GB2595434A - Witness pack - Google Patents

Abstract

A witness pack 9 for detecting low velocity and/or energy and/or non-metallic projectile fragments comprises a plurality of alternating witness layers 1 and impact absorbing layers 2. The first layer of the plurality of alternating witness layers 1 and impact absorbing layers 2 is a witness layer 1 The material composition of the witness layers 1 and impact absorbing layers 2 are individually selected to perform their respective functions. The witness layers 1 may be formed from a flexible polymer, for example polyethylene, polyvinyl chloride or polypropylene. The impact absorbing layers 2 may be formed from a rubber or foam rubber material, for example neoprene foam, silicone foam, silicone rubber or butyl rubber. There may be 3 to 10 witness layers 1.

Description

WITNESS PACK

Technical Field of the Invention

The invention is concerned with a witness pack for modelling physical injury, more specifically a witness pack for detecting and analysing low velocity, low energy and/or non-metallic projectiles such as fragments. In preferred embodiments, the invention enables eye and/or skin injury to be modelled. The witness pack comprises a plurality of alternating witness layers and impact absorbing layers, wherein the first layer of the plurality of alternating witness layers and impact absorbing layers is a witness layer and wherein the material composition of the witness layers and impact absorbing layer(s) are individually selected to perform their respective functions. The invention is also concerned with related systems and uses.

Backaround to the Invention

Witness packs are used in the field of munitions, charge and explosives testing for the detection and analysis of penetrating explosive fragments. Information obtained from witness packs can be used to assess potential damage in different scenarios, for example damage to armour, buildings, personnel and so on. Types of witness pack range from packs comprising spaced metal layers, through strawboard packs, to packs comprising a tissue simulant such as gelatin. Typically, the packs are designed to detect penetration by metal fragments, usually high velocity metal fragments.

It is an object of the invention to provide an improved witness pack, particularly an improved witness pack suitable for use in detecting low energy and/or low velocity fragments and/or non-metallic fragments.

Summary of the Invention

According to a first aspect, the invention provides a witness pack for detecting low velocity, low energy and/or non-metallic projectile fragments, the witness pack comprising a plurality of alternating witness layers and impact absorbing layers, wherein the first layer of the plurality of alternating witness layers and impact absorbing layers is a witness layer and wherein the material composition of the witness layers and impact absorbing layer(s) are individually selected to perform their respective functions.

By first layer is meant the layer of the plurality of alternating layers which will be first impinged/penetrated by fragments (secondary blast). "First layer" does not exclude other functional layers being positioned between the blast and the first layer.

In its simplest form, the witness pack may comprise two witness layers and a single interposed impact absorbing layer. However, there is usually a plurality of impact absorbing layers. The final layer of the plurality of alternating witness layers and impact absorbing layers may be a witness layer or it may be an impact absorbing layer. In various preferred embodiments, other functional layers may be positioned in front of, at the rear of and/or in between the plurality of alternating witness and impact absorbing layers.

Witness packs are used in explosive trials, for example in trials for assessing the damage caused by explosive penetration and/or fragmentation. Amongst other requirements, there may be a need to assess personnel injury using a witness pack. Personnel injury of interest may be generated by fragments, for example lower energy, lower velocity and/or non-metallic fragments.

It is known to use muscle tissue simulants such as gelatin to assess personnel injury, but this can be impractical because large areas often need to be assessed, multiple fragments are generated, and simple analysis is required. In addition to this, the temperature dependence of gelatin may restrict its practicality for this type of application.

Multilayer witness packs are known which capture fragments; subsequent analysis of the number of layers perforated (along with fragment mass) can then be used to back-calculate impact velocities. One example is a multilayer witness pack comprising aluminium or steel plates separated by polystyrene foam. The metal plates act as witness layers and also attenuate the projectile or fragment velocity, whereas the polystyrene layers are used as spacers to provide a suitable pack depth. Prior art witness packs of this type are suitable for detecting and analysing high velocity metal fragments (which fragments can cause lethal damage to personnel), but are generally unsuitable for detecting injurious, low velocity or low energy fragments and/or non-metallic fragments.

In contrast to prior art witness packs whereby the witness layer(s) serve both to detect fragments and to attenuate the velocity of the fragments, in the invention an alternative approach has been taken in which the material compositions of the alternating witness layers and impact absorbing layers are individually selected to perform their respective witness (detection) and impact absorbing (velocity or energy attenuation) functions. This provides a witness pack structure which can be optimised both to detect and analyse multiple fragments generated over a large area by relatively simple analysis, and to select an impact absorbing layer which simulates human tissue. Typically, the witness layers and impact absorbing layer(s) each comprise a different material, usually materials having different physical properties.

The materials forming the alternating witness layers and impact absorbing layer(s) are preferably specially selected to allow penetration by low velocity, low energy and/or non-metallic particles, and subsequent analysis and 30 measurement thereof.

