Jtam.dvi


JOURNAL OF THEORETICAL

AND APPLIED MECHANICS

48, 4, pp. 957-972, Warsaw 2010

DYNAMIC FRAGMENTATION OF LASER SHOCK-MELTED

METALS: SOME EXPERIMENTAL ADVANCES

Thibaut De Resseguier, Didier Loison, Emilien Lescoute,

Loic Signor, Andre Dragon

Institut Pprime, CNRS–ENSMA–Université de Poitiers, Département Physique et Mécanique

des Matériaux – ENSMA, Futuroscope – Chasseneuil Cedex, France

e-mail: resseguier@ensma.fr; didier.loison@ensma.fr; andre.dragon@ensma.fr

While spall damage in solid materials has been one of the most widely
studied shock-driven phenomena, very little data can be found yet abo-
ut spallation in liquidmetals. In recent papers, some of present authors
have reported on exploratory investigations of liquid spall – sometimes
called ”microspall” – in tin samples melted upon laser shocks of high
intensities. Modelling approaches have also been discussed and compa-
red with experimental observations. In particular, theoretical fragment
size distributions have been compared with experimental distributions
inferred from image analyses on fragment-impacted recovery polymer
shields. While the curves representing the number per unit area vs. si-
ze of fragments present a similar aspect, the numbers predicted by an
energy-based model are significantly greater than those observed expe-
rimentally. This is due predominantly to the characterisation technique,
limited to surface examination, while a number of fragments penetrate
the polymer shield and lay below its surface.This is why a new fragment
recovery technique is being tested in this paper. This technique, based
on a highly transparent low density gel, allows soft recovery and easier
observation (compared e.g.with foams or aerogels) of the fragments size,
shape and penetration depth, with improved spatial resolution. Coupled
with shadowgraphy, scanning electron micrography and optical micro-
graphy, this recovery technique progress allows for an advanced insight
into the phenomenology of microspall related fragmentation. To show
it, laser shocks upon a tin target and an aluminium target are investi-
gated, leading to different scenarios regarding melting conditions and
fragmentation processes.

Key words: laser shock, fragmentation, liquid spall, shadowgraphy, frag-
ment recovery, tin, aluminium



958 T. De Resseguier et al.

1. Introduction

The development of high energy laser facilities dedicated, for example, to in-
ertial confinement fusion, has prompted the question of debris ejection from
metallic shells subjected to intense laser irradiation (and consecutive mecha-
nical shock effects) as an evenmore important issue for both fundamental and
applied research. Whereas spall fracture in solid materials has been exten-
sively studied for many years, see for example the synthesis by Antoun et
al. (2002), scarce data can be found yet about fragmentation occurring after
partial or full melting upon shock compression or release. In this context, we
have set off a research programme based on laser driven shocks on tin tar-
gets, tin being chosen as a reference metal due to its low melting curve in
the pressure-temperature diagram. The dynamic fragmentation ranging from
the microspall process (ejection of a cloud of fine droplets) occurring upon
reflection of the compressive pulse from the target free surface to the late
rupture observed in the unspalledmelted layer (leading to formation of larger
fragments) has been in the centre of interest. These experiments provided so-
me insight into the fragmentation process by combining time-resolved velocity
measurements with post-shock observations of the recovered targets and frag-
ments collected on polycarbonate (PC) shields (de Resseguier et al., 2007a)
or in low density foams (Signor et al., 2007). Transverse visualization of the
fragments ejection has been further developed (de Resseguier et al., 2008). At
the same time, somemodelling issues have been addressed (Signor et al., 2008,
2010). In the second paper cited, an energy-based fragmentation model inspi-
red byGrady (1988, 1996) was introduced in the form of a cumulative failure
criterion in a one-dimensional hydrocode allowing for comparisons between
theoretical predictions and experimental observations. Further analyses, both
experimental and interpretative, in relation to the aforementioned model, are
given in de Resseguier et al. (2010). Figure 1, coming from this paper, shows
fragment size distributions after a 140GPa laser shock applied onto a 50µm-
thick tin target. The dotted line traces results from a numerical simulation
employing the energetic criterion of Grady corresponding to the above laser
shock. The experimental distributions inferred from the image analysis on the
fragment-impacted surface of the PC shield match fairly with the aspect of
the theoretical curve thus indicating a sort of qualitative consistency between
the experiment and theory. However, fragment numbers per unit area pre-
dicted by numerical simulations are significantly greater than those observed
experimentally. The major cause of this discrepancy is the penetration of a
large percentage of tin fragments into the PC shield, i.e., below its surface.



