DTICS ELECT' i"NSWCDD/TR-92/280 U .3T"AUG 4AD-A267 391 S C

IMPACT TESTING OF EXPLOSIVESAND PROPELLANTS

BY CHARLES S. COFFEY

RESEARCH ANC TECIINOLOGY DEPARTMENIT

and

V. F. DEVOSTADVANCED TECHNOLOGY AND RESEARCH, INC.

0JUNE 1992

Approved fur public release, distribuition is unlimited.

NAVAL SURFACE WARiFARE CENTER

Dahigren, Virgisia 22449-5000 v 5ilverprini. Moryalarid 20903-5000

>1i

NSWCDDITR-92/280

IMPACT TESTING OF EXPLOSIVESAND PROPELLANTS

BY CHARLES S. COFFEYRESEARCH AND TECHNOLOGY DEPARTMENT

and

V. F. DEVOSTADVANLED TECHNOLOGY AND RESEARCH, INC.

JUNE 1992

Approved for public release, distribution is unhlifnted

SACreI 101., 1 or - - -

NliC fAC i.A

NAVAL SURFACE WARFARE CENTER U ,°...'"J d "

Dahlgren, Virginia 22448-5000 0 Silver Spring, Maryland 20903-5000 .. - . -

By

. . i D Ibuti, o' li

A ,, C

NSWCDDiTR-92/280

4

FOREWORD

This report represents part of the results of an effort at the Naval SurfaceWarfare Center Dahlgren Division to understand the physics of impact inducedignition in energetic solid materials and from this to develop meaningful small scaleimpact tests. In this effort we have been fortunate to have received the support andencouragement from a number of individuals and organizations includingMs. S. C. DeMay and Mr. H. Richter of the Naval Weapons Center; Dr. D. Liebenbergof the Office of Naval Research; Drs. C. Dickinson, L. Roslund, R. Doherty,S. J. Jacobs,* H. Haiss, D. L. Woody; Mr. J. Davis; and Ms. B. A. Yergey of NSWCDD,and Professor R.W. Armstrong of the University of Maryland.

Approved by.

KURT F. MUELLER, HeadExplosives and Warheads Division

Presently at Advanced Technology & Research, Inc., Laurel, MD 20707.

i/ii

NSWCDD/TR-92/280

ABSTRACT

The drop Nei ght impact test is the simplest and easiest test that can beperformed on small quant:ties of explosives or propellants, and yet it has only aminimal role in assessing explosive sensitivity or performance. This report examinesthe drop weight impact test as it is currertly used, describes its major flaw, andsuggests an alternative test that holds the promise of providing the impact energyrequired to ignite an energetic material. Other impact tests are described. One ofthese, the Blý.listic Inpact Chamber Test, measures the rate of reaction and extent ofreaction during impact. This test demonstrates that during impact there are twoforms of initiation reactions that occur: one that is very fast and is likely due to directimpact-shear initiatiGn of the crystalline solids in the sample and, the other, a muchslower component, is thought to arise ,.ue to burning of tne sample.

V @

iii/i v

NSWCDD/'TR-92/280

CONTENTS

INT RODU CTION ................................... 1................

ANALYSIS OF THE DROP WEIGHT IMPACT TEST ................... 2

THE ENERGY TO IGNITION TEST .............................. 3SA M PLE SIZE .................................................. 6TIHE IMPACT SHEAR TEST .................................... 7TIlE BIC TEST ................................................. 9SOME EXPERIMENTAL RESULTS ............................. 10

SUMMARY AND CONCLUSIONS .................................... 12

v/vi

N SWCDlYrR-92/280

ILLUSTRA'TIONS

Figqri Page

1 SCHEMATIC OF TIHE NSWCDD IMPACT MACHINE ........ 14

2 SCHtEMATIC OF IMPACTOR-AN VIL FOR THE ENERGY TOIGNITION TEST ........................................ 14

3 TYPICAL ENERGY TO IGNITION TEST RECORDS OFTHE IMPACTOR DECELERATION AND TILE ONSETOF REACTION ......... ................................ 15

4 SCHEMATIC OF THE IMPACT SHEAR TEST ............... 15

5 SCHEMATIC OF THE SHEAR PROFILE DURING TILEIMPACT SHEAR TEST, T. ................................ 16

6 SIMULTANEOUS ACCELEROMETER AND INFRAREDOSCILLOSCOPE RECORDS FROM THE IMPACT SHEAR 0T T .. .......... .... ,........................................ .LU

7 BALLISTIC IMPACT CHAMBER (BIC) TEST ................ 17

8 TYPICAL PRESSURE-TIME RECORDS FROM THE BICT E ST ...................... ........................... 17

9 NORMALIZED FLUCTUATIONS IN THE INITIAL ENERGYRELEASE RATE, ((dE/dt)/< dE/dt >), AS A FUNCTIONOF THE VOLUME FRACTION IN INERT MATERIAL,V1idVo, FOR A NUMBER OF EXPLOSIVES ANDPROPELLANTS ........................................ 18

10 RATE OF PRESSURE INCREASE IN THR.EENON DETONABLE PROPELLANTS, LABELED 1THROUGH 3, AND ONE EXPLOSIVE, LABELED 4, ASA FUNCTION OF IMPACTOR ENERGY IN THEBIC TEST .............................. ............... 19

11 ENERGY RELEASED MEASURED IN THE BIC TESTAS A FUNCTION OF IMPACTOR ENERGY FORSIX DIFFERENT PROPELLANTS ....................... 19

vii

NSWCI)DITR-92/280

TABLE'S

Table Page

ENERGY TO IGNITION VALUES OF A PROPELLANT-LIKEM ATERIA L ............................................... 5

2 RESPONSE OF EXPLOSIVES TO THE BIC TEST ............ 12

viii

Mill

NSWCDD/TR-92/280

INTRODUCTION

Perhaps the simplest and easiest test that can be performed on small samples ofexplosive or propellant materials is the drop weight impact test. Yet for all of itssimplicity and ease of operation, this test plays a disappointingly small role inassessing the safety or predicting the performance of energetic materials.

The drop weight impact test goes back to at least. the early nineteen hundredswhen it was first used to assess the relative impact sensitivity of explosives.*,l-4A direct descendent of this test is still uses today in the Uni .ld Kingdom to obtain ameasure of the impact "sensitiveness" of ei.ergetic solids. Snimilar tests have beendeveloped in several other countries and mony of khese have become the respectivenational standard required for assessing the handling safety of energetiL solidmaterials. 5-8 Since the 1940's much of the investigation in impact ignition ofexplosives bas been conducted at the Cavendish Laboratory by Bowden, Field, andcoworkers.**,9 ,10

Generally, the various drop weight impact tests all seek to determine the dropheight at which the explosive or propellant samples react during some fraction of thenumber of imnacts. Unfortunately, this concept of impact sensitivity is seriously S,lawed. Thle report will show thatt clic Go NI GT drop h.eight eritteron are an

inappropriate measure of impact sensitivity. This should be no surprise, for otherthan as a means of obtaining a crude relative ranking of the impact sensitivity ofvarious energetic solids, it has been impossible to find any othc, meanin,-ful use forthis data. Een the relative ranking by the 50 percent Go-No Go drop heigit has tobe examined with care. Often the fluctuations in the 50 percent height a),e :uff .cientto change the relative ranking of materials whose impac,. ,.ensitivities ar, .imilar.More important, from a safety standpoint, the relative violence of the response ofdifferent materials to impact may differ markedly although their 50 percent Go-NoGo ignition drop heights are similar.

In this report, several new impact tests and concepts are pre'strd that seek toavoid the failure of Lhe standard impact test and attempt to obtain izformationrelevant to the behavior of impact induced ignitien in energetic solid& under anyarbitrary impact loading. The experimental and theoretical backgrou..nd thatunderlies mout o the work presented here shows that shear and shear rate, duringimpact or shock, are responsible for establishing the energy localization or hot spo?,sites from which ignition starts.1 1-13

* Reportedly, G. Rotter first used a drop weight impact apparatus to test explosives in about 1905.• The many papers from the CavendisH Labor'.tory on the impact initiation of

explosives are t)o numerous to list here; however, see References 9 and 10.

NSWCI)1)ITl-92/280

ANALYSIS OFTI1E DROP WEIGIII IMPACT'TESTI

As with any &.est, it is wise to obtain an analysis of the expected physicalbehavior in order to better understand the test results. The physical behavior of tiledrop weight impact test can be modeled on a computer with all of the ancillarymotions that such efforts seem to involve. The test also lends itself to an analyticalsolution which, while approximate, allows a more physically transparent representa-tion of its behavior during impact. 14 Ultimately, both techniques will eventually failbecause of our present inability to accurately describe the behavior of the sam plematerial during impact. However, for most impact tests which use only a smallsan tple size, the presence of the sample introduces only a minor perturbation into theoverall behavior of the impact machine. Indeed, as it is currently used, the principalfailure of the drop weight impact test will be shown to be independent of sample size.

A typical impact machine consists of an impactor of mass, M 1 , and an anvil ofmass, M 3 . Often between the impactor and the anvil is a striker on mass, M 2 , thattransfers the impact force from the drop weight to the sample and anvil. An analysistreating the drop weight, striker, and anvil as a collection of mass-spring systems isgiven in Appendix A. A simple and versatile impact machine that has been usedfrequently at the Naval Surface Warfare Center Dahlgren Division (NSWCDD) isshown in Figure 1. Impact velocities in excess of 40 m/b can be achieved with thismachine by accelerating a low mass impactor wi-h elastic shock cords.

