International Journal of Impact Engineering 39 (2012) 1e7

Contents lists available at SciVerse ScienceDirect

International Journal of Impact Engineering

journal homepage: www.elsevier.com/locate/ijimpeng

Impact response of high density flexible polyurethane foam

E. Zaretskya, *, Z. Asafb, E. Ranb, F. Aizikb
a
Department of Mechanical Engineering, Ben Gurion University of the Negev, P.O.Box 653, Beer Sheva 84105, Israel
b
Plasan Sasa, Kibbutz Sasa, M.P. Merom Hagalil, 13870, Israel

a r t i c l e i n f o a b s t r a c t

Article history: The impact response of high density flexible polyurethane-based foam was studied in a series of
Received 13 June 2011 symmetric (both the impactor and the sample made of the same foam) planar impact experiments, with
Accepted 9 September 2011 continuous VISAR monitoring of the velocity of the rear sample surface. The impact velocities in these
Available online 16 September 2011
experiments varied from 43 to 605 m/s providing a sample compression over the 0.36e51-MPa pressure
range, with the strain rates changing, respectively, from 4
103 to 6
105 s1. The linear shock velocity-
Keywords:
particle velocity Hugomiot of the foam, US ¼ US0þsu ¼ 14.8 þ 1.318u, was determined on the basis of the
Flexible foam
recorded velocity histories. The rise times of the velocity histories allows one to conclude that under
Shock compression
Crush up stress
shock compression above 3.2 MPa, the initial structure of the foam is completely crushed and the foam
resistance to the propagation of the shock is determined by the void-free foam material. The dynamic
tensile (spall) strength of the foam, determined in a separate impact experiment with 1-mm thick foam
impactor was found equal to 0.3 MPa. Such unexpectedly low spall strength is possibly the result of
substantial damage having taken place in the foam during compression.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction
The ability of polymeric foams to absorb energy of impact stands
behind a wide variety applications in automotive industry, civil
engineering, packaging and transportation of fragile goods. The
quasi-static mechanical properties of the foams, which belong to
a large group of materials with cellular structure, were studied
intensely over last fifty years. The results of these multiple studies
were carefully analyzed by Gibson and Ashby [1], who suggested
a series of useful phenomenological relations between the densities,
moduli, Poisson’s ratios, and collapse stresses of the foams and the
properties of the bulk material the foam was made of. Similar rela-
tions for foams’ moduli, yield or collapse stress, etc., were obtained
as the result of the micromechanical study of the struts of the foam
skeleton [2]. The dynamic response of the foams, essential for their
applications, was not addressed in these studies. An important
feature of the dynamic response of the polymeric foams is their
strong strain rate sensitivity [3]. The influence of the strain rate on
the mechanical response of foams was revealed in drop-weight
[4e6] and impact sleds [7] experiments and was found to be
crucial for their energy absorbing ability. Unfortunately the results of
these experiments do not provide data for a constitutive description
* Corresponding author. Tel.: þ972 8 6477102; fax: þ972 8 6477100.
E-mail address: zheka@bgu.ac.il (E. Zaretsky).
of the studied material. This problem was partly solved by use of the
SHPB (Split Hopkinson Pressure Bar) [8e11] study of the dynamic
response of the foams. This method allows obtaining the stress-
strain relations of the foams at the relevant (102 s1 - 103 s1)
strain rates providing valuable information for the constitutive
modeling of the foams and for energy absorption estimates.
Gas gun driven planar impact experiments are widely used for
studying the dynamic response of metals, ceramics and polymers
up to strain rates of about 106 s1 and higher [12]. The adjustment
technique used in such experiments allows one to create exact
uniaxial strain boundary conditions in the impacted sample. As
a result, the initial parameters of the shock-induced stress pulse
traveling through the studied sample are precisely defined. The
response of the impact-loaded sample being monitored by
a Velocity Interferometer System for Any Reflector (VISAR) [13]
provides accurate constitutive information about the studied
sample. One motivation for the present work was to explore the
possibility of using the VISAR-instrumented gun-driven planar
impact experiment for obtaining such constitutive information
about flexible polymeric foam. Another reason was dictated by the
need to close the gap between the highest loading rates of SHPB
technique, w103 s1, and the lowest of the planar impact,
w104 s1. Addressing these two issues prompted us to perform
a series of planar impact experiments with high density flexible
polyurethane foam, accompanied by VISAR monitoring the
velocity w of the free sample surface.

