Effects of Al/O on pressure properties of confined explosion from aluminized explosives

Xiao-yu DUAN,Xue-yong GUO,Qing-jie JIAO,Jing-yuan ZHANG,Qing-ming ZHANG

Beijing Institute of Technology,Beijing,China

Effects of Al/O on pressure properties of confined explosion from aluminized explosives

Xiao-yu DUAN,Xue-yong GUO*,Qing-jie JIAO,Jing-yuan ZHANG,Qing-ming ZHANG

Beijing Institute of Technology,Beijing,China

Aluminized explosives Al/O Confined explosion Characteristic parameters

Pressure histories were tested in a 500-L chamber to identify the pressure load in confined explosion from aluminized explosives.Different aluminized explosives with Al/O,ranging from 0.25 to 1.23,were used.The recorded pressure curves could express the reflection of initial shock wave and the after burning combustion of aluminum.As there is no objective way to gain quasi-static pressure(PQS),method of multipoint averaging was used in smoothing the original pressure curves to gain thePQS.ThePQS,rising time of pressure(tQS)which stands for the duration of the initial reflected shock wave,and attenuation coefficient(ω)which stands for the supportive effects of the combustion of aluminum to thePQSare used to characterize the pressure load in the confined explosion from aluminized explosives.The research results showed that the Al/O significantly affected the three characteristic quantities.With the increase of Al/O,thePQSincreased at first and decreased later,gaining maximum at Al/O=0.99;thetQSsustained growth and theωdecreased at first and increased later,gaining minimum at Al/O=0.99.

1.Introduction

Confined explosions may occur for various reasons,such as an ammunition storage explosion,a charge explosionwithin a room in terrorist action,a warhead explosion following its penetration into a closed space and so on[1].The damage to structural element may be extremely severer than a similar external free-field explosion[2].The properties of confined explosion are apparently different with free field explosion[3].The restriction of structure causes the reflection of shock wave which will promote the after burning effect[4-6]and develops a rather long time constant residual quasistatic gas pressure.

The earliest literature on explosion in confined space was conducted by Weibull[7]in 1968,both he and Trott[8](1975)considered the mean pressure as the blast load acting on the walls because the quasi-static pressure had a long duration compared to the eigen-period of the structure,whereas the initial shock wave had a short duration,which should be ignored.The maximum gas pressure has been proven to be a function ofQ/V(kg/m3),which was unrelated to the vent[1,9].Swisdak(1975)[10]showed that the static pressure in confined explosion decayed along with time,which consisted of heat conductivity of the chamber.The quasistatic pressure was described as the subsequent pressure wave immediately after the initial shock[11],which reflected the accumulation of the residual pressure after the initial shock wave[12]and depended on charge properties and other factors[1].Dong[13]studied the interactive mechanism between the internal blast loadings and response of spherical containment vessels.Wu[14,15]considered that in a fully confined chamber,a constant static pressure exists;the quasi-static pressure took place sometime after the reaction of the explosive product,but how long the quasi-static pressure occurs is unclear.Weibull[1]and Kingery(1978)[16]regarded the backward pressure att=0 by the averaging of the pressure as a quasi-static pressure within the chamber.Anderson(1983)[17]stated that allowing some time to establish the maximum pressure was appropriate,chose the pressure some time after initiation as the quasi-static pressure,and only when the vent area being zero,the quasi-static pressure time was att=0.Trzcinski[18,19]gained the quasi-static pressure by fitting the average pressure curve.Dengwang[20]considered quasi-static pressure as the stable pressure after some time.Penggang[21]regarded the average of the pressure in the time interval of 10-20 ms as the quasi-static pressure,but at 10 ms,there still existed some reflected of the shock wave.

Even through Naval Surface Weapons Center used the quasi-static pressure as the only parameter to judge the explosion power in relatively confined spaces[22],whether it is appropriate to represent the pressure load for aluminized explosive in confined explosion maybe suit for discussion.Therefore,a general cognition of the reading rules of quasi-static pressure has notbeen formed.To represent the pressure acted on the wall in confined explosion,a reasonably objective way to gain quasi-static pressure is needed.

