NOLTR 62-197

- THE SHOCK-TO-DETONATION TRANSITIONIN SOLID EXPLOSIVES

C..DJ

13 NOVEMBER 1962

NITED STATES NAVAL ORDNANCE LABORATORY, WHITE OAK, MARYLANn

Ez a*

Reproduced FromBest Available Copy

NO OTS

NOICE: When goverment or other drawings, speci-fications or other data are used for any purposeother %ban in connection with a definitely relatedgoverment procurment operation, the U. S.Government thereby incurs no responsibility, nor anyobliation vhatsoever; and the fact that the Govern-ment may have fornulated, furnished, or in any waysupplied the said drawings, specificationsq or otherdata '.z not to be rega-red by implication or other-wise as in any ranrr licensing the holder or anyother person or corporation, or conveying any rightsor permission to manufacture, use or sell anypatented invention that may in any way be relatedthereto.

S......UNCLASS IFIED• ' •NJLTR 62-197

THE SHOCK-TO-DETONATION TRANSITIONIN SOLID EXPLOSIVES

Prepared by:S. J. Jacobs. T. P. LiddiardJr., and B. E. Drimmer

"ABSTRACT, Experiments on the initiation of detonition incast and pressed explosives (TNT and various cyclotols),subjected to plane shocks introdaced at a plane surface, arepresented and discussed. Shock amplitudes ranging from 28 to140 kilobk:s in the explosive were generated by using planewave lensts with various combinations of donor explosive andbarrier composition and geometry. Vie shock velocity wasobtained in each sample as a function of distance of travelinto the wedje-shaped specimen from distance-time recordsobtained with a smear camera. The resulting curves for castexplosives axe found to be quite different from those foi

pressed explosives. Furthermore, cast TNT exhibited a resultancmalous to the other cast charges. The observatio,ý,a areinterpreted as showing evidence that "hot-spots" must bepresent behind the shock to explain the rapidity with whichthe dotonation is established. Sources of hot-spot formation"are suggested, The reaulta obtained at the lo-wr shockamplitudes when compared to results on the NOL gap test lendsupport to the idea that peak pressure and pressure historyin the shocked elements of explosive are far more importantthan wave shape in determining the time for transitionto detonation.

PUBLISHED APRIL 1963

EXPLOSION DYNAMICS DIVISIONEXPLOSIONS RESEARCH DEPARTMENT

U. S. NAVAL ORDNANCE LABORATORYWHITE OAK, X"YLAND

irj UNCLAS S IF IED

.1

UNCLASS IFI ED

NOLTR 62-197 13 November 1962

WA THE SHOCK TO DETONATION TRANSITION IN SOLID EXPLOSIVES

This report Aescribos recent NOL experimentel] work and thinkingconcerning the phenomenon af initiation of detonation in anexplosive subjected to a plane shock of known peak ainplitude.The necessary information on shock amplitude it obtained by meansof direct measurement of shock velocity in thu experiment. Theseresults taken in conjunction with relates experiments andtheoretical work on the transition from ihock to detonation arebelieved to contribute subptantially to our understanding of howexplosives are initiated. This understandinS is of utmostimportance in defining the sensitivity of explosive and prtpellantmaterials, and in eefining the variables which are in.portant tocharacterizing sensitivity. The ultimate objective of workalong these lines is the establishment of a bavis for safety inhand1ing and reliability ir the use of military ordnance.

This roport was presented as a discussion paper at a DetonationPhenomena session of the Ninth symposium (International) on(nmIbe.4-nn, held in- Corvine 1 Ui 4,vereit on, hr.,¶n- 27 - Se"a--a --... --...... .. - -.- - - - - - - - - - - -.. - - --... ..- .... .. .-,

1962, and sponsored by tha combustion Inutitut4. It willeventually be published aa a contribution to tie Proceedings ofthat Symposium.

4- The work wan porformed under NOL Yask 260, Project LACE.

A previous NOL report to which this work is related appeared isNAVORD Report 5710, confidential.

R. E. ODENINGCaptain, USNConmar.der

C. J ARONSONBy direction

UNCLASSIFIED

UNCLA6SIFIEPNOLTR 62-197

CONTEZ' SPage

INTRODUCTION ........................ 1EXPE.RIMENTAL ........................ 2INFORMATION OBTAINED ................... 4RESULTS . ........................ :.5

D I S CULN. ... ... ... O ... ....SU•MRY.. ...................... .......................... 12ACIK•OWLEDGMENTS ................ ......................... 13REFERENCES.. .................... ........................... 14

TABLES

Table Title Page

1 The Explisives Studied, Their State andDensities ................................ 16

2 Experimental Constants for Equation 1 (U and uin mm/microsecond) ............ .................... .17

ILLUSTRATIONS

Figure '. it e Pgqe

1 Wedge-Test Arrangement ..... ......... ........... .182 Typical NOL Wedge-Test R.-7o.d ........ ............. .193 Pressure-Pa._ticle Velocity Diagram for Determining

Shock Pressure in. the Explosive .... ................ 04 Shock Veleciti, v4 fliu,-ance Three 1 na

(for each Se^: ef Curves, Reading from Left toRight, the Brass Attenuator Thickness ws 0.5,1.0, and 1.5 inches) ......... ....................... 1

5 Shock Velocity vs Distance for Several TNT -.harges . .226 Effect of Input Pressurn (p,) on Distance (S)-.

to-Steady-State Detonation ?n Cast Composition B-3 . .237 Experimental Set-up for Detonation Quenching by

Preshockjand Sketch of Smear-Camera Tracet .......... .24

iiiUNCLASSIFIED

UNCLASW IFIEDNOLa 62-197

XNTRO'rJCT ION

SIt: -,ay be o-b' historical interest to note that many year-•ag3 Cornell Usi.vrsivy was the ozene of another discussion onthe problems of shozk-to-d~tonation traneition. In 1945 asmall grvup of scientists from the OSRD, the National ResearchCouncil of Canada, the Lriay, and the Navy met here to exchangeideau concerning datonation in explosives. G. Hertzberq (1)described to that group some interesting smear-camera recordshe had obtained concernIng the it.itiation of detonation i.solid and liquid explosives. At the same meeting he describeda cardi-gap test which wae probably the first of many to follow.E. Bogja (2) also presented a number of importanit, and at tIhattimn perplexing, experimental observations on the cransitionfrom shock to detcnation. The work dJscussed in that ineeting,and nuch of the work that followed, suffered for lack ofquantitative description of the forces and energies present in;the incoming shock which cause a detonation to form.