The witness pack of the invention is particularly suitable for assessing personnel damage caused by an explosive event such as the blast from an Improvised Explosive Device (IED). IEDs produce a significant quantity of nonmetallic debris (including particles or fragments of earth, stones and so on) in addition to metallic debris from the munition and/or any target. The explosive debris may also comprise a significant proportion of low velocity and/or low energy fragments. Low velocity, low energy and non-metallic blast debris may all pose a significant threat of penetrative injury to personnel, particularly eye and skin penetrative injury. It is crucial to be able to obtain information on injurious, but not necessarily lethal, fragments of this type, so that information can be fed into injury prediction models. This allows, amongst other applications, Personal Protective Equipment (PPE) to be designed and improved Other events which produce low velocity, low energy and/or non-metallic debris are behind barrier events & explosions (glazing, masonry, etc.), or devices specifically designed to produce non-metallic or low energy metal fragments. The invention may be used in safety assessments of; explosive devices, such as IEDs and Active Integrated Protection Systems (AIPS), Explosive Methods of Entry (EMoE), door breaching (ballistic), render safe procedures, ricochet of small arms rounds, behind barrier effects, including within vehicles and structures, Lethality/effectiveness assessments and PPE/ballistic material performance assessments.

By low velocity is meant a velocity up to about 800 m/s, more typically up to about 600 m/s. The velocity may be in excess of about 50 m/s. By low energy is meant an energy up to about 200 J, more typically up to about 120 J. The energy may be in excess of about 0.1 J. The skilled person will realise that the energy and velocity of a particular fragment depends on the fragment size and mass.

By way of example, a 6 mm glass sphere having a mass of 0.28 g may have a maximum velocity around 379 m/s and/or a maximum energy around 20.3 J to remain within the preferred 9 impact absorbing layer embodiment of the witness pack.

The purpose of the witness layer is to detect impact locations across the area of the witness pack, and along the length of the witness pack. After use, holes formed in the various witness layers can be analysed to obtain information on location and velocity of impact.

Preferably, the witness layers are formed from a flexible polymer, for example a flexible polymer selected from polyethylene, polyvinyl chloride and polypropylene. A flexible polymer is a preferred material for recording the impact locations because it remains intact, and can readily be removed and analysed.

The thickness of the witness layers may be selected to allow penetration by fragments having a velocity and/or energy range of interest. Typically, the thickness of the witness layers is less than or equal to 125 micron. The thickness of the witness layers may lie in the range 30 to 125 microns, more preferably in the range 50 to 75 microns, even more preferably in the range 55 to 65 microns. In one specific example, 62.5 micron (250 gauge) polyethylene was used to form the witness layers.

The purpose of the impact absorbing layer(s) is to attenuate the velocity and/or energy of the fragments in a similar manner to human muscle. Accordingly, the density of the impact absorbing layer is selected to exhibit, in use, impact absorbing performance similar to human muscle tissue. Preferably, the impact absorbing layer(s) are formed from a material having a density in the range 50 to 2500 kg/m3, preferably in the range 100 to 1500 kg/m3, more preferably in the range 150 to 500 kg/m3.

It is desirable that the impact absorbing material is non-frangible at likely incident velocities and/or energies. Suitable choices of material include saturated and/or unsaturated rubbers and elastomers, including rubber or elastomeric foams. Closed cell foams are preferred over open cell foams because they are typically less frangible. Rubber materials may be natural or synthetic.

Preferably, the impact absorbing material is selected from the group consisting of neoprene foam, silicone foam, silicone rubber and butyl rubber.

Typically, the thickness of the impact absorbing layer(s) lies in the range 5 to 25 mm, more preferably 5 to 12 mm. In one specific example, layer thicknesses of 10 mm ± 1 mm were selected.

The resolution of impact velocity predictions is typically determined by the thickness of the impact absorbing layers. In the invention, the witness pack is preferably similar to, but at least as easy to penetrate as, muscle tissue, so that any potentially injurious penetrating projectiles are not artificially discounted by injury models using the pack. Compared to strawboard packs of 3.8 mm layer thickness, and metal spaced witness packs, the preferred layer thickness of the invention provides approximately 2.5 times better resolution in terms of impact velocity predictions for the projectiles used as part of pack calibration.

In theory, the resolution of the witness pack can be increased even more by reducing the impact absorbing layer thickness. However, this would add additional burden to post-test analysis of the model where fragments have to be identified and recovered from the model to obtain the maximum layer perforated and fragment mass. The preferred thickness range of the invention is therefore a trade-off between resolution and analysis burden.

The number of alternating witness layers and impact absorbing layers are chosen to suit a desired application. In general, for measuring and analysing low velocity and/or energy fragments and/or non-metallic fragments, up to 10 witness layers are used. This means that up to 10 impact absorbing layers are used if the impact absorbing layer is the final layer of the plurality of alternating witness and impact absorbing layers, and up to 9 impact absorbing layers are used if the final layer of the plurality of witness and impact absorbing layers is a witness layer. In each case, this provides a sufficient number of witness layers to capture the majority of fragments being measured and analysed.

