Sei sulla pagina 1di 8

High temperature impact response of 998 alumina

E. B. Zaretsky

Citation: Journal of Applied Physics 114, 183518 (2013); doi: 10.1063/1.4830014


View online: http://dx.doi.org/10.1063/1.4830014
View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/114/18?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in


Impact response of potassium bromide in 166880 K temperature range
J. Appl. Phys. 108, 073528 (2010); 10.1063/1.3486015

The failed strength of ceramics subjected to high-velocity impact


J. Appl. Phys. 104, 013533 (2008); 10.1063/1.2955456

Delayed failure in a shock-loaded silicon carbide


J. Appl. Phys. 97, 113513 (2005); 10.1063/1.1923161

Characterization and evaluation of silicon carbide for high-velocity impact


J. Appl. Phys. 97, 093502 (2005); 10.1063/1.1881798

Direct evidence for compressive elastic strain at ground surfaces of nanocomposite ceramics
Appl. Phys. Lett. 86, 061909 (2005); 10.1063/1.1862754

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50
JOURNAL OF APPLIED PHYSICS 114, 183518 (2013)

High temperature impact response of 998 alumina


E. B. Zaretsky
Department of Mechanical Engineering, Ben-Gurion University of the Negev, P.O. Box 653,
Beer Sheva 84105, Israel

(Received 9 September 2013; accepted 28 October 2013; published online 13 November 2013)
The dynamic compressive strength (amplitude of the elastic precursor wave) of alumina of 99.8%
purity has been studied in a series of planar impact experiments with the 0.25 to 3-mm thick
alumina samples, preheated to 1300 K. The dependence of the dynamic tensile (spall) strength of
alumina on the ultimate compressive strength was studied with 2-mm thick samples at 300 and
1300 K. The compressive strength of alumina was found to decline by some 30%-35% between
300 and 1300 K, while the absolute values of the spall strength and its dependence on the peak
stress were found to stay almost unchanged between 300 and 1300 K. Over the studied temperature
interval, the elastic precursor waves in alumina decayed with propagation distance with the decay
parameters almost independent of temperature. The availability of decay parameters made it
possible to determine the strain rate function sc_ ; T of alumina and to derive some conclusions
regarding the onset of inelastic deformation in this ceramic. V C 2013 AIP Publishing LLC.

[http://dx.doi.org/10.1063/1.4830014]

INTRODUCTION where E0 q0 c2l is the material longitudinal (constrain)


