Science Engineering Technology at AWE Discovery25 September 2014 This issue: Explosive Materials Development Micromechanical Modelling Personal Protective Equipment Testing Slapper Detonator Modelling Hazard and Safety Studies AWE’s Outreach Discovery Contents Overview2 Explosive Materials Development 4 Micromechanical Modelling 10 Personal Protective Equipment Testing 14 Slapper Detonator Modelling 22 Hazard and Safety Studies 28 AWE’s Outreach 34 Science Engineering Technology at AWE Discovery25 September 2014 This issue: Explosive Materials Development Micromechanical Modelling Personal Protective Equipment Testing Slapper Detonator Modelling Hazard and Safety Studies AWE’s Outreach Cover image: Explosive microstructure representation; colours depict stress contours on particles (courtesy of Imperial College London). 25 I am delighted to be introducing the 25th issue of Discovery which is dedicated to explosives research at AWE. AWE plays a major part in materials science and I have enjoyed seeing some of the pioneering research that is undertaken as part of my role in supporting the long established AWE Cranfield University Strategic Alliance. AWE’s collaboration with Cranfield has an enviable history, an alliance of which I am proud and one that has evolved and developed steadily over the last decade. The collaboration between AWE and Cranfield University brings mutual benefits through educational programmes and the delivery of research and consultancy services. I had the pleasure of attending the AWE Cranfield Technical Showcase in October 2013 at AWE Aldermaston and was impressed by the extent of the collaborative work being undertaken. Over 100 AWE and Cranfield University scientists were present, clearly using the opportunity to improve knowledge and understanding of activities under the Strategic Alliance and their value both to AWE and Cranfield University. In addition, four new areas of future collaboration were identified. This further endorses the strong relationship between the two parties allowing continued access to unique facilities at the Cranfield and Shrivenham campuses. that AWE does working with UK academia. I also hope the articles give you an appreciation of the esteem in which AWE’s researchers and scientists are held and their contribution to the UK deterrence programme, so vital to protecting the UK and keeping us all safe. This gives me great confidence in the future of the Alliance and all the people at AWE and Cranfield University involved that make it work through their expertise, commitment and dedication. I would particularly like to thank my Cranfield University colleagues Professor Ian Wallace (Head of Cranfield Defence and Security), Professor Dimitris Drikakis and Professor Rade Vignjevic (both AWE William Penney Fellows) for their continued and excellent contribution to the collaboration. I hope you enjoy reading the scientific articles and get a sense of the range of explosive science 1 Professor Jacqueline Akhavan Head of the Centre for Defence Chemistry, Cranfield University Discovery Discovery25 • The Science and Technology Journal of AWE Overview AWE's interest in explosives is very broad and covers the formulation of new materials for future applications through to the manufacture, in-service support and ultimately breakdown and disposal of explosive components. AWE is not only focused on explosive components but provides assessments, backed up with experimental and modelling data, to support process and facility safety cases for its operations. Such a wide range of activities presents a number of technical challenges for the explosives community at AWE and the articles contained in this edition touch on some of those key areas. High explosives are compact sources of chemical energy that function by decomposition to produce gaseous products at very high temperatures and pressures. It is the rapid expansion of these gases that produces the blast effects that are characteristic of explosives. High explosives can explode in a matter of microseconds and it is this very rapid release of energy that causes damage. In order to use the energy created from a detonating explosive in an efficient manner, it is necessary to form the explosive into an appropriate shape. Unfortunately, many high explosives exist as crystalline powders. In the Second World War, ‘melt cast’ explosives were used widely. High powered, high melting explosives such as RDX or HMX were dispersed in molten TNT (trinitrotoluene) at lower temperature and then poured into a mould, cooled and solidified into a desired shape. 2 Technology has since moved forward and has resulted in a type of explosive formulation known as a Polymer Bonded Explosive (PBX). The advantages of PBXs over melt cast explosives are that a high power explosive can be incorporated into the formulation, resulting in higher energy output per unit volume, and better safety properties are noted; this is due largely to the fact that a protective layer of non-explosive binder prevents frictional contact between crystals if subjected to external stimuli. This balance between explosive performance and safety is critical for modern explosives where volume constraints are experienced. AWE re-established a formulations team in circa 2006 to develop new PBX formulations for future applications. The Explosives Material Development article describes ongoing activities of this team and the tools and techniques developed to support development in this area. results of these trials are described in the PPE Testing article. One of the challenges to the explosive synthetic chemist is to try and design molecules that have high explosive performance in a safe manner. Trying to predict the properties of new explosive materials is therefore difficult. At AWE we always start new material development at a very small scale to protect the operator and scale up in a staged process, carrying out safety testing as we progress. Once explosive composites have been developed they must be subjected to a suite of tests to understand their engineering, physics, performance and safety through life. This testing is essential to support material and system qualification activities and a large range of tests are carried out at AWE and at offsite explosive range facilities. The Hazard and Safety article provides an overview of the testing undertaken to determine if an explosive is safe to use. Recently we have undertaken a number of explosive trials on various items of Personal Protective Equipment (PPE) to provide further protection to the operator in the unlikely event of an abnormal occurrence during these high hazard experiments. The In the Comprehensive Test Ban Treaty (CTBT) era a move towards a model-based approach to understanding system performance is becoming increasingly important. The modelling requirements for explosive initiation trains are multi-facetted. The Slapper 3 Detonator and Micromechanical Modelling articles are examples of this model based approach. AWE has undertaken to invest significantly in its explosive facilities. The strategy is to refurbish current facilities and for the long term, a series of new builds is planned that will provide enduring UK capability. One example of this is the High Explosives Fabrication Facility (HEFF), that will rationalise a number of production facilities, spread across AWE, under one roof. This edition of Discovery provides an overview of some of the work undertaken at AWE related to explosives. Discovery Discovery25 • The Science Engineering and Technology Journal of AWE Explosive Materials Development The majority of High Explosive (HE) materials used in modern munitions are Polymer Bonded Explosive (PBX) composites. These consist of explosive particles (filler) which are bound together by a polymeric binder. PBX material properties such as performance, safety, processing and ageing can be targeted by purposeful selection and formulation of the constituents [1]. This includes the physical properties of the explosive particle filler, the polymeric binder component and density of the finished component. binder undergoing tensile testing; Figure 2 shows the results of binder testing. This is the principal reason for carrying out development work on the binder component before any composite formulation work is started. Figure 1 is a picture of a Energetic binders in explosive compositions have been the subject of much research over recent years. Conventional polymer binders can offer well characterised physical attributes to the formulation; energetic polymer binders can also contribute to the overall energy release. The benefit is enhanced safety of the formulation by trading the explosive filler with energetic binder whilst delivering FIGURE 1 FIGURE 2 Modulus/Failure Stress (MPa) Modulus Explosive particle characterisation and modification consists of the physical analysis and processing of the raw explosive materials. This aims to produce and characterise particles that are optimised in terms of particle size, shape and distribution for the chosen application. Traditional methods of producing explosive particles, such as crystallisation and milling, may be used to produce a wide distribution of particle sizes. Whilst the particle size and chemical content are more easily controlled, the morphology and density of the particle produced are strongly influenced by the preferred crystal habit and the method of producing particles. Crystallisation and milling can produce angular particles that are complex in shape. Failure Stress (MPa) Failure Strain (%) 1000 2.5 900 800 2 700 600 1.5 500 400 1 300 200 0.5 Failure Strain (%) The choice of polymer binder component is important as it has the potential to offer increased explosive performance over traditional melt-cast and wax bonded composites in terms of energy, higher melting temperatures, improved mechanical strength and ageing properties. the same overall explosive performance. 