Dynamic fragmentation of laser shock-melted metals... 959

It is confirmed by side views through the transparent shields reported in de
Resseguier et al. (2010) and earlier in de Resseguier et al. (2007a). Smaller
fragments should be ejected at higher velocities and undergo less drag during
penetration than larger ones. They should imbed more easily into PC, while
larger fragments are foundon the shield surface.This is consistentwith theda-
ta of Fig.1, where the discrepancy between theoretical and observed fragment
numbers decreases with increasing fragment size.

Fig. 1. Fragment size distributions after a 140GPa laser shock applied onto a
50µm-thick tin sample. The dotted line results from a one-dimensional simulation of
the experiment using a global energetic criterion to account for the micro-spall
process. The circles and triangles show experimental values provided by image
analysis of Scanning ElectronMicrographs (SEM) at highmagnification. The very
goodmatch between two different pictures (both in the central zone of the impacted
surface) seems to indicate that the results do not dependmuch on the choice of the
observed area. The squares are obtained from a lowermagnification SEM view,
where larger fragments can be observed but sub-µmparticles cannot be properly

detected (coming from de Resseguier et al., 2010)

This is the reason for searchingmore effective recovery techniques making
possible a better quantification of the fragments size and number including
particles which penetrate the shield. In this way, soft recovery of the ejecta is
performed in this work within a low density gel where the fragments can be
observed by post-shot optical microscopywith a resolution of µm-order. Such
a post-shock analysis is coupled with other diagnostics to investigate frag-
mentation processes like microspalling andmicrojetting (this term designates
three-dimensional effects governedmainly by target surface roughness leading
to the ejection of thin ’jets’ at very high speed). Currently, themost common
diagnostic to studydynamic fragmentation is theVISAR-interferometer (Bar-
ker and Hollenbach, 1972) which allows time-resolved measurements of the
target free-surface velocity. However, it is based on the probe light reflected



960 T. De Resseguier et al.

from this surfacewhile the reflectivity at stake is lostwhen the aforementioned
processes producea cloud ofmicrometric debris.The transverse shadowgraphy
is an alternative technique giving quasi-instantaneous pictures of ejected frag-
ments at controlled delay times after the laser shot. Mean ejection velocities
can be estimated from such pictures but low spatial resolution and integra-
tion across the transverse direction limit the characterisation of the fragments
population.To discriminatewithin a field of velocities, as e.g. for a cloud com-
posed of particles with different velocities, the Heterodyne Velocimetry (HV)
technique has been recently developed and seems promising. It has been suc-
cessfully applied to laser shock induced fragmentation (Mercier et al., 2009).
These comments put stress on the complementarity of themethods employed.
Here, shadowgraphy and optical micrography images are being exploited in
order to work out a more precise phenomenology of (i) reference microspall
produced in a thin tin sample and (ii) spallation in a partially melted alumi-
niumtarget. Inboth cases, the ejected fragments are recovered in a lowdensity
gel. The shock pressure decay in the aluminium target is computed in order
to explain the fragmentation scenario different from the reference case (i).

The microspall phenomenology and the experimental setup employed are
summarized inSection 2.Themainobservations regarding the test (i) are brie-
fly given in Section 3, with a focus on the use of the new recovery technique.
Amore extensive analysis of the test (ii) is proposed in Section 4. The corre-
sponding interpretations of fragmentation processes are presented inSection 5.
Concluding remarks (Section 6) and references end the paper.