modulus. Equations (1) and (1a) imply that measuring the
In the course of the last five decades, the impact response
variation of the amplitude rHEL of the elastic precursor wave
of aluminum oxide ceramics (Al2O3, alumina) has been the
with propagation distance offers the simultaneous determina-
subject of numerous studies.18 As a result, the impact
tion of both the shear stress and the initial plastic strain rate
response of different aluminas is well defined and the influ-
ence of the structural features, such as porosity or purity on at the very beginning of the inelastic deformation. The avail-
the dynamic mechanical properties, is established.914 At the ability of the material strain rate function s sc_ enables
same time, the mode, either ductile or brittle, of the inelastic deriving some conclusions regarding the mode of the mate-
deformation of alumina shock-loaded above its Hugoniot rial behavior at this stage. A number of attempts to capture
elastic limit (HEL) is still unclear. The cusp or kink which the decay of elastic precursor wave in alumina3,20 were not
appears as HEL at the stress or velocity histories recorded in systematic and qualitative only. The main problem of these
planar impact tests with ceramics does not allow determining works lay in the use of the samples with thickness that varied
the character of the inelastic deformation in the tested mate- within a relatively narrow range while the samples them-
rial. Some attempts to resolve the issue by loading the selves were relatively thick, 3 mm or more. The thickness of
pre-stressed ceramic sample15 or by loading the plane-parallel the samples imposed the rHEL h measurements over propa-
sample by convex impactor16 were relatively successful. Such gation distances in which the stress derivative drHEL =dh is
methods, however, are hardly applicable when the test tem- small and the scatter of rHEL values is relatively large.
perature differs from the ambient. The experimental tech- Marom, Sherman, and Rosenberg21 carried out the most suc-
nique that is capable of providing information about the cessful attempt to study the precursor decay in alumina, even
character of the elastic-inelastic transition in a given material though they considered the decay as an artifact caused by the
both at room and elevated temperatures is the measurement slow response of the employed gages (manganine). The scar-
of the decay of the elastic precursor wave with propagation city and uncertainty of the available information about decay
distance.17,18 In acoustic approximation,19 the decay of the of elastic precursor wave and the potential of such measure-
stress at HEL is related to the plastic strain rate c_ as ments to enhance our understanding of the onset of inelastic
deformation in alumina constitute motivation for the present
drHEL 4 G_c
 ; (1) study. Another motive was the lack of information about
dh 3 cl high temperature alumina response. Under quasi-static load-
where h is the Lagrangian coordinate, G q0 cs is the mate- ing conditions, the compressive strength of alumina
rial shear modulus, q0 is the initial material density, and cs decreases four-fold between 300 and 1300 K, while the ten-
and cl are the shear and longitudinal speeds of sound, respec- sile strength stays almost unchanged.20 No information
tively. The latter is assumed equal to the velocity of elastic regarding the dynamic compressive or tensile (spall) strength
precursor front. In isotropic material, the shear stress at HEL of any commercial alumina is available.
sHEL GrHEL =E0 and Eq. (1) may be rewritten as The purpose of the present study of the high temperature
impact response of alumina ceramic was to determine the
dsHEL 4 G2 c_ relations between the shear stress and the rate of the alumi-
 0 ; (1a)
dh 3 E cl nas inelastic deformation at its very beginning. It was also

0021-8979/2013/114(18)/183518/7/$30.00 114, 183518-1 C 2013 AIP Publishing LLC


V

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50
183518-2 E. B. Zaretsky J. Appl. Phys. 114, 183518 (2013)

