Presented at the Joint Meeting of the I

I

#iInternational ,Association for the Advancement of High PressureScienceand Technology

' and the

American Physical Society Topical Group on Shock Compression of Condensed MatterJune 28- July 2, 1993

i_--_ jl-_"-' ' '*_-- -_"..................................... _" ...... --_ i i/i I

TIME.RESOLVED SHOCK COMPRESSION OF POROUS RUTILE:WAVE DISPERSION IN POROUS SOIJI}S , .| -.

M. U. Anderson, Ii- A. Graham, and G. T. HolmanSandia National Laboratories, Albuquerque, New Mexico 87185-5800

Rutile (TiO=) samples at 60_ of sold density have been shock-loaded from 0.21 to 6.1 GPa with samplethickness of 4 mm and studied with the PVDF piezoelectric polymer strem-rate gauge. The technique uses acopper capsule to contain the sample which has PVDF gauge i_admges in direct contact with the front and rearsu_ A precise measure is made of _ .ro.topr.es_ stress wave velocity through the ..smnple, as well as theinput andpropagated shock stress. The in/"tia]density ts known from the _ _epm__ on Inecem, and theamoum of shbcl_-compremion is calculated from the meamn'ement of shock _el0_i'ty and input stress. Shockstates and re-shock states are meam_ed. The observed data are _ with _ou_ lmblish, ed .highpremp_data. It is observed thnt mtfie has a ¢_ush.st_.ngth'near6GPa. Propagaied_nse_._vary f_om 234 to 916 nsec. Propagated stress-puLseme t/rees ofshock_itMX, 2AI + Fe20_3Ni +.ld, and5Ti + 3Siarepreseillixt, . . • . ! I................... . . . ! I

INTRODUCTION contm at the PVDF gauge pact_e/uunple interface.'Ilm piemeleazic polymer PVDF meamres mess-rate

Shock compression of highly porous rutile (TiO2) directly and is semitive to sample hnperfections suchhas been investigated with a time-resolved shock as demity gradients, or _m size voids at the PVDFcompression technique. The rutile material was _

chosen to: (1) develop an understanding of a highly I The powder samples were pressed in threeporous single component system under shock I uniform layers using a tool steel die at samplecomp_ and fully qualify the time-resolved shock pressures of 500 MPa for each layei.. The coppercompression technique, (2) provide data for the I conta/nment ring began to mq_and radially at sampleextensive numerical simulation data base on rutilei, presmres above 350 MPa unless externally supported.(3) compare with the inforL_uttion learned through Sample demities achieved averaged 59.59fi, _+ 0.9%.previous sample preservatiou (recovery) work on ! At shock-pressures below_i2 GPa, the

i rutflei, and (4) compare with the e_'-_ng high pressure ! PVDF/sample ".nm2e___nee mismatd!.'_ use of a

shock compression work on rutile as reported by i Imffernmterial toisolate thefix__rearmrfaces of

Bugaeva et aLi The present work is part of an overall the sample from the mpper_'_ TPX wasstudy of shock-induced solid state dtemistry.S-, chosento impedance match the rutile at low pressure.

l avcanvlEtrAI, litn observedtocompre -by..ai.,..S00MrssampleThe lime-resolved shock compression technique premure, clanging the "I_._ed _le to 56.8%

used in the present study subjects the powder sample dense, with a sample thi_ _ of_+_75_m.

i to controlled shock-loadingf=om eithera compressed- "l]heex_ _ _ _ in Figuregas gun, or high explosive loading.i The sample was 1, comists of a powder sample encased in a copperencased in a copper capsule with PVDI_ gauge intrude with PVDF gauge packages at both input andpackages at both input and propagated locations as in propagated locationL A precise measure is made ofFigure 1. High explosive loading is described in the compressive stress-wave velocity through theHolman'ss paper in the present proceedings, sample, as well as the inlmt and propagated stresses.

The material used in the present study was The initial density is known from the sample"Puratronlc"grade, high purity powder from Johnson- preparation process, and the amount of shock-Matthey. Part/de sizes range from 1 - 800_m. Powder compression is calculated from measurement of wavesamples were pressed to 60% of 4.26 gm/cma, velocity and input stress.Difficulty was encountered when pressing the samples The conservation of momentum assumption used

. due to the high crush strength of rutile, consistent with in the calculation is known to be violated by, shock-compression observations, observation of the propagated wave dispersion. The: pressure-versus-relative volume states identified using

Sample Preparation this technique provide a reasonabl_ measure ofThe powder samples were pressed directly into materi,tl response under shock compression even

the instrumented copper capsule to achieve intimate though ti:e exact volume is not identified.

_:k _&_ Dt_TRIig, UTION OF THIS DOOUMENT IS UNLIMITIFt-) i"ft ,..

* 4'" " •

!

._ i; _

.__ _ ..................................... ,

For the toe-to-toe measurement, the transit times are_-eo% chosen from the current-versus-time data as the

s ,;_, leading edge of the pulse referred to as time=0 in.... Figure 1. For half max-to-half max measurement, the

time at which the current returns to the baseline is

_" j taken to be the time and level of propagated stress,a.

