Development of tuneable effects warheads

Mark REYNOLDS*，William HUNTINGTON-THRESHER

QinetiQ，F(xiàn)ort Halstead，Sevenoaks，Kent TN14 7BP，UK

Development of tuneable effects warheads

Mark REYNOLDS*，William HUNTINGTON-THRESHER

QinetiQ，F(xiàn)ort Halstead，Sevenoaks，Kent TN14 7BP，UK

The tuneable effects concept is aimed at achieving selectable blast and fragmentation output，to enable one charge to be used in different scenarios requiring different levels of blast and fragmentation lethality.It is a concept QinetiQ has been developing for an energetic f i ll consisting of three principal components arranged in co-axial layers，two explosive layers separated by a mitigating but reactive layer.The concept was originally designed to operate in two modes，a low output mode which only detonates the central core of high explosive and a high output mode which detonated both the central core and outer layer of the explosive.Two charge case designs where manufactured and tested；one of these designs showed a reduction in blast and fragment velocities of ～33%and ～20%，respectively，in the low output mode.

Tuneable effects；Multiple effects；Blast；Fragmentation；Hydrocode；Modelling；Trial

1.Introduction

The tuneable effects warhead concept is based on QinetiQ patented ［1］technology，previously explored using bare explosive charges ［2］.This work showed signif i cant differences in peak blast pressures between two detonation modes （35%）while maintaining quasi-static pressure.The study reported here looked at developing a metal cased variant with the aim to demonstrate a tuneable fragmentation output，whilst maintaining the demonstrated blast performance.

This next step for the concept was to test it in a more representative conf i guration generating fragments and blast.To ensure the exploitability and the relevance of the study，the warhead was designed to generate fragments with a lethal effect in the high mode.

2）Reactive，but non-detonable composition （aluminium powder loaded rubber）

3）Highly-aluminised explosive composition （RDX/Al/PB）Two modes are available:

·Mode 1 （lower incident pressure）-initiate central charge （1）only，

·Mode 2 （higher incident pressure）-initiate both charges （1）and （3）

2.Tuneable warhead concept

The concept consists of an energetic f i ll constructed from three principal components arranged in co-axial layers （Fig.1），namely:

1）High-performance High Explosive （HE）（HMX/PBPolymer Binder）

3.Case design

Both analytical codes and QinetiQ's Eulerian hydrocode GRIM were used to help develop two possible designs for the steel case.The cased designs were required to perforate a 5 mm steel target when operating in the high mode.Simulations were also used to predict the theoretical difference in case fragmentation between the low and high modes.Fragmentation，driven by external groove designs，was explored together with a more novel option combining a 3D printed plastic insert to initiate internal fracture circumferentially around the case，with axial external grooves.

The diameter and length of the charge were kept at the dimensions of the previous uncased study ［2］，the compositions oftheenergetic layerswere also keptnominally the same.

Fig.1.Tuneable effects charges.

4.Split-X and hydrocode modelling

The design study applied Split-X to calculate the required steel case thickness to perforate a 5 mm steel plate in the high mode.Split-X ［3］is an analytical code for the assessment of fragmenting warheads.Given the functionality of Split-X，the explosive，detonating in the mode，was modelled as a single cylinder with an inert surround.Given the available explosive mass，the assessment indicated that a case thickness of 10 mm with def i ned fragment sizes （approximately 10 mm cubes）was required to perforate the plate.The case steel selected was EN24 condition W.It was chosen based on the expected strength and ductility properties preferred for the case.

Hydrocode modelling with QinetiQ's Eulerian code GRIM was then applied to assess design options to control fragmentation and to assess the differences between the two detonation modes.The high mode was modelled with the detonation of both explosive components.The low mode was modelled with only the inner core of explosive detonating.The non-detonating components were modelled as inert throughout the timescales of detonation and initial fragment f l ight.

The typical arrangement for the GRIM hydrocode simulations is shown in Fig.2.The central core of PBXN110 was 35 mm in diameter，the annulus of HTPB-Al was 15 mm thick and the next annulus of PBXN109 was 15 mm thick.This gave a total explosive diameter of 95 mm.The length was 200 mm.

For constitutive models to describe metals，the physicallybased constitutive model due to Armstrong and Zerilli and modif i ed by Goldthorpe et al.［4］，Equation 1，is the preferred model used by QinetiQ.The Body Centred Cubic （BCC）form of the equation，relevant to the metals （Steel）in this study，is shown.

In this equation Y is the f l ow stress，T is the temperature，ε′is the strain rate，and εPis the plastic strain，with C1through C5，n and a1and a2，which describe the temperature dependence of the shear modulus，constants derived from the characterisation tests.