of interest to determine the effect of the temperature on these In the preheated tests, the impactor-sample misalign-
relations and on the dynamic strength parameters of alumina. ment was not controlled due to lack of space. The estimates
This was done by measuring the yield and spall strengths of the misalignment based on the 2-ns rise time of the elastic
and the decay of the elastic precursor wave in alumina sam- front in low-velocity tests allow us to assume that the mis-
ples preheated to different temperatures. alignment did not exceed 1 mrad. The pin-based measure-
ments of the impact velocity were backed by fiber-optic
MATERIAL AND EXPERIMENTAL measurements of the projectile velocity.18 As result, the
velocity uncertainty did not exceed 1% at RT tests and 2% at
Square 10 mm  10 mm samples with 0.25, 0.5, 1.0,
the tests with preheating.
2.0, and 3.0 mm nominal thickness cut from one block of
The 998 alumina has been tested in two series of the
99.8% Al2O3 glass-free low-porosity (less than 0.5%) alu-
planar impact experiments. In the first series, the samples of
mina were obtained from Micro Ceramics Ltd., Carmiel,
all nominal thicknesses were shock loaded by tungsten
Israel. An average diameter of globular Al2O3 grains of the
impactors accelerated in 25-mm bore 6-m length gas gun up
block measured based on the fractured block surface did not
to velocities of about 1050 m/s. In these experiments aimed
exceed 2 lm. Both 10  10-surfaces of the samples (of
at measuring the amplitude of elastic precursor wave (HEL),
0.2-mrad parallelism) were polished, one surface was coated
the velocity histories of the free surfaces of alumina samples
with 1-l layer of aluminum (tests at 300 and 800 K) or
were recorded with use of VISAR (Velocity Interferometer
gold (1300-K tests). Prior the impact testing, the density
System for Any Reflector24) with 407.2-m/s per fringe con-
q0 , the longitudinal cl , and shear cl speeds of sound were
stant. In addition to the room temperature measurements, the
measured with 3-mm thick samples. The average alumina
tests of the first series were performed with 800 and 1300-K
properties are q0 3:95932 g=cc, cl 10:541km=s, and
preheating. In the second series, the 2-mm thick alumina
cs 6:231km=s. The resulting bulk speed of sound and
samples were loaded by 0.5-mm copper impactors with
Poissons ratio were cb 7:691km=s and  0:231,
velocities between 80 and 500 m/s. In these tests aimed at
respectively.
measuring the dynamic tensile strength of alumina, the initial
The samples assigned for impact tests with preheating
temperature of the samples was 300 and 1300 K. Keeping in
were assembled with Kanthal heaters as it is shown in Fig. 1.
mind a possible degradation of the spall strength of alumina
The test assembly was similar to that used previously with
with compressive stress, the VISAR constant in these series
the 58-mm gun in impact experiments with preheating23 but
was 96.4 m/s per fringe.
slightly modified in order to fit the sample/heater dimensions
for use with a 25-mm gun and to minimize the non-uniform
EXPERIMENTAL RESULTS
temperature caused by these modifications. In particular, the
charged pins were made of 2-bore ceramic rods of 1.4-mm The velocity histories recorded in the tests with differ-
outer diameter with 0.3-mm tantalum wire as conductors, the ently preheated alumina samples of different thickness are
heater coils were made 4 mm longer than the sample and the shown in Figs. 2(a)2(c). All recorded waveforms contain
lateral sample surfaces were screened. All joints, including the elastic precursor wave followed by the ramp terminated
that between the K-type thermocouple junction and the sam- at the ultimate velocity. At all three test temperatures,
ple surface, were made using a high-temperature cement. the decay of the precursor amplitude with the propagation
The uniformity of the temperature field on the preheated distance is apparent. Note that at the front of the elastic pre-
sample was tested in separate, with no impact loading, cursor waves recorded with 0.25- and 0.5-mm samples at the
experiments with a 3-mm sample. Inspection of the front sur- highest, 1353 K, temperature modest spikes make their
face of the sample preheated up to 10001300 K by optical appearance. Such spikes are typical for high temperature
pyrometer has shown that at this preheating the temperature (close to melting point) velocity histories recorded in pure
was uniform within 5 K. The temperature difference between FCC metals18 where they may be attributed to the intense
the rear and the front sample surfaces measured with two multiplication of the mobile dislocations that had limited ini-
thermocouples was about 20 K at 800-K preheating and tial density.
about 10 K at 1300-K preheating. For thinner samples, the The stress at HEL of the 998 alumina was estimated
difference should be lower. using relation

FIG. 1. (a) Schematics of the experi-


mental assembly for planar impact
experiment with the sample preheat-
ing. The cement fixture of the heater
spirals is shown by the dashed line. (b)
Sample view from the impacted side.

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50
183518-3 E. B. Zaretsky J. Appl. Phys. 114, 183518 (2013)

FIG. 2. Free surface velocity histories recorded with 998 alumina samples of different thicknesses (shown on the waveforms) recorded at 300 (a), 800 (b), and
1300 K (c). The values of free surface velocity uHEL used for estimating the stress at HEL in alumina samples are shown by arrows.