More usually, the number of witness layers lies in the range 3 to 10, meaning that the number of impact absorbing layers lies in the range 2 to 9 if the final layer of the plurality of witness and impact absorbing layers is a witness layer, or in the range 2 to 10 if the final layer is an impact absorbing layer.

The witness packs of the invention enable back calculation of fragment velocities from their depth of penetration (DoP) and recovered mass. The multiple layers used in the pack allow assessment over a range of velocities and/or energies, and information derived from the pack can be used to assess injury via known models, for example the Sperrazza and Kokinakis Incapacitation Criteria.

Witness packs according to the invention provide a number of advantages, including: a long shelf life (validated for at least 4 years), velocity estimate calibration for a wide range of projectile properties (geometries, diameters, masses and densities), possible semi-automated analysis using image analysis of holes in preferred flexible polymer witness layers (if fragment densities are known), relatively inexpensive construction materials, weatherproof and waterproof using preferred materials and constructions, and lightweight.

In a preferred embodiment, the witness pack is specially adapted to model eye penetration and skin perforation and (optionally) corneal abrasion. For the assessment of skin and eye injuries, the fragment velocity and/or energy range of interest is particularly low, and data of interest can be acquired by providing an additional strike layer and monitoring the penetration of the strike layer and the first impact absorbing layer. The prediction of eye penetration risk is independent of the foam layer thickness (instead it is dependent on the strike layer) and prediction of skin perforation risk is dependent on the first layer of impact absorbing material in the witness pack, allowing the performance to be matched to human skin.

Accordingly, if the witness pack is required to model eye injury, an additional strike layer may be applied to the front face of the pack. (By front face is meant the face forward of the first layer.) The strike layer may comprise a material which is specially selected to model corneal penetration. The strike layer may be flexible or rigid. Preferably, the material chosen for the strike layer matches the mechanical and/or physical properties of the human cornea. Suitable materials are polyester, silicone elastomer film or silicone sheet.

For a polyester strike layer, the layer has a preferred thickness in the range 15 to 30 micron, more preferably 20 to 26 micron. In one specific example 23 ± 2 micron polyester was used. More generally, however, the chosen thickness of the strike layer depends on the material selected and the corneal property(s) selected for matching. The depth of the strike layer may be up to 300 micron for materials other than polyester.

The strike layer may comprise an outward facing adhesive layer so as to trap particles and fragments having a velocity and/or energy insufficient to penetrate the witness pack, but sufficient to cause corneal abrasion. In use, for the assessment of corneal abrasion injuries, any particles adhering to the adhesive layer after testing indicate a potential abrasion risk. Corneal abrasion is a low level injury which normally self-heals with no long term consequences. However, the use of an outward facing adhesive layer nevertheless makes an assessment possible. The adhesive layer may be a double sided adhesive, for example double sided polyacrylate adhesive.

If the witness pack is only required to model eye and/or skin injury, the number of layers used in the witness pack can be (but is not necessarily) constrained. In one preferred embodiment of the invention, the witness pack is adapted to model eye and skin penetration, and comprises 3 witness layers alternating with 3 impact absorbing layers. Alternatively, the pack may comprise 4 witness layers alternating with 3 impact absorbing layers, where the final layer of the plurality of layers is a witness layer.

In the invention, the witness layers are preferably formed from a single sheet of flexible polymer and the sheet forms windings around the impact absorbing layers and, optionally, an external winding around the witness pack. This provides a simple method of pack construction which holds the layers together and may obviate the need for separate means to join the layers together. Typically, pack embodiments where the witness layers are formed from a single winding of flexible sheet polymer comprise a final layer which is witness layer.

In order to ensure secure windings, the first witness layer may be a double winding of sheet polymer and hence, twice the thickness of the other witness layers.

The witness pack may comprise a frame around the perimeter of the alternating layers and/or other fixings adapted to hold the pack together. Any suitable fixing may be used, for example nails or bolts through the layers of the pack.

The witness pack may additionally comprise a retaining layer towards the rear face of the witness pack to capture and retain any residual (high velocity or high energy) projectiles. The retaining layer may be a strawboard layer, for example nominal 3.8 mm strawboard (Defence Standard 93-59, type D). The retaining layer may alternatively comprise materials such as fibre board, chipboard, plywood, thin metal sheet and so on.

According to a second aspect of the invention, there is provided a witness system comprising two or more witness packs according to the first aspect positioned side by side in an array. The witness system may form a wall for use in blast trials. Multiple small witness packs may be tiled to cover large or complex areas. The system may be deployed inside and/or outside vehicle platforms, buildings and so on.

According to a third aspect of the invention, there is provided the use of a witness pack according to the first aspect for injury prediction. Preferably, the witness pack comprises a strike layer as described in relation to the first aspect and the witness pack is used for the analysis of eye penetration and skin perforation. In the preferred use embodiment, a simple and quick interpretation of risk of eye penetration and skin perforation can be obtained, valid across a wide range of low velocity and/or low energy fragments and/or non-metallic fragments.

