A3.2.1, in a function of nose shape

of the tute, and the yaw angle to streamline direction,

V1 "C VZIF x3.2.6

where
3 are constants.
(Ref. 1 1). showed that for cylindrical
Gracey, note blunt tube,

the ranee of insensitivity to ctipalignment vast 11 to 23 degrees

+
according to ratio of inner to outer diameters of the nose for 0.125

to 0.96 respectively, Figure A3-4. The tube used in this work was

1 mmo/d hypodezzzic tube with L/d 0.020 inch with i/d to o/d ratio

of 0.72. So in equation A2.1.4 can be assumed very close to

unity.

The factor f in A3.2.6 depends on approach stream Mach number.

The value of is effective for sonic and supersonic flows due to

shock formation ahead of the nose. The mAxi- approach velocity

within experiment was such that the approach )ach number swtito exceed

0.115 so, the factor can be considered 1.0. Equation A3,2.1 can

be applied satisfactorily, without considerable error.

(iv) ß. 3'M

For compressible flow relations

if [
P "ý P+ Y2 1+ +
21 ...,
 79.

and for Mach Number 0.115, an error of 0.66, * arises in velocity

calculation by equation A3.2.1 for the assumption of I. 1.0.

 80.
APPENDIX 4

COJTRMATIO;
i Off' GA i11 IHARDJA WORK (RW. 7)
,

Experimental results for circular holes (Ref. 7) showed that

deflection angle 4) of a het issuing into atmosphere from the wall

of a duct with parallel now, is a function of bleed a* U*, and

area ratio a* as shown Figure ll, where

a* .. ratio of hole area to duct approach area

U* . het velocity component normal to the plane of hole.

With replottiv data of ri&ure ll, versus a* with constant

a* U* values, Figure A4-2 shows a aeries of straight lines produced

with a co=on mating point at 0= 80 degrees. This means that

for all a* U* values 4) -> 9 d% a a*--&. 0 and for a given value

of a*, 4) ---º90o as a* U* 0.
-il.
The negative elopes of these lines 4)/a* or deflection parameter

(aa- d))/a* decrease increasing

s with values of a* U* as shown

Figure A4-2, which when replotted on a logarithmic scale gives log )

as a unique function of log a* U* for the range 0.1 F a* U* . 0.7,

and a parabolic relation for the range 0.7 <a* U*ý< 1.0, as shown

FigureAý! -a "

(i) 0.1<a*ü*45-0.7
For the straight line part of Figure A4 3, its equation is in

the form

log 'A - log a+b log a* U* A4.1

A-a (a* U*)b

or A4.2
 81.

where a and b are constants, and by substituting data from

Figure A4-3 in equation A4.1 gives

2.263 loio a-b

2.023 s lo$o a- . 699b

1.888 n lof a- . 523b

'a
1.818 s lo$a a -. 398b

1.734 s loqa a- . 301b

1,686 Q loia a- . 222b

1.602 as lofa a- . 155b

adding first three and last four equations

2.058 = logo a- . 777b

1.710 .. logo a- 269b
.

solving Wand ansously

a= 33.42

b .. -0.69

and the emperical equation for 0.1 a* U* 0.7 becomes

U* a4
(aII)0.69 ý

\
For this range, equation A3.3 gives higher values of so

that let is the corresponding value on the straight line,.,

it
extension, and assume the difference al - .. With
J.
log f versus a* U on a semi logarithmic
plotting scale, gives

a unique function as shown Figure A4-4

 82.

i. e.

log E"C+d (a* ü*) A4.4

which is a logarithmic form of the equation

C+d
ý= 10
A4.5

where C and d are constants,, and by substituting data ! Tara

Figure A4-4 in equation A4.3

2.5 C+ . 7d

7.5 .. C+ . ad

10,2 aC+ . 9d

19.1 -C+d

adding first two and last two equations

. 641 -C+ 75d

.
1.145 .C+ . 95d

solving simultaneously

C. -1.249 =-1.25
da2.52

and the omperical equation for 0.7 <a* U* . 1.0 becomes

Sm + 2.52(a* U*)
10 '"1.25 Aß.6

combining A4.3 and A4.6 by the rolation ý"ý- )/ gives the

general emperical equation for all a* U' range within the area ratio

range 0.0344 C a" s. 0.314

i. e
[.. s4' (330)a* U~
" 6- . 0575 A4.7
(at Ut))'69
 83.

equation A4.7 compared to experimental reeultn gives a good

correlation as shown Figure A4-5.

In terms of deflection angle q) gives

Z3.43 (330)a* U+ºj

69 - . 0575 A4.8
(aýºv*)

With no cross flow

I)o 88 14.43
.. - a* A4.9

i. e. ep is a unique function of a*.

Comparisonof Gandumihardjs work, and author's work for

circular holes gives quite a good agroement as shown FiGurca 412-1.13

and 119-120.
 84.

APPENDIX 5

rý--TtJi)Y

General discussion}

With the development of high output combustors, the problem of

obtaining adequate nixing has become increasingly difficult an

combustor reference velocities and heat release rates have

increased. In the primary zone of the combustor, the rapid

mixing required for high heat release rate is usually obtained by

mans of high velocity air jets issuing into the combustion space

from awirlers and openings in the dome and liner, resulting high

turbulence level. Also adequate penetration and good mixing are

required in the dilution zone for achieving a satisfactory

temperature profile at turbine inlet. This requirement is also

obtained by suitable shape and number of dilution jets in the

secondary zone.

Jet structure

The flow pattern of gaseous jot can be divided into two main

regions Figure A5.1. The region near the jot exit called as the

potential core (or cone), where jet parameters of temperature,

velocity and jet fluid concentration are maintained approximately

constant. Outside and downstream of the potential core, a mixing

region exists in which the jot expands and interaction occur

between the jet and the surrounding fluid. Jet parameters of

interest are usually, velocity, temperature and jet fluid concentration,

 these are varying in this mixing region. Downstream jet exit,

the jot draws in its boundary fluid from the surrounding.

So although the velocity downstream decreases, the mass of the

fluid increase with increasing distance downstream, Figure A5.2.

Jet profile

Bidone (Reference 14) made some investigations for the study

of the behaviour of water jets issuing horizontally. He showed

that circular jets retain their circular profile along the jet

axis. For elliptical jets with major axis horizontal, the jet

lose its ellipticity with increasing distance till it becomes

circular and elliptical again further downstream with major axis

vertical. Finally the jet reduce to flat sheet in the vertical

plane, very broad and thin. For equilateral triangle apetturs;,

the jet resolve itself into three sheets disposed symmetrically

round the axis. The planes of sheets being perpendicular to

the sides of"tho aperture. Magnus (Reference 14) investigated

the for= of the jet from rectangular and square aperturee; -

Plates 12-14 show the jet profile for the smallest circular

holes No Cross flow, V1 = 80 130 ft. /sec. The two

with and

regions of the jet are clear and also with the effect of approach

velocity on jet deflection angle 4) Plate 17 for the

. -15-.
largest circular holes at same velocities, the effect of area

ratio on jet deflection is also clear. Plate i. 8. -_19 for the

small oval clot, both regions are still with a clear profile

transformation compared with circular holes. Plate 20 for largest

 86.

oval slot with aspect ratio m 2.0. Plate 21, for oval slot

with aspect ratio = 8.0. Plate 22 for gall rectangular slot,

also clear profile transformation can be observed. Plate 23

for rectangular slot with aspect ratio = 8.0.

Comparison of different profiles

Results of these profile show that smaller holes have high

turbulence level than larger holes. This is advantageous with

using small holes for good nixing and less engine smoke, but the

main disadvantage with using small holes is the high pressure

drop accompanying thew. Also slots with high aspect ratios have

adequate penetration due to the stronger produced jet. Plate 24

shows a comparison of four configurations at same flow conditions.

The smallest slot (01), the jet is weak with less penetration depth.

For the other equal area slots, circular an oval slots (C2,02)

to have more adequate penetration depth compared with rectann^ular

slots (R2).
 87.

TABLES
 0u1 N0ý
". tný """"
Kl 1'ý- ýD NM ýt)
Co r- %, ra ttl -: t M
Mý
" v- K1 tC1 N K1 Oý u1 O\ iOý- fý
AN 66 0 NN
Oý a0 ti C)% ýýý K1
CO CO CO CO CO ti """"""
Co ti
Or K1 n0 l0 ý"". NW%, 0 O
Q1 ON l0- t0 tl\ u1 M
0 CYN m ON Co %D - r
N 1'-- ""
týý""
--t Zt Je _Z Mr tt1 N
""""rr "
C%it- NNN
ut CO -It N CO CO
Co ýNimM --t r"
i ni
%D 00 %0 N.
Nr P- '- K1 co cV
to
X4
CN
c:
.ý
-: P- CC)
OO
N
tý1 ON rý

t1'1 r
Ull

%O MAýýý"ýr
%D
a tö
" r-
0OM
.,.,
.. ý "
'd CY% ti
p t3l tT en NNr tý
p. ON ai eä 1ß N .ýOO
Oý ON Co
b" tc\ Co
-t O"
ý 0N .ý ýýr N\ "" " 1ß
;gÖ t NN,
--:
Ný
ÖÖ Ni Co
"ý
. -pi 4ý-t N
r- _:
t týIl OO tc"%
43 tu %D ýON""ýýa tL1 ý +
,ý..
Co :3
ý 'e
4» 4) ýý0N
m t^a ttý T1
ºý 1r- r týD O W% N
p" dOO t- N1 ""ý+M O\
CO N tt1 r tD Nk
-*t

tä ý 01 q of 0 t[1 w
*NdOO lý Oa Oa Nr
tn pOOON""A N t0[1
.. ý +f 00 t CO N tt1 r tf\ N\ N K) CO N""a

ýW 9NOO000 ^ý N 0
%D
OW e-! OOOOOOOO000Or
p tt) tti N0 e-" .ý VJ tD ýN ýa000 ANN tf\ "
Hv0ý4 ao 000 +- W% t
--: a" ttl
4""0"6öN N- 644ý
t*
t-- N ý-w 00001
O t11 .ý k
OO"""" ,"a fý P- 1` h" P- P- ö
N- LEN NrrN0 ý+""""""x F

