WIZAR

Wear186-187 (1995)388-394

Liquid-jet impact on liquid and solid surfaces

T. Obara, N.K. Bourne, J.E. Field
Shock Physics, PCS, Cavendish Laboratory, Cambridge CB3 OHE, UK

Abstract

Experiments were conducted to investigate the mechanisms of the liquid-jet impact on liquid and solid surfaces associated with cavitation
damage and rain erosion. In this work, a liquid jet of 3 mm diameter, generated using a single impact jet apparatus, was impacted at ca.
600 m SC on the surface of water and polymethyl-methacrylate (PMMA) placed 1.5 mm below the nozzle exit. The various phenomena
which occurred were photographed using an image-converter high-speed camera. These include (i) the shock wave generated at the moment
of the liquid-jet impact on the surface, (ii) shock waves reflected at the liquid/air or solid/air interfaces, as release waves, (iii) the damage
to the PMMA, (iv) the overlap of release waves on the central axis, causing central damage in PMMA and cavitation in liquids. The cavitation
behaviour and the process of the damage is directly related to the behaviour of the compressive wave caused by the impact of the liquid jet.

Keywords: Liquid-jet impact; Cavitation; Erosion; Rain

1. Introduction it takes the release wave, which is generated at the contact

edge of the jet, to reach the central point, i.e.

As is well known, the high-velocity collision of a liquid

mass on a solid material generates extremely short high-
pressure transients which cause serious damage to the surface
and interior of the material [ l-71. A high-speed liquid jet is
also produced within bubbles when they collapse asymmet-
rically. Many researchers have already observed the creation
of high-speed liquid microjets in the final stage of the bubble
collapse [ S-161. The problem of high-speed liquid-jet impact
is of significance in a wide range of technology concerned
with rain erosion, cavitation, turbine-blade erosion and jet
cutting techniques, etc. However, in spite of the considerable
amount of research in the past, there remain unexplained and
poorly understood phenomena associated with liquid-jet
impact and the resulting damage processes. The reason for
this is that the diameter of the liquid jet is generally too small
to observe directly and the damage takes place in an extremely
short time period, i.e. the order of 1 ps.
It is possible to generate extremely high pressure in liquid
impact as a result of the water-hammer effect [ 171. The
water-hammer pressure, for impact on a rigid target, is gen-
erally expressed by
Pwh = Pew
where p and c are the density and shock wave velocity for
the undisturbed liquid respectively, and u the impact jet veloc-
ity. This water-hammer pressure lasts for the short period that
A,=$
where d is the diameter of the impacting jet. After the release
of the water-hammer pressure, the stagnation pressure of the
liquid jet lasts for a relatively long time and is of magnitude
psi= lPV2 (3)
The damage pattern on the PMMA was investigated in
Ref. [ 31 and the three types of damage created are described;
ring fracture, central damage and scabbing near the rear sur-
face of the target. This work accounted for the failure mech-
anisms in the solid as due to the extremely short duration of
explosive compression and the shear stress generated by
the high-speed radial flow caused after the jet impact. Theo-
retical treatment of the collision of a liquid sphere on a solid
surface was developed [ 18-201 in order to investigate the
pressure distributions around the impact area, and the highest
pressure was found to be generated at the contact edge of the
liquid jet. The contact edge pressure progressively increased
to reach a value about three times higher than the water-
hammer pressure on the central axis (Eq. ( 1) ) at the instant
(1)
the shock overtook the contact periphery. These results were
used to explain the ring crack found on the materials and the
short circumferential cracks around the contact area. How-
ever, at present, there is no analytic treatment which can
explain all the damage found in target materials, in particular

0043.1648/95/$09.50 0 1995 Elsevier Science S.A. All rights reserved

SSDIOO43-1648(95)07187-3
 T. Obara et al. /Wear 186187 (1995) 388-394 389

the mechanism resulting in the central crack near the liquid-

jet axis. In his discussion of subsurface damage in PMMA,
Brunton pointed out that in the static loading case as analyzed
by Hertz (see for example, Johnson [ 211)) the maximum
shear stresses occur at a point 0.5 of the radius of contact
El0
Solmold

below the surface. The maximum shear stresses intersect at

angles of 45. However, as shown below, the dynamic case
is more complex.
The mechanisms of damage caused by the cavity collapse
have been investigated extensively [ 10,15,22]. It is generally
concluded that liquid-jet impact has a major role in the dam-
I
Finng

