Professional Documents
Culture Documents
Mechanisms of Brittle Material
Mechanisms of Brittle Material
Mechanisms of Brittle Material
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
University M c r c t . n 's i
A Be."1 j. H o w 0 : " V\ V
U - M - I
300 N. ZeebRd.
Ann Arbor, MI 48106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
M ECHANISM S OF BR ITTL E M ATERIAL EROSIO N ASSOCIATED
BY
JIYUE ZENG
DO C T O R OF PHILOSOPHY
IN
1092
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
D O C T O R OF PH ILO SO PH Y DISSERTATION
OF
JIY U E ZENG
APPR O V ED :
Major Professor
D EA N OF TH E G R A D U A T E SCHOOL
U N IV ER SITY OF RH O DE ISLAND
1992
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
ABSTR A C T
published experim ental and theoretical studies of both ductile and brittle material
erosion. The review is presented in a tabular format for quick referencing. A study
using scanning electron microscopy (SEM ) is conducted to give the first-hand informa
tion on the erosion m echanism s associated with abrasive waterjet processes. It reveals
that the erosion m echanism s of brittle materials include plastic flow at the imm edi
ate impact site and a surrounding crack network. For polycrystalline ceramics, the
theory is used to m odel the brittle material removal applied to abrasive waterjet pro
cess. By analogy to the damage patterns by small detonation, the network cracking
crack network m odel to evaluate the fractured volum e is derived in terms of the input
stress wave energy and the required fracture surface energy. The stress wave energy is
expressed w ith a modified H utchings’ equation for normal incidence and with an equa
tion derived in this study for low incidence, respectively. The crack network model
combined w ith B itter’s deformation wear model gives the total material removal for
a single particle im pact at normal incidence, and, combined with Finnie’s m icrocut
ting m odel, gives the total material removal for low incidence im pacts. Observations
on the abrasive waterjet cutting front reveal that the cutting process is associated
with abrasive particle im pacts at glancing angles. The energy dissipation phenomena
removal m odel for low incidence im pacts, combined with the analytical results from
the energy dissipation study, is used to derive an equation which predicts the depth
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
of cut. By analogy to this theoretical equation, an empirical equation is also derived
which narrows the gap between the theories and applications. Based on this empirical
equation, a new m aterial parameter, called “M achinability Num ber”, is defined. The
processes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
First of all, I would like to thank m y parents for giving m e wisdom and teaching
me of striving for success and being patient, w ithout which I would not be able to
face any failure. I would also like to thank my wife for her love, encouragement and
understanding. Besides, I want to say a “Thank Y ou” to m y little boy who, though
not yet spoke, m ade m e happy when I was upset. A very special thank is given to
Dr. Thomas J. Kim, m y advisor in m any aspects. His wise guidance and friendly
attitude become driving forces for me through the course of this investigation and
are unforgettable in m y life. My gratitude is extended to Dr. Arun Shukla and Dr.
O tto J. Gregory for serving on my com m ittee, to Dr. David G. Taggart, Mr. Ray
McLaughlin, Mr. Manuel Merril, Mr. Rodney W . Knight and Miss Minnie Sagar
for their m any essential helps, to Mr. R. Heines, Mr. R. J. Wallace and Miss H.
A. Costantino for conducting some of the experim ents. Finally, I thank my labm ate
Shaoyan for his support and friendship. I wish a good luck for his future.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
Page
C H A PTER 1 IN T R O D U C T IO N .............................................................................................. 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4.1 An E lasto-P lastic M odeling Approach Applied to Brittle M aterials 97
VI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C H A PTER 8 CONCLUSIONS AND RECOM M ENDATIONS 186
A P PE N D IX C C O M PU T ER PROGRAM S 193
Vll
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Number Page
5.1 Experim ental data of the kerf width and taper ......................................................... 136
5.4 C utting test parameters and the critical jet exit angles ........................................ 146
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Number Page
1.2 A survey on the papers presented at the 1st - 10th International Sym posium of
1.3 A survey on the papers presented at the 1 st - 10th International Sym posium of
1.4 A survey on the papers presented at the 1st - 10th International Sym posium of
3.1 Experim ental configurations used in sweeping and grooving erosion tests .. 64
3.3 A cutting m echanism dom inates the material removal process of SS 304. (x 95,
3.4 A typical crater observed on the SS 304 sam ple, (x 310, im pact angle = 20")
..................................................................................................................................................... 67
3.5 A typical crater observed on the SS 304 sam ple, ((a) x 125 and its enlarged
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.6 Two partially overlapped craters observed on the SS 304 sample, (x 310, im pact
angle = 20 °) ..............................................................................................................................
...................................................................................................................................... 69
3.7 A raised lip at the exit end of the crater observed on the SS 304 sample. (x l9 0 ,
3.8 An enlarged view of the raised lip shown in Figure 3.7. (x 310, im pact angle =
2 0 ° ) .............................................................................................................................................70
3.9 An enlarged view of the m essy crater shown in the center of Figure 3.7. (x 1260,
3.10 M ajority of the craters on the SS 304 sample were created by an indentation
3.11 A typical indentation crater observed on the SS 304 sample, (x 190, impact
3.12 An indentation crater on the SS 304 sample shows inward grooving traces and
3.13 An indentation crater on the SS 304 sample shows evidence of ductile fracture,
3.14 A plastic im pression on the SS 304 sample due to a longitudinal im pact, (x 190,
3.15 An indentation crater on the SS 304 sample with particle fragments em bedded,
3.16 Overall appearance (a, x 490) and a partially enlarged view (b, x 2500) of a
typical crater observed on the AD 99.5 sample, (im pact angle = 20") ........ 75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.17 Overall appearance (a, x 240) and a partially enlarged view (b, x 970) of a
crater due to an erasing action, observed on the AD 99.5 sample, (im pact angle
= 20°) ................................................................................................................................... 76
3.18 Overall appearance (a, x 500) and a partially enlarged view (b, x 1500) of a
3.19 Overall appearance (a, x 340) and a partially enlarged view (b, x 990) of a
crater observed on the AD 99.5 sample, (im pact angle = 90°) ...................... 79
3.20 Tw o fractured craters (a) &; (b) observed on the A D 99.5 sample (x 490, impact
3.21 Top view (a, x 33) and side view (b, x 40) of the groove created on the AD 99.5
3.22 General m orphology of the groove surface on the AD 99.5 sample, (a, x 1500,
3.23 Some isolated craters and semi-craters observed on the edge of the groove of
the A D 99.5 sample, (a, x 290 & b, x 860, im pact angle = 90°) .................... 83
3.24 General damage pattern on the centered area (a, x 160) and the exit edge (b, x
40) of the eroded AD 99.5 sample, (im pact angle = 20°) ................................ 85
3.25 Two individual craters observed on the AD 99.5 sam ple, (x 840, impact angle
= 20°) 86
3.26 The top (a), middle (b) and bottom (c) areas on the vertical groove of the cut
3.27 A chip formed by a cutting action observed on the cut SS 304 sample, (x 500)
.................................................................................................................................................. 38
XI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.28 The vertical groove on the cut AD 99.5 sample, (a) The whole view (x 1.5); (b)
3.29 The top (a ), m iddle (b) and bottom (c) areas on the vertical groove of the cut
and scratching trace (area C) on the cut AD 99.5 sam ple, (x 500); (b, x 1400)
3.31 The very top edge of the vertical groove of the cut AD 99.5 sample, (x 500) 91
3.32 Steps (indicated by arrows) formed on the profile of the cutting front of the cut
3.33 The entire cutting front is impacted by the original jet or the deflected jet at
glancing angles..................................................................................................................... 93
3.34 A hypothesis for the step formation in AWJ cutting processes ...................... 95
4.1 Scanning electron micrograph of isolated im pact site on sintered alumina (No.
46 grit, 75 m /s). (A ) bar = 45 /rm, (B ) bar = 15 /im . (after Ritter et al., 1984)
.................................................................................................................................................. 98
(b) polarized reflected light micrograph, (after Evans et al., 1978) .............. 98
4.3 Optical reflected light micrographs of orthogonal sections through ZnS targets
im pacted by WC and glass projectiles, (a) The extent of the radial and lateral
fracture (parallel to the surface) for a WC projectile target; (b) the same im pact,
but highlights the formation of lateral cracks from radial cracks (e.g. at arrowed
location); (c) the extent of the radial and lateral fracture for a glass projectile
xii
with permission o f the copyright owner. Further reproduction prohibited without permission.
impact; (d) the same as (c) but indicating the zone of microfracture (arrowed)
beneath the center of im pact, (after Evans et al., 1978) ................................. 100
4.4 Dam age from a DIAjet particle im pact on glass, (x 1000) (after Summers et
4.5 Fractures in a ‘Perspex’ cone produced by 0.4 g of lead azide. (actual size)
4.6 Stress wave pattern and energy partition, (after W oods, 1 9 6 8 ) ..................... 102
4.9 Comparison of normalized erosion ratios for six alumina ceram ics...................114
4.10 Scanning electron micrograph of hand-m ade fracture surface of AD 85. (x 800)
116
4.11 Correlation of the grain size and fracture energy with the erosion ratio. .. 116
4.12 Idealized picture of an abrasive grain removing m aterial by scratching and net
work cracking..............................................................................................................................
118
4.14 Comparison o f the elasto-plastic m odel and the experim ental results 132
4.15 Grain size and fracture energy are the two major m aterial param eters. .. 132
5.2 Tapered kerfs on an alumina (AD 85) plate cut by an AW J............................ 136
xiii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5.3 Kerf width vs stand-off distance.................................................................................. 137
5.15 Effect of water pressure on the criticla jet exit angle........................................... 148
5.17 Effect of abrasive flow rate on the critical jet exit angle.................................... 148
5.18 Effect of workpiece m aterial on the critical jet exit angle................................. 149
5.19 Effect of workpiece thickness on the critical jet exit angle................................... 149
5.20 D ata of critical jet exit angles for 38 mm thick workpieces.............................. 149
xiv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6.3 Depth of cut versus traverse speed........................................................................... 158
7.1 Depth of cut versus water pressure for different traverse speed....................... 163
7.2 D epth of cut versus water pressure for different abrasive flow rate............... 163
7.3 D epth of cut versus water pressure for different orifice diam eter...................... 164
7.4 Depth o f cut versus water pressure for different nozzle diam eter................... 164
7.5 Depth of cut versus water pressure for different m aterials................................ 165
7.6 Depth of cut versus water flow rate for different traverse speed..................... 165
7.7 Depth of cut versus water flow rate for different abrasive flow rate.............. 166
7.8 Depth o f cut versus water flow rate for different water pressure..................... 166
7.9 Depth of cut versus water flow rate for different nozzle diam eter.................. 167
7.10 Depth of cut versus water flow rate for different m aterials............................... 167
7.11 Depth of cut versus nozzle diameter for different traverse speed....................... 168
7.12 Depth of cut versus nozzle diameter for different water pressure................... 168
7.13 Depth o f cut versus nozzle diameter for different abrasive flow rate 169
7.14 Depth o f cut versus nozzle diameter for different water flow rate.................. 169
7.15 Depth of cut versus nozzle diameter for different m aterials.............................. 170
7.16 Depth of cut versus traverse speed for different water pressure...................... 170
xv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7.17 Depth of cut versus traverse speed for different abrasive flow rate...................171
7.18 Depth of cut versus traverse speed for different water flow rate..................... 171
7.19 Depth of cut versus traverse speed for different nozzle diam eter...................... 172
7.20 Depth of cut versus traverse speed for different m aterials................................ 172
7.21 Depth of cut versus abrasive flow rate for different traverse speed............ 173
7.22 Depth of cut versus abrasive flow rate for different water pressure........ 173
7.23 Depth of cut versus abrasive flow rate for different water flow rate...........174
7.24 Depth of cut versus abrasive flow rate for different nozzle diam eter 174
7.25 Depth of cut versus abrasive flow rate for different m aterials...................... 175
7.26 Correlation betw een the predicted depth of cut from the empirical model and
7.27 Normalized depth of cut versus water pressure for different materials. . . . 177
7.28 Normalized depth of cut versus water flow rate for different materials. . . . 177
7.29 Normalized depth of cut versus nozzle diameter for different materials. .. 178
7.30 Normalized depth of cut versus traverse speed for different m aterials 178
7.31 Normalized depth of cut versus abrasive flow rate for different materials. 179
7.33 Cut surfaces of AD 94 (a) and an unidentified steel plate (b )....................... 183
............................................................................................................................................... 189
xvi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
NOM ENCLATURE
A loaded area
as curve-fitting coefficients
b„ curve-fitting coefficients
c constant
C, scale factor
c, curve-fitting coefficients
Cv orifice efficiency
d diameter of particle
xvii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ep m odulus of elasticity of im pacting particle
e coefficient of restitution
G„ real part of Yv
h depth of cut
Hd dynam ic hardness
xviii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
kl see Equation 4.48
Nm m achinability number
p„ water pressure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
rj radius of the im pacting circular disk
T = 7r /u lo a d in g contact tim e
t tim e
u surface displacement
xx
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
W stress wave energy transm itted into the target
Wi work done by the vertical force component over the superposed vertical
displacem ent
horizontal displacement
horizontal force
vertical force
a im pact angle
xxi
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
e angle of slope of the cutting front
VP
P oisson’s ratio of im pacting particle
Pw water density
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
CH A PTER 1
IN TR O D U C T IO N
The idea o f cutting w ith water could be as old as the Chinese phrase “dripping
water penetrates rock”. In the modern waterjet technology, the “dripping w ater”
is replaced by a high velocity water stream, which can penetrate a rock in seconds.
The modern waterjet technology was initiated by Dr. Norman C. Franz in 1968,
who was awarded with the first patent for a high pressure waterjet cutting system .
Based on this idea, the first commercial waterjet cutting system was developed to cut
lam inated paper tubes in 1971. In the following year, the First International Sym po
sium on Jet C utting Technology was held in England to bring together the leading
ous growth. During the following decade, significant im provem ents were m ade on
the overall reliability of the high pressure system with em phasis on extended ser
vice term, simplified m aintenance, and reduced costs. M eanwhile, research interests
were greatly expanded, which led to the m eeting of the first American Water Jet
Conference in 1981. The rapid advance of autom obile industry, m aterial science and
space technology in 70’s and 80’s demanded and also stim ulated the outgrowth of
new ideas and novel technologies in manufacturing. In 1983, the idea of entraining
abrasive into the water stream was patented and was im m ediately followed by the
years of efforts from both academia and industry, abrasive waterjet technology has
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
grown from being a showcase technology to becoming a viable tool for various in
dustries. The abrasive waterjet technology branched out in 1984 with an alternative
technique, i.e., the so-called premixed abrasive slurry jet(A S J) technique. ASJ used
the idea of cutting w ith abrasive/w ater mixture in a different way. Instead of entrain
ing abrasive into a high velocity water stream and then producing an abrasive/water
The primary potential commercial application for this technique was for offshore and
com petitive with the conventional AWJ. Figure 1.1 shows the important historical
ture survey was conducted based on the papers published in the proceedings from the
first through the ten th International Sym posium on Jet C utting Technology, spon
sored by the BH R Group of UK and held every two years since 1972. The first task in
this survey is to categorize the work (including research, system design, applications)
into six categories: (1) waterjet (W J), including continuous and discontinuous jets;
(2) cavitated waterjet (CW J); (3) conventional i.e., entrainm ent) abrasive waterjet
(AW J); (4) abrasive slurry jet (ASJ); (5) waterjet assisted processes (W JAP); (6) oth
ers. The percentage of the number of papers in each category is shown in Figure 1.2.
Plain waterjets were predom inantly used during the early years. It is interesting to
note the steady emergence of the conventional abrasive waterjets since 1982 and the
abrasive slurry jets since 1986. At the 9th Sym posium (1988), the number of papers
and by 1990 (the 10th Sym posium ), abrasive waterjets had became dominant. Figure
1.2 also shows a trend of research and developm ent works on the cavitated waterjets
with permission o f the copyright owner. Further reproduction prohibited without permission
1966
Patent on ttie
c o n c e p t of
w aterjet cutting
1971
1st com m ercial
1972 waterjet cutting
1st Int. Symp. on system
Je t Cutting
Technology
1981
1st American
W ater Je t
C onference 1983
1st com m ercial
abrasive w aterjet
cutting system
1984
1st commercial
abrasive slurry jet
cutting system
1990
1st Asian Conf. on
R ecent A dvances
of Jetting Tech.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1001
90
C3
80
t) 70
u
u
60 ■ OTHERS
<+-
O
50 m WJAP
ASJ
40
H AWJ
> 30
5 □ CWJ
20 □ WJ
10
0
(S ■*+ NO oo o <N CO o
r- r-- r- r" 00 00 00 00 00 ON
ON ON CN CN QN GN O'
Cn ON ON
1—H
•a •C *£ JZ •C
c T2 •*—*
in <o
4—1
r-
W
o
CN cn ON
Figure 1.2 A survey on the papers presented at the 1st - 10th International Sym posium
- cavitated w aterjets, AWJ - conventional (i.e., entrainm ent) abrasive waterjet, ASJ
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This survey also reveals the application diversity of waterjets and abrasive wa
terjets. The applications are generally divided into three catogeries: (1) machining,
including cutting, drilling, m illing, and turning of various m aterials for metalworking,
construction, rock excavation, food processing, etc.; (2) cleaning, including decoating
and descaling; (3) m ining, including tunneling. M achining overwhelm ingly dominates
the waterjet and especially abrasive waterjet applications (Figures 1.3 &; 1.4).
OPERATING PRINCIPLES
W hen water is pressurized to a high pressure (upto 414 M Pa or 60,000 psi) and
discharged from a small orifice, the velocity of the water stream can reach as high as
800 m /s. Im pingem ent of such a water stream causes damage to materials such as
rocks, plastics, woods, or even some metals by shearing, cracking and delamination. A
long-chain polymer is som etim es used to enhance the j e t ’s cutting efficiency (Franz,
1972). A dram atic increase of the cutting power is achieved by adding abrasive
particles into the high velocity water stream. In this case, water serves primarily as
an accelerating m edium while the abrasive particles take over the role of material
removal. M ixing and acceleration of abrasive particles in the water stream can be
abrasive into the water stream upon its exit from an orifice. The abrasive particles
is mixed with and accelerated by water through a m ixing tube (also called focusing
w ater/abrasive slurry through a single nozzle in which both water and abrasive are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ss;S*
^ □ Mining
^
C/5
oe
•a
C3
O H Cleaning
a
<
o
‘35
o 3 E=3 Machining
>
1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th
International Symposium o f Jet Cutting Technology
Figure 1.3 A survey on the papers presented at the 1st - 10th International Sym posium
B Cleaning
c
o
_o
”3.
£
<
o Machining
o
>
3
Figure 1.4 A survey on the papers presented at the 1st - 10th International Sym posium
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
EQ UIPM ENTS
component parts. T hey are a high pressure pum p, a nozzle assembly, an abrasive
delivery system , a catcher system , a m otion system , and a control unit (see Figure
1.5).
The first stage of the high pressure pump is a hydraulic radial displacement pump,
M Pa (3,000 psi). The pressurized hydraulic oil is then fed into and drives an intensifier
pump (Figure 1.6), which in turn pressurizes the water. T he output water pressure is
equal to the oil pressure m ultiplied by the area ratio of the two ends of the intensifier
piston (usually 20:1). The water supply to the intensifier pump is required to be
filtered to 0.5 micron and m aintain a stable pressure o f 0.45-0.55 M Pa(66-80 psi)
which can be provided w ith a booster pump. A typical arrangement also includes an
The main functions of the nozzle assembly (Figure 1.7) are to house the waterjet
orifice and the m ixing tube and to provide an o n /o ff control of the high pressure
water. For convenient alignment of the orifice and the m ixing tube, an adjustment
m echanism is helpful. Abrasive is fed, by vaccum suction, into the mixing chamber
located between the orifice and the m ixing tube. The key parts of the nozzle assembly
is the orifice and the m ixing tube. The orifice is usually m ade of sapphire or diamond.
A sapphire orifice could last up to 200 hours while a diam ond orifice lasts ten times
longer (Burnham , 1990). The mixing tube is typically m ade of tungsten carbide,
which gives a nozzle life less than four hours. A new nozzle m aterial (com posite
tungsten carbide) has been recently developed to extend the nozzle life 10 to 100
and a delivery hose. The metering valve turns o n /o ff the abrasive flow and controls
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
High P r e s s u r e P u m p A b r a s iv e D e liv e r y S y s t e m
I I
A b ra siv e
In te n s if ie r Hopper
E le c tric
Motor Pump
N o z z le
A sse m b ly
T an k C a tc h e r
S e ttlin g
T ank
C o n tro l U n it M o tio n S y s te m C a tc h e r S y s te m
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Low P ressure High Pressure
Oil O u t Oil In
Low
Pressure
High
W a te r
P ressure
In
W a te r
O ut
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
W ater J e t
O n/Off Control
High Pressure
W ater Inlet
W ater J e t
Orifice
Mixing 23 A brasive
C ham ber Inlet
Alignment
M echanism
Mixing
Tube
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the flow rate. The delivery hose transports abrasive from the metering valve to the
The catcher system serves the functions of dissipating the residual energy of the
jet as well as collecting the waste. It could be a tank catcher or a point catcher,
alumina balls are norm ally used for energy dissipation. The wastes including water
and abrasive are pum ped into a settling tank, where the solid wastes are settled for
disposal.
or a high precision robot is desirable. There are several options to accomplish the
relative m otion between the nozzle and the workpiece. The combination of a moving
nozzle and a fixed workpiece usually requires a tank catcher. A fixed nozzle combined
complicated configuration, seen in some X Y cutting system s, both the nozzle and
the workpiece m ove in perpendicular directions. This configuration is best used for a
The control unit encompases the overall system control including the cutting mo
tion, water and abrasive on/off controls and emergency protection. Normally the
water and abrasive control can be integrated into the m otion control system . The
ASJ System: An abrasive slurry jet system distinguishes itself from an entrainment
abrasive waterjet system in its nozzle assembly and abrasive delivery and m ixing
operation. Figure 1.8 shows a simplified abrasive delivery and mixing circuit. The
restrictor creates a small bypass flow, which is fed into the abrasive slurry vessel. The
slurry is then discharged into the water flow down stream of the restrictor. The final
slurry exits a single nozzle to form a jet. The configuration of the nozzle is designed to
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A brasive Slurry
Supply H opper
A brasive
Slurry
V essel wmmmmam
llllllllll
77777777,
W ater
Restrictor
Nozzle
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
focus the abrasive particles into the central line of stream to prevent excessive nozzle
wear. W hile the pressurized slurry vessel is discharging, slurry supply is drawn in
from an adjacent open slurry hopper, where dry grit abrasive is fed in and fluidized.
parameters are water pressure, waterjet orifice diam eter, abrasive type, abrasive size,
abrasive flow rate, m ixing tube diameter, nozzle stand-off distance, traverse speed,
The effects of water pressure on an AWJ cutting process are m ainly reflected on
the velocities of waterjet and abrasive particles. Theoretically, the velocity of waterjet
water and the orifice efficiency loss increase with the pressure, so does the loss of the
actual waterjet velocity. On the other hand, increase of the water pressure yields a
favorable effect on the abrasive m ixing efficiency. All these effects are included in an
expression (Hashish, 1989, p221) for the abrasive particle velocity at the nozzle tip:
_ vCy j 2 P w . .
v ~ T + r \Iw ( , )
where Pw is water pressure, p w water density, R ratio of abrasive/w ater mass flow
coefficient. Values of 77, C„ and Cv are given in A ppendix A. The overall effect of
The size of waterjet orifice combined with the water pressure, determines the water
flow rate which directly affects the performance of AWJ cutting. As a general rule,
the depth of cut increases as the water flow rate increases. Therefore, for a given
constant water pressure, a large orifice size usually results in a deeper cut. However,
since the water flow rate is lim ited by the pump capacity, the size of water orifice is
often within the range of 0.23 - 0.56 mm (0.009 - 0.022 inch). In addition, the size
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of orifice must be selected such that it is compatible to the size of the m ixing tube.
An oversize water orifice will result in water accum ulation in the m ixing chamber
and w etting the abrasive delivery hose while the jet formed by an undersize orifice
may not be well focused by the relatively large m ixing tube. Normally, the ratio
of orifice diam eter to m ixing tube diameter is chosen w ithin the range of 0.3 - 0.4.
m only used abrasives include garnet, alumina, silicon carbide, silica sand, steel grit,
glass bead. Since different com binations of abrasive/target m aterial have distinct
interaction and m aterial removal m odes, a certain typ e o f abrasive m ay behave very
differently for different target materials. For exam ple, garnet abrasive and alumina
abrasive have alm ost equal cutting power when cutting ductile m aterials. However,
alumina abrasive is much more effective than garnet abrasive when cutting brittle
The abrasive size in the ordinary range of mesh 50 - 120 has a minor effect on AWJ
cutting efficiency, but it is critical to the surface finish. Oversize grit abrasive reduces
the m ixing efficiency while undersize grit tends to cause clog in abrasive delivery line.