The important behavioral characteristics of the impact machine are determinedby the natural frequency of each of its elements, w = (KIM)I, where K is the effectivespring constant. of the element and M is its mass. For cylindrical elements, K =

(EA/L), where E is the elastic modulus, A is the cross sectional area of the element,and I is the length dsually, for most standard impact machines M3 > M 1 M2 so *that a ,, < (-)2. In Lhis limit, the force on the anvil fnd nsamnle given byequation (A-15) is approximately

F2() KK 2 (2gh)" Msintt t

where w I is the natural frequency of the drop weight, h is the drop height, and KIand K12 are the spring coostants of the drop weight and striker, respectively.

In the above limiting Lis. ovhich is typical of most well designed impactmachines, the duration of the inimact is essentially determined by the drop weightand is independent of the amphtiid& of the impact. Once the force starts to gonegative, rebound occurs. The mnote t.%act expression given by equation (A-15) inAppendix A shows that the half sirie wave loading determined by the drop weight hassuperimposed on it an ociPlating contpe-,.Pnt due to the natural vibration of thestriker, In a clean, well d&ýigned impact machine this is usually minimal. The massof the anvil is usually so large that it has little or no role in the impact response. Inthis regard, most of the elaborate an Qil base5 constructed for the current standardimpact machines are unnecessary. A compui.er analysis of the impact machine wouldshow all of these features plus the various wa, es reflecting back and forth throughthe system until equilibrium is reached. However, this added detail is unwarrantedmainly for the reason given below.

2

NSWCI)I)flIR 92/280I

The major flaw in any attempt to measure explosive sensitivity by using the50 percent Go-No Go drop height arises because, under these conditions, ignitionalways occars at or near the maximum force levels. This is most easily seen bysinuItaneIously monitoring the time of ignition with a photo-sensitive detector andthe applied stress load with an accelerometer o" strain gage. At the maximum forcelevel, the elastic energy stored in the machine is also a maximum and it is this energythat gives the rebound. The energy put into the machine by releasing the dropweight from a height, h, isjust M]gh which is partitioned between the elastic energystored in the machine, the plastic energy required to deform and heat the sample toignition, plus the small but inevitable amount of energy lost in the machine duringthe impact. Unfortunately, there is no easy way to separate or distinguish theseelastic and plastic energies from one another. As far as ignition is concerned only theplastic energy is important. Furthermore, it will be shown shortly that at or near the50 percent ignition level, the amount of elastic energy stored in the machine oftenexceeded the amount of plastic energy required to deform and ignite the sample.

This inability to separate the elastic energy stored in the machine from theplastic energy dissipated in the sample essentially renders the drop weight impactmachine useless for anything other than giving a crude ranking of sensitivity basedon the Go -No Go drop height. Even this information is dominated by the storedelastic energy and, therefore, is machine dependent. Thus, it is unlikely that eventhe ranking of explosive sensitivities, based on their 50 percent Go-No 6o dropheight, will be accurately reproduced among different impact machines.

TI1I ENERGY TO IGNITION TEST 1,1 1

Confronted with the failure of the standard 50 percent Go-No Go drop weight )i11T:n,-t test, we have chosen to take advantage of the physics of the impact situationand seek an impact ignition test that minimizes the amount of elastic energy storedin the machine at the moment of ignition. Modern, plastic bonded explosives andpropellants are much softer than the Rockwell 64 hardness of most impact machinetools. The sequence of events during the impact involves first a deformation of thecomparatively soft sample material, followed later by a buildup of stress in the anvil,striker, and drop weight system. This stress buildup usually occurs after the samplehas deformed and spread considerably. The static stress levels within the sample canhe determined by the analysis developed by Schroder and Webster. 17

In the energy to ignition test, we take advantage of this sequence of events bycausing ignition to occur early during the impact when the stress levels are low andUI~ lyL.IV U Like d[ILJ 1• UL9 l •UII• ICUIM iCSui-n. L•of .C c I I .• rgytransferred from the impactor goes into deforming and heating the sample, and verylittle is stored in the machine as elastic energy. This requires a relatively highvelocity impact of perhaps 10 m/s or greater, depending on the material. To simplifythe impact, we have eliminated the striker and let the impactor impact directly onthe sample and anvil. Tihe sample is usually a pellet 5 irn in diameter and 1 mnmhigh with a mass of approximately 40 mg.

The light from an ignition is monitored by three photosensitive diodes arranged120 degrees apart around the outside of the anvil. The change in velocity of theimpactor is determined by an accelerometer or some similar means. This is shownschematically in Figure 2. These measurements are recorded on a dual channeloscilloscope, a typical record of which is shown in Figure 3. The change in velocity of

3

NSWCDI)rl'-92/280

the impactor at the moment of ignition is dctermined by integrating theacceleronmeter record from first ;ipact to the moment of ignition as determined by 'Ithe photo detectors. •

lfVo is the velocity of the impactor at the moment of first impact and Av is thechange in its velocity at the moment of ignition, the change in the kinetic energy ofthe impactor at the moment of ignition is

A=MV - 4M(Vo - Av) . (2)

Our experience with the more impact sensitive explosives and propellants indicatesthat Av < Vo so that

AE = MVO (AV). (3)

If the velocity of the impactor is large, Vo > 10 m/s, then for at least the moreimpact sensitive explosives and propellants, ignition will occur early during theimpact when most of the energy transferred from the impactor is going intodeforming and heating the sample, and very little is stored as elastic energy in theanvil-drop weight system. At this time, the stress on the sample-anvil is usually verysmall, on the order of a few Mpa. On this basis, we make the assumption thatessentially all of the kinetic energy lost by the impactor is transferred to the sample.The energy required to ignite the sample is now approximately AE. It is convenientand informative to normalize the energy to ignition by dividing by the sample mass,m, to form the critical impact energy density required to cause ignition, AE/m. This Ihas to be one of the essential inputs into any calculation that purports to predict theimpact response of an energetic material.

Table 1 gives a listing of the energy required to cause ignition of an explosive,composed predominately of ammonium perchlorate and aluminum, over a wide rangeof impactor velocities and masses. Interestingly, for this material at least there is nota large range of variation of the energy to ignition value over this rather large rangein impact velocity and energy. Some of this variation is due to problems associatedwith proper gage operation at high impact velocities. At high impactor velocities, Vo.> 50 mts, ignition occurs within 1 or 2 ps of apparent impact, which is uncomfortablynear the limiting response times of the best currently available gages.

Finally, as noted earlier, the energy required for ignition is often less than theenergy available from the 50 percent ignition drop height. For the above material,the drop height for 50 percent ignition using the standard Explosive ResearchLaboratory (ERL) impact test machine at NSWCDD is approximately 15 cm. TheNSWCDD ERL machine uses a 2.5 kg drop weight, so that the energy availableduring impact is 3.7 J. If it is assumed that at least to first order the energy toignition for this material is independent of impactor mass or velocity, as indicated byTable 1, then only approximately 2 J of energy are needed to cause ignition. Theremaining energy is presumably stored as elastic energy during the impact or isdissipated by some other means.

The major problem that we have had with this test is that of insuring that thegage survives the impact at velocities in excess of about 20 na/s. This is the case,

4

N SWCI)I)/l''P-92/280

TABLE 1. ENERGY ro IGNITION VALUES OF A PROPELIANT-LIKEMATERIAL

(Sample size, 5 num dia. x 1 minm high; approximate mass, 35 to 40 ing.)

V() (1is) -V (m/s) E ]J)

1.4 .073 1.031.4 .063 .881.4 .065 .9!1.4 .118 1.652.8 .03 .842.8 .053 1.472.8 .064 1.84.43 .035 1.754.43 .041 1.85.37 .023 1.233.13 .098 2.973.13 10kg .087 2.663.13 Impactor .079 2.413.13 .047 1.483.13 .076 2.313.13 .065 1.973.13 .027 .843.13 .535 15.311 Newly polished anvil3.13 .157 4,81J and impactor surfaces.3(AQO 1.23.13 .041 1.273.13 .082 2.563.13 .085 2.623.13 .085 2.623.13 t .072 2.24

90.7 2.37 3,42167.4 16kg 1.63 4.37168.4 Impactor A48 1.3169.6 2.65 7.269.1 2.87 3.17

<2,14>*

*Average ignores the two polished anvil values.

particularly with the insensitive explosives and propellants, which require highimpactor velocities to achieve ignition before significant stress buildup o~.curs. Oftenwith these insensitive materials the impact velocity must be held below 20 m/s toprevent gage failure. This involves some compromise with the test requirement thatvery little energy be stored elastically in the machine before ignition occurs, but itallows an upper limit to be put on the impact energy required to ignite the sample.Clearly, the assumption that very little of the kinetic energy lost by the drop weightup to the moment of ignition is stored elastically in the impart machine, needs to becontinually monitored by examining the stress on the components of the machineat the moment of ignition.

5

NSWCD.D/TR-92/280

SAMFLE SIZE

For the energy to ignition test the sample size is important. Partly for historicreasons and partly for convenience we have chosen to work with 35 to 45 mg sizesamples. There appears to be nothing unique about this sample size. However,samples below 20 ing are difficult to ignite on impact. This will certainly be the case Awith the more insensitive explosives. On the other hand, samples of energeticsecondary explosives such as HMX, RDX, and PETN may transit to a detonation ifthe sample size exceeds 150 mg. This was deemed unacceptable because a detonationwould destroy the tools and reduce a simple lab scale test to one requiring a highdegree of protection for the operators.