4 [ from which the time at half max stress is chosen. Thesolid and open symbols indicate toe-to-toe and half

max-to-half max calculations respectively. The highpressure work of Bugaeva covers the stress range from

al _ ] 20.2 tO 192 GPa for 60% dense rutile with the 20.2

GPa data point shown in the upper left corner.

j • _tar_0 lo

o A I -Figure 2. The dispersed structure of an input shock " " _imt apropagated through 4mm of 60% dense mille shows n/isedmeincreasinginverselywithstress. _.. ,:, .,..

.:.:

The shock-compression behavior of 60% dense | n'-..•mille shows a high crush strength with a well-defined _g , ....."-......path toward solid density that shows good agreement "_-,Lwith existing high pressure data. ........ ....

The pressure-versus-relative volume data are z .......

I ......./shown in Figure 3 for rutile at 60% dense with the P-V v.,_points calculated from known initial density, measured , I I i "?'"z i ._-_-....input stress, and measured wave velocity using u u _ _= _, _ _ zconservation of momentum and mass relationships. ' _ v, duro i

The risetime of the propagated wave is sufficiently Figure 3. Measured stress-versus-relative volumestates for 60% dense Ti02 using toe-to-toe and half !long to require different methods of analysis, the wave max-to-half max wave speeds in calculationsvelocity from toe-to-toe, and half max-to-half max. ............. represented with solid and open symbols respectively

Table I. Summary of time-resolved shock compression experiments

Experiment Input Config- Input Wave Prop. Output Relative Vonumber* ConditionS uratione StressD VelocityE RisedmeF Stressa Volumes .....

_ _ (GPa) (km/s) (nsec) (GPa) _

2496(a) 0_1 2 0.21 1_ (1.1s2) 916(40o) 0.32TPX 1.674(1._9) 1.7602474 (a) 0.420 1 1.29 1.964 (1.822) 490 (194) 1.98 KeI-F 1.477 (1.440) 1.7062479 (a) 0.704 1 3_02 2.099 (1.998) 382 (132) 3.68 KeI-F 1.223 (1.176) 1.676BE055 (b) TNT 3 5.10 2.293 (2.132) 348 (116) 5.11KeI-F 1.035 (0.935) 1.671

; 2471 (a) 1.045 1 _ 2.514 (2.407) 234 (94) 6.52 KeI-F _ 1.662BE061 (b) Comp-B 3 6.10 22313 (2.138) 300 (132) 6.08 KeI-F 0.925 (0.796) 1.684

/la (a) impact loading; (b) explosive loading; B: symmetric impact velocity (km/s) Cu > Cu or explosive type;C: configuration, (nominal thickness): 1: Cu(6.35mm)/Kel-F(125_m)/PVDF(25,m)/Kel-F(125_m)/TiOz(4mm)/Kel-F(125_,m)/PVDF(25_m)/Kel-F(0.74mm)/Cu(9.52mm); 2:Cu(6.35mm)/TPX(9_4mm)/PVDF(25_m)/FE_(25"m)/Ti_2(4mm)/FEP(25_`m)/PVDF(25"m)/TPX(9.4_mm ) .,/Cu(6.35mm); 3: P-O40/explosive(25.4mm)/Cu(12.7mm)/Cu(6.35mm)/Kel-F(125_,m)/PVDF(25_m)/Kel .... :F(125,um)/TiO2(4mm)/Kel-F(125/am)/PVDF(25_m)/Kel-F(0.74mm)/Cu(9.52mm) D: PVDF measurement "E: wave speed toe-to-toe(half max-to-half max) F: Propagated wave toe-to-peak(10-90%) G: PVDFmeasurements (in polymer backing material) H" calculated from PVDF input stress and toe-to-toe wavespeed(half max-to-half max wave speed)

B !. "

|

..... a ..............

,ce ...... _ Properly packaged PVDF provide recording times[ i i. _ _,_ _ o _ +_(_.,_ of 6 microseconds, and unusual sensitivity to details! ! i CIO_oNA_ +F.c,OS• ...................".....................i* r_ o 511+_ while measuring stress-rate profiles.

I _ .... i..__ ......._._._. _: The data provide a significant data base oni i i i materials behavior adequate for the advanced

_| ..........._ ...................._......._'"-'i_i .............i........ modeling required to realistically describe shock

i i _ --_._e_ compression loading of porous solids..........."......................._........................'_.... I I , 12: "F,, i i i I I 1

•. ...................i.......................!.......\i I 1 I o.,! ACKNOWLEDGEMENTS

_'Xe i i _ [ ] i | The authors would like to acknowledge the.......illllrJ _ fabrication support of Heidi Anderson and the

, o_ o i _ o explosive firing site support at New Mexico Tech.". 2 , • Supported by the U.S. Department of Energy underInlmtStreea(GI_) • o.4 ut_mm,--_ contract number DE-AC04-76DP00789.