Abrasive flow rate is another major parameter. Increasing abrasive flow rate
increases the cutting depth. However, this relation is nonlinear. Since excessive
abrasive flow will reduce the m ixing efficiency, a flow rate larger than 7.6 g /s (1
lb /m in ) usually doesn’t produce proportional cutting power. Often it can clog the
The effect of the m ixing tube diameter is relatively small. A larger diameter
mixing tube usually produces a wider kerf and a smaller depth of cut. In terms of
material volum e removal, the effect of the mixing tube diameter is minimal.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Variation of the nozzle stand-off distance within a sm all value, say, 3 mm, has
little influence on the cutting performance. However, in some situation such as pocket
m illing, a large stand-off is used such that a spread jet is produced and the depth of
Traverse speed is the m ost critical parameter. Theoretically, the depth of cut is
inversely proportional to the traverse speed. Due to its significance to the depth of
cut and its ease o f control, the traverse speed is often used as the sole control variable
to be adapted to the change in workpiece m aterial as well as the thickness while the
Cutting angle is referred as the angle between the jet and the top surface of
al., 1990) recom m ended that an angle slightly less than 90 degrees in the opposite
direction of the traverse direction is ideal to com pensate the curvature change of the
The erosion resistance of target m aterial is not a well defined parameter, since no
single defined m aterial property can be used to characterize the erosion resistance.
For ductile m aterials, the erosion resistance is closely related to the hardness. For
brittle m aterials, more m aterial parameters are involved, such as fracture toughness,
grain size, flaw distribution parameter, etc.. It has been suggested by the author that
a new m aterial param eter, called “Erosion R esistance” , is defined (see Chapter 7 for
details). Prior to the introduction of such a m aterial parameter, some general trends
on the effect of target material on the cutting performance can be noted: (1) for
ductile m aterials, the depth of cut increases as the hardness decreases; (2) for brittle
process. However, current ASJs use a much higher abrasive flow rate (up to 180
g /s) combined with a low water pressure (less than 100 M Pa). ASJ does not use a
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m ixing tube. Instead, it introduces a new parameter, the suspension concentration.
The suspension in the slurry vessel contains about 50% abrasive by weight. The final
APPLICATIONS
Though abrasive waterjet technology is still in its infancy stage of developm ent,
it has been widely adopted by various industries such as aerospace, defense, foundries,
glass, autom otive, fabrication, hazardous environment, etc.. For exam ple, LTV Aerospace
and Defense C o.’s Aircraft division (Dallas) has em ployed a robotic AWJ system
to trim oversized com posite parts for B-2 bomber (Ashley, 1991). Rockwell Inter
national’s North American Aircraft Operations Division (Colum bus, Ohio) used an
vanced Technology System s, Inc.(M onroeville, Pennsylvania) has dem onstrated the
where sparking is prohibited (Agbede, 1990). The foundry industry uses abrasive
waterjets for cutting gates and risers off cast parts or removing burned-in sand from
castings.
ADVANTAGES AN D LIMITATIONS
high velocity water stream, any heat generated is im m ediately carried away by the
water. As a result, tem perature rise in the workpiece is less than 60°C. This rep
resents an improvement over not only traditional machining techniques, but also
m any non-traditional ones such as laser, plasma arc, electron beam , flame, electri
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in the m achining process, given appropriate abrasive selection, virtually any kind
ing m any “difficult-to-m achine” m aterials. Titanium , high strength ceramics, glass,
com posites of any m atrix m aterial, and honeycom b structures are just a few good
examples.
Small m achining force: The force applied on the workpiece by the jet in a typical
kerf cutting operation is less than 10 New ton (Burnham & Kim, 1989). This small
force allows m achining of fragile or deformable materials and structures such as glass
and honeycom b structures. Turning a long rod dem onstrates another advantage since
the vibration and deformation problems caused by the large force in traditional turn
ing or grinding process are elim inated (Hashish, 1991, p43). It also elim inates the
High flexibility: Characteristic of all point cutting tools, abrasive waterjet cutting
plicated cutting operations can be performed. For exam ple, abrasive waterjet contour
cutting has been widely used. Since the energy carrier is fluid, the cutting head and
the pum ping system can be flexibly connected. Therefore, it can be applied in re
decommissioning.
Like all other m aterial removal processes, abrasive waterjet m achining has its
lim itations:
where the cutting edge continually receives energy com pensation during the entire
machining process, an abrasive waterjet continually loses its energy due to dissipation
along its path. Therefore, in a typical abrasive waterjet kerf cutting process, the
cutting power of the jet decreases from the top of workpiece to the bottom , leaving
a tapered kerf and striation marks on the lower portion of the cut surface. This
17
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission
phenom enon has also been observed in other beam cutting processes such as laser,
electron beam , plasm a arc, flame, etc.. Decreasing the traverse speed can minimize
this energy dissipating phenom enon. Other m ethods, such as slightly inclining the jet
in a direction opposite the traverse direction (M atsui et al., 1990), have been proposed
High initial capital cost: Primary investm ent of an abrasive waterjet system will
require $50,000 - $500,000, depending on power and com plexity (M ason, 1989). Oper
ating/m aintenance cost is estim ated to be $10 - $30 per hour (M ason, 1989). However,
of this technology remain a m ystery to the world. W ith the m otivation of advancing
this new technology, this study will investigate the erosion m echanism s, energy dis
M aterial removal m echanism s are the key to understanding any machining process.
becomes the main concern and the first step of this study. Chapter 2 will present
a comprehensive review on the existing erosion theories, including both ductile and
brittle m aterial erosion studies. A SEM (scanning electron m icroscopy) study is then
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
conducted (Chapter 3) on AWJ eroded samples to disclose the m aterial removal modes
extensive erosion m odeling study for brittle m aterials, ■which leads to m athem atical
m odels of single particle m aterial removal under the conditions of AWJ impact at
reveal and characterize the energy dissipation phenom ena in an AWJ cutting process.
Chapter 6 uses the results of Chapters 3, 4 and 5 to derive an equation for depth
of cut applied to an AWJ kerf cutting process. This is followed by an experim ental
parametric stu dy (Chapter 7), in which the derived kerf cutting m odel is verified
on the justified kerf cutting model. These results are used to develop a parameter
prediction m ethod for AWJ kerf cutting processes. Chapter 8 summarizes the major
conclusions of all these studies and proposes recom m endations for future studies.
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
CHAPTER 2
A L IT E R A T U R E R E V IE W OF E R O SIO N ST U D IE S
fields such as particles (or slurry) transportation, equipm ent protection in a dusty
environm ent, turbine engineering, etc., prevention of particle erosion is the task. In
other applications, it is used as a tool for desirable m aterial removal, surface clean
ing, controlled destruction. Numerous studies on this subject have been conducted
in the book by Engel (1976) and articles by Tilly (1979), Evans (1979), and Ruff and
Wiederhorn (1979).
pact, a review is conducted based on the literature collected on the subject of erosion
by solid particle im pact. The reviewed works are sorted into two sections (ductile m a
terial erosion and brittle material erosion) and three categories (erosion mechanism,
parametric studies, and material removal modeling). For easy reference, a tabular
format is used.
Tables 2.1, 2.2, and 2.3 present the erosion m echanism s, parametric studies and
Tables 2.4, 2.5, and 2.6 present the erosion m echanism s, parametric studies and
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5 0 0 jim T illy
H46 ste e l 244 SEM A chip was cut out.
q u artz (1973)
R ip p le s w ere form ed a f t e r
a p e r io d o f e r o s i o n . T hin
poly 70 & 200 61 Brown
30, torq u es form ed on the
cry stallin e pm glass a ir SEM e t a l .
dow nw ind side of rip p les
a-Fe spheres 122 90 <1981,
ruptured ev en tu ally and
v70)
caused m aterial lo ss.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 ................... ..
T able 2.2 P aram etric stu d ies of Dv i c t i l e M aterials
copper,
10 Finnie
alum inum , m esh 60 R
76 (1958)
SAE 1 0 2 0 S i C
90
ste el
0 a 90
10
alum inum ,
copper, 3 0 0 |1
R
3AE-1055 cast
(as iron
10 B itte r
received) p e lle ts
0 a 90 (1963,
Im pact 90 pi 6y )
A ngle mesh 60
107
SiC
p a r ti
SAE-1055 R
cles
(fu lly
hardened)
0 a 90
210 P
alum ina N eilsen
p a rti 79 10 &
R
alum inum c les, a ir G ilch rist
475 P 150 90 (1968)
glass
0 a 90
soheres
Al allo y ,
60-125
chrom ium T illy
pm 10
steel, R (1969)
q uartz a ir 104
polypro
p a r ti 90
p ylene,
cles
nylon 0 a 90
2024 A l, 50-60 P
ri-6 A l-4 V alum ina 152 S m eltzer
410 SS dust 20 R
et a l .
a ir
(1970)
17-7 PH 0-210 P
90
s t a in l e s s A rizona 198 0 a 90
steel R d.D ust
H46 T illy
11% 135 Pm R (1973)
366 90
chrom ium quartz
ste e l
3 a 90
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2.2 E rosion param etric stu d ies (continued)
a 90
F innie
1100-0 280- 10
e t a l.
alum inum 3 5 0 Jim 78
(1979,
SiC 90
p36)
Im pact
a 90
A ngle rree-
1.58 mm f a i l i n g
R ickerby
coarse d ia . p article 10
g rain ed WC-
&
im pact 50
M acM illan
alum inum Swt%CO e d b y 90
(1982)
s p h e r e s m oving
soecim er a 90
several Levy
kinds of m esh coal- 10 15 & Yau
ste e l & 120 kerosene (1984,
tain less coal slurry 30 90 v98)
ste el
a 90
23 &
w ind S hayler
50 Jim 130 20
AISI 303 tunnel & Yee
ash
ste el 4 sand (1984,
p a r ti 310 90
b la ste r v98)
cles
a 90
several
1 2 0 Ji m
kinds of Shida 4
s ilic a Argon 15
steels Fujikaw a
p a r ti 120
at 300 gas (1985)
cles 90
650°C a 90
M orrison
37-270 4
304 Jim 10
S catter-
s t a in l e s s alum ina 100 good
steel p a r ti 90
(1986,
cles v l l l )
90
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2.2 E rosion param etric stu dies (continued)
pure
alum inum
a 90
hot
ro lled 5wt%
1020 150
steel g rit Lin &
quartz 10
Im pact 47.1 Shao
w ater
sand &
Angle (1991,
w ater 90
vl41)
high slu rry
chrom ium
cast
iro n
mullite
&
glass
3 mm
2.83
Al 6061- glass
TO w o r k shot
hardened 2.3 mm
steel 2.80
surface
impact shot
velo 20
3 mm
city Al glass 2.52
120 velocity Sheldon
gas
6061-TO shot exponent
gun &
annealed 2.3 mm 363 Kanhere
surface steel 2 .34
(1972)
shot
Al 3 mm
6061-TO glass 2.41
work shot
hardened 90
& 2.3 mm
steel :.19
an n ealed
surface shot
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2.2 E rosion param etric stu dies (continued)
H46
25 &
11% 61 T illy
125 p m 90 v el. exp. = 2.3
chrom ium (1973)
q u artz
steel 610
steel
im pact 70 2.5
g rit
v elo f r e e
c ity 1.58 mm
p olycry f a l l l i n g
R ickerby
WC- p a r t i c l e
10
s ta llin e &
6w t % C o 90 v el. exp. = 3.3
im p a c te d
alum inum M acM illan
spheres b y m o v in g 200
s p e c im e n
(1980)
0) 70 30 4.07 Brown
p olycry to n
61- & et a l.
to <D p ir 90 3.07
sta llin e G a ir v el. exp.
it)
r~ 1 Q. 200 122 30 2.7 (1981,
a-Fe tjl CO
v70)
p m 90 2.86
f r e e
R ickerby
1.58 mm f a l l l i n g
30 2.98
coarse 10 &
p a r t i c l e
g rain WC- M acM illan
im p a c te d
alum inum
6w t % C o 150 (1982,
D y m o v in g
90 3.3
spheres v 7 9)
s p e c im e n
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2.2 E rosion param etric stu d ies (continued)
R decreases as
qu artz
82 45 the value of CpAl
vario u s sand
in c re a se s. H utchings
pure
(C: sp ecific heat (1975,
m etals SiC
p: d en sity , AT: v35)
p a r ti 136 20 tem p, rise above
cle the m eltin g D t.)
plain car
bon steel 2 5 0 |xm
R slig h tly G reen
under diff SiC
erent heat a ir 158 30 increases w ith e t a l.
p a r ti
treatm ent hard ness. (1981)
cle
conditions
S p ecific erosion
76
various energy is propor
SiC 20 M alkin
target pure 137 tio n a l to the sp
(1981)
m at m etals e c ific m elting
e ria l q u artz 82 45 energy.
prop AISI 4140
steel R increases w ith
e rtie s
in creasin g hard A m brosini
under a ir
1 25 flm &
various b la s t 50 ness & ultim ate
SiC 30
heat stren g th & decr Bahadur
ing
treatm ent easin g d u c tility . (1987)
conditions
a) The ag in g or
w ork-hardening
several treatm en t has no
70-200 e ffe c t on R of
abrasion
|1 m a ir h i. tem p, allo y s;
re sistan t 30,
q u artz, N inham
allo y s & b la s t 60 b) D ifferent hi.
60, (1988)
SiC ing tem p, allo y s have
high tem - 90
p a r ti only sm all d if f
D erature
c le erence of R;
a llo y s
c) A b rasiv e-resis
tan t allo y s exhi
b it higher R than
h i. tem p, allo y s.
A new p a r a m e t e r t|
S rosion E fficiency t
ls d efin ed as the
ratio of the vol- Sundara-
ime rem oval to the ra jan
any any
vary 90 ;olum e d i s p l a c e d . e t a l .
kind kind
q = 2EvH /u2 (1990,
E : vol. lo s/u n it v!40)
V
m ass of erodent
H: hardness
u: im pact v el.)
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2 .2 E rosion param etric stu d ies ( c o n tin ued)
U=305
T illy
H46 0-200 R
128, //^ u ^ 2 4 4 (1973)
11% pm 244, 90
chrom ium q u artz 305
ste e l 0 200
d ( p m)
erod
gold,
ent M isra &
copper, SiC R
size Finnie
copper abra 120 20
(1981)
single sive
cry stal 0 250
d (pm)
M orrison
37, 63,
304 &
130,
10 R S catter-
sta in 270 pm
100 good
less alum ina
90 (1986,
ste el p a rti
0 300 v lll)
cle
d (pm)
5wt%
hot 150 10 R L in &
q u artz
ro lled g rit Shao
sand- 47 . 1
1020 quartz 90 (1991)
w ater
ste el sand
slu rry 0 300
d (Um)
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Im pact
E rosion T arget E rosion E quation
E rodent ^ngle u Com m ents R eference
M echanism s M aterial ( d e g r e e
(R: s i n g l e p a r t i c l e m a t e r i a l r e m o v a l'
1) For a s a 0 ,
p _ m (v s in a - K )
2e W ell c o rre la te d w ith
2 m C ( v s in a - K ) 2 C (v sin a - K )2 , t h e ero sion trend s of
+ ------------ ---------------------------- ---(vcos a — -----------------------------—
both d u c tile and
vvsin a vvsin a
cu ttin g b r i t t l e m ate ria ls.
wear d u c tile 2) For B itte r
& 0 <x > cxq, The ero sion ra te of
& (1963,
CO d efo rm atio n e la s tic p m ( v s in a - K ) 2 d u c tile m aterials is
oo b r ittle p5 & pl69)
wear sphere 90 2 e
peaked a t a low
m aterials 3 in cid en ce an g le w hile
m (v 2c o s 2 cx - K i ( v s i n cx - K ) t . th a t of b r ittle
+
m a te ria ls is peaked
E: sp e c ific energy of defoxrm ation a t norm al in cidence.
Im pact
E rosion T arget E rosion E quatio n
E rodent A ngle u Com m ents R eference
M echanism s M aterial (R: single p a rtic le m aterial rem oval)
(deg ree
tw o-stage
erosion
R = e , m ( - V )2 [1 - ( ^ r . ) o K ]2 + r 2 m F j , v( ^ ) 2 W e l l c o r r e l a t e d w ith
d u c tile fra g ile experim ental d ata in T illy
due to 90
m aterials p a rtic le tl and E2 : sp e c ific energy of th e term s of p a rtic le (1973)
p a rtic le
1s t & 2nd stage ero sio n . F d,v: size and im pact
fragm enta
p a rtic le frag m en tatio n co n stan t, v e lo c ity .
tio n
d th r: th resh o ld d. V r: te s t v elo city
R = (sin 2 a - 4sin 2 a ) + l ^ l l Y 2s i n l a
4 r p (o3
w here /ap|> K |b The o v e rall v e lo c ity
irre g u la r
0) = V m e x p o n e n t was found
to be g re a te r than 2. F innie &
cu ttin g d u c tile
sm all T his ex pression is only v a lid fo r
m aterials rig id E xperim ental v e r if i MaFadden
the case of non-zero h o rizo n tal
abrasive c atio n was not (1978)
v e lo c ity component as th e p a r tic le
p ro v id ed .
e x its th e su rface. For th e o th er
case, a m ore c o m p lic a te d f o r m u la t io n
can be found in th e o rig in a l paper.
R _ Ae 2 « rm p v 2
er C ay E xperim ental v e r if i
low cycle irre g u la r A lp: p la s tic s tra in in tro d u ced per H u tchings
d u ctle c atio n was not
rig id 90 (1979)
fatiq u e p ro v id ed .
m etals cy cle. Ef: fa ilu re s tra in . H r: ratic
p a rtic le
of deform ed volum e to in d en ta tio n
T able 2 .3 M aterial rem oval m odeling of d u c tile m ate ria ls (co n tin u ed )
Im pact
E rosion T arget E ro sio n E quation
E rodent 'V n g le a Com m ents R eference
M echanism s M aterial (R: single p a rtic le m aterial rem oval)
(degree
0 .0065m v2-5pp 25
R ~ -------------------- -—
p la te le t r tO.75i_jO.25
' -' P 1 111 IIs W ell c o r r e l a t e d w ith
form ation rig id
m etals 90 Tm: t a r g e l t a r g e t t h e r m o p h y s i c a l 3 u n d a r a r a j a r
Cp: t a r g e t s p e c i f i c h e a t .
a t c r i t i p a rtic le m e ltin g p o i n t. Hs: s t a t i c hardness. p ro p e rtie s . (1983)
cal s tra in
T h is i s a s i m p lif ie d e q u a tio n . See
th e o r i g in a l p a p e r fo r com plete
fo rm u latio n .
The liq u id drag force is considered
irre g u la r in ad d itio n to the ta rg e t m aterial
rig id r e s i s t a n t f o r c e in th e m o tio n eq u - E xperim ental B en ch aita
d u c tile t i o n o f th e p a r t i c l e . The d e r iv e d
c u ttin g p a rtic le sm all v e rific a tio n was e t a l .
m etals fo r m u la tio n f o r t h e m a t e r i a l rem oval
o in a not provided for (1983)
liq u id re q u ire s n u m erical e v a lu a tio n . In
ero sio n r a te . B en ch aita
stream t h i s m o d e l , R cc v2-5 . B u t t h e e f f e c t
(1985)
of p a r tic le siz e is not in cluded.
a) P a rtic le e x its m aterial (txcixo)
m ( ) n ° - 25 v 2-5 ______
R = 0 .564( — ----------------)sin 2uv,sin a
a rL 25 R r a 7 5
b) P a rtic le h o rizo n tal v e lo c ity Y |,= ()
w hile th e p a rtic le is p ro p elled E xperim ental v e r if i
irre g u la r 0 H ashish
c u ttin g d u c tile c atio n was not
rig id outw ard (ao<cx<ai) (1987)
m etals p ro v id ed .
p a rtic le 90 r = _my2. cos 2 a ( i + R « - a o Um2 (x)
2Rsa f «i-cx0
c ) Vh = 0 w hile th e p a rtic le is s t i l l
T able 2.3 M aterial rem oval m odeling of d u c tile m ate ria ls (con tin u ed )
p _ ^ 2 llC m v 2 s i n 2 a F ( t ) j f j ]
p la te le t nCp
W ell c o rre la te d w ith
form at!on rig id j ( n + l ) ( n / n c ) ( 2 - n / n c) c? Sundararajar
m etals 90 ta rg e t therm ophysical
a t c r i t i p a rtic le (1983)
p ro p e rtie s .
cal stra in 4(l+A .)tan2a F ( t )
T able 2 .3 M aterial rem oval m odeling of d u c tile m ate ria ls (co n tin u ed )
Empact
E rosion T arget E rosion E quation
E rodent \n g le a Com m ents R eference
M echanism s M aterial (R: single p article m aterial rem oval)
;d e g r e e
0 .2nCmv2sin2a F (l). , .
R = ( ~ ){1 +
nCp T his model is ab le to
( n + l ) ( ) i / p c) ( 2 - p / p c) ex p lain th e ero sio n -
L | - e2} a n g le r e l a t i o n , w hich
p la te le t irre g u la r 4(l+X )lan2cxF(t) is stron g ly dependent
form ation or 0 n: strain -h ard en in g exponent
m etals on p. & X . T h e p r e d i c
a t c r i t i sp h erical F (t ): num erical c o n sta n t
ted v e lo c ity is g re a S undararajar
c al stra in p a rtic le 90 C: a can stan t c h a ra c te riz in g
te r than 2. I t supp (1991)
th e tem perature dependence
o rts the ob serv atio n s
of flow s tre ss.
t h a t ( 1) the lin e a r
|.l: c o e ffic ie n t of fric tio n re la tio n bew teen R &
hardness; (2) R i s
|.lc: c r itic a l c o e ffic ie n t of
to in se n sitiv e to ta rg e t
fric tio n
stren g th .
X: p a rtic le shape c o e ffic ie n t
e: c o e ffic ie n t of r e s titu tio n
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1000 |im g l
h.p - S i 3 N4 500 No frac tu re.
ass sphere
d)
i—1
4^ T? 3 mm d i a .
4^ CO 0 9 4% glass
co n ical, G 70 K irchner
0 rH alum ina sphere
rad ial, ■P Q, H ertzian cone cracks. &
r\j (X
& la te ra l 90 G ruver
crack Z CO
h .p . 1 .59 mm (1978,
S to
sysLem 0 S i 3N 4 dia. WC 50 p25)
o M
o p sphere
«H (0
In terg ra n u la r R itte r
AD 99.5 1 0 1 6 Ji m slin g er ra d ia l cracks
90 crack in g . et a l .
SiC type Large g rain
In terg ran u lar (1987)
CR6 & abrasive => in te rg ra
106 & som e ra d ia l
CR30 nular cracks
crack in a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
AD 9 9
alum ina
1100-0
R ip p les.
alum inum
to o l ste e l
(Rc 64)
m olded
120 m esh B rittle erosion.
g rap h ite
SiC No r ip p le s .
96.5%
p a rtic le
rip p le MgO
MgO p a rtic le
99.5% No rip p le s .
alum ina
m olded
g rap h ite
P la stic i n d e n t a ti o n accom
Wada &
m ixed mode 7 1 0 |im p an y in g l a t e r a l crack in g ,
alum ina a ir SEM W atanabe
damage SiC 300 80 c ru sh in g and in te rg ran u la r
(1987)
abrasive crack in g .
T able 2 .5 P aram etric studies of B rittle M aterials
to o l
ste e l
Rc 64
high
d en sity
MgO
1 2 7 [izn
(96.5%)
im pact
angle S heldon
9[i;n
10
m olded |I m a ir 152
F innie
g rap h ite an g u lar 90 (1966,
SiC
127 [1/7? p387)
high
d en sity
alum ina
(99.5%)
p late
glass
210
108
[1 m
297
glass 96.3
|l m
N eilson
475
79.2 40-90
)I m
G rich rist
a ir
(1968)
297 [Xm
g rap h ite 128
alum ina
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2.5 P aram etric studies of B ritt le M aterials (cont;in u e d )
N eilson
20 &
210 pm R
a ir 128 G ilch rist
perspex alum ina
90 (1968)
3 a 90
60-125
glass, pm 10 T illy
R
fib re - irreg u a ir (1969)
104
glass la r 90
guartz
im pact 0 a 90
a n g le ■
30 p m
an g u lar R
65
alum ina
co rn in g Sargent
20 0 a 90
7740 et a l .
a ir
pyrex (1979)
90
glass
10 p m
R
an g u lar 84
alum ina
0 a 90
reactio n 10 R outbort
23 p m slin g er R
bonded ang u lar type 151 et a l .