After a lapse of over ten years, work began to be reportodin tho open literature which describad in quantitative termsthe build-up to deconation from shocks of kr.own prissuream.tplitudem (3-7). The list of papers has growi. rapidly in morerecent years (8-15). In the majority of these papers the build-up to datcnatio. hts been attributed to an initiation ofchemicai reaction by zither a uniform or a locaiized temperaturerise ar-ociated with the adiabatic comrression, foll.wed by groth M--deterinajed by the cont.ini.ied speed-ý.p of the reaction once begun.An alternate h:,pothesie which postulates the development ofhigh thermal conductivity behind the shock leading to a heatpulse has also Lppeared (1.6-17). The latter hypothesis makesno clear distinction between the behavior of liquids and poly-crystalline solids. The former, more prevalent, viewpointsu plies a framework for explaining differences in behaior:(7 between solids and liquids, (b) between salids formed bydifferent tecbniques (such as by casting cr by pressing), (c)o•2-• due to geometric config-ralt|ons zof the medium under &tidy, and _(d) due to spatial &nd temporal distribution of pressure andflow. The transitioit to detonatioi_ in P liquid exý,losive, when

la pane step shock is induced in it, apppa.'s to be the simplestto explain in its physical aspect- (8, 19-20). Here thetemperaturr rise in a homogeaeous compression %eems sufficientto account for the build-up to det-nation. The meager evideicefrom experim.ints on single crystals, carried out in such a waythat rarefacticn effecto mav be considered negligible, are ina-cord %ith this mclel (9..

UNC2LASS IFPIED

Iwo C

OW 4~

UUNCLASSIFIEDWOLTR 62-197

The responea cf polycryscalline solid explosivre to theentering shock is not as cloar as in the case of liquld explosives.Solid explosive- tre formed into a mass which contains numercuscrystal entities, And both mecroscopic and microscopic voids.h•zn initiated by plana shocks of low shock &mplitudn, thodepenlence of build-up time on crystal size and void contentmakeu it faixly evilentt that the low temperature rise calcul't-*dfor a homogeneous compression cannot accov.-it for the abservedtransition. Thus a hot-spot mechanism of the typo suggested byBowden and othery (21-24) io required. The rconfitmatfun ofearly work by Winn"- (7) and Marlow (6), who reported thatinduced shocks witl ,-,asurea as low as 20 kilobars would causetransition-to-detonation, has est.iblishad important support of amechanism centexed around a relatively small number of initiationsites. Studies to show how a detonation develops when both thephysical state of the explosive and the shock amflitude arevaried, are beginning to lead to a better understanding of thenature, magnitude aisd behavior of the initiation sites.

The shock-to-detonation transition has veen studied at NOLby the use of a plane-wave system zrranged in such a way as tomake it possible to follow continuously the wave font withinthe shocked sample (3, 4). The experiments to be doscribedhave made it possible simultaneously to establish the initialpressure in the shock and to observe the growth-to-detonationas it develops. Thc observations are made on a wedge-shapedtest sample, the wedge permitting observations without grosslyPffect;n h oedmesos f!-~. in tj!e -e-ge- ^F inte-Arest:

"Th is .s equivalent to the observatiot. of grnwth wj.&~jil anexplooiive charg. of much laiger dimensions. Thus, the resultsappear to agree reasonably well with shock-initiation work onlon- cylinders of cross-sectional area compaTable to the areaof the face af the test wedge, provided: (a) the observationsin the cylinder are nade in the region of its axis (no% on ivsexterior surface), and (b) the presoure-tim. histories of theentering shock are similar. Tho results of these experimentsare in accord with the explanation that growth-to-detonationin polyzrystalline solids is the rerult of pressure build-upfrom temperature-triggered chemical ieas"ion spreading from")ocalized siteb.

EXP3RIM3JTAL

A typical set-up for generating 20-to 180-kilobar, planeshocks in the test cpecim n is illustrated in Figure 1. 1,ithis example the 11-cm diameter plane-wave generator davelopeda detonation wave that was flat to 1 0.3 rxn over a dia-eter of9 cm. K slab of explosive, 12.5 cm x i2.5 cm x 2.5 cm wasplazed between the generator Lnd a 20-cm diameter disc of inertbarrier, or shock attenuator. A sample of the test explosive,

2-UNCLASS IFIPD

UNlCLA~SS IFIEDNOLTR 62-j9l

in the form of a 25* wedge (apex angle 900) w&s then placed (;nthe opposite face of the attenUator. (A th'in film of siliconegrease was generally placed 1,etwet-i the attenuator and thes test

'F- . wedge to minimize the possibility of accidentally causing ahot-spot by a small antount of entrapped air in the regiors.) Ingereral the teat wedgee had facts 3.2 cm by 3.2 cm, and thni,%forewere 1.4 cm high- For the leor sensitive explosives, or where

very low-amplitude shocle were to be used, larger wedyes were

Ibis different pressure levels in the test explo.;ive wedgewere obtained by varying the di.tforent. conwponants of the shock-gctierating system. Thus, the slab of explosive, between the

plafle-wava generator and the inert shock attenuator, was eithercast Composition S, Baratol, ot TNT; in addition, thie thicinessof this slab was increased to as much as 5 cm, as the~ ncad~fE warreinted. The attenuator was either solid brass, aluminuti', orPlexi~jas, o~r was made from 1-cm thicknesses of such materialsin various ;aminated configuratiors, to produced the desiredshlock prossu,ýes.

The phase vslocity of shock arrival along the wedg.o freesurface was determiiaed wit~h a smear camera having a writingspeed of 3.8 mni/microsecond. The ar..ival of the wave was re-corded by t1'.e comer& by using an aluminized Mylar fi~lm on the

surface of the specimen, and reflecting lig'it from an electrically-.exuolo.d wire confined in. a alises capillary. When the wave reached

was brutlyredu.e anshow in Fig~ure 2, permitting pr-ecise

int th wege.11'loctyof shock propagation was then ob-

L to be aw small as possible consistent with the desired height,so that rarefactions from thi region previously shocked wouldnot penetrate into the region behi-nd the yet-unshc'cked portionof the wedge in time to effect the desired obveirvations.

7cr e&a~h experimental arrangement the initial froie-surf acevelocity of the attenuator, without the A.iplosive sample, wasdetermined by direct measurement in an ide..itical lmns-donor-attenuator system. similar prolilainary experiments determinedthat, at t1te center of the plate, over a diamuet~er of 5 cm, orVt more, the tims-of-arrival of the shock was simuitaneous to within30 nanopeconds, while the free-curf ace velocity was constant towithin ± 2%. The particle velocity in the attenuator, at themetal-specimen interfdce, i's then given by the usual assumptionthat it was one-half the' moaaured initial free-surface velocity.'rho shock Hugoniots for Plexiglas and for the Naval bras% uced

3UNZLAsSSIFIED

.f

.7

NOLTH 62-197

° in these ex.-1,r.entp w ee uo..tain-.d by cirect measurement of4, thock and fice-sirf&ce velccity by tho methods described by

Rice et. 1 (24), and Iy Co;*burn (25). The Hugoniat for 24ST"'ljuminum was taken from f.hc report by Rice (24).

"INFORMATION eTITAINSD

The oxpljeivcs sLudied and their pertInont properties arelist.d in TaLle I. 'ilie obseLved time-of-arrival of the sho-kis eturbance &t the waedce free surface was converted to a

vAlocity-distanTe (U-s) curve within the wed"e, by careful slope

measuremea1t of the smear-cav,.;xa ruc.rd, assuming p'ane-wavepropagation Jnside the wedqc. As shown below, the value of theobsenvud shvck velocity at zero wadge thickness was used withtLe Hugoaiot data for the shock attenuator, to determine theinitihl pkessure in the exp~orive. Hugoniots for the urreacced

. explosives were then constructed from these data, assuming that"negligible chsaitJcal reactior. had occtrred at this zero wedgethickneas dxicing pasiage o± the shock. Shock velocities at lowpressure %p- 2KB) were established from measurements made on"the same explosivos, using a simple aquarium method for shocktransit-time observations (26).

