CP620, Shock Compression of Condensed Matter - 2001
edited by M. D. Furnish, N. N. Thadhani, and Y. Horie
© 2002 American Institute of Physics 0-7354-0068-7/02/$ 19.00
EFFECTS OF INITIAL TEMPERATURE ON THE SHOCK AND
RELEASE BEHAVIOR OF FILLED AND UNFILLED EPOXIES
M. U. Anderson, R. E. Setchell, and D. E. Cox
Sandia National Laboratories, Albuquerque, NM, 87185
Abstract. The shock and release behaviors of both alumina-filled and unfilled epoxies are examined in
planar impact experiments over an initial temperature range of -50 to +70 C. The materials under
investigation can be used as encapsulants in applications such as shock-activated pulsed power supplies.
Characterizing the effect of initial temperature on their dynamic behavior under shock loading and
subsequent release is part of a larger effort to develop predictive numerical models for the performance of
these power sources. In the current study, transmitted waveforms were recorded using laser interferometry
(VISAR) to examine initial temperature effects on the inelastic behavior in these materials. Current results,
combined with the results of previous studies of these epoxies at ambient temperature, show a clear trend of
increasing inelastic features in compressive waveforms with decreasing temperature. Temperature effects
are more pronounced in the filled samples. In both cases, release wave speeds decrease with decreasing
temperature.
INTRODUCTION
The materials we have examined include Epon
828 resin with 16.7% (by weight) Z hardener (2)
and a 0.42 volume fraction of alumina filler, made
with the mixing and curing procedures reported by
Munson et al. (3). For simplicity, this material will
be denoted by Fl in the text and figures that follow.
A second filled epoxy, denoted by F2, uses Epon
826 resin (2), 25% (by weight) of an amine hardener
mixture, and the same alumina volume fraction and
curing process as Fl. Two unfilled epoxies, denoted
by UF1 and UF2, have also been examined. UF1
epoxy is F2 without the alumina filler. UF2 epoxy
is 828/DEA, consisting of Epon 828 resin (2) mixed
with DEA (diethanolamine) hardener in a 100:12
weight ratio, and cured at 90°C for 16 hours.
In ambient-temperature studies (1) we examined
the dynamic response of epoxy samples using two
types of planar-impact experiments conducted on a
63.5-mm diameter gas gun (Fig. 1). In reverseimpact experiments, epoxy samples backed by lowdensity carbon foam were mounted as projectile
facings. The targets consisted of a 1.6-mm thick
The shock-induced depoling of an encapsulated
ferroelectric ceramic has been utilized for pulsed
power generation for many years. More recently,
interest in numerically simulating the operation of
pulsed power devices has motivated a number of
new experimental and theoretical efforts. The stress
history experienced by a ferroelectric element
during device operation is strongly affected by the
shock compression and release properties of the
surrounding encapsulant. However, only a limited
data base is available for developing improved
models for the dynamic response of these materials.
Consequently,
filled
and
unfilled
epoxy
encapsulants initially at ambient conditions were
examined in a previous study (1). In the current
study, we examine the differences in dynamic
behavior under shock loading and subsequent
release that result from varying the initial
temperature.
669
polymeric solid polymethyl methacrylate (PMMA),
which led to the development of nonlinear
fused-silica buffer followed by a 12.7-mm thick
fused-silica window. A diffusely reflecting film of
-54C(Ref.5)
20 C (Ref. 1)
FUSED SILICA BUFFER/WINDOW
EPOXY SAMPLE
CARBON FOAM
0.2
0.3
0.4
0.5
0.6
0.7
0.8
PARTICLE VELOCITY (km/s)
FIGURE 2. Hugoniot data for filled and unfilled epoxy, with
initial temperature effects shown for filled epoxy Fl (5).