0734-743X/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijimpeng.2011.09.004
2 E. Zaretsky et al. / International Journal of Impact Engineering 39 (2012) 1e7
2. Material and experimental
Planar impact experiments with the polyurethane foam were
performed with 59-mm bore 4-m long gas gun of the Laboratory of
Dynamic Behavior of Materials at Ben-Gurion University. The 8.9
(0.1)-mm thick sheets of high density (r0 ¼ 409  4 kg=m3 , some
65e66% open porosity) flexible polyurethane foam were received
from PLASAN Ltd., Sasa, Israel. The structure of the studied foam, the
dense packing of interconnected hollow spheres of 100e150 - micron
diameter, is shown in Fig. 1a. Prior the impact experiments, the foam
was tested in quasi-static compression using a 5587 Instron testing
machine equipped with the Instron 2501 150-mm compression
platens and an Instron 2601 deflection sensor. The results of this test
with a 55 mm thick (6 layers of 8.9-mm thick) foam sample are
shown in Fig. 1b.
The presently studied material has a stress-strain diagram that
is typical for foams [1], with an inflection point at approximately
0.13 MPa and collapse stress, determined as shown in the insert of
Fig. 1b, equal to 0.07 MPa. The initial, straight, segment of the
diagram has a slope of about 0.78 MPa which may be interpreted as
the foam’s Young’s modulus.
For 8 of 10 planar impact tests (marked as PFA to PFH in Table 1)
both the samples and the impactors were prepared from 60

60 mm2 square pieces cut from the foam sheets. The projectiles
equipped with the foam impactors were prepared according to the
following procedure: the foam square was glued (two-component
DEVCON 5 min epoxy) to the front part of 55-mm diameter, 11.8-mm
thick plane-parallel polymethylmethacrylate (PMMA) disk, and the
foam surplus over the 55-mm diameter was removed by turning. In
order to provide shorting of the velocity and trigger electrical
charged pins, the 14-m thick aluminum foil ring was glued, using the
same epoxy, to the front surface of the foam impactor. The back
surface of the PMMA disk was glued with Loctite Super glue to the
front edge of the hollow aluminum cylinder sabot. Finally the rear
edge of the sabot was closed with a PMMA lid with an O-ring. In these
eight tests and in the ninth test PFE1 where the sample was made of
5.5-mm (instead of 8.9-mm) foam layer, a similar sample assembly
was used. The rear surface of the square foam sample was glued on
a polyvinyl chloride (PVC) 100-mm diameter and 5-mm thick disc
with a 45-mm diameter central hole. In order to provide reflection of
VISAR beam, a piece of 14-m aluminum foil was glued on the rear
surface of the sample. The disk with the glued sample was fixed on
the base ring of the double-tilt sample holder, whose parallel
Fig. 1. SEM image of the cross-section of the studied polyurethane foam (a), and stress-strain diagram obtained after quasi-static compression tests performed with a 55-mm initial
thickness foam sample. The strain rate in the test was ε_ ¼ 1:5
103 s1 . The insert shows the determination of the collapse stress.
orientation to the front of the projectile had been preliminarily
adjusted with an accuracy of 0.1 mrad. The schematics of these
experiments are shown in Fig. 2a. In the tenth experiment, aimed at
measuring the dynamic tensile (spall) strength of the foam, the
impact was produced by a free-of-foam PMMA disk (primary
impactor) on the 1-mm thick foam sheet (secondary impactor)
separated from the sample by a spacer ring of 5-mm thickness. As
result, the foam sample (Fig. 2b) was struck by a thin foam impactor.
The impact velocity, ranged from 43.5 to 605 m/s, was controlled
by electrical charged pins. The uncertainty of the measurement of
the impact velocity did not exceed 1% of the measured velocity
value. The impactor-sample misalignment controlled by the trigger
pins did not exceed 1 mrad in all experiments. Depending on the
impact strength, the velocity of the rear sample surface was moni-
tored by VISAR with delay lines providing velocity constants of 96.4,
224.0, and 407.2 m/s per fringe. The parameters of the ten planar
impact experiments are listed in Table 1.
3. Experimental results
The VISAR-recorded velocity histories obtained after symmetric
planar foamefoam impacts are shown in Fig. 3aec.
Except for the waveform obtained after the weakest impact, the
waveforms shown in Fig. 3 are characterized by a two-wave
structure marked as P1 and P2 in Fig. 3a. The PFE1 test was per-
formed in order to show that the presence of the P2 wave is caused
by interaction of the unloading wave generated at the rear surface
of the sample, with the reloading wave generated at the interface
between the foam impactor and the PMMA backing. The stress s -
particle velocity u diagram and the time t - distance (Lagrangian) h
diagram of Fig. 4 are to illustrate such interaction in the case of the
PFE and PFE1 tests. Since both the samples and the impactors are
made of the same foam, the particle velocity u1 behind the shock
fronts P1 propagating through both the sample and impactor with
velocity US is equal to one half of the impact velocity. Accepting for
both shocks the same impact velocity equal to v0 ¼ 312 m/s, yields
for the particle velocity u1 ¼ 156 m/s. The amplitude of this impact-
generated shock is s1 (Fig. 4a). At the arrival of the shock at the
sample free surface it acquires velocity equal to 196 m/s. In the
absence of the PMMA backing, the rear surface of the foam
impactor should be decelerated from 312 m/s to 106 m/s. The
presence of the PMMA backing results in the reloading of the
impactor material from state with stress s1 to state with stress s2.
 E. Zaretsky et al. / International Journal of Impact Engineering 39 (2012) 1e7 3