Furthermore,along with the widespread use of aluminized explosives,the use of quasi-static pressure as the only parameter that could represent all details in the confined structure may need discussion,as the longer reaction time of aluminum compared with that of high explosives[23].To identify the blast load in a confined explosion from aluminized explosives,six types of RDX-based aluminized explosives weigh 100 g whose Al/O(ratio of gram atoms of aluminum to gram atomsof oxygen)[24]ranged from0.25 to 1.23 were used to test the pressure histories in this study.The methods to identify the quasi-static pressure were raised and three characteristic quantities were used to depict the energy release of aluminized explosives.The energy release versus the aluminum content of the aluminum-containing explosives was studied.

2.Experimental materials

2.1.Samples

The tested samples were composed of phlegmatized RDX,aluminum powder,and a small amount of wax.The aluminum,which was purchased from Beijing Dk Nano S&T Ltd.(Beijing,China),had a median diameter of 50μm.The SEM image of the 50-μm Al particles,conducted using Hitachi S4700(Hitachi Ltd.,Tokyo,Japan),as well as the particle size distributions are shown in Fig.1.

The samples were prepared 100 g by pressing technique;the aspect ratios were 1.A mass of 20 g JH-14 was used as the primer charge and a standard fuse was used to initiate the charge.The compositions of mixtures are listed in Table 1.The sketch of the sample assembly is shown in Fig.2.

2.2.Chamber lay out and instrumentations

There is not relevant standard of the chamber in confined explosion and the 500 L special-steel fully confined chamber of Xi'an modern chemistry institute were used.The chamber consisted of a cylinder with a base with diameter of 900 mm and height of 800 mm,and a spherical crown.The pressure gauges were located in the pressure-transmitting tube on one side of the wall and 450 mm above the base.The schematic of the chamber is shown in Fig.3.

The chamber was filled with air at a normal pressure of 0.1 MPa.The sample was hanged in the center of the chamber which was 400 mm high above the bottom with nickel wire.

The sensor of Kistler 603-BQ-01(Kistler Instrument AG,Berne,Switzerland)was used.The sensitivity of the sensor was 0.5 V/MPa.The VXI data acquisition with frequency of 200 kHz/s(Chengdu Nanhui Science and Technology Co.Ltd.,Chengdu,China)was applied.The recorded time was 2.5 s.The room temperature was 15°C.

Table 1Compositions of samples.

3.Results and processing

3.1.Tested results

The piezoelectric pressure sensor was adopted,and the tested plane displayed the voltage as shownin Fig.4.The value of pressure is equal to the value of voltage divide by the sensitivity of the sensor.The pressure time curves for HL-02 are shown in Fig.5.

As shown in Figs.5 and 6,within the recorded interval,the pressure history consisted of three phases.The first stage of the pressure history presented these short duration-reflected waves of high pressures,as reported inprevious studies[7,25].It consisted of the initial high pressure,short duration-reflected wave,and several following reflected shocks.The second stage was the almost steady pressure wave with a slight noise but obvious decayed,which was almost a process of approximate exponential attenuation.The third phase was the slow decrease of pressure approximating a linear decrease.Fig.6 shows the tested pressure curves of the six samples.

As shownin Fig.6,all the pressure histories of the samples could also be divided into three categories.In the first phase,overlapping occurred because of severe oscillation,which lasted for almost 20 ms.Even through there are some differences from various samples,it is hard to distinguish because of overlapping.In the second phase,the pressure curves obviously differed mainly because of the properties of the samples.For this experiment,the main difference of the samples caused by the after burning of aluminum which support the decaying of shock wave[26].In article[27],the duration of the combustion of aluminum powder was about 400 ms.The end of the second phase was about 200 ms-300 ms which is connected the combustion of aluminum powder.In the third phase,as time goes on,the pressure and temperature decrease and where the reaction of aluminum could not continue and the third phase of the pressure was mainly decided by the thermal conduction of the system.In this study,the second phase was the main concern,including the beginning time and pressure and the decaying characteristic.

As shown,the Al/O affected the pressure curves apparently.And the pressures of HL-05 were the highest in the second and third phases,whereas those of HL-01 were the least.

3.2.Data processing

The pressure curves at the initial stage of shock were of severe oscillation,and the characteristic parameters were difficult to accurately obtain,as shown in Figs.5 and 7.At the same time,determining the quasi-static pressure is mostly through averaging,with a certain degree of subjectivity.The multipoint averaging method was used in this study.The principle of this method is averaging the adjacent appointedn,points and the pressure of No.iis calculated as follows:

wherePiis the originalvalue of No.iandPi′is the calculatedvalue of No.i.As the recorded time was 2.5 s and the frequency of the gauzes was 200 kHz,the maximumvalue ofiwas 5×105.The value ofndecides how smooth the curves are after processing.