SiC 90 (1980)
alum ina
0 « 90
15 M orrison
270 pm slin g er R
m u llite 80 e t a l .
alum ina type
90 (1985)
0 « 90
0 a 90
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2 .5 P aram etric stu d ies of B rittle M aterials (continued)
590 - 30
alum ina 7 6 0 urn
(KP-95) an g u lar 90
g lass Wada &
a 90
W atanabe
im pact
angle (1987)
300
500 -
20
8 5 0 Jim
S i 3N4 alum ina
90
( 9S)
a 90
glass 3.0
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 2 . 5 Parametric studies of Brittle Materials (continued)
o 2 .3
a> 25 C
u
•H 500
JZ 2.4
a
a
fO
CO 1000 3.3
°C
2.3
2 5 °C
a> (0 500
u c 2.8
a) -h
.u £
♦aH Hd
1000 2.7
CO 03 W ieder-
° C 1 5 0 |I m
37 horn &
co SiC air
o 2.3 Hockey
im pact o 25 C parti stream 90 vel. exp.
co (1983,
velo co cle 94
CD (0 500 2.1 p 7 66)
city
4-> D 2.3
O *H 1 0 0 0
.C <0 ° C
2.9
2 5 °C
c
o 500
o 3.3
•H °C
r-H
•H
CO 1000 3.4
°C
H .P. o 1.8
SiC 25 C
V o 2.1
<D 25 C
CO
CO
(I) 500 2.5
u
a
n
4-)
1000 2.4
O -H
JZ CO
°C
37 15 2 . 84 M orrison
p. m slin g er 31-90 2.80 et a l .
60- vel. exp.
m u llite 15 2 . 97
270 type 100 (1985)
dm 31-90 2 .23
A l20 ; 2.9
Shipway
a ir 2.7 &
sin tere d SiC 40-
b la s t 90 ve x . exp. Hutchings
b4 c silica
120 9.4
ing (1991)
600-850 2.8
JU- 1:
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able 2.5 P aram etric stu d ies of B rittle M aterials (continued)
glass 4.25
21. 6 & 3 . 95
MgO
.0
a oAO
u 91
g rap h ite Jim 3.78
Sheldon
ang u lar 152
steel,Rc64 erodent 3.58 £
SiC air 90 diam eter
alum ina Finnie
3.86
exponent (1966,
glass 282 - 5.12
45.7 p 3 93)
940 pm
MgO - 3.3 9
ste e l
76.2
g rap h ite shot 3.1 4
55
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Im pact
E rosion T arget E rosion E quation
E rodent 'in g le a Com m ents R eference
M echanism s M aterial (R: single p article m aterial rem oval)
[degree
a) For sp h e rica l p a rtic le
( 3 m \ /2.4m\
R — k q l p m _2 V m _2
sp h erical sm all
Some c o rre la tio n fo r
H ertzian or b) For an g u lar p a rtic le S heldon
b r ittle the v e lo c ity exponent
cone an g u lar /3.6m-. / 2.4m-, & F in n ie
m aterials and p a rtic le rad iu s
crack in g rig id R = k 5 f p ( m- 2 >v( ni-2 * (1966)
ex p o n en t.
p a rtic le
m: ta r g e t m a te ria l flaw
d is tr ib u tio n param eter.
K4, K5: co n stan ts.
T h is equation was
r p‘l G 0 l V i | , l, 1-2
d eriv ed fo r low
Evans &
la te ra l K d 1 5 H d 0 *5 v e lo c ity erosion.
Cn b r ittle so lid W ilshaw
Oi cracking 90 Lack of s u ffic ie n t
m aterials p a rtic le G: shear m odulus of ta rg e t (1976)
experim ental
Kd: dynam ic fra c tu re tou g h n ess
v e rific a tio n .
Hd: dynam ic hardness
r 3 . 7 v 3 . 2 n 0.25
R oc FP V PP
la te ra l Some c o rre la tio n Evans
b r ittle so lid
90 K i: 1 3 II.s0 2 5
w ith experim ental e t a l .
crack in g m aterials p a rtic le
Kc: fra c tu re toughness d a ta . (1978)
Hs: q u a s i-s ta tic hardness
Im pact
E rosion T arget E ro sio n E quation
E rodent ^ngle a Com m ents R eference
M echanism s M aterial (R: single p a rtic le m aterial rem oval)
(degree
In teractio n
C o m p licated fo rm u latio n was o b tain
of H ertzian
ed. See the o rig in a l paper fo r
cone cracks sph erical M ehrotra
b r ittle 90 d e ta ils . A sim p lified equation is : Poor c o rre la tio n
due to t r i - rig id e t a l .
m aterials fo r n.
p let(3 cone p a rtic le R oc ( v 4 ' 5 - v04/5)n ( 1979)
cracks)
Vo: th resh o ld p a rtic le v e lo c ity
form ation
n: co n stan t
la te ra l
crack in g p p D°-22Ea 6 7 v2-4d0-67
R = (l - f)Y4 P
plus
h 0 56k 1c>-33
sp allin g Johansson
+ l z P ii” ME v -d In su ffic ie n t
due to b r i tt l e so lid
experim ental e t a l .
m ate ria ls p a rtic le 90
co n ical (1987)
I10 3 3 K ic2 v e r if ic a tio n .
crack
in te ra c f: fra c tio n
tio n Y4, Z4: g e o m e t r i c a l and s ta tis tic a l
facto rs
2.3 Sum m ary
The erosion m echanism s of ductile materials have been classified into seven cat
of cutting, the im pacting particle displaces, along its trajectory, a portion of target
m aterial into a chip by shearing. This mechanism is usually associated with erodents
with sharp edges and relatively small im pact angles. For erodents with rounded edges
or large im pact angles, an extrusion mechanism is usually induced, where the target
m aterial is displaced to form a lip (or ridge). These lips (or ridges) remain attached
to the target surface, but are vulnerable to subsequent im pacts. In some cases, even
sharp erodents im pacting at small angles may result in ridges instead of chips (e.g.,
Tilly, 1969) due to the particle orientation. Impact by the flat face of a particle causes
little dam age (e.g., Smeltzer et al., 1970). A particle oriented with a large negative
rake angle tends to produce a lip while one with a positive or less negative rake angle
tends to produce a chip (e.g., H utchings, 1977). Therefore, both cutting and extru
sion m echanism s often exist in a m ultiple particle erosion process. On the other hand,
the extrusion m echanism observed for single particle im pacts often transforms into
particle erosion, especially at large im pact angles. This is because overlapped impacts
produce highly strained regions where thin platelets spall off from the target surface
when a critical strain is reached (e.g. Rickerby & M acM illan, 1980). Some less signifi
cant m aterial removal m echanism s include secondary erosion, m elting and splattering
and particle em bedm ent. Secondary erosion m echanism is due to th e secondary im
pacts of particle fragments (e.g., Tilly, 1973). This mechanism becomes significant
when fragile and large size particles are used for erodent. A m elting and splattering
m echanism is caused by the local high temperature in the im m ediate impact zone.
58
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
where the target and/or the particle material is m olten and m olten debris are ejected
(e.g. Smeltzer et al., 1970). Particle embedment produces m aterial removal only
when the em bedded particle is subsequently detached so that some bonded material
is removed (e.g., Smeltzer et al., 1970) or when the em brittled surface layer by the
particle em bedm ent is eventually fractured away (e.g., Ives & Ruff, 1979). Ripple for
m ation is a very im portant erosion phenomenon. Ripples are often formed when the
erosion conditions include a relatively soft target m aterial (e.g., Stringer & Wright,
1987), a relatively small impact angle (e.g., Reddy & Sundararajan, 1986, v l l l ) or
a large stream impact angle with tangential flow com ponent (e.g., Rao et al., 1983).
A typical m aterial removal m echanism in a ripple form ation process is the forma
tion and detachm ent of macro-flakes (torques) at the ripple crests (e.g., Brown et al.,
Parametric studies of ductile materials have been focused on the effects of these
five erosion parameters: impact angle, impact velocity, target m aterial properties,
erodent size, and erodent hardness. Various studies on ductile m aterial erosion have
shown a consistent trend in which the variation of erosion rate versus impact angle
is peaked at a relatively small impact angle (e.g., Finnie, 1658). This phenomenon
ductility of the target m aterial decreases, the angle associated with the peak erosion
rate shifts toward 90° (e.g., B itter, 1963, pl69; Lin & Shao, 1991). The effect of impact
T he values of the experim ental velocity exponent scatter within the range of 1.5 - 4.0.
The data collected in Table 2.2 give a mean value of 2.46 w ith a standard deviation
of 0.7. Different aspects of the effects of target m aterial properties on the erosion rate
have also been studied. Some general trends are that (1) the erosion rate increases
59
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
with increasing target material hardness and decreasing ductility (e.g., Ambrosini &
Bahadur, 1987); (2) the erosion rate decreases as the value of the product C p A T (C:
specific heat, p: density and AT: tem perature rise above the m elting point) increases
(e.g., H utchings, 1975). A new parameter “Erosion Efficiency” , defined as the ratio
of the volum e removal to the volum e displaced, has been proposed by Sundararajan
et al. (1990, v l4 0 ) to account for the overall erosion resistance of the target material.
Studies on the effect of erodent size have revealed that the erosion rate increases with
increasing erodent size and there probably exists a transition erodent size at about
100 pm (e.g., Tilly, 1973). The trend of the erosion rate as a function of the erodent
hardness also exhibits a transition point where the ratio of target/erodent hardness
approaches unity and the erosion rate increases dram atically w ith increasing erodent
hardness.
The m aterial removal models listed in Table 2.3 can be classified into two cat
egories: cutting m odel and deformation wear m odel (failure at critical strain). All
the cutting m odels have been originated from or similar to F in n ie’s m odel (Finnie,
1958). The concept of deformation wear was first used by B itter (1963). Its physical
meaning was clarified in the subsequent modeling studies (e.g., H utchings, 1981; Sun
dararajan, 1983). The cutting m odels are associated with sm all im pact angles while
the deformation wear models are associated with large im pact angles (m ost of them
with 90°). B itter’s model (1963) combined both the cutting and deformation wear to
give a better representation of the erosion trend for the im pact angle varying from 0
to 90°. All of these models give the values of the velocity exponent between 2 and
3. Target m aterial properties were included in the cutting m odels in terms of plastic
flow stress. In the deformation wear models, they are included in terms of specific
energy (e.g., B itter, 1963), hardness (e.g., Sheldon and Kanhere, 1972), critical strain
(e.g., Hutchings, 1981), and thermal properties (e.g., Sundararajan, 1983). Erodent
geometric parameters and erodent material properties were included in some models
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(e.g., H ashish, 1987), but neglected in the others (e.g., Finnie, 1958).
The erosion process of brittle materials is m ore com plex than that of ductile ma
terials. The erosion m echanism s observed in the existing brittle m aterial erosion
studies have been classified into seven categories: (1) conical, radial and lateral crack
system , (2) intergranular and transgranular cracking, (3) ring fracture, (4) microchip-
ping, (5) plastic deform ation and m elting, (6) ripple form ation, and (7) m ixed mode
dam age. The conical, radial and lateral crack system is the m ost commonly ob
served im pact dam ages on brittle m aterials. Its general appearance usually displays
a central plastic impression surrounded by conical, radial and lateral cracks though
some o f these dam ages are som etim es missed. It has been revealed that the conical
and radial cracks form during the loading period of th e p article/target interaction
while the lateral cracks form during the unloading period (e.g., Chaudhri and Walley,
1978). M aterial rem oval is usually due to lateral cracking. Such damages have been
primarily observed for normal (90°) im pacts. The m ajority of the target materials
used in these tests are glass or fine grain ceramics. Various erodent materials, sizes,
and types (angular or spherical) have been used in these tests. The impact velocity
was ranged from 37 m /s to 1000 m /s. Intergranular and transgranular cracks have
been observed on ceramics, especially coarse grain alum ina ceram ics (e.g., Ritter et
al., 1984). T hey have been observed for both large and sm all im pact angles. Ring
fracture dam age includes a m ain ring crack and m any short circum ferential fractures
(Bow den and Field, 1964). This type of damage is associated with impacts by softer
erodents. P lastic deformation has also been observed in m any im pact events of brit
tle m aterials, involving both angular and rounded erodents. Ripple phenomenon was
absent in m ost erosion tests of brittle materials, but an exception was found on a
silica glass sam ple eroded by 1000 mesh SiC particle (Sheldon and Finnie, 1966).
E xisting param etric studies of brittle materials have revealed that the erosion rate
of brittle materials tends to peak at a large im pact angle. However, this behavior may
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
change, under appropriate erosion conditions, into a typical ductile erosion behavior,
i.e., the erosion rate peaks at a small impact angle. Sheldon and Finnie (1966) have
provided such evidence, e.g., the erosion rate of a glass sample peaks at almost 901'
when using 127 iim SiC particles, but when 9 /im SiC particles were used, the peak
occurs at about 25°. In some cases, twin peaks have been observed on the erosion rate -
im pact angle curve (W ada and W atanabe, 1987). This m ay be due to a combination
of ductile and brittle erosion behavior. The velocity exponents listed in Table 2.5
give a m ean value of 2.66 with a standard deviation of 0.63. Very few parametric
studies have been conducted to study the effects of target m aterial properties of brittle
m aterials. In one recent study, Srinivasan and Scattergood (1988) disclosed that the
erosion rate increases as the target hardness decreases. The effect of erodent size has
and Finnie (1966, p393) give the mean value of the erodent diam eter exponents to be
3.45 with a standard deviation of 1.33. Other investigations have presented smaller
values of the erodent diameter exponents. The effect of erodent properties has been
studied by W ada (1991). Wada concluded that the erosion rate is proportional to the
The brittle m aterial erosion models can be roughly classified into three types:
conical crack m odel, lateral crack model, and intergranular crack m odel. The Hertzian
cone crack m odel of Sheldon and Finnie (1966) was derived by evaluating the volume
of a spherical cap which is fractured by many Hertzian cone cracks. The critical step
in deriving the lateral crack models is to determine the depth of the lateral fracture
and the impression radius (e.g., Evans et al., 1978). The intergranular crack model of
62
with permission o f the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3
a rapid erosion process. Although the information derived from the literature review
covers a board range o f erosion conditions and m aterials, no one has system atically
investigated the erosion process involved in a high velocity (up to 600 m /s ) abrasive
The experim ents were conducted in the University of Rhode Island Waterjet Lab
oratory. A high pressure abrasive waterjet was generated with a 9X W A TER N IFEr'u
dual intensifier high pressure pump and a PA SER TA/ abrasive waterjet system . An
ASEA industrial robot was used to perform the traverse m otion. A 9 mm thick
sintered alum ina (A D 99.5) plate and a stainless steel (304) plate were used as the
target m aterials, respectively representing brittle and ductile m aterials. The erodent
particle was garnet abrasive (Barton Mine) of m esh No. 50 and 80, which has an
average diameters of 300 and 180 /tm, respectively. Three different erosion conditions
(sweeping, grooving and cutting) were used to generate the eroded surfaces. In the
sweeping and grooving tests two different configurations of the jet impingement an
gles, 20° and 90° with respect to the target surface, were used as shown in Figure
3.1. In the cutting tests, the jet was set perpendicular to the sample surface (Figure
3.2). The cutting process was instantaneously interrupted to retain the cutting front,
which appears as a vertical groove on the side of the sam ple. The parameter settings
for these three erosion conditions are listed in Table 3.1. The differences of traverse
63
with permission o f the copyright owner. Further reproduction prohibited without permission.
AWJ Nozzle
T raverse
d irection
Workpiece
Figure 3.1 Experimental configurations used in sweeping and grooving erosion tests.
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 3.1 Param eter settings of erosion tests.
* see A ppendix A.
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
speed and abrasive flow rate should be noticed for these three erosion conditions. The
eroded sam ples were coated with carbon and then exam ined w ith a scanning electron
microscope.
the m aterial removal process. Figures 3.4 and 3.5 show two typical craters by cutting
actions. A partially enlarged view of the crater in Figure 3.5 (a) is shown in Figure
3.5 (b), indicating clear evidence of plastic flow and m elting (notice the string-like
frozen m etal). Sm all lips are seen on the side edges of the crater shown in Figure
3.4, im plicating a plowing action. However, absence of lips at the exit end of the
crater strongly suggests that majority of the crater m aterial is removed by a cutting
action. Figure 3.4 also shows som e micro-craters distributed around the main crater,
abrasive fragm ents. However, it appears that these micro-craters make negligible
contribution to the global m aterial removal as compared to that of the main crater.
Some abrasive fragm ents are evident in Figure 3.5. T hey adhere to the surface, but
not em bedded. Figure 3.6 shows two partially overlapped craters. Existence of the
the first crater (on the left) does not change the cutting action of the second im pact.
The upper portion of Figure 3.7 and an enlarged view in Figure 3.8 show a raised lip
at the exit end of the crater, indicating existence of an extrusion m echanism , probably
caused by a rounded face of an abrasive particle. T he center of Figure 3.7 and its
enlarged view (Figure 3.9) show a m essy crater. Its dam age m echanism s are probably
66
with permission o f the copyright owner. Further reproduction prohibited without permission
Figure 3.3 A cutting mechanism dom inates the m aterial removal process of SS 304.
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
00
Figure 3.5 A typical crater observed on the SS 304 sample, ((a) x 125 and its enlarged
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3.6 Two partially overlapped craters observed on the SS 304 sam ple,
Figure 3.7 A raised lip at the exit end of the crater observed on the SS 304 sam ple.
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3.8 An enlarged view of the raised lip shown in Figure 3.7.
Figure 3.9 An enlarged view of the m essy crater shown in the center of Figure 3.7.
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SS 304 Sample at 90° Impact Angle
Figure 3.10 shows that m ajority of the craters were created by an indentation
m aterial loss in the form of flaking (deformation wear) when the critical strain is
reached. T he upper left portion of Figure 3.10 and Figure 3.12 show another inden
tation crater. Some grooving traces remaining on the left hand side of the crater
evidences inward m otion of the im pacting particle. On the right hand side, it can be
seen that a large portion of m etal has been raised. In some situations, such raised
portions might have been fractured away (See the two craters at the lower portion of
Figure 3.10, also Figure 3.13). An enlarged view of area A in Figure 3.13 shows some
dim ension. Particle em bedm ent is a very common occurance. Figure 3.15 shows an
exam ple. Besides, Figure 3.15 shows that the edge of the crater has been fractured
A typical crater is seen in Figure 3.16. It appears that the particle was sliding on
the surface. Many fragments are observed around the crater. This m ay be traced back
to the specimen preparation process which includes grinding and polishing. These
processes tends to yield a top layer of the target surface with some crushed grains.
The im pact of the particle caused loosening of these crushed grains. Some grain
boundary cracking also accompanies the im pact. An interesting feature of this crater
is the feather-like surface marks forming an array on both sides of the crater with a
unique angle. A closer look of these surface marks (Figure 3.16 (b )) suggests that
they were caused by plastic flow. Shown in Figure 3.17 (a) and (b) are another impact
site and its enlarged view, respectively. Surface fragments are observed. They may
71
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3.10 M ajority of the craters on the SS 304 sample were created by an inden
Figure 3.11 A typical indentation crater observed on the SS 304 sam ple,
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
..
Figure 3.12 An indentation crater on the SS 304 sam ple shows inward grooving traces
Figure 3.13 An indentation crater on the SS 304 sample shows evidence of ductile
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3.14 A plastic impression on the SS 304 sam ple due to a longitudinal im pact,
Figure 3.15 An indentation crater on the SS 304 sample with particle fragments
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(b)
Figure 3.16 Overall appearance (a, x 490) and a partially enlarged view (b, x 2500)
of a typical crater observed on the AD 99.5 sample, (im pact angle = 20°)
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
.
IVW
(b)
Figure 3.17 Overall appearance (a, x 240) and a partially enlarged view (b, x 970) of
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
be caused by an erasing action, in which the im pact particle slides over the surface,
loosening and rolling the crushed grains on the top layer of the surface. Since this
sample is covered with a top layer of crushed grains due to the polishing process
in sam ple preparation, the observations here m ay not be true for the body material
An impact crater is shown in Figure 3.18. A piece of m aterial was dug out, which
is com posed of small or crushed grains. At the bottom of the crater, as shown in
Figure 3.18 (b), clear evidences of plastic flow and m elting are observed. Another
such exam ple is shown in Figure 3.19. Figures 3.20 (a) and (b) show two other impact
sites, where plastic flow is not observed, but pieces of small (or crushed) grains are
still evident. Similarly, the small (or smashed) grain phenom enon observed on this
sample m ay not be applicable to the body m aterial erosion. However, the information
Grooving tests were done on AD 99.5 samples only. A traverse speed much lower
than that in sweeping tests was used. A shallow groove, as shown in Figure 3.21 (a)
and (b), was formed on the eroded sample. The general m orphology of the eroded
surface is shown in Figure 3.22 (a) and (b). The “rock-eandy” appearance suggests
that the grain boundary cracking is the failure m echanism (T he cracks appearing on
the grain facets in Figure 3.22 (b) are cracks on the carbon coating). Also shown is
some m essy material, which is probably deformed grains. Individual craters are not
isolated craters and semi- craters are evident on the edge of the groove (Figures 3.23
77
with permission of the copyright owner. Further reproduction prohibited without permission.
(b)
Figure 3.18 Overall appearance (a, x 500) and a partially enlarged view (b, x 1500)
78
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
00
Figure 3.19 Overall appearance (a, x 340) and a partially enlarged view (b, x 990) of
a crater observed on the AD 99.5 sam ple, (im pact angle = 90°)
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
( b)
Figure 3.20 Two fractured craters (a) & (b) observed on the AD 99.5 sample
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(b)
Figure 3.21 Top view (a, x 33) and side view (b, x 40) of the groove created on the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(b)
Figure 3.22 General m orphology of the groove surface on the AD 99.5 sam ple,
82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(b)
Figure 3.23 Some isolated craters and semi-craters observed on the edge of the groove
of the AD 99.5 sample, (a, x 290 & b, x 860, im pact angle = 90°)
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
AD 99.5 Sample at 20° Impact Angle
Figure 3.24 (a) shows the general damage pattern on the central area of the eroded
surface, where numerous pits are randomly distributed. Figure 3.24(b) shows that the
distribution density of pits is fading towards the exit edge of the groove. Figures 3.25
(a) and (b) reveal that these pits are actually individual craters. Again, the failure
was caused by intergranular cracking. Some deformed grains are also present. The
craters have approxim ately semi-spherical shape and the num ber of grains forming a
crater is of the order of 5. It is interesting to note that these craters do not elongate
SS 304 Sample
Figures 3.26 shows the eroded surface on the top (a), m iddle (b) and bottom area
(c) of the vertical groove on the SS 304 sample. T hey uniquely dem onstrate cutting
action due to abrasive particles im pacting at a glancing angle in respect to the eroded
AD 99.5 Sample
The overall appearance of the vertical groove on the AD 99.5 sample is shown
in Figure 3.28 (a) (whole view ) and (b) (upper portion). B y scanning through the
groove, similarly, it was found that the surface m orphology appears identical for ma
jority of the groove, as evident in Figure 3.29. The erosion mechanism is primarily
tically deforming actions. Figure 3.30 (a) exhibits all of these features. Area A is
grain is seen in area B. Area C and its enlarged view in Figures 3.30 (b) and (c) show
a scratching trace, where plastic flow and m elting are evident. Figure 3.31 shows
the very top edge of the groove, where more grain facets are exposed, indicating
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(a)
(b)
Figure 3.24 General damage pattern on the centered area (a, x 160) and the exit edge
(b, x 40) of the eroded AD 99.5 sample, (im pact angle = 20°)
85
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission
(b)
Figure 3.25 Two individual craters observed on the A D 99.5 sam ple,
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(c )
Figure 3.26 The top (a), m iddle (b) and bottom (c) areas on the vertical groove of
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3.27 A chip formed by a cutting action observed on the cut SS 304 sample.