Table 1
Parameters of the planar impact experiments performed with dense flexible polyurethane foam.

Test impactora samplea imp.velocity, m/s Test impactora samplea imp. velocity, m/s
b
PFA foam foam 43.5 PFE1 foam foam 314
PFB foam foam 72.5 PFF foam foam 384
PFC foam foam 141 PFG foam foam 499
PFD foam foam 235 PFH foam foam 605
PFE foam foam 311 PFASc foam foam 355
a
Thickness of both foam impactors and samples was 8.9  0.1 mm. The impactor diameter is 55 mm, the samples were 60 mm
60 mm squares.
b
The foam sample was a 60 mm
60 mm square of 5.5 (0.1)-mm thick.
c
The thickness of the primary PMMA impactor and of the secondary foam impactor were 11.80  0.01 mm and 1  0.1 mm, respectively.

The reshock R of the amplitude s2 travelling towards the sample
rear surface meets the unloading wave U1, generated at the rear
surface of the 5.5-mm sample or U2 generated at the rear surface of
the 8.9-mm sample. As a result, the partially unloaded shocks R1
and R2 (for 5.5-mm and 8.9-mm samples, respectively) with
amplitude s3 arrive at the sample rear surface, accelerating it from
w1 ¼ 196 m/s (the free surface velocity corresponding to the P1
wave amplitude) to w2 ¼ 354 m/s (the free surface velocity corre-
sponding to the P2 wave amplitude). The later arrival of the P2
wave (Fig. 4b) at the rear surface of the 5.5-mm sample, 47.1 ms with
respect to 44.1 ms in the case of the 8.9-mm sample, is the result of
certain relations between the speeds of unloading (U1,U2) and
reloading (R, R1,R2) waves. It is apparent from Fig. 4, that for accu-
rate description of the second, P2, wave knowledge of the s1  s2,
s2  s3, and s3  0 paths is required. Since reliable data are avail-
able only for the description of the s1  s2 path, the analysis in the
next section will be limited to the shock states s1 behind the P1
wave and to those produced by unloading from the s1 states.
The waveform recorded after a single spall-oriented impact is
shown in Fig. 5. Since the foam impactor used in this experiment
was very thin, a part of the waveform amplitude is eliminated by
hydrodynamic decay. (The 355-m/s velocity of the primary PMMA
impactor corresponds to some 600-m/s impact of the secondary
foam impactor on the foam sample.) Although the spall signature
(the velocity pull-back Dw) at the waveform of Fig. 4 is very small,
of about 3 m/s, the post-spall oscillations of the free surface suggest
that the foam sample was fractured by the tensile pulse caused by
collision of the rarefaction waves generated at the free surfaces of
the sample and the secondary impactor.
4. Impact response of the foam
4.1. Hugoniot of the foam
In order to obtain the Hugoniot states at the top of the P1 wave
from the free surface velocity histories shown in Fig. 3, the velocity
US of the propagation of the wave front should be determined. Since
the wave fronts are not step-like, we assume US for the velocity of
propagation of the wave half-height. The wave leading edge trav-
elling with velocity ULE higher than US gives rise to the multiple
wave reflections, as it is shown in Fig. 6a. These reflections distort
the timing of the arrival of the wave half-height at the sample
surface. Accounting in that the ULE is the velocity of propagation of
low-stress perturbations, we assume that all the multiple reflec-
tions shown in Fig. 6a propagate with the same velocity equal to
ULE. In such case, the Lagrangian velocity US is determined by the
times t1 and t2 of the arrival of, respectively, the wave leading edge
and the reflection from the wave half-height at the free surface of
the sample of thickness d.
d  ðt2  t1 ÞULE =2
US ¼ 2 (1)
t2 þ t1
The principal Hugoniot of the foam, namely the stress s1 and the
specific volume V1 at the top of the P1 wave, may be obtained by
applying the mass and momentum conservation laws to the wave
front propagating with the velocity US :
V1 ðUS  u1 Þ US u1
¼ ; s1 ¼ (2)
V0 US V0
where V0 ¼ 1=r0 ¼ 2:44
103 m3 =kg is the initial specific
volume of the foam. Using conservation of mass and momentum
across the unloading wave (path u1w1 in Fig. 4a), yields the esti-
mates of the average velocity of the unloading wave UUL and of the
specific volume VUL of the foam unloaded after loading up to the
stress s1. The corresponding estimates are given in Table 2.
The velocities ULE, US, and UUL and the difference ULEUS are
shown in Fig. 6b as a functions of the particle velocity u1 ¼ v0/2 at
the top of P1 wave.
The linear expression US ¼ 14.8 þ 1.318u (dashed line in Fig. 6)
fits the data on the velocity of the propagation of the wave half-