1 In contrast to the temperature dependence of rHEL of


rHEL q0 cl uHEL ; (2)
2 pure FCC metals,16,17,24 the compressive strength of the alu-
mina does not display a monotonous increase with tempera-
where the values of density q0 and longitudinal speed of ture. The inelastic deformation in FCC metals is determined
sound cl are temperature-corrected based on the recom- by a unique glide system, 1=2h100if111g with no additional
mended values of Ref. 22. The values of rHEL measured with glide system appearing with the increasing temperature. In
the samples of the same thickness decrease with temperature, alumina, the glide system with the easiest glide is the basal
Fig. 3. It is noteworthy that the trend of the temperature vari- system 1=3h1120i0001. This system, however, does not
ation of rHEL resembles that of the temperature dependence satisfy the von Mises criterion of homogeneous deformation
of the stress rst 1   Y=1  2 . The latter has been of a polycrystalline aggregate and is combined usually with
calculated based on the quasi-static (_e 7  105 s1 ) meas- the pyramidal slip 1=3h1102if1120g to provide five inde-
urements of compressive strength Y T 25 and temperature pendent deformation modes.27 Increased temperature brings
dependence of Poissons ratio  T of alumina.22 into operation additional pyramidal slip systems, namely
1=3h2021if0112g and 1=3h1101if1011g, each of which is
capable of providing 5 independent deformation modes. As a
result, the slip in Al2O3 at 1300 K is predominantly pyrami-
dal. The introduction of new glide systems results in the
decrease of dynamic strength. This indeed takes place in
pure HCP titanium with switching on the pyramidal slip28 or
in pure cobalt with the HCP-FCC phase transformation.23
The velocity histories recorded in spall-oriented tests
with 2-mm alumina samples are shown in Fig. 4. The spall-
related velocity pull-back signals Dupb are present on all
the waveforms except on those obtained in the strongest

FIG. 3. Stress at HEL measured with the 998 alumina samples of 0.25 and
3-mm nominal thickness a function of temperature (circles). Dashed lines
are square fits of presently obtained data. Filled triangles and the dotted line
show the temperature dependence of the stress rst Y 1   =1  2 cal-
culated based on quasi-static compressive strength data of Lucalox ceramic
Ref. 25 and temperature dependence of aluminas Poissons ratio from
Ref. 20. Open triangles are RT values of rst calculated based on the refer- FIG. 4. Free surface velocity histories recorded with 2-mm samples of 998
ence compressive strength data of 998 aluminas of Coors Ceramics and alumina shock loaded at 300 and 1300 K by 0.5-mm copper impactors. The
McDanel companies. impact velocities are shown along the waveforms.

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50
183518-4 E. B. Zaretsky J. Appl. Phys. 114, 183518 (2013)

(450-m/s) tests. In the latter case, the pull-back amplitude temperatures, the variation of the stress excess rHEL  rst
is comparable to the scatter of the VISAR signal. The spall with propagation distance h is fit well by the power function
strength of alumina was calculated based on the measured  a
pull-back signals as h
Dr rHEL  rst Dr0 ; (4)
h0
1
rsp q0 cl Dupb : (3)
2 where Dr0 and a are the fit parameters and h0 1 mm. The
dependences Drh obtained based on the rHEL measure-
For the 450-m/s tests, Eq. (3) gives apparently overestimated ments with differently preheated 998 alumina are shown in
values of the spall strength. The peak stress values in these Fig. 6. The corresponding decay parameters, Dr0 and a, are
experiments exceed the material HEL. The uncertainty of the listed in Table I.
spall strength values determined based on these tests is larger As follows from Fig. 6 and Table I, the Dr0 and a values
than for weaker shots, Fig. 5. obtained with the rHEL sets of different temperatures are
It is apparent from Fig. 5 that the 1000-K preheating close. All the tree sets of point may be fit by a single curve
does not affect strongly the dynamic tensile strength of 998 with average Dr 0 5:160:5 GPa and a 0:4660:04 with
alumina; at both the temperatures, the alumina spall strength Pearsons correlation coefficient 0.98. Combining (1a) and
stays almost constant and is equal to approximately 0.5 GPa (4), one has
as long as the loading stress does not exceed the material
HEL. Stronger loading is accompanied by rapid degradation  a1
a
c2s 4 h0 G
of the alumina spall strength. Ds sHEL  sst Dr0 _
c A_c m ; (5)
c2l 3a cl Dr0