(x 500)
0 5 0k % ge& F
Figure 3.28 The vertical groove on the cut AD 99.5 sample, fa) T h e whole view (x
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(c )
Figure 3.29 The top (a), m iddle (b) and bottom (c) areas on the vertical groove of
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(b) (c)
(area B ), and scratching trace (area C) on the cut AD 99.5 sam ple, (x 500); (b, x
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29 . 0 V 49 6
Figure 3.31 The very top edge of the vertical groove of the cut AD 99.5 sample,
(x 500)
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
dom ination of intergranular cracking due to large angle im pacts.
were developed on the cutting front, which distinguish the upper zone as a steady-
state cutting zone and the lower one an unsteady-state cutting zone. To verify this
phenom enon, the AD 99.5 sample obtained under the cutting condition was cut into
halves along the cutting front. Figure 3.32 shows that steps, indicated by arrows, do
D uctile m aterials such as stainless steel 304 exhibit primarily a cutting m echanism
for low im pact angles. Other mechanisms such as plowing, extrusion, secondary
erosion by particle fragm ents occur on a minor scale. Under large impact angles,
surface dam age includes indentation, edge fracture, particle em bedm ent and some
intergranular cracking, applicable to both cases of normal im pacts and low incidence
im pacts. Transgranular cracking occurs in some individual grains, which does not
affect the dom ination of intergranular cracking m echanism . P lastic flow and m elting
The steps observed on the cutting front of the AD 99.5 sample resemble those
H ashish’s conclusion that these steps lead to large angle im pacts is questionable. To
explain this point, a typical jet/ta rg et interface is illustrated in Figure 3.33. Direct
impact of the original jet only covers the top portion of the cutting front, below
which secondary and tertiary im pacts of the deflected jet occur. Therefore, the entire
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
Direct
Im p ac t
Zone
Secondary
or
steps Tertiary
Im pact
Zone
Figure 3.32 Steps (indicated by arrows) formed on the profile of the cutting front of
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cutting front is im pacted by the jet at low incidence angles, except the very top
edge where, as the jet advances into the material, im pacts at large angles occur, as
shown in Figure 3.31. The existence of the steps slightly changes the local curvature
of the cutting front. However, impacts at these steps are clearly not due to large
angle incidence. Actually, this step phenomenon is very sim ilar in form to the ripple
phenomenon observed by Finnie and Kabil(1965, p .60) and m any others. Both the
steps and the ripples develop under low incidence erosion conditions. They both move
downstream as erosion proceeds. Although most of the ripple phenomena have been
observed for ductile m aterials, this does not eliminate the possibility of its occurrence
for brittle m aterials under appropriate conditions. For instance, according to Sheldon
and Finnie(1966), ripples were observed on silica glass eroded by 1000 mesh SiC
flow of surface m aterial(Brown & Edington (1982, v. 79), Naerheim (1985, v. 105),
Carter et al. (1980), which is obviously not the case for most brittle materials. Finnie
and I<abil(1965, p .60) proposed that ripples are initiated from a surface with random
roughness. This surface can be described by an infinite series of sinusoidal waves for
The wave with this optim um wavelength eventually grows and produces a ripple
pattern. This theory is based on a ductile material removal m odel and thus may not
Based on Figure 3.32 and the results of Hashish’s visualization experim ents, the
following hypothesis (illustrated in Figure 3.34) is developed to explain the step phe
nomenon in abrasive waterjet cutting. As the jet advances into the workpiece from
the right angle edge, a sm ooth cutting front is developed and continuously grows
in size until the rear side of the jet arrives on the edge of workpiece (Figure 3.34
(a)). During this stage, the entire cutting front is directly im pacted by the origi
nal jet. This stage corresponds to the so-called “steady-state” interface observed by
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CO
Figure 3.34 A hypothesis for the step form ation in AWJ cu ttin g processes.
H ashish(1988, p. 159). As the jet moves inward further, the area behind the rear
side of the jet is exposed to the secondary im pact by the deflected jet, resulting in a
sudden change in the material removal rate. As a result, the curvature of the cutting
front also has a sudden change (Figure 3.34 (b )). This is the initialization of a step.
Then, the step tends to grow because the inner-most layer of the jet is deflected away
by the front side of the step and becomes a protecting layer for the rest of the step.
However, as the step grows into a certain height, it is exposed to a greater flux of
erosive m edium and its top side begins to erode. B y then, a m ature and stable step is
formed. As the m aterial on the front side o f the step is rem oved, the step appears to
m ove downstream (Figure 3.34 (c)). During the process of step growth and m oving
downstream , another cycle of curvature change will start trailing the jet (Figure 3.34
(d )). However, the next step does not im m ediately follow the previous one. The
reason for this is probably due to the “gouging” effect which often occurs in a pipe
bend. Since the jet is confined by the larger curvature of the step, similar to the bend
profile, the region at the front of the step experiences an optim um erosion condition,
which elim inates the possibility of curvature change. As the step m oves downstream
for a certain distance, far enough to isolate the optim um erosion condition, another
cycle of curvature change starts and leads to the formation of another step, which
Based on the SEM observations and the analysis above, it is concluded that the
AWJ cutting condition is associated with abrasive particle im pacts at glancing angles,
regardless of the type of the target materials. M icrocutting is the sole erosion m ech
anism for ductile m aterials. Ceramic material removal is due m ainly to intergranular
cracking and plastic flow. Sequential steps are formed on the cutting front trailing the
directly im pacted region by the original jet. However, these steps do not significantly
change the im pact angles and thus the low incidence erosion m echanisms continue to
dom inate.
96
with permission o f the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4
due to intergranular cracking and plastic flow. Usually, ceramics have much higher
resistance to plastic deformation than most metals because much fewer slip planes
exist in ceramics than in m etals. However, in the event of a high velocity im pact, the
localized pressure at the im m ediate impact site reaches a very high value so that even
a ceramic will exhibit certain degree of plasticity. Furthermore, the heat generated
from initial plastic deformation is confined to the local zone because of the poor
on solid particle im pacts of brittle materials in the literature have revealed impact
damage in the form of Hertzian cone, radial and lateral cracks, which resembles the
for brittle m aterials have been derived based on quasi-static assumptions. However,
another typ e of im pact dam age has also been reported. Intergranular chipping with
out evidence of radial or lateral cracks for alumina ceramics has been observed by
Ritter et al. (1984), Ritter (1985), Ritter et al. (1986), Ritter et al. (1987, p 55-l),
Wada et al. (1988, p l l3 ) , Graham & Ball (1989), as evident in Figure 4.1. High
velocity im pacts by deformable projectiles such as lead, nylon, liquid produce a dam
age m orphology such as that shown in Figure 4.2, characteristic of many short and
discrete circum ferential fractures (Bowden & Field, 1964, Evans & Wilshaw, 1977,
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4.1 Scanning electron micrograph of isolated impact site on sintered alumina
(N o. 46 grit, 75 m /s). (A ) bar = 45 fim, (B) bar = 15 fim. (after Ritter et al., 1984)
Figure 4.2 Circumferential cracks outside a central undamaged zone produced by the
98
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Evans et al., 1978). Figure 4.3 shows orthogonal sections through ZnS targets im
pacted by W C and glass projectiles (Evans et al., 1978). Though the conical, radial
and lateral cracks are similar in form to the traditional quasi-static indentation dam
ages, the fractures are more intense and these cracks tend to intersect and form a
network. A subsurface microfracture zone is also present. Figure 4.4 (after Summers
abrasive w aterjet (directly pum ped). The appearance of the fracture surface is dif
number of discrete fractures. All these phenom ena strongly suggest the existence of
a m echanism in which discrete fractures occur sim ultaneously in the vicinity of the
impact site. T hese discrete fractures m ay eventually coalesce to form a crack network
This typ e of im pact damage resembles the dam age in an explosive loading condi
tion. Figure 4.5 (after Kolsky, 1963) shows that the fracture pattern in a “Perspex”
cone produced by detonation of 0.4 g lead azide. A surface crater was formed at
the site of detonation (the center of the cone base) while the tip of the cone was
fractured off. It has been indicated that if the specim en is large enough, the surface
crater will be the only damage. The surface crater consists of a very large number of
m inute hair-like cracks radiating from the seat of the charge. This damage has been
attributed to stress pulses. Since the duration of the stress pulses is so short, any
cracks formed will not be able to propagate, and, instead of running cracks, a large
number of discrete fractures occur and coalesce to form a continuous but irregular
Due to the similarity of the network-type im pact dam age and explosive dam age,
connection of the network-type im pact damage m echanism to the stress wave induced
fractures is natural. Elastic half-space theory has proven the existence of three types
of elastic w aves, compressive wave (P-w ave), shear wave (S-w ave) and Rayleigh wave
99
with permission of the copyright owner. Further reproduction prohibited without permission
400 um :' / ' ■ '••'o 4
i^ Y ,- i
T1
8 0 0 fim
Figure 4.3 O ptical reflected light micrographs of orthogonal sections through ZnS
targets im pacted by WC and glass projectiles, (a) The extent of the radial and lateral
fracture (parallel to the surface) for a W C projectile target; (b) the same im pact, but
highlights the formation of lateral cracks from radial cracks (e.g. at arrowed location);
(c) the exten t of the radial and lateral fracture for a glass projectile impact; (d) the
same as (c) but indicating the zone of microfracture (arrowed) beneath the center of
im pact, (after Evans et al., 1978)
Figure 4.5 Fractures in a ‘Perspex’ cone produced by 0.4 g of lead azide. (actual size)
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(R-wave) in the event of a dynam ic point loading on a half-space. The P-wave and
S-wave propagate radially outward from the source along hemi-spherical wave front
while the R-wave propagates radially outward on a cylindrical w avefront. The P-wave
is a longitudinal (or dilatational) wave and the S-wave is a transverse (or distortional)
wave. The wave w ith the highest velocity is the dilatational. Its velocity is expressed
Cl C° \ l ( l + v ) ( l - 2 v ) (4-1)
where C„ is the velocity of longitudinal waves in a thin rod and is expressed as:
Co = J - (4.2)
VP
The velocity of the distortional waves is given by (Kolsky, 1963, p60):
a‘ = ci w ^ ) ( 4 -3 )
There is no an explicit expression that describes the exact solution for the Rayleigh
= y * (4.4)
l + v
For Poisson’s ratio v = 0.2, it is obtained that C-y/Cy — 0.612, C n / C 2 = 0.912,
As they propagate outward, the waves encounter a rising volum e of material and
therefore, the energy density as well as the displacement am plitude decrease as the
travel distance (r) increases. The energy density as a function of travel distance is
called geometrical damping law, which is 1/r for P-waves or S- waves and l / y f r for
The partition of energy among these three types of waves have been computed by
Miller and Pursey (1956) and well illustrated by W oods (1968) in Figure 4.6.
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
circular footing
/v \
RAYLEIGH WAVE
horiz.
■comp.
> £ £ ^ p RESS10N
WAVE TYPE PERCENT OF
shear
TOTAL ENERGY
w indow
RAYLEIGH 67
SHEAR 26
COMPRESSION
Figure 4.6 Stress wave pattern and energy partition, (after W oods, 1968)
102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The P-wave arrives before the S-wave and R-wave. Since it passes as a compressive
pulse (p ositive), it only causes fracture under certain condition (such as when the
wave front expands rapidly causing radial tensions). In contrast, since tensile pulse
(negative) is associated with the wave front of the S-wave and R-wave, these two
Bowden Sz Field (1964), has revealed that fractures are in itiated closely behind the
materials includes two com ponents of material removal. One of them is caused by
plastic flow at the im m ediate im pact site. The m aterial removal com ponent due to
plastic flow (P),) can be evaluated with existing ductile erosion m odels. The other
stress waves. T he material removal component due to network cracking (V/) will be
evaluated w ith a crack network m odel derived in the following section. The total
V = Vp + Vf (4.5)
The stress wave energy is dissipated in two ways: fracture and internal friction.
The partition of the stress wave energy between these two dissipating mechanisms is
still not known. In this study it is assumed that a constant fraction /,„ of the stress
wave energy is converted into the fracture energy in the process of forming the crack
network. It is further assumed that the crack network is formed by separate fractures
aiong the grain boundaries, and is modeled by a cluster of cubic ceils of size a (a =
103
with permission of the copyright owner. Further reproduction prohibited without permission
grain size or flaw distribution param eter) as shown in Figure 4.7. T he total surface
2 A f = (6 a2) N (4.6)
where N is the number of grains w ithin the fractured volum e. The cubic cellular m odel
D* = <4 J )
where Vj is the fractured volum e due to a single impact w ith the cubic network model
expressed as:
Vf = a:iN (4.8)
Dc = - (4.9)
a
or
2A j = - V f (4.10)
a
The fracture energy required to form a crack surface of 2 A f can then be expressed
as:
W c = 2 A n = — Vf (4.11)
a
The total fracture surface energy is assumed to be a fraction (/,„) of the total
stress wave energy (W ). Thus, the volum e removal due to fractures is solved to be:
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SINGLE TARGET
MATERIAL CHIP
REMOVAL VOLUME
105
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
where the total stress wave energy (W ) needs to be determined. The stress wave
energy will be evaluated in the following two sections for the two cases of normal
The total energy absorbed by the target during impact can be expressed in terms of
the plastic deformation energy, stress wave energy, and residual energy. The plastic
deformation energy is dissipated in plastic work transferring into heat and stored
energy in the form of residual strain. The stress wave energy transm its into the
body of the target m aterial and is dissipated by fractures and internal friction. The
residual energy is the balance of the total input energy. It is partially dissipated
by the fragmentation of the particle and the rest becomes the kinetic energy of the
The propagation of stress waves in an isotropic elastic half space due to a point
and Pursey(1954, 1956) studied wave propagation in: (a) an infinitely long strip of
finite width vibrating normal to the free surface; (b) an infinitely long strip of finite
width vibrating parallel to the free surface and normal to the axis of the strip; (c) a
circular disc of finite radius vibrating normal to the free surface; (d) a circular disc
of finite radius rotating about its central axis on the free surface. Utilizing Miller
and Pursey’s result for case (c), Hunter(1957) derived the elastic stress wave energy
produced by a normal particle impact on a half space. H unter’s equation was modified
106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
derived the stresses and displacem ents produced by a m oving line load over an elastic
half-space. Eringen and Sam uels(1959) extended the investigation to a slightly curved
elastic half space. A three-dimensional stress wave propagation m odel was derived
by Chao(1960) for the cases of a tangential and by Chao et al. (1961) for a vertical
point load on an elastic half space. The radiation patterns induced by a horizontally
The stress wave energy derivations for normal im pact by Hunter (1957) and by
pulse of arbitrary profile P (t). The stress wave energy was equated to the work done
W = (4.13)
The expression for surface displacement u, derived by Miller and Pursey (1954),
m = j|j I I fc (4.14)
* = ( 4 - 1 5 )
where £(i/) is a function of the P oisson’s ratio (i/). Its value is 0.5324 for v = 1 /4 and
The Hertz theory of impact was applied to derive the approximate expression of
the pressure pulse for the case of a particle of mass m im pacting on an infinitely large
107
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
t(t\ - / ~™-Z0u 20 cosu;0t, |t| < t / 2 u „
m ~ I 0, |(| > ir/2w0 <U 6 >
where,
_ .15 . 2^5 1 —v 2 1 —v 2
~ i6 m ii’ g = ~ E ~ +- ^ r (4ir)
W ith Equation 4.14 replaced by Equation 4.16, the stress wave energy transm itted
( - )
PQ'C'o
where
r = 1.068=(1 + „ ) ( i - j £ ) i ( l ? ) l c ( „ ) (4.20)
Equation 4.19 was further expressed as the ratio of stress wave energy vs total
W , 2 t v ~’
. 2 = = , , . (4-21)
2m
Equation 4.21 is valid when the Love criterion is satisfied, i.e., (v j C „ y ^ < < 1 .
For im pacts at a velocity high enough to cause plastic deform ation, as pointed out
H ertz’s equation. Hutchings derived a new expression for the pressure pulse in the
case of elastic/p lastic im pact, based on the assum ption that the loading phase of the
im pact is perfectly plastic while the unloading phase is purely elastic. The half pulse
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
f ( t ) = F0sinujpt, 0 < t < tv / 2 u p (4.22)
where
“' =7vlr
r p V 2 Pp (4'24)
Since the force level must be continuous at the point of transition from loading
to unloading and, as proven by Hunter(1957), the elastic force history can be ap
proxim ately expressed by a sinusoidal function of tim e, the half pulse of the elastic
where u;R = u p/ e .
T he pulse expressions 4.22 and 4.25 were used in evaluating the integral 4.14,
yielding:
where
(4.27)
1+ e
a (e ) is a dim ensionless function of the coefficient of restitution (e) and has been
Based on Equation 4.26, the ratio of stress wave energy vs total input energy was
determined as:
4 / \ >/ w 1 + v 1 - V 2 • 47TV/6 P p ,H d 3 / A rt o \
Ar = a e ) C ( v ) — - ( - ----- — — )2 ( 4 -2 8 )
1+ e 1 - Iv L* p pp
109
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Since the im pact velocity associated with the present study is high enough to
produce plastic deformation, H utchings’ equation will be used in the m odel proposed
H utching’s equation.
tj( u ) = CM (1 + (4.29)
J L = Ar = / „ , M ( ^ ) * ( § ) t (4.30)
“ 7TIV pp L
The volum e removal due to plastic flow for normal incidence is evaluated with
v, = (4.3D
where K is the threshold impact velocity at which the elastic lim it is just reached. It
is expressed as:
( 4 -3 2 )
_ (4.33)
127 pp £/
110
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The total volum e removal for a single impact at normal incidence is obtained as:
= p , Hd a m(v - K f
27 KE ' 2e v ‘ '
AD 99.5, AD 96, AD 94, AD 90, and AD 85 are used as the target materials. The
m echanical and physical properties of these ceramics are listed in Table 4.1.
The erosion experim ents are conducted with the abrasive water jet system at the
University of Rhode Island Waterjet Lab. W ith a water pressure of 241 M Pa and
Garnet abrasive particles of mesh no. 50 (pp = 4000 k g / m 3) are entrained into the
w aterjet with a flow rate of 16.2 g /s . The velocity of the entrained abrasive particles
is estim ated to be 479 m /s (see Appendix A). The abrasive water jet exits from a
Prior to the abrasive erosion experim ents, a plain water erosion experiment is
conducted, to determ ine the effect of plain waterjet on the m aterial removal. The
same experim ental conditions as those for the abrasive erosion are set up except no
abrasive feeding. The plain water jet impinges on the specimens with a traverse
speed of 20 m m /s. The weight losses of the specim ens are determined by means of
a weighing scale w ith resolution up to 0.001 gram. The cutting tim e is calculated
by measuring the traverse speed and the cutting length on the specimen. Then the
water consum ption in the cutting period can be calculated with cutting tim e and the
water flow rate. The ratio of the specimen weight loss to the water consumption is
defined to be the erosion ratio. The resulting erosion ratios from the water erosion
experiment are listed in Table 4.1. It appears that water alone can hardly erode the
111
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.1 Properties of alumina ceramics.
M o du lus of Elasticity E (GPa) 221 276 303 303 372 386 [1]
F ra c tu r e E n e r g y Y (J/m 2 ) 1 7 .0 1 5 .8 1 8 .1 2 2 .4 2 5 .5 *
20 .1
2 0 d e g . Model 1 .2 4 6 3.0 2 1 3 .2 4 8
0.8 8 9 2 .1 8 7 0 .5 4 8
[1] Material Property Standard 990, Coors Ceramic Co., January, 1989.
[2] Bertolotti, 1973.
[3] Swanson, 1972.
[4] Barker, 1978.
112
with permission of the copyright owner. Further reproduction prohibited without permission
alumina ceramics. However, it is anticipated that water plays an im portant role in
To verify the repeatability of the erosion experim ent, two abrasive erosion ex
perim ents are conducted by varying the traverse speed. T he first experim ent uses a
traverse speed of 20 m m /s and the second, 15 m m /s. The erosion ratio is defined
as the ratio of the specim en weight loss to the abrasive consum ption. T he latter is
determined w ith the cutting tim e and the abrasive flow rate. Since, theoretically, the
erosion ratio is independent of the traverse speed, these two experim ents should give
identical results. The experim ental results, plotted in Figure 4.8, clearly indicate the
Since the m aterial removal by plastic flow is m inim al for normal im pingem ent and
for hard target m aterials, the predicted volume removal neglects the contribution by
the plastic flow. Therefore, Equation 4.33 is used to represent the predicted volume
removal.
Then the volum e removal per gram of abrasive can be expressed as:
The mass removal per gram of abrasive, i.e. the erosion ratio is:
< « • >
The results from the second experim ent are chosen to compare with the analytical
erosion m odel. Since the results of the erosion m odel contain an empirical coefficient
f, a direct comparison of the m agnitudes of results betw een the derived m odel and the
experim ent is not appropriate. Therefore, the erosion ratios from both the derived
model and the experim ent are normalized with their respective average values so that
the trends can be observed for both approaches. T he norm alized values of erosion
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o - 0 ----- EXPERIMENT 1
o
— EXPERIMENT 2
o>
O)
2 -
MODEL
4 ■a— EXPERIMENT 2
"ET"
0
AD85 AD90 AD94 AD96 AD995 AD999
Alumina Ceramics
Figure 4.9 Comparison of normalized erosion ratios for six alum ina ceramic
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ratios are plotted in Figure 4.9 against the different types of alumina ceramics.
Considering the complication of the erosion process as well as the omission of the
other minor effects such as the porosity and the plastic deform ation, the consistency
of the trends arising from the theoretical and the experim ental results is surprisingly
good. A large discrepancy occurs at the data point for A D 85. One of the reasons for
this discrepancy might be the omission of porosity in the calculation of the derived
m odel. A previous study with scanning electron microscope on a hand-m ade fracture
surface of AD85 has indicated the existence of a large am ount of pores in the AD85
microstructure as shown in Figure 4.10. The pores tend to enhance the erosion process
the water trapped in the pores may generate local fracture network due to the water
wedge effect and it increases the total material removal significantly. The derived
One interesting phenomenon arising from the trend of the erosion ratio versus the
alumina type is that there exist a strong correlation between the erosion ratios and
the grain sizes as well as the fracture energy. The erosion ratio, the grain size and
the ratios of grain size to fracture energy ( a / 7 ) are normalized with their respective
average values and plotted in Figure 4.11 for the purpose of comparison. It appears
that the larger the value of a / 7 , the larger the erosion ratio.
4 .3 .4 S u m m a r y
The modified H utchings’ equation has been used to evaluate the stress wave en
ergy. The crack network model has been applied to predict the relative erosion ratio
among different types of alumina ceramics eroded by the normal impact of an abrasive
water jet. Good correlation with experim ental results is observed. A strong correla
tion also exists between the erosion ratio and the grain size as well as the fracture
energy of alumina. It is also found that erosion of aiumina ceramics by water alone
115
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4.10 Scanning electron micrograph of hand-made fracture surface of AD 85.
(x 800)
5
MODEL
U) 4 EXPERIMENT 2
O
m GRAIN SIZE
CC
</) 3 i- a/ y
O
LU
T3
CD
2
_N
"ca
E
o “0 -
2
0
AD85 AD90 AD94 AD96 AD995 AD999
Alumina Ceramics
Figure 4.11 Correlation of the grain size and fracture energy with the erosion ratio.
116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
is very small compared to erosion by abrasives.
Figure 4.12. The top layer of m aterial is removed by a cutting action through plastic
flow. The stress waves transm itted into the vicinity cause fractures along the grain
boundaries, which form a network of cracks. Evaluation of the stress wave energy
Since the plastic flow occurs in the im m ediate vicinity of an im pact site, Finnie’s
m odel (Finnie, 1958; Finnie & M nFadden, 1978) gives the im pact loading pattern as:
where
Pvo = m u jv s i n a , (4.38)
Pho = (4.39)
Kf
» = (4.40)
m
where ay flow stress of target material, b w idth of particle cutting face, if; ratio of the
depth of contact I to the depth of cut y t, and K f ratio of vertical to horizontal force
com ponents.
The total stress wave energy radiated by the im pact is the sum of the work done
by the vertical and horizontal force com ponents over the corresponding superposed
117
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
e.g .
Figure 4.12 Idealized picture of an abrasive grain rem oving material by scratching
118
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IF = Wy + W 2 = ~ P 2voG v + j Pa
20Gh (4.41)
where T is the loading contact tim e, T = 7r/u/, and G„ and (?/, is the real part of the
radiation adm ittance Yvand 1/,, defined as a (com plex) ratio of m ean displacement
rate to the applied force (Miller & Pursey, 1956). They are expressed as:
n - r» = p t , (« •« >
PvoC1 Piloe
where ux and u z are the superposed m ean displacements in the horizontal and vertical
direction, respectively.