100 0 0 A B C D E F Isocyanate Samples Effect of different isocyanate samples on the mechanical properties Image of a binder under tensile test. of a Polyurethane using the same base polyol. 5 Discovery Discovery25 • The Science and Technology Journal of AWE The relationship between the morphology of HE particles and the inherent safety and performance properties has been widely studied. Recent work has shown that a significant improvement in the shock sensitivity of two widely used explosives, HMX and RDX, can be made by tailoring the chemical and physical properties of the HE crystals [2, 3]. It was suggested that shock sensitivity is related to sharp edges, surface cracks, crystal shape as well as inclusions or impurities inside crystals and the crystal size. Reduced sensitivity HMX, for example, is now available commercially and as shown in Figure 3, has a more rounded nature and contains less internal porosity. Spherical particles produced with a smooth surface are ideal from a powder handling and ageing perspective. They have the minimum surface area to mass ratio. A spherical shape also provides scope for designing and tailoring the particle packing properties and surface area for any given application i.e. as a high or low density filler. The production of reduced sensitivity HE materials and a desire to gain more control over the shape and morphology of HE particles is driving research at AWE and academic outreach partners. Recent work with Cranfield University has been undertaken to process the explosive HNS (2, 2’, 4, 4’, 6, 6’-Hexanitrostilbene) into spherical particles. The method has proven to be a reliable and controlled means of producing a range of fine particles in narrow size distributions. The challenge was to establish the optimum conditions for control over the particle sizes whilst maintaining a spherical shape. The use of the spherical HNS particles in its desired application requires them to be in a compacted and geometrically confined state. HNS has been traditionally processed by wet chemistry techniques and the particles were highly irregular in nature, as shown in Figure 4, and were not ideal for pressing to high densities. FIGURE 3 Figure 5 shows the surface of a high density pellet pressed from spherical HNS particles. HNS spherical particles press readily to a high density and the compacted solids produced are robust enough to be handled. This technique is being developed and applied to other explosives including PETN (pentaery thritol tetranitrate) and HMX. Reduced sensitivity HMX (top), and HMX type A (bottom). 6 The science and technology of PBX formulation involves the careful combination and processing of the explosive filler with polymer binders and other ingredients. The result is a precursor for manufacturing; explosive moulding powders contain the essential ingredients of the formulation prepared into low density granules and are ready for pressing and machining into the finished PBX components. FIGURE 4 Spherical HNS particles (left) compared to HNS type IV (right) (courtesy of It is possible to control the product quality in terms of size, distribution and strength of the granule in the moulding powder to favour manufacture. This can be achieved through selection of the mixing technology and the right operating conditions appropriate to the product required. Examples of typical processing techniques available include: • Solvent-paste mixing • Water-slurry (PBX process) Cranfield University). •Emulsion/coating • Resonant Acoustic mixing • Extrusion (paste explosives) Newer mixing technologies such as acoustic mixers (Figure 6) are attractive as they use non invasive, low-frequency, high-intensity sound energy that is capable of mixing high viscosity materials. They do not possess many of the FIGURE 5 hazards associated with traditional mixing technologies, particularly those that may be attributed to shear of the HE from the mixing blades. The development of the mixing process involves stages of scalingup and maturing the processing steps by a controlled increase in the mass of explosive material. This ensures consistency in safety, mechanical and explosive performance properties at each FIGURE 6 Surface of die pressed pellet produced from spherical HNS (courtesy of Resonant acoustic mixer. Cranfield University). 7 Discovery Discovery25 • The Science and Technology Journal of AWE stage from laboratory scale, to pilot plant and finally to manufacturing plant. Figure 7 shows a test setup during manufacture and Figure 8 shows some of the equipment that is used in the scale up manufacturing. FIGURE 7 It is important to produce pressed compacts that are close to the theoretical maximum density (TMD) of that composition in order to maintain performance [4] and at the same time strive towards reducing the sensitiveness of the composition. An example of a test carried out on new explosive compositions to measure detonation velocity. The development of such highly filled compositions that are pressed close to their TMD is a relatively unique situation when compared to other industries. FIGURE 8 1 litre (left) and 25 litre (right) planetary mixer used to scale-up explosive compositions using the solvent paste mixing process. 8 FIGURE 9 20 mm Small particles coating larger particles. The choice of average particle size and particle size distribution used in a composition will affect the formation of granules in the moulding powder, as shown in Figure 9 and hence the processing behaviour of the resulting powder and so are essential when optimising the density of the final pressing. Summary The challenge to produce well characterised explosive particles, to formulate them into tailored compositions and eventually manufacture into components intended for a desired purpose involves the use of a wide variety of techniques, utilising some established processes and some more state-of-the-art concepts. Only by developing an understanding of raw materials, processing and its influence on final product throughout its service life, will total control be possible. Optimisation of the particle modification processes for easier processing and targeted properties, such as producing specific sizes of spherical HNS particles, has been achieved. Particle characterisation techniques are available to confirm the desired effect. Formulation techniques aim to produce explosives with the appropriate power, safety and mechanical properties to satisfy the requirements of the application. It also supports the manufacturing process to produce homogeneous mixtures with granule properties that make subsequent pressing and machining processes more efficient. 