\ FUSED
SILICA
FIGURE 1. Configurations for reverse-impact and transmittedwave experiments at ambient temperature.
aluminum was deposited at the interface, and a dualdelay VISAR system was used to record the particle
velocity history at this location. The well-known
elastic properties of fused silica were used in a
method-of-characteristics calculation to find the
corresponding particle velocity history at the impact
interface. The final value of particle velocity at this
interface, the corresponding axial stress given by
fused silica properties, and the measured impact
velocity provide a direct measurement of an epoxy
Hugoniot state.
Examples of Hugoniot data
obtained in this fashion are shown in Fig. 2. In
transmitted-wave experiments, fused silica was
impacted into a stationary epoxy target backed with
a fused silica buffer and window. Impact conditions
were nominally the same as in the reverse-impact
experiments. Typical wave profiles after shock
waves had propagated through 6-mm thick unfilled
and filled epoxy samples are shown in Fig. 3. The
final shock pressure in both cases was
approximately 1.6 GPa. The unfilled sample shows
an initial shock followed by a progressively slower
increase as an equilibrium state is approached. This
profile is very similar to those observed in the
0
0.2
0.4
0.6
0.8
1
TIME (MCROSECONDS)
FIGURE 3. Transmitted waveforms in 6-mm thick, unfilled
(UFl) and filled (F2) epoxy samples at ambient temperature (1).
viscoelastic material models (4). The filled sample
shows an extended rise time and more pronounced
rounding, characteristic of strongly dispersive
behavior. Waveforms in this material were first
recorded by Munson et al. (3), who modeled their
results using a rate-dependent formulation for a
Maxwell solid. Initial temperature variations were
not explicitly addressed in this formulation nor in
the nonlinear viscoelastic modeling.
Initial temperature effects on the Fl shock
Hugoniot between approximately 0.5 and 6.0 GPa
were examined by Lee et al. (5). Quartz pressure
transducers located behind projectile facings were
used to determine impact stresses in cooled and
heated Fl targets. At -54 C they found that the
shock pressure at a given particle velocity was
670
bonding fused silica windows resulted in cracking of
the window due to large tensile stresses that
developed during target cooling. To avoid this
difficulty, a "floating-window" technique was
developed for cold experiments. The floatingwindow design is shown in Fig. 5.
higher than the ambient value by an amount that
decreased from 10% to less than 4% as shock
pressures increased. Conversely, at +74 C shock
pressures were lower than ambient values by similar
amounts.
Initial density variation over this
temperature range is less than ± 0.2%. Polynomial
fits to these results are included with the ambienttemperature Hugoniot data in Fig. 1.
"FLOA TING-
EPOXY SAMPLE
FUSED SILICA BUFFER/WINDOW
/
TARGET DESIGN FOR COLD AND HOT
EXPERIMENTS
II
n
<4——
1
/ 1
SPRING-LOADED
PLATE
——^
VISAR
i ia -^"••••"•r -..____
The focus of the present study was to examine
how propagated wave structures depended on initial
temperature.
Preliminary experiments were
conducted with target configurations that utilized
high-impedance sapphire buffer/windows, which
previously had been used successfully in cold
experiments on high-impedance ceramics. Figure 4
shows profiles obtained in this fashion using Fl
samples, with other experimental conditions
matching those used for the profiles in Fig. 3. A
more dispersive response is apparent for the cold
VACUUM GREASE
i
il^"C
s
THERMALLY
CONDUCTIVE
GREASE
1
1 COPPER FLOW CHANNEL
\
ALUMINUM TARGET CUP
FIGURE 5. "Floating" fused-silica window design.
Target heating results in compressive stresses in
bonded fused silica windows, and normal potting
procedures could be followed for high-temperature
experiments.
A modified NESLAB ULT-80
recirculating heat exchanger was used to flow either
denatured ethanol or distilled water through the
copper flow channel for cold or hot experiments,
respectively. Three thermocouples imbedded into
the epoxy samples near their outer radius were used
to continuously monitor sample temperature.