I
reech

I Nitrogen
Bottle

age processes occurring in many engineering materials [ 221 Computer

and it was found that the diameter of the damage pit produced
by shock compression of a bubble on a surface is much rT!!

smaller than that of the minimum bubble base. A detailed

investigation of bubble collapse and damage-pit formation
was conducted [ lo] and it was concluded that the impulsive
pressure, causing plastic deformation of the material, was
closely related to the behaviour of the liquid jet inside the
bubble, which contributed to damage-pit formation. How-
ever, a difficulty arises since the diameter of the microjet
inside a bubble is generally too small to observe, i.e. about ~ Liauid-iet AoDroximate
one tenth of the initial bubble diameter [ 231.
The present paper describes an experimental investigation
into the mechanisms of destructive action caused by liquid-
jet impact. This paper also aims to identify the cause of each
damage type and to clarify the damage mechanisms using
high speed photography and optical techniques. A liquid jet
of about 3 mm in diameter travelling at a velocity of about (b) Streak Axis
600 m s- , in air, was generated by a single impact jet appa-
ratus (SIJA) [ 1 ] and was impacted onto the surface of water Fig. 1. (a) Schematic diagram of the experimental set-up. (b) The field of
or PMMA. These phenomena were visualised by means of view of the high-speed camera.

an image converter high-speed camera with the aid of shad-

firing pressure of 2.0 MPa. The experiments were all con-
owgraph optics.
ducted in air at room temperature and atmospheric pressure.