where sst Y=2 and Y is the quasi-static compressive


DECAY OF ELASTIC PRECURSOR WAVE IN 998
ALUMINA strength of the alumuna, and m a=a 1 is the material
strain rate sensitivity exponent. The values of A and m corre-
As follows from Eq. (1), the stress at HEL should sponding to the different test temperatures are listed in the
approach some constant value rst , while the plastic strain two rightmost columns of Table I. Since the dependences
rate approaches zero. For the three testing temperatures, the Dsc_ corresponding to the different test temperatures virtu-
rst value of the presently studied ceramic was determined ally coincide the dependences sHEL Dsc_ sst , shown in
based on the room temperature value of compressive Fig. 7 are based on average values of decay parameters (bot-
strength of 998 alumina ceramics of the McDanel Company tom row of Table I). The presently obtained dependences
and on the temperature dependence of compressive strength sc_ are shown in Fig. 7 along with the shear stress values
of slightly stronger Lucalox alumina.25 For all the studied derived from compressive strength Y of Lucalox ceramics,
obtained in quasi-static experiments in Hopkinson pressure
bar tests, and in a planar impact test.25
The apparent from Fig. 7 agreement between the pres-
ently obtained dependence sc_ and the results of the meas-
urements of alumina strength under low to intermediate
strain rates (from 7  105 to 2  103 s1 ) should be

FIG. 5. Spall strength of 998 alumina estimated as an average of the rmax sp


and rmin max
sp values based, respectively, on the maximum Dupb and the mini-
min
mum Dupb amplitudes of the velocity pull-backs of the noisy signals
shown in Fig. 4. The length of the error bars of the weaker tests is the differ-
ences between the rmax min
sp and rsp . The bottom end of the error bars of the
two strongest shots correspond to rmin min
sp 1=2q0 cb Dupb . The peak stress is FIG. 6. Stress excess Dr rHEL  rst as a function of propagation distance
normalized using the rHEL values obtained with 2-mm samples at the corre- h at different test temperatures. The dashed line corresponds to the average
sponding test temperature. decay parameters Dr Dr  0 h=h0 a .

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50
183518-5 E. B. Zaretsky J. Appl. Phys. 114, 183518 (2013)

TABLE I. Parameters of elastic precursor decay in 998 alumina. the lower values of the strain rate sensitivity exponent,
m 0:25  0:28. Correspondingly, the power in Eq. (4)
Test temperature K rst ; GPa Dr0 ; GPa a A m
should be of about a 0:33  0:39 which, even with regard
300 3.79 4.66 0.49 24.3 0.329 to the measurements uncertainty, is much smaller than that
800 2.99 5.47 0.45 23.6 0.310 found presently for 998 alumina, a 0:4660:04. The close-
1300 1.65 5.05 0.42 22.8 0.296 ness of the measured value of a to 0.5 allows us to assume
Average 5.05 0.46 23.6 0.315 that a substantial part of the initial inelastic deformation in
998 alumina is ductile. At the same time, we are unable to
assert that the brittle component does not participate in the
considered with caution. The strength data of Ref. 25 were onset of inelastic deformation in this material.
Linking the elastic precursor decay in alumina to the
obtained under conditions of uniaxial stress loading whereas
dislocation-related processes allows estimating the density of
the planar impact loading of the present work corresponds to
the uniaxial strain. At the presence of purely axial stress, the mobile dislocations in these processes involved. The contri-
bution of mobile dislocations into plastic strain rate is gov-
onset of inelastic deformation in alumina is brittle.29 The
erned by Orowans equation
planar impact loading implies the presence of radial com-
pressive stress suppressing the crack initiation. As soon as c_ qm bhvi; (6)
the radial compression exceeds a 1.3 GPa threshold (and this
takes place in all presently performed tests with 998 alu- where qm is the density of mobile dislocations having
mina), the response of alumina becomes ductile-like.16,29 Burgers vector b and hvi is the average velocity of these dis-
The ductile character of alumina inelastic response finds locations. It is unlikely that the average velocity hvi of the
some confirmation in the measured values of the decay dislocation glide is determined by thermal activation of the
parameter a. As was previously mentioned,16 the relaxation motion dislocation segments. The highest shear stresses cor-
of shear stress caused by motion and multiplication of dislo- responding to the top of elastic precursor wave in the present
cations should lead to a  0:5 and to the strain rate harden- experiments are of about 67 GPa. At this stress level, the
ing exponent in Eq. (5) equal to m  1=3. dislocation velocity controlled by thermal activation cannot
Based on general energy considerations, Grady and exceed 1 m/s.33 Even in the case of the ultimate dislocation
Kipp30 and Grady and Lipkin31 have shown that in the case density, this velocity is still insufficient for providing the
of brittle failure the exponent in Eq. (5) should also be equal required rates of plastic deformation. It was shown that in a
to m 1=3. Their approach, however, implies that cracks wide variety of metals, the only mechanism that controls the
propagate with the speed of sound. A more accurate account motion of dislocations under shock loading is the phonon
of stress dependent velocity of crack propagation32 suggests drag, namely, the interaction of the cores of mobile disloca-
tions with the lattice phonons.34,35 In such case, the average
dislocation velocity is