RESULTS
-0.1
0.0
0.1
02
0.3
0.4
0.5
0.6
TIME (MICROSECONDS)
0.7
0.8
0.9
Measured waveforms from transmitted wave
experiments on unfilled epoxy samples (UF2) at
different initial temperatures are shown in Fig. 6.
The nominal impact velocity was 0.547 km/s,
resulting in a final shock pressure of 1.63 GPa for
the 20 C case. The most apparent effect of initial
temperature is the significant reduction in release
wave speed with decreasing temperature. The
transition amplitude between the initial shock front
and the following viscoelastic relaxation to an
equilibrium state also decreases with decreasing
initial temperature, as seen in Fig. 7. The time scale
in this figure is expanded compared to Fig. 6.
1.0
FIGURE 4. Transmitted waveforms in 6-mm thick filled epoxy
samples (Fl), recorded using sapphire windows.
temperature case.
The impedance mismatch
between sapphire and alumina-filled epoxy is large,
however, resulting in a strong compressive wave
reflected back into the epoxy sample that
complicates interpretation of observed waveforms.
A better window choice is fused silica, which
provides a much closer impedance match to the
epoxy samples. Unfortunately, the coefficient of
thermal expansion of fused silica is very small
compared to most other materials. Initial attempts at
671
Measured waveforms from transmitted wave
experiments on alumina-filled epoxy samples (F2) at
different initial temperatures are shown in Fig. 8.
the 20 C case. As indicated in the preliminary
experiments using sapphire windows (Fig. 4), a
more dispersive behavior is apparent as initial
temperature is reduced. Even the high temperature
experiment shows only dispersive behavior, with no
shock front present. Release speeds are observed to
decrease with temperature, as in the unfilled epoxy.
SUMMARY
0.00
-0.20
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
The shock response of both unfilled and filled
epoxies show increasingly inelastic features as
temperature is reduced. The unfilled epoxies
display viscoelastic features, and release speeds
decrease with decreasing temperature.
Filled
epoxies show highly dispersive behavior, with no
elastic regime. Release speeds also decrease with
decreasing temperature. These results show the
need for incorporating temperature dependence in
dynamic response models for these materials.
2.00
TIME (MCROSBCONDS)
FIGURE 6. Transmitted waveforms in 6-mm thick unfilled
epoxy samples (UF2).
0.25
ACKNOWLEDGMENTS
0.00
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
The authors would like to thank Howard Arris and
Doug Adolf of Sandia for the epoxy samples.
Sandia is a multiprogram laboratory operated by
Sandia Corporation, a Lockheed Martin Company,
for the United States Department of Energy under
Contract DE-AC04-94AL85000.
0.20
TIME (MICROSECONDS)
FIGURE 7. Wave fronts from Fig. 6 shown on an expanded time
scale.
REFERENCES
1. Anderson, M. U., Setchell, R. E., and Cox, D. E.,
"Shock and Release Behavior of Filled and Unfilled
Epoxies", in Shock Compression of Condensed Matter
- 1999, edited by M. D. Furnish et al, AIP Conference
Proceedings 505, New York, 2000, pp. 551-554.
2. Registered products of Shell Chemical Company.
3. Munson, D. E., Boade, R. R. and Schuler, K. W., J.
Appl Phys. 49, 4797-4807 (1978).
4. Schuler, K. W. and Nunziato, J. W., Rheol. Acta 13,
265-273 (1974).
5. Lee, L. M., Jenrette, B. D. and Greb, A., Air Force
Weapons Laboratory Report AFWL-TR-87-133 (1987).
0.16^
[ 0.14:
'0.12
> 0.10 -
; o.os:
: 0.06:
»
:
0.04
0.02:
0.00
-02
0.0
02
0.4
0.6
0.8
1.0
12
1.4
1.6
1.8
2.0
22
2.4
2.6
TIME (MCROSECONDS)
FIGURE 8. Transmitted waveforms in 6-mm thick filled epoxy
samples (F2).
The nominal impact velocity was 0.357 km/s,
resulting in a final shock pressure of 1.74 GPa for
672