Fig.2.GRIM model setup left-low mode，right-high mode.

As part of the drive to develop a system of physically based material models，Goldthorpe developed a path dependent ductile failure model ［5］.The QinetiQ algorithm used in the code is Equation 2

In this equation S is the measure of ductile deforγmation/ damage，σnis the stress state （pressure/stress）andis the shear strain withAsderived from characterisation tests.The material fails when S reaches SF，which was also derived from characterisation tests.

The parameters for the EN24 W condition steel are listed in Table 1.

The polymer composite materials，aluminised HTPB and PBXN109，were both represented with tabular equations of state，and for PBXN109 the QinetiQ Porter-Gould constitutive model ［6］.Table 2 lists the parameters，and the initial moduli are provided for the dynamic regime of interest （i.e.in an unrelaxed condition）.

The three explosive materials were modelled using JWL（Jones，Wilkins and Lee）equations of state；the parameters applied are listed in Table 3.

It was acknowledged that a highly aluminised outer layer would have a lower brisance in comparison to non-aluminised compositions.The consequence was that case fracture wouldrequire careful design to introduce features to generate suff icient stress concentrations to promote cracking.

Table 1EN24 W condition constitutive data.

Table 2Inert polymer model data.

Given the potential issues to ensure reliable fragmentation from the lower brisance explosives deployed in high conf i nedblast performance warheads，there was a desire in this low TRL （Technology Readiness Level）project to explore nontraditional/new methods of controlling fragmentation and new/ lower cost manufacture methods.Methods that had the potential to fragment distinctly differently in the two modes were highly desirable.

Fig.3 shows the GRIM modelling predictions of the two designs in both modes；the external groove or “Helical”design and the plastic insert and groove combination or “Hybrid”design.The predictions indicated the potential for signif i cant differences in the size of the fragments between the two modes.

Table 3Explosive JWL properties ［7］.

Fig.3.Case breakup predictions.

Fig.4.Velocity predictions.

Fig.4 shows predictions from both Split-X and GRIM of the prof i le of fragment velocity along the length of the case.They both show a signif i cant difference in fragment velocity for the low mode.

5.Gap testing

To ensure the outer layer of explosive did not detonate，gap testing was used to guide the required thickness of the nondetonable composition.

Gap testing used cylindrical samples.The samples were arranged with a detonator on a set length of PBXN110，with the selected thickness of the reactive，but non-detonable composition and then 15 mm of a PBXN109 “mimic” explosive （a QinetiQ formulation）in contact with a 5 mm steel witness plate.

Based on the gap tests （Table 4）a barrier thickness was selected at a nominal 22 mm；this thickness proved to be greater than that used in the blast only charges，due to a variation in the composition of the outer explosive layer and the steel case.

The gap tests provided data to obtain an indicative shock level required to detonate the PBXN109 mimic （QRX-293-M6）.By modelling the gap test，the shock level in the explosive was observed.This level in turn was then used to assess the charge designs using 2D models.These models showed a potential issue since the peak shock level is enhanced when it hit the steel case and then was further enhanced when it combined atthe end of the charge with shock from the central explosive，F(xiàn)ig.5.This assessment was used to modify the design of the charge by increasing the thickness of the inert layer at the base of the charge.

Table 4Gap test results.

Fig.5.Four simulation times showing pressure localisation at the base of the charge.

6.Charge manufacture

The two charge designs chosen are detailed below and shown in Figs.6 and 7:

·Helical cased charge:

○10 mm thick case with 10 mm spaced 1/3 depth helical grooves

·Hybrid cased charge:

○3D printed plastic Buxton type liner

○10 mm thick case with 10 mm spaced 1/3 depth and four equally spaced 1/2 depth vertical grooves.

Fig.8 shows the high mode version of both designs at the top with a full diameter disc of sheet explosive and the low mode of both designs at the bottom with a small disc of sheet explosive designed to only detonate the central core of explosive. Examples of the assembled charges are shown in Fig.9；the right hand charge has the case painted black with a white grid applied to enable the case expansion to be calculated.

Fig.6.Charge cases-helical（left）-hybrid （right）.

7.Trial setup

The charges were detonated in the two modes at the MoD Pendine range.The trial setup had a 5 mm witness plate，four strawboard packs with velocity foils，four blast gauges and two Phantom high-speed cameras to capture the early case deformation and fragment f l ight.

Fig.10 shows a view of the trial setup showing the blast gauges，velocity foils，steel witness plate，and fragment packs.

Fig.7.Filled charges-helical cased （left）-hybrid cased （right）.

Fig.8.Detonation control with sheet explosive discs:high mode （top）-low mode （bottom）-helical case （left）-hybrid case （right）.