2.2. The target

2. Experimental

In the first set of experiments, the liquid jet was impacted

2. I. Single impact jet apparatus
Fig. 1 (a) shows the schematic diagram of the experimen-
tal set-up. A vertical liquid jet produced using SIJA, was
impacted onto the surface of PMMA. SIJA uses a high pres-
sure reservoir to project a 5.5 mm diameter lead slug down a
barrel to impact onto a specially designed stainless steel noz-
zle ( 1.6 mm in the exit orifice diameter). The nozzle is care-
fully filled with water to ensure that there are no bubbles
within the chamber and that the meniscus at the end of the
exit orifice has a convex shape. When correctly adjusted, the
water is extruded from the nozzle at any chosen velocity
between 90 m s-l and 1 km s- controlled by the reservoir
pressure. The impact velocity of the liquid jet is accurately
recorded by electronically measuring the interval taken for
the jet to interrupt two light beams. The velocity of a liquid
jet used in these experiments was about 600 m SC with the
onto the surface of water, in a transparent walled chamber
(200 X 50 X 150 mm3) below the nozzle exit. The water used
in this research was tap water.
In the second series of experiments, PMMA was used as a
target to observe the details of the compressive wave and the
fracture process. The density, longitudinal speed and dynamic
yieldstressofPMMAare1.17X103kgm-3,2.7kms-1and
220 MPa respectively. Therefore, the water-hammer pres-
sure, calculated to be about 900 MPa, greatly exceeds the
yield stress of PMMA. Preliminary experiments determined
that a distance of 15 mm between the nozzle exit and the
surface of the target caused most destructive damage on the
PMMA surface. The sample of PMMA has the dimensions
of 50 mm width and 5.9 mm or 11 mm thickness. Therefore,
any compressive wave cannot reflect from the side surface to
disturb the visualisation within an observation period of about
7 l.~s.The side surface of the PMMA were carefully polished
390 T. Obara et al. /Wear
in order to view the stress waves. The samples were mounted
on a sample holder perpendicular to the central axis of the jet
nozzle.
2.3. High-speedphotography
The phenomena caused by the impact of a liquid jet were
photographed by means of an Ultranac programmable high-
speed image converter camera (Ultranac FS501, Imco Elec-
tro-Optics Ltd.). The field of view of the camera was set as
shown in Fig. 1 (b) to view the propagation and interaction
of waves within the target. The camera was operated in fram-
ing or streak mode. The framing rates were varied from lo7
to lo6 frames per second. The great advantages of this camera
are that the interframe and exposure time for each frame can
be independently adjusted to the particular high-speed phe-
nomena. The interframe time and the exposure time of each
frame were controlled by custom software written in our
laboratory. The sample was mounted in a conventional shad-
owgraph system. The flash source was a Xenon tube (QCAS)
which delivered 150 J in approximately 100 ps. The camera
and the flash source were operated by a delay generator trig-
gered by the liquid jet cutting the light beam on nozzle exit.
2.4. Pressure measurements
In order to obtain the time dependence of the impact stress,
a pressure gauge was constructed. This gauge was made using
poly-vinylidene difluoride (PVDF) piezoelectric film of
25 pm thickness (Piezotech Co. Ltd.). The PVDF film was
punched to produce a disk of 1.5 mm diameter and connected
to brass electrodes of 25 pm thickness using conductive tape
(Scotch 9703 Conductive Adhesive Transfer Tape, 3M). The
total thickness of the gauge was estimated to be about
150 Fm. The PVDF gauges were calibrated with manganin
piezoresistive stress gauges (LM-SS-125CH-048, Measure-
ments Group Inc.) by impacting the flat-ended projectile
utilising a single stage gas gun. The rise time of the manganin
gauges was 20-30 ns [ 241. The projectile was accelerated
up to 600 m s-l, so the PVDF gauge was calibrated up to
approximately 1 GPa. The PVDF gauge was used in current
mode, i.e. a current viewing resistor is placed across the gauge
and the voltage across it is recorded directly. In this case the
stress derivative is measured and a time integration must be
carried out to recover the stress signal. The signals from both
gauges were recorded by a digital storage oscilloscope (Tek-
tronix TDS 460). The rise time of the PVDF gauges was
found to be comparable with the manganin gauges and
showed excellent repeatability. The PVDF pressure gauge
was mounted flush to the surface of the PMMA target with
epoxy resin and the distance between the surface and the
PVDF gauge was less than 50 pm.
186-187 (1995) 388-394
3. Results and discussion
3.1. Liquid-jet impact on the su$ace of water
Fig. 2 is a high-speed sequence showing the liquid-jet
impact on a water surface. The water surface, W, is visible
as a horizontal line near the top of each frame. The speed of
the liquid jet was 430 m s- and the liquid jet impacted from
top to bottom. Since this picture was taken without using
shadowgraph optics, the propagation of the compressive
wave is not visible. In frame 1, the liquid jet, J, has already
impacted on the water surface and penetrated into the water.
The dark area, Ct,, in front of the penetrating liquid jet is a
cavitation bubble cloud. These cavitation bubbles are gener-
ated by the tensile stresses caused by the overlap of release
waves from the free surface. The detailed mechanism of bub-
ble generation will be discussed later. In frame 2, the cavi-
tation area expands downward and the bubble cloud also
grows. From frame 3, the shape of the penetratingjet suddenly
changes and a precursor jet, Jr, is produced in the head of the
jet, directed towards the bubble cloud. The cavitation cloud
is rapidly collapsed from the top just after the generation of
the precursor jet.
Fig. 3 shows the high-speed photograph taken with the