a foil
b
foam impactor spacer ring
PMMA backing velocity and primary PMMA
trigger pins impactor

from laser

to VISAR

holed PVC disk

secondary foam
foam sample
impactor
base ring of
the sample holder

Fig. 2. Schematics of the symmetric planar impact tests with primary foam impactor (a), and of the spall-oriented experiment with the secondary foam impactor (b). In both cases
the velocity of the free rear surface of the foam sample was monitored by VISAR.
4 E. Zaretsky et al. / International Journal of Impact Engineering 39 (2012) 1e7

Fig. 3. Experimentally recorded waveforms obtained after "strong" (a) and "weak" (b) impacts. The corresponding impact velocities are shown along the waveforms. The waveforms
recorded after stronger (72.5 m/s and higher) impacts have a two-wave structure containing P1 and P2 waves propagating with different velocities. The 141-m/s waveform is shown
in both (a) and (b) as a scale. The 311-m/s waveform (PFE test) of (a) is shown in (c) together with the velocity history recorded after impacts of similar velocity (314 m/s, PFE1 test)
on the sample of 5.5-mm thickness. The values of the sample thickness are shown near the bottom part of the waveforms.

height with Pearson correlation coefficient p ¼ 0.9998 implying
more than four-fold compression of the foam at strong shocks. It is
apparent from Fig. 6b that the difference between ULE and US
decreases with the strength of the impact and for the particle
velocities higher than 200 m/s, it becomes negligibly small, 2e3 m/
s. The rise times of the waveform fronts (the time intervals between
0.1w1 and 0.9w1) decrease from approximately trt ¼ 126 ms for the
weakest, 43.5-m/s, impact to less than trt ¼ 0.43 ms for the
strongest, 605-m/s, one. The rise time of the shock front is usually
associated with the material strength [14]. It is plausible to assume
that in the tests with relatively low impact velocity the rise time is
controlled by the foam crush up, while at the stronger impacts the
steepness of the shock front is controlled by the dynamic viscosity
of the crushed (free of voids) foam.
The linear dependence of US on u yields the foam’s principal
Hugoniot on the stress s - particle velocity u plane
 E. Zaretsky et al. / International Journal of Impact Engineering 39 (2012) 1e7 5

a b Table 2
Foam states behind compressive and unloading waves.