Origin8 was used in this study to achieve the smoothness of the curves,and the adjacent averaging in signal processing module was used.The point of windows in that module stands fornin Eq.(1).In this study,the data processing of HL-02 was used as an example to introduce the method of multipoint average data processing.The peak pressures of the smooth curves with different numbers of windows,n,are listed in Table 2.

As shown in Table 2,with the increase ofn,the peak pressure of the curve gradually decreased and the decreasing scope became less whennwas more than 600 ms.

The smooth curves with window pointsnof 200,400,and 600 are shown in Fig.8.Compared with the oscillations shown in Fig.6,those in Fig.8 were fewer and not sharp,and the multipoint averaging has played a certain role.With the increase ofn,the curves became more slippery.A certain degree of shock is still obvious in the smoothing curves in Fig.7(a)and(b).

According to the definition of quasi-static pressure[8],the pressure curves,after reaching the quasi-static pressure,would not significantly vibrate.Choosingnbe in accordance with having no further vibration after quasi-static pressure,which manifests lower noise,as shown Fig.7(c).More noise exists after the peak as shown in Fig.7(a),and the noise is reduced as shown in Fig.7(b).In Fig.7(c),almostno noise is evidentafter thepeak.WhenTable 2 and Fig.8 are incorporated,nshould be set as 600 in this study.Other samples also perform well in 600 points of window.And in other case,the multipoint averaging method could be also used but with differentn.The smoothing curves of the tested samples are shown in Fig.8.The curves after processing could still express the features of the load on the structure,similar to what is shown in Fig.5.

Table 2Peak pressures of smooth curves with multi n.

ThePMof the smoothed curves was defined as quasi-static peak pressure,PQS,in this study and the corresponding time as the rising time of pressure,tQS.ThetQSstands for how fast could the pressure in the chamber achieve the approximate constant value of pressure which means the duration of the first reflected stage.

As Figs.6 and 8,there is a fast-falling phase after reachingPQSuntil about 200 ms-300 ms in both the original curves and the smoothing curves.The pressure curve in the third stage was connected with the thermal conduction of the system as article[5].Therefore,this paper studies the second phase,P1(t).This stage contact the combustion of aluminum,as the combustion duration of aluminum could reach 400 ms.This phase manifested that an approximation exponential decay process,as in Equation(2),may be determined by the combustion of aluminum.The attenuation coefficientωin this stage is chosen as a feature quantity which stands for the attenuation of thePQSin the second phase and within this study it may be connected with the reaction of aluminum in the samples.

whereP1(t)was the pressure in phase II,Pxwas the initial pressure of phase III,andtxwas the initial time of phase III.

To identify thetx,theωof different time intervals were fitted as shown in Table 3.

As shown in Table 3,the fitting correlation coefficients are all above 0.994,which means that the exponential decay is suitable.The value ofωdecreases along with the increases in time interval.The value ofωversus time interval is shown in Fig.9.

As shown in Fig.9,the value ofωwithin 250 ms sharplychanges and slightly varies after 250 ms.The trends ofωfrom the other samples were the same as HL-02.Theωin time interval of 250 ms was taken as the third feature quantity.

4.Results and discussions

4.1.Feature quantities

According to the data processing method in Section 3,three feature quantities exist:the quasi-static peak pressurePQS,risingtime of pressuretQS,and attenuation coefficient in time interval of 250 ms,ω,of each sample are listed in Table 4.

Table 3Attenuation rate of different times.

4.2.Relationship of PQSversus Al/O

The relationship ofPQSversus Al/O is shown in Fig.10.

As shown in Table 4 and Fig.10,PQSincreases first and decreases later along with the increase of Al/O,achieving a maximum when Al/O was 0.99.Article 26th given the static pressure from TNT which was a certain functions of W/V and the static pressures for other explosives should multiply by the complete combustion equivalency factors.That means that for the single W/V,the mainly determined factor for the static pressure is the equivalency combustion factors.For the situation of this study,the equivalency combustion factors determined the quasi static pressure lonely.That means the HL-05 released the max energy.