sb
hvi ; (7)
B
q
where B B= 1  hvi2 =c2s is the effective phonon drag
coefficient accounting for the fact that the dislocation veloc-
ity hvi has to be subsonic.36 There is no available informa-
tion in the literature regarding phonon drag coefficients of
either alumina or sapphire. It is known, however, that credi-
ble estimates of B may be obtained using Leibfrieds
formula37

d
e
B ; (8)
10cs

where e is the phonon energy density and d is the effective


dislocation width which is of order of magnitude of the
Burgers vector b. Comparison of the drag coefficient calcu-
lated using Eq. (8) and the values of B obtained either by
experiments or by molecular dynamic simulations38 reveals
FIG. 7. Shear stress at HEL as a function of plastic strain rate. Solid lines are
based on the dependence (5) with average decay parameters Dr  0 5:1 GPa that below 0.8Tm Eq. (8) with d  2b provides the drag coef-
and a 0:46. Open circles correspond to the material state after 0.25 and ficient estimates which within 10%15% agree with the
3-mm traverse. Filled symbols correspond to the Ref. 25 measurements of molecular dynamics results. Using Leibfried approach with
Lucalox ceramic compressive strength at 300 and 1200 K and room tempera-
ture HEL data of Munson and Lawrence, Ref. 4, (square) to which in Ref. 25
the data of Ref. 22 on alumina specific heat and thermal
the strain rate of 104 s1 was attributed. Open triangles correspond to the expansion, the values of the drag coefficient in alumina equal
temperature corrected compressive strength of McDanel 998 alumina. to 1.9  105, 5. 8  105, and 10.2  105 Pa s were found

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50
183518-6 E. B. Zaretsky J. Appl. Phys. 114, 183518 (2013)

show unambiguously that the decay takes place in alumina


both at room temperature and at 1300 K and its parameters
may be successfully quantified. It is noteworthy that a
recently published study on the evolution of the shock waves
in SiC ceramic has shown that the amplitude of the elastic
precursor wave in the silicon carbide stays constant and
equal to rHEL 8:72 6 0:17 GPa over 0.5 to 8.3-mm propa-
gation distance.39 Evidently, in the case of the SiC, establish-
ing of the steady two-wave structure has been completed
over a less than 0.5-mm wave passage. At the same time,
neither the cause nor the mechanism of so fast relaxation of
the shear stress is clear.
Another objective of the present study has been con-
cerned with a possibility to recognize, based on the parame-
ters of the elastic precursor decay, the mode, either ductile or
brittle, of the onset of the inelastic deformation. In the case
of the presently studied alumina, the ductile yielding compo-
FIG. 8. The density qm of mobile dislocations in 998 alumina (solid lines)
and in copper of 99.999% purity (Ref. 24, dashed lines) as a function of ini- nent plays substantial role in the immediate post-HEL relax-
tial plastic strain rate c_ at different temperatures. The open and filled circles ation of the shear stress. At the same time, an unequivocal
correspond to the states at the top of the elastic precursor in 3 and 0.25-mm ascription of the relaxation mechanism to one or another
alumina samples. The inset shows the variation of qm with temperature at
mode is hardly possible.
constant strain rate whose values are show next to the dependence.