A more fundamental and more germane issue is that of the appropriate samplesize required to relate the results of this test or, for that matter, any small scale testto the initiation of large scale charges. Bowden has shown that, at least for someprimary explosive materials, the minimum hot spot size is of the order of a few tens ofmicrons.1 8 Our earlier heat-sensitive film experiments with plastic bondedexplosives and propellants have shown that, at threshold, ignition begins in theregions of high shear and often at the site of the largest explosive crystal in thisregion. 19,20 In these experiments, in which the impact velocity was much less thanthat required to achieve a 50 percent ignition probability, pressure appeared to haveno direct role other than as the forcing function that created the shear. These weresimple symmetric impact experiments and threshold ignition always occurred nearthe outer edge of the sample during impact, where the pressure was a minimum andshear was a maximum. It never occurred in the regions near the center of the samplewhere pressure achieved its maximum value and shear was a minimum. From this,we infer that shear i4 the mechanism responsible for energy localization andsubsequent ignition and that, for a practical test, the sample must be of sufficient sizeto allow shear to develop in a measurable way. 0

As the amplitude of the impact was increased by increasing the impact velocity,the number of ignition sites also increased and the ignitions occurred earlier duringthe impact induced deformation. 2 0 At impact velocities of 50 to 100 mi/s on impact,sensitive materials' ignition occurred within 2 or 3 ps of impact. At these higherlevels of impact, ignition often appeared nearer the center of the sample, but thesewere likely to be due to the correspondingly higher initial shear levels causing energylocalization and ignition in some opportunistically aligned crystal earlier in theimpact. However, in several thousand mild impacts, where the impact velocity neverexceeded 5.5 ps and heat sensitive film was used to locate ignition, never was ignitionobserved to have occurred at the center of a symmetric impact experiment whereshear is a minilmlullll nldl ll v psul-uf-e ib a Jn1axuiiUiY.

The evidence cited above demonstrates that ignition occurs at local sit.:s withinthe 35 to 45 mg sample. Since these local ignition sites occupy only a small portion ofthe sample, the implication is that the energy required to deform and heat the wholesample, as measured by the energy to ignition test, is an over estimate of the energyrequired to cause ignition which can occur atjust one hot spot site. However, forpractical reasons it is not possible to deal with this process on a single ignition sitebasis. Rather, it is necessary to treat this problem on a statistical basis in which theenergy to ignition is established as the average energy required to cause ignition in asmall sample, but one still large enough to contain a statistically significantrepresenta-tion of the material of interest and its potential hot spots-ignition sites.

6

NSWCl)D,'R-92/280

From these and similar results from other small scale tests, we conclude thatthe 35 to 45 mg sample size is adequate to determine the impact energy density a1required to ignite an explosive or propellant. As long as the large scale sample issufficiently uniform to be adequately represented bN these small samples, the impactignition process should be the same in both. Equally important is the manner inwhich this initial reaction either grows to consume the entire sample, or eventuallyfades and dies away. Tllis is a concern of the Ballistic Impact Chamber (BIC) test tobe d1iscussed later in this paper.

Obviously, what is missing from these small scale tests is the confinementavailable in a large charge. Any form of restriction, either inertial or otherwise, thatalters the ignition or its growth will, to rome degree, influence the final outcome of atest. In practice, there are so many possible impact events involving large chargesand so very few large or small scale tests available to evaluate the hazards of any ofthese, that there is no choice but to combine small scale tests that give the energy toignition and the rate of reaction growth with the appropriate data from a few largescale tests, along with the correct theoretical understanding of the ignition event.Together, the combination of large and small scale tests and the proper theoreticalunderstanding should allow us to predict the shock or impact response of a largeexplosive charge.

Finally, since shear is the mechanism of ignition, the surfaces of the anvil andimpactor are important because it is the friction at these surfaces tha t gives rise tothe shear forces as the sample material deforms and spreads during impact. Experi-ments with the energy to ignition test showed that freshly polished surfacesconsiderably increased the amount of energy required to achieve ignition. This isshown in Table 1. However, a careful examination shows that with these polishedsurfaces ignition occurred later than with rougher surfaces, and *t. is likely that most, 0if not all, of the apparent increase in the energy to ignition was really eiastic energystored in the machine. From a practical testing point of view, we have found that it isnot necessary to regularly resurface the anvil and impactor. After resurfacing, theinitial three or four impacts gave anomalously high but decreasing measurementsthat asymptotically settled out to a repeatable value. This suggests that thereactions from the first few tests after polishing, burn and roughen the anvil andimpactor surfaces to create a standard surface roughness which does not changemuch over a large number of subsequent reactions.

TIlE IMPACT SHEAR TEST

The energy to ignition test has been modified and extended in a number ofways. One of these, the impact shear test, is particularly relevant to the abovediscussion on ignition site size. The impact shear test was designed to introduce welldefined, localized, shear induced hot spots into large samples, typically 1 to 50 gramsof hard, difficult to deform materials. In order to maintain a controllable laboratoryscale test, only non-detonable materials were used, chiefly nitrocellulose gunpropellants. However, if required, the test could be performed with explosives byusing an appropriate shelter. The NSWCDD impact machine, Figure 1, served as thetest vehicle with which a 10 kg drop weight was used to impact and drive a 5 Mnthick steel wedge into the sample. The angle of the wedge was 90 degrees. Reactionwas detected by liquid nitrogen cooled, infrared detectors. The test is shownschematically in Figure 4.

7

NSWCDD/TR-92/280

The steel wedge created local shear in the sample during impact in the mannerindicated in Figure 5. Without the shear and energy localization by the wedge, thecritical energy density could not be achieved because of the sample size, and ignitionwould be impossible in any laboratory scale impact test.

Typical oscilloscope records of the acceleration and infrared emissions from animpact on a nitrocellulose propellant sample are shown in Figure 6. The presence ofthe wedge introduces a significant amount of ringing in the accelerometer record thatis not present in the more streamlined energy to ignition test. However, integratingthe accelerometer record from the moment of first impact until ignition occurred,gives an estimate of the change in velocity of the impactor up to the moment ofignition. From this, an estimate of the upper limit of tile energy needed for ignitioncan be determined. For the nitrocellulose propellant in Figure 6, which was impactedby 10 kg released from 1.5 m, the change in the impactor velocity at ignition wasapproximately .12 mJs. From equation (2) the upper limit for the energy to ignitionwas approximately 7 J. The critical impact velocity to cause an occasional ignitionwas measured to be approximately 2.5 m/s (0.35 m drop height).

For the 1.5 m drop height (5.54 mIs impact velocity), ignition took placeapproximately 250 ps after the first impact. Since the drop weight slowed onlyslightly, Av = .12 m/s, the depth that the wedge was driven into the sample whenignition occurred was about 1.4 nun. The volume of material displaced by the wedgeat the moment of ignition was approximately 7 mm 3. It has been shown that for anideal material, the volume of material plastically deformed by a 90-degree wedge isapproximately five times the volume of the material displaced by the wedge. 2 1Recognizing that nitrocellulose has a non-isotropic fibrous structure, and that, aswith all impacts, some portions of the deformed material have been more severelyworked than others, the volume of plastically deformed material was approximately .35 amm. Taking the density of nitrocellulose as 1.2 gmcicm 3, the upper limit of theenergy density required for ignition was about 160 J/gin. Keeping in mind that thisrepresents an overestimate on the energy needed for ignition, it is not too differentfrom that of a similar material in the more accurate energy to ignition test, E --110 J/gmn. To establish an energy density of 110 J/gmn throughout an entire 20 gmsample would require a drop weight of 140 kg released from 1.5 m. The use of such adrop weight mass exceeds even the broadest definition of a small scale test.

As with the energy to ignition test, the actual hot spots-ignition sites occurredin much smaller volumes than the total volume of the nitrocellulose sample deformedby the wedge during impact. Typically, ignition was located in the heavily plasticallydeformed regions near the tip of the shear wedge. Often, ignition appeared associatedwith cracks propagating longitudinally away romathe tip of the wedge, although itcan not be definitely stated which came first, the crack or the reaction. However,crack surfaces formed in this same region during impacts in which no infraredemissions were observed, and by inference no reaction occurred, contained noevidence of reaction products that could be detected by X-ray Photo ElectronSpectroscopy (XPS) in post-impact analysis.22 The XPIS analysis can detect productmolecules present at levels of 1010 gins. When reaction did occur as indicated byinfrared emissions, noise, and smell, the XPS evidence showed that substantialamounts of product molecules existed on the crack surfaces in the plasticallydeformed region close to the wedge. Here, as with the crystalline materials, it is mostlikely that shear in the nitrocellulose sample caused ignition to occur in small regions

NSWCDDPI'R-92/280

of the deformed material. No substantial force build up was observed to haveoccurred at the time of ignition on any of the force measuring gages that monitoredthe force on the grain during impact.

TrHE BIC TEST 6,23

Initiation of a chemical reaction is rot sufficient by itself to predict the responseof an explosive or propellant charge to an arbitrary shock or impact. The growth ofthe reaction in the energetic material has a major role in determining its response tothese stimuli. As with the energy to ignition test, the number of real lifecircumstances that could effect the growth of reaction is too large to deal with inanything but a general way. The object of the BIC test is to obtain information on thegrowth and extent of reaction in a simple and, hope'ully, understandable impact test,and to rely on our developing understanding of the impact ignition processes toextend this data to more general circumstances.

The BIC test is a quasi-confined impact test.* It consist primarily of a cup-likeanvil which holds the sample and provides guidance to a carefully fitted striker. Thestriker is equipped with an "0" ring which fits tightly against the walls of the anvilcup. ReboL:nd of the striker is prevented by mech anicalmeans so that, together withthe anvil cup, it forms a quasi-confined volume in which the impact induced reactionoccurss. The anvil cup contains two ports: in one is mounted a pressure gage, and inthe other is mounted a 0.177 caliber pellet gun pistol barrel. A 0.513 gram, 0.177caliber pe'llet is accelerated down the barrel 'by the hot reaction gases. This servesboth as a means of measuring lihe mechanical'energy available from the hot reactiongases arid as a means of'control ling the pressure in the quasi-confined volume. TheKanvil cup and striker are mounted on the base anv.: block of the NSWCDD ImpactMachine as shown in Figure 7.