However,the explosion heats of samples were calculated by EXPLOS 5,shown in Fig.10,and the results showed that the explosion heat achieved a maximum in Al/O of 0.67 and decreases with the increase of Al/O.That indicated the energy release in confined explosion was not only the explosion energy,but also the after burning energy.There were some aluminum reacted with the oxygen in the chamber.ThePQSfor HL-06 did not increase right along HL-05.That means that not all of aluminum reacted,because HL-06 had the maxcomplete combustion heat.It accounted for that even through there were enough oxygen in the chamber(4.46 mol which is far more than that need to fully oxidize all combustible components in HL-06 under some condition),the aluminum in HL-06 did not fully combustion with the oxygen in the chamber.It may be caused by the fast attenuation of the pressure and temperature in the chamber of the second phase.With the use of several explosives and blasting tanks,the best formula to achieve maximum quasi-static peak pressure is HL-05.

Table 4Feature quantities of samples.

4.3.4.3Rising time of pressure tQS

The relationship oftQSversus Al/O is shown in Fig.11.

As shown in Fig.11,with increasing of Al/O,tQSincreases persistently.As the reaction time of aluminum powder was longer than high explosives,in the process of achieving quasi-static pressure there existed some combustion of aluminum,and the more aluminum cause the longer combustion time which directly leaded to more time to attain static.On the other hand,more aluminum in sample will cause the lower initial temperature resulted in the slower reaction of aluminum.As thetQSstands for the time needed to achieve thePQS,it indicated how long could the blast load act on the chamber and the tested results indicated that the lower Al/O caused shorter time toPQS.For confined explosion from aluminized explosives,Al/O affects not only the quasi-static pressure but also the rising time of pressuretQS.

4.4.Attenuation coeff i cientω

The relationship ofωversus Al/O is shown in Fig.12.

As shown in Fig.12,attenuation coefficientωdecreases at first and increases later with the increase of Al/O and reaches a minimum of Al/O 0.99.When Al/O is less than 0.99,the attenuation coefficient decreased along with the increase of Al/O.Then,ω increased when Al/O was 1.23.Excess aluminum powder did not attenuate the evolution of quasi-static pressure.Withω,the preferred aluminum powder explosion in a confined space-based RDX aluminized explosive content was 40%.

PQSandωstand for the value of quasi-stic pressure and it attenuation,and both of them preformed best in Al/O 0.99,which indicated that the combustion of aluminum in HL-05 was more prefect and released more energy in this situation of this study.That because on the one hand,more aluminum will decrease the initial temperature and make it difficult for the reaction of aluminum,on the other hand,excess aluminum will act with the oxygen in surrounding air and release more energy.At the same time,thetQSfor HL-05 was 41%more than HL-01.In actual use,maybe it is appropriate to combined the three feature quantities for realizing effective injure to various targets.

5.Conclusion

The tested pressure histories of RDX-based aluminized explosives in the confined explosion were affected by Al/O significantly.To identify the reading rule of quasi-static pressure,method of multipoint averaging was used.Three characteristic quantities were gained to investigate the pressure load which were also affected by Al/O.

Along with the increase of Al/O,PQSincreased at first and decreased later,and then reached its maximum when Al/O was 0.99.With the increase of Al/O,tQSpersistently increased.The attenuation coefficientωdecreased at first and increased later with the increase of aluminum content and reached its minimum at Al/O 0.99.

[1]Feldgun VR,Karinski YS,Edri I,et al.Prediction of the quasi-static pressure in confined explosions and its application to blast response simulation of flexible structures.Int J Impact Eng 2016;90:46-60.

[2]Grif fith H,Pugsley A,Saunders O.Report of the inquiry into the“collapse of flats at ronan point,canning town”.HMSO;1968.

[3]Hu HW,Song P,Zhao SX,et al.Progress in explosion in con fined space.Chin J Energ Mater 2013;21(4):539-46.

[4]Michael D,Andrews WS,Jaansalu KM.The fragmentation of met-al cylinders by thermobaric explosives.Kingston,CANADA:Royal Military College of Canada;2005.

[5]Kuhl A L,Oppenheim A K,Ferguson R E,et al.Effects of confinement on combustion of TNT explosion products in air[C].28th international symposium on combustion.Scotland;2000.p.1-16.