9 References [1] J. Akhavan, The Chemistry of Explosives, The Royal Society of Chemistry, 2004. [2] Ruth M. Doherty, Relationship Between RDX Properties and Sensitivity, Propellants, Explosives, Pyrotechnics 33, No. 1 (2008). [3] RDX and HMX with Reduced Sensitivity Towards Shock Initiation –RS-RDX & RS-HMX, Øyvind Hammer Johansen, Jørn Digre Kristiansen, Richard Gjersøe, Alf Berg, Terje Halvorsen, Kjell-Tore Smith, Propellants, Explosives, Pyrotechnics 33, No. 1 (2008). [4] E. C. Abdullah, A. M. Salam and A. R. Aziz, Cohesiveness and Flowability Properties of Silica Gel Powder, Physics International, vol. 1, no. 16, 2010. Discovery Discovery25 • The Science Engineering and Technology Journal of AWE Micromechanical Modelling The micromechanical model is a material science finite element analysis (FEA) model that has been developed through a strategic partnership between Imperial College London, University of Cambridge and AWE. It is a predictive capability to support the understanding of material safety, particularly relating to the mechanical integrity of energetic materials. Polymer Bonded Explosives (PBX) are energetic composites that are designed to meet safety and performance requirements. Not only are they exploited for their explosive power but their role is as an engineering component when subjected to in situ stresses and strains during the manufacturingto-disassembly lifecycle. The explosive and engineering safety of PBXs depends on acceptable strength to resist sudden cracking or damage. A typical PBX comprises of stiff crystalline explosive particulates bonded within a polymer matrix, called the binder, see Figure 1. The rigid particles reinforce and stiffen by imparting rigidity to the composite and in turn limiting the amount of stretch before failure occurs. The required mechanical strength of the composite can be tailored by varying the type of materials and ratios of material ingredients. In order to achieve target material mechanical properties, a cycle of trial and error that can result in the need to produce numerous formulations. This repetition raises safety issues regarding the increased exposure, handling and cost. FIGURE 1 Modelling Capability The micromechanical model capability was developed to support the explosive formulations effort at AWE. The aim is to predict a composite’s mechanical response from properties of the constituents and their interactions; reducing the need for experimentation and improving safety. The concept behind a simulation of PBX composites is essentially to represent the microstructure (i.e. < 150 μm in length) as a representative volume element (RVE) to simulate the response of a continuous material. The particles are assigned properties attributable to that of the explosive crystal and the material between boundaries as binder properties. Stresses and strains (tensile or compressive) can be applied to the boundaries and using finite element modelling methodology the resulting cracks or weak areas can be realised. There are two input methods to determine the geometry of the RVE, either a simulated structure or representative structure in the form of an image. Cracked explosive microstructure (courtesy of University of Cambridge). 11 Representative images of PBXs are produced by Scanning Electron Microscopy (SEM). However, by using this method not all of the particles are accounted for. The very fine particles are not always captured using the microscopy methods; additional detail from the images is lost in the binary image conversion. The representative structure results in an approximate 60% packing Discovery Discovery25 • The Science and Technology Journal of AWE particles are normally assumed to be simply elastic, as they are orders of magnitude less compliant than the associated binder material. FIGURE 2 The University of Cambridge has provided AWE support on explosive safety for a number of years through bespoke experimentation. One success has led to characterisation of the adhesive interaction between the binder and fill particles. Scanned (left) and simulated (right) RVE (courtesy of Imperial College London). fraction, which is significantly less than the intentional packing fraction in PBX materials (90% ‒ 95%). An alternative is to simulate the structure using particle packing software; in this case a higher packing fraction can be simulated at the cost of idealised particle shapes. Figure 2 shows the difference between scanned and simulated RVE. Once an RVE structure is created and meshed, various material models are applied that describe the individual constituent material properties and their interactions. These material and interaction models are derived through bespoke methods from our strategic partners Imperial College London and University of Cambridge. A number of binder constitutive rheological models have been developed by Imperial College London. Such material models are often complex and are dependent on a variety of factors such as strain rate and temperature. They are enhanced to make up for the ‘lost’ volume fraction in the images and as a result stiffen the rheological properties of the pure binder, for example using the Mori-Tanaka method. Material models for the crystalline fill They have developed two methods; the first is a method called the Wilhelmy Plate Method which introduces binder coated surfaces into reference liquids characterising their surface chemistry. From such values, a thermodynamic work of adhesion, between the binder and crystalline fill value, can be derived. This effectively equates to the FIGURE 3 Position-sensitive Photodetector Laser Diode Cantilever Spring Sample Tip Atomic force microscopy binder pull off (courtesy of University of Cambridge). 12 FIGURE 4 Failure propagation within the microstructure. energy requirement to separate the two surfaces from each other. The second method employs Atomic Force Microscopy (AFM) to measure the adhesive interaction by binder pull off experiments. This method uses cantilevers that have been doped with a hemisphere of binder material at one end, see Figure 3. These cantilevers are fully calibrated then slightly pressed into samples of the crystalline fill. The direction of the cantilever is then reversed and the force required to remove the samples is recorded, thus fully characterising the interaction. Both methods lend themselves to characterise the interaction between any crystal or binder system without the need to physically formulate the composite. The Model in Action Analyses have been carried out using finite element software. Once all the material model and interaction specific variables have been inputted, test specific boundary conditions that mimic real life mechanical test regimes are applied to the model and it is set to run over a specified time. This is being achieved by the incorporation of new materials, new failure mechanisms and additional interactions such as friction between the binder and particulate fill. The extra data will refine the predictive capability and choice of material models for numerous explosive compositions of interest. Simulations have provided excellent agreement both visually and numerically for a variety of materials. Figure 4 demonstrates various methods of failure, including adhesive debonding and material failure which coalesce to mimic composite fracture, features known to occur in real life PBXs. Due to such promising results and the effective working partnership between AWE, University of Cambridge and Imperial College London, the micromechanical modelling capability is being further developed. 13 Discovery Discovery25 • The Science Engineering and Technology Journal of AWE Personal Protective Equipment Testing The management of hazards by a hierarchy of control measures is a widely accepted system promoted by the Health and Safety Executive. The use of Personal Protective Equipment (PPE) is the least effective control in this hierarchy and is usually implemented as a final defence. PPE includes gloves, safety glasses and work wear, the selection of which is of critical importance in an explosives laboratory. In the laboratory, hands-on techniques are employed to prepare new explosives at the small (i.e. gram) scale. In addition to minimising the hazards through other controls, it is vital that the PPE provides suitable and sufficient protection against personal injury should an unintended explosion occur. The main hazard from an unintended small scale explosion in the laboratory is the fragmentation of the experimental apparatus such as glass or ceramic flasks, funnels etc. A review of the literature on this subject revealed that there was minimal evidence to support the selection of PPE to protect against this threat [1]. Consequently, a series of realistic trial scenarios were designed to test a range of PPE and thus provide data to enable appropriate PPE to be selected. As part of the safe system of work for the development of new explosive materials, the process begins with very small quantities. Quantities are scaled up once the new material has been shown to possess the required level of safety to a range of stimuli such as impact, friction, heat etc. In order to mimic the hazards associated with the scale up process, explosive charge masses of 0.3 g, 1.0 g and 7.5 g were used. Additional charge masses of 0.05 g and 5.0 g were used to simulate specific experimental scenarios. All charge masses given in this paper except the 0.05 g charge were in addition to the explosive mass in the detonator. Deliberate firings of high explosives are initiated by a detonator but these are not representative of an accident scenario in a chemical laboratory. Care was taken to select a detonator arrangement that minimised the release of metal fragments, as these could otherwise adversely influence the test results. A Reynolds Industries RP-2 detonator, mounted within a thick cardboard tube, was shown to be effective in this regard. This detonator also has an extremely low explosive mass of 0.05 g, thus any impact on the trial results were minimal. The fragment source consisted of selected items of laboratory apparatus, with a small PETN (pentaerythritol tetranitrate) explosived charge placed inside. This charge was then deliberately initiated by the RP-2 detonator, to break up the apparatus and propel the resulting fragments. The containers selected to act as fragment sources included glass test tubes, glass round bottomed flasks and ceramic Buchner funnels. The different fragment threats produced by the varying containers and charge masses allowed the degree of protection FIGURE 1 Casting of ballistic gelatine hands inside protective gloves. 