aid of shadowgraph optics in order to view the stress waves.
Frame 1 shows the moment 0.9 ps after the impact of the
liquid jet on the water surface. The exposure time for each
frame is 20 ns. The hemispherical wave, S, travelling down-
ward is the shock wave generated by the liquid-jet impact.
The dark concave wave, R, behind the shock wave is a release
wave. The shock wave generated at the moment of the liquid-
jet impact was diffracted at the contact edge and the release
waves reflected at the contact edge of the liquid jet, since the
acoustic impedance between the water and air is greatly dif-
ferent. The release waves, travelling to the central axis of the
liquid jet, overlap, as shown in frame 1, and generate strong
Fig. 2. High-speed photograph showing liquid-jet penetration of a water
surface and the induced cavitation bubble cloud. The iiquid jet impacts from
top to bottom. Speed of liquid jet, 430 m s-l; exposure time per frame, 750
ns. W, water surface; J, liquid jet; C,,, cavitation bubble; J,, precursor jet.
 T. Obara et al. /Wear
Fig. 3. High-speed photograph taken using shadowgraph optics showing the
generation of the cavitation bubble cloud and its collapse. Exposure time,
20 ns; speed of liquid jet, 510 m SK. S, shock wave; R, release wave; C,,
cavitation bubble; Se, secondary shock wave; J,, precursorjet.
tensile stress in front of the penetrating liquid-jet along the
jet axis. In frame 2, it is obvious that the cavitation bubble,
Ct,, was generated just after the overlap of the release waves.
In frame 3, the cavitation area expands downward following
the overlapped area of the release waves. Between frame 4
and 5, the above-mentioned precursor jet, J,, penetrates the
main jet. It is clearly visible that the secondary shock wave,
S,, was generated by the impact of the precursor jet (see also
below). The bubble cloud in the top of the cavitation area is
rapidly collapsed after interacting with the secondary shock
wave. Furthermore, the multiple shock waves, S, propagate
from the area, where the bubble cloud collapsed, as shown in
frame 6. Thus, the bubble cloud was collapsed in a chain
reaction by interaction with the shock waves generated by
the top bubble. No rebound is seen in the bubble cloud
because of the multiple impingement of the shock waves. The
life time of the bubble cloud is very short, ca. 4 ps.
The mechanism causing material damage by single bubble
collapse was investigated [ lo] and it was concluded that the
secondary cavitation, which is caused by the impact of the
microjet within the bubble to the contracting bubble wall was
responsible. The cavitation cloud, generated by the impact of
the liquid jet, might contribute to the damage of the material,
since rapid collapse of these bubbles could produce localised
high pressures near the surface of solid materials.
Fig. 4 shows a streak picture taken with a streak rate of
200 ns mm- . The total observation period is equivalent to
14 us. The horizontal axis indicates time and the vertical
distance along the streak axis, which accurately coincides
with that of the liquid jet. The speed of the liquid jet was
186-187 (1995) 388-394 391
Fig. 4. Streak photograph showing the generation of shock, release waves
and induced cavitation cloud. Speed of liquid jet, 520 m s-. The cavitation
bubble is collapsed by interacting with the shock wave, which was produced
by the precursor jet. J,, liquid jet in ah; W, water surface; S,, bow shock
wave; S, shock wave; R, release wave; G, cavitation bubble; Jr, precursor
jet.
520m s- and the liquid jet travelling in air, J,, is clearly
seen in the left-hand side of this photograph. The liquid jet is
accompanied by a bow shock wave, St,, which transmits into
the water before the jet impacts. The speed of the transmitted
bow shock wave is equal to the speed of sound in water
( 1.47 km s- ) . The impact of the liquid jet on water creates
a relatively strong shock wave, S, travelling with the speed
of 1.95 km s-. The release wave, R, is visible 0.6 ps after
the jet impact and propagates behind the shock wave. The
cavitation bubble cloud, Ci,, was generated by the overlap of
the release waves. After 4 p,s from the jet impact, the pene-
trating jet, Jr, is suddenly accelerated and generates a shock
wave directed towards the bubble cloud. The cavitation cloud
was collapsed from the top at this moment.
The mechanism causing the precursor jet might be
explained as follows; the release wave, which causes the
cavitation cloud in front of the penetrating jet, also propagates
inside the liquid jet and also produces the cavitation cloud
within the liquid jet. Evidence for this has been found [ 251.
The bubble generated inside the jet collapses to drive the
microjet within the bubble. The precursor jet as shown in
these pictures is the penetration of this microjet and its impact
on the bubble wall produces the secondary shock wave, col-
lapsing the bubble cloud. However further work is necessary
to investigate the mechanism which causes the precursor jet
and the secondary shock wave.
3.2. Liquid-jet impact on the suvace of PMMA
Fig. 5 shows a high-speed sequence of the liquid-jet
impact on a PMMA target. The thickness of the target was
5.9 mm and the liquid jet impacted from the top downwards.
The impact and rear surface are visible as parallel lines in
each frame. The speed of the liquid jet was 590 m s- , the
exposure time per frame was 20 ns. Frame 1 is 0.6 p.s after
the liquid-jet impact and shows that the shape of the liquid
jet produced by SIJA, J,, is not perfectly hemispherical but
the head of the jet is flattened by aerodynamic ablation. The
impact shock wave, S, is clearly visible from frame 1. The
release wave, R, is also generated at the contact edge of the
liquid jet. At this point the surface of the target is already
damaged in a ring, D,. In frame 2, the radial jetting, J,, is
created near the contact edge of the liquid jet and travels on
392 T. Obara et al. /Wear 186-187 (1995) 388-394

