H.N. Presles and P. Vidal- Detonation Generation and Propagation in Homogeneous Liquid Explosives

8/3/2019 H.N. Presles and P. Vidal- Detonation Generation and Propagation in Homogeneous Liquid Explosives

1/10

JOURNAL DE PHYSIQUE IVColloque C4, supplBment au Journal de Physique m, Volume 5,mai 1995Detonation Generation and Propagation in Homogeneous Liquid ExplosivesH.N. Presles and P. VidalLuboratoire de Combustion et de Ditonique, UPR 9028 du CNRS, ENSMA, BP. 109, 86960 Futuroscopecedex, France

Foreword

The goal of this paper is to present the main features of detonation generation andpropagation in liquid explosives. Most of the results are related to N itromethane (NM) since it isthe most studied liquid explosive.I - Introduction

Depending on whether they are homo geneous or heterogeneous, explosive media havespecific detonic properties which are related to different heat release mechanisms. Beforedescribing the ignition, propagation and extinction of a detona tion in a liquid explosive charge,it is useful to recall som e specific features of liquid explosives.Th e latter are homogeneous in norm al conditions but by add ing heterogeneities, such assolid particles or bubbles, they can exhibit solid explosives properties.

Hom ogeneous liquid explosives are transparent to the visible light so that the radiationsemitted by the detonation products can be recorded w hile the detonation wave propagates insidethe explosive charge.

Liquid explosive densities are of order 1 so that the detonation pressure is in therange 10to 20 GPa. With such a pressure level the com pressibily of the confinem ent has a greatinfluence on the shape of the detonation front and on the critical diameter.

Wh en liquid explosives are submitted to a shock or a detonation wave they exhibit aphenom enon called "electrical polarization effect". For instance, this effect is very useful tolook at the consecu tive events leading to the ignition of a detonation.

If cavitation is generated in the liquid explosive charge under precursor effectspropagating through the confinement, an other detonation propagation regime called "lowvelocity detonation" can be obtained. Fo r safety reasons, it is very important to understand thisparticular regime.

11 - Detonation ignition in liquid explosives submitted to a plane shock waveTh e processes involved in the detonation ignition in liquid e xplosives submitted to a

plane shock w ave have been clearly established in 196 1 by Ca mpb ell et al. /l/ after theArticle published online by EDP Sciences and available at http://dx.doi.org/10.1051/jp4:1995412
http://www.edpsciences.org/http://dx.doi.org/10.1051/jp4:1995412http://dx.doi.org/10.1051/jp4:1995412http://www.edpsciences.org/


2/10

C4-144 J O U R N A L DE P H Y S I Q U E IVpreliminary work of Chaiken 121. Using a n other experimen tal technique, Hardesty in 1976 I31confirmed all these processes. When a liquid explosive is submitted to a plane shock wave ofproper characteristics, a therma l explosion occurs after an induction time T Fig. l ) , eading to adetonation wav e which propagates into the compressed explosive at a velocity D* higher thanthe shock wave velocity U. W hen this detonation wave overcomes the shock wave it becomesan overdriven detonation propagating in the exp losive at rest, its velocity decreasing toward thenormal detonation velocityD.

Assuming that the chemical reaction follows an Arrhenius rate law, the induction timecan be approxim ated by the relation

E2 = K T2, ex p (m)

Tc is the shock temperature of the explosive and E the activation energy.In agreement with this relation, experimental results show a great dependance of the

induction delay on the initial temperature To of the liquid explosives (Tcdepends on To).Measuring T for different shock strenghts and calculating the corresponding liquid

temperature Tc, it has been shown I41 that the activation energy of NM compressed by a shockwave is about 25 kcallmole, that is half of the value for the gas-phase unimoleculardecomposition.

In most of the studies the diagnostic information during the subsequent events ofinitiation was obtained from a streak camera looking towards the compressed explosive throughthe shock wave. In some explosives, like NM, light from the detonation propagating in thecompressed explosive is weak but readily photographed . With so me other explosives, there is alack of luminosity. In 1965 TRAVIS 151 has shown that electrical signals are generated byinitiation processes occuring in dielectric liquid explosives filling a plane capacitor. A few yearsbefore, EICHELBERGER and HAUVER I61 and HARRIS 171 have reported studies on thecharge generation by sh ock in inert dielectric. They have considered the effect to be d ue tomechan ical polarization of the molecules produced a t the shock front.