15 Discovery Discovery25 • The Science and Technology Journal of AWE FIGURE 2 Example of high speed video stills of a 0.3 g explosive charge inside a ceramic Buchner funnel. afforded by the following items of PPE to be evaluated: • Four types of glove • Two types of wrist protector • One type of face shield • Two types of safety spectacles • Two types of safety screen The items of PPE were placed at a representative stand-off from the fragment source. Ballistic gelatine (see Figure 1) was used to simulate human tissue in order to assess the damage done to ‘hands’ both with and without PPE. TABLE 1 Container type Explosive mass (g) Peak observed velocity (ms-1) Buchner funnel 0.3 85 Buchner funnel 1.0 162 Buchner funnel 7.5 590 Round bottom flask 0.3 200 Round bottom flask 1.0 475 Round bottom flask 7.5 1230 Test tube 5.0 1260 Approximate peak fragment velocities. 16 smaller mass of explosive was also assessed. As part of the early stage explosives formulation process, a very small quantity of explosives and binder (0.02 g) in solvent may be hand-mixed in a nickel crucible, using the heat of the hand to drive off the solvent from the mixture. FIGURE 3 Damage to gelatine hands from 0.3 g (Left), with the palm protected, and 1.0 g (Right) explosive in a glass round bottom flask. Inspection of the gelatine enabled an assessment of the damage done to ‘hands’ inside protective gloves. Fragment flight and impact was visualised using a Phantom V710 camera recording at 25,000 frames per second. Hands The high speed videos of the various firings revealed the speed and relative sizes of the fragments produced, see Figure 2. Fragment velocities and sizes were measured using high speed video footage and the highest recorded velocities are shown in Table 1. The glass round bottom flask firings resulted in a large number of small, high velocity fragments. At 0.3 g and 1.0 g, the ceramic Buchner funnel produced fewer fragments, which were generally larger in size and with lower velocities. At 7.5 g the ceramic Buchner funnel was extensively fragmented, resulting in many small fragments; however the fragment velocities were still significantly lower than for the round bottom flask with the same charge mass. Trials to assess the damage of unprotected gelatine ‘hands’ showed a large number of fragment penetrations from the glass round bottom flask fragment source even with a 0.3 g charge. At the 1.0 g scale, the injuries to a real hand would be extensive, see Figure 3. The hazard to unprotected hands posed by the detonation of a much Two gelatine filled nitrile gloves were arranged, one supporting a nickel crucible and the other suspended above the crucible holding a bone spatula, see Figure 4. The test configuration represents a significant over-test, as the charge mass is more than double that which would be present in a real accident (the 0.05 g in a RP-2 detonator was the smallest charge that was capable of being fired), the explosive is a consolidated charge (rather than lower density powder) and full detonation occurs, which is extremely unlikely in this accident scenario. Examining the gelatine ‘hand’ after the nickel crucible firing revealed a shallow area of damage FIGURE 4 Nickel crucible with a 0.05 g explosive charge (RP-2 detonator only). 17 Discovery Discovery25 • The Science and Technology Journal of AWE (Figure 5), which, if considered on its own, could lead people to assume that this scenario would only cause a shallow surface injury to an unprotected hand. FIGURE 5 The high speed video showed the nickel crucible being forced down into the gelatine hand, which would likely result in soft tissue damage, bruising and potentially broken bones. The recovered crucible was found to be severely damaged with the end splayed open, resulting in sharp metal edges which were driven into the gelatine hand, see Figure 6. This highlighted the need for a holistic assessment of the data gathered, as focusing on just one element created an unbalanced view of the full effect. Damage to gelatine hand from Nickel crucible test. Gloves FIGURE 6 Splaying of Nickel crucible. FIGURE 7 Filter Paper Explosive Charge Built in Ceramic Gauze Explosive located in Buchner funnel, not in contact with external walls. 18 Low levels of surface damage were observed on gloves and wrist protectors using 0.3 g and 1.0 g charges, with the ceramic Buchner funnel as the fragment source. Extensive fragment damage to the surface of the PPE was observed under the same conditions with a glass round bottom flask fragment source, furthermore at the 1.0 g scale it was clear that penetration of some test articles had occurred. The same observation was made when repeating the experiment with a 7.5 g charge; the glass round bottom flask produced significantly more damage than the ceramic Buchner funnel. The degree of damage caused by the different fragment sources under the same conditions can Table 2 Penetrations, number/depth of deepest penetration Test article Unprotected hand 0.3 g ceramic Buchner funnel 1.0 g ceramic Buchner funnel 7.5 g ceramic Buchner funnel 4/moderate 18/moderate Severe/deep Red glove No penetration No penetration 6/deep White glove No penetration No penetration 2/shallow Blue glove No penetration No penetration 9/shallow Grey glove No penetration No penetration 11/shallow Yellow wrist protector No penetration 18/shallow Severe/deep Grey wrist protector No penetration No penetration 18/shallow Summary of results from Buchner funnel trials. Table 3 Penetrations, number/depth of deepest penetration Test article 0.3 g glass round bottom flask 1.0 g glass round bottom flask 7.5 g glass round bottom flask Unprotected hand 56/moderate Severe/deep Severe/deep Red glove No penetration 26/moderate 80/deep White glove No penetration 57/moderate Severe/deep Blue glove No penetration 16/moderate 65/moderate Grey glove Not tested 6/shallow 45/moderate Yellow wrist protector 2/moderate 17/deep Not tested protector Grey wrist Not tested 10/shallow Severe/deep protector Summary of results from glass round bottom flask trials. 19 Discovery Discovery25 • The Science and Technology Journal of AWE FIGURE 8 Example of high speed video stills of a liquid filled Aluminium bath with a glass round bottomed flask containing a 7.5 g explosive charge. be attributed to two main factors. Firstly, there is a significant air gap between the explosive and the relevant external structure of the ceramic Buchner funnel (see Figure 7), whereas the geometry of the round bottom flask means that the flask wall is in very close proximity to the explosive that was projected towards the PPE. The air gap in the Buchner funnel attenuates the shock and reduces the effect of the explosive on the fragment source. Secondly, the walls of the ceramic funnel are much thicker and more massive than those of the glass flask and are therefore not accelerated to the same degree. This explains the results in Table 1, where the peak observed fragment velocities from the glass round bottom flask are higher than those for the ceramic Buchner funnel at all charge masses. An indicative summary of the protection offered by the gloves and wrist protectors against the ceramic Buchner funnel and glass round bottom flask fragments is given in Tables 2 and 3. The relative depth and number of penetrations is shown, with ‘Severe’ indicating multiple areas of conjoined penetration that could not be divided into individual impact points. Aluminium Pan Aluminium pans are often used in oil baths to allow explosives synthesis processes to be heated. The ability of a oil filled aluminium pan to contain fragments from a glass round bottom flask using 1.0 g and 7.5 g charges was assessed (see Figure 8). The liquid (silicone oil in a laboratory environment, water in a test) transfers the explosive 20 shock directly to the aluminium pan. With a 1.0 g charge the pan remained intact providing effective radial fragment protection from the glass flask, whereas a 7.5 g charge was sufficient to