A reduced thickness version of the witness pack may be used in scenarios where only the potential for corneal abrasions, risk of penetrating eye injury and/or risk of skin perforation need to be assessed. For such use, the witness pack may comprise 3 witness layers alternating with 3 impact absorbing layers. Alternatively, the pack may comprise 4 witness layers alternating with 3 impact absorbing layers, where the final layer of the plurality of layers is a witness layer. The purpose of reducing the number of layer is to simplify the pack construction and analysis. However, more witness layers and impact absorbing layers can be used if desired.

According to a fourth aspect of the invention, there is provided use of a witness pack according to the first aspect for testing the effectiveness of personal protective armour and/or personal protective equipment. The witness pack may be used to back protective materials and/or equipment to study overmatch or compare to an unprotected case.

Any feature in one aspect of the invention may be applied to any other aspects of the invention, in any appropriate combination. In particular product aspects may be applied to system and use aspects and vice versa. The invention extends to a system or method substantially as herein described, with reference to the accompanying drawings.

Brief Description of the Drawings

The invention will now be described, purely by way of example, with reference to the accompanying drawings, in which; Figure 1 shows a comparison of the predicted perforation velocities for metal spaced witness pack, strawboard, Post Mortem Human Subject (PMHS) eyes and skin; Figure 2 shows a schematic, exploded figure of a multi-layered witness pack according to a preferred embodiment of the invention having 7 witness layers and 6 impact absorbing layers; Figure 3 shows raw penetration data for the materials evaluated in comparison to the average animal data; Figure 4 is a calibration graph for a witness pack according to the invention showing the V50 velocity against normalised DoP over density using 3, 6 and 9 mm steel spheres and 3 and 6.35 mm glass spheres; Figure 5 shows a schematic, exploded figure of a multi-layered witness pack according to a preferred embodiment of the invention (MDFPIM V2.0); Figure 6 shows a schematic, exploded figure of a multi-layered witness pack according to a preferred embodiment of the invention (MDFPIM V2.1); Figure 7 shows ballistic testing results for MDFPIM V2.0 and V2.1, with layer 1 30 \Ras compared to the ideal model performance and 95% confidence limits; Figure 8 shows ballistic testing results for MDFPIM V2.0 and V2.1, with layer 2 Vsos compared to skin perforation performance metrics. Error bars represent the 95% Confidence Interval (Cl) on the \hos; and Figure 9 shows MDFPIM V2.0 normalised DoP over density calibration curve, showing 95% confidence and prediction intervals on the linear fit.

The drawings are for illustrative purposes only and are not to scale.

Detailed Description

Comparative Example

Known types of witness pack were assessed for their utility in investigating the injury risk from penetration by soil and stones projected from buried explosive devices. Short descriptions of each of these attempts are given below.

Five layer strawboard packs (each layer nominal thickness 3.81 mm) were used to observe whether or not fragments generated from buried Home-Made Explosive (HME) devices penetrated the packs. Packs were placed 2 m from a 10 kg device, which was buried 300 mm in the soil. Two firings were conducted and, in both cases, the packs showed evidence of being impacted. However, there were no embedded fragments and no perforations of the first strawboard layer. Nevertheless, experts observing the trials were of the opinion that an unprotected person standing at the pack location would have suffered severe soft tissue injury.

Metal spaced witness packs were constructed from the layered metal sheets listed in Table 1 interspaced with 25 mm Polystyrene to BS 3837-1:2004.

Plate number Material Plate thickness (mm) 1 Aluminium (1050A temper H14 to BS EN 485-2) 1.0 2 1.0 3 3.0 4 Mild steel CR4 to BS 1.5 1449-1.1:1991 1.5 6 1.5 7 1.5 Table 1: Metal plates used in witness packs.

A pack 2 m tall and 0.5 m wide was placed at 1.5 m from a 5 kg HME device buried to 300 mm. After the blast, the pack showed evidence of impacts and some dents, but no perforations of even the first aluminium layer. Again, subject matter expert opinion was that an unprotected person standing at that location would have suffered severe soft tissue injury.

In an attempt to have a material that could be more easily penetrated, a witness pack comprising expanded polystyrene was used for a range of buried HME tests. This pack was constructed from 10 layers of 10 mm polystyrene-divinyl benzene (SDVB) at a density of 15 kg m-3, backed with two layers of 1 mm aluminium 1050A, each separated by 10 mm polystyrene. Whilst it was penetrated by the fragments, there were several major issues with this pack: it was overly easy to penetrate, it was frangible and it was frequently massively disrupted during the blast, preventing meaningful analysis.

Figure 1 shows a comparison of the predicted perforation velocities for the strawboard (first layer -layer 1), metal spaced witness pack (first layer -layer 1), PMHS eyes and PMHS thigh skin.