2. Microspall phenomenology and experimental setup

When a high power pulsed laser beam is focused onto the surface of an opaque
sample, a thin layer ofmaterial is vaporised into aplasmacloud.Theexpansion
of this plasma cloud towards the laser source drives by reaction a compressive
pulse into the solid sample (target), see Figs.8a and 9a. This is an essential
feature regarding laser-matter interaction summarizing the transformation of
laser irradiation into pressure pulse. The resulting strain rates are about 107

to 108 s−1, higher than those commonly attained under explosive loading or in
plate impact experiments. If the diameter of the irradiated spot is much gre-
ater than the sample thickness and if the laser energy distribution is spatially
uniform, the assumption of a planar wave front with conditions of uniaxial
strain in the target region around the laser axis is a good approximation. Du-
ring propagation through the sample, the compression wave gets steeper and



Dynamic fragmentation of laser shock-melted metals... 961

the releasewave spreads, so the pressureprofile becomes essentially triangular,
and the peak pressure decays. As far as the amplitude of this pressure load
is beyond the melting pressure (about 60GPa for tin), the material is melted
on compression. For lower amplitudes (between 20 and 60GPa for tin), it can
bemelted fully or partially on release from the compressed state (Elias et al.,
1988; Eliezer et al., 1990).

Upon shock breakout at the free surface of the target, three-dimensional
effects governed predominantly by surface roughness can lead to ejection of
thin jets at very high speed, usually termed asmicrojetting (Asay et al., 1976).
Employing targets with smooth surfaces, as in our previous experiments (the
order of magnitude for roughness of ∼µm), this process is usually limited to
extremely small masses, see Fig.8b.

The seminal monograph by Nowacki (1978) gives detailed insights into
wave propagation issues; the propagation of plane shock waves is dealt with
in Chapter III of this book.

The reflection of the shock from the free surface produces a release front,
and the interaction of this reflected wave with the incident unloading wave
gives rise to tensile stresses over a layer of finite thickness beyond the free sur-
face. For the solid material, this interaction of the two release waves can lead
to the well-known spallation fracture by nucleation, growth and coalescence
of voids or microcracks. If the material has been melted on compression or
on release, the interaction occurs in a liquid state, which cannot bear intense
tension. Instead of the aforementioned spallation (solid state), a microspall
(or liquid spall) takes place (Andriot et al., 1984), leading to material frag-
mentation into fine droplets ejected from the whole layer subjected to tensile
loading. In Fig.8c, concerning a thin tin sample (50µm) subjected to high
loading pressure, this layer matches the sample thickness, so the whole target
is fragmented into a microspall. This is not the case in general for thicker
targets andmoderate pressure (see de Resseguier et al., 2010). The unspalled
molten layer may separate into large liquid secondary fragments, turning into
spherical drops under the action of surface tension.

In theexperiments reportedhere, two complementary techniqueshavebeen
used to investigate shock-induced dynamic fragmentation. First, optical trans-
verse shadowgraphy provides successive pictures of the debris ejected from the
sample, using a continuous laser probe to light up the gap behind the target
and the afocal system (Fig.2). The beam is divided with a pellicle beamsplit-
ter and sent to two cameras with controlled delay times and an acquisition
duration of 10ns, which ensures minimum motion blur (typically 10µm for
a particle ejected at 1km/s). Mean ejection velocities can be estimated from



962 T. De Resseguier et al.

Fig. 2. Schematic of the experimental setup. Fragments ejected from the laser
shock-loaded sample are observed by transverse shadowgraphy then collected in a

transparent gel (Lescoute et al., 2009)

such pictures, but low spatial resolution, speckle background due to the cohe-
rent laser probe light, and integration across the transverse direction strongly
limit the characterisation of the fragments population.Therefore, soft recovery
of the ejecta has been performed to complement the results. In previous tests,
such recovery has been achieved in foams (see Signor et al., 2010), but their
analysis requires X-ray radiography and/or tomography. Here, a gel compo-
sed of hydrocarbon and polymer is set about 1 to 2cm behind the sample
free surface (Fig.2). Its low density of 0.9g/cm3 favours soft collection of the
debris, while its high transparency allows easy observations of the recovered
fragments with a µm-order resolution. Both the gel and the sample are pla-
ced in a vacuum chamber (10−4mbar) to avoid laser breakdown in air before
reaching the sample surface.