The shock Hugoniots for both the inert barriers %nd thenop-reactina epl~ouives ?.ra conveniently expressed by relatingshock velocity (U) to particle velocity (d) in the simple form:

U-a+bu+cu 2

•UJ' where a, b, and c aru constant&.. Whlen Ibib eqkitnion is applied

to the experimental data by the method of least squares, theM. value of the constant, c, is often co small, that in the region

--f interest, the U-u relation can generally be considered linearto acceptable accuracy. The valuis of a, b, and c in thisequetion are listed in Table II for a nunber of ,materials usedin our wurk. Pressure, density, and energy jumps across theshock tront tre derived by the well-known hydrodynamic relations"-ftr a shock (assuming initial pressure i~egligible)s

.- p 0 U U, 2)

P- 0 u/(U-u), 3)0

_v 2E- 0e-p•Vo-v)/2--u /2, 4)

where p iu pressure, cis density, E is specific energy and v isspecific volume (reciprocal of o). Subscript "o" refers to theunshocked state. The particla velocity in the non-r!actincexplosive was determined by boundary-value matching of p and u,as illuntrated in Figure 3, using the calibrated value: of

V 4UNCLASSIFTED

4

UNCLASSIFIEDNOI-TR 62-197

particle velocity and pressure in the barrier (at the barrier,test-explosive interface) and the observed shock velocity in '

the test explosive (at the same interface)- In the Figure,

suibscript "e" refers ý, states within the shocked explosivelsubscript m" refers to states within the shocked m!etal barrier.

In aplyng hismethod, the p-u curve for the reflected ravre-fr'-ction (or shock) wave within the barrier (in this illustration,brass) is approximated by the reflection of the shock p-u curve

for ýhe barrier, about the line: u equal to 1/2 the free surf aceEL velocity for the barrier in the given experiment. Since thepressure and the particle velocity across the interface must becontinuoua, the desired solution is the interanction of thisrarefaction line. with the straight lizie for the explosivepassing t.hrough the origin and having a slope

P/u 0 0 U l5

where U is the measured shock velocity in the test wedge of

pprxiato~~inoled tismethod is a sbtnilimprovemeuitover the linearized impedance equation assumption often made,4 and E-sviously used in this Laboratory (3). The latter methodleads to a larger system~atic erzor in the pressure and particle4 velocity than the present approach.

RESULTS

The experimental observationa may be conveniently shown asgraphs of shock velocity in the explosive sample as a functionOr distazice traveled trorn the m-tal Intert,%ce. Figures 4 annl 5arc tyj4..... of the results found &t NOL. In Figure 4, thoresults for the three cast cyclotols are shown for three jijI ialshoo~k amplitudes using brass attenuators. The first point to benoted is that the initial wave velocity increases as the brassplate thickness decreases. The initial values for the twop Compoaition P types are the 3ame for a given brass thickness.These velocitieu were converted to the pressures shown. in Fig-urE'4 by the proc-idure previously described. The second feature of

the curves is that the dislance to build-up-to-detonation in afunction of the initial pressure. Composition B-3 shows a

o$4 shorter transition distance tian Composition B at each pressurelevel. The difference most proba~bly is due to an RIAX particle-fr ~iize effect. The curves of Figure 4 are typical of the largest

I m~ajority of records obtained in this Laboratory on over a hundredtrials with a number of cast and plastic-bonded explosives at ~

r ~ bulk densities in excess of 97% of theoretical maximum.

TNT, when cast, exhibits a somewhat different shock propa-il gation history. For initial pressures in the explosi-e below

100 kilobare our records consistently show evidence of what

5 '. 1UNCLASS IFIED

.4 W-

AVO

UNCLASS IFIEDNOLTR 62-197

appears to be a levIlling off of velocity in the neighborhoodof 5.3 nmm per microsecond, followed by a second rise, to normaldetonation velocity, Figure 5a. This type of observation hasappeared too frequently for us to attribute it to reading error.J. M. Majowicz (3) first observed iL'ais initial step, but failedto see the later tianaition- to-detonation because his wedgewas only 14 mm in height. The final transition is seen to occur Iat greater distances in the pressure range shown in Figure 5a.When the initial shock in the cast TNT exceeded 130 kilobars, theinitial wave velocity exceeded this intermediate, plateau valueand only one transition was observed, with a considerablyshortened distance to oetonation. For pressed TLIT, the resultsfor low initial shock pressure show new features, Figure 5b. Theobserved initial shock velocity is out of line when compared withthe results from cast TNT, being too high for a non-reactive shock.Furthermore, the transition-to-datonation occurs in a distanceconsiderably shorter than in the case of cast TNT charges shockedby the aame shock generator system. One may note that one of thecurves in Figure 5b involves a pressed charge at a density highi;z.than that oi the cast TNT, yet the growth distance to oetonati.nis still only 4 mm as compared to 15 to 20 mm for the cast chargesshocked in a similar manner. It is therefore quite clear that-harce norosity per se is insufficient to describe the effect ofphysical state on the transition history. lbe effect of pressing,shown here for TNT, in shortening the transition distance is alsopreaent in the cyclotols and in other explosives° we may citeComposition B as an examples with a 1-inch brass barrier (initialP.07 7 kilobars) pressed Composition B reached full detonationvelocity in less than 2 nun, compared to 4 mm for the castexplosive.

The "overshoot" shown in the velocity-distance curve ofpressed TNT, Figure 5b, requires comment. In p-essed explosivesour camera Kecords consistently have shown this irregularity,which we have interpreted as a transient rise to velocities inexcess of the *,ormal detonation rate. In some records theSvelocitits appear to be as mach as 50% over normal, but morefrequently, as shown in Figure 5b, the excess is about 20 - 30%.Such overshoots are entireiy possible, we believe, on hydro- 4

dynamic grounds, depending on the nature of the reaction-rateprofil3 behind the shock front. i•n the other hand the distanceover which enuess velucity has been observed in our records is

mnmll, of the order of 1-3 mm. While reading errors, madeduring measurements of phenomena rapidly changing over such smalldistance&, are aggravated by the mathematical process of differ-entiation. careful examination of the photographs indicate th&tthe records definitely exhibit such super-velocities. We believe,therefore, that the photographic evidence of the overshoots isbeyond reading error, although the magnitudes of the overshootscannot be precisely determined. Campbell, et aL (9), have also

tUNCIAS IFIED

UNCLASSIFIEDNOLTR 62-197

studied presaed charges. They have been quite emphatic that noovershoot had been detected in any of their experiments onpressed solids. We admit that there is room here for honestdifferences, awd these may be du6 to differencau in the two setsof experiments. More refinement of the experiments are neededto settle the question.