Figure 1 shows the regions where the strawboard or the metal spaced witness pack will not allow assessment of potentially injurious fragments. Layer 1 of the strawboard requires approximately 3 times the velocity required to perforate skin before it is perforated. The metal spaced witness pack is very dependent on the fragment sectional density as to whether the velocity to perforate layer 1 is much higher, or similar to that required to perforate PHMS skin.

Selection of impact absorbing layer material Following the poor performance of polystyrene-containing witness packs in trials, other impact absorbing materials were assessed. Non-frangible materials were chosen so that they would not fracture during blast tests. Materials were considered based on whether they might provide a similar penetration response to animal muscle tissue.

Testing was performed on witness packs comprising between 4 and 10 layers of the impact absorbing material (nominal thickness between 9 and 10 mm), and formed using the preferred inter-wound polymer sheet construction for the witness layers. The polymer selected for the witness layers was 62.5 pm (250 gauge) thick polythene sheeting. All witness packs were constructed to 300 mm width, 500 mm height.

Figure 2 shows an exploded view of a typical witness pack 9 used in the testing.

The pack has 7 witness layers 1 alternating with 6 impact absorbing layers 2, and the witness layers are formed from a single sheet of polythene sheeting 3 wound around each impact absorbing layer 2 and the perimeter of the pack. The front face, or strike face 4, of the witness pack (the first witness layer) is formed from a double layer of polythene due to the pack construction. It should be noted that Figure 2 is not to scale, and is expanded along the pack depth A and width B to show the winding construction. Witness packs according to the invention typically have a tightly bound construction with no gaps or substantially no gaps between the impact absorbing and witness layers. Figures 5 and 6 below are similarly expanded.

Nine different material combinations were evaluated, including a selection of low, medium and high density polystyrene-divinyl benzene (SDVB), a cast silicone and a selection of foams. The materials used, along with basic properties are shown in Table 2.

The packs were impacted with individual fragments (3, 6 and 9 mm steel spheres and 3, 6.35 and 20 mm glass spheres) over the velocity range 30 to 500 ms-1. Impacts were conducted into each pack, ensuring the impact location was distant from previous shots. A pack of each material was also simultaneously impacted by 20 g of 3 mm steel spheres (approximately 175 in number) at an average of 115 m s-1 to ensure it would hold up against simultaneous penetrations and could be subsequently analysed.

Analysis of the number of layers of the polythene sheeting perforated for each shot was converted into a discrete depth of penetration (DoP) into the pack (number of layers perforated times individual layer thickness). That was then converted into the Normalised DoP over density.

The normalised DoP over density is calculated by dividing the DoP by the projectile diameter, then additionally divided by the projectile density to get a normalised DoP over density' function. This function enables comparisons between projectiles of different diameters and densities.

Three materials were found to give a good match to the animal penetration data, as shown in Figure 3. Of these, the cast silicone Dragon Skin® material was less preferred due to production time and thickness variability of the material. Neoprene and silicone closed cell foams were found to have no statistically significant difference to the animal perforation data at the 95% confidence level (cl) in terms of a comparison of their best fit lines over the velocity range investigated.

The closed cell neoprene foam was chosen for further pack development because there was more data from this initial testing, and the material is readily available and specifiable (in terms of its mechanical properties).

Retarding material Layer thickness (mm) Nominal density (kg m-3) Low Density (LD) SDVB 15 10 Medium Density (MD) SDVB 10 15 Medium Density (MD) SDVB 15 15 High Density (HD) SDVB 10 20 High Density (HD) SDVB 15 20 Closed cell neoprene foam 9.5 160 Closed cell silicone foam 10 200 Closed cell polyethylene foam (Plastazote) 10 45 Cast silicone sheet (Dragon SkinO) (a silicone rubber) 10 1080 Table 2: Materials for initial pack down selection Witness pack calibration Calibration was conducted on a witness pack (denoted MDFPIM V1.0) constructed according to Figure 2, with 6 layers of neoprene foam as the impact absorbing layers and inter-wound polythene sheeting as witness layers.

The calibration was conducted with a 3, 6 and 9 mm steel spheres and 3 mm and 6.35 mm glass spheres over the velocity range required to bracket no penetration up to full penetration of the pack, with an average of 25 shots for each projectile. A total of 132 valid shots were completed, allowing calculation of 27 V50 velocities for different fragment and layer combinations. A V50 calculation gives a better statistical representation of the penetration data, which would otherwise be very sensitive to the velocities of the shots conducted during the experiment if raw data was used to construct the calibration relationship (and would appear to have steps in velocity at higher normalised DoP over density values).

The model was calibrated whilst backed 25 mm around its periphery (e.g. as a frame). Figure 4 shows the V50 velocity required to perforate each layer, converted into the normalised DoP over density. All the points agree well to a linear fit with an R2 of 0.970.

The large confidence interval on the highest V50 was due to it being at the limit of the weapon capability, not allowing the desired velocities to be achieved in order to reduce the confidence interval, rather than large variability in that data point.