[6]Wolanski P,Gut Z,Trzcinski WA,et al.Visualization of turbulent combustion of TNT detonation products in a steel vessel.ShockWaves 2000;10:127-36.

[7]Weibull HRW.Pressures recorded in partially closed chambers at explosion of TNT charges.Ann N Y Acad Sci 1968;152:357-61.

[8]Trott BD,Backofen J,Joseph E,et al.Design of explosion blast containment vessels for explosive ordnance disposal unit.ADB 016707.1975.

[9]Esparze ED,Baker WE,Oldham GA.Blast pressure inside and outside suppressive structure.ADA025504.1975.

[10]Swisdak M M.Explosion effects and properties:part I-explosion effects in air[R].ADA018544.

[11]Lee KB,Lee KD,Kim JK,et al.Relationship between combustion heat and blast performance of aluminized explosives.36th International Annual Conference of ICT&32nd International Pyrotechnics Seminar.Karlsruhe:ICT;2005.

[12]Li HB,Jin PG,Yan JJ,Wang JJ.Calculation method of quasi-static pressure of blast waves in confined chamber.Initiat Pyrotech 2014;1:45-8.

[13]Dong Q,Li QM,Zheng JY.Interactive mechanisms between the internal blast loading and the dynamic elastic response of spherical containment vessels.Int J Impact Eng 2010;37(4):349-58.

[14]Wu CQ,Fattori G,Whittaker A,Oehlers DJ.Investigation of air-blast effects from spherical-and cylindrical-shaped charges.Int J Prot Struct 2010;1(3):345-62.

[15]Yi Hu,Cheng-qing Wu,Matthew Lukaszewicz,et al.Characteristics of confined blast loading in unvented structures.Aust Int J Prot Struct 2011;2(1):21-43.

[16]Kingery CN,Schumacher RN,Ewing WO.Internal pressures from explosions in suppressive structures.ARBRL-MR-02848.1978.

[17]Charles E,Anderson JR,Baker WE,et al.Quasi-static,duration,and impulse for explosions(eg.HE)in structures.Int J Rock Mech Min Sci 1983;25(6):455-64.

[18]Trzcinski WA,Paszula Jozef.Confined explosions of high explosives.J Tech Phys 2000;4(4):453-70.

[19]Trzcinski W,Stanisław Cudziło A,Paszula Jozef.Studies of free field and confined explosions of aluminium enriched RDX compositions.Propellants Explos Pyrotech 2007;32(6):502-8.

[20]Deng-wang Wang,De-zhi Zhang,Yan Li,et al.Experiment investigation on quasi static pressure in explosion containment vessels.Acta ArmamentarII 2012;33(12):1493-7.

[21]Jin Peng-gang,Guo Wei,Ren Song-tao,Wang Jianling,Gao Zan,LI Hongbin Research on TNT energy release characteristics in enclosed condition explosive materials.2014,2(43).

[22]Peng-gang Jin,Wei Guo,Jian-ling Wang,et al.Explosion pressure characteristics of TNT under closed condition.Chin J Energ Mater 2013;36(3):39-41.

[23]Zheng-qing Zhou,Jian-xin Nie,Zhuo-cheng Ou,et al.Effects of the aluminum content on the shock wave pressure and the acceleration ability of RDX-based aluminized explosives.J Appl Phys 2014;116:144906.

[24]Michael M.Swisdak.Explosion effects and properties:part II-explosion effects in water[AD].A018544.

[25]Waldemar A,Trzcinski,Cudziło Stanisław,Paszula Jozef.Studies of free field and confined explosions of aluminium enriched RDX compositions.Propellants Explos Pyrotech 2007;32(6):502-8.

[26]Jun-xiang Zhou,Guo-hui Yu,Peng Li,et al.Experimental study of the aluminized explosive RDX/Al explosion under water.Blasting 2005;22(2):4-6.10.

[27]Ju Huang,Bo-liang Wang,Qian Zhong,et al.A preliminary investigation on energy output structure of a thermobaric explosive.Explos shock wave 2012;02:164-8.

5 January 2017

in revised form 18 April 2017 Accepted 24 May 2017 Available online 27 May 2017

©2017 Published by Elsevier Ltd.This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

*Corresponding author.

E-mail address:nust@bit.edu.cn(X.-y.GUO).

Peer review under responsibility of China Ordnance Society.