for the temperatures 300, 800, and 1300 K, respectively. The ACKNOWLEDGMENTS
densities qm of mobile dislocations in alumina as a function
of initial plastic strain rate c_ at different temperatures are The author is grateful to Professor Gennady Kanel for
shown in Fig. 8 together with the densities of mobile disloca- his comments and fruitful discussions. Financial support
tions estimated for the top of elastic precursor wave in pure from Israeli Ministry of Defense via research Grant No.
copper.26 8429793 is gratefully acknowledged.
As apparent from Fig. 8, the density of mobile disloca-
1
tions in alumina at the same rate of the inelastic deformation T. J. Ahrens, W. H. Gust, and E. B. Royce, J. Appl. Phys. 39, 4610
is by an order of magnitude, at least, smaller than that in a (1968).
2
W. H. Gust and E. B. Royce, J. Appl. Phys. 42, 276 (1971).
typical metal. It is possibly, even lower since the estimates 3
W. H. Gust, A. C. Holt, and E. B. Royce, J. Appl. Phys. 44, 550 (1973).
of qm were performed with the assumption that the measured 4
D. E. Munson and R. J. Lawrence, J. Appl. Phys. 50, 6272 (1979).
5
inelastic strain rate is entirely dislocation-based. J. M. Staehler, W. W. Predebon, and B. J. Pletka, AIP Conf. Proc. 309,
745 (1994).
6
D. P. Dandekar and P. Bartkowski, AIP Conf. Proc. 309, 733 (1994).
CONCLUSION 7
N. H. Murray, N. K. Bourne, Z. Rosenberg, and J. E. Field, J. Appl. Phys.
84, 734 (1998).
The response of alumina of 99.8% purity on the shock 8
N. H. Murray, N. K. Bourne, Z. Rosenberg, and J. E. Field, J. Appl. Phys.
loading has been studied in a series of VISAR-instrumented 84, 4866 (1998).
9
planar impact experiments with alumina samples preheated F. Longy and J. Cagnoux, in Proceedings of the International Conference
up to 1300 K. The dynamic compressive strength of the alu- on Impact Loading and Dynamic Behavior of Materials, Bremen, 1987,
edited by C. Y. Chiem, H.-D. Kunze, and L. W. Meyer (DGM-
mina has been found to decline by some 30%35% between Informationsgesellschaft, Oberursel, FRG, 1988), pp. 10011008.
300 and 1300 K. The similarity between the temperature 10
F. Longy and J. Cagnoux, J. Am. Ceram. Soc. 72, 971 (1989).
11
variations of the dynamic and static strengths suggests that N. K. Bourne, Z. Rosenberg, I. G. Crouch, and J. E. Field, Proc. R. Soc.
the mechanism responsible of the onset of inelastic deforma- London, Ser. A 446, 309 (1994).
12
N. K. Bourne, Z. Rosenberg, I. G. Crouch, and J. E. Field, AIP Conf. Proc.
tion in the alumina is virtually independent of the loading 309, 769 (1994).
rate. It follows from the measurements of the decay of the 13
N. K. Bourne, W. H. Green, and D. P. Dandekar, Proc. R. Soc. London,
elastic precursor wave with propagation distance that this 14
Ser. A 462, 3197 (2006).
mechanism is also virtually independent of temperature; D. P. Dandekar, J. W. McCauley, W. H. Green, N. K. Bourne, and M. W.
Chen, Global Mechanical Response and Its Relation to Deformation and
over the entire interval of the studied temperatures, the Failure Modes at Various Length Scales Under Shock Impact in Alumina
decay parameters stay almost unchanged. Between 300 and AD995 Armor Ceramic (Army Research Laboratory, March 2008).
15
1300 K, both the absolute values of the dynamic tensile E. B. Zaretsky and G. I. Kanel, Appl. Phys. Lett. 81, 1192 (2002).
16
(spall) strength and their dependence on the ultimate loading G. I. Kanel, E. B. Zaretsky, A. M. Rajendran, S. V. Razorenov, A. S.
Savinykh, and V. Paris, Int. J. Plast. 25, 649 (2009).
stress do not vary. It is interesting to note that the static ten- 17
E. B. Zaretsky and G. I. Kanel, J. Appl. Phys. 110, 073502 (2011).
sile strength of alumina stays at the same temperature inter- 18
E. B. Zaretsky and G. I. Kanel, J. Appl. Phys. 112, 073504 (2012).
19
val constant and equal to 0.25 GPa.22 G. E. Duvall, in Stress Waves in Anelastic Solids, edited by H. Kolsky and
W. Prager (Springer-Verlag, Berlin, 1964), p. 20.
One of the objectives of the present work was to provide 20
N. H. Murray, N. K. Bourne, and Z. Rosenberg, AIP Conf. Proc. 370, 491
an answer to the question whether the elastic precursor wave (1996).
21
in alumina decays with the propagation distance. The results H. Marom, D. Sherman, and Z. Rosenberg, J. Appl. Phys. 88, 5666 (2000).