ý# ý
000000+ OO pO Op pO
*W tc1 O- 000O r- a) m u1 000O O- 11
O r- N t0 a0 e-! 0 N- N .ý ta Co
.ý
"""""Mm"".. fQ
4 ýý N
-- ----..
r ý
-I
p
Qf
H
H
 tn f- ýN
P0000 M tt1 tf1 VIN tt1 tt1 © CC 1D
"-G-

ro ý
"
ti
""""
Nýý
""""" "
ý
"""1""
(ý.
r,
tf1 "- tp lC1
co Co cc ti f- %D Co Co Co Co i-- r-

tt\ LM OO Cr O Co OUOO I-
CO VN OMN .- Co W0 K1 f"- cv 0,% Co ý C) CY N-+
Ch ON Co m1 Co M O1 Cfý OA ýO (714 ON tT Q, G"vi Co
"""""" I"""""
"ý"ýý"
tLl P- N
tn MM CO co " 1!1 M D N.
%, UN " %f) kn
ý ti M0N f` tt1 "'" ýO N 1ý "" tý U\
UN ý 10 .7 O tf1 t11 (`
,7ý "" ""
N %D V- U1 M C1i ttý ý- M r" tt1 Kl T- N kD r- co tf1
tt1 CO 9- iNO
t11 Mý ý " tf1 !tý Co u1 t11 N vD
ýr co "MM .- CO Cý OC) lD N U4ý
CO tV ap NN6"6
4v- rM
t+i CO N s- 6 CV f` .-e CV U1 CO t N
0 -:
to N
t+ U-%CO
A NNý. u P'" _*Ln o ý. iii. ß VVN n U'% t.O
ý0 N ýt1 N"0 7 Co Ln
v ON "" MN., -ý .ý"" ýp O .7 1ý
r ýO ttý r" tr1 ol cm lD N
, r., _"
tt1 tn N
ýý
W0 N r, :g N- ý , ýºý ýi ýi
ýp tý11
ý
am ýº
o ý+ Ný"
v. tl1
,-""
"9 rn N
IM
ý
.ýN -,0 "-
0
--t
O
- .7Nf
ii 0 O Je M o
of-4 N ýýýO9N
ýD ý oý
tre _Oý co 01r
ýý "" (V ul Kl N r N
, rý.; -: .- ý'o tMD .7 .ý
Co N '" 11
a
C% OO
ýtt1
Co ý Lno.
9 01
up
NM"""" rn v,
/ýý
""" - ý
N
K1 ý ?+
ýº trMýM r- r- N fý 1i1 MN
"- tr1 Kl
v- -4
U1
!t
14 cm C1
r4
c13 cJ
rn o Co t% O" 00
V 0 N 0o N CC) " C% ýNN0" GO "
co ONN Co tt% t- N-a
t 0-4 . r1 Nr MCo Nýý "r' .ýNr
.
M CO tr1
43 . r+ O
.ý 0
00 0 00 ° OO tt1 O tt1
000 ýo
0 W Lc1 qN"""" f`-
u tý NY
0 CO e- ""
ýr
ýO Mrri T tp....ýN .. T'
4N .7N
HV
ý Z 0 'kO %O t0 t.n l0
.- %O
O

:gý44 UN lC1 UN Lt1 t[1 S

mi:! NNNNNN
,{
43
U1
9- r i! t- -w- q Co 94 N4NN NN
f+ A ö
'20 ý ý
00000 ý0ýö0öö
0ö080g
ÖNýý vtn00000 0 0 p
c0 p n'. ý co ý 0""0.0V- N %O Co
--t
""ý""" i:.; oa N

d
r1 ri
A
ý Co to
co
H E4
H
 a

CID 1ý- %D n
-2 -1

(, - t11 NM %I0 M
Oh N f- Co cu .7 X,
r-
0e UN n t'o'1 4,4 N
606060
.1

le I11 O
LM N '', ý CT N h. A
ý CO ,y"" _. fý t`
0 '8' ýýý Co N """" .rN NN
ý'
tio
sa
L'1 v- CO lf1 Lii O ni N Co
U ONN V- ON kD K1 ei rd) w
0 V to 0 kD tä1 t11 cn +N"" f- N
" M CD N"""
rc3 "
4-4 t- f- OM >b t+1 ti CO 4ý
M "+ co %O
00
. -4 bC N i-
i0 t11 1- ""
N* 9
CO "ýr0
0 N, 0r CO it1 N l0 1l1 466
[ßd"' e-
0ý K
. .ý,.
ý . -.
ý
c0 T co O %o O
t[1 O
N M º+1 co ko kD --:t ý.D
KN co MN 6 r- Öý
ý[! Co tt1 r N>8
X, co ~ýNt
eM-
f.. .-*00 f"
ý"+ý O t11 --:

co o
qý. a
O tr1 M ON MM ý ON tt1
Oi ÖO t-i co 04 NÖ 0i NÖÖý
ýý + Oý Kl
rr". """ -ý"""
r .-rrr
.ý aaý" N
rnr-
Nr r- r`oOO
ýD NO
NOO UN
ei co N ul OOOMN

; ýcl OOr
ýý""""
.r NNM CO NýNaO

f4
:Pw
= .ýý
ýPlj
4" e-i
Lnooooo
N0
O
a
CO CO CO CO CO CO
ý'- rr !" tt-
10
r-
.
m ib -03 .©000" ý 'h ... a". a
ý e- r e- rrU
od ýr
ýý
W- ,.
161 -R F-
O-N ON
qpl
Wý0 ýý ýN 0000 00 \Q X
ýNr 43
as NNN tiý". 1
oý tn N t- ý-
"
lv a.. r."v
G7

i
O
v #p oo$ooo
Ö0Nýýý
ý *ao öÖNÖÖÖý
u
p"" "" " "< " ºQ `q -: r %D CD
a6aa64m

1D
O 0
H
ý ý
ý
H EH
 ti
01 Co M Co tt1 tt1 p(1 tt1 ýW ýJ
ý """"i"
"ý ýO Mý Oý ý ti ýý tf' t-
Co CO !ý
t0
tD t11 Co Co
ý
cc Cr .7v
CO Co ýn cv
P- I, -
ý
d
@d tt1 fý- N Co N VN tt1 NOO 1%- t1\ N rn r1 O U\
ýM NNO f- CIJ (%J V- co -,. ý
ÄVýý U-1 l0 ýD ý tt\ t11 .ý
ý `U 47 1t1 t[
""""""
A4 00.40 ". """"""

02 N tt1 Co N (Y% 1ýo Co +n co t

tl1 ,ýn tn "6"Qý
.:
"" CO
.7
CO "ýO co
AC ýD N ý- """ :n1,0 (\t ,-""
N ,0 e- j 1,0
-: a- . -j -1

Co %M ý 11
N ýn .nOO " t- f- N1 CO ý M Oý . CN
ýczi ý: ýý°nA ö CO "" CV la (- N N"i .
Q) N ". " CD CV Co i". " tt1 r K'1 q N N
ýN "_plt
G: M CJ Nr". CV t` r U1 (V e-
t+ ýN ýý- tc1
>b O
t[1 %. co O
ýN ý Co 0.0. ö 8 CO i
t) " ON Co ý- CO N -t Co Kl
47, tD "". Co P" .. to ýO 00 cý O ob
erl
+d
"NýO Kl W ý- " A Co
Co tn
M N fý K'1 N 4r- r QJ N to
Crrfý tt1 "- N tt1 Co rn Ul%
ýý ONO In TT P- N lA N C'j
OO %D o" O*-
VN- m1 C1 Nr" Nº M e- NýÄ ti ý
ty
.1rN
00
oý^
ýý
i*. sf'JO Ný NO o UN
ci, ko
Ci N CO N ýi
tJT itl CT o 0 W
aqj
ti)
co M cý
0 O
o t" a N r-g Q " ,to tSl
43
"
Orrý %. CO tf1 n cm --:r Nr0tý
i
.4" .ý""
43 C% -9 CY%
p0ra CO OO O% O Co t11 O ýNNO ttl Op
,)nN N CO UN CO N CO ". 6, "
Nýr ýN CO N t- 9- .: t c4 .- tt1 r- M Co MN ?i
w4'bý
v
Ö vw 0
Oý OOOýO
ýOOO N 00
". ". tn"6 0 r-tn ýO "ý"ý"
rl +i CO .7N f- ". ".
9- t0 N% e- r- ".
e- Co je N t- " '+
.NO0 0 0
" wo %JO ý
0
t4 44
0 44 tf1 tt1 t11 t1N ttl tt1 0 co co co co co co 0
NNNNNN r 9- rrr r-
q, 000000
"". ""f" "". """"
N vcs
ý
4"N""
NN
""
NN n K1 M en nnn
ý ýýýý ýý N
O l0 l0 l0 to t0 ý
°äýcö .i U1 tl1 tt1 til tl1 tl1
43 " ._""". " AJ "". aý0 t
ýý.rrrT .. 'q- -. -
S WNNNNNN co NNNNN
ia
N

qd
000000 !)000000 Ki
v0 00000 p
0 u1 00000 C) tt1 OOOOO a)041
tn 7;o0000
QN.? to CO Oýý-: t %.
0 CO N"ý" %D" CO" ý
f"ý"". " 4+''+ Cß
.. .