stress exceeds the failure strength. In frame 6, another dam-

Fig. 5. Liquid-jet impact
Exposure time per frame,
in air; S, shock wave; R,
D,, ring crack; D,, central
middle damage.
on a PMMA surface and the damage processes.
20 ns; speed of liquid jet, 590 m s-r, J,, liquid jet
release wave; Jr, radial jet; S,, bow shock wave;
damage; C, compression wave; D,, spahing; D,,
age, D,, occurs in the middle portion of the target. This
damage was caused by the interaction of the release wave
with another wave, originated from the tip of the central
damage. The details of the mechanism will be discussed later.
Fig. 6 is a high-speed photograph showing more detail of
the liquid-jet impact and fracture process. The velocity of the
liquid jet is 520 m s- and the thickness of the target is
11 mm. The exposure time for each frame is 50 ns. The mag-
nification for this photograph is high enough to view the
details of the wave interactions and the processes of fracture.
In frame 1, the impact shock wave, S, is clearly seen. The
toroidal-shaped release wave, R, is generated from frame 3
near the contact edge of the liquid jet because of the large
difference of the acoustic impedance at the interface of
PMMA and air. It is obvious that the release waves are prop-
agating in the same direction as the initial shock wave and
overlap three dimensionally near the central axis of the liquid
jet, from frame 5 onwards. The central damage, D,, is initiated
from frame 6 following the overlapped area of the release
waves and grows with time. The shape of the central damage
is ellipsoidal, in which the major axis is coincident with the
direction of the overlapped release wave propagation. The
central damage starts to close from frame 8 onward by inter-
action of the compression waves associated with the stagna-
tion pressure.
Fig. 7 is a streak picture, the streak axis accurately coin-
ciding with the central axis of the liquid jet. The horizontal
axis indicates the time and vertical axis the position along the
central axis of the jet. This photograph was taken with the
streak rate of 100 ns mm-, the total observation period is
about 7 ps. The thickness of the target is 5.9 mm. The impact
and rear surface of the PMMA target are visible as the hori-
zontal lines. The three horizontal lines, L,, L,, Ls, between
the two surfaces are marker lines at a spacing of 1.5 mm from
each starting other at the impact surface. Their purpose is to
identify each wave as a compression or a tensile wave by the
different direction of line shift as the density gradients pro-
duced by compression or tensile waves have opposite signs.
The speed of the liquid jet is 580 m s- and is visible at the

Fig. 6. High-speed photograph showing the generation of the ring crack and
central damage in PMMA. Exposure time, 50 ns; speed of liquid jet, 520 m
s-. S, shock wave; R, release wave; D,, ring crack; D,, central damage.