The transition distance vs initial shock amplitude withinthe er~ploeive has now teen %etermined for cast Composition B-3,over the iange 30 - 13G kilobars. It has been found that astraight line very nearly fits the results if one plots thereciprocal of the distance against the initial pressure, asshown in Figurn 6. This line extrapolates to an infinite distanceat p -28 kilob&rs. This can be interpreted as an indication ofthe throahold pressure for initiating this explosive with thegiven shock generator. A cursory examination has b-en made ofthe rate of pressure decay behind the shock in the shockgenerator system used to obtain tl.e data of Figure 6. Our bestestinate is that the pressure will fall to about 60% of peak ina time of 2 microseconds after passagf of the shock into theel:plosive, in the absence of chemical reaction. This decay rateis comparable to that estimated to occur in the NOL gap test(12). When Composition B-3 was tested in that gap experimsnt,the S0% point for detonation was found to require an initial peak

lam vlue in thm ^vpl1ive of 20 kilobara. The closenesi of thethrishold shock pressures in the two experil ntm for the sanr.,aexplosive may te used to infar that the long cylinders will, nearthe sensitivity limit, show the same uniform initial velocityas we have found in the wedges at very nearly the same pressurelevel. The velocity would be near ecoustic because at pressuresin the neighborhood of 20 kilobars the shock wave velocity innear to the limiting acoustic valve. Cachia and Whitbread (5)have actually observed this initial "constant" velocity withionication probes embedded in cylindrical charges at somewhathigher pressure levels (U-4 •mn/microsecond). The 50% gapexperiment has ehown a very sharp cut-off oetween •go" and "no-go" in Composition B. It now appears that the range betweenpractically 0% probability of detonation build-up and practically100% probability is about 1 to 2 kilobars in the donor shock.We could interpret this result to mean that in the constantvelocity region the shock pressure is actually increasing byabout this amount in the cases wnere a detonation is established.

DISCUSSION

The clarification of the nature of the growth-to-detonatinfrom a mechanical shock has required that quantitative measure-ments of the initial shock pressure be estabalished. Since thefirst work at NOL we have mapped out (non-reactive) shockHugoniots for a itumber of sclid explosives, two are 2resented

.7i UN"CLASS IFIED

UNCLASSIFIEDNOLTR 62-191

here. Our work has shown that both time and distance for growth-to-detonation is monotonely related to this initial shockamplitude. In 1956 an ad hoc theory was presented in a note byone of us (27) to relate the growth process to shock and react'onvariables. It was pointed out in that note that the history ofthe build-up would probably depend not only on the initial shbckamplitude but also on the nature of th& rarefactions behind theinitial shock. A subsequent paper by Majowicz and Jacobs (3;concluded that the build-up in experiments such as thosepresented here, must have involved a substantial inductionperiod before any" chemical reaction occurred, because tLeobserved shock velocity was initially constant, innofar as wecould detrmine. Our present position has changed in regard tothis delay mechanism. It is now clear to us that our initialshock was followed by a rather steep pressure decay. In thepresence of this rarefaction, the velocit", of the leading &hockin the explosive should have fallen by a messureablv arount ifit had not been supported by energy contributions from reactionswhich must have occurred shortly after passage of the shock. Theabsence of such a velocity decay indicates that the rate of thereaction closely behind the wave must be increasing as the shockprogresses, so that ultimately, the reaction rate, increasingnon-linearly, caus'es the shock to build-up very rapidly toa detonation.

*-rnere now have beer. reported a numLber of related st-udioeby several groups to describe in quantitativt terms, the growth-to-detonation in solids. Through these studies it is apparentthat a unified picture is emerging. In discussing the problemwe find important support in the work of Campbell et al (9),who used plane shocks as we did, but employed donor chargeswhich were considerably larger, thus leading to a much slowerdecay of pressure behind the incident shock. Brown and Whitbread(15), who studied initiatirn by the impact of disks made ofseveral materials, 3h.-wed quite clearly thAt the threshold foxinitiation depends on both the amplitude and duration of theinitial shock within the explosive but not on the propertiesof the disk. Favier and Fauquignon (14) have also shown adependence of the build-up distance on the pressure induced inthe explosive irrespective of the attenuator composition.Similar findings have been reported by Sultaroff and Boyle (10)for shocks through various attenuating media including air, andfor shocks induced by cylinder impact. Jaffe, Beauregard andAmster (12) have established the relation between barrierthicknese and shock peak pressure in a controlled gap-testexperiment and have thereby established thresholds for initiationwhere the duration of the incident shock is somewhat longer thanthat of Brown and Whitbread.

8UNCLASS IFIED

I ,

UNCLASSIFIEDNOLTR 62-197

If w3 confine our attention to cast and plastic-bondedexrlosives, the conclusion reached by Cachia and Whitbroad (5)and by Campb.ll. Davis, Ramsay. Travis (9) for build-up todetonation is, with minor modifications, the interpretationwhich we find acceptable. This may be stated as follows. Theincident shock initiates a small amount of chemical reaction(in localized regions) with essentially no delay. The growthdepends on the pressure effect due to the initial shock plusthc pressure contribution due to the reaction. If the netpresaure behind the wave increases, the leading stock wllgrow to a detonation. If it decreases the detonation will fail.The first pcint of complete reaction will depend on the reaction-time history oxptrienced by the explosive layers after the shockhas passed. If re..ction in regions behind the shock are sloweddown or stopped by adiabatic expansion or heat conduction, thedetonation wave will probably form at or near the shock front,if it is foried at all. The extent o0 initial reaction and itssubsequent growth in a given region will be strongly dependenton the sliock amp>.tude entering that region.

Before exploring the mechanism further, we would like topoint c it a few facts and their i;,nlication concerning shockinitiation of detonation k.ear threshold pressures. It ir nowquite clear tlat detonations can be initiated in solid explcziiesby shocks with peak pressures between 20 and 40 kilobars, in

square charges (6, 7, 10, 12). By comparison, liquid nitro-methane requires 86 kilobars, liquid TNT aboxt 125 kilobars.and Dithekite 13 (HNO 3 /nitrobenzene/i20:63/24/13) about 85kilobar6 (8). In these low denbity liquids, tb"e averagetemperaturt rise at these pressures, in the reighborhood of800 to 100 *C, appears quite adequate to account for transitionto detonation by an initiation process involving homogeneousreact on kinetics. Vie hydrodynamic calculations made byHubbard and Johnson (18), Boyer (19), and Enig (20), clearlyshow a direct correspondere-i between what is observed experi-mentally and what is predicted from computer runs. In particularthe ranid growth to a detonation behind the leading shock, theovershoot in velocity and pressure when this detonation over-takes the leading shock, and the subsequent decay to normaldetonation, appear in both the experiments and in the computa-tion&. it is less clear that the temperature rise associatedwith a homogeneous compression can ba sufficient to initiatereaction in solids at pressures of 80 kilobars, and at 20-30kilobars such a possibility is out of the question. At 80kilobars the Hugoniot energy jump given by equation 4 is 122calories per gram for Composition B. If we assume that all ofthis energy is thermal and the cpecific heat is as low as 0.35cal/g/deg the temperature rise would be 3500C on the average.At 30 kilobars the Hugoniot energy is only 28 cal/gm and the

9,UNCLASS IFIFD

UNCLA63IFZEDNOLTR 62-197

average temperature rise using the above assumptions is 79'C.There is ample opportunity to consider the localizing of energyin microscopic regions (but lar~o relative to molecular dimensions)within the solids under comprerilon. Many workers fallowing Bordenhave noted the existence of small voids in solids, and haveaccepted the simple hypothesis that qas in such voids woul4 getsufficiently hot undeo shock comression to supply the neededinitiation temperature rise. This arqument was tated 1y Cachiaand Whitbread by comparing the 50% gaps for an explosive con~ain-ing in its voids, various gases or a vacuum (5). The same 50%point was found in every case. Is it not possible that the voidact in other ways? We think the answer in yes.