Fig.9.Assembled charges-helical case （left）-hybrid case （right）.

8.Helical cased charges

The early case expansion of the helical cased charges during/ following detonation is shown in Fig.11.This f i gure shows a Helical cased charge operating in the high mode on the left and operating in low mode on the right；with the time post the f i ring trigger shown in each frame.The shape of the expansion of the case looked different when the two modes were compared，with the low mode showing a more barrelled shape.

The last frame for the low mode （Firing 4 （F4））shows the start of possible case fragmentation near the top of the charge；similar to the middle frame on the high mode （F2）but with the possible start of fragmentation near the bottom of the charge instead.This appeared to suggest that the lower part of the outer layer of explosive was burning/def l agrating at a high or very high rate perhaps verging on detonating.

Fig.10.Trial setup.

Fig.11.Phantom images of helical cased charge expansion；high mode （left）-low mode （right）.

The helical cased charge formed strips in the high mode and in the f i rst low mode f i rings.The case cracks appeared to form independently from the helical grooves （Fig.12）.For the second low mode helical cased f i ring the charge conf i nement was modif i ed at the base to investigate/change a postulated reactive behaviour of the outer layer.This modif i cation appeared to result in little or no signif i cant difference to the fragment velocities or peak blast pressures.It did，however，show signif i cantdifferencesto thecasefragmentation. Although the case still split into strips，some of the strips were partially def i ned by the grooves （Fig.13）.This was most likely due to the difference in conf i nement at the base of the charge.

Analysis of the recovered fragments showed that the case had stretched considerably before fracturing.This indicated the case did not fracture/shatter due to the explosive shock/brisance and therefore experienced signif i cant stretching pre-fracture.

Fig.12.Helical case fragments split across grooves.

To promote early case fracture it is likely that deeper grooves would be required.Alternatively a change in the explosive element with a higher brisance explosive or an increased explosive content could be considered.The experimental data generated in the tests of both concepts can be used to validate an updated modelling methodology for this type of warhead.This updated capability can then be used to revise future case designs.

It was noted that the thickness of the middle layer，which was set to mitigate the shock from the inner layer，would remain approximately constant for larger charges.Thus a larger charge would be expected to exhibit a different fragmentation.In that sense it was recognised that the charges under test should be considered small rather than small scale.

9.Hybrid cased charge

The early case expansion of the Hybrid cased charges during/following detonation is shown in Fig.14.This f i gure shows a charge operating in the high mode on the left and in the low mode on the right，with the time post the f i ring trigger shown in each frame.The high mode charge showed a conical shaped expansion，whereas the low mode showed a more barrelled shape.The difference in early case expansion was an indication of the difference in explosive energy release rate.As was expected the low mode was shown to be more akin to a pressure burst，whereas the high mode showed a more typical conical shape with radial case displacement linked to detonation time.Given the low brisance of the outer layer and warhead geometry，snapshots of the early case shape were not expected to equate to a signif i cantly different fragment scatter between the modes.

Fig.13.Helical case fragments from the modif i ed low mode f i ring.

Fig.14.Phantom images of hybrid cased charge expansion；high mode （left）-low mode （right）.

Fig.15.Hybrid case fragments，high mode f i ring.

Fig.16.Hybrid case fragments，low mode f i ring.

The Hybrid cased charge split along the axial grooves producing large heavy fragments （Fig.15）.The Buxton liner in the high mode generated regular shallow cuts into the case，but the cuts did not appear deep enough to promote regular fracture across the strips def i ned by the grooves.Occasionally the strips had fractured at a location of a Buxton liner formed cut，but the timing of the fracture was unknown.

When the Hybrid cased charge operated in the low mode，the case split in a similar manner.The Buxton liner again generated regular small cuts into the case （Fig.16）.

10.Case expansion

The Phantom camera images of the charges allowed the observation of the case expansion at early times.Measurements were taken to calculate the case radius at the grid position marked on the case of the charges，this gave up to four radius values at each position along the case length.At later times，obscuration prevented some measurements and limited how many times were measurable.The case radius from the trial has been compared with a GRIM 2D simulation for the high modes of both charge designs.Fig.17 shows three times from the Hybrid high mode from round 7.The GRIM modelling results are plotted with a time offset from the experiment to take account of the detonator and pellet delay.This is assumed to be approximately 7 μs.The points from the experiment show some variation which is due to errors in measurement；however they show good agreement with the modelling.These data will also help validate the early expansion of the case for 3D modelling of the design.Low brisance explosives typically continue to accelerate fragments during the expansion of the explosive products；due to the early obscuration this assessment method would not be expected to yield a reliable measure of fragment velocity.