3. Soft recovery of microspalled ejecta from a tin sample

In this short section, some exploratory results are given as a reference illu-
stration regarding the application of the new recovery technique based on the
low density gel. The target is a 50µm-thick tin sample of 99.99% purity. The
experiment was performed in the Laboratoire pour l’Utilisation des Lasers In-
tenses (LULI, UMR CNRS 7605, Ecole Polytechnique, France). A laser pulse
of 1.06µmwavelength, 4nsdurationand868J energywas focusedonto a2mm
diameter spot in the target surface, leading to a 6.9TW/cm2 intensity. The
corresponding laser-matter interaction computation indicates a ∼ 135GPa



Dynamic fragmentation of laser shock-melted metals... 963

peak loading pressure applied on the irradiated surface. Thus, the test is ve-
ry similar to that reported above (Fig.1, de Resseguier et al., 2010), but the
ejecta are collected here in a gel set 2cm away from the target free surface
instead of a PC shield.

Fig. 3. Optical micrographs of fragments recovered in a gel set behind a 50µm-thick
tin target subjected to a 4ns laser shot at 6.9TW/cm2. The impact surface in the

gel is vertical, with debris penetrating from left to right

Figure 3 shows optical micrographs through the gel recovered after the
experiment. Fragments penetrated up to 10mmover a zone of 2-3mmdiame-
ter in the transversal direction. Most fragments are approximately spherical
which indicates that they result from the fragmentation process in the liquid
state, so confirming full microspalling expected in this shot. The observations
strongly suggest that most ejected particles are present in the gel, and that
their shapeand sizehavenotbeendeeply affectedby their impact andpenetra-
tion. Thus, this new technique for fragment collection appears softer than the
former PC shields, where large uncertainties had arisen from possible effects
such as secondary fragmentation upon impact or rebounds fromthe shield sur-
face. Although few solidified droplets of ∼ 100µm diameter can be seen, the
widemajority of observed fragments have a size of about 10µmand less. This
is fully consistent with previous results concerning tin debris collected in PVC
foam for similar shock conditions, as shown in Signor et al. (2010). Smaller
fragments are not resolved due to the limitations of optical microscopy em-
ployed here. Actually, the stream-like wakes of dark-grey traces are filled with
sub-µm tin droplets. To enter into these wakes and characterise such small
embedded debris, high-resolution (up to 0.3µm)X-raymicro-tomography has



964 T. De Resseguier et al.

been planned. Furthermore, calibration work is needed to be able to correlate
the impact velocity of a particle of known size and mass with its penetra-
tion depth in the gel, so that ballistic properties of the fragments within the
microspall can be estimated.

4. Dynamic fragmentation of aluminium in the vicinity of

melting point

Controlling spallation processes – including liquid spalling in the full meaning
of the term ’spallation’ here – is crucial in many fields where structural ma-
terials are subjected to intense shock loading, such as pyrotechnics, armor
design, or inertial confinement fusion, forwhich the question of debris ejection
from metallic shells submitted to intense laser irradiation has become a key
issue in order to anticipate and reduce damage to optics and other equipment
impacted by ejecta. This is why research programmes on spallation concern
a large class of engineering materials. As regards our specific programme on
dynamic fragmentation of metals below and above melting, it is evident that
starting from tin, considered as a reference material, further research effort
is to be made for more commonly applied structural metals. In this section,
some exploratory results are given for aluminiumusing the setup described in
Section 2.

The target is a 200µm-thick aluminium foil. The experiment was perfor-
med in the Centre d’Etudes Scientifiques et Techniques d’Aquitaine (CESTA,
CEA, Le Barp, France), on the Alisé laser facility. A laser pulse of 1.06µm
wavelength, about 3ns duration and 157J energy was focused onto a 1.7mm
spot, leading to a 2.6TW/cm2 intensity and an 80GPa peak loading pressu-
re on the sample surface, according to laser-matter interaction computations
with the code CHIC of the Centre Lasers Intenses et Applications (CELIA,
UMR CNRS 5107, France). Information regarding the fragmentation process
is provided by means of (i) scanning electron micrographs of the rear (free)
surface of recovered target (Fig.4), (ii) optical micrographs of fragments reco-
vered in the gel described above (Fig.5), (iii) optical transverse shadowgraphs
of the fragments ejected from the sample free surface; one shadowgraph is
shown in Fig.6. The central part of the target has been removed, the edges
of the hole show clear evidence of resolidification after melting, and the ring
between thismelted edge and the outer unaffected zone presents a rough frac-
ture surface typical of ductile fracture (Fig.4). Such features, very similar to
previous observations in tin samples (see de Resseguier et al., 2007b, 2010),