Figure 4. The 4 mm propagated wave risetime-versus.- REglgRENCES

input stress is shown for five different material L R.A. G_.Sofids .under High Press_e Shocksystems. Strongly material-dependent response is uompre_on: Mechanics, l-nyslcs aria t:nemistry,observed. Springer-Verlag (1993).

The risetime-versus-input stress relationship of 2. V.A. Bugaeva, M. A. Podurets, G. V. Simakov, G.S. Telegin and R. F. Trunin, "The dynamicthe propagated wave through 4 mm thick samples is compressibility and equations of state of rutile-shown for four different highly porous powder systems structure minerals', zn lzvestiya, Academy ofin Figure 4. The single component systems rutile and Sciences, USSR, Physics of the Solid Earth, Vol.

15, No. 1, 1979, pp.19-25.HMX high explosive_O, the unreactive two component 3. R.A. Graham and N. N. Thadhani, Solid Statesystems, ZAI + Fe2Os s and 3Ni + AI, and the reactive Reactivity of Shock-Processed Solids', in Shocktwo component system 5Ti + 3Sizz. Waves in Materials Science, edited by A. B.

The HMX has long risetimes at low stress Sawaoka, Springer-Verlag (1993). '_ ......followed by a rapid decrease in risetime due to the low 4. N.N. Thadhani, "Shock-Induced Chemical

Reactions and Synthesis of Materials', Progress incrush-strength, and the onset of reaction. Rutile Materials Science, _ 117-226, 1993.demonstrates a more constant trend of risetime 5. R.A. Graham, B. Morosin, Y. Horie, E. L.decrease with increasing stress and a very high crush- .Venturini, M_ Boslough, M. J. Carr, and D. Lstrength without the effects of reaction. The 2AI + Williamson, Chemical Synthesis Under High

Pressure Shock Loading, in Shock Waves mFe_Os mixture exhibits a rapid decrease in risetime i Condensed Matter, edzted by Y. M. Gupta,due to the aluminum crush-up followed by very little Plenum (1986) pp 693-711.risetime decrease due to the FetOs crush-up. The 5Ti 6. R.A. Graham, M. U. Anderson, Y. Horie, S-K.+ 3Si mixture crush-strength is initially similar to 2AI You, and G. T. Holman, "Pressure Measurements+ FezOs followed by decrease in risetime due to _ in Chemically Reacting Powder Mixtures with thereaction. The 3Ni + AI mixture show 13 and 26 J Bauer Piezoelectric Polymer Gauge', Shock

1 Waves, in press.nanosecond risetimes for 1.5 and lmm sample 7. R.A. Graham, M. U. Anderson, F. Bauer, and R.thicknesses respectively. E. Setchell, "Piezoelectric Polarization of the

Ferroelectric Polymer PVDF from 10 MPa to 10DISCUSSION & COI_!CLUSION GPa: Studies of Loading-Path Dependence',

Shock Compression of Condensed Matter - 1991,The shock-compression of rutile shows a pressure- eds. S. C. Schmidt, et al, North-Holland, (1992),

versus-volume relationship that crushes toward solid pp 883-886.density in the pressure range of the present study, and 8. G.T. Holman Jr., R. A. Graham, M. U.

Anderson, present proceedings, in press.shows good agreement with existing high pressure 9. D.E. Wackerbarth, M. U. Anderson, R. A.work. The propagated wave risetimes show a Graham, SAND92-0046, February, 1992, Sandiaconsistent trend of increasing inversely with input National Laboratories. -'stress. 10. S. A. Sheffield, R. L. Gustavsen, R. R. Alcon, R. _'

The rutile, HMX, 2AI + Fe2Os, 3Ni + AI, and A. Graham, M. U. Anderson, presentp,_roceedings, in press.

5Ti + 3Si data illustrate the strongly material- 11. r_..W. Dunbar, R. A. Graham, M. U. Anderson,dependent risetime characteristics which appear to be G.T. Holman Jr., N. N. Thadhani, presentboth morphologically and material dependent, proceedings, in press.

DIS

of their cns oyccs._ _,y ..... 1, -=Twtbcrcm',aor _aY pi ........... _an_tzs_aesLor use"--- '---J lisbilit" _ .._,_,td_uty XOfuw _-'-I, .....fuincss of any tmormauu_, a_m,-_-,_ v-,--'- _. " tO spc-

• " C%_ Of SaVSCBoy Lsa_,__,,-.._ ..,•,,,.,,_ mr_ eoes tlOta_l_camuuy _ _'-" _ .... ,-v thcs'coL'_'"" "T. • " _d_iSZS1_ _ assY'_.. "_mca-d_oa. or rsvormS t__thc.U_ _...mdo aotoa:eaSy su_tc or"r_,c views and opinions of zuusors _ ,,'-'--1_ _..,..._.,s'a_cc%those of the Umted Stn_s Govzs'ms_a% suuy

°O

II