Since the standard: 50 percenL drop height test has only a limited meaning, itwas decided to initiate the sampIe with a fix-dir p act that L w0n1u ye aU-iiCitli v

ignite all explosives and propellants regardless oftheir sensitivity to impact. In thisway it would be possible to measure the response of all materials to the same impact.For explosives, the impact is achieved by releasing a 10 kg mass from 1.5 in dropheight. To insure a large enough shear to cause ignition in all Lossible energeticmaterials, a piec, of 180 grit garnet paper is placed on the anvirand the sample isplaced on top of it,*" For impact sensitive materials such as propellants, a 2.5 kg

* A c')fined volume impact test, for liquids and prepellants has been developed and marketed byTechorlu;)roduets Olin-Mathieson Inc- 1) N. Griffin, American Rocket. Society Technical Paper 1706-61, Palm Beach, Fla., 1961. This'test'ha,: been further refined by J.T. Bryant and S.E. Wood, AIAAJ ourn'l, No. 10, 1975, p. 1410.

* 4'l'he presenc( or garnet payer ot the anvil is, an attempt to insure that all samples experience thesu, vhear forecs. duri2ng, ,npawt. T his shear iA devnlnr nS tho samnle, fonws across the surface of

the axnvil dormg the imnpact and arises due to the fricton force between the sample and anvil. ToilIust ra .e the effect o'this surface friction a series of standard impact experiments were performednit the impact sensitive explosive PETN which had a 5 p particle size. The explosive was in a loosepowder and carefully drir, ( under vacuum. The tests were conducted over several humid days inJune 1991 at NSWCIj)O . inmediately after drying, the 50 percent Go-No Godrop height wasme:asVured to be 8 ± .5 cm on our machine as it was configured at that time. Over a 2-hour timeinterval, the PlT'N powder had absorbed enough moisture from the humid air so that the 50 percentdrop height had increased to 12 ± 2 cm. When the material was again dried under vacuum at the50 percent drop height, it again ret.urned o 8 ± .5 cm and, over a 2- hour interval, increased to 12± m2 c. This sequmn:,., of tests.. were repeated a number of times.

I-ETN is not a partic, tiarl y hydrosco-mic material but the effect of even a small amount of water inreducing the surfac arid inferx-aticle'friction is evident. Most modern PIX and propellantluaterials have. binders that ,re impervious to water. 'Ihllt: garnet paper is an attempt to establish asufficiently larg•e rfe fserict.ion force to overwhelm the smaller, uncontrollable and oftenunrecognized forces- that perturh the deformation of the sample during the BIC impact.

9

NSWCDI)/TR-92/280

mass impactor is used. For these materials, the BIC test results are independent of:-he presence or absence of the garnet paper. The sample size is 45 ± 3 mg in the formof a 5 nun diameter disc about 1.2 mm thick. a.

The hot gases from the impact induced ignition accelerate the pellet down thebarrel and the resultant work and rate of work provide a measure of the extent andrate of growth of reaction. Several representative pressure-time records are shown inFigure 8. The pellet velocity, V(T), is obtained from the pressure-time records

V(T) = (A/h) P1() dt (4)

where A is the cross sectional area or the gun barrel, m is the mass of the pellet, P(t)is the time-dependent pressure, and T is the time measured from the moment ofignition. A large number oftests were run with a velocity gate located at the end ofthe gun barrel to measure the exit velocity of the pellet. The measured pellet velocityand the velocity calculated from equation (4) always agreed to within better than10 percent. The kinetic energy of the pellet is just Ep = 4-mV(T) 2. For comparison, itis helpful to divide this energy by the sample mass to obtain an energy density(J/gm•).

SOME EXPERIMENTAL RESULTS

An examination of the BIC test pressure-time records reveals that there are twodistinct types of reactions occurring during impact as shown in Figure 8- One of theseis a very fast reaction that occurs at the beginning of the impact induced ignition.The second is a much slower reaction that follows the initial fast reaction. Inaddition, often the slower reaction has superimposed on itl bursts of'pulses fromio lateoccurring fast reactions. Typically, the rate of pressure increase in the fast reactionwas 1.0 to 10.0 Mpa/ps, while the pressure rise of the slower reaction was usually .01to .1 Mpa/ps. Depending on the material, the peak pressures reached in thesereactions are of the order of 10-1 Mpa for insensitive explosives, and 10 Mpa forsensitive materials like IIMX and RDX.

A number of experiments were performed using heat sensitive film anddecreasing the drop height so that reactions were quenched just after they hadstarted. These showed that the initial fast reactions were due to the impact initiationof large single crystals or an agglomeration of crystals located in the high shearregion near the outer edge of the sample disk. In several experiments the sampleswere seeded with up to five large crystals prior to impact. The number of fastpressure pulses due to the fast reaction component almost always corresponded to thenumber of crystals added to the sample. With the heat sensitive film, it was easy toshow that these fast reactions occurred at the sites of the large crystals. Because ofthe mild impacts very little of the sample mass was consumed by burning.

Our present understanding of these two types of chemical reactions suggeststhat the initial fast reaction component is due to impact induced solid state reactionin the larger crystals located in the high shear regions of the sample. This isbasically the shear band-energy localization-hot spot ignition picture.] 1,12,13 Thistheoretical picture states that, depending on the amplitude of tue applied shearstress, the energy localization can be a very fast solid state process that accounts for

10

NSWCDD/TR-92/280

the initiation of reaction in crystalline solids during shock or impact. For high levelshocks, it can account for the very fast molecular dissociation rates required fordetonation. Here the BIC test measures these solid state reaction rates for lowvelocity impacts. The second slower reaction is believed to be m'Iinly due to burning.

The initial very fast reaction is an important measure of the violent explosion Al

and detonation hazard of an energetic material to low-level impact. Both the peakpressure a, more importantly, the rate of pressure increase, provide measures ofthe extent aad rate of the initial reaction of an energetic material to mild impact.For an energetic, fast reacting explosive like PBX-9404, the peak pressure istypically 10 Mpa and the initial rate of reaction is 1.3 Mpa/ps. At the oppositeextreme, for the insensitive explosive TATB these quantities are .3 Mpa and.02 Mpa/ps. A good measure of insensitivity to shock or impact appears to be thedegree to which this initial fast reaction is minimized o- even eliminated. This isunder-standable from a theoretical point of view since it minimizes or eliminates thefast acting impact or shock driven solid state reactions.

The fluctuations in the rate of pressure increase are important in determiningthe susceptibility of an energetic solid to a violent response from a shock or impact.The fluctuations provide a measure of the interaction of an initiation site with thesurrounding explosive. As such, th2y can potentially provide a link between smallscale tests and the ability to predict the response of large scale charges to anarbitrary shock or impact. 24 Experimentally determined fluctuations for a number ofPBX explosives as a function of explosive content are shown in Figure 9.

'T'o illustrate the importance of the initial rate of reaction, Figure 10 shows theinitial rate of pressure build up for four different materials for which the standard13C test nrocedrure was mnodifid to allow the impact energy to be varied. Three ofthese matei ials (1 through 3) were nondetonable propellants aud the fourth (labeled4) was a detonable propellant. The BIC input energy was varied by reducing themass from 10 to 2.5 kg, or reducing the drop height from 1.5 m to .6 m. Each datapoint is an average of at least ten impacts. For the mildest impacts, the soft PBX-likedetonable propellant was the least responsive of the four materials. However, as theenergy and velocity of impact increased, the detonable propellant developed a veryrapid increase in reaction rate as reflected in the rate of pressure rise. The non-detonable propellants showed only a more or less linear increase in reaction rate,much the same as if only a slow increase were taking place in the number of particlesundergoing reaction as the energy and velocity of the impact increased.

Piupel11ants are desi gnedU to burn. Consequently, In thc BIC test, the propellan-tsamples are entirely consumned during the burning reaction. Therefore, the totalenergy in the samplhe is available to accelerate the pellet. Thishasbeen confirmednumerous times in the BIC test where the total energy transferred to the pellet for agiven propellant type seldom varies more than 10 percent from its averaged value.Fig-tre 11 shows the pellet energy for several different propellants in which the BICimpact energy was changed by changing the drop weight mass and the drop height.Each data point represents an average of the results of at least ten impacts.

This regular behavior of the pellet energy does not occur with explosives since,generally, they do not burn but rather tend to extinguish reactions at low pressures.Consequently, explosive material usually remains on the anvil after the impact whileno material remains after impacts on propellants. For explosives, the failure tototally consume the entire sample is reflected in the variations observed in the pelletenergy, which often ipproach 50 percent or more. Using the standard BIC test

11

NSWCDD,TR-92/280

configuration, the initial pressure build up rate has given a reasonably good rankingof the bullet impact sensitivity for a number of explosive and propellant materials.This is shown in Table 2. p

TABLE 2. RESPONSE OF EXPLOSIVES TO THE BIC TEST

Explosive <dp/dt>(psi/ps) <E>(J/gm) MOST SAFE

NTO (10 percentRDX) 2.2TATB 7.2 10.7C-4 13.N-109 21.4 65.4N-110 23.6 59.4N-106 27. 117.W-121 32, 57.F-108 37. 66.N-103 50. 215.W-114 50. 146.W-201 51.N-5 53.W-108 54. 987W-9 66 54.W-115 79. 163.Comp-B 113. 163,1H-6 176. 90.HMX 203. 163.

LEAST SAFE

im-SUMMARY AND CONCLUSIONS

The small scale tests described here were developed as part of an ongoinginvestigation of the impact and shock initiation processes that has been undertakenat NSWCDD. These tests represent perhaps the best of several novel impact teststhat have been developed in the course of the inves~igation because they most clearlyprovide some of the data necessary for the correct prediction of the response of largescale charges to arbitrary ,hock or impact.