The velocity required to perforate the second witness layer (layer 2) of the witness pack was found to be within the scatter of the skin perforation data for PMHS skin, over the range of projectiles used in the calibration. This means any projectiles not perforating through witness layer 2 are not likely to be injurious and can be ignored. The model is unlikely to miss capturing any injurious projectiles that hit the pack.

Witness packs for assessing eve and skin injury The above-described witness pack is suitable for injury assessment, but the assessment requires back calculation of fragment velocities from their depth of penetration and recovered mass.

Witness packs denoted MDFPIM V2.0, MDFPIM V2.1 and MDFPIM V2.2 were constructed to allow quick and direct estimations of the risk of eye injury (eye penetration and optional corneal abrasion) and skin perforation, in addition to estimating impact velocities based on fragment mass and penetration depths into the pack.

All witness packs had the inter-wound sheet polymer construction. MDFPIM 5 V2.0, denoted 19 in Figure 5, was constructed as follows: * 9 layers of neoprene closed cell foam as impact absorbing layers 10, each 10 mm thick, density 150 ± 10 kg/m3.

* The last impact absorbing (neoprene foam) layer had an additional piece of strawboard in front of the neoprene as retaining layer 11.

* 62.5 pm (250 gauge) thick polythene sheeting 12 was placed between each neoprene layer to act as a witness layers 13, with a double layer (witness layer 1) at the impact (strike) face 14 due to the inter-wound construction.

* A layer of layer of 23 ± 2 pm polyester film, with optional double sided polyacrylate adhesive, was applied directly to the strike face 14 as a strike layer 15.

MDFPIM V2.0 is shown schematically, in exploded view, in Figure 5. MDFPIM V2.0 allowed back calculation of velocity following an impact from a penetrating fragment. The additional polyester film layer additional allowed the assessment of eye injuries by mimicking the penetration response of the human cornea.

MDFPIM V2.1 is shown in Figure 6, as an exploded view, and was a thin version of the MDFPIM V2.0 described above. The thin version 20 comprised 3 layers of the neoprene foam as impact absorbing layers 21, with 62.5 pm (250 gauge) thick polythene sheeting 25 acting as witness layer 22 between each layer, and with a double layer (witness layer 1) over the impact (strike) face 23. The 23 ± 2 pm polyester-film was also applied directly to the strike face as strike layer 24. This was to provide a pack in situations where only eye penetration and/or skin perforation assessments are required. The V2.1 pack provides the same response as V2.0, but reduces material requirements.

For applications where corneal abrasion injuries that indicate potential for low level' eye injury are of interest, a third pack version was constructed. MDFPIM V2.2 is identical to V2.1, apart from the feature that the polythene layer encasing the impact face is transparent and the remainder of the polythene is black. All the polythene is to the same grade and changing its colour was merely to assist with the analysis process. The MDFPIM V2.2 allowed assessment of eye penetration and/or skin perforation risk. The construction of the MDFPIM V2.2 is not illustrated.

For assessment of corneal abrasion injuries, any particles adhering to the polyester film after testing indicate a potential abrasion risk. Corneal abrasion is 15 a very low level injury, and will normally self-heal with no long term consequences.

Witness pack testing for eye and/or skin iniury The MDFPIM V2.0 and V2.1 were tested to determine how their penetration response related to penetrating eye injury, skin perforation and to generate a calibration curve for penetrations deeper into the model.

MDFPIM V2.2 was not specifically tested as it was assumed that there was no performance difference between the different colours of the polythene sheeting.

Ballistic testing of the MDFPIM V2.0 for the eye penetration response, skin perforation response and penetration depth calibration deeper into the model used many of the same projectiles. Table 3 shows the projectiles and their properties (listed in order of increasing mass), identifying which were used in each of the eye penetration, skin perforation and DoP calibration assessments for the MDFPIM V2.0 and V2.1.

Geometry Diameter(mm) Material Mass (9) Density (g cm-3) 0 c 0_ >-. co o I-I-I in Sphere 1.6 Acrylic 0.003 1.24 X Sphere 1.0 Steel 0.004 7.85 X X X Sphere 4.2 HDPE 0.03 0.95 X X Sphere 3.0 Glass 0.04 2.50 X X X Sphere 3.2 Aluminium 0.05 2.70 X X Sphere 5.0 Cellulose acetate 0.08 1.30 X X Sphere 5.7 HDPE 0.09 0.95 X X Sphere 3.0 Steel 0.11 7.85 X X X Sphere 3.0 Tungsten carbide 0.22 15.63 X X Sphere 6.0 Glass 0.29 2.50 X X X Sphere 6.4 Glass 0.34 2.50 X Sphere 4.4 Steel 0.35 7.85 X X X Cube 4.0 Steel 0.50 7.85 X X Irregular 7.5 Limestone 0.50 2.30 X Sphere 6.0 Steel 0.89 7.85 X X X Sphere 9.0 Glass 0.95 2.50 X X X CN FSP 5.4 Steel 1.10 7.85 X Sphere 9.0 Ceramic 1.45 3.80 X X X Cube 6.0 Steel 1.68 7.85 X X Sphere 9.0 Steel 3.02 7.85 X X X Cube 8.0 Steel 3.97 7.85 X Cube 12.7 Aluminium 5.44 2.70 X X Sphere 12.7 Steel 8.30 7.85 X X X Geometry Diameter Material Mass Density 0 c ri >, o (mm) (9) (g cm-3) Lu co C Sphere 20.0 Glass 10.50 2.50 X X Table 3: Projectiles and properties used to assess the MDFPIM V2.0 for each of the eye penetration, skin perforation and DoP calibration tests.