0
ý .-
ý Q?
ý
co .0
N ý
EH
 tl1 Mm0 1- 0 tf1 NU0 tf% UN u1 t- 0
ýp t11 .ý Qý ý 1ý K1N Qlý lD UN Vr e ýD f_
a-
Co Co Co f` Co CO ti Co Co Co
06006"0"f466"60

ý NMr N
e- NN0 U'% t1J ONM tt1 "-
co co CO ý co , -^ P- Pý- 1- I- ýý ti NN

N ct1 Co Co nN CD
u-, ýý""" C? Lr1 ý""
N lD tf1 cV "" ýn ,^l. ý """ r- tt'ý
D
%. --I
1- f- N1 C' ^! M MUi CO
kD kD r r- r-

titi`i° ö fý t! 1 in N "l O CO r- %O O
Cý! Co rý e- U'1 tS1 VN tt*% co
C) ýýý ýn ti .::
ý U1 ýIn ,D 91U
.ý """"" tt1 tt1 tf1 to
00006

N S.
.c
r, %n ru m tf1 N f` r CO ß tt1 NOOO
A Qrn 2; Ö Ü1 N t11 f- %p t11 Qa 9
-; n ý' ý tý
ä trl
. Zt
#
--: .

yJ N
%D l0 U'% r4,% i! 1 CM M CO c0 c0 K1 (ý r-
Cl
O 1`
%.
Lin
O n;
%. O% M
e-
t[1 Ný Nl \D e+ Q1 (`+ N Co
ý 9- NM-: t
.:
t r (V Kl '"i
,..
ý' N Ntd... -+ . Zt

.. 4 NM1! 1 t11 tl1 U1 00o cn N t11 OOO

CO t11 Co MN %D U1 ..Y N ýp nMN0
"
-0ýsQ%+''
ÖÖÖvÖ

""ýýý
00000
600""
r.. tý- P- ý-
ý000
""a
L-
0

"ýV
d CIN
0 Ln 00000
43 0 lA Cä OOOO tn MMMM
(V 4) .- ýO N1 Co CO- co Co co Co ý.. r .-r .-
.ý
ý
zJ ýGL O %O to u1 Co t! 1 lf1 M K1 Irl ON "- 01 Co
-0 ý Co Co Co Co %D kO D
ti lQ %. 411 UN
ýÜ
ÖÖÖÖÖ pÖÖOÖ
ci 0'1 O0000
49 H ý""""
vý v
N' ý ¢ý
p ýý
.. ý
W4 v r- co% qLsn in tt1
a)
4 UN to N O1 0
O! 000
CO N ti1 ON
ý v q-a N0 ici NM l0 N t11 K1 t- e- NNN
t! aNN Ln rn %D `
t-
i-3
vp 4{
0 440 UI
43
O 000
iý
1
WOOOOO
aö T C% ID 111 tn UN o C) 000
m° cs AI rti IA
. .V.-ft \A
. L, O tt1 tiý tt1 tf1 tt1 U1 LCl LCN ul tcý
A ._..04
r N 04 M CO e- e- a- r 'e- NN NNN
11 q

ý0 ýN NN
ýa
o cm QA 0r% tt1
r- t 4. :t tD O1
ýý<, rýý~" r- r -.: ýýf .00
r, N Kl Kl . '- e- e- r
0
" e- äi
0
aýcs ý0 ý tt1 . ý1
w\oo0o0 P
CV y jN
oý
co
0
"
00000 .1
O1 Oh "" "Or
Mý r- ýr- r t- r -4 TTrM .
ri
.ý
h- 40 r e-
W v
"o lt1 Q*% NON al Lll t`
ýý .43 LCI% -57% O tn -A N
vý a W- M t-
OOOOV!
ON to
P4 000O
A4 . r. t.0 CO
p, p000
C) rP4 4 "" ý""""" 640r
;22 t! ' ý` A66r V'r
V-
. 40 .-
q-
ýt
d'
P-1.
 OM f- tm O MN (ý- K'1 itl K! t! l, ttl tC1 O
(p tilt tell ICA O i--
%. N t co
f- -.:
ti t--
"""""

tl% "- n1 M, e- t` Q1 01 lCJ

h- -1
ýb9 CoCD
CO
1.0 f, - º- t` le D
%. lD %D r-

N M
+Yl CO """ -ý UN o (T tn
xao M .7 K) 14
:n f- C%j
tt1
r "-- rT
-1
s--
µ
wN
CO .-N IN-
v- r1.... t- cU -. ý
ý
.- ,-

CO NQ ep N1 C1 ti co NN ON
U%A-: t CO co CO WN r- ti ýU r O, týl`l11
i1 ýn ý7 ýA ýO ýýý ýD
"""""

ýIi
0 1- K1 0 O TN tin 0 00
co UN o0 e- Co
4 .-
f-
P" n trN
Ir, %0
1`rn
fý
rn
O
tf1
-:
ýD
j
Oý (ý
NOO
Co CO ý- ýº-
., tn %D
-ý
aý %D 0 0O t,*-
41Y
ra ý'
I CT1
0 ý
. -- (.
%D (V Ch
v-
Lf% M
o
N rn o an
r- 1- 0
"ý N e- NN.. Ki I N1
K%D tC1
_"` -zt

O a
k. ý tr%n O N1 r- 00
ýN .-I.
.i.ý4.
CO N t- r NN ý-
ý ti ti N -..
ti
o000 OO 0 00000 OOoO0
0 000 ...
416 .0
ý
ým
ýC)
0

r4
ý,
ýJ
IN N
P-
ttl
ý
%O
P- Co
%O
n 0000© 00000
MMM Nl Kl
.do y'
. e" CV NNM Co Co q Co Co ý- rrr
All
co CO O t! > tI1 O CO %D r- lt1
c» Iy o Co oo t`
1t1 .7M UN %D
Ci F+ NNNNN
Co OOOOO OO00p ööööö
s r-1
.#.
r4 0 ý"i"ý

C3 V
. ri % n v
c) ++ M
O in O tt'º 00 co t+'1 N 0 Ö
. rd V *- Co a)wN M

ýi,
WN T- N tt1 \N0ýý ti %O

ö
v
ýü
C: o
a
ý _ M
.ý ul tt1 t0
4-V
44
OOOOO
. ý'

000000
n %D

4) 0ö ýýýý
T
O
ý
tt1 O_ r-
tc% ýý
t11 ONNNN
Co tt1 tll Lf1 tin
cV
LC1
NNNNN
NNNNN
ý UN eý 4- e- e- r
ý ý a It 11
ýý pa lD ý
19 H ti ti NNN NM ý co O Co
r4 Nr [- M \D ý K1 NM
in x ýfTTTT
ý-1 "ý"ww
.ý
ý CYN
00 CO Co CO t) rrNMN o. " N
A 'ý r: Co 0 0 ,+V.. (V
J 0 r-1 ý
NN tfIN IK1 4) N CC) N
w "\ 00000 tý """"" _' --t""""
co N C)
O0000 v- .jr NN NNN
C9 v .ýM
O c3 q
výN O co N0NN O0 CO ^+ CV tn
ý ºOC1tOD %D O 94 ONO UN ta N l[1 ýD O
00 0O N- AOpOr0 Pý OOO
O p4 """""
.- .-
ý..
"""""
pý """""
ýT
fý pý rr r- ýr- r . e; !ýrrrr f" T` f" !"

"i
HI
 NO t\: O f-- OCo r^ N tt1 OOO
fý O fý tt1 ONA r.
CD Co r. CO t11 tT o'ý
VN
L[1
Co Co r- f- f`
""""" "" -" """""""" Co

0
OO ý- N K+ ti u
CO Co Co Co Co tt1 Co o r-N
. oti ý
O¼D.

ýý""" M º+'1 ti\ tf1 Co O tCl UN

co "" t t- t[l co "
OO Co f- O` """ f` 1ý -: """M
t
-. MMM ý- ý[1 týrN CV Kl U'% Q1 e-
.2 ,

troo tr. C', a .-O cV ý LC% LM

ýýýý ý.D .- Cw K> ý I+1 aý- ý
0 tM0 14.1 tt1 t.C) ý tt1 u1 4n
.,
"i""" ""i+" 6 d> 060

tt1 0O U i3 O
0000
riý
ý=n O kl Co tn N.
.a t11 OO
aN e- rrN
10 -
Ln O ýD ýO
M .-C, ) tt1 O
tf\ QM .-r 04
oV-
ýrr 0 co
aýNN

1
M
ýM N LCN Qº MMN K1
-.
LH

CV MM O tr1 f- N (n
+ý
:1ýrN ON K1 NNM
TT N Zt .2 Je
ýýýýMý
- -- -- -"
ýib
4
t+l co co OO OO UN CO N r. u1 N u1
/- S l0 1t1 O Kl (ý tý lC1 rN N
V ÖýÖÖÖ ÖÖÖ
Or. n
00000 OOO cp O
0"""0 00"60
6 """a0
4.1
4) ý 03 K1
'" w,
ýM tD MM
M MR
G1
y M;
c% """"
W.,
N
tý u~1 fl- 00000
co co co co co r
tý
Ai
tý
rýr
ý_s
R tt1 M u1 MN ~Nýý
Coä :T LCl UN OO t- N O
%. ti UN
43 r4
vv
lý
;s..
01 WN
ÖÖÖÖÖ
Ul f- ÖÖ
ý"""
ÖÖÖ
00000
""""
ÖÖO

O r+ U a 00 41 00

G1 ; v
. +0 ý"
4)0
O rH 'ý
0 =iýti
rLnrno. ;20
, r.Ln
..
ý
7 .-, v, v 5y
a
%Z
43
ýpf
ýri ýOr
u1 cv 1` cu
tt1 'wo G1 r
tl

43
4-t
ONON
t! N
ýa
N
CO r
V-
K1

e-
v W
R? tý 00OOO
i=
VN N,
tn o 000000
tt1 iý in t- o 0
o.
mN
04
F. 0
r4
ýý l,CIA
p 0
q-
t
--.
T
co AA
rrrrr
N'1 OOOOO
V- tC1 t11 Ul l.1 u1
Qw
00 ä U 0NNNNN
Ln N 0f) O
v r4 010 e l0 l0 t' r+ ýM ý
oo W ýOOOOO f3 :1. týf'º 43 ?(N C%
.W4 0 .i
--t
a. .ý
"
A 'd ý. 7.7 ýA Ö" sº rNM 6 (10
ýr s- t-

0 ý Qý tP1 tl1 (`-

° ýooooo aýý"""
.i
aý oo cn o0 T ýD ýpý t11 N Co Ný
+' M Kl %.O Co
rr t- e- r .9
t) rJ .7 ý1
p
c20 c3 ci
t`! NO !ý- O ON tf1 0 NCo QNO tt1
y.," N
ýp arNMQ! C3 O i" r äöööö ON Pý O Oý
ýM
Or
""""a P+ ".. ". Pt ""ý""
ý ý,,, "t- r e- rr s7; ý- rr eý e- e- r a- rr
,pa
ýz
r

r4
r1
ßl
0
 a P M h
v1 o p O O
ý a
tl° ý ý 9% r cr
s.
to Lo
r, r " . "I - . - - - ,i
i

ýo b ý- N Q' cýo
r ý,,
-
r-
i I

ý' I
cý', r ý " r, -

la
m
h N 1ln V) lrý
rJ
"

Z-3-- r e ° ° a a Q
S
_ s N \,I
Q f" '° M M
46 9 c-1
W
Ä M
N (>0 ^
N - o 10 ol 00 o .o. o 'a) cl o 17", rA
N
z N
ý° a N
N N
G
cJ M M

\r
r-, Öf-, N o
Q
r
r-
cj .i N G 9
p tir
Ö
O Ö
a%+
a ý" ý
O
0
0
.
,,
- .
. .