A'- this stage, it is Too early W asserth iataL LhiesCt Lb llU a Inth11 outpULts I UPIeprIset

the final form necessary to describe impact initiation and subsequent reactiongrowth. It is expected that as our understanding of these processes develops, the testsand their interpretation will also change somewhat. However, because of thefundamental importance of their basic data to the physics and chemistry of theinitiation and growth processes, it would seem unlikely that significant changeswould occur in the present tests. More than likely, it will be necessary to supplementthese tests with other tests in order to obtain sufficient data to predict the response ofan energetic material to an arbitrary shock or impact.

Finally, over the 90 years or so that have passed between the time of the firstimpact tests on explosives and today, a considerable amount of frustration andskepticism has been built up over the failure to obtain meaningful and repeatable

12

NSWCDD/TR-92/280

impact test data. In large part, this is due to the failure to understand the physics ofthis seemingly simple test and, in particular, how, when, and where the impact ,ignition first occurs. In point of fact, the impact test is not a simple test but, rather, afairly complicated one, and one that is further clouded by the inevitable variabilitythat seems inherent in the material properties of all physical materials. In spite ofthis, the drop weight impact test is perhaps the simplest possible impact test. If wecannot understand this impact test and its results, then there is little hope that wewill ever be able to understand and predict the impact response of nergetic materialsin more complicated and realistic situations.

1 0

13

NSWCIDDiTR-92/280

LUUl t4jt

0~ 0 ý

U LUU _j Z c

LU a- cif:Qa- Uw-> LU

LL' _

o I QC,LUU

LU

0~0)

EO E

I-ui__j CZ

~LU Q-LJ.

LU

L4-H > O

L)

LUUo -i

14 4

NSWCDD/ITR-92/280

0U L

__ LU .

0< ito LU -

0 0 LL l

tZI Z C-ZUJI < 0-0 L

-j-j

UJ L9

Ucr 4 ___

LU LL]

LU 00>0 E

I--

-J L E>0CL,

Lm 0 ' E-

NSWCDD/TR-92/280

E t E- I)-E E

~0 W

- E- W_ - -'

4-4

LULU el -

0.L Lu 0

LU .Ln

LU

-C C

16p

N SWC DD/TR-9 2/2 80

LUU

CL x

4<LJ V)LU -j x 0

/ r0V.-E0 E cc

u 2 U)

CLI0E

-J SI0

F-LU/

0 0

LU 0

LUU 0

17

NSWCDD/TR-92/280

A

0 W9-Q

• INCREASEDINSTABILITYF (h) >1 PREDICTED FLUCTUATION

h/•e= .13oN-106 F(h) z 1g Lx-•,4 ~W-108// ,1

------------ -- -- - - -------oN-103 - 1 - ^OSTDS4M(11) 0 o (F-108S-D)-- - NJ I-0HIIX-6 --- :

D 8o PROP__W W-113 :o 88,,- .5 - 1 2/4) -o0 v 88

INCREASED "J'vW l14 N-109(VARIATIONS)VSTABILITY - 14

F (h) <1 F-108 (VARIATIONS)

0 .05 .1 .15 .2 .25 .3 .35 .4 VlnVo

(Each data point represents the average of at least ten BIC impacts.)

FIGURE 9. NORMALIZED FLUCTUATIONS IN TIlE INITIAL ENERGY RELEASE RATE((de/dL/(dE/dt)), AS A FUNCTION OF TIlE VOLUME FRACTION IN INERTMATERIAL, VIn/Vo.FOR A NUMBER OF EXPLOSIVES AND PROPELLANTS

18

NSWCDDIT R-92,/2800

00

-j ll n L2 E

-j 0

iiL~ . .L

~~+*oCL :i LII lz

.1 3

00

L U C

to

z -j

LU <

00 :9WýPE,

QQ1:

.3

jLLJ~~E E- ~-

C-4

1912C:: Ln

NSWCIDfIR-92/280

I

REFERENCES

1. R. Robertson, Trans. Chem. Soc., p. 18, 1921.

2. J. Wilby and H.N. Mortlick in Proc. of the International Conference on Sensitivity andHazards of Explosives, Eds. C. K. Adams and E. C. Whitbread, London, England, 1963.

3. W. Will, Zeits. Schiess-u. Sprengstoffwesen, 1, 209, 1906.

4. H. Kast, Zeits. Schiess-u. Sprengstoffwesen, 4, 263, 1909.

5. 11. Xoenen, K. H. Ide, and W. Haupt, Explosivstoffe, 8,178, 1958.

6. C.T. Afanos'ev and U.K. Bobolev, "Initiation of Solid Explosives by Impact," IsraelProgram of Scientific Translations, Jerusalem, 1971.

7. L. Avrami and R. tIutcbhnson, Energetic Materials, Vol. 2, E.D. Fair and R.E. Walker,

Eds., Plenum Press, N.Y., 1977.

A. A- Macek, Chem. Revs.. 62- 41-63, 1962.

9. F.P. Bowden and A.D. Yoffe, Initiation and Growth of Explosion in Liouids and Solids,Cambridge University Press, 178.

10. J. E. Field, G. M. Swallowe, and S. N. Heavens, in Proceedings R. Soc. London, A383,1982, p. 231.

11- C.S. Coffey, Phys. Rev. (B), Vol. 32, No. 8, p. 5335, 1985.

12. C.S. Coffey, "Initiation of Explosives by Shock or Impact," Proceedings of the 9thSymposium on Detonation, OCNR Pub, 113291-7, p. 1243, 1989.

13. C.S. Coffey, "Localization of Energy and Plastic Deformation in Crystalline SolidsDuring Shock or Impact," to be published in the J. of Applied Physics, Oct 1991.

14. VyF. DeVost and C. S. Coffey, Evaluation of Equipment Used to Impact Test SmallScale Explosive Samples, NSWC TR 81-215, 1981.

15. C.S. Coffey and V.F. DeVost, Critical Initiating Impact Energy ofExplosives,NSWC TR 86-14, 1986

16. C.S. Coffey, V.F. DeVost and D.L. Woody, Towards Developing the Capability toPredict the Hazard Res' nse of Energetic Materials Subjected to Inpact, inProceedings of the 9th • posium on Detonation, 1989.

21

NSWCDD/TR-92/280

REFERENCES (Cont.) 017. W. Schroder and D. A. Webster, J. Appl. Mech., 16, 289, 1949.

18. F.P. Bowden and A.D. Yoffe, Fast Reactions in Solids, Academic Press, London,England, 1958.

19. C.S. Coffey and S.J. Jacobs, J. Applied Physics, 52, 1981, pp 6991-6993.

20. C. S. Coffey, M. J. Frankel, T. P. Liddiard, and S. J. Jacobs, in Proceedings of the 7thSymposium on Detonation, Jun 1981, NSWC 1\IP 82-334, 1982.

21. R. Hill, The Mathematical Theory of Plasticity, Oxford University Press, p. 215, 1956.

22. J. Sharma and B.C. Beard, "Fundamentals of X-Ray Photoelectron Spectroscopy (XPS)and Its Applications to Explosives and Propellants," in Chemistry and Physics ofEnergetic Materials, Ed. S.N. Bulusu, NATO ASI Series C, Kiuwer AcademicPublishers, Boston, 1990, Vol. 309, p. 569.

23. C. S. Coffey, V. F. DeVost and D. L. Woody, "The Ballistic Impact Chamber (BIC)Test," in Proceedings of the JANNAF Hazards Mtg., Mar 1988.

24. C.S. Coffey, "On The Fluctuations in the Initiation and Growth of Reactions inExplosives and Propellants During Shock or hnpact," (to be submitted to J. ofExplosives, Propellants and Pyrotechnics, 1991).

22

NSWCDD/TR-92/280

44

APPENDIX A

ANALYSIS OFTHE EXPLOSIVE RESEARCHLABORATORY/

NSWCDD IMPACT MACHINES

The Explosive Research Laboratory (ERL) Impact Machine has been used in theUnited States in various modified forms since the early 1940's. The machine wasdesigned at the ERL, and the original design has changed very little over theyears.A-1 The machine described in reference A-1 is the current version.

The principal components of the machine consist of interchangeable dropweights of 2.5 and 5 kgs, a 535 gm striker, a 500 gm anvil, and a massive base ofapproximately 500 kg. The striker and the anvil are nearly identical columns, both3.2 cm in diameter. The striker is 9 cm long and has a slightly rounded top. Theanvil is 7.6 cm. long and is rigidly attached to the base mass.

In impact tests, the explosive sample is placed between the striker and theanvil. A complete description of the response of an explosive or propellant sample toimpact is.well beyond our prsnt, analytical capaulities, and certainly far beyondthe predictive capabilities of any computer code. However, much can be learned byexamining the response of the impact machine when no sample is present. This is the"bare tools" case.

The duration of the "bare tools" impact is approximately 350 to 400 us. Thetime required for a disturbance to travel the diameter and length of the striker andanvil column is approximately 6 and 20 ps, respectively, where the speed of sound insteel is taken as 5 x 10 m/s. These transients have a duration of at least one order ofmagnitude shorter than that of the duration of the "bare tools" impact. Therefore, itcan he assumed that the drop weight, striker, and anvil can be treated as lumpedmass-spring elements and that, during the compressive phase of the impact, theseelements remain in contact with each other. Only in tension do the striker, anvil anddrop weight separate. This occurs only during rebound when the motion of thecomponents are no longer of interest.

The upper surface of the striker is rounded to lessen the alignment problemwith the impacting drop weight. The presence of this curved surface results in theadditional complication of a Hertzian impact. Here, for simplicity, the effect of thecurved surface will be neglected on the basis that it makes only a minor contributionto the overall response of the drop weight-striker/anvil system.