fragment the pan. Safety Spectacles, Face Shields and Screens Safety spectacles and a standard polycarbonate laboratory face shield were tested against a 1.0 g charge in a glass round bottom flask, with complete protection being provided. A 5.0 g charge was subsequently tested with a combination of safety spectacles and face shield using a test tube as the fragment source. In this situation the fragments which impacted the PPE were directly in contact with the explosive charge, maximising the energy of the fragments. The against realistic explosive laboratory accident scenarios. AWE has used this evidence to select more appropriate PPE and has thus significantly improved the level of protection against these hazards. In general, protection, or at least a significant reduction in harm, was possible for the smaller charge sizes providing an appropriate degree of stand-off was maintained. FIGURE 9 References Glass test tube firing of a 5.0 g explosive charge with face shield, safety glasses (left hand side), 3 mm safety screen (right hand side) and 6 mm [1] Klapötke et al, Safety Science, 48 (2010), 28 – 34. safety screen (front). Acknowledgements face shield suffered multiple penetrations; however fragments that defeated the face shield were stopped by the safety spectacles, which were behind the face shield. The same combination of PPE was tested with a 7.5 g charge in the glass round bottom flask. The face shield failed catastrophically and although the safety spectacles were not penetrated by fragments, the arms of the spectacles were fractured at the hinges. A safety screen constructed from 3 mm polycarbonate was subjected to fragments generated by glass round bottom flasks with 1.0 g and 7.5 g charges and a glass test tube with a 5.0 g charge. The screen provided complete protection against the 1.0 g charge; however it failed catastrophically when subjected to the much more severe 5.0 g and 7.5 g threats, resulting in large fragments of polycarbonate with velocities of up to 66 ms-1. A 6 mm polycarbonate screen provided complete protection against the 5.0 g charge in the test tube. Figure 9 shows still images from the high speed video of the 5.0 g explosive charge in a glass test tube. On the left hand side of the experiment are the face shield and safety specticals, on the right is the 3 mm polycarbonate safety screen and at the bottom is the 6 mm thick polycarbonate safety screen. The initial stages of failure can be observed for the 3 mm safety screen in the lower middle image shown in Figure 9. The authors would like to thank the AWE Media Group for their support and use of their high speed video facilities and the Electro-Explosives Team for the use of their firing chambers and all the preparation work needed to conduct these trials. Conclusions These trials have allowed quantitative assessment of the mitigation provided by PPE 21 Discovery Discovery25 • The Science Engineering and Technology Journal of AWE Slapper Detonator Modelling Detonators are an essential component for the safe, repeatable and reliable operation of an explosive train. Slapper detonators (or Exploding Foil Initiators) were first developed in Lawrence Livermore National Laboratory in 1965 by John Stroud. Slapper detonators were designed mainly for military applications but have since been adopted by several fields including explosive welding and hardening, oil and mining. At AWE several modelling techniques are employed to help design Slapper detonators. Traditionally these models included empirical fits to past experimental data but with recent leaps in computational power more predictive, first principle based techniques are being developed. The functioning of a Slapper detonator can be sectioned into three parts: the electrical system, the flyer launch and the shock initiation of the explosive pellet. This article describes the techniques used in the empirical based approach and provides a brief description of future activities for more predictive techniques. plasma. Inertial forces restrain the expansion of the metal for a period of time, after which the metal expands rapidly; this point is known as the point of burst. The expanding plasma punches out a flyer that travels down a short barrel. At the end of this barrel the flyer shock initiates an explosive pellet such as HNS (hexanitrostilbene) or PETN (pentaerythritol tetranitrate). Slapper detonators require an electrical firing system, called the Fireset, to supply the correct electrical input to the bridge. The Fireset and detonator can be carefully designed to achieve the correct timing and output pressure from the device. A Semi-analytical Slapper Detonator Model The bridge has large variations in resistance due to changes in the thin foil when a large amount of current is passed FIGURE 1 Bridge Lands Dielectric material Functional Overview Figure 1 provides a side schematic Tamper view of a typical Slapper detonator and the 3D representation of a chip slapper detonator. Typically a few thousand amps of current Explosive (usually HNS or PETN) are passed through the bridge Explosive (usually HNS or PETN) Flyer travels down barrel material, causing a high current Barrel density and electromagnetic Air fields. This in turn causes Joule Flyer travelsDielectric down barrel material Lands Heating within the metal and Barrel Air subsequently the resistance Tamper Dielectric Bridge significantly increases. material Lands While the metal heats it rapidly changes state from a solid to Schematic representation of a Slapper detonator. Tamper Bridge 23 Discovery Discovery25 • The Science and Technology Journal of AWE through it. Modelling this behaviour from first principles is extremely computationally expensive as thermodynamic and electromagnetic effects would need to be taken into account. Empirically fitting to past data is a much faster method of determining the electrical response of a bridge, providing a semi-empirical model rather than a first principles approach. specific action, and therefore accumulated energy, has been shown to increase as a function of initial current rate of rise [2]. This suggests that there are inertial forces which initially restrict bridge expansion. Several authors have tried to understand and characterise the point at which the bridge bursts in Slappers detonators. Anderson and Neilson were the first to use the concept of action which is based loosely on classical mechanics [1]. Action is defined as FIGURE 2 Exp current Exp voltage Sim current Sim voltage Current (A) Voltage (V) t A(t) = ∫ tI 2 (t) dt A(t) =0 ∫ I 2 (t) dt Past experimental data can be used to construct specific action-resistivity curves for a particular bridge material. Simple scaling laws can then be used to adjust the shape and position of this curve to different Fireset conditions. The electrical current within the circuit can be modelled using a first order differential equation which is supplemented with the appropriate specific action-resistivity curve. A prediction can be made of the 0 A more general term is defined as specific action t 2 I (t) 2 g(t) = ∫ t I (t) dt g(t) =0 ∫ a dt a 0 Time (s) Validation of electrical response prediction made by the model. Velocity (ms -1) where I(t) is the current flowing FIGURE 3 through the wire and a is its area. Action is a convenient quantity to characterise the1point of burst E �as = J = voltage and resistance 1 are difficult E� =E c + � c =� Ec + E � c to obtainJ experimentally. EE c + �� c � Ec + E � c � E Anderson and Neilson found from several experiments that the specific action at burst as a function of current density was invariant to within 10% for Simulation each respective bridge material. Experiment Therefore, specific action can Distance travelled (mm) potentially be used to find the point of burst for a bridge material, irrespective of the Prediction of flyer velocity and comparison with experimental trace. bridge geometry. In practice 24 Using the point of burst and energy data, the velocity of the flyer can be calculated. The flyer is pushed down the barrel by the rapidly expanding bridge which has changed state. The model assumes that the density of the expanding gas is uniform, there is an infinite tamper behind the flow, the flyer only has kinetic energy and the gas flow is ideal. A one-dimensional, incompressible Navier-Stokes formula is used to express the pressure in the gas and the ideal gas equation of state is used to relate the pressure of the gas and input energy. As a result, an analytical expression can be found for the acceleration of the flyer which is solved numerically using a Runge-Kutta algorithm. Experimental flyer velocities can be obtained from a technique called Photonic Doppler Velocimetry (PDV) and can be used to validate the model. Figure 3 shows that the model can achieve a very good