A few of the possible ways for localizing energy are thefollowing:

a. Micro-- cughness of the shbcck anI rhock-waveinteraction (Campbrli).

b. Elastic-plastic changes behind the shock frontwith localised shear or fracture.

c. Discontinuity of flow near voids le-Singto Uhear.

d . Dicairut of rli~ at gri cnfi

e. Spalling or spray into voids. (Jobannson).

f. Phase change under shock loading.

g. Defects in the crystallities.

We do not have evidence to support unambiguously any of theabove as tha mechanisms. Our thinking has strongly le&ned toshock-produced micro-shear or micro-fracture at or near voids,as the path by which the explosive in locally ignited, but wedo not yet know precisely how to characterize these variables.

In the paper by Campbell on initiation of solids (9),evidence is cited to the effect that the explosive near theentering boundary reacts to only a small extent, transmits itsexcess pressure, and then apparently stops reacting. They wtate,on thn basis of these experiments, that the explosive in thatregion not only fails to react to compketion but also will notsustain further reaction when subjected to a second shock (asfrom the region where detonL ion final]y is established). Thisargument is plausible, we believe, for some solid explosives inthe wedge type experiment used by Campbell and by us. Two bitsof infomation will be used to discuss this point. First,

10UNCLASSIFIED

UNCLAdSIFIED

NOLTR 62-197

Bayer (28) ham used a model to compute the transition to detona-tion in solid explosive& in whtich two mechanisms for reactionare assumed to proceed simultaneously. One is an ignitionreaction based on first order homogeneous reaction kineticsithe second is a surface burning reaction in which the Arrheniusterms contained in the equation are the shme as in the ignitionreaction. An arbitrary limit of 1% of the total mass is allowedto react according to the ignition mechanism. The computedresult showed a shock velocity vs distance cur,'e very similar tothese shown in Figure 4. The result also showid reaction tofirst go to completion at points in the explosive whict were nearthe accelerating shoc% front. No baais is given for limitingthe amount of material reacted by the ignition reaction to 1%.It is possible, however, that heat transfer from the reactedsites could, in fact, cause a limitat.on of reaction to thisorder of magnitude provided that thA initial shock were not toostrong. If it is assumed that localized reaction can quenchafter a very short time an explanation must still be found toaccount for failure to re-ignite and propagate a detonationbackward after detonation is established in the forward direction.The following observation on detonation failure in preshockedsolid explosives seem pertinent to this problem.

It has been established by repeated experiments in ourLaboratory, that a sten0y-state detonation in a sheet of EL 506C1

between 0.05 and 0.24 inches in thickness, can be quenched if thedvtonation encounters a regirin in the explosive wnich is beingcompressed to a high density by a second shock wave having a peakpressure between about 10 and 20 kilobars. In one series ofexperiments using the set-up of Figure 7, the explosive in twoparallel layers separated by a plastic gap were initiated simul-taneously at opposite ends. As both detoaations propagated thebow *hock behind the detonations moved toward the alternatelayer of explosive. Each layer of explosive was thus compressedby a shock frcin the alternate layer of ixplosive. Whin & detona-tion reached the preshocked zone it was sean to fail very quicklyin smear camera records. Undetonated explosive could be pickedup from the floor of the test chamber 2 . Other explosives haveshown similar quenchout, e.g. cast RIW/tMT, IM0/Plastic. It .5very likely that the shocked explosive rescved to some small degree

- BL-S6, is a pliable, sheet explosive, manufactured by E. I.duPont Co., containing approximately 70% PETN, and 30% inertmaterial.2. Johannson (31) has described experiments on dynamites in whichair shocks, leading a detonation, eculd cause the detonation tofail.

UNCLASSIFIED

UNCLAS IFIEDNOLTR 62-197

because of the first shock, but then, not only was this reactionquenched in the manner suggested by Campbell, but also theexplo3ive in this shocked, compressed state war unable topropagate a detonation already estsblished in the unihockedregion. These observations lead us to conclude that the hypo-thesis of Campbell can to valid under appropriate conditions.In the case of lonq c-,linders shocked from one end, the retona-tion can be explained by the fact tVat rarefactions behind thegrowing shock (originatinq mainly at the cylinder cides), willreturn the explosive to a condition where it is again receptiveto a build-up to detonation by a shock origirating at the regionwhere detonation is extablished. Some very interesting experi-ments described by Clay. et al (29) in which a shock inComposition B is found to grow to a detonation after passagethrough a preshocked zone also may find explanaticn in theobservation of preshock qrench"•g.

SUMMARY

We have examined the qrowth-to-detonation from mechanicalshock for TNT and 3 cyclotols. Both cast and pressed chargeshave been studied. Experiments have been conducted over apressure rangs from 30 to 120 kilobars. in the cast chargesthe initial wave in the shocked specimen 1-.s the character of anon-reactive shock. The initial wave velocities in these chargeshave been used to compute the peak prejsures behind these initialwhocks. The bmi'd-up to detonativn he* been 'ouid tO be. iei&6it1v6to the RD" purticle Size in two cyclotols of very similar composi-tion aiad density. The build-up to dotination has further beenfound to occur more rapidly in pressed charges than in castcharges of the same composition and density. These observationslead us 'o conclude that in this range of iritial shock pressures,the init~ation occurs at localized centers from which thereaction spzeads. Before wN can be suri that the hot-spotmechanism is the only mechanism for polycrystalline eolids athigher shock pressures more inrormation is needed in the higherrange. In particu'.ar it will be necessary to develop au equationof astate for solid oxploaives in which the temperaturw can beaccurately defined. It remains a possibility ii the higher rangeof pressures, between about 120 kilobars and the detorationpressure, that there may be competing processes going on. Moreor less homogeneout reaction nay Le taking place when the Hugoniotenergy jump exceeds about 100 cal/gm, phzticulaely if theactivation energy can be decreased by compres&ion as has beenrecently suggested by Taller (32). It also remains for futurework to establish the details of ths process of localized initLa-tilon of reaction near the threshold limits o* rhock premcures,that is, in the range of pressure* bolow k00 k.ohbara for moztsolid military explosivas.

12UNCLdASS IF ED

UNCLASS IFIEDNOLTR 62-197

ACIMOWLrDOMENTS

The authors wish to acknowledge the painstaking work ofJ. Schneider, whose careful preparation and execution of theexplosive tests permitted the development of the refined measure-ments reported herein. This work was supported, in part, bythe Atomic Energy Commiusion through the E. 0. Lawrence RýIiati3nLaboratory, Livermore, California.

13UNCLA8SS IF IELD

UNCLASSIFIEDNOLTR 62-197

REFERENCES

1. G. Hertsberg and C. R. Walker, "Optical Investigations oiInitiation and Detonation", Nat. Res. Council, Canada.Project XR-84, (March 1945-August lV46). Also Nature(London' 1, 647-8 (1948).