As th e initial shock wave en ters the liquid explosive and propagates, it po larizes themolecules in a thin layer (Fig. 2). Then the polarization relaxes rapidly. Using the ALLISONtheory 181to analyse the electrical response of NM compressed by shock waves in the range 3 to8GPa, we have show n that the polarization relaxation time is of the order of 10 ns 191.

When detonation starts at the interface between the barrier and the explosive at time T, tpolarizes a second laye r of mo lecules which leads to a positive signal and w hen it catches theleading sho ck wave only on e polarized layer remains. A negative signal corresponds to thisevent. As the detonation wave gets closer to the second metallic plate the electrical pulse risesrapidly because of the rapidly dec reasing effective electrode spacing. TRA VIS has sho wn thatthe times of occurence of the initiation events measured by both the polarization and opticaltechniques are the same. Th is electrical technique is very useful to measure very sho rt inductiontime.


3/10

Mechanical o rientation of the dipoles submitted to shoc k wave is a possible cau se of theelectrical phenomena associated with shock or detonation propagation, but it has not beenproved and s om e more fundamental researches should be started to really understand it.I11 - Detonation propagation

Ma ny exp eriments have shown that the relationship between explosive charge size andsteady state detonation velocity is qua litatively different f or hom ogeneous and heterogeneousexplosives.

Fo r both kind of explosives, th e detonation pressure is so high that part of the chem icalenergy is used to produce lateral material motion. As a consequence, the detonation front iscurved and its velocity decreases with the charge diameter down to a minimum value called thefailure or the critical diame ter, below which no steady wave can propagate. T he correspondingcharge diameter is termed critical diameter.

The detonation velocity of liquid explosives decreases linearly with respect to thereciprocal of the charge diame ter (Fig. 3) and is nearly inde pendant of the confinem ent nature.Th e detonation velocity decrem ent between infinite and critical diameter is very sm all (about1%). The corresponding curve for heterogeneous explosives shows a strong downwardconcavity, a larger velocity decrement and a strong sensitivity to the co nfinement nature.

Th is is the case with heterogeneous m ixtures based on NM and containing beads I101 orglass micro-balloons (GMB) (Fig. 3 ) /l l/.Following CA MP BE LL and EN GELK E 1121, the differences between the curves forhomo geneous liquid a nd heterogeneous ex plosives is due to different mech anisms supp ortingwave propagation.

Homogeneous burn is the mechanism involved in the detonation of homogeneousexplosives .~hilen the case of heterogeneous ex plosives it is re-inforced by "hot spots". Th elack of con cave part in the homogeneous ex plosive s curv e should be du e to th. ;ack of the hotspot mechanism.

The influence of chemical sensitization of NM on its steady detonation propagation hasalso been studied by EN GE LK E 1131. H e has shown that very sm all amount of amines (DETA)produces a significant reduction of the NM critical diameter. For instance, adding 0.03 weightper cent of DET A to NM reduces the critical diame ter by an amount of 40%. A so small amountof chem ical impurity has no influence on the initial density nor on the specific heat of reaction ofpure NM so that only the chem ical kinetics is altered. From detonation velocity meas uremen tsof mixtures containing 0.03 weight per cent of DETA, ENGELKE deduced that the one-dimensional reaction zone length of the detonation in this mixture is 80% of that in pure NM.

T he sh ape of the detonation front can also be used for obtaining information about thechem ical kinetics because it is determined by the interaction between fluid mecha nics and theheat release rate inside the reaction zone


4/10

JOURNAL DE PHYSIQUE IV

BDZ IL et a1 1141, using exp erimental measurements of detonation wav e shap e andvelocity in combination with a quasi-one dimensional theory of the processes inside the reactionzone, have evaluated the parameters of the Arrhenius law for NM. They got a rather high valueof the activation energy (92k23 kcallmole) in co mparison with that ob tained from initiationexperiments.

Abo ut the detonation velocity of NM-DETA mixtures, WALK ER I151 found, with one-dimensional experiments, an increase with DETA up to 0.1 weight per cent. Measuring thedetonation velocity and temperature of theses mixtures confined in brass tubes, we observedthat these characteristics are continuously decreasing while the DETA concentration i sincreasing. S o these results obtained with two-dimensional axisymetric detonation do no tconfirm those of WALKE R.I V - T h r e e - d i me ns i ona l de t ona t i on f r on t s t r uc t u r e

The detonation front in som e liquid explosives such as NM is not smooth 1151 and streakcamera records of the emitted light show s a nonun iform burning of the explosive throughout thecross section. O ne can d educe that detonation in NM is unstable. To discern the details of thedetonation structure it is usual to decrease the specific energy of the explosive by diluting it withsome non-explosive molecule. For instance NM is usually diluted with ac etone 1171 1181 1191 .Using such a m ixture, we can observe that the detonation wave structure behaves sim ilarly as ingaseous explosive w ith respect to d ilution with a n inert ad ditive, to chem ical sensitization and todetonation overdriven degree.