Fig.17.Round 7 Buxton high mode case expansion.

Table 5Fragment velocities and peak pressures.

Similar data for the low mode experiment might be used to help calibrate a model for the low mode using a method to account for the def l agration of the explosive.

11.Velocity and blast results

In total the trial consisted of 9 f i rings；the f i rst was a PE4 bare charge followed by the eight test charges.Table 5 shows a summary of the fragment velocities recorded by the velocity foils and the peak blast pressures recorded by the f i rst two blast gauges.

The Helical cased charges showed a small decrease in fragment velocities ～5%between modes but negligible differences in the peak pressures at the f i rst gauge location.

The Hybrid cased charges showed a decrease in fragment velocities and peak pressures with reductions of ～20%and～33%respectively.

The Helical and Hybrid cased charges showed similar peak pressures but differences in fragment velocities.These differences would be consistent with the differences in case fracture.

12.Conclusions

The Helical cased charge design showed only a small variation in fragment velocity between the modes and no discernible difference in the peak pressures.The grooves in the case had little if any effect on the case fractures.

The Hybrid cased charge design showed differences in fragment velocities and peak pressures between the two modes. However，the Buxton liner did not appear suff i cient to promote regular case fracture.

The earlier case fracture for the Hybrid cased charge appeared to be required to realise the differences between thehigh and low modes.This was evidenced by the thinner case fragments and the lower fragment masses from the Helical cased charges.

The realisation of lower than predicted differences between the modes in the experimental trials suggested that in the low mode，extra energy was released above that represented in the modelling.This suggested that while the barrier stopped propagation of the detonation，it did not stop ignition of the outer layer.

The reduction in peak pressures for the Hybrid design of～33%was very similar to the value observed in the previous bare charge study ［1］at ～35%.The similarity in the peak pressure reductions suggested that，when the case fracture is not delayed，both cased and uncased conf i gurations operate in the same manner.

The insight these cased tests provided through the effect on the fragment velocities was that the outer layer burned/ def l agrated in these designs.This indicated that an additional feature of this Tuneable Effects Warhead concept was that the barrier thickness could be potentially adjusted to vary the explosive output by designing a burn rate for the outer layer. Adjusting the mitigant layer would thus enable a much more varied output from the charge concept than the two modes originally envisaged.

The study has therefore demonstrated a tuneable warhead concept and increased understanding of its operation，plus it has indicated future developments.It has therefore helped demonstrate the art of the possible with tuneable effects warhead concepts.

13.Recommendations

A subsequent requirement is an evaluation of the operational requirement for tuneable effects warheads.This should be completed，considering the results of this study.

This warhead concept should then be further developed to deliver the user requirement identif i ed.

Acknowledgments

The authors would like to acknowledge the f i nancial support of the Anglo-French Materials and Components for Missiles，Innovation and Technology Partnership （MCM ITP）program jointly funded by UK MoD （Dstl）and DGA.

［1］Haskins P.Controllable output warhead，EP2564150A1，2011.

［2］Colclough E.A novel tuneable effects explosive charge，2012 Insensitive Munitions&EnergeticMaterialsTechnologySymposium，LasVegas，2012.

［3］N.N.SPLIT-X v5.2.10，An Expert System for Design of Fragmentation Warheads，NUMERICS GmbH，Petershausen，Germany，2005.User's Manual.

［4］Goldthorpe B，Butler A，Church P.A wide ranging constitutive model for BCC steels.Journal de Physique 1994；C8-471.

［5］Goldthorpe B.A path dependent model for ductile fracture.Journal de Physique 1997；7:C3-705.

［6］Cornish R，Porter D，Church P，Gould P，Andrews T，Proud B，et al.Elert M，F(xiàn)urnish MD，Chau R，Holmes NC，Nguyen J，editors.Comparison of Porter-Gould constitutive modelwith compression testdata for HTPB/SUGAR.Shock compression of condensed matter-2007.Melville，NY:American Institute of Physics；2007.p.777-80.

［7］Carleone J.Tactical Missile warheads，vol.155.1993.p.120-1 Table 19.

Received 7 October 2015；revised 20 January 2016；accepted 21 January 2016 Available online 24 March 2016

Peer review under responsibility of China Ordnance Society.

*Corresponding author.Tel.:+44 1959514911.

E-mail address:mwreynolds@qinetiq.com （M.REYNOLDS）.

http://dx.doi.org/10.1016/j.dt.2016.01.006

2214-9147/? 2016 China Ordnance Society.Production and hosting by Elsevier B.V.All rights reserved.

? 2016 China Ordnance Society.Production and hosting by Elsevier B.V.All rights reserved.