7 7

J 0 O o M M
s o M cf1 tý
M to 'co Q) 03 '30 - -
7 - - ,
ý
N 94 fIý N
M tý Ul M ý
i fr M r
pl ü P- 'ý ý ý r r r`
C1 ý- ý-
ao r-
o
N
° º- r o ro 10 r
ý^+ . . . . ..
> "
y: S IS
ý ýsC ý y
y rý N1
Q p J [ý1 J, o ý 0 V1 oA
ý F+` b ° s (
] >ý M `^ r4' M
.,.
ýw ý r r- ý ,
cJ ý c1 c4
_f
ý, O 7 ý
L1'.. ýz 0 o
M S $ 0 0
M ý 0 CJ CJ OO
a Vý rn
ýA N

r I
c!o ro e `ý °
O' co V' ýv
ý e ý ý v ý ýo
Z- N 9o
lý y -Z,

Hý?
N
wäy ' `T
M
S 1 f
-ýä
ö ö ä r' iN
e ý `O c o
°-^' +n ý ý^ L- rý1 y r^ ý , r'
W0 ý ö ý ý g cd
o o 0 0
o a a _ ý ý o ý v

7
ý- ýr"
-
r
ý
s
ý
 (ý tt1 NO u'l CJ C) Ln (-) U liN N tI1 t11 N u1
CO Cý (M --t ý 1- 9- Ol e- e-
CO Co ti ti fý Co a` Co 0. t,-- 00 N Co Co
00464
0"0"0

0
e- ýOO ý1 rl r-- vi co w0 OJ f- r- Co
r- tt` ko kl? N" co tf1 ýD IID 1` f-
co co cr CC)

tä1 l! 1 0 CO 4ä1 OK1

r Nl N'N Co
CT"" " cäj "w
Q-) """" (Äi
CO O CO CC? G'ý """ý
t "- tV qr- i"
_

CV f` tfý co O
rn º_1 cr f- c o cc ri M-%
rn =le rn C-
r- iA N't Kl cY2 r
U 9- Co O r- n- Ký, Z il, ý U-, zt tt1 tt1 W
ýýwý ýý 0"000

ýf s0000 UO0 0v000

U"%U V- 9C
co ýý 0 Co e.c.N ac) u1
ýr0 iV
t11 CO N
sC- --:ý
PNt
:; trl if "'ý
C~ý tý"1 CV r
1- CT r CV CV r- r r- C Kl
4- Co t- .-NM
v
. ý+ ILO
Q Co t
4N ý tr\ ts1 ti x3 N ti -: rN
z9 tý ý! CO CV 1ý "r tf1 u1 r 00 ýO 01
r'a (Y Ir A .: 7 ºý 01 rNNM
:ý. TT r e- .7

f- C) ti1 O UN O (ý- ti\ lD K1 l- N tt1

fi %D ON OG NN ýD t! 1 Co N tT
f`ü ti NN ID
r'.. lD t-O ti I©`-
00 O OOOOO p0000
" o:,-ý o """"" "". ".

M t! 1
0 M"o"t 00000 00000
hl MM t(ý M
0f1 co CO CO co co r 42 r
cq Wý tn CO .- e- ý-
N ko nm tt1 (ý" i`ý OO0
C3
C5 S o
Iý- r- Ln o
ýD rM ýý vý u1 M kO tl1 t! 1 in Ln
P-- NNNN
r. s rf

a)a
. r4 94
C*.;- rz
00000 ....
t" ý ti ti
0
00000
ÖÖÖÖÖ
"90""
t). +
. r4 v %N oO N. CYNý
00
w 0Omc 0r ý CC)rn
dÖ I ONýO \ N m0 --:t C's
ra 1` ON N to r- s-
Oa a tn .7 to rn .--
4-2
C)aö &A Cl
ý'z N00
00
Co 0 0
ýAýýýý
00 v
CO 0 0000
v- "rV
ý. U1 tll% t11 tt1
63Id - r- v
r"
IL-4 ..ýi
p1 rN
v
r V- T- IT- N NNNN
cl
Ör-! vo v 1 Ln
K1 t+1 N1 tt1 M t11 Co co o tKl 1` I` Q NM
,4 4ý)Pý'i tiO l9 t1? ý lo "- e-""" N lb CC) N ,- tn
ý e- N"
A'dx .ýr. .
NNN4 "" e- NN "ý e- a-
Ev
t.-
0 ýD M tC1 1- co
1-4 ý't 00000 .af Cý e- a"
4, . tf1 M " jý. N t- r
00o00 T- s- P t"' r N r t0 1` 04 _:
t tt% co
-:
CO
m
öä ý CV CO ca
tn
N U\ läl 410 tOt
O M
r. .- A
cV %. (V ý- K1 ýD
o'00 OÖÖýÖ OOOOT
"""""
0- 0 +- "ý"""
0 9.. 1- 1r e- "r V- t-
ý ZA
ýI
ý
3.
ý
E'
 0 ° N M °o rn oý
ýp Co Co UN %.
0 _:t K, aýo Co
.............
t- 1- !, - w co hº r` tý
".

ýýýä O fý ý- M (ý O O K1
& ýO ýO !ý 1` iý lt1
r-
0
r
h ti
Co ýA %,

tl1 tf1 . -:t t! 1 CO Kl N

cn .ý CO f, -
x o .ý"f .ýý"" tCý t1ý ýA "
ý° r0-0%M ýii ý 10
"""Ný
M 01
cc) r` ti rN e- a- ""

N Co 0M U1 ýöN° ~n tý NO
Vi CM w- CO .7M
CT Co t- O-
U
t ako
ý. . K1 ý.t7 ýD
_: . n .ý tf1

^ýz
.n . cº tn nn O C., ý ti
: ý'+ 1` t11 lp N (n
rn o r- r- r- ýý
41
ý11 N N0
4-3 a- N t+1 i --+ ýM --:t U'% CV ý' tfý
.,. ý
a-----_.,.
"ý " äý 0 Co M ý rn 8 rn ti
NMM ccg0ýý
oýv
t` ýo ý
.ý ýe- . .- .- Kl tr'1 .7 N
,7
. ý+
N ýý 1l- tt% tt1 K1 N CO tD tt1 Nr UN t,f1 W- CJ ON
ýI
V
ÖOOp0 (- N r. (- %O
OÖOOO 00000
ý, ý """"" 04644 0"466

ýJ d MNýý! t11 111 CC) M

43 ý -1 """N
ý r ý0r f'ý1 Pº ýNN 06, 0000
MMMM
ra "
"d co
y tt1 CO rrr

0aS
o° A.Co Co tz r- rý to t! 1 u1 tl1 WN
ti
öööö0 ööööö 1- 1- f- 1-
43
ý r+ o000 O
0 a- ...... ....

.. 0 ý
t> > n
..4 o `ýz O UN O U1
ý
a WN 1- K1 \NnNNM
N &r
t) co aý e- NN f+1 ern-
CO
0W 43
4/ 4.3
v w00000
ap OOOOO
-I x8 O UN O U-% ONNN cm N
ý m O U1 UN l(1 t11 tCl
mýn co cL1 111 Ic1 111 111
:rrýNN KMl la _8 la _8
n
ýýMÖ A tl1 NN
0 r443 CD CO Co CO Co
434-,0 (M +3 OW %D ON T-
"ýw6466 l0 \O t0 ýO ort tt1 CO """
.. ýa)r-d e- rr t- e- 06 v- N
Ki K1 V00r r- 04
'dcn
ý
9 0
ri u1 Co t+1
0
ý lD Nr ti l0
i
ooooo a Co ~M CV
0r CO f N
C? 00000
r -t ý- rr
Co
AM
& 40
t[1 Oý ý- T
CO rrN ý-
m 0 vö
o= b
1+1 Ö ýý O UN t- OU1 UN
öý öö0ö T
T
R0
iý9
P4 OOOr
e- ý0 Kl f- 01 V-
4OOO
P.
za r d. tý
""""" P. .. 66 r
ý
._ ,= ._ ,_ ý- rrr .cg r it- f- e-

%D

ý
 cv º- O u1 tt1 co 0 t11 in tt1 tl. - Lill
LD D
%. f- >n to %D P- O
1` co .ý .ýOOco cc
co
."""""""""""""

0
%D 1- w- N1 tip tt1 N tt1 4-
(ý- Cn Co Co Co Ul kD P- f`ý 1"ý U) %D t,0 f-

tr\ N Ö ý+
Cn ý""" ,1 Co LL'~1N
ýb N. N. ""ýý w"" Ký
e- %- .0- V- r CV Co T- N- M º^ j' r
.
CO
NNM CO O N ti U1 N ! +1 ýº c11 t- K1
UN O CC) N 1+1 "- M l11 cX] C,)
UN %O \o %D Zi' t11 it1
0""00 Je40600
ý"""

"ý If'
ýz iM t- N O 1,O
cm /-'ý ° rn rn rn OOOo
u1
i '
ON tt` 0 o°c
s`. tM
ý' tl1 tiD tý ý- U\ \O CO e-
.
a
ýN
""I ';1ý O
tf1
C7 f,- r--.
rn A ~- Ö co u\ rNM
%f? (ý.
?'ý..
o ;A 111 V- l0
NM :7 rN Pq Mý NN K1 zt
.,
ýýý o 0000 CO u1 ul
,, ý"- ýO tt1 ý M kp ºý O
: -I ü co ýw. .. .ý
ý-NNti d0OOO O Cy OO
cý, O 0QG0 400f0 """.
....