2. E. M. Dogga, 0. H. Messerly, and H. A. Strecker, OSRDRaport No. 5617 (December 1.4, 1945).

3. J. M. Majowicr and S. J. Jacobs, Classified Naval OrdnanceLaboratory Report (Marah 1958). Also Bull. APS. Ser. 11,1: ;.. 293 (1958).

4. S. J. Jacobs, ARS Jour. 30, 2, 151-8 (1930).5. G. P. Cachia a4n E. 0. Whitbrsad, Proc. Rov. Soc. 4

269-273 (19158).6. W. R. Marlow and I. C. Skidmore, Ibid. 284-8.7. C. H. Winning, Ibid. 288-247.S. A. W. Campbell, W. C. Davis, and J. R. Travis, Phys.Fluids,

A. 498-510 (1961).

9. A. W. Campbell, W. C. Davis, J. B. Ramsey, and J. R. Travis,Ibid, 511-521,

10. M. Sultanoff aid V. M. Boyle, ONR Symoposium Report ACR-52,(September 26-28, 1960) p. 520-532.

11. E. L. Kendrew and E. 0. Whitbread, Ibid, 574-583.12. I. Jaffe, R. Reauregard and A. Amster, ARS Jour., 3l,

22 (i962).13. J. R. Travis, k. W. Campbell, W. R. Davis, and J. H. Ramsay.

Colloquium Reporc on "Lee Ondes de Detonation', 28 August-2 September 1961; Gi1 s/Yvette, Editions du CNkR, Paris,(1961)op. 45

14. J. Favier and C. F. Fauquignon, Ibid. p.5915. S. M. Brown and 3. 0. Whitbread, Ibid, V. b916. M. A. Cook, D. Z. Pack, L. N. Cosnor, and W. A. Gay,

J. Appl. Phys, J.Q, 10, 1579-94 (1959).17. M. A. Cook, D. H. Pack, and W. A. Goy, 7th Symposium onCombustion, V. 820-35, Butterworth S¢iantific Publications,

London (1959).18. H. i. Hubbard and M. R. Johnson, J. Xppl. Phys, J1, 765-9

(1959).19. X. H. Boyer (and othors)"Study of Detonation behavior,

Pub. No. U 187, Aeronutronics Systems, Inc. (Ford Motor Co.)Navy Contract No. NOrd-17945 (May 1958).

20. J. W. Enig, See Referance 10, _p. 534-59.21. F. P. Bowden and A. r. Yotfe."Initiation and Growth of

Explosion in Liquids and Solido' Cambridge Univ. Press (1952).22. K. K. Andreev, See Reference 5, p. 257-67.23. C. R. Johannso3I and others, See Reference 5. p. 160-167.24. H. H. Rice, R. 0. McOueen and J. M. Walsh, p. 1-64 in=Solid

Stilte Physicau Vol. G, F. Seitz and D. Turnbull. Editors.Academic Press, New York (1958).

14UNCLAksIFIED

WI1CLASSZ1ZZDWOLTR 62-19i

RZKEREINCB (Cort ad)

25. N. L. Coleburn, MPUN Report 60;6, (October '.%1. 1460).26. J. it. Kajowicy, Unpublished Dota (Ucval Ordnance

Laboratory) (19!8).27. S. J. Jacobs, Unclassified Note in Conference on Wave

Shaping, Pasadena (Jun. 5-7, 1956), sponsored by PicatinnyArsenal.

28. 14. H. layer (and others)"Stsidy of Detonati.,n Behavior!Pub. No. U 369, keronutronics systems, Inc. (Ford MotorCo.) Navy Contract No. VOrd-17945 (.eb 1959).

29. R. 9. Cl&y, K. A. Cook, R. T. KMyes, and 0. K. Stupe,bull. APB, I er. 11,1, N0.6, P.3"4 (1959).

30. W. D. Garn, J. Chem. Phys, Ia. 3. 819-22 (1959).31. C. H. Johannson, H. L. SelberI, A. Peremon, and T. Sjotlin,

31at Int. Congress of Industr al Chemistry, Liege,September 1958.

37. Z. Teller, J. Chem. Phys., 36, 4. 90-3 (1962).

I

3

-tai: I.;NCLAS ZFZ8D

uo~.Tm 62-197

Of TABLE 1The Explosives Stuuied,

Their State and Dsnsities

. " JE::p-ltaVtw Density , Q/_ -N31 . Composition D

RD,/TdT/Wax i 59/40/1 Cast 1.71

2. Composition 9-3R •/TNT 60/40(Mean REK Particle sitea60-80 microns) Cast 1.72

3. Cyc otolRWN/TNTY 75/25 Cast 1.73

4. Trinitrotoluene (TNT) Cast 1.58(Microcrystalline) 1.62

5. Tririttotoluene Pressed 1.51(Mean Pad-ti.le Sitea 1.6140-80 microns) 1.*64

UNCLASS IFIED

in..

UNCLASSIFP aDNOLTR 62-197

TABlLE 2

Fxperimental Constants for Equation I1(U and u in mwm/miczoa~cond)

;aterial -_ e. ab c

Naval Brass 8.37 3.560 1.833 0

Plexiglas 1.18 2.710 1.568 -0.037

, Lucite1 1.29 2.588 1.51

TNT 1.60-1.62 2.39 2.05 0

TNT (liquid)2 1.472 2.00 1.68 0

Composition P B-3 1.72 2,71 1.86 0

1. Rleference 12

2. Reference 30

Z: 17UNCLASS IFIED

J-14

NOLTR 62 197

4w

t,1 -7

.A; -A

71A

APAR

NOLTR 62-19T

'1

-.- I Ale

II..

Lu

19

rnWL ý d.,m -Jm -. -t~s;j

NOLTR 62-197

300

HUOOXIOT

NAVAkL MAS8

200 , p5

LINEARIZED/I"SEDANCE

RELATION•

IR ACTIONi~oo LOCUS p

COMPOSITION B-3

02

0.0 0.5 1.0

:.• PFIGURE 3 PRESSURLE-PARTICLE VE.LOCITY DIAORAM FORS~DETEM3ININO SHOCK PRESSURE IN TH•E EXPLOSIVE

20

NOLTR 62-197

8 IXITIAL) ,Po- -

4 95 kb* 67

SiCAST CiC.APOSITION B

S (mm)

?5( kb. I

77

).I- I- CAST C75.POSITION B-1

4 41 1K

0 2 •6p7 kb -

- /869I I 6

2 CAST T5/,5 CYCLOTUL

PIOUIE 4 SHOCK VFL•Y-1T" V- DISfANCE FOR ThREE, CYCLOTOLS (FOREACH SET C¢, CURVU, READING FROM LEFT TO RIGKT, THE BRASS

Am22NUATjR THICKNESS WNS 0.5, 1.0, AND 1.5 INCHES)

21

NOLTR 62-197

58

00m 137 kb

6

I 75 kb

I CAST I d

10 20

,-I 5b

= 6 /7,o68

4 ~PRESSE TNT1.51 g/oGN

PKB4ZTYi 1.61

1:-1L I 1 , I I I 1

) 5 10l

P TOUE 5 SHOCK VELOCITY VS DISTANCEFOR SEVERAL TNT CHARO.E

22

NI-T 2-C

IO

I j I I\<

(w Ut S/0

23

NOLrP 62-197

0 PA

,CN

UNCLASSIFIEDNOLTR 62-197

DISTRIBUTIONCopies

Chief, Bureau of Naval Weapon@Department of the NavyWashington 25, D. C.