Since the traverse distance of the m oving load corresponding to the abrasive tra
jectory is very small, it is assumed that the stress energy radiated by a stationary
load applies to the present case, provided that the loading history is equivalent. W ith
reference to cases (a) and (b) in Miller and Pursey’s work (1956), the displacement
components corresponding to a vertical load and a horizontal load can be, thus, ex
pressed in the form of Fourier transform as: (Note: Equations (49) and (50) in Miller
& Pursey’s (1954) can not be used because of some calculation errors in derivation of
(4.44)
W = 2y^ ^ -1 (* § -
(4.45)
c „ ( m
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
where
n o = w - k i f - 4? v p - h ^ e - k i (4 .4 7 )
h = J -§ ^ , k2= u i S z ’ <4-48>
r, is the half width of the strip. At the abrasive and target m aterial contact surface(i.e.
_ 2i G j Q s i n t r ,
xvF ~ c , 4f ( o (4-50j
2 kn \M'2 — k%3in£r3
- - W ( 0 ( 4 -5 1 )
w = * ± J g ( 4. 52)
2 z fT (fl« n fr a
(7 , ^ ( 0 ( 1
where
Using inverse Fourier transform, one can obtain the displacem ent functions at z
= 0 as:
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
_ _1_ r 2 i G { t ) s i n t r s ,.g
2tt J .oc (7,^(0 *
—2 z*00 G(
= ~ r~ / f ? 7 \ 3 in ( r s s i n i x d i (4 .5 6 )
■
kC ^ J o f'(S )
(N oticing that the first term is odd function and the 2nd term is even function.)
1 /■» 2 y ^ 2 ~ kl kl s i n i r * icX J ,
= * L c ^ m ‘ *
2 r°° \ / Z 2 — k 2k 2
~ n ~ / ~s i n Z r ° c o s ( x d Z (4 -57)
TrC^iJo ? ^ (?)
2 ^ 2- - 4 f ) iJ 7 i( r 3
Uzv = J- [ ' ■
2 tt j — < 7 ^ (0
— — ------ a i n c ^ r 3 c o s c , x d c , (4.58)
ttG. 14 -/o
1 2ifT( 0 « n ^ r a _if
"*A = ^ L ~ C ^ W ~ ?
121
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
- 1 f r3 j H ( 0 ■ >1 — c o s r J ,
v-zh = — / uzhdx = —— / — —•.szn7\,£ -------- -----------------------(4.63)
r 3 Jo trC.,4 -/u F (£ ) r a£ ' K '
_ —2 H0{£) . , . 1 — cosr3k[£
“ '/i = ~ 7 ; To / p / f \ s in r 3ki£ - (4-67)
TrC'.i.iAr ./u ,F0(£) r\,£
where
Cu h 2(1 - v ) , „ ,
^ — fe7 — V 1 - 21/ ( )
Therefore, for the loading P„(i) = Pv0e>ul and P/,(t) = P/,0e!“( applied over the
contact area A,
—2r3I \ P voelult
« « = " T tTtAO,,
W " (4 -72)
_ _ 2r3I 2Pi:0elu;t ,.
“ ** ~ TrA^M ( }
122
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
lutt
2 r sI 3Pvoe"‘
U,„ = (4.74)
k AC, ii
- 2 r J . tPhoeiut
= * A C .U (4’75)
where
r cosrak xi ^ 3& tA ^
/ i = y„ m (~ ^ r )( t m w ( 6)
/, = (4.78)
Jo ^o(?) T*a«iC
I i = /» w (^ r )( r .* ,« )<i? (4-79)
The superposed displacements on the horizontal and vertical coordinates are, re
spectively:
Ux — Uxu T Iti/i
u z = U 2„ + U z h
2 r3zw - [ R e ( I i ) + i I m ( I i ) ] P vo + \Re(I2) + i I m ( L ) ] P h„ , .
n = ^ ftT ( 4 ' 8 2 )
123
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2raiuj [Re(I3) + i I m ( I 3)\PV0 - [Re(I4) + i I m { I A)\Pho fA ooN
= ^ jr (4-83)
2rsu I m ( I i ) P m - I m ( I 2)Pho
G‘ = T Z c Z K (4 '84)
W = - I m ( h ) - / m ( / 2) ( ^ ) 2} (4.86)
± vo *vo
For r 3k\ < 1 (e.g., r 3kx = 7.85 x 10 - 3 for mesh 50 abrasive and AD 99.5 target),
it can be approximated:
s i n r 3k [£ ss r 3k i£ (4-87)
1 — cosr3k i ( % ^ 3 (4.88)
<4-M>
(4 J 1 )
T'ski f x £2H o( 0
L = % (4.92)
I Fo( 0
124
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
These integrals can be num erically evaluated using the approach described by
Miller & Pursey (1954). The poles of the integrands in Iy - Iy occur at the zeros of
( 2 e - v 2y = m i ( e - m 2 - v 2) ( 4 .9 3 )
After elim inating the term 16£3 on both sides, this equation is reduced to a cubic
C, this equation is num erically solved for the range of Poisson’s ratio 0. 1 < v < 0.49.
The solutions are listed in Table 4.2. It is assumed that principal values of the radicals
are used in integrals Iy - Iy. Therefore, the only significant zeros of F 0(£) are those
for which £ is real and |£| > |p|, i.e. those corresponding to the 3rd poles (p;)).
Equation 4.86 indicates that only the imaginary parts of these integrals areof
interest. Since only the range 0 < £ < /x and the small semi-circle path above the
pole produce imaginary values, the following expression can be applied to evaluate
Let
<4 -9 5 >
A = /m f « ( 4 ' 9 6 )
0 (- W
= I m t™ r i( (4,98)
125
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.2 Poles of the integrands.
P o isso n ’s
1st Pole 2nd Pole 3rd Pole
Ratio v
0.1 1.500 0.674 1.000 1.680
0.15 1.558 0.715 0.999 1.727
0.2 1.633 0.771 0.996 1.793
0.21 1.651 0.786 0.994 1.808
0.22 1.669 0.802 0.992 1.825
0.25 1.732 0.866 0.975 1.884
0.3 1.871 0.975 ± 0.1 OOi 2.017
0.333 2.000 1.027 ± 0.147i 2.145
0.35 2.082 1.061 ± 0.171i 2.226
0.4 2.450 1.218 ± 0.253i 2.600
0.45 3.317 1.605 ± 0.400i 3.495
0.49 7.141 3.375 ± 0.947i 7.485
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The M athCAD software and Programs 2-5 in Appendix C are used to numerically
evaluate the integrals 0 i - 04 for the range of 0.1 < v < 0.49. The results are listed
in Table 4.3.
I m ( I i ) = ^ r sk\0i, I m ( I 2) = fa (4.99)
I m ( I 3) = 0 3, Im{I.i) = ^ r 3k x0 A (4.100)
^ = j4.Cz.J4 2 * VO + ( j* V
rO)2]] (4'101)
‘3 t ( f t + | )) ( 4 ' 1 0 2 )
where
The values of 0$ and 0 6 are calculated for the range 0.1 < u < 0.49 and listed in
Table 4.4.
As observed in the SEM photos, this fraction is in the order of 10- 1 . Therefore,
2ra = b = 0.1rp was used as the preliminary assumption. According to Finnie (1958),
both of the values of 0 and K j are estim ated to be 2 for small angle im pingem ents.
127
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
W = m v 2s i n 2a ( — )^( - - ) ? [k + 2/?(— ) z (— ) 5] (4.104)
pp t p crj
where
s = 0.0309 ( f t + f t ) f t , (3 = - f t f t ( f t + 4 f t ) . (4.105)
In general, the m agnitudes of pp and p are in the same order and E > 0 7 . There
fore, the value of k is negligible compared to the 2rd term of Equation 4.104. This
implies that the contribution of the uxv and uzh on W is negligible. The expression
w= (4.106)
E
The volum e removal due to plastic flow (4p) is evaluated with the modified Finnie’s
model (Finnie & M acFadden, 1978), considering only the case where the particle is
not imbedded. The volum e removal for a single impact can, then, be expressed as:
bv2 . 2 15Trbvls i n la
VD = —- ( s m 2 a — 4sin a ) -j ----- ------ (4.108)
u;2 4rpur
2 f
V„ — ( si n 2 a — 4 s i n 2a + 38.12 v s i n 3a. — ) (4.109)
407 y <t/
128
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.4 Numerical results of j3$, 0a, k , & 0.
P o isso n ’s P5 p6 K P
Ratio v
0.1 1.088 1.011 0.027 12.526
0.15 1.119 1.028 0.025 11.749
0.2 1.138 1.054 0.022 11.002
0.21 1.140 1.061 0.022 10.857
0.22 1.142 1.068 0.021 10.718
0.25 1.141 1.095 0.021 10.305
0.3 1.120 1.160 0.020 9.678
0.333 1.089 1.224 0.020 9.303
0.35 1.066 1.267 0.021 9.135
0.4 0.956 1.464 0.023 8.707
0.45 0.745 1.948 0.031 8.402
0.49 0.360 4.137 0.046 8.264
16
12
10
8 Real Values
-«— - Approx. Values
129
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The volum e removal due to fractures ( V j ) is obtained by substitutin g 4.106 in
4.12, yielding
= M a < r j T n f 8 in*a
1 37 E K' ’
The total volum e removal by a single im pact is, therefore,
f wf3a(rfmv23in2a
37 E
9 7
77X17** IO
4 ( si n2 a — 4 s in 2a -f 38.12v s i n 3a j — ) (4.111)
4 <rf V °7
The experim ental conditions are similar to those described in Section 4.3.3. The
two sets of erosion test results are plotted in Figure 4.14 in terms of “erosion ratio” .
pVf C f wf3acrfpv s i n 2a
(4.112)
m 37 E
pV = PvL + e v!, }
m m m
where C is an efficiency coefficient, which accounts for the energy loss due to particle
0.978, R = 0.176, the particle velocity prior to im pact (v), according to Appendix
A, iscalculated to be 477 m /s. The value of flow stress is estim ated with twice the
130
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To estim ate the coefficient C, a similar erosion test is conducted on a target m ate
rial of SS304, which has a Brinell hardness of 149 ( “Sourse Book” , 1978) (equivelent
Vickers hardness 156 kg/ m m 2) and a density of 8027 k g /m m 3. Its flow stress is esti
m ated to be twice the Vickers hardness, i.e., 1529 M Pa. Since the m aterial removal
m echanism on a ductile material is solely due to plastic flow, Equation 4.113 is used to
calculated the total theoretical erosion ratio, which yields 0.1503C g /g . B y equating
this value to the experim ental erosion ratio, 0.01563 g /g , it is determined that C =
0.104. To estim ate the value of the data for AD 99.5 is used in equation 4.112 to
Using the values C = 0.104 and f w = 6.65 x 10 1, the total theoretical erosion
ratios for the six aluminas are calculated and plotted in Figure 4.14 to be compared to
the experim ental results. A good correlation betw een the elasto-plastic m odel and the
erosion experim ents is observed. As shown in Figure 4.15, the normalized grain size
correlates the erosion data trend for all ceramics tested except AD 85. In addition,
plotting the ratios of a / 7 in Figure 4.15 shows an even better correlation. Therefore, it
appears that the grain size and fracture energy are the two major material parameters
4 .4 .4 D iscussions
The theory of Miller and Pursey on the elastic response of normal and tangential
loads on an elastic half space has been used to derive the stress wave energy. The
obtained expression of the stress wave energy has been incorporated into the crack
network m odel, leading to an expression for the single particle material removal due
to intergranular network cracking. Finnie’s m icrocutting model has been used to eval
uate the volume removal due to plastic flow. Combining these two m aterial removal
com ponents yields a complete elasto-plastic model for a single impact at low incidence
angles. This model has been verified with erosion experim ents. Good correlation is
131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Water Pressure: 241 MPa Water Flow Rate: 4.784 1pm
Orifice/Nozzle Diameter: 0.457 mm/1 ..575 mm
Abrasive: mesh no. 50 garnet, 16.9 g/s
5 -
Traverse Speed: 15 and 20 mm/s
o Stand-off Distance: 5 mm
©
o' Incidence Angle: 20 degrees
X! 4 -
60
"cS • ----- Model
3 - □— Experiment 1 ( 1 5 m m / s)
a n- - Experiment 2 (20 mm/s)
Oi
c
_o
ou<
W
A D 85 A D 90 A D 94 A D 96 A D 995 A D 999
Alumina Ceramics
Figure 4.14 Comparison of the elasto-plastic model and the experim ental results.
>-
sa
£
C
o
£ 0 ----
Alumina Ceramics
Figure 4.15 Grain size and fracture energy are the two major m aterial parameters.
132
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
observed.
In a study on the erosion dam age of several ceram ics, Ritter (1985) proposed an
analytical m odel based on the assum ption of grain boundary cracking, which was
described as:
amv2
V oc-- ------ (4.115)
7
This m odel was recently modified by W ata and W atenabe (1987, p5) incorporating
a H vm v 2
V oc (4.116)
-
7
where Hv = Vickers hardness.
However, they were unable to find a correlation betw een this m odel and exper
im ental results. Therefore, an empirical model was proposed for alumina ceramics
as:
(4'117>
The two analytical models, i.e. Equations 4.115 and 4.116 are very similar to the
expression of the volume removal due to fracture derived in this study, i.e. Equation
4.110. It isinteresting to note that the inverse proportionality of the m aterial hardness
in the empirical model( Equation 4.117) is analogous to the F innie’s m odel, i.e.,
both brittle fracture and plastic flow, the proposed dynam ic m aterial removal model
is more representative.
Although the study by Wada and W atanabe did not dem onstrate a correlation be
tween the grain size and the erosion rate, the proposed m odel shows a good correlation
between them.
133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4.5 Sum m ary
m aterial removal effects of both the fractures and plastic flow. The fractured volume
individual grains or fragm ents. A crack network m odel is thus derived by relating
the fracture surface energy required for the network form ation to the input stress
wave energy. The elasto-plastic m odel has been applied for evaluating the material
removal of single particle im pacting at normal incidence or low incidence. The stress
wave energy is expressed with a modified H utchings’ equation for normal incidence
and with an equation derived in this study for low incidence, respectively. The crack
network m odel com bined with B itter’s deform ation wear m odel gives the total ma
terial removal for a single particle impact at normal incidence, and, combined with
Finnie’s m icrocutting m odel, gives the total m aterial removal for low incidence im
pacts. The obtained expressions of material rem oval have been verified by erosion
experim ents. Good correlation is observed. It is also found that the erosion rate is
strongly correlated w ith the grain size and fracture energy of target materials.
134
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 5
E N E R G Y D IS SIP A T IO N P H E N O M E N A
as well as other beam cutting processes such as laser, plasma arc, flame cutting,
etc. A typical AWJ cut workpiece displays striation marks on the cut surfaces as
shown in Figure 5.1, and a tapered kerf such as those in Figure 5.2. To characterize
these energy dissipation phenomena, two sets of AWJ kerf cutting experim ents are
T he objective of this experim ent is to provide some information on the kerf width
and taper angle variation in AWJ cutting. A 6.09 mm thick alumna (AD 85) plate
is used as the workpiece. Stand-off distance, traverse speed, water pressure and
orifice/nozzle diameters are fixed. The resulted kerf widths and taper angles are listed
in Table 5.1 and are displayed in Figure 5.3 - 5.10. The general trends are that (1) the
kerf w idth is directly proportional to the exit nozzle diameter and stand-off distance
w hile increasing water pressure and decreasing traverse speed also slightly increase
the kerf width; ( 2 ) the taper angle increases with stand-off distance, exit nozzle
diam eter and the traverse speed, but remains relatively constant with variations in
w ater pressure.
A series of cutting tests is conducted using four different types of m aterials repre
senting both ductile and brittle categories. These materials with different thickness
are cut by varying the cutting parameters as shown in Table 5.2. The striation marks
135
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5.1 Striation marks on an Figure 5.2 Tapered kerfs on an alumina
136
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.0 2.0
1 .8 -
(m m )
E
E
1.6 -
Width
§
Kerf
0 2 4 6 8 10 0.0 0.2 0.4 0.6 0.8 1.0
S ta n d -o ff D is ta n c e (m m ) T raverse R a te (m m /s)
Figure 5.3 K erf w idth vs stand-off distance, Figure 5.4 K erf w idth vs traverse speed.
w 2. 0 -t----------------------------------------------------------------- 2.0
-j
1. 8 -
E
E
5
r(O
1 .2 -
Figure 5.5 Kerf w idth vs water pressure. Figure 5.6 K erf w idth vs exit nozzle diam eter.
7 7
lission of the copyright owner. Further reproduction prohibited without permission.
6 6
T aper(degree)
0) 5 5
(D
o>
©
4 4
k_
©
Q.
© 3 3
H-
2 2
1 1
2 4 6 8 10 0. 0 0. 2 0.4 0.6 0. 8 1.0
S ta n d -o ff D is a n c e (m m ) T raverse R a te (m m /s)
Figure 5.7 Taper vs stand-ofF distance. Figure 5.8 Taper vs traverse speed.
CO
OO 7
Q) 5
0)
T aper(degree)
cn
©
k_
©
4
CL
m
H 3
1
200 250 300 0.6 0.8 1.0 1.2 1.4 1.6 1.8
W ater P r e s s u r e (ksi)
E x it N o z z l e D i a m e t e r ( m m )
Figure 5.9 Taper vs water pressure. Figure 5.10 Taper vs exit nozzle diam eter.
Table 5.2 C utting test parameters.
Plexiglas 2 6 3 .3 6 .4 5 7 /1 .1 6 8 IGE 1 1 .4 1 .0
Maine 1 .0
Marble 1 5 1 .0 6 .4 5 7 /1 .1 6 8 1 1 .4
Garnet
Marble 2 4 2 .4 4 .4 5 7 /1 .1 6 8 # 60 1 1 .4 1 .0
241 1 .5
A I6061-T 6 1 3 8 .2 .3 5 6 /1 .1 6 8 1 1 .4 1.5
A I6061-T 6 2 9 .6 7 .3 5 6 /1 .1 6 8 Barton 7 .6 1 0 .5
A I6061-T6 3 8 9 .3 .3 5 6 /1 .1 6 8 # 80 7 .6 0 .4
Polypropylene 39 .1 . 3 5 6 /1 .1 6 8 7.6 5 .2 5
PLEXGLAS 1
6- PLEXGLAS 2
MARBLE 1
E
E MARBLE 2
c 4- ALUMINUM 1
a
ra ALUMINUM 2
'>
Q 2-
ALUMINUM 3
POLYPROPYLENE
0 20 40 60 80 100
Depth in workpiece (mm)
139
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
are etched and plotted in Figure 5.11 in terms of the depth in the workpiece and
the deviation referred to the depth coordinate. For a qualitative comparison of these
curves, each curve is normalized with respect to the workpiece thickness (Z maT) in
the z direction and the m axim um deviation (X max) in the x direction, and plotted
linearly and logarithm ically in Figure 5.12 (a) k (b ), respectively. Interestingly, all
of these curves are nearly parabolic. Curve-fitting the data w ith parabolas yields the
correlation coefficients between 0.990 and 0.999. The curve-fitted equations and the
corresponding correlation coefficients are listed in Table 5.3. Therefore, the striation
where x is the horizontal deviation of the curve, 2 is the depth in the workpiece and
This type of generalization can be used as a first approxim ation for quantifying the
geometric nature of the striation. It is noticed that the curve-fitted expression 5.1 for
striation curves is not applicable in the neighborhood of the top of workpiece, where
an improved curve-fit is achieved by combining a straight line for 0 < z < b and
Equation 5.1 for z > b. Since energy dissipation is insignificant for small values of 2 ,
Equation 5.1 is used in this study. For a group of preset waterjet cutting parameters,
the striation curves are parallel as shown in Figure 5.13. Therefore, a continuous
cutting process can be m odeled as a curve (parabola) m oving with a constant speed
shown in Figure 5.13, are related by Aa; in the horizontal direction and A h in the
A h = AxcoaO (5.2)
140
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
PLEXGLAS1
PLEXGLAS2
MARBLE 1
MARBLE 2
ALUMINUM 1
ALUMINUM2
ALUMINUM 3
RCLYPRCPYLENE
PARABOLA
0
.01 10
N o r m a liz ed d e p th in w o r k p ie c e , X /X m a x
COEFFICIENT OF
CURVE LABLE CURVE-FITTED EQUATION CORRELATION R2
141
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
AWJ Cutting Head
Workpiece
max
max
max
142
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
where 0 is the angle of slope of the cutting front, d x / d t is the traverse speed (u) and
M = ucosB (5.4)
or
— = , 1 (5-5)
The linear displacement rate u is the desired cutting speed. The linear material
removal rate M is the actual local cutting speed. The ratio M / u represents the cutting
efficiency at the localized zone. Equations 5.4 & 5.5 explicitly indicate the trend that
the cutting efficiency decreases as the angle of slope (9 ) or the depth z increases.
W hen the cutting efficiency is reduced to a certain degree, the abrasive entrained
jet stream becom es unstable and a striation mark starts to develop. The cutting
efficiency eventually reaches a critical point at which the jet stream significantly loses
its cutting power and is “kicked-back” by the target material. This critical point is
associated w ith the m aximum cutting depth for a given cutting condition.
The angle between the vertical coordinate and the jet exiting the bottom of the
workpiece (see Figure 5.14) is the so called "‘jet exit angle”. This angle will increase
with the traverse speed provided that the other parameters remain constant. As the
traverse speed increases to a critical value, the jet will not be able to cut through
the workpiece and the jet exit angle suddenly becomes very large and unstable, i.e.,
143
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5.14 Jet exiting the bottom of workpiece.
144
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the jet is “kicked back”. The jet angle beyond which a “kicked back” phenomenon
occurs is defined as the critical jet exit angle ( 6C). From a series of cutting tests,
it is onserved that the value of the critical exit angle (9C) nearly remains constant
glass, 901 Tool Steel and W hite Marble, are chosen to include a group of materials
from both ductile and brittle categories. In addition to m aterial properties, the cut
ting process is also tested by varying waterjet parameters such as: traverse speed,
water pressure, abrasive flow rate, orifice/nozzle size, workpiece thickness, and abra
sive type and size. The traverse speed is chosen to be the major controlling parameter.
For each of the three m aterials, Aluminum 6061-T6, Polypropylene and W hite Mar
ble, four sets of cutting tests are conducted. In each set of cutting tests, one of the
four param eters, water pressure, abrasive flow rate, orifice/nozzle size, and workpiece
thickness, is varied while the others are held constant. The values o f these parameters
For each test cut, the cutting speed is started at zero and then increased gradu
ally until the critical point of deflection (kick- back) is reached. At this point, the
cutting speed is im m ediately decreased slightly until the cutting is resumed. Several
photos are taken of the exiting jet with a camera equipped with a telephoto lens and
positioned just far enough from the cutting area to avoid the cloud of water vapor
and dust created during the cutting process. A pressurized air nozzle is also posi
tioned near the cutting area so as to disperse this cloud and increase the clarity of the
photographs. The critical jet exit angle can be measured directly from these pictures
as shown in Figure 5.14. However, this m ethod serves only as an indication of the
feasibility of m onitoring the exit angle during the cutting process. For the purpose
of analysis a simpler m ethod for obtaining the exit angle is used by simply etching
145
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 5.4 Cutting test parameters and the critical jet exit angles.
146
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the cut surface onto a piece of paper and then m easuring the angle from the etching.
T he results are listed in Table 5.4 and plotted in Figures 5.15 - 5.20.
From Figures 5.15 - 5.19, it appears that the only param eter which exhibits a
significant trend in its relationship w ith the critical jet exit angle is the workpiece
thickness (see Figure 5.19). The curves in Figures 5.15 through 5.18 shows that the
critical angles for the m ajority of the samples lie in a band betw een fifteen and twenty-
five degrees. This can be seen more clearly in Figure 5.20, which is a plot of all the
data obtained from the 38 mm thick workpieces. The m ean value and the standard
deviation of these data are found to be 20.4 degrees and 4.08 degrees, respectively.
T he deviation includes m easurem ent inaccuracy since the values of the exit angle are
measured when the cutting speed had been reduced by a sm all, but varying amount
from the critical point. The instability of the critical cutting condition should be
another im portant factor which makes the jet exit angle fluctuate.
The energy dissipation phenom ena in AWJ processes such as striation marks and
tapered kerfs have been docum ented. It is found that the striation marks on the
abrasive jet cut surface can be characterized by parabolic curves. A simple analysis
indicates that the cutting efficiency depends solely on the slope of the striation curve.
A preliminary experim ental study reveals that the critical jet exit angle at the condi
tion of kick-back decreases as the workpiece thickness increases while its dependence
147
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80 - 6 0 6 1 -T6 ALUMINUM
WHITE MARBLE
S’ 70 *
S- 60- POLYPROPLYENE
50-
40 -
30 -
20-
10-
Figure 5.15 Effect of water pressure on the criticla jet exit angle.