Dynamic fragmentation of laser shock-melted metals... 965

might suggest microspalling. However, the ejecta collected in the gel (Fig.5)
are clearly different fromthose recovered fromthe tin target (see Section 3 and
Fig.3). Indeed, a large (∼mmsize), single, roughlypenny-shaped fragmenthas
penetrated about 7mmdeep into the gel. Large amounts of much smaller de-
bris are foundmostly behind that main fragment.While most of these debris
have irregular shapes, some spherical particles, indicative of fragmentation in
liquid state, are detected (see detailed views in Fig.5). The shadowgraph in
Fig.6 has been recorded 1µs after the laser shot. It shows the expansion of
a debris cloud from the free surface. Despite the speckle background due to
transverse illumination by a laser source, such images provide a good estimate
of the mean ejection velocity of the fastest fragments, about 3500m/s in this
experiment.

Fig. 4. Scanning electronmicrographs of the free surface of a 200µm-thick
aluminium target after a 2.6TW/cm2 laser shot

Fig. 5. Optical micrographs of fragments recovered in a gel set behind a
200µm-thick aluminium target subjected to a 2.6TW/cm2 laser shot. The impact

surface in the gel is vertical, with debris penetrating from left to right



966 T. De Resseguier et al.

Fig. 6. Shadowgraph behind a 200µm-thick aluminium sample 1µs after a
2.6TW/cm2 laser shock applied onto the left surface (not shown). A debris cloud
(delimitated by the white line for clarity) is ejected from left to right, with a peak

ejection velocity of about 3500m/s

To interpret the experimental data reported above, 1D computations of
shock wave propagation using the code CHIC with a tabulated BLF equ-
ation of state (Bushmann et al., 1992) for aluminium have been performed to
estimate the peak pressure values throughout the sample thickness, notably
near the spalled regions, taking into account the decay of the pressure pulse
during its propagation from the loaded surface (Fig.7). It can be seen that
shock pressure decays from ∼ 80GPanear the irradiated surface to ∼ 33GPa
beneath the free surface. From the pressure-temperature-phase diagram and
Hugoniot curve of aluminium (Henis andEliezer, 1993) one can infer that this
metal would fully or partially melt on release from shocked states between
about 60GPa and 100GPa (where the Hugoniot would meet the melting cu-
rve). Hence, a ∼ 65µm-thick layer beneath the irradiated surface of the laser
shock-loaded sample would be partially melted, but the reflection of the com-
pressive pulse fromthe free surfacewould take place in solid aluminium,where
the interaction of the incident and reflected release waves would induce tensile
stresses of increasing intensity, until reaching the so-called ’spall strength’ of
thematerial. This value, considered as the threshold tensile stress for damage
initiation, is known to strongly depend on the strain rate (see Antoun et al.,
2002; Klepaczko and Chevrier, 2003). Under laser shock loading conditions,
it has been reported to be about 1.6GPa for solid aluminium (Tollier and
Fabbro, 1998). Considering the shorter duration of the laser pulse leading to a
higher strain rate in the present experiment, it has been set to 1.8GPa in the
simulation.According to that simulation, several spalled layers would separate
successively from the free surface. The first, fastest one would be ejected at
3300m/s. This prediction is in fair agreement with the maximum velocity of
about 3500m/s inferred from the shadowgraph (Fig.6).