Test u1, ULE, US, s1, V1/V0 w1, UUL, VUL/V0

m/s m/s m/s MPa m/s m/s
0 0 1.0
PFA 21.75 104.7 40.7 0.36 0.465 43.5 40.7 1.0
PFB 36.25 129.0 61.2 0.91 0.408 53 70.7 0.534
PFC 70.50 157.8 109.5 3.16 0.356 99 124. 9 0.461
PFD 117.5 192.6 174.0 8.36 0.325 152 226.9 0.383
PFE 155.5 236.1 222.7 14.2 0.302 198 288.3 0.354
PFE1 157.0 236.1 219.4 14.2 0.295 199 275.4 0.348
PFF 192.0 270.5 266.4 20.9 0.279 240 345.8 0.325
PFG 249.5 345.0 342.3 34.9 0.271 315 419.0 0.321
PFH 302.5 415.9 413.0 51.0 0.266 384 488.0 0.319

s ¼ ð6:053u þ 0:539u2 Þ
103 MPa, where the particle velocity

units are m/s. The foam Hugoniot in stress s - relative volume V/V0
coordinates is shown in Fig. 7 along with the static compression curve
Fig. 4. Stress-particle velocity (a) and time-distance (b) diagrams of the tests PFE and
of Fig. 1b. The departure of the Hugoniot data from the isothermal
PFE1 corresponding to the waveforms shown in Fig. 3c. The impact velocity is assumed
to be equal to 312 m/s in both tests. Waves’ timing (b) corresponds to the arrival of the compression curve, Fig. 7, seems to be caused by large irreversible
leading wave edge at the sample rear surface. heating of the shocked foam. An accurate estimate of the temperature
caused by this heating is, however, hardly possible. Within assump-
tion of constant specific heat CV and constant ratio of Gruneisen
parameter to specific volume G/V, the temperature TH behind the
shock front is related to the temperature T0 ahead of the shock as [15]
  
1þs 2
TH ¼ T0 exp Gε 1 þ ε þ. (3)
6T0 a
Here a is the coefficient of the volumetric thermal expansion, s is
the slope of the material Hugoniot, and ε ¼ ðV0  VÞ=V0 . The
thermal expansion coefficient a of polyurethane is of about
2
104 K1 [16], For the polyurethane foam of close, 300 kg/m3,
density Maw et al. [17] suggest the value Gz1. Using such values of
G and a yields even for modest, V=V0 z0:6, compression
TH z1000 K. Another approach based on the direct thermodynamic
definition G ¼ aCb2 =Cp with characteristic for polyurethane specific
heat Cp z500 J=kg [18] seems to underestimate the TH; for the
strongest, V=V0 z0:27, foam compression in the 605-m/s test it
yields TH z500 K with G ¼ 0.064.

4.2. Foam’s compressive strength

Fig. 5. The waveform recorded after impact of 1-mm foam impactor on the 8.9-mm
foam sample (shot PFAS of Table 1). The insert shows vertical magnification of the As mentioned above the material strength mechanisms are
spall-related signal with the velocity pull-back Dwpb z3 m=s. responsible of the shock front rise time trt. The latter may be
quantified through the assessment of the average strain rate during