90
80 6 0 6 1 -T6 ALUMINUM
m
CD
oID> 70 WHITE MARBLE
n POLYPROPLYENE
60
UJ
a 50
2
< 40
<
O 30
EC 20
O
10
0
.2 2 9 /.7 8 7 .3 5 6 /1 .1 9 4 .4 5 7 /1 .5 7 5
ORRIFICE/NOZZLE DIA (mm)
80 - 6 0 6 1 -T6 ALUMINUM
CD
2□> 70 - WHITE MARBLE
CD
T> 60 - POLYPROPLYENE
50 -
a
2 40 -
<
—I
< 30 -
O
b 20 -
E
O 10 -
2 4 6 8 10 12
ABRASIVE FLOW RATE (g/s)
Figure 5.17 Effect of abrasive flow rate on the critical jet exit angle
148
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10-
LU
UJ
C/3
<
WORKPIECE MATERIALS
Figure 5.18 Effect of workpiece m aterial on the critical jet exit angle.
6061-T6 ALUMINUM
WHITE MARBLE
U) POLYPROPLYENE
£
111
o
z
<
-J
<
o
h-
CE
o 10-
0 20 40 60 80 100
WORKPIECE THICKNESS (mm)
Figure 5.19 Effect of workpiece thickness on the critical jet exit angle.
80 - • 6061-T6 ALUMINUM
70: « WHITE MARBLE
at
B POLYPROPYLENE
_i
V-
0 5 10 15 20 25
SAMPLE NUMBER
Figure 5.20 Data of critical jet exit angles for 38 m m thick workpieces.
149
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 6
M O D E L IN G OF A W J C U T T IN G P R O C E SSE S
(1984, p249 & 1988, p l5 9 ), it was observed that the total depth of cut is divided
into two distinct zones associated w ith different m odes of abrasive/target interaction.
angles. In the lower zone, sequential steps are formed which lead to large angle
im pacts. A modeling study (Hashish, 1984, p249) characterized the upper zone as
a cutting wear zone and the lower one a deformation zone. F in n ie’s m icrocutting
m odel (Finnie, 1958) and B itter’s deformation wear model (B itter, 1963, p5 & p l6 9 ),
respectively, were used to evaluate the material removal in the cutting wear zone
and deform ation wear zone. A global cutting equation was then derived for AWJ
cutting of ductile materials. More recently, Hashish (1987, p 6 6 -l) im proved Finnie’s
m odel by incorporating the effects of particle size and shape, resulting in a change
in the particle velocity exponent from 2 to 2.5, which provides better correlation
w ith experim ental values. A corresponding modified version of the cutting equation
Good correlation between the modified equation and cutting experim ents with various
The step phenomenon was also observed in the SEM study described in Chapter
3. However, in contrast to H ashish’s conclusion that these steps lead to large angle
im pacts, Chapter 3 concludes that the entire cutting front is im pacted by the original
jet or the deflected jet at glancing angles and the existance of the steps slightly changes
150
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the local curvature of the cutting front, but does not affect the low incidence cutting
mode. Therefore, the elasto-plastic model for the low incidence im pact, derived in
Section 4.4, applies to the individual particle removal during the cutting process.
The equation for depth of cut is derived by relating the m acro material removal
abrasive particles entrained in the jet. According to Section 5.1 on curve-fitting the
striation marks (traces of the cutting front), a parabola can be used to approximately
represent the striation curve, and thus the curve of the cutting front. The local cutting
efficiency along the cutting front is defined to be the ratio of the linear m aterial
removal rate (M ) to the traverse speed (u ), which is a cosine function of the angle of
The total m aterial removal rate along the cutting front can be calculated by
integrating
Q = / Mdjds (6.1)
Jcp
where dj is the effective jet diameter. As the cutting efficiency decreases from top to
bottom along the curve of the cutting front, the jet diam eter is also reduced. It is
assumed that the effective jet diameter is related to the focusing nozzle diameter and
dj = DcosQ (6.2)
where D is the jet diameter at the top of workpiece which is assumed to be equal
to the focusing nozzle diameter. The variation of the effective jet diameter causes
151
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
= Du J cosddz (6-3)
since dz = dscosO.
and
dz — —----- — . (6-5)
2Acos2#
Substituting Equation 6.5 into 6.3 and integrating along the parabolic curve with
D
/ 7u
7/ trVc riH
dc dO
Q = 2 i L ^ = m )D n k (6-6)
where
A= ^ (6.7)
and
f L( 0c) = - 2 — l n ( 1 + 3™ d c ) (6.8)
t andc cos9c
The total m aterial removal rate along the cutting front can also be interpreted as
the accum ulated effect of individual particle removal. Since the individual abrasive
particles in the jet im pact on the different locations of the cutting front, differences in
the distance that individual particles travel prior to impact lead to a loss in efficiency.
152
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
B y incorporating an average efficiency into the idealized volum e removal of single
Q = — ( 4 / c os 9ds) V (6.9)
771 *5 JCp
where V is the idealized volum e removal by individual abrasive particle, m the mass
of single abrasive particle, m the abrasive flow rate, and the term within the brackets
stands for the average efiiciency(S = total curve length of the cutting front).
Since
[ cosOds = f dz — h (6.10)
Jc p Ju
and
dz
= / ds = f
Jcp Jo cosd
d6
= — j
2A Jo cos3 9
tanOe .1 + si n9c
2t an9c cos9c + M I'l ' i r )1»
cos9c (6-n )
777
Q = f2(0c)(~)V (6.12)
m
where
M V = ^ («-W)
cosOc 'r “ H cosOc )
Combining Equations 6.6 and 6.12 yields:
i = (6.14)
Du v !
where
153
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 ta n 2Qr
m ) = rtanOj (6.15)
Lc o sO (
Section 5.2 indicate that the value of Bc is nearly constant regardless of variations in
workpiece m aterial, water pressure, orifice/nozzle size as well as abrasive flow rate,
except that it slightly decreases with increase in workpiece thickness. In any event,
the value of 6C doesn’t exceed 50 degrees. In the range of 0 < 8C < 50°, the variation
of f ( 8 r:) given by Equation 6.15 is less than 2% (see Table 6.1). Therefore, the value
9C, deg. 0 10 20 30 40 50 60 70 80 90
m 1.000 1.000 1.000 0.998 0.993 0.981 0.953 0.891 0.752 0.053
C f m(3ao'fmv2s i n 2a
V
3jE
C m v 2, . „ . . n „ . rp7>\
4 ------------ ( s i n 2 a — 4 sin a + 3 8 .1 2 u .s z 7 t a. — ) (6.16)
4oy V crf
The contributions of the secondary and tertiary impacts are assum ed to be propor
tional to the m aterial removal of the primary im pacts. They are incorporated into
Equation 6.16 with a coefficient of im pact efficiency (C ), which also accounts for the
efficiency loss due to particle m utual im pacts and particle fragm entation. Since the
incidence angle a is considered to be very small, thus, s i n a « a , and the 2nd and third
terms in the bracket are negligible compared to the first term. It is further assumed
that the variation of the incidence angle along the cutting front is sm all. Therefore,
the value of a is taken to be constant. W ith these sim plifications, Equation 6.16
reduces to
154
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
+ {en)
with the velocity v calculated w ith Equation 1.1, Equation 6.17 is rewritten as:
_ , r?C'»C'v . 2 C mP , L. . 2 f w/3a<rja2
1 -f R pm Z'yE cf
Substituting Equation 6.18 into 6.14 yields the final equation for the depth of cut:
L ( t] C v C ,j v2 C m P w 2 f wf3acrf a 2 a ,
4 = ( t t r ] + { ]
99.9, AD 99.5, A D 94, AD 90 and AD 85). The theoretical values of depth of cuts
are predicted w ith Equation 6.19. To apply Equation 6.19, the coefficients need to
be determ ined. According to A ppendix A, the orifice efficiency C v varies from 0.932
to 0.911, the com pressibility coefficient Cy from 0.985 to 0.974 and the momentum
transfer efficiency rj from 0.805 to 0.936 within the usual water pressure range, 150
- 300 M Pa. The product of these three coefficient TjCvC y varies between 0.739 and
0.831. For sim plicity, the average value of 0.79 was taken. The stress wave energy
coefficient f w was determ ined to be 6.65 x 10--1 in Section 4.4.3. W ithout losing
0.175 rad. Correlation of Equation 6.19 with the experim ental result gives a coefficient
of impact efficiency (C) of 0.6, instead of 0.104 determ ined in Section 4.4.3. The higher
value of the coefficient of im pact efficiency is probably due to each abrasive particle
interacting more than once with the target material along the cutting front. W ith
these coefficients, the predicted depth of cut is calculated using Equation 6.19 and is
155
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
As shown in Figure 6.1, the derived m odel qualitatively predicts the relative re
sponse of the five different alum ina ceramics. Figures 6.2 - 6.6 indicate that, although
certain discrepancies exist, generally good correlation between the m odel and exper
im ental results in the general trends of depth of cut as a function of the five major
process parameters is observed. The m odel over estim ates the effect of increasing
abrasive flow rate. As abrasive flow rate increases, the probability of particle mutual
impingem ent also increases, which is not taken into account in the present study.
Usually, an abrasive flow rate exceeding 11.34 g /s (1.5 lb /m in ) does not produce ad
ditional m aterial removal, as evident in Figure 6.4. W ith the abrasive flow rate below
this value, the derived model provides close estim ates. A large discrepancy is found
in Figure 6.6 for small nozzle diam eters, which is probably caused by the om itted ef
ficiency loss in the case of excess flow passing through a small focusing nozzle. Again,
this indicates that an optim um range of nozzle diameters should be used to reduce
the efficiency loss. W ithin this range, the model becomes applicable.
6.3 Summary
B y relating the macro m aterial removal rate along the AWJ cutting front to the
accum ulated micro m aterial removal rate due to individual particles, the equation for
depth of cut is derived. This equation has been verified by kerf cutting experim ents.
It is concluded that the derived kerf cutting m odel correlates well with experim ental
results although certain lim its are required in the selection of certain process param
eters.
156
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
w ater pressure: 207 MPa
orifice/nozzle diam eter: .457 m m/ 1.27 mm
traverse speed: 0.8 and 1.2 mm/s
stand-off distance: 1.5 mm
16 -
abrasive: mesh 80 garnet, 7.56 g/s
w ater flow rale: 4.432 1pm
u = 0.8 ram/s Model
u = 1.2 mm/s Model
12-
u = 0.8 mm/s Exp.
u = 1.2 mm/s Exp.
8-
4 -
1 , ,-- 1----
AD 85 AD 90 AD 94 AD 99.5 AD 99.9
10
target material: AD 99.5
orifice/nozzle diam eter: .457 m m /1.27 mm
stand-off distance: 1.5 mm
abrasive: mesh 80 garnet, 7.56 g/s
8-
traverse speed: 1.2 mm/s
6-
3
CJ
4 -
q. _ - Exp.
2- D Model
157
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
target m aterial: AD 99.5
w ater pressure: 207 MPa
orifice/nozzle diam eter: .457 m m /1.27 mm
sta n d -o ff distance: 1.5 mm
15 abrasive: mesh 80 garnet, 7.56 g/s
CJ 10-
Q
5 -
♦ - ' Exp.
Model
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0
8-
target m aterial: AD 99.5
w ater pressure: 207 MPa
orifice/nozzle diam eter: .457 m m/1.27 mm
7 - stand-off distance: 1.5 mm
traverse speed: 1.2 mm/s y^
abrasive: mesh 80 garnet
6 -
/ S*---------------- •
s ' y
£ / yy
3 5-
O
U
O y r '
/ /
S* 4- / /
a / /
/ -----♦ — Exp.
{ ------H----- Model
3 -
158
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
10
target m aterial: A D 99.5
w ater pressure: 207 MPa
orifice diam eter: .228, .305, .356, .406, .457 mm
8- nozzle diam eter: 1.27 mm
stand-off distance: 1.5 mm
abrasive: mesh 80 garnet, 5.67 g/s
traverse speed: 1.2 mm/s
6-
u
4-
Q
—*
—- —
2-
*■ — - Exp.
ta Model
10
target m aterial: AD 99.5
w ater pressure: 207 MPa
orifice diam eter: .305 mm
stand-off distance: 1.5 mm
8- abrasive: mesh 120 garnet, 3.78 g/s
traverse speed: 0.8 m m /s
6-
«■ — Exp.
cj
C
m B—— Model
o
4 -
c.
u
Q
0 .6 0 .8 1 .0 1 .2 1.4 1 .6 1.8
159
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 7
M A T E R IA L M A C H IN A B IL IT Y B Y A W J
In the past years, abrasive waterjet technology has gone from being a showcase
of AWJ have grown, so have the needs for improved machining processes. In an age of
“space-age” m aterials which demand precision manufacturing, people who once were
astonished by the fantastic cutting edge of AWJ now demand even better control of
the process. A survey report of industrial needs for waterjets(K lavuhn & Baker, 1989)
indicates that the top priority is to establish performance standards and a complete
cutting database. As a preliminary effort to address these needs, this chapter presents
m ethod. T heoretical and experim ental studies(K im & Zeng, 89; Zeng &: Kim, 89,
90, 91, 92; Zeng et al., 91; Zeng et al., 92; Hashish, 84, 86, 87, 88, 89) provide a
pilot study(Z eng &; Kim, 1989) which used the concept of “M achinability Num ber”
to certain m achining operations and conditions. In this study, the kerf cutting test is
study can then be extended to various other abrasive waterjet applications including
cutting, drilling, m illing, turning, and aeburring. The primary response in an AWJ
160
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
kerf cutting operation is the depth of cut. The machinability of a certain material is
established by the relation between depth of cut and the preset AWJ parameters as
well as the m aterial properties such as hardness, flow strength, fracture toughness,
density, etc..
The kerf cutting m odel described in Equation 6.19 relates the depth of cut to the
material properties of workpiece and the five major process parameters in a simple
manner. The way the m aterial properties appear in this equation makes it feasible
to define a single parameter to account for the combined effects of several material
( Re), which has a dimension the same as that of stress. It is formulated as:
Rp = \ -------- (7.1)
2fu, Satrra- , _a_ ' 7
3~/E (Tj
The value of R e can be determined using Equation 7.1, granted that all the m ate
rial properties are known, or more practically, by conducting erosion or kerf cutting
p _ f 7l ^ ' v Cy ^ Cr h Pw /-m
R e - {T T R ) ' ^ M (7’2)
If the value of R P is determined, the following kerf cutting model becomes appli
cable:
= Thml Pw
2670(m + m w)2D u R e
where metric units should be used for all the variables,i.e., k g / s forabrasive mass
flow rate (m ) and water mass flow rate (ni,v ), pa for water pressure (Pw) anderosion
resistance (J2e), m for nozzle diameter ( D ) and depth of cut (h), and m / s for traverse
speed (u).
cutting process.
161
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7.2 A n E m p irica l M o d el
To verify the m odel given by Equation 7.3, an empirical m odel is proposed. As the
first step, a large number of AWJ kerf cutting tests are conducted to generate the data
trends of depth of cuts in terms of the five major process parameters: water pressure
(Pw), water flow rate (rhw), focusing nozzle diameter (D ), traverse speed (u) and
abrasive flow rate (m ). W hile the target material evaluated is primarily Aluminum
6061-T6, other materials such as nylon, granite, stainless steel (316L) and alumina
ceramics (A D 99.5) are also tested. These tests are conducted at the University of
Rhode Island Waterjet Lab, equipped with a dual intensifier pump and a P a s e r
abrasive waterjet system . The test results are presented in Figures 7.1 through 7.25.
In Figures 7.1-7.5, water pressure is not the only parameter which is varied, though
it appears so, since the water flow rate is a function of water pressure for a given
orifice diameter. In Figures 7.6-7.10, the water flow rate is regulated by using orifices
Similarity is found among the data trends for each of the process parameter.
To quantify this observation, the test data with Aluminum 6061-T6 as the target
materials are analyzed with a m ultiple regression m odel based on Equation 7.3. The
D n1no
where n(J, n,[, no, 7x3 , 7z_i, 715 are regression coefficients.
analysis. The following regression coefficients are obtained: n„ = 0.0234, rzj = 1.25,
n 2 = 0.687, 7x3 = 0.343, tx .| = 0.618, 7x5 = 0.866. The correlation coefficient is 0.954
and the determ ination coefficient is 0.911. The correlation between the m odel and
162
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
u = 2 mm/s
u = 3 mm/s
40- u = 4 mm/s
u = 5 mm/s
u = 6 mm/s
30 -
U
o 20 -
£
Or
Figure 7.1 D epth of cut versus water pressure for different traverse speed.
40
■0----- m = 5.67 g/s
m = 7.56 g/s
■tt m = 9.45 g/s
30 tit = 11.34 g/s
<--s m = 13.23 g/s
s
3
U
•o 20
a,
CJ
a
Figure 7.2 D epth of cut versus water pressure for different abrasive flow rate.
163
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
nozzle diameter; 1.27 mm * Do is orifice diam eter
traverse speed: 4 mm/s
stand-off distance: 1.5 mm
50 abrasive: mesh 80 gam et, 7.56 g/s
target m aterial: A L 6061-T6
Q' — Do = .229 mm*
40 - • Do = .305 mm
Do = .356 mm
Do = .406 mm
3
u 30 - Do = .432 mm
o
•S
20 -
10 -
04
100 200 300
Figure 7.3 Depth of cut versus water pressure for different orifice diameter.
60
orifice diam eter: .305 mm
traverse speed: 4 mm/s
stand-off distance: 1.5 mm
50 -
abrasive: mesh 120 gam et, 3.78 g/s
target m aterialL: AL 6061-T6
40 -
D = .762 mm* * D is nozzle diam eter
D = 1.016mm
30 - D = 1.270 mm
D = 1.524 mm
D = 1.778 mm
20 -
10
04 —I— —l—
100 150 200 250 300
Figure 7.4 D epth of cut versus water pressure for different nozzle diameter.
164
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
------a ---- • Nylon orifice/nozzle: .457 m m /1.27 mm
stand-off distance: l.S mm
" —• — • Granite
abrasive: m esh 80 gam et, 7.56 g/s
------ 1— ■ AL6061-T6 traverse speed: 8 m m /s(N ylon),
■ SS316L 1.2 mm/s(AD 99.5), 4 mm/s for others
60
------ ■— ■ AD 99.5
3
u
40 -
■3
c.
20-
T
100 150 200 250 300
Figure 7.5 Depth of cut versus water pressure for different materials.
— 40
Ej
c
u 30
u_
O
H,
=> 20
2 mm/s
10
u = 5 mm/s
6 mm/s
0
1 2 3 4 5
Figure 7.6 D epth of cut versus water flow rate for different traverse speed.
165
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
w ater pressure: 207 MPa m = 3.78 g/s
orifice diam eter: .229,
m = 5.67 g/s
.305, .356, .406, .457 mm
5 0 - nozzle diam eter: 1.27 mm m = 7.56 g/s
traverse speed: 4 mm/s m = 9.45 g/s
stand-off distance: 1.5 mm
m = 11.34 g/s
40 - abrasive: mesh 80 gam et
target m aterial: AL 6061-T6
o 30-
o
.5
tj*
u
a 20 -
10 H
Figure 7.7 Depth of cut versus water flow rate for different abrasive flow rate.
60
orifice Diameters: .229, .305, .356, .406, .432 mm
nozzle diam eter: 1.27 mm
P = 138 MPa
traverse speed: 4 mm/s
5 0 - stand-off distance: 1.5 mm P = 173 MPa
abrasive: mesh 80 gam et, 7.56 g/s P = 207 MPa
target m aterial: AL 6061-T6 P = 242 MPa
40- P = 276 MPa
30 -
■c-s
a 20 -
10 -
—
—.--- —I------- ■
------ 1----
2 3
Figure 7.8 Depth of cut versus water flow rate for different water pressure.
166
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
60
w ater pressure : 207 M Pa
orifice diam eter: .229, .305, .356, .406, .457 mm
traverse speed: 4 m m /s
50 -
stand-off distance: 1.5 mm
abrasive: mesh 120 gam et, 3.78 g/s
target m aterial: AL 6061-T6
40
D = .762 mm*
D = 1.016 mm
* D is nozzle diam eter
u 30 D = 1.270 mm
D = 1.524 mm
D = 1.778 mm
20
10 -
Figure 7.9 D epth of cut versus water flow rate for different nozzle diameter.
80
water pressure: 207 M Pa
Nylon
orifice diam eters: .228, .305,
Granite .356, .406, .457 mm
AL6061-T6 nozzle diam eter: 1.27 mm
S S 3I6L stand-off distance; 1.5 mm
60 -
abrasive: mesh 80 g a m e t, 5.67 g/s
AD 99.5
traverse speed: 8 m m /s(N ylon),
1.2 mm/s(AD 99.5), 4 mm/s for others
cj 40 -
O
20 -
Figure 7.10 Depth of cut versus water flow rate for different materials.
167
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
u « 2 mm/s
w ater pressure: 207 M Pa
u = 3 mm/s orifice: .305 mm
50 u = 4 mm/s stand-off distance: 1.5 mm
abrasive: mesh 120 gam et, 3.78 g/s
u = 5 mm/s
target m aterial: A L 6061-T6
u = 6 mm/s
40
£
<-> 30 B-
a 20
10
0
0 .6 0 .8 1 .0 1.2 1 .4 1 .6 1 .8
Figure 7.11 D epth of cut versus nozzle diameter for different traverse speed.
60
orifice diam eter: .305 mm
traverse speed: 4 m m /s
stand-off distance: 1.5 mm
50 - abrasive: mesh 120 gam et, 3.78 g/s
target m aterial: AL 6061-T6
40 - P = 138 MPa
P = 173 MPa
P = 207 MPa
u 30 - P = 242 MPa
P = 276 Mpa
20 -
10 -
■’ 1 1 1 1 1 > 1 1 1 1
0 .6 0 .8 1 .0 1.2 1.4 1 .6 1
Figure 7.12 D epth of cut versus nozzle diameter for different water pressure.
168
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
w ater pressure: 207 MPa
orifice diam eter: .305 mm
traverse speed: 4 mm/s
50 - stand-off distance: 1.5 mm
abrasive: m esh 120 garnet
target m aterial: AL 6061-T6
40 -
5 ------a — m - 1.51 g/s
£**
Sm ..- • ■■■■■ m - 3.02 g/s
3
u 30
....■— ■ m = 4.54 g/s
o m = 6.05 g/s
ja
o. m - 7.56 g/s
a
a 20 -
10 -
— ,--------, , 1 1 1 . 1 . 1------ 1—
0 .6 0 .8 1.0 1.2 1 .4 1.6 1.8
Figure 7.13 D epth of cut versus nozzle diameter for different abrasive flow rate.
60
w ater pressure: 207 MPa
traverse speed: 4 mm/s
stand-off distance: 1.5 mm
50 - abrasive: mesh 120 gamet, 3.78 g/s
target m aterial: AL 6061-T6
40 - o— m w = 1.109 LPM
• ----- m w = 1.968 LPM
■— mw = 2680 LPM
30 - 1 '0 — mw = 3.501 LPM
b mw = 4.432 LPM
a.
e
G
20
10 -
Figure 7.14 D epth of cut versus nozzle diameter for different water flow rate.
169
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
80 -
w ater pressure: 207 MPa
orifice diam eter: .305 mm
70 - stand-off distance: 1.5 mm
abrasive: m esh 120 gam et, 3.78 g/s
60 - traverse speed: 8 m m/s(Nylon),
.8 mm/s(AD 99.5), 4 mm/s for others
E
3 50 -
1,1 1Q----- Nylon
20 - ' ♦
10 -
Figure 7.15 D epth of cut versus nozzle diameter for different materials.
50
P = 138 MPa
P = 173 MPa
P = 207 MPa
40 -
P = 242 MPa
P = 276 MPa
30 -
o
uo
20 -
10 -
orifice/nozzle: .381 mm/1.27
stand-off distance: 1.5 mm
abrasive: mesh 80 gam et, 7.56 g/s
target m aterial: AL 6061-T6
T
Figure 7.16 Depth of cut versus traverse speed for different water pressure.
170
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
m = 5.67 g/s
m = 7.56 g/s
m = 9.45 g/s
m = 11.34 g/s
m = 13.23 g/s
30
u
(M
O
£
a.
o
a
20
Figure 7.17 D epth of cut versus traverse speed for different abrasive flow rate.