üý. i NN400 tC1 Co IM

43 91
n
0000
ý T U-s N"i 00000 MMMM
ra " ý" %- NN co Co co co Co rrrr
to
10,14
CV CO If) cv (V n1 MO N1 1- N1
00
s
1ý1
ý
fitifi
lD l0 kD ý

+am
ri CO
o VVVV
i""""" OOOOO
ißÖýÖ
0 >; """"
0 tj v
.. + tJ
Q)
P.
.. ý ONO
ý`i O fý O tt1 m ttº OM0
W %0 ýrN.. N tn
r, äýMý0 '1' t0Cl lO
Q N N1 t0
VO 43
0004 4,1
ýR tt1 tl1 t[l UN ih 00000
ýd ylxPý. Or'. Oum 0 Ih in 141%KN
0
fr\
mýv° äý nýý Op tt1
11 t-
tfl tCl t11
ei e- +e- *+ s. N
ýý
ýý 43
0
tm- tiN tff't %A tn oet, - aý o u, e a%'n
mNý cV cm ca N Un cv
"4" s" . tD' wa00
cv 43 N: ýO
r4) ýý Co Co Co 6 t- NN ON ö"rrN
Aý
ri
01 ti %D cd' ~ Co Ki
;. o0000 ý" "` """
r"` N Kl O% tt1
tº CO CO "
ý00000
r e" ý- r s- CO NM tK1 ýrrN

ci
O cv m% OM Co 0 O
%.
,-
lC T ig 22 m ö_ ýrN, l0, O
ýÖÖ
"f"
O'T o"
".... o` -C-It,
o' o a- pe OOO ýº-
ý..
""
b'ý
9:4 """"
r TTTT
.- .- cý rrr e-
ti a ý- *-

A
Co
H'
 ti Ln O Co OM Co t+1
LIO m +-
0
Co
N
Co
q-
Co
CO 1-
CO Co
D
Co
e-
Co
ö CD
0""" 4060 "m
,+004 co

e
NNMM t- º+1 Q% K1 ON
-e- Co CO Co Co t` ti co ýý ti ti
-1

+il tf1 t!1 tf1 CD

ýýýýý ý"ý t` W" t O"
.: ""O
"ý .ýO !'.
r CO N Kt

W% N O Co ýÖ
U
r,
i1 0 fý -" ý ý,' r,.
tgtl M tf1 tn q0 Nn u'1 l! 1
"ý"" 6660

Qgn tt1 OO 00Q

I;t& 0 O (` j"- ao
irl

p
Ln O t- .7Nco V- N C%
ä .-
N tt1 V- T- N qr- r- N
CYN
-
Lv
WI N tC1
'-a N tt1 co fl-- Co ýN &J
Q~ OO N O tr1 O tD
O' N N 7 t N
y .; ,: -::
ýIý ý OO u1 O 0o, ON t11
N rOO
N 04 %O IA NN ýO tt1 ý" O
or. 'llý ýÖÖ ti ÖÖÖ ÖOOO
I"" O
4060 ý"""
43
ý yIJ
ýv N
Oh N
ra ""
ý CY%NN OOOO ca o CYNo
r N tt1 NM
ý tV ttl lp CO CO ca cc) f" f` r- t"

L
0 o UN 0 N lt1
Co o NN tt1 tf1
s ti ýö0 #0 I'D
titi rz u1
91
4)
u N- 11
- ti ÖÖÖÖ
ar'ý- 0 00bö
0 . 0""* ""ý"
v
ýü
O tt1 O OD 0
=hr N WNON (TI tD N ti O \Nd tý
P- ON c0 MON .ý
" r- li% Co z9
ON
Uli 10 «ä im r- e- ±a t- r f-
Co
e-1 O
4a
nxm o tn O
tt1
U1 ON ONo 0000
t+l nMM
0 P000
Kl
43 1- ON u1 u1 u\ u1 r ýNNý
aN t11 rr
tt
00 cm
Co
ý
M ?, o T 7. an
. rýta
" po on Oýi t~p CO rvV .t%ý- n
ýöö ý """
v"r
a+ ý" NM ý-, N
. O°
ýPýM
Z,--Z NMN t^ N
C) 000 N
""""
r+ ýM " lD M
" I
2ý. 0000 ,v- .:
t l0 0D
V
W
* tt1

ýpNý Co Ný o" cV L11 UNI c0

R. 0 c0
p©O OOO ý- WO cm Or
"""" tߺ """"
r"ý"" r e- .- ý
H
FL rýrr .- ý- ý- s- t-
c0
ý
ý
ý
ý
H
 ONN Co tM 0 tt1 ml cu OO u1
f- CV N1 00 CO Cli tT 0 tt1 tt1 0
CO Co Co Co " Co Co tý " (n Co CO
""""Tý""T"""

0
ýOO K1 ºý'1 I,-- r ti Cr
CO CO cz Co : tD I- i- Mýtifi
N M fý ýp
Co f- " M t+1 .ý"
ON rn V,.
m *..rn ýD "ý ! t1 ""fN
q"' T
"ý ý- "- rN (` r

NN Cf1 ! fl r+1 1`- M ttl tf1 ON CO

0 QN rn O fý t11 rn o M f- t- O
vwý UIN M tfý tt1 ýO UN %O
ý"6"
.... """ý

ýý
ý: v_- 000 rn OO
tc-A :: r- 0000
. -Z O %, 0 N Co
ý tt1 Nr
rrN rf'1 1ý- NN ý- t- N K1

ý! Iv
ý CIN - O N
co Öý ON I` te1
'
`i
.1
u\ A 0 fM t11 O tf1 ý
4- r rt r ýM º+1 NM .ý
U1 1ý- N
(0-
u0N0 0 t11 N CO t! 1
ý ý{O., N
1I47
-'ü fý,.
w. ".,
ýý
0000
t-
0öbb
ýýh (`ý
OOOO
fM

..,. " 6064

ýv i- Pº M
43 ý+i N a00 t OT 000
v ýý _: rn o'% 00
rs 0
r
N t(1 t+'1 f4l
ý
1swOý
-ý t- (- Co Co r 'r T. r
ý41 L ýýýN N (V %D 0 tf1 ON ttl%
l0 O
%. l0
ý
cs.ýJ, i
42
10
ý _1 `' ÖÖÖÖ öööö 1ý N-
OOOO
{ý- f`"
ck-ý r
d r-1 ... ... ý """"
Ö ?ý C)
ti N N CJ
. r4 O rI1 e_
O t[1 OO G'% o tn
t7 co 1-
.1O N CC) ON t^ ý lp 0
0
Ln n
aö t ko 'R- N 1- -+ t- CV N .eo
TN
v .:
0 a10 ýd
+2
ä-1
'-I' OO 0000 Oý 00
4) Ö ýº O UN NN rr t- 00ö 1ýf1
Co f- .TO tA Ln LrN tý
tO. r4 O
%O -r- N rrrr
Zt tý NN LrNN (V
.7
ý
43, 0
0 co 41- ý ý..
q- 143. N
Uf-f %O t0 %D co N NM týt1ýÖ
0 lý4 ý 43
NNNN ý R. 4k " :. ý
0. r N N . rl f" ý- N
A "d 0
it 0
o ý_ tl1 N tL1 t11 i Ui co
.ýö 0 000 ýj Ab 4Hi
w 000 Vp 0 aq rn rn
ýýý -:
t CO CO V* N Ln ti
co ý
0
to NN t11 NO t11 O tSi NN 1` ttl
ý ÜOýÖ OO r+ tl1 T
cý'i ý0 OO .- ii OOOT
...
6 lk 0.. K 0- p+ "sý"
^T ý ýr
N- rrT R+ !-
.=
.1ýLa -1ý

o'
r
b
ý
ý
H
 (V CU tt1 O %p I,*-
r 1` oO
-:
I N ü co O +- N O
:.?
co
f-
kD
N
V'1
+11 N- 1-
ko t- N ," --1
+º 660 " ýn w 1`
..... rýý"

D
i` t- NoM
$' t- Q? tü
tf1 tfN ko N
1p 160
ý
N
ti
M
ý

%D
oUN 00 *- t: % nO 00
o P- C% CM CO K1 Z- Co N k1 ct
b4 ."ý.. """"" " """
0 ý ti ý CCINO t r- N V1 CD
r- ý ti

1` N VII N tt1 O 1`- t! 1 oo

I
ttN r
a co c"cN mo M CYI%ýý o tý rn rn
t11 V1 l11 tt1 W MM L11 LA
040&" ý"""
". """
Npýý
//ý
.. ^^ 0000 00000 OOOO
y' N tf1 W0 C13V-~-:: ti
;; rn N tV O 0 cv -+ - ýM
Q% T- N M-JA
a .- r- .- cIJ s~-

yJ H
tt1
" UN 1- N M r- MN
y, lM0 ýý ýo MONO .ý
1- l11 %O Co U1
wr lV N C71 "- r- t- N K'1
tct rNN
-s CO l0 O QO CO co O©
ýI la ýN UN tt1 Ul%00
ýÖ to ILO %0 %I0 CC)
0 Ö ÖÖ
p.. 7 OOO ö N. N. N. ö Cý
0" Äb "" *... .0 s"""
yJ
ývý Hý
43
C) 1` 0000
N1 .7 co
ýrs -0
>ý Kl (ýº ON r 00 00 r
-:
C1
MMMM
ý'
a Co Co Co Co t- rr
od
t! co ý. º
O 4-a u.
CO co tn o h- ýý UN O0 Co Co tt% CY
0 te ti Co ti ti
4J H vO öö00 ÖÖ ÖÖÖ
C3 °r' OOOO
ý
47 ..:. . 0
f"""
.r1 r-1
tI C3 ^'f rJ
-r4 V, UN ix ! 11 N 000 N -zu, %O
tn O
N
M 00 lD
ý ti t* r+ tn N CqCY>. N cc) N

00
C+ 0
¢i A uýi 43
{.s
w
{- rN
v 4a 0 A-4
v ooe oa
00 u1 Nýýý ý ö NN
tOrt0 44 'ýý ttZ tý-
u1°iuýiý u1 r ix ig
0 VýM -+ NN
nýr
ý Vý
td tD tD ýtC1 OO l0 t%- NN
0ý toto .7
0ri
t0 tD to tD ý 43 N M N UN
-l 640 S" -V-4
T' T 4+
q- r tý ý"ýt. r Aa "'ý
Amý O q- .-
ýd o ý-1 H
C3 la
rl
4ý V `: ': f:
"ý o 0000 0
0000 Co %m ti N
C2
C3
0
ö°ý V, rr e- ,QrM
t)
ÖN tJ
.2
CO e"
ý
M
NNN ý

t'a N tl1 Co NO UN ul O 0 ºý'1

s. O MN O
v rO OOO
e-
ý-
ä0öööö O QÖÖ
QI
ýO " LL ".... """
0 ý r" -f
T`
,"
TT
.-"
1- Qr +ý"" e" rr '.4 rrrr
ý
p.4
 ti-A O tt1 l1'1
tf1 tf1 LM
kD Q'ý tn O r-enLr%
Nl O tr1 ý1
ý t- ti
Co p- C*\ P- ON co r- r-
4000 """º 6"0"

0
2ý" a, 0 (NJ U O i- lf1 t11
. t0%n `D
--+- f--. G:'
CO Co I,-- c'", r, Kl tfll NN
,

IL" tfl N'1 M U1 tr1 -,"s.