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50
183518-7 E. B. Zaretsky J. Appl. Phys. 114, 183518 (2013)

22 32
R. G. Munro, J. Am. Ceram. Soc. 80, 1919 (1997). G. Ravichandran and G. Subhash, Int. J. Solids Struct. 32, 2627 (1995).
23 33
E. B. Zaretsky, J. Appl. Phys. 108, 083525 (2010). S.-Y. Yoon and J.-Y. Lee, J. Korean Cryst. Growth 10(4), 337 (2000).
24 34
L. M. Barker and R. E. Hollenbach, J. Appl. Phys. 43, 4669 (1972). V. A. Alshitz and V. L. Indenbom, Sov. Phys. Usp. 18(1), 1 (1975).
25 35
J. Lankford, J. Mater. Sci. 16, 1567 (1981). U. F. Kocks, A. S. Argon, and M. F. Ashby, Progress in Materials
26
E. B. Zaretsky and G. I. Kanel, J. Appl. Phys. 114, 083511 (2013). Science, 1st ed. (Pergamon, 1974), p. 72.
27 36
J. D. Snow and A. H. Heuer, J. Am. Ceram. Soc. 56, 153 (1973). J. J. Gilman, in Physics of Strength and Plasticity, edited by A. S. Argon
28
E. B. Zaretsky, J. Appl. Phys. 104, 123505(2008). (MIT Press, Cambridge, 1969), pp. 313.
29 37
H. C. Heard and C. F. Cline, J. Mater. Sci. 15, 1889 (1980). G. Leibfried, Z. Phys. 127, 344 (1950).
30 38
D. E. Grady and M. E. Kipp, Int. J. Rock Mech. Min. Sci. Geomech. A. Yu. Kuksin and A. V. Yanilkin, Phys. Solid State 55, 1010
Abstr. 16, 293 (1979). (2013).
31 39
D. E. Grady and J. Lipkin, Geophys. Res, Lett. 7, 255, A. S. Savinykh, G. I. Kanel, S. V. Razorenov, and V. I. Rumyantsev,
doi:10.1029/GL007i004p00255 (1980). Tech. Phys. 58, 973 (2013).

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
141.209.144.159 On: Thu, 18 Dec 2014 17:18:50

Potrebbero piacerti anche