a b
Lagrangian velocities US, ULE and UUL, m/s

time
Velocity difference ULE - US, m/s

400 80
t2
ULE

300 60

ULE
US t1 200 40

ULE 100 20

h, mm
0 0
0 8.9 0 100 200 300
particle velocity u, m/s

Fig. 6. Time t - Lgrangian distance h diagram of the propagation of the spread shock front through the foam sample of 8.9-mm thickness (a) and velocities of propagation of the
leading edge ULE (filled circles), of the wave half-height US (open circles), and of the release wave UUL (open triangles) as a function of the particle velocity at the top of P1 wave (b).
The dashed line is the linear approximation US ¼ US0 þ su. The crosses show the difference ULE  US (right ordinate).
6 E. Zaretsky et al. / International Journal of Impact Engineering 39 (2012) 1e7
Fig. 7. Hugoniot of the foam on the s  V/V0 plane. The dashed line is the maximum
foam compression V/V0¼(s1)/s ¼ 0.241. The insert shows the low pressure Hugoniot
data together with the static compression curve of Fig. 1.
the foam compression: h_εi ¼ ðV0  V1 Þ=V0 trt , where V1 is the foam
specific volume at the top of P1 wave. The average strain rates h_εi,
estimated on the basis of the recorded waveforms are shown in
Fig. 8 as a function of the Hugoniot stress (coordinates are loga-
rithmic). It is apparent that the point with s ¼ 3.2 MPa marks
a boundary between two modes of the foam impact response.
Below this point the foam structure participates in the establish-
ment of the shock front form. Beyond the 3.2-MPa stress the foam
structure is completely crushed and the shape of the wave front is
maintained by the viscosity and the thermal conductivity of the
crushed material. While the stress s ¼ 3.2 MPa marks apparently
the point beyond which the voids cease to exist in the compressed
foam, it is hardly possible to fix the stress of the onset of the foam’s
crush up. Recalling that in the weakest test, PFA test of Table 1, the
impact velocity, v0 ¼ 43.5 m/s, and the maximum waveform
velocity coincide, and that the departure of the corresponding
Fig. 8. The P1 wave rise time (from 0.1u1 to 0.9u1) as a function of the P1 amplitude s1
in logarithmic coordinates. The 3.2-MPa point marks the change of the foam response.
Hugoniot point from the isothermal compression curve is negli-
gibly small, we can conclude that the foam deformation during this
test is still reversible. Respectively, the Hugoniot stress in this test,
s ¼ 0.36 MPa, may be considered as the stress of the onset of the
foam’s crush up under shock loading.
4.3. Dynamic tensile behavior
The experimentally measured value of the velocity pull-back
Dwpb allows estimating the dynamic tensile (spall) strength ssp of
the foam [19]
1
ssp ¼ r0 C Dwpb (4)
2
where r0 ¼ 409 kg/m3 is the initial foam density, and C is the slope
of the Raleigh line connecting (on the s  u plane) the unloaded
after loading state of the foam with that corresponding to the
collision of the two release waves, generated, respectively, at the
free surfaces of the foam impactor and the foam sample. The slope C
is determined by the compressibility of the foam under negative
stress. Since the latter is unknown, the value of the P1 wave
propagation velocity U1 z540 m=s corresponding to the impact
velocity 355 m/s (the impact velocity in the spall-oriented experi-
ment), was used as the C estimate. This yields for the dynamic
tensile strength of the foam ssp z0:3 MPa. The tensile strength of
bulk flexible polyurethanes is within 8e40 MPa [20e22]. It is
plausible to assume that the tensile strength of the polyurethane
foam with some 65-% porosity should not be lower than 2.8 MPa.
The measured value of the foam dynamic tensile strength, 0.3 MPa,
is tenfold lower. The difference seems to be related to serious
damaging of the foam during the compressive part of the loading
cycle.
5. Conclusion
A flexible polyurethane foam with initial density r0 ¼ 409 kg/m3
was tested in a series of gun-driven planar impact experiments
accompanied by VISAR monitoring of the free surface of foam
samples. The velocity of symmetrical (foamefoam) impact in these
experiments was varied between 43.5 and 600 m/s. The average
compressive strain rate in these experiments ranged from
3.6
103 s1 to 6.2
105 s1.
The recorded velocity histories made it possible to establish the
principal Hugoniot of the foam in a linear form US ¼ US0þsu ¼
14.8 þ 1.318u, where US is the velocity of propagation of the wave
half-height and u is the particle velocity behind the wave front.
Such Hugoniot implies that under strong shock loading maximum,
the foam compression is V/V0 ¼ 0.241. The velocity of the release
(unloading) wave UUL, determined from the same velocity histories
are found to be some 25% higher than the corresponding US values.
The experiments show that under impact with velocity higher