fit to experimental PDV traces. This gives confidence that the correct impact velocity is being used for the prediction of shock initiation within the detonator explosive pellet. The final part of the model predicts whether the energy from the impact of the flyer is sufficient to cause a detonation within the explosive pellet. The one dimensional initiation threshold of explosives can be quickly evaluated using the James Initiation Criteria [3] developed at AWE. The James Criteria uses the observation that many explosives require a minimum amount of energy to initiate. This phenomena is taken into account using a critical specific energy, ∑ c , where P is the pressure imparted into the explosive and τ is the duration of the shock. Experimental data have shown t that for=a given there is a A(t) I 2 (t)explosive dt critical value of E, denoted E c and 0 Σ c , below which no detonation occurs, as illustrated in Figure 4. ∫ This means that a single value, t 2 J, can be calculated I (t) and used = whether ordtnot the tog(t) predict a explosive0pellet within the detonator will detonate or not. The James Criteria parameter is defined as: ∫ ∑c = u 2/ 2 and u is the particle velocity in the shock transmitted to the explosive by the flyer impact. The energy per unit area of the shock imparted into the explosive is also needed and is defined as J= 1 Ec � c + � E = E� � Ec + E � c The semi-analytical Slapper detonator model developed at AWE calculates the threshold firing voltage needed for the E = Puτ FIGURE 4 Increased shock pressure dynamic electrical response of the bridge and hence the energy supplied to the flyer when launched. Figure 2 shows a comparison between experimental data and predictions made by the model. Typically electrical predictions are made within 15% error with experiment using this technique. Σ Det. J=1 Initiation Cut-off Σc Non-Det. Ec E Increasing shock duration Critical values found for a typical explosive below which no detonation is seen when impacted with a thin flyer. 25 Discovery Discovery25 • The Science and Technology Journal of AWE FIGURE 5 Magnetohydrodynamics model showing temperature. explosive HNS and showed within 10% agreement with experimental data. High Fidelity Predictive Models The techniques described so far are empirical in nature making them dependent on existing experimental data. The first principles approach demonstrates application of our understanding of the detonator function and could provide greater accuracy in the predictions of Slapper detonator performance over the semi-empirical approach. However, high fidelity modelling of Slapper systems is not trivial as electromagnetic, thermodynamic and hydrodynamic processes are all prevalent and these also demand experimental data. Much like the semi-analytical approach described above, the physical phenomena in a Slapper detonator can be broken down into stages. Magnetohydrodynamics (MHD) has long been used in astrophysics to model magnetic fields in stellar objects. This technique can also be applied to Slapper detonators as extremely large electromagnetic fields are seen that also need to be coupled to a thermal conduction model. AWE has recently been using a code to successfully perform MHD simulations, as illustrated in Figure 5. In future this technique will be applied to Slapper detonators to further facilitate the design of new Slapper detonator geometries and materials. In addition to an MHD component, predicting the initiation of the explosive is also required. Presently there are very few models which could simulate the build up of detonation in either PETN or HNS. A technique which can simulate the build up to detonations is called reactive flow modelling and although more predictive than the James criterion, still requires experimental data and certain assumptions to supplement the equation of state. 26 It is understood that the important reaction stages of the detonator running to detonation takes place over multi-length scales. In the future a multiscale modelling and experimental approach may be needed to obtain a fundamental understanding of the explosive, as illustrated in Figure 6. FIGURE 6 Hydrodynamics Metres Seconds Mesoscale Micrometres Microseconds Molecular dynamics Nanometres Nanoseconds Equation of state, P(V, E, T) needed for continuum simulations 0.1 nm P(V, E, T) of crystalline high explosive obtained from simulations 0.1 nm A multi-scale modelling approach to understanding the buildup to an explosive detonation. Summary References Acknowledgements Slapper detonators can provide both safety and performance benefits to an explosive system. They are often part of a complex initiation train which requires both precise timing and output pressure from the detonator. Modelling these systems gives a great deal of insight into performance, given its complex and coupled internal processes. Currently AWE uses empirical methods to explore the design space and parameters needed to achieve a specific outcome. However, in future more high fidelity physics based approaches will be used to obtain a more predictive capability. [1] The authors would like to thank Hugh James for the use of his diagram to explain the James Criteria. Also the authors would like to thank Jonathan Baker for the images showing magnetohydrodynamic simulations. Many thanks also to Scott Aitken for the experimental data used to validate the semi-analytical Slapper detonator model. Anderson, G.W. and Neilson, F.W. - Use of the “Action Integral” in exploding wire studies. Exploding Wires, Plenum Press Inc., first edition, 1959. [2] Lee, R.S. - Fireset. Lawrence Livermore National Laboratory Report, UCID 21322, 1988. [3] James, H.R.- An extension to the critical energy criterion used to predict shock initiation thresholds. Propellants, Explosives, Pyrotechnics, vol. 21, p 8 – 13, 1996. 27 Many thanks to Peter Bolton, Colum O’Connor and Adam Hazelwood for providing an SEM image of PETN. Discovery Discovery25 • The Science Engineering and Technology Journal of AWE Hazard and Safety Studies To determine whether an explosive is safe to use, it is subjected to UK standard tests to determine the materials sensitiveness (a measure of the stimulus required to ignite or initiate an explosive material) and explosiveness (a measure of the degree of violence shown by an explosive material in response to a given stimulus). UK standard tests for both approval and qualification form part of AWE’s capability to demonstrate the hazard and safety characteristics of explosive materials. These are in addition to bespoke experiments, simulation and material modelling efforts. The tests commonly used at AWE examine the response to impact, heat and Electro Static Discharge (ESD). Tests include impact sensitiveness (Rotter test), response to incendive sparks (Ease of Ignition test), response to prolonged flame exposure (Train test), heat stability (Temperature of Ignition test), sensitiveness to impact between a variety of surfaces (Mallet Friction test), response to exposure to electrostatic discharges of different energies (Spark test), explosiveness response when impacted and ignited (LabSET test), impact with crush and pinch (Steven test) and response due to penetration by a metal rod (Small Scale Spigot test). The tests are undertaken on either powder or relatively small pressed charges. Figure 1 is a picture of an explosive powder undergoing exposure to a prolonged flame, also know as the Train test. If the material passes the tests then it may undergo qualification tests designated by the Defence Ordnance Safety Group (DOSG) for approval. These tests use pressed charges to examine the response of the explosive to potential threats that the material may encounter during its lifecycle. The two most likely threats are impacts e.g. during transportation or as a result of drops or falls and thermal e.g. fires. A number of approval tests have their origins in the conventional munitions arena. Examples of some of the threats and stimuli FIGURE 1 that the qualification tests subject the explosives to are; glancing blows against gritted surfaces (Oblique Impact test), crush, pinch and extrusion (Susan test), intrusion (Spigot Impact test), localised ignition whilst confined (Tube test – Internal Ignition) and bulk heating whilst confined (Tube test – Fast Heating). Performing these approval tests on a variety of explosive materials offers an opportunity to ‘rank’ materials in terms of sensitiveness and explosiveness. The extent of the reaction in the explosive is dependent on a number of factors such as the composition of the explosive (explosive type, binder concentration and material etc), nature of the threat (for impact tests speed, mass, shape, material and orientation) and degree of confinement around the explosive in the region of the