A schematic diagram detailing the lumped components of the impact machine isshown in Figure A-1. The mass and spring constant of the impactor are M 1 and K1 ,respectively. Similarly, M 2 , K 2 and M 3 , K 3 are the mass and spring constants of the

A-1

NSWCDDAFR-92/280

striker/anvil and base mass, respectively. The equations of motion of this coupledsystem are

I

MIX1 = -Kl(Xi - X2- X- ) (A-d)

M2 X2 = - 2 (X - X3 ) + K1 (X, - X 2 - X3) (A-2)

MaXs = _Ka Xa + K2 (X2 Xa) (A-3)

where

X, el W)-El (0)

X 2 C e2 (t) - C 2(0)

X= e (t) - e (0)

1 1

The mass of 1he base is so large compared to the other masses involved that it cansafely be taken as infinite, leading to the simplification that

X= 0 (A-4)

and

- 2 - (A-5)"'33- k + h3 2

For a steel cylinder, the spring constant is given as K = EA/L, where E is the elasticmodulus, A is the cross sectional area, and L is the length of the column. For thelarge base system, K3 ;> K2 , so that X 3 = 0. This allows considerable simplificationof the equv, tions of motion which now become

MIX,= -KI(X 1 -X 2 ) (A-da)

A-2

NSWCDD/TR-92/280

M2X2= - K2X2 + KI (Xi -X 2). (A-2a)

It is convenient to solve these by Laplace Transforms since the impact starts at timet = 0. Equations (A-la) and (A-2a) transform as

MIP2X 1 - M I [PX 1 (0) + X1 (0)] = -K 1 (X1 - X2) (A-6)

MP2 p 2 X 2 M 2 [PX 2 (0) + X 2 (0)] = -K2X 2 +KI(X1 - X2 ), (A-7)

where the over head bars denote the transformed quantity. The appropriate initialconditions are xl(O) = x2(0) = 0 and x2(0) = 0 since the system initially is not understrain nor is the striker moving. The initial velecity of the drop weight released froma height h is xi(0) = (2gh). Solving equation 6 for xj gives

1 + 1(0)2+W12x21 (A-8),2 + W1 2

Combining (7) and (8) givesK1 (A-9)

Sx (0)

2 M 2 1

(p2 + 1,2) (p2) (p2 + p•2)

where 6)12 K /Ml, and

2 + 21,2 1+ 12 1 1• 4 2 2 4 ( -0-- 2 + -2 (1 + 4 M-)(0l 2w) (01 +t 5)a2

2 2 M 1 112 12'

and2 +W2

2 1 12 1 2 2 4 (A-11)- + (+ + 4 A 0+4 -)W 2 112 )+G2 2 M 1"2 1 A12'

12W (A-12){12 -- M2

Equation 9 can be recast by partial fractions as

2 2= 1 - I W ()2 2+A- 13)(-il2M 2-i 2---2 Xd0)I 2 211 i) r (P - x) p(P + iA} )(P - ip)

Taking the inverse Laplace Transformation of Equation (A-13) gives the incrementalchange in length of the striker as

A-3

NSWCDD/TR-92/280

( 1 11i "I I ( -4

.X2 (t) = --1 X (04 2 J sinit -- sinl't (A 14)

The force that the striker exerts on the base anvil is F2(t) = K2X2(t). For the originalERL-designed impact machine, M2 <. MI and K2 > K1 so that P ; co,- r = W12

and w•12 > w3 . In this limit, the force on the sample is approximately

W • ( 1 (0) sinlt I I - sinfa)12tK I KK2 1() 1 w 12

K1XI(O)

_- siC.) (A-5)

S..

[*t DROP WEIGHT

I ) K2

M2Cl t STRIKER AND ANVIL.

C2 K2

M3 BASE

-- _=_ K3 -

FIGUlI, A-1. LUMPIE1) MASS-SPRING ELEMENTS OF SIMPLIFIED IMPACT MACHINE

A-4

NSV'CDDITR-92/280

4j

DISTRIBUTION

Copies C~e

ATTN ONR 1132P (R MILLER) 1 ATTN 6-A-145 1ONT2OT(L V SCHMIDT) 1 DDESB-KT 1ONT213 (D SIEGEL) 1 J WARD 25ONT 23 (Y J FAULSTICH) 1 CHAIRMANONT 232 (D HOUSER) 1 DEPARTMENT OF DEFENSE EXPLOSIVES

CHIEF OF NAVAL RESEARCH SAFETY BOARD800 N QUINCY STREET UCT I HOFFMAN BUILDING 1ARLINGTON VA 22217-5000 2461 EISENHOWER AVENUE

ALEXANDRIA VA 22331ATTN R SIEWARC 1OUSDRE/R&AT MST ATTN OP-098 1THE PENTAGON OP-981 1WASHINGTON DC 20301 OP-982 1

OP-983 1ATTN D ANDERSON 1 OP-987 1OUSDREITWP-NW&M OP-02T 1TIlE PENTAGON OP-22 1WASHINGTON DC 20301 OP-22b 5

OP-32 1OUSDREfTWP--OM OP-35 1ATTN C KOPCSA* 1 OP-37 1TIlE PENTAGON OP-374 1WAStHIN(GTON fIC 2(.3iP-C OP-501 1

OP-502 1ATTNq I MEN? 1 OP-503 1USD(iQ.':;IY i k(t/AT/ET) OP-72 ISTAFF SPECIALIST FOR WEAPONS OP-74 1

TECHINOLOGY OP-75 1TIlE PENTACON CHIEF OF NAVAL OPERATIONSWASHINGTON DC 2G0301 NAVY DEPARTMENT

WASHINGTON DC 20350ATTN SURFACE WARFARE I

AIR WARFARE 1 ATTN SPAWAR-05 1SUBS/ASW 1 COMMANDER

OASN/RE&S SPACE AND NAVAL WARFARE SYSTEMSNAVY DEPARTMENT COMMANDWASHINGTON DC 20301 WASIIINGTON DC 20363-5100

(1)

NSWCDD/TR-92/280

DISTRIBUTION (Con,.)

copies Copies

ATTN SEA-05R I ATTN CODE 177 1SEA-55 1 TECHNICAL LIBRARY 1SEA-55X I COMMANDFRSEA-55X1 1 DAVID TAYLOR RESEARCH CENTERSEA-55X2 I PORTSMOUTH VA 20375SEA-55Y 1SEA-66U 1 ATTN TECH NICAL LIBRARY 1SEA-62 1 COMMANDING OFFICERSEA-62Y 1 NAVAL RESEARCH LABORATORYSEA-62Z 1 WASHINGTON DC 20375SEA-63 1SEA-63D 1 ATTN TECHNICAL LIBRARY 1PMS-402 I COMMANDING OFFICERPMS-406 1 NAVAL SURFACE WARFARE CENTERPMS-407 1 DAHLGREN DIVISION

COMMANDER COASTAL SYSTEMS STATIONNAVAL SEA SYSTEMS COMMAND 1ANAMA CITY FL 32407-5000WASHINGTON DC 20362-5105

ATTN TECHNICAL LIBRARY 1ATTN AIR-5004 1 COMMANDER

AIR-51623 1 NAVAL UNDERSEA WARFARE CENTERAIR-540 1 DIVISIONAIR-5404 1 NEWPORT RI 02841-5047AIR-93 IAIR-932F 1 ATTN CODE 3917 1AIR-93211 I CODE 38 (I DERR) IAIR-932K 1 CODE 389 (T BOGGS) 1AIR-932T I CODE 32 1PMA-242 1 CODE 3205 1

TECHNICAL LIBRARY 1 CODE 32051 (S DEMAY) 1COMMANDER CODE 3208 1NAVAL AIR SYSTEMS COMMAND CODE 3211 1WASHINGTON DC 20361 CODE 326 1

CODE 3261 1ATTN TECHNICAL LIBRARY 1 CODE 3263 1

CODE 17 1 CODE 3264 1CODE 172 1 CODE 3265 (J PAKULAK) 1CODE 1740.3 (R GARRISON) 1 CODE 3266 1CODE 1740.4 (S WANG) 1 CODE 327 1CODE 175 (W SYKES) I CODE 381 1CODE 1750.2 (W CONLEY) I CODE 385 1CODE 1740.2 (F FISCH) I CODE 3850 1

COMMANDER CODE 3853 1DAVID TAYLOR RESEARCH CENTER CODE 3891 (J COVINO) IBETHESDA MD 20084 CODE 39 1

TECHNICAL LIBRARY ICOMMANDERNAVAL AIR WARFARE CENTERWEAPONS DIVISIONCHINA LAKE CA 93555-6001

(2)

NSWCI)DITR-92/280 V.

DISTRIlBU rION (Cont.)