the surface. Since the speed of the radial jet is very fast (ca.
1.3 km s-l), the radial jet is travelling in air accompanied
by the bow shock wave, St,. The central damage, D,, is also
generated by the overlapping of the release waves. In frame
4, the impact shock wave reflects as the release wave, R, at
the rear surface of the target. Spalling, D,, is not found at the
rear surface but inside the target when the resultant tensile
Fig. 7. Streak photograph showing the wave interactions and the process of
damage in PMMA. Speed of liquid jet, 580 m s-. The three lines, L,, L,,
and L,, between the two surfaces indicate marker lines at 1.5 mm intervals.
J,, liquid jet in air; S,, bow shock wave; S, shock wave; R, release wave; D,,
ring crack; D,. central damage; C, compression wave; D,, spalling; T, shear
wave: D,, middle damage.
 T. Obara et al. /Wear
left hand side of this picture, J,, travelling downwards. The
bow shock wave, Sb, travelling with the jet, is also clearly
visible and transmitted into the target material. The speed of
this transmitted shock wave is equal to the longitudinal sound
speed of the PMMA. The shift direction of the three marker
lines behind the shock wave is in the same direction as the
shock-wave motion. The magnitude of the line shift is quite
small indicating the weakness of the bow shock wave. How-
ever, the shock wave, S, generated by the liquid-jet impact is
clearly visible and the line shift is quite large. The release
wave, R, can be seen travelling faster than the impact shock
wave. The line shift direction is the reverse of that for the
shock wave. The ring crack, D,, on the surface was generated
behind the release waves. Eventually, the central damage, D,,
initiated following the overlapped release fan. Since the spall-
ing, D,, developed when the two release fans interacted with
each other, the rear surface of the target is not damaged at all.
The compression wave, C, associated with the stagnation
pressure, can also be seen to start to close the central damage.
It is obvious that when the central damage closes, another
wave, T, is produced and propagates downward, interacting
with the release waves produced by the reflection of the
impact shock wave. The middle damage, D,, is caused just
after the interaction of these two waves. The speed of the
impact shock wave and release wave are estimated to be about
2.7 km s- and 3.2 km s- respectively. However, the speed
of the wave, originated from the closing of the central dam-
age, is quite slow, about 1.4 km s-i, i.e. equivalent to the
speed of the shear wave in PMMA. Furthermore, the line shift
associated with this wave is almost negligible, so this wave
is neither a compression wave nor a release wave. This wave
could be a shear wave, which is generated by the closure of
the central damage. However, further work is necessary to
investigate the production of this shear wave by the closure
of the central damage and also the mechanism which causes
the damage found in the middle of the target.
In earlier studies, the pressure measurement of the liquid-
jet impact was conducted utilising piezo-electric pressure
transducers [ 11. However, the rise time of this type of pres-
sure transducer was relatively slow and it was not rapid
enough to measure the impact stress. Fig. 8 shows the typical
stress signal measured by a PVDF gauge when a liquid jet of
570 m s-i impacted on the target surface. The stress signal
rose very quickly and reached a peak stress of about 850 MPa,
i.e. the peak stress is equivalent to the water-hammer pressure
calculated by Eq. ( 1) . This stress lasted only about 1 ys and
rapidly fell off. The period of the high pressure is also equiv-
alent to the duration calculated by Eq. (2) assuming that the
head diameter of the liquid jet is 3 mm. After the releases of
the water-hammer pressure, the above mentioned stagnation
pressure was found to be applied to the target for a further
4 p.s. The stress of the order of 200 MPa is approximately the
same stress level as calculated by Eq. (3) as the liquid jet
produced by SIJA was impinged continuously on the surface
for a period of ca. 10 ps. Since the pressure gauge was embed-
ded with epoxy resin, the small difference in the acoustic