After the firings had been completed on each pack, analysis of the number of layers (of the polythene sheeting) perforated for each shot was recorded. The Vso's were calculated using the statistical program R using a bias reduced generalized linear model. This enabled the V50 to be calculated, along with 95% confidence intervals on the measurement.

Results of eye penetration ballistic testing A total of 504 fair shots were completed across the different targets and eleven different projectiles (all spheres). This included repeating the 3 mm steel sphere 15 and 9 mm ceramic sphere testing against the MDFPIM V2.1.

The Vso data are shown graphically in Figure 7, compared to the ideal model performance and 95% confidence limits on the ideal performance, the human eye performance reported by Kennedy et a/ (KENNEDY, E.A. et al. Risk functions for human and porcine eye rupture based on projectile characteristics of blunt objects. Strapp Car Crash Journal, 2006, Vol 50, pp 651). The error bars on the V5odata are the 95% confidence intervals.

Figure 7 shows that for the 11 projectiles tested for the MDFPIM V2.0 over the sectional density range 0.08 to 6.55 g/cm2 and the two projectiles for the MDFPIM V2.1, the physical model performs very closely to the ideal performance curve (indicating a 50% risk of human eye penetration). There is no statistically significant difference at the 95% confidence level between the performance of the model V2.0 and V2.1 for eye penetration assessments (P>0.5).

A power fit applied to the MDFPIM V2.0 layer 1 data sits just below the ideal human eye performance (and below the lower 95% confidence interval line for projectile sectional densities <3 g/cm2). Although this indicates the MDFPIM V2.0 (and V2.1) slightly overestimate the risk of eye penetration (i.e. the MDFPIM V2.0 layer 1 will be perforated at slightly lower velocities than average human eyes), this is the desired way round for an injury model that could be applied to safety cases. The model needs to ensure that if it determines a scenario to be 'safe', then the residual risk of being incorrect minimised.

Due to this general requirement to err on the side of caution for application to safety case type assessments and within the human eye ideal performance 95% Cl for projectile sectional densities >3 g/cm2, it was not deemed necessary to tune the performance of the MDFPIM V2.0 further to get closer to the human eye ideal performance.

MDFPIM V2.0, V2.1 and V2.2 validation for skin perforation The V50 response of layer 2 (second witness layer, positioned behind first impact absorbing layer) of the MDFPIM V2.0 and V2.1 was also determined.

Perforation of layer 2 was taken to indicate a 50% risk of skin perforation.

The same methods and analysis procedures were used as detailed for the eye penetration assessment. 19 different projectiles were used to assess the skin perforation performance of the model, with details listed in Table 2.

2 separate performance metrics were used to bound the risk of skin perforation for different scenarios, which could be applied to the MDFPIM V2.0: Vso performance curve for skin perforation of PMHS adult thigh, skin intact and fresh. This could be taken to represent a reasonable most likely case' for an adult population or military personnel and may be most suitable for situations where the risk does not want to be over-predicted, i.e. lethality or effectiveness type prediction on a military population.

Vs° performance curve for skin perforation of PMHS child thigh, skin intact and fresh. This could be taken to represent the 'worst likely case' civilian or vulnerable group population and may be most suitable for situations where the risk does not want to be under-predicted, i.e. a collateral damage type prediction for a vulnerable population or general public. [Ref: Missliwetz, J. (1987). "Zur Grenzgeschwindigkeit bei der Haut. (E me experimentelle ballistische Untersuchung mit Geschossen vom Kaliber 4 mm und 4.5 mm) [The limit velocity to the skin (An experimental investigation with ballistic projectiles calibre 4 mm and 4.5 mm)]." Beitr Gerichtl Med 65: 411-432.] The Vso values calculated from the experimental data for each of the MDFPIM versions are shown in Figure 8 compared to the performance metrics detailed 20 above (16 Vsos for the MDFPIM V2.0 with spheres, 3 Vsos with cubes and 2 \Ras with spheres for the MDFPIM V2.1).