reaction. Confinement can protect the explosive from the threat; generally higher levels of confinement give more violent reactions than bare or lightly confined tests. For bare or lightly confined test vehicle impacts, the explosives are generally disrupted by the impact to such an extent that the sites of ignition are either extinguished or that the extent Explosive powder burning steadily in the train test. 29 Discovery Discovery25 • The Science and Technology Journal of AWE FIGURE 2 Frames from a video camera showing the reaction of some explosives, encased within a lightly confining test vehicle, impacted by a bullet. of the reaction does not grow past the burning stage. Heavy confinement tends to keep most of the explosive intact within the test vehicle and traps the gasses produced from the reaction of the explosives. The trapped gas increases the pressure within the test vehicle and causes the burn rate to increase, which increases the pressure rapidly and which further increases the burn rate. If the explosive has extensive cracks throughout its structure, as a result of the damage caused by the impact, then the flame front can rapidly travel throughout the explosive causing the majority of the volume of the explosive to contribute to the rapidly increasing pressure within the confinement. Usually, this rapid rise in internal pressure, which can occur in just a few milliseconds, will overcome the strength of the confinement and the test vehicle will disassemble in a manner that can appear to resemble an explosion. If the confinement provided by the test vehicle is sufficiently high, this rapid increase in internal pressure can transition from a rapid burn to detonation. 30 AWE's hazard and safety capability is extended by the work with Cranfield University. The effect of insult type and severity on the reaction growth and subsequent violence has been researched and implemented at a firing facility at Cranfield. This work was undertaken under the AWE technical outreach programme. The research contract has enabled both impact and thermal threat type tests to be developed. Tests have ranged from drop weight impacts, bullet impacts (see Figure 2) and projectiles fired from FIGURE 3 Stills from a video showing a gas gun projectile impacting a spigot that causes a reaction in a heavily confined explosive powder. gas guns into bare or confined explosives. Figure 3 shows stills from a video showing the response of gas gun projectile impacting a spigot that causes a reaction in a heavily confined explosive powder. beneath, through the glass surface, by high speed video cameras and any ignition of the explosive is recorded as a flash of bright light, Figure 4 shows schematics of the Trolley test and some target plates after impact. A recent development in impact testing (Trolley test) complements the Oblique Impact Qualification test which is used at approval. Whereas the Oblique Impact test examines the explosiveness of a large explosives test sample by dropping it at an angle onto a gritted surface from different heights, the Trolley test examines the sensitiveness of a small sample of the explosive to a similar threat by using a gas gun to propel the trolley, with the explosive hung beneath, down a track so that the explosive impacts an inclined, gritted toughened glass surface. The impact is filmed from Thermal tests developed as part of the research contract with Cranfield University have enabled explosive test samples to be heated, in different test vehicles, which provide various levels of confinement, at different heating rates until the explosive reacts. Blast gauges, thermocouples and video cameras are used to record the development and growth of any reaction produced by the explosives. In one test a disc of explosive was lightly confined in a hermetically sealed test vehicle and subjected to a propane fuelled flame on the top surface of the test vehicle. It was found that the The trapped gas increases the pressure within the test vehicle and causes the burn rate to increase, which increases the pressure rapidly and which further increases the burn rate. 31 Discovery Discovery25 • The Science and Technology Journal of AWE test vehicle separated relatively benignly and most of the explosive was recovered unreacted. FIGURE 4 a. To model the response of explosives to impact, a predictive capability called HERMES (High Explosive Response to Mechanical Stimulus) has been developed. The HERMES material model comprises of several sub models including a constitutive model for strength, porosity and surface area through fragmentation, an ignition model an ignition front propagation model, and a model for burning after ignition. b. Ignition is based on a purely mechanical criterion depending on a time integral of a function of the shear, equivalent stress, pressure and strain rate as follows: c. Ign = ∫ d. t 0 2– 27 s1s2s3 2Y 3 5 p + s2 2 P0 1/2 έp dt Here s 1,2,3 are the principal stress deviators, Y is the equivalent stress, p is the pressure, P0 is a prescribed constant value of pressure, and έp is the plastic strain rate. The mathematics are designed to capture the cumulative effects of shearing, compressing and rapid straining through the respective terms in the time integral. e. The Trolley test. (a) A schematic of the test layout with the target, trolley and test track on the left of the wall and the gas gun on the right of the wall. (b) A schematic of the explosives under the trolley impacting the inclined target surface. (c) The target plate after impact with HE smeared across its surface. (d) Shows the remains of the explosive attached to the trolley after impact. (e) Frame from the high speed video showing ignition (the bright flash of light in the upper centre of the picture) of the explosive as it impacts the target. 32 Ignition is deemed to commence when I gn reaches a particular value called the ignition criteria. The ignition criterion vary between different types of explosives and are determined from observations of experiments with these different explosives. FIGURE 5 Projectile Cover Plate Test Vehicle Base The predicted ignition location (left image) for explosives in the “Steven” test. The projectile is deforming the test vehicle cover plate which is crushing the explosive against the test vehicle base. A “Steven” test vehicle base, recovered from a test (right image) shows annular scorch patterns on the base plate. In order to validate the results obtained from HERMES, a simulation was constructed of an impact test (Steven test) with the results obtained being compared to the results of experiments. For a range of impact velocities, a number of model parameters were varied and their effect on the predicted explosive response was noted. The predicted onset of ignition due to extreme shear and material deformation occurs in an annular region similar to those seen in experiments. Figure 5 show the predicted ignition locations and the base plate for a Steven test. For the prediction of explosives response to thermal threat, there have been a number of heat flow codes which have been modified to include chemical reaction routines. These codes have ranged from the relatively simple early codes such as TRUMP and TOPAZ to more recent and considerably more complex coupled codes such as CALORE and ALE-3D. These codes are constantly under development with incorporations such as more complex chemical reaction kinetics for the explosive materials. Summary AWE maintains a modern hazard testing capability and the ability to develop new test tools to examine the response of explosives to particular threats. This is needed to continue to assure that the explosives that are used at AWE are suitable and safe to use throughout their lifecycle. Future Work The models are planned to be improved by carrying out tests that determine the physical mechanisms which determine the onset of reaction and its subsequent growth. This will entail the inclusion of improved diagnostic techniques in current tests and the design and implementation of bespoke test methods. The development of an enhanced predictive capability based on computer modelling enables better understanding of the behaviour of the explosives when subjected to threats as well as providing the opportunity to reduce the requirement of fielding expensive, large scale trials. 