186-187 (1995) 388-394 393
1 .o ---~~----v
0.8
0.6
0.4
0.2
0.0 I,# ,4
0.0 1.0 2.0 3.0 4.0
Time (p.)
Fig. 8. Typical stress signal measured by a PVDF pressure gauge. The gauge
was mounted on the surface of the target with epoxy resin. Speed of liquid
jet, 570 m SC, The signal of the pressure gauge shows a fast rise time and
the peak stresses are equivalent to a water-hammer pressure of ca. 850 MPa.
impedance between the epoxy resin and PMMA may have
interfered with the reflection of the shock wave. As a result,
neither the ring crack nor central damage were observed on
the surface of the target or underneath the pressure gauge.
4. Conclusions
Liquid jets were impacted on the surface of water and
PMMA. The phenomena were visualised utilising an image
converter high-speed camera. The results obtained are sum-
marised below.
A release wave was produced by the reflection of the shock
wave at the interface. In impact upon a water surface, a cav-
itation cloud was generated at the central axis of the liquid
jet just after the overlap of the release waves. The cavitation
bubble collapsed very rapidly in a chain reaction by interact-
ing with the secondary shock wave, which was generated by
the impact of a precursor jet. The generation of the cavitation
bubble and its rapid collapse might contribute to damage,
since the bubble collapse might produce localised pressure
concentrations on a materials surface due to microjet impact.
The direct impact of the liquid jet on the PMMA causes
destructive damage near the impact surface of and deep inside
the PMMA. The process of the failure is directly connected
with the behaviour of the compressive waves travelling
within the material. Particularly, the release waves produce
the tensile stress and cause the ring crack on the surface. The
central damage was produced behind the three dimensional
overlap of the release waves.
The impact stress was measured using PVDF pressure
gauges. The stress signal rose very quickly and reached a
peak stress, which was equivalent to the water-hammer pres-
sure.
394 T. Obara et al. /Wear 186-187 (1995) 388-394
Acknowledgements
We thank Drs Y. Me-Bar and W.G. Proud for encourage-
ment and advice, and D. Johnson and R.P. Flaxman for their
technical help. T.O. thanks the Japan Society for the Pro-
motion of Science for a Postdoctoral Fellowship for Research
Abroad. N.K.B. acknowledges funding from EPSRC. The
research on liquid impact has been carried out with the sup-
port of the DRA( M) and MOD.
References
[ 11 F.P. Bowden and J.H. Brunton, Proc. R. Sot. Land., A 263 ( 1961)
433.
[2] F.P. Bowden and J.E. Field, Proc. R. SK. Load., A 282 (1964) 331.
[3] J.H. Brunton, Philos. Trans. R. Sot. Load., A 260 ( 1966) 79.
[41 J.E. Field, Philos. Trans. R. Sot. Land., 260 (1966) 86.
[5] J.E. Field and M.B. Lesser, Proc. R. Sot. Loud., A 357 (1977) 143.
[61 J.E. Field, M.B. Lesser, and J.P. Dear, Proc. R. Sot. Lond., A 401
(1985) 225.
[7] J.E. Field, Phys. Med. Biol., 36 ( 1991) 1475.
[81 T.B. Benjamin and A.T. Ellis, Philos. Trans. R. Sot. Land.. A 260
(1966) 221.
[91 W. Lauterbom and H. Belle, J. Fluid Mech., 72 (1975) 391.
[lo] Y. Tomita and A. Shima, J. Fluid Mech., 169 (1986) 535.
[ 111 J.P. Dear and J.E. Field, J. Fluid Mech., 190 (1988) 409.
[ 121 Y. Tomita and A. Shima, Acustica, 71 (1990) 161.
[ 131 N.K. Boume and J.E. Field, Proc. R. Sot. Load., A 435 (1991) 423.
[ 141 N.K. Bourne and J.E. Field, J. Fluid Mech., 244 ( 1992) 225.
[ 151 N.K. Boume and J.E. Field, Cavitation damage by bubbles collapsed
by shock-waves, Proc. Inst. Mech. Engrs, Cavitation, December 9-
11, 1992, Cambridge, UK, 1992, pp. 49-54.
[161 N.K. Boume and J.E. Field, J. Fluid Mech., 259 ( 1994) 149.
[171 S.S. Cook, Proc. R. Sot. Land., A 119 (1928) 481.
[I81 F.J.Heymann,J.Appl. Phys.,40(1969) 5113.
[ 191M.B. Lesser, Proc. R. Sot. Land., A 377 (1981) 289.
[20] P.A. Lush, .I Fluid Mech., 135 (1983) 373.
[211 K.L. Johnson, Contact Mechanics Cambridge University Press,
Cambridge, UK, 1985.
[221 CF. Naude and A.T. Ellis, Trans. ASME Dr J. Basic Eng., 83 ( 1961)
648.
1231 M. Plesset and R. Chapman, J. Fluid Mech., 47 (1971) 283.
f241 Z. Rosenberg, D. Yaziv, and Y. Partom, J. Appl. Phys., 51 (1980)
3702.
[251 J.-J. Camus, PhD Thesis, University of Cambridge, UK, 1971.