ý N. N'1
4Q
s
C In " ti a"- kC """NtiCa
ý e- N P: ýe- N Co
p+1

00 UN UIN VN o cs e- K1 CO CO
v°o kn u; es1 C% UN W
wvýý ý cV tt1 ýO ý
ý"""

00
tr, ON Ný/ý ý000 Ný 000
ý= o 43
ý CO
ON R-
"
N ai ý r-
Iý
q-
c0
9- 1ý
co
t- N

J
hý
CL'

ra ý ö Q~i 1Co
-
ON
.-Nm
Co O
r4
C+! rNM
1!1 N Cn
` Co
ý" N :I
N N1 rN
%ý
y,
ýI co O t! 1 O0 t- 0
t! 1 NO
Vý
ÖÖÖ r- ti ý ko Ln0rn
4'Yn OOOO
0
%. N. t- r- k7
o000
ý ý-ý hN
a 000

I`ý MN
.... .. 0.

C) """ý 0000
ý 0O00 M rt1 M K1
0 CO CO Co Co e- .- r- 1-
1ü
d +ý
O --.
Co
Co 00 00 00 00 0
w ti n UN
.N r1 fXý ööö
L2q 0
00 00
O .0 00 00
e-#
0a0 .. C
00

v I> v 0
to _, x tc% o ä rn 0
W O%O
rm co 040 *Sý 0 trN oM
N
ý US CO
oT %D.ýT \ e- .r- r- .- ti N
ýO 64 N ýTT N %1 .- 9- ±a' N tit V-
00 w t-
43
0wb W
ýIýý t11 ti1 t! 1 UN O
c)o ýýý O (- t`- P- 0000
mOU VN ti1 u1
g4 -A
ä tt1 O
N t0
+- .ý Cp tt1
r e-
ýýýý
eý a-
av43 43 p 11
0
U t'J rý f- P- fl.- f- ý~ ý Co :ý K1 K1 Co
q r-I o0 m Qº ý""" l0 +ý KM !"Q\ Qt
o%
N
.d 4ý ý""" °% Q% "rj ri n
.ö"" ".
C4 0 M Kl Kl Kl tl " t- N en are N
rd 0
mco
t- il%-
äöööö A""'
r-
iM CJr 1ý
U-N
TTTT Co .0N -*
.aý
a -i a
ý ÖMNO q
pf 0 0 1+1 UN O0
o P: 1ý- rn Pý CO t
ý ö0 Q, göö 0 67 O
°.
a a- 624 0 ... p, """"
4 f- t t-
fT F
q-
d rrr r- v-
 .n00 UN CD 00 cc cv Uli N. N N. u1 N.
ýp O C) C\ ýý Co ý co .- ýi: CM 0 .-M CO
h-
" Co6"000"06600000 !- !- CO t- Co (- CO Co CO f`-
N- i-

tJ 0 tr t- fl- CO n CTti tll O ý ýw ti

-G- h- co co '' CC) n ý,O 9,0 N. Co 1f1

ull ooLn0 o ti oo°

x ýý u, .ýý.. N CO til
öý ... hý w 0""N .7"
.ýc,
-- ,. cm 't. Sý
..
N (V NN Kl Q% e-

1`- O0 fl.- O tý 0 %a Lr%

ON MrN O C3 © LM "9aN
tt1 l, 0 kD ý10 ýfy tt1 tSl ta lfl
""ý"" """"-"

000 0,0 0 tf1 OOO

ý0 %I0n ti °N Öý0
0 rL? r --CO OÖ
r q. N is - %- N L'ý sý rrN
a*'N C9 co -"

Iü
w h
rh =t .-M UN tr% ti Ln ul, %O (- MMM CIJ
Ln ý NO rN
tilt
ON
ý N .ýý -t-
N :t
cc" 00 (t1 o Ln o0 O tt1 tn O
tt1 MO
ti ÖÖÖÖÖ
OO 00000
c}, """"" "....
0f660

A3

en co00 00000
K1 N1 MMM
s1
"Ä Co NP% t0 P- W co Co co Co ý- e- e- v- ý-
"ii co
q4-31 tf1 O V.
0 M
NNNN
co 0% 00 MO M Co Co ult t11
r4 tt1
.7ý E:, '1O
.4.5 N-N ÖÖÖÖ Ö.
CO 00
g: ýoooo o 000 """""
r4
as
P. 2 ýx 0
aCA)
ArA ý O tn 0
%. Ln o
N1 O CO ON !n %a - oooLnrn
N D
%. ý, ON M co
o M ti ý ON
uýi? ýýý v-
00 43
výn '`00000
ý
Iz 0 0 0 t- r" r` r- 00000
m"' v C3 UNCA -r- en ,n COCo Co Co Co
tM0 tý1 r M co tn 0 Ln ýýýN
-9
rs ä ý' ý r T
q
t- t- s- r
t-
äi
P.- ýO r*-
c) r- h- t- h- co N'1 td :1
ýi Qi 0,% C% 01 CMýi 7% M
(fr% CO .M- O tD
MW ý l0 p1 ý-
M'4
ý 'M W4 CO ""s"
Fi MAMA afTTNM ý" ý- N
O- ý-
ý r-1
C .ý Cti rý1 M 1`
r-1 ý\ý 00000 w .ýNN
4.4 00000
TTTTT ANýýNý ix
p a _I
O . -s
O ý N if1 N u1 tt1 t uh n0
ti O .- N or) tV ÖÖÖÖý ir- N Co
0 OO ß1 0000 9-
r oo+- ý""""" 4"""
a a t- ýe-
.". TTT 1'- ý" r- ý- ý- .r
.- ý- - ý-
 U Cý ýM U1 N iü1 O(V t11 iV U-N N en
ýýýN ti ti ý Co ti NoOM P- ti ti
Co
. 0606460000.66

00r .-R ºrl lD lD Co N

UN
ýýýN
rn ýa .o
ti t lD lii N -_t
(-
.:
-%'qqcoq
VN UrN 0
ýO3CO .9 In N"
CO O"
tfý ý t11 ýD "" LO CO lV ""f0 _t
(ý N t+1 ý CC) ý- .-
r- w to ým .ý e- e-

r- r- u,% o tf1 O N OOOON

i- t-
I-D 1` CO i
o "- UN cn ti tfN u1
tn N
lkO tf1
9-
ýa n
"""" ". .n ýn tiD %ü
"0*"s

ýIý,
ý2tf1ýýM
J.:
N Q 0
NO %D f-
CO
NN
NM
lýD
%D
0
CV
-4
äMM ý' Lf1 N9 0:3: tt1 4- n :t zt Ln
.
. t .

vü0
.i. 1ý^ Co
'}I-
Co ct1 N U*%r- 1,-- 0
ttl 0 10 rNMýN ilO CO CT Mm
" cV M ý r- NM -I
y
-2 -1
:;i' -'I

tf1 U1 N t11 N 111 Lf1 O ýQ OMO ýf1

S COt1O MM ti ti ti N h. 1D N1 N
_pI ý ÖO0 V

0OOOß 00000
caý- 6..
0*100 0 0"6"0

.ý yIý
7N
0 ý
Y-a O
" CD Co
ý- M tt1 tý- OOOOO N NýNN
y. ý" 1- co co co co Co
b ý, %- I" rrrrr
C
Q 0 ýcr 0 N. 1` N1 N1 n cu Co U1 N Co
ýNNh
'Iý ÖÖÖÖÖ
00000
ýO
000 ý"ý""
.... """""

um t) U
.. 4
. ai=-
Q} tt1 tf1 N00
ý

oa
ný
ýT

o0 tr%
co
'I-)
O kD O
C1t NM
KI
K1

ýa
W
ý
t9 O K1 Nýý
eM-
trl tV co
N.
43
ß-1
0
Al
N. t0
V- NNM
00 l0C%

OOO
00 OOOOO

ýb
00
Q1v t-
awNMM
p
t00 O
Co 0
c0

p
ýý
rrrrr
I%
-L-,
C:
O
MI
rnrno. rnrn

Co CO Co Co cot NN O tf1 tt1 co

ia .: t UN t11 P- V- i r- Oý
le. 4 4-i 0 --: t0
p ".ý """" .rt
v"
TTTT v r ý- NMM rrrN
ý 0 0
0 r-i
ý CO N- {'J NN tt1 N
4i OOOO CO Co "" I* t0 CO """
erký0 000 "" tt1 0N ý
ý+ TTT %0 CID rNN X4, Mý0 eý-
C7
n n
0 0
F+ ra C3 NN
0 tt1 CO P- 0 %O -* fl- oO Co
F4 a-! N tt1 0" o C+ 01 9-
Sö ö a P40000
z P+
00o
"... r- T
" P. """"
.-"""" !'TTT Trrr
N WT "- t- T 4 r sj'

ý
ý

H
 ýoo tý eo VN t- o t[10 !, - tC1 VN O Cý
do m r- m "- CO ti Co ti Co ý e7 t1ý MN QIN
Co t- +a co cu Co r-
00060 00006

e
ý OdN N OO Vj K1 tflý C7 NN lD q-
.N LrN ko ko h
Co Co Co Co Co

P- tt1 CO n .ý srý eý vD
z4 ti ti Ným N6"
t- U "
"N ý* tý
Co W1 -z
""
.". _z M
rrr ý r- rr e- Prl lp CJ eý
W. V-

tn u'% cV co a co O K1
co 00
NNNN CD P- r- r q-
r
tt1
tt1 tt1
?(N 1-
tt1
r
V ýD ý ýD W ta :I 111 I'D %D W " 4'\ tf\ ý7
ýý
0f400 f"""ý