60
water pressure: 207 MPa
■O m w = 1.109 LPM
nozzle diameter: 1.27 mm
stand-off distance: 1.5 mm -* m w = 1.968 LPM
u
o 30 -
a.
u
Q
20 -
10 -
Figure 7.18 Depth of cut versus traverse speed for different water flow rate.
171
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
40
a— D = .762 mm
♦— D = 1.016 mm
■----- D = 1.270 mm
30 - *— D = 1.524 ram
E D = 1.778 mm
S
u
<•*
o 20-
■5
a,
£
10
w ater pressure: 207 MPa
orifice: .305 mm
stand-off distance: 1.5 mm
abrasive: mesh 120 gam et, 3.78 g/s
target m aterial: AL 6061-T6
Figure 7.19 D epth of cut versus traverse speed for different nozzle diameter.
u 40 -
u*
O
Nylon
20 - Granite
AL 6061-T6
SS 316L
AD 99.5
0 2 4 6 8 10 12
Figure 7.20 D epth of cut versus traverse speed for different materials.
172
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
— -o — ■ u - 2 mm/s w ater pressure: 207 M Pa
■ u - 3 mm/s orifice/nozzle: .381 m m /1.27 mm
50 ■» ■
—• u = 4 mm/s stand-off distance: l.S mm
abrasive: m esh 80 gam et
• u - 5 mm/s
target m aterial: A L 6061-T6
■■■ » • u- 6 mm/s
40
u
30
o -*■
o.
t) -*
a
20 -
10 -
—I—
4 6 8 10 1122 14
Figure 7.21 D epth of cut versus abrasive flow rate for different traverse speed.
60
P = 138 MPa orifice/nozzle: .381 m m /1.27 mm
P = 173 MPa traverse speed: 4 mm/s
50 - P = 207 MPa stand-off distance: 1.5 mm
abrasive: mesh 80 gam et
P = 242 MPa
target m aterial: AL 6061-T6
P = 276 MPa
40 -
u 30 -
o
20
10 -
—r~ —T-
10 12 14
Figure 7.22 D epth of cut versus abrasive flow rate for different water pressure.
173
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60'
w ater pressure: 207 MPa
riiw = 1.109 LPM
orifice diam eter: .229,
mw = 1.968 LPM .305, .356, .406, .457 mm
50-
mw = 2.680 LPM nozzle diam eter: 1.27 mm
mw = 3.501 LPM traverse speed: 4 mm/s
stand-off distance: 1.5 mm
mw = 4.432 LPM
40- abrasive: mesh SO garnet
target m aterial: A L 6061-T6
o 30-
U-i
o
£
a,
u 20 -
a
10 -i
Figure 7.23 D epth of cut versus abrasive flow rate for different water flow rate.
60
w ater pressure : 207 MPa
orifice diam eter: .305 mm
traverse speed: 4 mm/s
50 stand-off distance: 1.5 mm
abrasive: mesh 120 garnet
target m aterial: A L 6061-T6
0 40
13----- D = .762 mm
• - D = 1.016 mm
^ 30 ■ D = 1.270 mm
U
0 o D = 1.524 mm
E
o, H D = 1.778 mm
Q 20
10
0• T T
0 2 4 6 8
Figure 7.24 Depth of cut versus abrasive flow rate for different nozzle diameter.
174
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
-a - — Nylon water pressure: 207 MPa
Granite orifice/nozzle: .457 m m /1.27 mm
stand-off distance: 1.5 mm
-+ ----- AL 6061-T6
traverse speed: 8 m m /s(N ylon),
60- ----- SS316L 1.2 mm/s(AD 99.5), 4 mm/s for others
-* ----- AD 99.5 abrasive: m esh 80 garnet
40 -
20 -
—r~
10 12 14
Figure 7.25 Depth of cut versus abrasive flow rate for different m aterials.
60
ex p
x
0
n 60
th e
N o te : h is t h e m e a su r e d d e p t h o f c u t (mm)
exp
h is t h e p r e d ic te d d e p th o f c u t (mm)
th e
Figure 7.26 Correlation between the predicted depth of cut from the empirical model
175
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To evaluate the validity of Equation 7.4 for target materials other than 6061-T6,
the data shown in Figures 7.5, 7.10, 7.15, 7.25 are normalized and reproduced in
Figures 7.27-7.31. The data trends for the different target materials are compared
to the norm alized values from Equation 7.4. A general similarity among these data
trends is observed. Therefore, Equation 7.4 can be applied to other engineering ma
inverse of the “Erosion R esistance” i?e in Equation 7.3 or the constant n0 in Equation
7.4. The value o f a m aterial’s “M achinability Num ber” is determined experim entally
= n z, r)U.si8u
,, 0.866
m p 1 .2 5 ^ 0 .6 8 7 ^ 0 .3 -1 3 '
where the constant Ca is a scale factor chosen to give N m values within a preferred
constant C„ has a nom inal unit to balance the dim ension. The values of (7., are given
The establishm ent of an AWJ cutting database requires that the “Machinability
total of 27 m aterials are selected from the entire spectrum of engineering m aterials,
176
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.0
Nylon
Granite
1.6 - AL6061-T6
SS 316L
3 AD 99.5
u
Model
1.2 -
aa,
Q
•3
0.8 -
O
Z
0 .4 -
0.0
100 15 0 200 250 300
Figure 7.27 Norm alized depth of cut versus water pressure for different materials.
2.0
..Q. Nylon
Granite
AL6061-T6
SS 316L
AD 99.5
v
o
Model
a
0.8 -
•a
c
0 .4 -
0.0
0 1 2 3 4 5
Figure 7.28 Normalized depth of cut versus water flow rate for different materials.
177
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.0
Nylon
Granite
AL6061-T6
SS 316L
3 AD 99.5
o Model
3.
a
a
-o
0.8 -
a
Z
0 .4 -
0.0
0. 6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Figure 7.29 Normalized depth of cut versus nozzle diameter for different materials.
10 -
Nylon
Granite
AL6061-T6
3
CJ SS 316L
AD 99.5
Model
Q
■3
0 2 4 6 8 10 12
Figure 7.30 Normalized depth of cut versus traverse speed for different materials.
178
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission
------e — Nylon
----- Granite
—* - — AL 6061-T6
----- SS316L
— AD 99.5
------a ----- Model
M---------- B
X •m
v /
/
/
4
0 2 4 6 8 10 12 14
Figure 7.31 Normalized depth of cut versus abrasive flow rate for different m aterials.
179
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
varying from very soft to extrem ely hard m aterials. To determ ine the “M achinability
Num ber” for each selected m aterial, three test cuts are m ade and the process param
eters are substituted into Equation 7.5 to calculate the value o f N m. The average
The m achinability number can be used in various applications. The most im por
tant application is in the selection of optional AWJ process parameters for kerf cutting.
In a typical AWJ kerf cutting operation, the workpiece m aterial and thickness are
readily determ ined. Typically, m ost process param eters are set using rule-of-thumb
values with the traverse speed used as the variable to be adapted to the given m a
terial and thickness. The optional traverse speed is determ ined using the following
equation:
where q is defined as a quality level parameter. The value of q can be chosen between
1 and 5, depending on the desired quality level. Exam ination of the wall surface of a
separation cut (i.e. maxim um depth of cut = workpiece thickness) has revealed that
striation marks are initiated at about 1/3 distance, m easured from the top, of the
total depth. Therefore, to estim ate the traverse speed required for a sm ooth cut, 3
times of the workpiece thickness should be substituted into Equation 7.6 for the value
of h, i.e. q = 3. The values of q for five different quality levels are defined as follows:
Quality Description
Levels
q = 1 Criteria for separation cuts. Usually, q > 1.2 should be used.
q = 2 Rough surface finish with striation marks at the lower half surface.
q = 3 Sm ooth/rough transition criteria. Slight striation marks may appear.
q = 4 Striation free for m ost of engineering m aterials.
q = 5 Very sm ooth surface finish.
It should be always kept in mind that selection of a high quality level results in
180
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hot P r e s s e d S i3N 4 11.1
AD 9 9 .9 223 1.6
B 4C ' 77 7 7 7 7 7 % 4.2
Ti3B 2 m m & 4.3
AD 9 0 '/77/77777777777X10.3
SiC w zm zzzzm 12.6
AD 9 9 .5 7/7/777/7/7777^7777 13.1
AD 9 4 7777777777777777777!?,1 17.3
AD 8 5 777777777777777777m17.3
Oil H a r d e n e d S te e l 777777777777777777777//777777777A so.4
S S 304 7////77/7/77/7/7/777777777777777A 81.9
S S 3 1 6L 77/77//777777777777777777////7777X83.1
ASTM A 36 S te e l W ////////////7 ///7 7 7 ///7 ///////A VJ&
C opper 77//77777777777Z////7////////7/7m 110
T itanium v/?///7/////7//7/////77/////mm\ 115
Z incalloy 777//77777777/7////////////////777Z7\ 1 3 6
AL 6 0 6 1 -T6 213
G ra n ite Z///////////////Z77/77////7//7///Z?V/m77\2Zl
A sp h alt C o n c re te V/////////////7//7777777777777ZZZmZZ!m 461
L ead 7////7/////////7/////7////////7//////////Z////A 490
W hite M arble '77777777777777777777777777777777//////////////A 5 3 5
Nylon W /777777777777777777777777777777777777777777A5 3 8
G la s s V///////////////////////////////////7Z/////77/77A^
P le x ig la s 690
G ra p h ite 7/////Z///7///77777//////////7/77777777//777/7777AW
P o ly p ro p y len e 777777777777777777777777777777777777777777777777777A9 8 5
P in e W ood 77777777777777777777777777777//////////////7//////////7//77A 2 6 3 7
1 10 100 1000 10000
M a c h in a b ility N u m b e r s
181
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
slow cutting speeds, and therefore increases cost. Selection of a value of q greater
than 5 is unnecessary.
For general cutting purposes, the following rule-of-thum b parameter settings are
suggested: water pressure 242 M Pa (35 kpsi), water flow rate 3.32 1pm (0.877 gpm)
(i.e. orifice diam eter 0.381 m m (0.015”), nozzle diameter 1.016 m m (0.040”) and abra
sive (garnet) flow rate 7.56 g / s ( l lb /m in ). In this case, Equation 7.6 reduces to
„ = (7.7)
where c = 2 for the M etric unit system and 24.7 for the Inch unit system .
Equation 7.7 has been successfully applied in the following two case studies:
As shown in Figure 7.32, the value of N m for AD 94 is 17.3. Using the rule-of-
thum b values for all th e other process parameters, the traverse speed is determined
w ith Equation 7.7 to be 0.69 m m /s for a separation cut and 0.19 m m /s for a smooth
cut(quality level = 3 ) . As shown in Figure 7.33 (a), the cut quality is as expected.
the appropriate N m is estim ated to be 82. Equation 7.7 yields a 1.63 m m /s traverse
speed for a separation cut and 0.46 m m /s for a sm ooth cut. These cuts are also very
Equations 7.6 k. 7.7 have not incorporated the effects of stand-off distance, abra
sive type and abrasive size. As a general trend, the depth of cut is reduced as the
stand-off distance increases. However, variation of the stand-off within a small dis
tance, say, 3 mm, does not cause significant changes in the depth of cut. On the
H ashish(1986, p297), variation of abrasive size within the ordinary range (mesh aG -
182
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 7.33 Cut surfaces of AD 94 (a) and an unidentified steel plate (b).
183
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150) has little effect on the attainable m aximum traverse speed.
result in great difference in depth of cut. The derived m odel has not incorporated
this effect because (1) the effect of abrasive type on the material removal m echanism
is not well understood and (2) the abrasive type is difficult to quantify and therefore
to incorporate into an analytical equation. However, Equations 7.6 and 7.7 are still
reliable when abrasives other than garnet are used. For instance, a recent study(K im
et al., 1993) has revealed that, in cutting a brittle m aterial, e.g. AD 85, alumina
abrasive produced 5 tim es the depth of cut as that achieved with garnet abrasive,
but when cutting a ductile m aterial such as AL 6061-T6, little difference between
these two abrasives was observed. Therefore, as a preliminary estim ation, Equations
7.6 and 7.7 can be directly applied for a ductile material cut with alumina abrasive.
- 5 to account for the greater power of alumina abrasive in cutting brittle materials.
For other abrasive m aterials, the following m odification procedure can be used: (1)
Conduct a few cutting tests with the given abrasive material; (2) Calculate the value
of N m using Equation 7.5; (3) Divide the new value of N m by that given in Figure
7.32, leading to a correction factor for this specific abrasive; (5) M ultiply Nm by the
An empirical m odel for AWJ kerf cutting has been form ulated using m ultiple re
gression analysis based on a large quantity of experim ental data. This model success
fully relates the depth of cut to the major AWJ process parameters. A “M achinability
184
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
simplifies the param eter selection procedure and elim inates or at least minimizes
185
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C H AP TER 8
CONCLUSIONS A N D RECOMMENDATIONS
The SEM study on the AWJ eroded samples of stainless steel 304 and alumina
(A D 99.5) has led to the conclusions that (1) Ductile m aterials are eroded by a cutting
ceramics, are due to a com bination of plastic flow and network cracking (intergranu-
lar cracking for polycrystalline ceramics); (3) AWJ kerf cutting process is associated
with abrasive particle impacts at glancing angles; (4) Steps exist in the AWJ cutting
front, but their existence doesn’t significantly change the particle impact angles.
The network cracking has been attributed to stress wave induced fractures. An
elasto-plastic m odel has been proposed to include the material removal by both the
fractures and plastic flow. A crack network model has been derived to evaluate the
fractured volum e by relating the required fracture surface energy to the input stress
wave energy. T he stress wave energy for a normal im pact has been evaluated w ith a
modified H utchings’ equation. T he stress wave energy for a low incidence impact has
been derived in this study. The crack network m odels for these two cases are thus
obtained. A com plete elasto-plastic m odel for a single im pact at normal incidence
includes the crack network m odel and B itter’s deformation wear model. For a low
incidence im pact, it includes the corresponding crack network model and Finnie’s
m icrocutting m odel. The obtained elasto-plastic m odels have been verified by erosion
experim ents. Good correlation has been achieved. It is also found that the erosion
rate is strongly correlated with the grain size and fracture energy of target materials.
The study of the AWJ energy dissipation phenom ena concludes that (1) The
striation marks can be characterized by parabolic curves; (2) The cutting efficiency
186
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
along the cutting front is a function of the slope of the striation curve; (3) The critical
jet exit angle decreases as the workpiece thickness increases while its dependence on
An equation to predict the depth of cut for AWJ cutting operations was derived.
This equation correlates well w ith AWJ kerf cutting experim ents provided that process
practical applications. The empirical m odel provides a relation between the depth
of cut and the major AWJ process parameters. The “M achinability Number” intro
engineering m aterials by AWJ cutting, which also greatly simplifies the parameter se
lection procedure and eliminates or at least minimizes tim e-consum ing and som etim es
Although these studies have yielded informative and useful results, further studies
are recommended for more accurate m odels and better m ethodology. Chapter 3 has
revealed th at, under the AWJ condition, erosion of the target material (AD 99.5)
includes contributions from intergranular cracking and plastic flow. However, the
proportionality between these two m aterial removal com ponents needs to be quanti
fied. The im pact damage on brittle materials other than AD 99.5 also needs to be
investigated. The steps formed on the AWJ cutting front has been noticed. Yet the
and experim ental verification. The cause of network cracking has been attributed
to impact induced stress waves. However, this hypothesis has not been supported
ducted using the same target material and the dam age m orphology be compared to
that caused by the particle impact under the AWJ condition. Photoelastic technique
may also be used on a transparent polvcrystalline ceramic to verify the stress wave
187
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
propagation and interaction with grain boundaries during the event of particle im
pact. The derived elasto-plastic model does not incorporate the effects of the type
and shape of the particle. The effects of the water droplet impact are also neglected.
To refine this m odel, these factors should be considered. Further m odeling efforts are
also needed to extend the current m odel to cover the cases where the im pact angle has
an interm ediate value (e.g., 45°). For non-crystalline brittle materials such as glass,
a m aterial flaw distribution parameter needs to be defined to replace the grain size
in the current model. The energy dissipation phenom ena in AWJ cutting have been
phenom ena is also neccessary. The striation marks are related to the instability of the
AWJ cutting process, which can be the topic of an extensive investigation. Although
the derived equation for depth of cut has been primarily applied to brittle materials,
it can be extended to cover ductile materials. The empirical equation for depth of
cut can be expanded to incorporate the effects of the typ e and size of abrasive. The
definition and application of the “Machinability Num ber” can be also extended to
188
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX A
DETERMINATION OF AWJ PARTICLE VELOCITY
According to Equations ( 8 ) and (10) in the reference by Hashish (1989, p221), the
1 + R V Pw
where Pw is water pressure, pw water density, R ratio of abrasive/w ater mass flow
coefficient. The values of 77, C„ and C v can be obtained from Figures (A .l) and (A.2)
0 99 -
0.94 -
0.92 -
0.9 7 *
c
0 .9 6 -
om-
ui
0.9 5 - 0.6 ft -
0
c 0.84-
b?
0.94 -
c
0.9 2 - E
0
3
0.91 •
0.76 •
0.74 -
0.89 0.72
50 150 250 350 450 120 140 160 180 200 220 240 160 260 320
189
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX B
EVALUATION OF COMPLEX INTEGRALS
Following the procedure described by Miller & Pursey (1954), the integrals in
/= f° W (B.l)
The integrand <f> has singularities at the points £ = 1 , /i and p;), which are such
that 1 < (i < p-i. The range of integration is divided into six parts such that
M rn 'P a -« rp:\+° n
ii = / (j>dZ,i2 = / <j>d£,i3. - / <j>d£,iA = / (f>d£,i5 = / <j>d£,i6 = <$>d(
J (J *1 •'A* vp3—<5 ''P3+(5 •'I
(B.2)
where 5 is a sm all, F a large positive number. Each of the com ponent integrals is
evaluated along the real axis with the exception of F[, for which the path is chosen to
be a small sem i-circle above the pole £ = p3. (The contributions from the semi-circle
i.e.
m = Z M ( - nT + 1 t t z Jz m (B-3)
rc=0 »=1 P^J
where
A “ = i2tti
r - Jcp
I (<z — p.t) + dz (B A )
190
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
where cp represents a closed path.
[ (z - p 3)n~l4>(z)dz = 0
J cp
m = £ M ( - » )” + 7 % (S-6)
n=0 Z - P3
i, = /
•/cir
m i i = n=(J
£ -*» / ( f
Jcir
- + C, / -
Jar £ —
~~ ( B .7 )
The clockw ise semi-circle path above £ = p,3 is represented by the equation £ =
r d£ i5e‘l . o\
L i ^ r r L ( B -8 )
/ (t-nT tt = f {6e"fi(Se'1) ( B .9 )
Vn+l
= -------- [1 — cos(n + 1W] (B.10)
n + 1
= 25, (B .U )
i , = 26 A a - triCi ( B . 12 )
Using the Residue Theorem, C\ — RESI DU E{( j> {p i)} . Since 26A„ is a real
number, therefore
191
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
/•OO fl ft*
I m / (f)d£ = I m / <j)d£ + I m / <j)d£ — irRESIDUE{(j>[p:j)} (B .14)
Jo Jo J 1
192
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A P P E N D IX C
T h is i s a s a m p le MathCAD p r o g r a m , w h ic h s o l v e s t h e f o l l o w i n g
e q u a tio n f o r a s o l u t i o n o f S t o b e r e a l and S > u ,
c o r r e s p o n d in g t o P o i s s o n 's r a t i o = 0 .2 2 .
2 - (1 - o )
V- :z i.e . ^ = 1 .6 6 9 0 4 6
1 - 2 -cr
G iv e n a g u e s s v a l u e S .‘ = 1 . 1
g iv e n
to l := 0 .0 0 0 0 0 1 (to le r a n c e )
4
2 2 4 2 2 2"
,2 ' s " F * 16- S • S - 1 s - V- .
(e q u a tio n t o b e s o lv e d )
S := f i n d ( S )
S = 1 .8 2 5 2 8 2 (s o lu tio n )
C h eck
r 2 2i
[2 -3 - n J = 2 2 6 . 074142
4 f 2 I f " 2 2
16- S • [S “ 1J • Ls ~ ^ = 2 2 6 .0 7 4 1 4 2
193
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
P ro g ra m 2 N u m e r ic a l E v a lu a tio n o f th e F irst In teg ra l
T h is i s a s a m p le MathCAD p r o g r a m , w h ic h e v a l u a t e s t h e
f o l l o w i n g i n t e g r a l f o r P o i s s o n ' s r a t i o = 0 .2 2
1 - <7
2 - ------------- i.e . jx = 1 .6 6 9
1 - 2 • <7
p := 1 .8 2 5 3 ( th e in te g r a n d p o le )
*1
2 2 2 2
2 -i s - 1- > s - jx - jx - 2 •S J_
l := ds
1
2 2 2 2
_1_2- s - jx 4- s • s - 1- s - |X .
(T h e 1 s t te r m o f t h e i n t e g r a l )
i = - 0 .2 4 9 3
1
ft*
2 2 2
2- s - 1- s - jx — |jx - 2 'S _
l := ds
2
2 2 2 2
2•s - jx 4- S s - l-> s - jx J
(T h e 2nd te r m o f t h e i n t e g r a l )
i = - 0 .2 0 9 8 + 0 . 1 7 9 2 i
2
194
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
2 2
DO : = U 2 V ] -4 . P - i - 'j P " V- .
dp
2 2 2 2
NU : = P ' _2-->|p - 1-r p - |X - t a - P ,
NU
RES : =
DO
i.e . RES = - 0 . 9 3 3 6 ( r e s i d u e a t p = 1 .8 2 5 3 )
-H
p = 3 .1 1 2 2 (s o lu tio n )
1
195
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
P rogra m 3 N u m e rica l E v a lu a tio n o f th e S eco n d In teg ra l
T h is i s a s a m p le MathCAD p ro g ra m , w h ic h e v a l u a t e s t h e
f o l l o w i n g i n t e g r a l f o r P o i s s o n ' s r a t i o = 0 .2 2
1 - <7
[a : = 2 - ------------- i.e . (j. = 1 .6 6 9
1 - 2-cr
p := 1 .8 2 5 3 (th e in te g r a n d p o le )
2 2 2
s - jx
d s
2 2 2 2 2
S - (A j 4- s s - 1- s - p. J
(T he 1 s t te r m o f t h e i n t e g r a l )
l = 0 .8 1 6 3 1
1
2 2 2
s - n -|A
i := ds
2
2 2 2 2 2
2 -s - p. 4- s s - 1- s - n JJ
(T he 2n d te r m o f t h e i n t e g r a l )
i = - 0 .4 5 2 3 + 0 .1 1 4 1 i
2
196
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
p "
2
[ 2 2' 2 2 2 2
DO : = .[2 -P ~ F . - 4 - p •\ p - l-> P “ F .
dp
2 2 2
NU : = WP - F 'F J
NU
RES : =
DO
i.e . RES = - 0 . 0 9 4 ( r e s i d u e a t p = 1 .8 2 5 3 )
p : = Im + Im - ir-RES
2
p = 1 .2 2 5 8 (s o lu tio n )
2
197
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
P ro g ra m 4 N u m e rica l E v a lu a tio n o f th e T h ird In te g ra l
T h i s i s a s a m p le MathCAD p r o g r a m , w h ic h e v a l u a t e s t h e
f o l l o w i n g i n t e g r a l f o r P o i s s o n ' s r a t i o = 0 .2 2
l ~ o
2 ------------- i.e . (A = 1 .6 6 9
1 - 2-0
p := 1 .8 2 5 3 (th e in te g r a n d p o le )
2
4- s s - 1
ds
2
4- s ^
(T h e 1 s t te r m o f t h e i n t e g r a l )
i = 0 .1905i
1
*F
2 2 2
ja - 4- s s - 1
ds
2 2
4- s ■\ s - 1-
(T h e 2n d te r m o f t h e i n t e g r a l )
i = - 0 .1 6 0 3 - 0 .3836i
2
198
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
DO : = - 4- p • > p - 1- P "
dp
DO = - 2 1 . 8 9 2 1
2 2
NU I* - 4- P J p - 1
NU
RES := —
DO
i.e . RES = 0 .7 3 5 3 ( r e s i d u e a t p = 1 .8 2 5 3 )
p = - 2 .5 0 3 (s o lu tio n )
3
199
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
P rogra m 5 N u m erica l E v a lu a tio n o f th e F ou rth In teg ra l
T h is i s a s a m p le MathCAD p r o g r a m , w h ic h e v a l u a t e s t h e
f o l l o w i n g i n t e g r a l f o r P o i s s o n ' s r a t i o = 0 .2 2
1 - cr
V
- :: 1 - 2-<7
i.e . [J. = 1 .6 6 9
p := 1 .8 2 5 3 ( th e in te g r a n d p o le )
2 2 2 2 2 2
s • s - 1- s - n + JJL - 2 s
ds
2 2 2 2 2
2•s - [X 4- s ^ s - l - \ s - [x .