Dol.-3 ........................ ................. .... 2RRRE-5 . .. . . . . . . . . . . . . . . . . . . . 1RUJK-11. ............... ........................ 1RUML-32 ................ ....... 1.

Director, Special Projects OfficeDapartment of the NavyWashington 25. D. C.

SP-20 .............. ..........................

Chief, Bureau of ShipsDepartmcnt of the NavyWashing]ton 25, D. C .............. ................... 1

Chief, Bureau of lards and DocksDepartment of the NavyWashington 25, D. C ............ ................... .

SuperintendentNaval Post Graduate SchoolMonterey, California ..................

Office of Naval ResearchDepartnent of the NavyWashington 25, D. C.

Power Branch (Code 429) ........ ..................

CommandantU. S. Marine CorpsWashingtop' 25, D. C ...................... ............ I

CommanderOperational Development ForceU. S. Atlanttic FleetU. S. Naval BaseNorfolk 11, Virginia-;& ........... ..................... 2

C ,mmanderU. S. Naval Ordnance Test StationChina Lake, California

Code 556 ...................... .................... 1Code 4572...... ............... 1

Aecbnif l Library .. . . . . . ...............

UNCLASSIFIED

HOLTR 62-197

copies

DirectorNaval Research LaboratoryWashington 25, D. C.

Technical Informnti-on Section . ....... .. . 2

DirectorDavid Taylor Model BasinWashington 7, D. C . .....

Dr. h. a. Keil. .......... .................. . . . 2

CommanderNaval Air Development Cent*LJohnaville, Pennsylvania

Aviation Armament Laboratory ..... ............. .. 1

CoxaanuerU. S. Naval Weapons LaboratoryDahlgren, Virginia

Technical Library ................ ................. 2Weapons Lab. . . ......... . . . . . . . ITerminal ballistics Lab ........ ................. ICode WLT ............... ........................ IL. Pruett . . .................... 1P. Altman . . . .................. 1J. Grey .............. .................... .. .

CommanderNorfolk Naval ShipyardPortsmuth, Virginia

Underwater zxr~oeion* Res. Div . .. .. .. .. . .. 1

Commanding OfficerU. S. Navil Weapons StationYorktown, Vii'ginia

R&D Division ................. .................... 2

conmandi ng OfficerU. S. Naval Ordnance LaboratoryCorona, California. .................. 2

CoaUtnding OfficprU. S. Naval Propellant PlantIZr-Ian Head, Maryland

Technical Library ............ ................. 1EODTC .................. ....................... 1

CommanderNaval Radiological Defense Lab.Sa F3gajcJlco, C3i•fornia

oSh l r. . . . . .......Me...1

NOLTR 62-197

Copies

Comi.anlirg OfficerU. S. Naval O.:dnance PlantIooon, Georgia .................... . 1

Commanding OfficezNaval Ammnition DepotCrane, Indiana . .................. .1

Comeanding OfficerU. 3. Naval Armunition DepotUNvy Number Six Six (66)C/o Fleet Post Officesan Prancisco, California

Quality Evaluation Laboratory ............ 1

Commanding OfficerU. S blaval Underwater Ordnance StationNeuport, Rbode island ................. ................. 1

Commanding OfficerU. S. Naval Weapons Evaluition FacilityKirtland Air Force BaseAlbuquerque, New Mexico ......... ................. .. 1

Office of Chief of OrdneoceDepartment of the AxayWashington 25, D. C.

ORDGU ..................... ....................... 1ORMD.. .............. ......................... . 1ORDTN.. . ............... .........................

Office of Chief of EngineereDepartment of the ArmyWashington 25, D. C.

JMGNB ................... ......................... 1ENGEB ..................... ....................... 1

Comw4mazietArmy Rocket & Missile AgencyRedtione Arsenal, Huntsville, Alabama

ORD2CR-R.. ................ ....................... 1

Conmanding OffficerDiamond Ordnance Fuze LaboratoryConnecticu, Avenue & Van Ness St., N. W.Washington 25, D. C.

Ord Developmer.t Lab ........... ................. .Code 005 ........................ .......... 1

NOLTR 62-197

Copies

Comanding OfficerPicatinny UsmenalDover, New Jersey

ORDBB-TH8 . .................. 1

ORDBD-TX1.. .......... . . .... 1ORDD?-T... . . . . . . . . ISORDIB-TP1 . 1ORDBB-TP2. .. 1 .ORD3X-TP3 ... .. 2ORDEB-TR2 . . . . . . 1ORDB--T1 . .. . . 1ORDIB-DR4(P. 8. Tweed) . . . ..

Comnanding OfficerOffice of Ordnance ResearchBox CMDuke StationD•rham, North Carolina ................ .

Commanding OfficerWatertown ArsenalWatertown 72, Massachusetts .......... . . 1

Office of Chief Signal OfficerResearch & Ieveiopw.nL DlviairaWashington 25, D. C .............. ................. 1

Conmanding OfficerEngineer Research & Development Lab.Ft. belvoir , Virginia

Tech. Intelligence B!anch. ....... ............... 1

Comnanding GeneralU. S. Army Proving GroundAberdeen, Maryland

Tech. Librur-y .................. .................. 1Dr. R. Hichelberger ........ .................. .IM4. Sultanoff .................... 1Dr. F. Allison ........... ..................... 1

Commanding OfficerChemical Corps'Themical & Radiological Lab.Army Chemical Center, Maryland ....... ............. 1

Commanding Officerfort Dietrick, Maryland ................

| _ _ _

IJ

NOI.TR 62-197

Copies :

Cojimanding GeneralU. S. Army Ordnance Aummunition CenterJoilet, Illinois ...................... 1

Commanuing OfficerHolston Ordnance WorksKinguport, Tennessee .................... ................ 1

Commanding GeneralFrankford ArsenalPhiladelphia 37, Pennsylvania

S. Picoli ................. ........................ 1

CommanderOrdnance CorpsLake City ArsenalIndependence, Missouri

H. Kline, Industrial Engr. Div. . . . ......... . 1

Commanding GeneralWhite Sands Proving GroundWhite Sands, New Mexico ............ ................. I

Chief of StaffU. S. Air ForceWashington 25, D. C.APORD-AR . . . .n. . .l ........................ .

APGC (PGTRI, Tech Lib)Eglin Air Force Base, Florida .. .. .. ... . .. .