For exam ple as shown on fig. 4 adding a small amount of DETA to a NM -acetonemixture reduces greatly the size of the detonation front structure 1201. Thus, as in gaseousexplosives, the three-dimensional structure is very sensititive to the chemical kinetics.

From these observations, chemical decomposition processes in homogeneous liquid andgaseous explosives look similar.V - Detona t ion fa i lu re and c r i t i ca l d iamete r

Looking with framing or streak camera at the detonation front propagating in liquidexplosive charge with size near the critical one, one can observe non lumino us areas spreadinginward from the confining wall 1211.

Because these ar eas look lik e reaction qu enching w aves they are called failure waves1221. They appear to be characteristic of homogeneous explosives. When the flow expansion istoo large compared to the rate of chemic al energy release, a reaction failure wave travels alongthe detonation fron t. Quasi-steady, quasi-one dimensional theoretical modellings 1261 associate acritical shock cu rvature and a critical velocity to this even t.Behind failure w aves, there is com pressed ex plosive which can thermally e xplode andinitiate again a detonation.


5/10

As the detonation process in charge size near the critical one is unsteady and the startingof failure wave random, the best way to look at the failure mechanism is to subm it a steadydetonation to a sudden confinement enlargement 1171(Fig. 5).

When the detonation passes over the corner of the confinement a failure wave takesplace at the periphery and propag ates symmetrically toward the axis of the charge at a constantvelocity.

According to the criterion of DRE MIN and TROF IMO V I231 the critical diameter of theunconfined ch arge is defined by the tube diam eter @ for which the re-initiated detonation in thecomp ressed explosive overtakes the failure wave exac tly on the axis.

Experiments sh ow that the critical diame ter of liquid explosives is very sen sitive to theconfinement nature (and also to the confinement thickness when this thickness is sufficientlysmall 1274 and to the initial temperature. For instance, the critical diameter of NM confined in asteel tube is around 2 mm and 14 mm in glass. It decreases as the initial tempera ture increaseswith a rate of abou t0,43mrn/C 1241.

Using the model of DREM IN and TROFIMOV , ENIG and PETRONE I251 have shownthat this can be explained by the fact the chemical kinetics is of an A rrhenius form.V1 - Conclus ion

Du e to different heat release mechanisms, liquid homogeneous explosives present som especific detonic properties in comparison with he terogeneous liquid or solid ones.A s there is no possibility to look straight to the chem ical kinetics inside the detonationwave of con densed explosives, researches have to be undertaken to solv e this problem.

One possible way is through inverse methods based for instance on realistic two-dimensional detonation mode lling comb ined with detonation wave sh ape measuremen ts.

Bu t as instabilities h ave been observed in the detonation of som e liquid explosives, thenext step would be to use a three-dimensional approach in order to understand the realdetonation and to determine for instance if there is some similarity between detonation waves ingaseous mixtures and in homogeneous liquid explosives.

V11 - References/l/ CA MP BE LL A.W., DAV IS W.C., TRAV IS J.R., Shock initiation of detonation in

liquid explosives, 3 rd Symp. on Detonation, ACR 52,469-498, 1960.121 CHAIKEN R.F., Kinetics theory of detonation of high explosives, Master's thesis,

Polytechnic Institute of Brooklyn, 1958.


6/10

(2-148 JOURNAL D E P H Y SI QU E IV

131 HARDESTY D.R., An investigation of the shock initiation of liquid NM, Comb. andFlame, 27, 229-251, 1976.

I41 CHAIKEN R.F., Correlation of shock pressure, shock temperature and detonationinduction time in NM , Symp. HD P, 41-53,1978.

151 TR AV IS J.R., Electrical transducer studie s of initiation of liquid explosives, 4b Symp.on Detonation, AC R 126, 609-615, 1965.

161 EICHEL BERG ER R.J., HA UV ER G.E., Solid state transducers for recording ofintense pressure pulses, Congrks sur les Ondes de dktonation, CNRS Paris, 363-381,1962.

l71 HA RR IS P., Mec hanism for the shock polarization in dielectrics, J.App1. Phys. 36,739-741, 1965.