(The 1 s t te r m o f t h e i n t e g r a l )
i = - 0 .0 0 3 4
1
2 2 2 2
s • s - 1- s -fx + [jx - 2- s _
d s
2 2
S - l-> s - jx j j
(T he 2nd te r m o f t h e i n t e g r a l )
i = - 0 .3 6 5 3 - 0 . 0 4 4 3 i
2
20 0
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
2 2 2 2
DO := — [, P - M - 4- p • > p - 1 m P - N
dp
r *
CO
2
CM
to
2
NU := p • 2 ’ ' p - l - i p - n + jx - 2- p JJ
NU
RES : =
DO
i.e . RES = 0 .2 4 6 7 ( r e s i d u e a t p = 1 .8 2 5 3 )
p : = Im + Im fi - 1T- RES
4 M L2
p = - 0 .8 1 9 3 (s o lu tio n )
4
201
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
P rogram 6 M u ltip le R e g r e ssio n A n a ly sis
LH := READPRN(LH) ( r e a d i n g exp. d a t a of
dept h of c u t s )
- 1
b : = ( X T ’ X) • ( XT - LH) ( co mpu t i ng the m u l t i p l e
r e g r e s s i o n c o e f f i c i e n t s )
H := 249 (number of d a t a s e t s )
-1.63
1.25
0.687
b = 0.618 ( c o mp ut i n g r e g r e s s i o n
0.866 c o e f f i c i e n t s )
0.343
= 1 0 = 0.0234
= b = 1.25
( r e s u l t s of
= b = 0.687 r e g r e s s i o n
c o e f f i c i e n t s )
= 0.343
= - b = 0.618
= - b n = 0.866
202
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
<1> <2> <3> <4> <5>
T HL H : = b + b •X + b ‘X + b -X + b ' X + b •X
0 1 2 3 4 5
( c o m p u t i n g t h e p r e d i c t e d v a l u e s of dep t h of c u t s
in the l og. form)
SSE
SSE = Z :[< ( LH - THLH) MS E : =
N - 2
MS E = 4 . 1 2 2 - 1 0
i : = 0 . . N - 1 2"
SST : = 'l h - mean(LH)
. i _
SSTO : = 2s ST
RS Q : = 1
SSTO
RS Q = 0. 911 ( c o e f f i c i e n t of d e t e r m i n a t i o n )
203
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BIBLIOGRAPHY
Ambrosini, L. and S. Bahadur. “Erosion of AISI 4140 S teel.” Wear, Vol. 117, (1987):
37-48.
Anand, K.. “Flux Effects in Solid Particle Erosion.” Wear, Vol. 118,(1987): 243-257.
Barker, L.M .. “Short Rod K \ c M easurements of AUOwT Sym posium on the Fracture
Mechanics of Ceramics, V. 3, Plenum Press, New York, (1978): 483-494.
Bellman, Jr. and Alan Levy. “Erosion M echanism in D uctile M etals.” Wear, Vol. 70,
(1981): 1-27.
Bitter, J. G. A.. “A Study of Erosion Phenomenon — Part I.” Wear, Vol. 6, (1963):
5-21.
204
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Bitter, J. G. A.. “A Study of Erosion Phenomenon — Part II.” Wear, Vol. 6, (1963):
169-190.
Brown, R. and Jeff W . Edington. “Occurrence of M elting During the Solid Particle
Erosion of Copper.” Wear, Vol. 73,(1981): 193-200.
Brown, R. and Jeff W . Edington. “Mechanisms of M aterial Loss during the Threshold
Period of Erosion by Solid P articles.” Wear, Vol. 77, (1982): 347-353.
Brown, R. and Jeff W. Edington. “Erosion of Copper Single Crystals under Condi
tions of 30" Incidence.” Wear, Vol. 79, (1982): 335-346.
Burnham, C.. “Abrasive Waterjets Come of A ge.” Machine Design (May 10, 1990).
Burnham, Chip. “On the Beams: Lasers and Abrasive W aterjets.” Modern Machine
Shop (February, 1992): 81-87.
205
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chalmers, E. J.. “Effect of Param eter Selection on Abrasive W aterjet Perform ance.”
Proceedings of the 6th Am ericam W ater Jet Conference, W JT A , Houston, Texas,
(A ugust 24-27, 1991): 345-354.
Chao, C.C.. “Dynam ical Response of an Elastic Half Space to Tangential Surface
Loadings.” Journal of Applied Mechanics (Septem ber, 1960): 559-567.
Chao, C .C ., H .H. Bleich, and J. Sackman, “Surface Waves in an Elastic Half Space.”
Journal of Applied Mechanics (June 1961): 300-301.
Cherry, J.T ., “T he Azim uthal and Polar Radiation Patterns Obtained from a Hori
zontal Stress Applied at the Surface of an E lastic Half Space.” Bulletin of the Seis-
mological Society of America, Vol. 52, No. 1, (January, 1962): 27-36.
Cheung, John S. T.. “Basic Equations for Water Jet O perations.” Proceedings of the
1st Asian Conference on Recent Advances in Jetting Technology, Singapore, (M ay
7-8, 1991): 37-42.
Christm an, T im othy and Paul G. Shewmon. “Erosion of a Strong Aluminum A lloy.”
Wear, Vol. 52, (1979): 57- 70): .
Cole, J. and J. Huth, “Stresses Produced in a H alf Plane by Moving Loads.” Journal
of Applied Mechanics (Decem ber, 1958): 433-436.
Engel, Peter A .. Impact Wear o f Materials. Am sterdam -O xford, New York: Elsevier
Scientific Publishing Company, 1976.
Eringen, A.C. and J.C. Sam uels, “Impact and M oving Loads on a Slightly Curved
Elastic Half Space.” Journal o f Applied Mechanics (Decem ber, 1959): 491-498.
206
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Evans, A. G.. “Im pact Damage Mechanics: Solid Projectiles.” Treatise on Materials
Science and Technology, Vol. 16, Erosion, C. M. Preece, Ed., New York: Academic
Press, (1979): 1-67.
Evans, A. G., M. E. Gulden, and M. R osenblatt. “Im pact Dam age in Brittle Materials
in the E lastic-P lastic Response R egim e.” Proceedings of the Royal Society of London,
Seriers A, Vol. 361, (1978): 343-365.
Evans, A. G. and T. R. Wilshaw. “Dynam ic Solid Particle Dam age in Brittle M ate
rials: an A ppraisal.” Journal of M aterial Science, V .12, (1977): 97-116.
Finnie, Iain, “Erosion of Surfaces by Solid P articles.” Wear, Vol. 3, (1960): 87-103.
Finnie, Iain. “Erosion by Solid Particles in a Fluid Stream .” Sym posium on Erosion
and Cavitation, ASTM S T P # 3 0 9 , (1962): 70-82.
Finnie, I. and D. H. M cFadden, “On the Velocity Dependence of the Erosion of D uctile
Metals by Solid Particles at Low Angles of Indence.” Wear, Vol. 48, (1978): 181-190.
Finnie, I. and Y. H. Kabil. “On the Formation of Surface Ripples during Erosion.”
Wear, Vol. 8, (1965): 60- 69.
Franz, Norman C.. “Fluid Additives for Improving High Velocity Jet C utting.” P ro
ceedings of the 1st International Sym posium on Jet Cutting Technology, Conventry,
207
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
England, (April 5-7, 1972): A 7-93/A 7-104.
Gat, Nahum and W iden Tabakoff. “Some Effects of Temperature on the Erosion of
M etals.” Wear, Vol. 50, (1978): 85-94.
Graff, Karl F.. Wave Motion in Elastic Solids Oxford: Clarendon Press, 1975.
Hashish, M.. “Steel Cutting with Abrasive W aterjets.” Proceedings of the 6th In
tem atioanl Sym posium on Jet Cutting Technology, BH RA , the University of Surry,
U.K ., (April 6-8, 1982): 465-487.
Hashish, M.. “A Modeling Study of M etal Cutting with Abrasive W aterjets.” Journal
of Engineering Materials and Technology, ASME, Vol. 106, (January, 1984): 83-100.
208
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hashish, M.. “Abrasive-W aterjet Cutting Studies.” Proceedings o f the 11th Con
ference on Production and Technology, Carnegie-Mellon University, Pittsburgh, PA,
(1984): 101-108.
Hashish, M.. “An Improved M odel of Erosion by Solid Particle Im pact.” Proceedings
of the 7th International Conference on Erosion by Liquid and Solid Impact, Cam
bridge, UK, (Septem ber 7-10, 1987): 66-1/66-9.
Hashish, M.. “D ata Trends in Abrasive Waterjet M achining.” presented at Autom ated
Waterjet Cutting Processes, SM E, Detroit, Michigan, (M ay 10-11, 1988.
Hashish, M.. “Turning, Milling, and Drilling with Abrasive-W aterjets.” Proceedings
o f the 9th Intem atioan l Symposium on Jet Cuttinq Technoloqy, BH R A , Sendai, Japan,
October 4-6, 1988): 113-131.
Hashish, M.. “A M odel for Abrasive-W aterjet (AW J) M achining.” Journal of Engi
neering Materials and Technology, ASME, Vol. I l l , (April, 1989): 154-162.
Hashish, M.. “An Investigation of Milling with Abrasive-W aterjets.” Journal o f En
gineering f o r Industry, ASME, Vol. I l l , (May, 1989): 158-166.
Hein, Le Khac and Paul G. Shewmon. “Effects of Hardness on the Solid Particle
Erosion M echanisms in AISI 1060 Steel.” Wear, Vol. 89, (1983): 291-302.
209
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hitchcox, Alan L.. “Vote of Confidence for Abrasive Waterjet C utting”, Metal
Progress (July, 1986): 33-42.
Hooker, Stephen V. and W illiam F. Alder. “Particle Impact Regimes in Single Crys
tals.” Sym posium on the Fracture Mechanics of Ceramics, Vol. 3, New York: Plenum
Press, (1978): 333-347.
Hunter, S. C.. “Energy Absorbed by Elastic W aves during Im pact.” Journal of the
Mechanics and Physics of Solids, Vol. 5, (1957): 162-171.
Hutchings, I. M.. “Deform ation of Metal Surfaces by the Oblique Impact of Square
P lates.” International Journal o f Mechanical Sciences, Vol. 19, (1977): 45-52.
H utchings, I. M.. “Some Com m ents on the Theoretical Treatment o f Erosive Particle
Im pacts” , Proceedings of the 5th International Conference on Erosion by Solid and
Liquid Impact (Septem ber 3-6, 1979): Paper 36.
Hutching, I. M.. “Energy Absorbed by Elastic Waves during Plastic Im pact.” Journal
of Physics D: Applied Physics, Vol. 12, (1979): 1819-1824.
210
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the Removal of Surface M aterial by Spherical Projectiles.” Proceedings of the Royal
Society of London, Seriers A, Vol. 348, (1976): 379-392.
Jahanm ir, S.. “T he Mechanics of Subsurface Dam age in Solid particle Erosion.”
Wear, Vol. 61, (1980): 309- 324.
Johansson, Stefan, Fredric Ericson and Jan- Ake Schweitz. “Solid Particle Erosion
— A Statistical M ethod for Evaluation of Strength Properties of Sem iconducting
M aterials.” Wear, Vol. 115,(1987): 107-120.
Kim , T. J., H. A. Costantino, and J. Zeng. “An Experim ental Investigation of N oz
zle Wear in the Abrasive W aterjet System .” to be subm itted for presentation and
publication at the 7th American Water Jet Conference, 1993.
Kim, T. J., J. G. Syvia, and L. Posner, “Piercing and C utting o f Ceramics by Abrasive
W aterjet.” Preceedings of the W inter Annual Meeting of the Am erican Society of
Mechanical Engineers, the Production Engineering D ivision, Miami Beach, Florida,
(Novem ber 17-22, 1985): 19-24.
Klavuhn, J.G ., and C.B. Baker. “Industrial Needs Survey — Waterjet C u ttin g.”
Proceedings of the 5th Am erican Water Jet Conference, W JTA , Toronto, Canada,
(A ugust 29-31, 1989): 263-273.
211
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Kolsky, H.. Stress Waves in Solids. New York: Dover Pub., 1963.
Lamb, H.. “On the Propagation of Tremors over the Surface of an Elastic Solid.”
Philosophical Transactions o f the R oyal Society o f London, A203, (1904): 1-42.
Laitone, J. A.. “Aerodynam ic Effects in Erosion Process.” Wear, Vol. 56, (1979):
239-246.
Lankford, J.. “Com pressive Microfracture and Indentation Dam age in AUO^F S ym
posium on the Fracture Mechanics of Ceramics, V .3, Plenum Press, New York, (1978):
245-255.
Levy, Alan V.. “T he Role of Plasticity in Erosion.” Proceedings o f the Fifth Interna
tional Conference on Erosion by Liquid and Solid Impact, Cambridge, UK, (Septem
ber 3-6, 1979): 39-1/39-10.
Levy, Alan V. and Pauline Chik. “The Effects of Erodent Com position and Shape
on the Erosion o f Steel.” Wear, Vol. 89, (1983): 151-162.
Levy, Alan V. and Paul Yan. “Erosion of Steels in Liquid Slurries.” Wear, Vol. 98,
(1984): 163-182.
Lynn, Randall S.. “On the Particle Size Effect in Slurry Erosion.” Wear, Vol. 149,
(1991): 55-71.
M alkin, S.. “Correlation between Solid Particle Erosion of M etals and Their M elting
Energies.” Wear, Vol. 68, (1981): 391-396.
212
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Marshell, D. B ., B. R Lawn, and A. G. Evans, “E lastic/P lastic Indentation Damage
in Ceramics: the Lateral Crack System .” Journal of the A m erican Ceramic Society,
Vol.65, N o.11, (November, 1982): 561-566.
M ashloosh, K . M. and T.S. Eyre. “Effect of Attack Angle in Abrasive Wear.” Metallic
Materials (Inst M et), Vol. 2, No. 7, (July, 1986): 426-430.
Mason, Frederick. “Water & Sand Cut It.” American M achinist (October, 1989):
85-95.
M atsui, S., H. M atsumura, Y. Ikem oto, K. Tsukita, and H. Shim izu. “High Precision
Cutting M ethod for Metallic Materials by Abrasive W aterjet.” Proceedings o f the 10th
Intem atioanl Sym posium on Jet Cutting Technology, BH RA , Am sterdam , Holland,
(October 31 - November 2, 1990): 263-278.
Mehrotra, P.K ., G.A. Sargent, and H. Conrad, “A Model for the M ultiparticle Ero
sion of B rittle Solids by Spherical Particles.” Proceedings of the 5th International
Conference on Erosion by Liquid and Solid Impact, Cambridge, England, (Septem ber
3-6, 1979): 127-145.
Miller, D. S. and E. J. Bloomfield. “The Future for Abrasive Jet C utting.” Proceedings
o f the 6th A m e ric a m Water Jet Conference, W JTA, H ouston, Texas, (August 24-27,
1991): 561-574.
Miller, G .F. and H. Pursey. “On the Partition of Energy between Elastic Waves in
a Semi-Infinite Solid.” Proceedings of the Royal Society o f London, A 233, (January,
1955): 55-69.
M ills, D. and J. S. Mason. “Particle Size Effects in Bend Erosion.” Wear, Vol. 44,
(1977): 311-328.
Misra, Ambrish and Iain Finnie. “On the Size Effect in Abrasive and Erosive Wear.”
Wear, Vol. 65, (1981): 359-373.
Misra, Am brish and Iain Finnie. “Correlations between T w o-B ody and Three-Body
Abrasion and Erosion of M etals.” Wear. Vol. 68, (1981): 33-39.
213
with permission of the copyright owner. Further reproduction prohibited without permission
Morrison, C. T ., J. L. Routbort, and R. 0 . Scattergood. “Solid Particle Erosion of
M ullite.” Wear, Vol. 105, (1985): 19-27.
Muller, Stefan. “Robots Optim ise W aterjet Cutting at Ford G enk.” The Industrial
Robot (June, 1988): 97-100.
N aim , M. and S. Bahadur. “T h e Significance of the Erosion Param eter and the
M echanisms of Erosion in Single- Particle Im pacts.” Wear, Vol. 94, (1984): 219-232.
Naylor, M. G. S. and Page, T. F.. “T he Effect of Temperature and Load on the Inden
tation Hardness Behavior of Silicon Carbide Engineering Ceramics.” Proceedings of
the 5th International Conference on Erosion by Solid and Liquid Impact, Cambridge,
UK, (Septem ber 3-6, 1979): Paper 32.
Ninham , Andrew. “The Effect of M echanical Properties on Erosion.” Wear, Vol. 121,
(1988): 307-324.
Norwood, J. A.. “New A daptations and Applications for Waternife C utting.” Proceed
ings of the 7th Intem atioanl Sym posium on Jet Cutting Technology, BHRA, O ttawa,
Canada, (June 26-28, 1984): 369-388.
Pain, H. J.. The Physics of Vibrations and Waves. New York: John W iley h Sons
Ltd, 1968.
214
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Posner, Louis A. and T. J. Kim. “M achining of Ceramics by Abrasive W aterjet.”
International W ater Jet Symposium, Beijing, China, (Septem ber 9-11, 1987): 4-45/4-
58.
Preece, C., editor. “Treatise on Materials Science and Technology.” Erosion, V ol.16,
Academic Press, New York, 1979.
Rickerby, D.G. and N.H . M acM illan, “M echanisms of Solid Particle Erosion in Crys
talline M aterials.” Proceedings of the 5th International Conference on Erosion by Solid
and Liquid Impact, Cambridge, UK, (Septem ber 3-6, 1979): Paper 29.
R itter, John E .. “Erosion Dam age in Structural Ceram ics.” Materials Science and
Engineering, Vol. 71, (1985): 195-201.
215
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
R itter, J. E., P. Strzepa, K. Jakus, L. Rosenfeld, and K. J. Buckman. “Erosion
Dam age in Glass and Alum ina.” Journal of the Am erican Ceramic Society, Vol. 67,
No. 11, (Novem ber, 1984): 769-774.
Sharma, Sunita. “Solid Particle Erosion of Steels at 30° Impact Angle.” Master
Thesis, The University of Rhode Island, 1984.
Shayler, P. J. and K. H. Yee. “Erosion of AISI 303 Steel by Fine (about 5 /im and
about 50 p m ) Ash P articles.” Wear, Vol. 98, (1984): 127-140.
Sheldon, G. L. and I. Finnie. “On the D uctile Behavior of Nom inally Brittle Materials
during Erosion C utting.” Journal of Engineering f o r Industry, ASM E, (November,
1966): 387-391.
Sheldon, G. L. and Ashok Kanhere. “An Investigation of Im pingem ent Erosion Using
Single Particles.” Wear, Vol. 21, (1972): 195-209.
Shetty, D. K., I. G. Wright and A. H. Clauer. “Effects oi Com position and Mi
crostructure on the Slurry Erosion of W C-Co Cerm ets.” Wear, Vol. 114, (1987):
1-18.
216
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Shewmon, Paul G.. “M echanics of Erosion of Alum inum A lloys.” Proceedings of the
7th International Conference on Erosion by Liquid and Solid Impact, Cambridge, UK,
(Septem ber 3-6, 1979): 37-1/37-5.
Shewmon, Paul G.. “Particle Size Threshold in the Erosion of M etals.” Wear, Vol.
68, (1981): 253-258.
Shukla, A., D. C. Holloway and W. W ilson. “Influence of Nose Shape of Ceramic Pro
jectiles on Dam age in Granite Rock.” Proceedings of the 5th International Congress
on Experim ental M echanics, Montreal, 1984.
Source Book on Industrial Alloy and Engineering Data. Ohio: American Society for
M etals, 1978.
Stringer, John and Ian G. Wright. “Some Views on the Formation of Ripples on
Eroded Surfaces.” Proceedings of the Seventh International Conference on Erosion by
Liquid and Solid Impact, Cambridge, UK, (Septem ber 7-10, 1987): 47-1/47-7.
217
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Sundararajan, G.. “T he Effect of Temperature on Solid Particle Erosion.” Wear, Vol.
9 8 ,(1 9 8 4 ): 141-149.
Sundararajan, G.. “A Comprehensive Model for the Solid Particle Erosion of Ductile
M aterials.” Wear, Vol. 149,(1991): 111-127.
Tan, D. K. M.. “A M odel for the Surface Finish in Abrasive W aterjet C utting.”
Proceedings of the 8 th Intem atioanl Symposium on Jet Cutting Technology, BHRA,
Durham, England, (Septem ber 9-11, 1986): Paper 31.
Tabor, D.. The Hardness o f Metals. London: Oxford University Press, 1951.
Tilly, G. P.. “Erosion Caused by Airborne P articles.” Wear, Vol. 14, (1969): 63-79.
Tilly, G. P.. “A Two Stage Mechanism of D uctile Erosion.” Wear, Vol. 23, (1973):
87-96.
W ada, Shigetaka. “Effect of the Fracture Toughness of Impact Particles on the Ero
sion Wear of Ceram ics.” Journal of the Japan Society of Powder and P ow der M etal
lurgy, Vol. 38, (1991): 55-56.
W ada, Shigetaka and Naoyoshi W atanabe. “Solid Particle Erosion of Brittle Materials
(Part 5) — The Effect of Impingement Angle.” Journal of the Ceramic Society of
Japan, International Edition, Vol. 95, (1987): 925-932.
W ada, Shigetaka, Naoyoshi W atanabe and Toshihiko Tani. “Solid Particle Erosion of
B rittle Materials (Part 6) — The Erosion Wear of AI 2 O 3 —S i C C om posites.” Journal
of the Ceramic Society of Japan, International Edition, Vol. 96, (1988): 113-119.
218
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
W ada, Shigetaka and Naoyoshi W atanabe. “Solid Particle Erosion of Brittle Materials
(Part 7) — The Erosion Wear o f Commercial Ceramic Tools.” Journal of the Ceramic
Society of Japan, International Edition, Vol. 96, (1988): 323-325.
W ada, Shigetaka and Naoyoshi W atanabe. “Solid Particle Erosion of Brittle Materials
(Part 8) — The Erosion Rate of ZrOn Ceram ics.” Journal of the Ceramic Society of
Japan, International Edition, Vol. 96, (1988): 587-590.
Wada, Shigetaka, Naoyoshi W atanabe and Toshihiko Tani. “Solid Particle Erosion of
B rittle M aterials (Part 9) — The Erosion Wear of AI 2 O 3 —S i C Particle C om posites.”
Journal of the Ceramic Society of Japan, International Edition, Vol. 96, (1988): 737-
740.
W inter, R .E. and I.M . Hutchings, “Solid Particle Erosion Studies Using Single An
gular P articles.” Wear, Vol. 29, (1974): 181-194.
W oods, Richard D .. “Screening of Surface Waves in Soil.” Journal of the Soil Mechan
ics and Foundations Division, Proceedings of the American Society of Civil Engineers,
(July, 1968): 951-979.
Zahavi, Josephand and George F. Schm itt, Jr.. “Solid Particle Erosion of Reinforced
Com posite M aterials.” Wear, Vol. 71, (1981): 179-190.
Zahavi, Josephand and George F. Schm itt, Jr.. “Solid Particle Erosion of Polymeric
Coatings.” Wear, Vol. 71, (1981): 191-210.
Zeng, J., R. Heine, and T.J. Kim , “Characterization of Energy Dissipating Phe-
nomenan in Abrasive Waterjet C utting.” Proceedings of the 6 th American W ater Jet
219
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Conference, H ouston, Texas, (A ugust 24-27, 1991): 163-177.
Zeng, J. and T .J. Kim. “A Study of B rittle M echanism Applied to Abrasive Wa
terjet Processes.” Proceedings o f the 10th In tem atioan l Sym posium on Jet Cutting
Technology, B H R A , Am sterdam , the Netherlands, (O ctober 31 - Novem ber 2, 1990):
115-133.
Zeng, J. and T. J. Kim. “An E lasto-Plastic M odel of B rittle Material Rem oval in
Abrasive W aterjet C utting.” subm itted for publication in th e Journal of Engineering
M aterials and Technology, 1992.
Zeng, J. and T . J. Kim. “Developm ent of an Abrasive W aterjet Kerf C utting Model
for B rittle M aterials.” accepted for presentation at the 11th International Conference
on Jet C utting Technology and publication in its proceedings, St Andrews, Scotland,
(Septem ber 8-10, 1992).
220
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.