CommanderAir Force Special Weapons CenterKirtland Air Force BaseAlbuquerque. New Mexico ........... .................. 1

CommanderAir Force Cambridge Research CenterL. G. Hanscom FieldBedford, Massachusetts . . ............ .. 1

Commanderholloman Air Development CenterAlamagordo, New Mexico. ............ ............... ]

Office of Technical ServicesDepartment of CommerzeWashington 25, D. C. ................... 100

R

NOLTR 62-197

Copies

DirectorApplied Physics LaboratoryJohns Hopkin4 University0621 Georgia AvenueSilve; Spring, Maryland .................. 2

Solid PropellarAts Agency ............... I

New Mexico Institute of Minir,g and TechnologySccorro, New Mexico

k. L. Kempton .................... 1

Atomic Mncrgy CommissionWashington 25, D. C.

Director, U. S. Bure.u of MinesDivision of Explosive Technology4800 Forbes stroetPittsburgh 13, Pennsylvania

Dr. R. W. Van Dolah ............. ................... 1

ChiefDefense Atnmi.c Support AgencyWashington 25, D. C ................... .................... 5

Armed Services Technical Information AgencyArlington Hall etationArlington 12, Virginia ................ .................. 10

National Aeronautics and Space Administration1520 H Street, N. W.Washington 25, D. C. . . ................. I

National Aeronautics & Space AdministrationGoddard Space Flight CenterGreenbelt, Maryland ................ .................. 1

Lewis Research Center, N.A.S.A.21000 Brookpark RoadCleveland 35, Ohio

Library ....................... ....................... 1

National Aer..,dutics & Space AdministrationAmee Research LaboratoryMoffett Field, Clifornia ........... ................. .. 1

7,1 NOJ.TR 62-191

Copie&

Lawrence Radiation LaboratoryUniversity of CaliforniaP. 0. Box 808

,- • Livermore, CaliforniaTech. Information Div....... .... .. . .. .. .. .........Dr. J. Kurry ........... .... ................... 1Dr. M. Wilkins. .............. ................... 1Dr. G. Dorough ........... ................... .

DirectorLos Alamos Scientific LaboratoryP. 0. Box 1663Los Alamom, New vexico

Library .................. ....................... 1Dr. Campbell ............... ..................... 1Dr. L. C. Smith .......... .....................

Sandia Corp.P. 0. Box 969Livermore, California ... . .. ............. ... 1

Sandia Coire.

P. 0. Box 5400Albuquerque, New Mexico .......... ................. 1

Atlantic Research Corporation19701 West Goodvale RoadSaugus, California ......... .................. .

Denver Research InstituteUniversity of DenverDenver 10, Colorado ........ .....................

Tho Franklin Institute20th & Benj&min Franklin Parkway"Philadelphia 3, Pennsylvania

Ow Mr. Gunther Cohn ............. ................... 1

Aerojet-General Corp.Ordnaaue DivisionDowney, California

Dr. L. Zernow ............................ 1

Allecjhany Ballistics LaboratoryCumberland, Maryland ............. .................. 1

AW-

NOLTR b2-1 0 7

L_ ,.•_CopiestSArmour Research Pound~ti-on

.Technology Center, Illinar Institute of Techn-ology

•-•=2' :10 West 35th street

Chicago 16, Illinois . 1

E. 1. duPont deNemoursEastern Laboratox i&.

SExplosivus Depart m ntGibbstown, New Jursey

,3r. L. Courson ................ ............ ....... 1Dr. W.inning ............. ..................... .

Lockheed Aircraft Corp.P. 0. Box 504Sunnyvale, California ........... ................ .

Stanford Research Instit.tcsPoulter Laboratories AMe.alo Paak, California ........ .................. .

University of UtahSalt Lake City, Utah

Dr. 1. Cook. Explosive Research Group .... ......... 1

Uarvard UniversityCam~bridge, Mastachusetts

Prof. G. B. Kisti&JvrK... . ..... ..................

Carnegie Institute of TechnologySche:tley ParkPittsburgh 13, Pennsylvania

E. M. Pugh ................ ...................... 1

Amherst CollegeDepartment of PhysicsAmherst, MassachusettsDr. A. B. Aroi................. . ... . .. .. .. .. .......

r Yale UniversityDepartment of Chemistry

I., New Haven. ConnecticutDr. John P. Chesnick ............ ................. 1

University of MichiganDepartment of PhysicsAnn Arbor, Michigan

Prof. Otto LaPorte ............ .................. 1

S• Jl(l•iJ•iiLIi iLd5 ,, i lHiL, in, ll ,,I, I,,, .el, I•4° NObTR 62-197

SD•partment of Physics

::• Penn State University

University /'ark, PennsylvaniaDr. R. G. •toner .. ................ I

Prlnceton UniversityPrinceton, Hew Jersey

Dx. Walker B1eakney .................. l

Atlas Powder CompanyTamaqua, Pennsylvania . ................ 1

Aeronutronlc, Division of Ford Motor Co.Ford RoadNewport Beach, Callfornla ................ l

SInstitute of Coopeiative Research

Johns •c•kins University1315 st. Paul StreetBali t imore, Maryland ................... ].

Institute for Air weapon ResearchUniversity of ChicagoChicago, Illnell .................... I

• i CalIEornia Institute 0£ Technology d

- ; •asaoena •° •aA•EUEEILaSDr. H. Davidson .................... i

i• Arthur D• Little, Inc.

Cambridge 42, Massachusetts : °Dr. G. R. handrlck .... . ..... 1 "-:"., '•Repot '*bary •v. ,•A• • : .... : : . " . . J. _ ....

L, Hercules Powde• company o•'.•:•-

Experiment Station •" (',Wilmington, Delaware •m•" .•,, ."

•L-'.: j J Roth . . . . . . . . . . . . . . . . . . . . . . . I ". .......,. Dr. Lawrence . . . . . . . . . . . . . . . . . . . . . . I , '-•"•• •" ",%"i• General .•lectric Company. •i•.• , ,i Missile & Ordnance Systems Department ,, • ".m .

2198 Chestnut Street • •... Ti :,•• hilad•]•hia, Pennsylvania ,::':. • . "

" T.W. Kennedy . . . . . . . . . . • !'-,,j. T•F

>,.• • ,,,•..

•. " .,,.,. ,,< -• .. • -'

• . •. ',, • :

z 10

Ir CI

-~~~~ -- -- 4

Lýf~ 9 .. I L, Cd oel .a 4 0 "t .- W v 0 -"C 90 3 w W 40 ~4,

k r. 46 o,

ot o

K 9 0 c .0

w -P tJ -4 o Ne 0. "J r943 4

"4- 0,, *' c6 .e ol N "14-Z -01 Vl

4' 4-b -

19 4IP FS-4 7 0 ý4i t

424,4 r- I";0 ()I4 4 3 r

*~cO u - -4'4 4) 0 .I.-A.6 4

s. r r o v> 4 2 C.4SI k14 4' cý ~;, . r

0 v

v C4 04' -

*4 r-4 N 0

C, I

r4 0,.4. r4.1-4

C' 4 )t' n E 1-I4' V,

9,,

9,-49 r- . U

- -~~~4 IT ), ~ J 4 9

'-0C-1 4)0.- 4 " -..oT!to- c'4 M. aM V 0 d %a 0'4 ' It .A I6. VC *1 "Vr'T -3

46 0'C 4-.~I . , C*44. y . v* a W)

Ns-e Js v 01-~~~i -S a~ .

1-4 91 9V r9 4'.4 W Cf" 04 IL e

24I3,0