Copies Copies

ATTN CODE 2730D 1 ATTN CODE 2145 1TECHNICAL LIBRARY I COMMANDERJ CHANG 1 PACIFIC MISSILE TFST CENTERP DENDOR 1 POINTMUGU CA 93042L NEWMAN 1G. NAU.LE'Lr 1 COMMANDING OFFICER

COMMANDER SEAL TEAM 2NAVAL SURFACE WARFARE CENTER FPO NEW YORK NY 09501-4633 1INDIAN HEAD DIVISIONINDIAN HEAD MD 20640-5000 COMMANDER

NAVAL UNDERSEA WARFARE DIVISIONATTN TECHNICAL LIBRARY 1 KEYPORT WA 98345-0580 1COMMANDERCENTER FOR NAVAL ANALYSES COMMANDER2000 SUITLAND ROAD NAVAL SURFACE WARFARE CE' I'TERWASHINGTON DC 20390-5140 PORT IIUENEME DIVISION 1

PORTtHUENEME CA 93043-5007ATTN LIBRARY1

SUPERINTENDENT COMMANDERNAVAL POSTGRADUATE SCIIOOL NAVAL WEAPONS EVALUATIONMONTEREY CA 93940 FACILITY

KIRTLAND AIR FORCE BASEATTN TECHNICAL LIBRARY I ALBUQUERQUE NM 87117 1PRESIDENT I *

A ,,^ A e . AT f. ,,-. AT-.N CODE 3031 (E N iE-AQ)NEWPORT RI 02841 CODE 50D (A NORRIS) 1

CODE 505 (J E SHORT) 1ATTN RDT OFFICER 1 CODE 90 (A E WIIITNER) 1

SEAL TEAM 1 COMMANDERUNDERWATER DEMOLITION NAVAL SURFACE WARFARE CENTER

TEAM I CRANE DIVISIONCOMMANI)ING OFFICER CRANE IN 47522-5000NAVAL AMPHIBIOUS BASE

CORONADO ATTN CODE 321 (M BUCIIER) ISAN DIEGO CA 92155 COMMANDING OFFICER

NAVAL WEAPONS STATIONAT'TN IDT OFFICER I CONCORD CA t •i20-5000COMMANDING OFFICERNAVAL AMPIIIIOUS BASE ATTN CODE 470A 1LIrITLE CREEK LIBRARY 1NORFOLK VA 23511 OFFICER IN CHARGE

NAVAL SURFACE WARFARE CENTERATTN TECHNICAL LIBRARY I INDIAN HEAD DIVISIONCOMMANDING OFFICER YORKTOWN DETACIIMENTNAVAL EXPLOSIVE ORDNANCE YORKTOWN VA 23691-5110

DISPOSAL TECHNOLOGY CENTERINDIAN HEAD MD 20640

(3)

NSWCDDrrR-92/280

DISTRIBUTION (Cont.)

Copies Ccpies

ATTN TECHNICAL LIBRARY ATTN CHIEF DOCUMENTS 1DIRECTOR D DREITZLER 1 4DEFENSE NUCLEAR AGENCY REDSTONE ARSENAL ARMY MISSILEWASHINGTON DC 20305 COMMAND

REDSTONE ARSENAL AL 35809ATTN C FOWLER1

DEFENSE SCIENCE BOARD ATTN SLC-BR-TB-EE 1THE PENTAGON SLCRBR-IB-I (P KASTE) 1WASHINGTON DC 20301 V BOYLE 1

ATTN LIBRARY 0 BLAKE 1

DIRECTOR R FREY 1

DEFENSE RESEARCH AND G MELANI 1

ENGINEERING M CIIAWLA 1

WASHINGTON I)C 20305 J TRIMBLE 1

TECHNICAL LIBRARY IATTN LIBRARY ARMY BALLISTIC RESEARCHDIRECTORLABORATORYDEFENSE ADVANCED RESEARCH ABERDEEN PROVING GROUND

PROJECTSAGENCY ABERDEEN MD 21005-50661400 WILSON BLVDARLINGTON VA 22209 AWN LIBRARY

COMMANDER OFFICERAWFN TECHNICAL IJBRARY I HARRY DIAMOND LABORATORYINSTITUTE FOR DEFENSE ANALYSES 2800 POWDER MILL ROAD1801 N BEAUREGARD STREET ADELPHI MD 2ML783 RAALEXANDRIA VA 22311

A'WIN LIBRARY AWTN IItIIB-EA-A

COMMANDING GENERAL ARMY ENVIRONMENTAL HYGIENE

MARINE CORPS DEVELOPMENT AND AGENCYABERDEEN PROVING GROUNDEDUCATION COMMAND

MARINE CORPS LANDING FORCE ABERDEEN MD 21005DEVELOPMENT CENTER ATTN J BARKELEY

QUANTICO VA 22134 ARMY MEDICAL BIOENGINEERING

ATTN IRSAR-IRC 1 RESEARCH AND DEVELOPMENTDRSAR-LEM (R FREEMAN) 1 LABORATORYDRSAR-SF (R YOUNG) 1 FORT DIETRICK MD 21701

ARMY ARMAMENT MUNITIONS AND ATN G R HUSKClIEMICALCOMMAND 1TTNDER

ROCK ISLAND IL 61299-6000 COMMANDERARMY RESEARCH OFFICE

ATTN TECHNICAL LIBRARY 1 P.O. BOX 12211ARMY ARMAMENT RESEARCH AND RESEARCN TRIANGLE PARK

DEVELOPMENT COMMAND NC 27709-2211DOVER NJ 07801 ATTN DRXTII-TE-D

ATTN SMCAR-AEE-WW (PAl LU) 1 ARMY TOXIC AND HAZARDOUSUS ARMY RESEARCH, DEVELOPMENT, MATERIALS AGENCY

AND ENGINEERING CENTER ABERDEEN MD 21005

PICATINNY ARSENAL, NJ 07806

(4)

NSWCDDfrR-92/280

DISTRIBUTION (Cont.)

Cpis Copiess

ATTN SMCCR-DDP I ATTN TECIHNICAL LIBRARY 1 4COMMANDER SANDIA NATIONAL LABORATORIESARMY CHEMICAL RESEARCH P.O. BOX 969

DEVELOPMENT AND ENGINEERING LIVERMORE CA 94550-0096CENTER

ABERDEEN PROVING GROUND MD ATTN STRUCTURAL AND SHOCK21010-5423 CHEMISTRY

ATTN WL/MNME DIVISION 1153

WL/MNMW I (MARK U ANDERSON)WL/MNOI 1 SANDIA NATIONAL LABORATORIESWL/MNME (G GLENN) 1 ALBUQUERQUE NM 87185-5800WIMfMNMF(RMABREY) 1WUMNME(R MCKINNEY) I ATTN RECORDS CONTROL FOR

WLJMNMW (J FOSTER) I RICHARD ANDERSON I

AIR FORCE ARMAMENT DIVISION TECHNICAL LIBRARYEGLIN AIR FORCE BASE FL 32542-6009 ARGONNE NATIONAL LABORATORY

9700 SOUTH CASS AVENUEATTN T MATUSKO ARGONNE IL 60439AIR FORCE OFFICE OF SCIENTIFIC

RESEARCH ATTN T W CHRISTIANBOLLING AIR FORCE BASE THE JOHNS HOPKINS UNIVERSITY

WASHINGTON DC 20332 APPLIED PHYSICS LABORATORYChIEMICAL PROPULSION INFORMATION

ATTN TECHNICAL LIBRARY 1 AGENCY 0M FINGER 1 JOHINS HOPKINS ROAD

C M TARVER LAUREL MD 20707E LEE 1P URTIEW 1 ATTN TECHNICAL LIBRARY

J D IIALLQUIST TIlE JOHNS HOPKINS UNIVERSITY

L M ERICKSON APPLIED PHYSICS LABORATORY

E JAMES 1 JOHNS HOPKINS ROAD

UNIVERSITY OF CALIFORNIA LAUREL MD 20707LAWRENCE LIVERMORE NATIONAL

LABORATORY ATTN CODE TERA (J JOYNER)P. O. BOX 808 TECHNICAL LIBRARY 1

- T.~.vflV~1,tfl ~r IMFni Tr"T'l (It' TiNT 101 XiNLIVERMORE CA 94550 NEW., IME,,,CO INI , II 1-i , ILI .... ,TECHNOLOGY

ATTN J. REPA 1 SOCORRO NM 87801M. J. URIZAR 1S. W. PETERSON 1 ATrN LIBRARYL.SMITII 1 E LIZKA

C. FOREST 1 APPLIED RESEARCH LABORATORYC. MORRIS 1 PENNSYLVANIA STATE UNIVERSITYA. W. CAMPBELL 1 P.O. BOX 30

R. ENGELKE 1 UNIVERSITY PARKP. C. CRAWFORD STATE COLLEGE PA 16801

LOS ALAMOS NATIONAL LABORATORYLOS ALAMOS NM 87545

(5)

NSWCDDITR-921280

DISTRIBUTION (Cont.)

Copies Copies

DEFENSE TECHNICAL INFORMATION ATTN R CHURCH 1CENTER TRW

CAMERON STATION SAN BERNADINO CA 92401ALEXANDRIA VA 22304-6142 12

ATTN G CHINATTfN LIBRARY 1 AEROJET ORDNANCE AND

B HAMMANT I MANUFACTURING COMPANYP HART 1 9236 EAST HALL ROADP HASKINS 1 DOWNEY CA 90241

ROYAL ARMAMENT RESEARCH ANDDEVELOPMENT ESTABLISHMENT ATTN A MELLOR

FORT IIALSTEAD SEVENOAKS KENT VANDERBILT UNIVERSITYUNITED KINGDOM NASHVILLE TN 37235

ATTN LIBRARY 1 ATTN R HODGES 1II R JAMES 1 J SMITH 1

ATOMIC WEAPONS ESTABLISHMENT LOCKHEED MISSILES ANDFOULNESS ISLAND SPACE COMPANYESSEX SS3 9XE P 0 BOX 504UNITED KINGDOM SUNNYVALE CA 94086

ATTN LIBRARY 1 ATTN G WILLIAMSP LEE 1 HERCULES INCORPORATEDR HI MARTIN 1 ROCKET CENTER

ROY.L ORDNANCE PLC PO BOX 210WESTCOTT AYLESBURY ROCKET CENTER WV 26726BUCKING!HAMSHIRE iHP18ONZ ATTN M KLAKKEN 1UNITED KINGDOM M BERGER 1

L LOSEE 1ATTN P JONES T SPEED 1THE BRITISH EMBASSY HERCULESBRITISH DEFENCE STAFF BACCHUS WORKS3100 MASSACHUSETTS AVENUE NW MAGNA UT 84044-0098WASHINGTON DC 20008

ATTN D JEFFJONESATTN M CHICK MATERIALS ANDMATERIALS RESEARCH LABORATORY PROCESS DEVELCPMENT IP 0 BOX 50 THIOKOL CORPORATIONA 0,,,r',,r IT A I Uý IF •T•"rI'r¢'DT A '~ •At• •'A!K• Ii'Al~'N'xlokN.m~rA V " YJNý'rDV 2, TPACTICAL OPRv1rArI1ANc

AUSTRALIA HUNTSVILLE DIVISIONP. O. BOX 400006

ATTN V. F. DEVOST 3 HUNTSVILLE AL 35815-1506S JACOBS 1W PICKLER 1 ATTN LARRY LEET P LIDDIARD 1 KTECH CORPORATIONJ W WATT 1 901 PENNSYLVANIA AVENUE

ADVANCEDITECIINOLOGY AND ALBUQUERQUE NM 87110RESEARCH INC

LAUREL TECHINOLOGY CENTER14900 SWEITZER LANELAUREL MD 20707

(6)

NSWCDDITR-92/280

DISTRIBUTION (Cont.)