181 ALLISON F.E., Shock induced polarization in plastics, J. Appl. Phys., 36, 7,2111-2112, 1965.

I91 D E I C A ZA - H ER R ER A M ., PRESLES H.N., BROCHET C., Polarisat ion dunitromethane sous choc , Revue de Physique A ppliquCe, 547-553, 1978.

1101 ENG ELK E R., E ffect of the number den sity of heterogeneities on the critical diameter ofcondensed explosives, Phys. Fluids, 26, 9, 2420-2424, 1983.

1111 H.N. PRESLE S, P. VIDAL , J.C. GO IS, B.A. KH ASAINO V, B.S. ERM OL AE V,"Influence of glass micro bdlon s s ize on the detonation of NM" bases mixtures, acceptedfor publication in Shock W ave Journal.

IlY CAM PBE LL A.W., ENG ELK E R., Th e diameter effect in high density heterogeneousexplosives, 6mSymp. on Detonation, A CR 221, 642-652, 1976.

1131 ENG ELK E M., Effect of chemical inhomogeneity on steady-state detonation velocity,Phys. Fluids, 23, 5, 875-880, 1980.

1141 BD ZIL J.B., E NG EL KE R., CHR ISTE NSO N D.A., Kinetics study of a condense ddetonating explosive, J. Chem. Phys., 74, 10,5694-5699, 1981.

I151 W AL KE R F.E., Initiation and detonation studies in sensitized NM , Astronautica Acta,6, p.807 ..., 1979.


7/10

ZELDO VICH Y.B., KOR MER S.B., KRISKE VICH G.V., Y US HK O K.B., Dokl.Akad. Navk. SSSR, 171, 67, 1965.DREMIN A.N., ROZANOV O.K., TROFIMOV V.S., On the detonation of NM,Comb. and Flame, 7, 153-162, 1963.MALLOR Y H.D., GREENE G.A., Lunimosity and pressure aberrations in detonatingNM solutions, J. of Applied Physics, 40, 12,4933-4938, 1969.UR TIE W P.A., KUS UBO V A.S., D U FF R.E., Cellular structure of detonation in NM,Comb. and Flame, 14, 117-122, 1970.PRESLES H.N., Nouveau critkre d'Ctude de la sensibilisation d'explosifs liquides,CRAS, 314, 11, 575-578, 1992.CA MP BEL L A.W., M ALIN M.E., HOLLAND T.E., D etonation i n homogenousexplosive, 2"d Symp. on Detonation, 454-477,1955.CO TT ER T.P., Th e structure of detonation in some liquid exp losives, Dissertation,Cornell University, 1953.DREMIN A.N., TROFIMOV V.S., On the nature of the critical diameter, 10h Symp.Int. o n Combustion, 839-843, 1965.CA MP BEL L A.W., M ALIN M.E., HOLLA ND T.E., Temp erature effect in the liquidexplosive NM, Journal of Applied Physics, 27,6,963, 1954.EN IG J.W., PET RO NE F.J., The failure diameter theory of Dremin, 5~ Symp. Int. onDetonation, AC R 184,99-104, 1970.STEWART D.S., BDZIL J.B., The shock dynamics of stable multi-dimensionaldetonation, Com b. and Flame, 72,311-323, 1988.FOR BES J.W, LEMA R E.R., BAKE R R.N., Detonation wave propagation in PBX W-115, IX h Symp. (Int.) on Detonation, OCNR 113291-7, I, 806-815, 1990.


8/10

JOURNAL DE PHYSIQUE IV

Fig. 1. Space-time representation of the events leading to the generation of adetonation in a homogeneous liquid explosive submitted to a plane shock wave

0 Z t time

Fig. 2. Electrical signals associated with detonation generation.At t= 0, the shock wave enters the nitromethane.7 and t have the same meaning as on Fig. 1.


9/10

Fig. 3. Detonation velocity with respect to the reciprocal chargediameter for NM and for a heterogeneous mixture made of NM and2% (mass fraction) of glass micro-balloons (GMB) confinedin PVC and steel tubes

radiation emitted bythe detonation wave radiation emitted b>-the shocked air 1Fig. 4. Streak camera records of the radiation emitted by thedetonation wave front propagating in NM-acetone mixture (upperrecord) and in the same mixture sensitized with DETA flower record)


10/10

JOURNALDE PHYSIQUE IV

Steel confinement

Fig. 5. Streak camera record of a detonation wave submitted to a confinementdiscontinuity :A conical failure wave moving with constant tranverse velocityreduces the size of the detonation front. Then re-initiation occurs.

Documents

H.N. Presles and P. Vidal- Detonation Generation and Propagation in Homogeneous Liquid Explosives