Ul% tt1
N co O Co N0 tl1 N WýONN
0
ti p tf1 rr fV (, )
Kl O r-
%. Or M tl1 tý co
.*N
a
v r
. ý. a
tf1 U1 t+1 tl1 O NMM N'1 K1
4J1 . t,0 CO r aD
O r- to ýO ýN %D n- ý UN N "- P- c3
NMM n tý t- N ºe1 K1
.ý --t
ý tD OOnO
Lin ý 0ýý O UN 0 UN o
tD iCN NO tt1 .ýN O\
'1ý ti ý t- V,-
Ö0ÖÖÖ
00000 O0000
6" !A4f "ýý""
Cl """""
sý M
0
CJ f-4
,&3 W
üI -, N It tf1
M
M MM0
-r-3
t+
"- ý
ti
ý Co
ý- LA

N
t+1 tf1
NN
th MO
00000
co CC)co CC)(a
r- rn KN O
rrrrr
OMM
Kýl

t11 O
elü ti rz ti ko ti ti ti f- "O tf1 tt1 ti1
f~ N- f- f- f- f-
0 OOOO
orZ-
6 "
o0000
0.000
OO000
"ý0
c C1
O 4.» ý
.. +" O tl ý
OO co ýM 0) KN CO NO NOOO
.ý ta
t-+
tý ý .9** WN co co
CIO N -1
* ý; %0 Nýýý
Vý 43
43 0-4
OW N 44
A i
O0 a co O K1 tt1 OT 00000
-e+ !"T 4"' 000000
00000
wo e- cr i u. W,1 ao t11 tf1 111 t11 UN tt1 t11 tC1 t11 0
"T NNNNN
.- CO 04 . Zt .: r 10 N- f"
00 .0+ !-
A 0
-oa n
G3 ým Co ý- ýNýN 0
42 NNNN (V tOD . UN
0 4406 .i Co """" O+ N 0 N
4) CO Co CO CO CO öý rNNM V t0
'd
r-I O
co ý i-1
ß) Co pN 10 Co N Ä ti M K1 t0
Zs "... ý
"\ooooo N c: ) ca "" CYN in
ý00000
V- rr N a- Kl M ýN tn
ON +- rN
n
ý
wC ýý O tf1 NMNN C3
O t1i NN c0
c°v tt01 Pý 910 C,4 R+ tn
ä"i"r"
OOOO O Oý O f- 0OýÖ ý-
Q1 ý " "' p,
Q
10 """""
rrrrr r. r. "r- e- s; rr .-rr

l
O
e4
p
d
ad
 tt1 OO r- O P- N I` tý O 1` N
CO Co ef1 Oý O t0
Co Co
"""r"""""
N. 1s- CD Co
0O co 1` 1`
ý-
" Co W
"""
I`

0
Oý ONM N a[) cm N c0
N Co CD Co ý ýD NN ca en t0 f- t`

ýG t11 1t1 Co VN N tt1 t11 tý1 tt1 zt

9- M ,1"" tt1 O l0 "
ýýý """ jer "" lfi CV
N .
t[1 Co NN ON t-
r
.,. "ý' .7

CO N (` f- hl r-- N f-- NO N
cli
V ON
ýO
ý
ýD
co co
111 t11
M Kl
ýN u'1 la %D
"0"""

fýT
00 O ý
K1 ýO zt UN oýo tQ_
+ýO O Co - 10 r- r CO e- {`+ N CO r
rN tt1 CO e- e- N ý`º Ire .rN

ýomNA cv CO r. tt'% ou co ýj, N0

Co K1 r- kD
" *. CJ ý+ " r- N*
.2

, -d"OO tn tl- C1A 0 t- v to tn t11 f-

, Co NN ýO t11 to r- N lp -.: O
" "' P- 1` NN Ö000
iR. O00 O 00 0Ö0
,
G2 0006
ý
O
.&2 r-i Lt1 (' LM ON
Co ý-' '^ """" 0MM0
ý
f. ý-
N
r OA QN
Ln tD
00000
. ýý.e 4 .L
COco cc) cc) co TTTT

e. `'" NQ UN N 1- MNN t` Co MOO

kO O kD
%. $10 ýb l0 ý ýD

ý ýL5
cis g ýfiý
ti N
00000
ti f`ý ÖÖÖÖ

ýý
04)
ý 9000
ýý t) 00 0ý ti ý0 ý
o co 09
C*j C? u1 q- UN r- K1 LCJ drM
Ö"M ä vD
;ý N tn l0 rý r 04
N.
ii
oaýa 43
Q tý a+ Op000 OOOO
ý1rýN O
ý. ý1 co r %D O t1
ý Oý 01 Oý Qý
.d tf tt tý
a Oil *' rrrr
MAD °ýO

- ýý°nýt°`ö
Etlß
äÖ0000
. to ....
ý ýüýöý°.
.+ ýý .0".
ýÖý "+ K1 """ -°
Cý
0 00 co .Zr -t .7 rrMM
0
"rrN
0
ef PN tc1 N
W 0000 .ý > acatiý0
Z
-, OOOo wM CY ýo M
2ý s- r Ný
w r- r CO
C]
"[
,5 ß7
tl
0 C) cis
"N t[1 N CO a tt1 ý tn tt1 O OM ü1 N Co
O Co
ps
\OOO.-N r-
9-
ä$ööör F. ON
p40009-
CO 9-
O "e- "00
SZ'i Pe " r- "e-
-"
n- a.. -. .. p. """"

ý 4r ý- rr

ý.
r-ý
ý
H
 il1 Vº ON VIN It u1 P- 1` N
ýv°ýý° ý ti rn
r°J
co co
ö
CO
(ý" N tt1 ON
p1 Q1 CO (-.
4006 """"

bf`
M vy %t5 t0 N :T C4
Co CO Co t! '1 sD (ý- CO I UN

M 111 Co O M
Oß
ö0ö
ýO tD
""
_zr
""
tt1 ;
ýýO0
"a&
i
cv
N lV CV Co*
N-- Ný ý- n- T- N %D %-

ON N t%-0 ýýý :t1 w r- N ! 11

U P- Co f- N ý M CO "-
tt1 tn UN %D to ttl %D
"""" """
0&

ýý
>
ý_ .
ri
000
tt1 i- tn N
o
t-
000
t- OM
+-
0000
ýýý
t11 Ný Kl NV-
ra p% r- N K1 M r- NN rrN i+'1

yIJ N ON
Ci ý
h
N C4 tr% fn
kO UN co
.ý .ý 0 r- o f- ý
Ul ý
t7 K1 '' 1 Kl tir- N- tn
ý q- e- ... rrt! .ý

ý111 1l1 tt1 r %I0Co %IO0 N- kO L[1 tt1

ýIvÖ ýÖÖ l0 t11 N N P: ti ý
OOO0 üOOO
6660 06"0
ý ý"""
ý
üIý N Qý C1 ttý11
3O
: vr+
.
-ým r
}
ti
NM
ýt1 tOý
I- "-
0000
Co Co CO
t0000
fý 1 'tr
Co
Co Co 0 rn CIO r- MN
J Iý NN fc- p O
%0 %. u1 KY tO tý ý t0
ti ti r- ti tý- N ti r-
OOOO OOO0 O o00
q, ý 0060 06f"
"""ý
mý q
. c) f` O ýO
:4Ö., ýRIZ
01 1:y ti0 l0 a) t11 ONO to
=. . ýýt~D O
v 'T v
ý
ýý
..
(%- q- C\I ýýýýr "ý- N
A4
H
ra ', 43 4ý
Iý H
Od 000
NNN 0
N 0
Nl Ul%UNIn
NNNN
tn
ýdp V 0 If1
ýý, r"" tn UN UN ý- tn tf1 tfy tt1
COÖ 4 ýN Co y+ PT T* N 04 N N
.ý .ý
a
dO ti0 ýO ýO ýC
oýh.
UN M
fl-. .'r"" S
43ON un o
NNNN 6 ý- N N výýýN
'A, ötA 0
ý
01 N !11 tlý
xý P
000
000
rrr
""""
N :3
cM0
ý
ý~N
O
tý11 Co
v
.rs cs
ýÖ NM tr1 t11 N lt1 O 0 1-
%O 14 .2 co
UN tn
r
OOr öö? ýö 0 "-
f3. OOOr
A. """
%- r- r ""..
f- f` Tr
Pt ý+"ý
.Q ý- rrr
 tn o0 UN Lfl%
Ln
er r cý No r- Co p
U% Co Co I.-604 co Co O1%Co 1,- co
*00 f460

"
C7% ON ti
$' ti ti CO Ul% NN 0 oý
%.
ý ti ýý

tl1 U1 ON
x ýO N f c% u1 O""
-o ""w
-:
N
ý"ON
t` f. fý- LA LA ý f-
.ýr

U'% M tt1 [- cV ti NO f-
U f` 110 Co t11 tf1 f- f-
u1 "n .ý ýýý Kl tfN t11 t11
trl, UA LCN
4".
'º2
000 000 0000
l0 tD CC) M
CO O0 co :4 co 0 Kl D1 N
.o
4
4-=
tv MO
e- M kD NMM . -t

"
fV
1,
,y oý cm u-N
QÄ ýn
CO
tt1 UN
V'1 rr
l.'º NN
NOV.
NN
N t--
kD %ýQ rn
O
N 10
"Nnn

OOo lD OOO
co ti o. tI1 N cm ttý tC\ N0
t`,- f` l0 (`- 1- %D
c OOO OOO OOOO
606 06" ""6"