Copies Copies

ATTN K L CHRISTIANSON INTERNAL DISTRIBUTION (Cont)J L TIOULTON 1 R13 1G JOHNSON 1 R13 (K D ASIIWELL) 1

ALLIANTTECIISYSTEMS INC (R N BAKER) 1

?225 NORTHLAND DRIVE (R D BARDO) 1BROOKLYN PARK MN 55428 (A BROWN) 1

(C S COFFEY) 6

AT'TN GIFT AND EXCHANGE (J DAVIS) IDIVISION 1 (D L DEMSKE) I

LIBRARY OF CONGRESS (J W FORBES) 1WASHINGTON DC 20540 (R11 GUIRGUIS) 1

(P K GUSTAVSON) I

INTERNAL DISTRIBUTION (R N HAY) !

E231 2 (It D JONES) 1

E232 3 (K KIM) 1

E342 I (R J LEE) 1

G13 (D L DICKINSON) ( (W W LEE) 1

G13 (T WASMOND) 1 (E R LEMAR) 1

G22 (W II IHOLT) 1 (P J MILLER) 1

(W MOCK) 1 (C T RICHMOND) 1

(S S WAGGENER) 1 (IH W SANDUSKY) 1

RIO 1 (C T SUTHERLAND) 1

Ri01 (D G TASKER) 1

RIO (R R BERNECKER) 1 (W 11 WILSON) 1 I 0RlOA (C DICKINSON) 1 (D 1, WOODY) I

RIOAI 1 (F J ZERILLI) 1

RlOA2 1 R14 1

R10B (11 S HIAISS) 1 R14 (J W KOENIG) 1

Ril 1 R15 1

R1l (A. DUONG) 1 R15 (S COLLIGNON) 1

(C. GOTZMER) 1 R42 (R DEW[I') 1(N. JOHNSON) 1 U10 (W WASSMANN) !(Y. TRAN) I Ull (R PLENGE) I

(J. LEARlY) 1 (D ItlNELY) I

(J. KELLEY) I U12 (C SM1TII) 1I"• 2 ( 1 HW INICKLE.Y) .

R12 (B A BAUDLER) 1 U32 (G PARRENT) 1

(B C GLANCY) 1 U43 (L LIPTON) 1

(L C HUDSON) 1(G LAIB) 1(C, L MENSI) 1(L J MONTESI) 1(P F SPAIIN) 1(T SPIVAK) 1

(7)

fornm Approved

REPORT DOCUMENTATION PAGE OMB No0 704 0188

Public rep."s tn burden for tis h Ollutior) of lnforniation is vtlim.ttc'O to average I hour x'r rermnse, InLldlchn the itme for ton icwlng irl ru 1ont. seoctrir.lly eitilnyj data-Oar-ct, 9' thea s, and rurirro •i ij th,- dota, noded. and onmplict,. ,rnd , viewing the collertton of rlorirotOn Send Lorfmrenfs regarding thit burden r irstnrnate or any otherasp"Ic of this zolletion of information. indudcing sugq.strons for reuirg this buideno, to WashingtonoIn eadqJarter, SeltOLt. tnrectorate for Inforrnrton Operotiont and

121orn Jeff Daf-t | 1scraolt ihway. Suite 1204. Arhn•lton. VA 22202-4JO2. -nd to the Offince of Manadgenent ard tudget, Paperwork .itnutim.on l'roJd (( 104-01)i.BWrihirylon. DC 2050?

1. AGENCY USE ONLY (Leaveblank) 2. REPORT DATE 1992 3. REPORTTYPE AND DATES COVERED

4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

Impact Testing of Explosives and Propellants

6. AUTHOR(S)

Dr. Charles S. Coffey (NSWCDD) and V. F. DeVost (ATR, Inc.)

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

Naval Surface Warfare Center I)ahlgren Division REPORT NUMBER

White Oak Detachment NSWCI)DITWR-92/28010901 New Hampshire AvenueSilver Spring, Maryland 20903-5000

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORINGAGENCY REPORT NUMBER

11. SUPPLEMENTARY NOTES

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Approved for public release; distribution is unlimited.

13. ABSTRACT (Maximum 200 words)Tihe drop weight impact test is the simplest and easiest test that can be performed on small quantities

of explosives or propellants, and yet it has only a minimal role in assessing explosive sensitivity orperformance. This report examines the drop weight impact test as it is currently used, describes its majorflaw, and suggests an alternative test that holds the promise of oroviding the impact energy required toignite an energetic material. Other impact tests are described. One of these, the Ballistic ImpactChamber T 'est, measures the rate of reaction and extent of reaction during impact. This test demonstratesthat during impact there are two forms of initiation reactions that occur: one that is very fast and is likelydue to direct impact-shear initiation of the crystalline solids in the sample and, the other, a much slowercomponent, is thought to arise due to burning of the sample.

14. SUBJECT TERMS 15. NUMBER OF PAGES

lExplosives Drop Weight Impact Test 38Propellants Ballistic Impact Chamber Test 16 PRICE CODE

Explosives Sensitivity Initiation Reactions17, SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION UF

OFREPORT OF THiS PAGE OF ABSTRACT ABSTRACT

IJNNCLA SSI VI E1) UNCLASSIFIED) UNCIASSIFIE1) SAI{5N750-01 20-55tor0 u 29 (Rev' 2 A'

"t' - -tlr -d by ANSI Stiu 19 ItII tr 10i2

V6

GENERAL INSTRUCTIONS FOR COMPLETING SF 298

The Report Documentation Page (RDP) is used in announcing and cataloping reports. It is important thatthis information be consisten with the rest of the report, particular y t ie cover and its title page.InstruCtions for filling in each block of the form follow. It is important to stay within the lines to meetoptical scanning requirements.

Block 1. Agency Use Only (Leave blank). Block 12a. Distribution/Availability Statement.Denotes public availability or limitations. Cite any

Block 2. Report Date. Full publication date including availability to the public. Enter additionalday, month, and year, if available (e.g. 1 Jan 88). limitations or special markings in all capitals (e.g.Must cite at least the year. NOFORN, REL, ITAR).

Block 3. Type of Report and Dates Covered. Statewhether report is interim, final, etc. If applicable, DOD - See DoDD 5230.24, "Distributionenter inclusive report dates (e.g. 10 Jun 87- Statements on Technical Documents."30 Jun 88). DOE - See authorities.

Block4. Title and Subtitle. Atitle is taken from the NASA - See Handbook NHB 2200.2

part of the report that provides the most meaningful NTIS - Leave blank.

and complete information. When a report is pre-pared in more than one volume, repeat the primarytitle, add volume number, and include subtitle for Block 12b. Distribution Cude.

the specific volume. On classified documents enterthe title classification in parentheses. DOD - Leave blank.

DOE - Enter DOE distribution categoriesBlock 5. Funding Numbers To include contract and from the Standard Distribution forgrant numbers; may include program element Unclassified Scientific and Technicalnumber(s), project number(s), task number(s), and Reports.woek unit number(s) Use the following labels- NASA - Leave blank

NTIS - Leave blank.C - Contract PR - ProjectG - Grant TA - lask Block 13. Abstract. Include a brief (Maximum 200PE - Program WU - Work Unit words) factual summary of the most significant

Element Accession No information contained in the report.

BLOCK 6. Author(s). Name(s) of person(s)responsible for writing the report, performing the Block 14. Subject Terms. Keywords or phrasesresearch, or" credited with the content of the report. identifying major subjects in the report.If editor or compiler, this should follow the name(s)

Block 15. Number of Pages. Enter the totalBlock 7. Performing Organization NamefsqaQnd number of pages.Address(es. Self-explanatory.

DIL,u.. ,E P ie o. __ __ __ ... ... .. :p ..... pi IE....U•Block 8. Performing Organization Report Number (NTIS only)

Enter the unique alphanumeric report number(s)

assigned by the organization performing the report. Blocks 17.-19. Security Classifications. Self-

Block 9. __on_.orinqgMonitoring AgencV Name(s) explanatory. Enter U.S. Security Classification in

and Address~es.. Self-explanatory. accordance with U.S. Security Regulations (i.e.,UNCLASSIFIED). If form contains classified

Block 10. Sp.Lnsorinn/Monitoring Agency Report information, stamp classification on the top andNumber. (If Known) bottom of the page.

Block 11. Supplementary Notes. Enter information Block 20. Limitation of Abstract. This block mustnot included elsewhere such as: Prepared in coop- be completed to assign a limitation to the abstract.eration with. , Trans. of...; To be published in-... Enter either UL (unlimited) or SAR (same as report)When a report is revised, include a statement An entry in this block is necessary if t he abstract iswhether the new report supersedes or supplements to be limited. If blank, the abstract is assumed tothe older report be unlimited.

Standard Form 298 Back (Rev 2 89)