than 43 m/s (the shock stress s1 ¼ 0.36 MPa), the loading of the
foam is accompanied by the foam crush up and by an irreversible
heating. A reliable assessment of this heating requires, however,
more accurate, than available at present, information on the foam
Gruneisen parameter.
Based on the rise times of the recorded velocity histories, one
can conclude that under impact with velocity higher than 141 m/s
(s1 z3:2 MPa), the foam is completely crushed and the foam
structure ceases to participate in the establishment of the wave-
form front. Starting from this point, the structure of the wave front
is controlled by the viscosity and thermal conductivity of the bulk
foam material.
 E. Zaretsky et al. / International Journal of Impact Engineering 39 (2012) 1e7
The dynamic tensile (spall) strength of the studied foam was
found to be very low, w0.3 MPa, some 10 times lower than expected
for such foam under quasi-static tension. Such low resistance of the
foam to the tensile stress may be caused by the considerable damage
suffered by the foam under shock compression.
To summarize, the VISAR-instrumented gun-driven planar
impact experiment is a useful tool for obtaining constitutive
information about low-impedance materials under strain rates
ranging from 103 to 106 s1.
References
[1] Gibson LJ, Ashby MF. Cellular solids, structure and properties. 2nd ed. Cam-
bridge: Cambridge University Press; 1999.
[2] Kraynik AM, Nielsen MK, Reinelt DA, Warren WE. Foam micromechanics. In:
Sadoc JF, River N, editors. Foams and emulsions. Netherlands: Kluwer
Academic Publishers; 1999.
[3] Lankford J, Dannemann KA. Strain rate effects in porous materials. In:
Schwartz DS, Shih DS, Evans AG, Wadley HNG, editors. Porous and cellular
materials for structural applications. Warrendale, PA: Materials Research
Society; 1998. p. 103e8.
[4] Progelhof RC. Impact measurement of low-pressure thermoplastic structural
foam. In: Kessler SL, Adams GC, Driscoll SB, Ireland DR, editors. Proceedings of
instrumented impact testing of plastics and composite materials. Houston
(TX): ASTM STP 936; 1986. p. 105e16.
[5] Shim VPW, Tu ZH, Lim CT. Two-dimensional response of crushable poly-
urethane foam to low velocity impact. Int Journ.of Impact Eng 2000;24:
703e31.
[6] Ramon O, Miltz J. Prediction of dynamic properties of plastic foams from
constant-strain rate measurements. Journ Appl Polym Sci 1990;40:1683e92.
[7] Sounik DF, Gansen P, Clemons JL, Liddle JW. Head-impact testing of polyurethane
energy-absorbing (EA) foams. J Mater Manufacturing 1997;106(5):211e20.
7
[8] Chen W, Lu F, Winfree N. High-strain-rate compressive behavior of a rigid
polyurethane foam with various densities. Exp Mech 2002;42:65e73.
[9] Song B, Chen W, Dou S, Winfree N, Kang J. Strain-rate effects on elastic and
early cell-collapse responses of a polystyrene foam. Int Journ Impact Eng
2005;31:509e21.
[10] Song B, Lu W-Y, Chen W. Strain-rate e temperature superposition behavior of
Polymethylene Diisocyanate (PMDI) based rigid polyurethane foams with
different densities. DYMAT 2009;2009:1251e6.
[11] Chen W, Song B, Frew DJ, Forrestal MJ. Dynamic small strain measurements of
a metal specimen with a split Hopkinson pressure bar. Exp Mech 2003;43:
20e3.
[12] Kanel G, Razorenov SV, Fortov VE. Shock wave phenomena and the properties
of condensed matter. New-York, Berlin, Heidelberg: Springer; 2004. p. 33.
[13] Barker LM, Hollenbach RE. Laser interferometer for measuring high velocities
of any reflecting surface. J Appl Phys 1972;43(11):4669e75.
[14] Grady DE. Structured shock waves and the fourth-power law. J Appl Phys
2010;107:013506.
[15] Duval GE, Graham RA. Phase transitions under shock-wave loading. Rev Mod
Phys 1977;49(3):523e79.
[16] Mateen A, Shamim A. Thermal expansion of rigid polyurethane, reinforced by
hammer-milled glass fibre. J Mater Sci Lett 1989;8:927e30.
[17] Maw JR, Whitworth NJ, Holland RB. Multiple shock compression of poly-
urethane and syntactic foams. In: Schmidt SC, Tao WC, editors. Shock
compression of condensed matter -1995; AIP Conference Proceedings 270.
Woodberry, NY: American Institute of Physics; 1996. p. 133e6.
[18] Liao SK, Jang SC, Lin MF. Phase behavior and mechanical properties of
siloxane-urethane copolymer. J Polymer Res 2005;12(2):103e12.
[19] Antoun T, Seaman L, Curran D, Kanel GI, Razorenov SV, Utkin AV. Spall frac-
ture. New-York, Berlin, Heidelberg: Springer; 2002. pp. 90, 99.
[20] Datta J, Kacprzy M. Thermal analysis and static strength of polyurethanes
obtained from glycolisates. J Thermal Anal Calorimetry 2008;93(3):753e7.
[21] Smith TL. Tensile strength of polyurethane and other elastomeric block
copolymers. J Polymer Science: Polymer Physics Edition 1974;12:
1825e48.
[22] Jellinek HHG, Martin F, Wegener H. Tensile strength of polyurethane exposed
to nitrogen dioxide. J Applied Polymer Science 1974;18:1773e8.