33 Discovery Discovery25 • The Science Engineering and Technology Journal of AWE AWE’s Outreach Major Events and Collaborative Activities In this section, we cover a number of high-profile events and conferences in which AWE has been involved. They represent a wide range of disciplines and areas of expertise and also put AWE’s relationship with key stakeholders into context. UK Policy Conference Some 120 newcomers and experts in the field of nuclear weapons policy gathered at the third AWE sponsored annual Project On Nuclear Issues (UK PONI) Conference in London. The event, held on 6 June 2013 at the Royal United Services Institute (RUSI), created a forum which combined fresh perspectives on nuclear policy with insights and experiences from established experts to explore the future of nuclear policy, both at home and abroad. The conference attracts representatives from across the nuclear community including the UK government, industry and academia. There were also international representatives from the US, Russia, France and China giving attendees a thoughtprovoking and lively debate. Other representatives included FCO, NATO, MOD, Rolls-Royce, King’s College London, Lockheed Martin and the Centre for Strategic International Studies. Former Secretary-General NATO, the Rt Hon Lord George Robertson, delivered a keynote address during which he described his journey on deterrence issues and some of the political challenges when he was UK Defence Secretary and during his time in office at NATO. He said Continuous at Sea Detterrence (CASD) is the ultimate deterrent to safeguard our country. Rt Hon Lord George Robertson describing his nuclear journey Institute of Mathematics Nearly 50 scientists, industry experts, academics and students had their first glimpse of the world leading Orion laser facility when the delegation gathered at the Institute of Mathematics and its Applications (IMA) Employers’ Forum, hosted by AWE Aldermaston on 4 March 2013. The one-day session centred on academic technical outreach, including some of the advantages and challenges that are faced by employers and universities when engaging in outreach. As well as AWE, other organisations represented included Dstl, Thales, EDF Energy, Schlumberger, Qinetiq, Heriot-Watt University, University of Reading, University of Cambridge, Loughborough University, University of Warwick, University of Stirling and University of Manchester. Inaugural Linear Solver Workshop A two-day workshop was held at Warwick University on 8-9 July 2013 to discuss a key challenge for exploiting the rapidly evolving high performance computing (HPC) architectures to solve the complex science and engineering models which underpin the techniques used by AWE for certification of the stockpile. Feedback from the meeting attendees highlighted the importance of bringing together experts in these diverse areas together in the UK to discuss the common challenges faced in 35 Discovery Discovery25 • The Science and Technology Journal of AWE exploiting the next generation of hardware platforms. To exploit the next generation of hardware, the workshop highlighted the importance of developing new algorithms which remains one of the main challenges hence the involvement of numerical analysts. This will be driven by the competing demands for greater power efficiency and increased performance. International Radionuclide Migration Conference Over 300 international delegates gathered at The Brighton Centre, for the Migration 2013 conference. This was the first time the conference had been held in the UK and gave our scientists the opportunity to promote their research to a wide number of universities, research institutions and government bodies. "the UK has a long history of leadership in nuclear forensics. While important in its own right, it also gives us the edge in a diplomatic and political context through raising the UK profile and enhancing credibility.” The Migration conference series provide an international forum for the exchange of scientific information on chemical processes controlling the migration behaviour of actinides and fission products in natural aquifer systems. AWE hosted a number of posters at the evening sessions and, throughout the event, an AWE exhibition stand showcased the core business areas and breadth of the science conducted at AWE. AWE presence at Migration 2013 Conference 36 The main focus of the conference was the development of actinide chemistry to enable the characterisation of activation and fission products in natural aquifer systems and the subsequent assessment of geochemical species and their migration in the geosphere. Presenters provided lectures on experimental investigations, method development and predictive modelling. A special session was held to discuss the programme in place for waste disposal in the UK. AWE supports the 14th international Conference on Radionuclide Migration Nuclear Forensic Workshop Nearly 80 experts in nuclear forensics from around the world gathered for the Nuclear Forensics Workshop held on 7-9 January 2014 at Lancaster House, London. The three-day event was co-sponsored by AWE, MOD, the Home Office and the Foreign Office under the aegis of the Global Initiative to Combat Nuclear Terrorism (GICNT). The event shared knowledge in nuclear forensics between the GICNT Member States to support global security efforts. The mission of the GICNT is to strengthen global capacity to prevent, detect and respond to nuclear terrorism by conducting multilateral activities that strengthen the plans, policies, procedures and interoperability of partner nations. The GICNT is co-chaired by the US and Russia. Minister for Crime and Security, James Brokenshire MP, said: “In recognition of the threat of nuclear terrorism, the UK is developing advanced nuclear forensics capabilities and practices. These capabilities will be embedded and integrated into our existing law enforcement and operational systems to provide a seamless end to end capability for managing nuclear security incidents. Chief Scientific Adviser to the Foreign Office, Professor Robin Grimes, said: “It is well appreciated that the UK has a long history of leadership in nuclear forensics. While important in its own right, it also gives us the edge in a diplomatic and political context through raising the UK profile and enhancing credibility.” capability that allows the UK to investigate criminal acts involving nuclear materials. The Conventional Forensics Analysis Capability can recover fingerprints, fibres, DNA and other traditional trace forensics markers from material that have been contaminated with radiological, nuclear or explosive materials.” As part of the forum, a structured table-top exercise called Blue Beagle was discussed by a panel composed of leaders in nuclear forensics from law enforcement and other agencies. The panel discussed various stages of a radiological crime scene investigation. “Building upon the knowledge and capabilities of AWE we have created a dedicated world-class nuclear forensics analytical 37 Discovery Discovery25 • The Science and Technology Journal of AWE Discovery Editor: Doctor Graeme Nicholson Editorial Board: Contributors: David Chambers Doctor David Geeson Doctor Norman Godfrey Doctor Katherine Grant Rashad Hussain David Murray John Roberson Overview Doctor Claire Leppard Doctor Steve Trussell Explosives Materials Development Doctor Peter Bolton Doctor Adam Hazelwood Michael Hopkins Till Graphic Design and Illustration: Micromechanical Modelling Doctor Claire Leppard Daniel Lewis AWE Media Group Photography: AWE Media Group Find out more about AWE at our website: www.awe.co.uk Comments and suggestions regarding this journal, please email: discovery@awe.co.uk For further copies of this journal and details of other AWE publications, please write to: Corporate Communications Office Building F161.2 AWE Aldermaston Reading Berkshire RG7 4PR 38 PPE Testing Doctor Peter Jenkins Stephen Miller Doctor Chris Murray Slapper Detonator Modelling Doctor Mary-Ann Maheswaran John Richardson Hazard and Safety Studies Doctor John Curtis Andrew Jones 25 Printed on Greencoat silk Logo here Discovery Discovery25 • The Science and Technology Journal of AWE REF MG/33420 If you require this document in an alternative format; such as large print, on alternative paper, or electronically, please contact the Media Group Graphics Team 0118 982 5249 who will be happy to help. © British Crown Owned Copyright 2014/AWE Published with permission of the Controller of Her Britannic Majesty’s Stationery Office. “This document is of United Kingdom origin and contains proprietary information which is the property of the Secretary of State for Defence. It is furnished in confidence and may not be copied, used or disclosed in whole or in part without prior written consent of Defence Intellectual Property Rights DGDCDIPR-PL - Ministry of Defence, Abbey Wood, Bristol, BS34 8JH, England.” AWE is the trading name for AWE plc. AWE is a Government Owned, Contractor Operated organisation. AWE is operated by a joint venture of Jacobs Engineering, Lockheed Martin, and Serco. AWE Aldermaston, Reading, Berkshire, RG7 4PR Discovery 25 • The Science Engineering and Technology Journal of AWE The Science & Technology Journal of AWE • Issue 25 • September 2014