Dynamic fragmentation of laser shock-melted metals... 967

Fig. 7. Pressure profiles computed in a 200µm-thick aluminium sample subjected to
a 2.6TW/cm2 laser shock, at increasing distance from the loaded surface

(from 3µm t0 197µm). The dotted line represents the 60GPa threshold pressure
for partial melting on release

5. Discussion

As pointed out earlier, high power pulsed laser irradiation produces the abla-
tion of a thin layer (micrometer thick) of target material. The expansion of
this layer, transformed into a plasma cloud, induces by reaction a compressi-
ve pulse onto the solid target (duration of ∼ 3-4ns for the two experiments
considered). The loading pressure pulse is imposed as a mechanical bounda-
ry condition on the irradiated target surface. During its propagation through
the target, the compression front gets steeper, turning into a shock front and
the unloading wave spreads. The pressure profile becomes roughly triangular
(sometimes called unsupported shock wave or Taylor like wave) and the peak
pressure decays. The reflection of this triangular compressive pulse from the
opposite free surface of the target generates tension whose intensity should
increase with distance from that surface. The dynamic fragmentation of the
target sample may involve the following processes: (i) microjetting, (ii) spall
fracture, (iii) microspallation after melting, depending on the loading con-
ditions, target thickness and surface roughness. Both shots analysed in this
paper by complementary techniques involve some melting, respectively in tin
and aluminium, but they present highly contrasted scenarios.

The 50µm-thick tin foil subjected to a 135GPa laser shock is fullymelted,
predominantly on compression, then completely microspalled over its whole
thickness. The corresponding gel is free from large fragment, fine droplets are
recoveredwith awide range of penetration depths indicative of awide range of
ejection velocities,mostly in theaxial direction, andedge effects donot seemto



968 T. De Resseguier et al.

strongly affect the debris population. These resultsmatch the phenomenology
schematically depicted in Fig.8, as well as previous experiments reported.
Quantitative analysis, similar to that shown in Signor et al. (2010) for debris
collected in PVC foams, is to be made with three-dimensional X-ray micro-
tomography to extract volumetric distributionof fragments (size andnumber).
High resolution is necessary to capture sub-micrometric fragments.

Fig. 8. Phenomenological interpretation of laser shock-induced ’micro-spall’ in a
50µm-thick tin sample. Melting is assumed to occur throughout the sample

thickness, either on compression or on release

The 80GPa laser shock on the 200µm-thick aluminium target leads to a
very different tableau (Fig.9). Some partial melting occurs near the loaded
surface, but due to shock pressure decay with propagation distance, most of
the sample remains solid, so that spall fracture is the predominant fragmen-
tation mechanism, with a spall strength of about 1.8GPa (tension) in solid
aluminium at the extreme strain rate involved here. It leads to ejection of
solid spalled layers from the free surface (Fig.9c). The fastest one is seen to
fly at about 3.5km/s, followed by a continuous debris cloud (Fig.6), then it
penetrates about 7mm deep into the gel. This picture is consistent with one-
dimensional computations of laser-matter interaction and shock propagation
(Fig.7). Further deformation of the target leads to full opening, explaining the
final hole (Fig.4), as well as ejection of both solid and liquid (coming from the
vicinity of the irradiated surface) secondary fragments (Fig.9d), which pene-
trate into the gel behind the main spall. To reach the microspall regime for
such a thick aluminium foil, the shock pressure (i.e. laser intensity) should be
increased, and/or sample should be preheated to lower the pressure required
for shock-inducedmelting.



Dynamic fragmentation of laser shock-melted metals... 969

Fig. 9. Phenomenological interpretation of laser shock-induced fragmentation in a
200µm-thick aluminium sample. Partial melting is assumed to occur only over a
limited depth beneath the loaded surface (left), then shock pressure decays below
the melting conditions, so that spall fracture takes place in the solid region

6. Conclusion

Dynamic failure and fragmentation of shock-loaded tin and aluminium have
been investigated usinga laser shockgenerator andcomplementarydiagnostics
methods (scanning electron microscopy, optical shadowgraphy, soft recovery
technique involving a low density gel). The energies involved in laser shock
experiments are lower than in conventional shock techniques, e.g. plate impact
experiments, so they are less destructive. This allows one to carry out an
original post mortem examination of targets and fragments. Analysing the
recovered ejecta leads to evaluation of the number and of characteristic size
spectrumof the fragments.As illustrated in thiswork and fromearlier studies,
the fragmentation in microspalled (liquid) tin involves micrometric and sub-
micrometric particles. In regard toaluminiumtested in loading conditionsnear
melting on release, the fragments sizes are found to range from some∼ 10µm
to somemm after a combination of spall fracture (solid state) and secondary
fragmentation in partially melted regions near the loaded surface. It is shown
that even in solid state, the fragmentation process is far more complex than
the idealized succession ofwell definedspalled layers. Furtherwork is needed to
better understand the evolution of this process from spalling to microspalling
in a variety of structural metals.