üIý M 111
o
""M 0000
MMMM
- Ln --t CO
Qh r-
OOo
co co Co TT T' t-
.IM
-0 Co tn MMO CO 0 tl1 N
ýo
ý1ýý Co 9 tt1 NO Ln U1 NN
r-
0 000 OOO0
0"
ýý w: a
00 4000
""
903
*ab b
00 ... 26 oO ýo ý O t11 O
j' VO N
.i Ln FE:
OýÖ
0 tf1
Co
M .7
CD
n CD
`ýTT 1- .-N
v ý
+N
044- jl. OO WOOO 0 0000
u O
T
"O
Cfl
O
-t O
Co
t11
Co Co
tt'% t1'1
Kl 00 0NNN9
e°ooý_ IR- -: Co TT T`
ýaNýý i!
,p
nf, >b N
Ch 01 (TN Co 1` r ýO Qý
n °o 43 N
44 ý ". .. ý (ý. "f 4"4
Ä 02 d"Tý arrr
r, rr
,d ý 0
0 MN U\
i f"&0
4 ööö t0rn ON t11 V- O
ý ý- r .- A Co N f- O\ r
C)
ý C3
0 " tn Co tt1 O trl% tC1 O
tf% rl ºn
s. ONM ta O 1- N o cl -olO lp
C., .-".. OO P1 OOO
0000
P4 0ss
0 Pt ý- v- ý- 4 "- rr r ... 1- t- r
ia
 OOON M t1ý tf1 O tt1 OON
ýO CO OO 1- +- rO p (- {- r Co
1` Co Co t- W CD cn Cfý C5 CC) f-
f640 64a" ""ýf

0 K1
ge- NN MO IM . - N Ký Mý
Er Co Co w 1- t` Co ýýNN
Co co

UN N
r-o
cn ýn r-
%lo,
rn o
"" rn
baý
ýý . CO !"" 4 4-
NN M (r
ý44 r1 IX) -r-' r

Q Co Il\ N o uIN h- 0 tv O (; ) aý
fý- N CO r
VNO
ýý
"-
ýO -2
tD ýýNý tV t[1 ttý ýl)
A""" .ý. " """t

OOO OOü
uti cn co
00 trN U\ -ýt C, ý
NO co +-
o"-ýtý
O r
ý, CO Ch Q CV U1 CV
ý, %O t- eý rN
aa e- e- N rrr

LV
('1N co ko ON {-
ý cO ýO t11 tr1 ON ý tt1 O t`
r'a ý" Ný I- N*-: .-N .J .7

O N 0 co cn
00 tc%o
" tA
q--ýä
U,
01% co
äö
K1
P-
O
NZ ti ti rn to
p
43
0
-" ." " 0000
6600 0000
0000
O
q3rI üIQ N. ý ý Co tn
EU q
000 0M
r-ý M,
M ýý
ý f.
M 0000
R1 ® >A- w Co Co Co r .- V. ..
r--4 Ln Co
Qý NOO ttt NMOO
ý ti Co Co ti ti co t- (7) (0 Co
CD r- 1- ÖÖÖÖ
43
0 a ar: o 00 0 OOOp
043a3 0000
"t (º "
.,ou a
21_ý LM 0 tD M O +r1
.4c. wýM tn o U1 kD O U1 u1 Co
:a. f+
-1 to tD
rný
t''1 ýD
T- N
r
N
00 .ry e- *- t- vi
ota V4 A
UN o 000 0 00
do 0 N LL'1 i- NNN 04 NN 00
NN
m: c r Co rM t11 ! t1 u1 t11 NN Ln LCI%
ta rý r %- e- .-r NN
a
ti u%Co to n K1
ma.ýo 0 00 oNoM t11 O co -1 N
Kl """
andm ....
ý- r t+1 M ýrrN

li ~
ýIlkýo
yo000
000 N
"""
r- crý
" u1
CO N
u1
tt1
"er r e- -1 Co . Zt
003
!! 1
tt1 cc1 LM N cu tn UN tt1 0 u1
v ä%D N CO
OO
CO (T ON
OOO
CO s-
!"
s-
ýr
0000
"""" " ti " ý
. ."
.-"
0 Pý''
T ý- t- It rrr t- 41" r- e-
Co ý
N
d
r4
ý
H
 N (- tt1 ýoý rn Oº
CC Cý o in o
Co (- 1'- CO o" ulý co 0
co co f- P- co t` CO
T"""
0600 0*00

0
ý- M '.ta CO co NNpN co (0 lf1
t! 1 i- Co .ý
bao Co M u--%
r, ti
Co Co
Co tn t[1 UN"Oo Ln tD OO kD
! M CO %O r"
D4 Kl .ý
"i .. ý
N1 """N
NpNN Tr- tri (7" t-

NONN 0N td1 P-
U "r- M ý- r n Co
l0 NN M
rn co t- cV O0
tD %D W 10
.ý tS1 ýO l0 M tt1 W %-0
000a

o0
tn CM 000 UN
o
NO .::r co .N- o
N-
cn
r
In 110
CO O
ýN
C9
ON
0 t
_-:
43 N q- rN %D rrN CO 40- rN

ý' 9iP4 ýa
ý
r
ýN
tt1
No
0N
ý
... 4
th W
eN-
NMI
01 tc1 ýO M 111 -1
fý t11 ý ON
ý- N .7.. 7

td1 t11 M nNN ON o t- cn o

co _zr 4 t- VD M ý+1 Co 10 NN

q
0
ö
ýi%.
0m
ýööbO
....
NNN NNNN
OOOO
6""6
öööö
0404

4a
+a O ä Lt1
4i) r-4 "'-A'I`^ r, r-
ri 00 " 0 M
+ýs C) N CO ü000 M to
M tl1 ýD t- rr
tit, co Co Co co t-
qA .a, o th Co ON
ý.ý.. ý+ (X)
Cn rn Ln un Co tr% Co r_
C3f-4 - ü Cp EIS- ti ÖÖÖ
v460 O 0000 0
119
qc
v0 q.; ....

.ý.. a v
tl tt1 0 4J O KN +r o
.0 ý 0 WO %0 Co N
., o% to
tt1 rM uli CO %0 "N ON

(1)ti 0:
00
n fl-+
_=o
a
zr
to rr
43
Ci
N t0 r e- MT

0
cn o 0 00 00
0C: 0 %IZ O rnýa Ln CC) LC%
tf%UN u1 tý±º M0MM
ýý % ý t1'1 M
tn Ul%ulk UN t! 1 tt1 t! 1 t11
ap,aC
aI tG e- r- n "P e- f- ßl r4 NN

mý ö %O kD t;0 sl0 43 P- %O O Co O O/ýýCO q- P-

C1 (IN Ot Co Co N OV ý

" tý "" fN /ý ý

.+w .ý-+ M +!ý MM ý0 " +r NM

UN yý""

6"wN
Aý
ö m
M %D K1
ä .ý 000 Je
cm M N
000 ý.". 9- th. ui
CA
ü't tt1 M
r- r t- .v -: cO T cv ,ý Ln
co
0
O
C3
'ý'9 KIN t1'1 0 tr% Co
tý a cv
"
N CO ý' N ON ö co
ON o
tn o0 ý
T
U ;
-ýrýÖ i" P. 00 000 T

ä` "0" a T
"
o ': "'
rr f" V. T' T
P.
-O, W-
 O ti N r. tl1 O0
\ýo (n N i1- K1 CO CO n
Co ti Co N Co r- ti Co
s004 6f00

0
LIN ri ýý P- %I0 t40
.gnM Co CO .7
CO CO t --t t- f-
%0
.7
tf1 rn tt`1 N t1'1 h- o tn
S.
ý ý- " Ln Ln "
9 0
O N fa6M
r ý
ý ý ý
-4.
ý o0 NN N Kl O) e-

N ti CO tn t+1 K1 f`K1 Lt1 OM

U (V MM l11 T
ff ý ý
UN to ýý
D ýO l0
%p %. %D
.. ""a"

ýiý
__
1
X.4
4.3
ra
tt1
Co
ti
tt1 OO
NN-:
O CO r'
r- "r- N
t tV
rn
U1
N
0
%"
ýa0

Cor
e" N
p tn
f`
000
tn
CO r
Uli 0
e- CO N
e- CV

.1. `^vgr../1 N M NN Öti M tC1 CC) N

>
iJ
ýZ-L a

ac_a -
YI- r.

N
01 N
1+1 ý V- N Zt, -.t
co z0
ON

Iv
ý. tl1 ýD tt1 cv o tn tno tn tM tf1 tt1
C4 tL1 NM ýp tt1 N e-
! NN Ö000
OO 00 00
Oý--'öit 0 *0 0"40 "60"
p
rs I U
O .. l0 M Ln
4J H OO
MMOt! 00
.sý
N co ý1 O 00 00 MM
-ra t, ý N rr
c~V týl Co Co coco
co O t11 o r- 1 NO
ti fZ: N

ý ýUN
co C8
co 0
iA-
. 4i
CC) NN
OO
N
O 00
a.
0ö
co
OO
ÖÖ
00
Orr
ý043 wýIlk
co
O
n
rl LM t11 Lt1 8 C, 0t Ul%
00 KA tfl O N
r/ i1 >ýý ý ý
M
Co o rý 01 --: M
v4t4
r- rný
N ýD rr
ý ý' V- r 43

u oco a a,
oa+a t! 1 ult 00 NCo\t
no ,ý,
O00
bo"+ý
,zr Co ;g 1-
r
tD
It1
T
NN
tt1 t! % U\
ý- e- fT
00000
m4.9
v
a -1

e t11 tn Ö r--
ON
n r+"3. OOOO co N +s O
I`- " " "ý ý
ý ür°4 ý: ý.. 6 "ý
tt'1 K'º
, "1
t) " V- I- N
" ý-
A 'd 02 b: -4 .7 ý
0
0 ý P- t- cm 0 co U-t
ý ýO --t
00 0 " zu% p_*"0
-X O ýi lD l0 lD M
00 0 ý." N ýý
C2 v
T O
ý0
ýN t1\ (V NO ý LC%LCI%
:+ 4-a f`- 0 N co fi ON .ý I- q-
pWno0 0 OO P4 0
-%. i: i
""
"
P4
00 "
r
t-..
0r- «-" q- r rr
qq- 4
ý,: p' r e- t- r
0
Be /VnLU rri'P, 90

CO 21 TFNTS
T1tl!
her
*. off
FIGUR M 126

APPETIDICES FIGURES 26

PiinT43RAPHIC PLAT FM 2h

Desip-n-drnwin? n (Separate donniment) 10