Acknowledgments

We thank all the staff of the LULI and the Alise team for technical support, as

well as Pierre-HenriMaire, JérômeBreil andGuy Schurtz for providing access to the

code CHIC.



970 T. De Resseguier et al.

References

1. Andriot P., Chapron P., Lambert V., Olive F., 1984, Influence of mel-
ting on shocked free surface behaviour using Doppler laser interferometry and
X-ray densitometry,Proc. Shock Waves in Condensed Matter, Asay J.R., Gra-
hamR.A., StraubG.K. (Eds.), North-HollandAmsterdam, 277-280

2. Antoun T., Seaman L., Curran D.L., Kanel G.I., Razorenov S.V.,
Utkin A.V., 2002, Spall Fracture, Springer, NewYork

3. Asay J.R., Mix L.P., Perry F.C., 1976, Ejection of material from shocked
surfaces,Applied Physics Letters, 29, 5, 284-287

4. Barker L.M., Hollenbach R.E., 1972, Laser interferometer for measu-
ring high velocities of any reflecting surface, Journal of Applied Physics, 43,
4669-4675

5. Bushmann A.V., Lomonosov I.V., Fortov V.E., 1992, Equation of state
for metals at high energy density, Technical report of Institute of Chemical
Physics, Chernogolovka

6. DeResseguierT., SignorL.,DragonA.,BoustieM.,RoyG.,Llorca
F., 2007a, Experimental investigation of liquid spall in laser shock-loaded tin,
Journal of Applied Physics, 101, 1, 013506

7. De Resseguier T., Signor L., Dragon A., Severin P., Boustie M.,
2007b, Spallation in laser shock-loaded tin below and just above melting on
release, Journal of Applied Physics, 102, 7, 073535

8. De Resseguier T., Signor L., Dragon A., Boustie M., Berthe L.,
2008,On the dynamic fragmentation of laser shock-melted tin,Applied Physics
Letters, 92, 13, 131910

9. De Resseguier T., Signor L., Dragon A., Roy G., 2010, Dynamic frag-
mentation of laser shock-melted tin: experiment and modelling, International
Journal of Fracture, 163, 1/2, 109-119

10. EliasP.,ChapronP., LaurentB., 1988,Detection ofmelting in release for
a shock-loaded tin sample using the reflectivity measurement method, Optics
Communications, 66, 2/3, 100-106

11. Eliezer S., Gilath I., Bar-Noy T., 1990, Laser-induced spall in metals:
experiment and simulation, Journal of Applied Physics, 67, 715-724

12. Grady D.E., 1988, The spall strength of condensed matter, Journal of the
Mechanics and Physics of Solids, 36, 3, 353-384

13. GradyD.E., 1996, Spall and fragmentation in high-temperaturemetals, high-
pressure shock compression of solids, In: High-Pressure Shock Compression of
Solids. II Dynamic Fracture and Fragmentation, Davison L., GradyD.E., Sha-
hinpoorM. (Eds.), Springer, NewYork Ch.9, 219-236



Dynamic fragmentation of laser shock-melted metals... 971

14. Henis Z., Eliezer S., 1993,Melting phenomenon in laser-induced shock wa-
ves,Physical Review E, 48, 3, 2094-2097

15. Klepaczko J.R., Chevrier P., 2003, A meso-model of spalling with ther-
mal coupling for hardmetallic materials,Engineering FractureMechanics, 70,
2543-2558

16. Lescoute E., de Resseguier T., Chevalier J.M., Boustie M., Cuq-
Lelandais J.P., Berthe L., 2009, Soft recovery technique to investigate dy-
namic fragmentation of laser shock-loadedmetals,Applied Physics Letters, 95,
211905

17. Mercier P., Benier J.,Frugier P.A., Sollier A., Rabec Le Gloahec
M., Lescoute E., Cuq-Lelandais J.P., Boustie M., de Resseguier

T., Claverie A., Gay E., Berthe L., Nivard M., 2009, PDV Measure-
ments of NS and FS Laser driven shock experiments on solid targets, Proc.
Shock Compression of CondensedMatter, American Institute of Physics,1195,
581-584

18. NowackiW.K., 1978,StressWaves in Non-Elastic Solids, Pergamon,Oxford

19. Signor L., de Resseguier T., Rot G., Dragon A., Llorca F., 2007,
Fragment-size prediction during dynamic fragmentation of shock-melted tin:
recovery experiments and modelling issues, Proc. Shock Compression of Con-
densed Matter, American Institute of Physics, 955, 593-596

20. Signor L., Dragon A., Roy G., de Resseguier T., Llorca F., 2008,
Dynamic fragmentation of melted metals upon intense shock wave loading.
Some modelling issues applied to a tin target, Archives of Mechanics, 60, 4,
323-343

21. Signor L., de Resseguier T., DragonA., RoyG., Fanget A., Faessel
M., 2010, Investigation of fragments size resulting from dynamic fragmenta-
tion in melted state of laser shock-loaded tin, International Journal of Impact
Engineering, 37, 8, 887-900

22. Tollier L., Fabbro R., 1998, Study of the laser-driven spallation process by
VISAR interferometry technique. II. Experiment and simulation of the spalla-
tion process, Journal of Applied Physics, 83, 3, 1231-1237

Dynamiczna defragmentacja metalu topionego impulsem laserowym:

wybrane rezultaty badań doświadczalnych

Streszczenie

O ile problem łuszczenia i wykruszania się materiałów pod wpływem obciążeń
udarowych został bardzo szeroko przeanalizowany, bardzo niewiele informacji spoty-
ka się odnośnie odpryskiwaniametali w stanie płynnym.Niektórzy z autorówobecnej



972 T. De Resseguier et al.

pracy przeprowadzili badania nad odpryskiem płynnym, czasem zwanego mikrood-
pryskiem, polegającym na dezintegracji materiału stopionego impulsami laserowymi
odużejmocy.Zaproponowalioni także sposóbmodelowania tego zjawiska,weryfikując
go z wynikami eksperymentalnymi. W szczególności, rozkłady teoretyczne rozmiaru
odprysku porównano z wynikami doświadczeń, gdzie metodami analizy obrazu zmie-
rzono rozmiar odpryśniętych cząstek uwięzionych w specjalnych ekranach polimero-
wychumieszczonych za próbką. Zauważono, że analityczne krzywe rozkładu reprezen-
tujące liczbę odłamków na jednostkę powierzchni w funkcji wielkości tych odłamków
pozostająwzgodzie z eksperymentem, jednak już sama ilość zdefragmentowanegoma-
teriałuwynikającego zmodelu energetycznego znacznie przewyższa rezultat doświad-
czalny. Ta nadwyżka wynika głównie ze stosowanejmetody zliczania ilości odprysku,
ograniczającej się jedynie dobadańpowierzchniowych, tymczasemczęść odpryskupe-
netruje ekranpolimerowyznacznie głębiej. Z tegopowoduwniniejszej pracy zapropo-
nowano i przetestowano nowąmetodę jakościowego i ilościowego określania odprysku
oraz przedyskutowano jej skuteczność. Nowametoda bazuje na zastosowaniu zamiast
ekranów polimerowych przezroczystego żelu o małej gęstości, umożliwiającego łatwą
obserwację (wporównaniudo pianek i aerożeli)wielkości odpryskiwanychcząstek, ich
kształtu oraz głębokości penetracji przy zwiększonej rozdzielczości przestrzennej ob-
razu. Technika ta, wspomagana cieniografią oraz mikrografią optyczną i elektronową
(mikroskop skaningowy), stanowi zaawansowane narzędzie do badań nad zjawiskiem
mikroodprysku.Do prezentacji skuteczności tej metody przeprowadzono doświadcze-
nia nad próbkami z cyny i aluminium przy różnychwarunkach topienia i defragmen-
tacji pod wpływem impulsu laserowego.

Manuscript received April 15, 2010; accepted for print June 7, 2010