Figure 8 shows that for the 19 projectiles tested for the MDFPIM V2.0 over the sectional density range 0.22 to 6.55 g/cm2 and the two projectiles for the MDFPIM V2.1, witness layer 2 of the physical model performs very closely to the skin performance curves. A power fit to the combined sphere and cube MDFPIM V2.0 data (R2=0.964) provides a very good match (within 5.5 m s-1 for the projectile sectional densities tested) to the performance curve for the vulnerable civilian population.

The witness pack may slightly over-estimate the risk of skin perforation for an adult population. For a vulnerable civilian / worst case general population, the physical models give a good prediction of the (50%) risk of skin perforation.

Given the inherent variability in PMHS skin data on which the ideal performance metrics were based, coupled to the scatter in the MDFPIM V2.0 individual V50 data points, the MDFPIM V2.0 layer 2 performance is within the desired performance range and is less variable than real skin.

Materials for strike layer A selection of materials were considered as potential alternatives to the 23 pm sticky polyester film to go on the front of the MDFPIM to modify its penetration response. The polyester film was the first material evaluated and gave good agreement to the required eye and skin penetration response as detailed above, with the additional benefit of allowing assessment of potential for corneal abrasion. Further testing on alternative materials was not conducted, but the following materials are considered to be alternatives based on matching material properties to the human cornea: * 0.125 mm thick silicone elastomer film * 0.45 mm thick silicone elastomer film * 0.5 mm thick silicone elastomer film * 0.6 mm thick silicone elastomer film * Silicone Sheet -30 shore A, 1 mm thick * Silicone Sheet -40 shore A, 1 mm thick * Silicone Sheet -50 shore A, 0.5 mm thick Also tested was a version of the MDFPIM V2.0 (with the 23 pm sticky polyester 30 film) using 125 pm polythene sheeting instead of the 62.5 pm, impacted with the 3 mm steel sphere. This was found to raise the penetration response of layer 1 above the upper limit of the eye penetration performance corridor and the V50 for layer 2 above the adult or military population skin performance. Due to this, the 125 pm polythene was discounted for use in the model, a thickness less than 125 pm is preferred.

Calibration curves MDFPIM V2.0 was calibrated by firing a variety of different projectiles at the model over a range of velocities. Velocities were from as slow as was achievable with the weapon system, up to full perforation of the pack (or the maximum velocity achievable if full perforation could not be achieved). Not all projectiles were evaluated over this entire velocity range.

After the firings had been completed, the MDFPIM V2.0 was dismantled and the maximum number of layers perforated was recorded for each shot.

This calibration testing was all performed at ambient test temperatures (24±3°C) 15 on the MDFPIM V2.0 constructed from foam from the same batch. The foam used to construct the models had been stored for 1 year since production, prior to this testing.

22 different projectiles were used with details listed in Table 3. A total of 728 valid shots were conducted against the model. This includes shots conducted as part of the eye and skin layer performance assessment). From this, the V50 for each layer of the pack was calculated (where sufficient data was available), resulting in 112 separate \hos. Of these, some were for layers 1 and 10, which have been excluded from the calibration curve, leaving 101 valid \hos. Witness layer 1 Vsos were discounted as they had a 0 depth of penetration and would have skewed the calibration fit (different projectiles will have a different velocity required to perforate layer 1 in Figure 6. Witness layer 10 Vsos were discounted due to the strawboard layer and discontinuity in material types in the model.

The number of layers perforated for each Vso were used to calculate the Normalised DoP over density and plotted against the corresponding V50. The calibration curve for the MDFPIM V2.0 is shown in Figure 9. Error bars are the 95% confidence intervals on the individual Vs() assessments.

Confidence and prediction intervals on the linear fit for all projectiles are shown in Figure 9.

The MDFPIM V2.0 calibration curve is given by Equation 1 (with an R2 value of 0.952).

Normalised DoP -0.0175v -1.312 Equation 1: MDFPIM V2.0 generic calibration equation Where: p is the projectile density (g cm-3) v is the impact velocity (m sA) The residuals of the linear regression of Equation 1 are normally distributed. Curves representing the 95% confidence intervals can be approximated by Equation 2 and Equation 3.

Normalised DoP -1.91 x 10-6v2 + 0.018v -1.42 Equation 2: Lower 95% confindence interval for the MDFPIM V2.0 generic calibration equation Normalised DoP -1.91 x 10-6v2 + 0.017v -1.12 Equation 3: Upper 95% confindence interval for the MDFPIM V2.0 generic calibration equation The curves representing the 95% prediction intervals can be approximated by Equation 4 and Equation 5.

Normalised DoP 4.90 x 10-7v2 + 0.018v -1.94 Equation 4: Lower 95% prediction interval for the MDFPIM V2.0 generic calibration equation Normalised DoP -4.90 x 10-7v2 + 0.017v -0.680 Equation 5: Upper 95% prediction interval for the MDFPIM V2.0 generic calibration equation Where: Normalised DoP = 10(L -1) Equation 6: Normalised DoP related to MDFPIM layer number, where each layer is nominally 10 mm thick d = projectile diameter (mm). For cubes, the average projected length in 3D is used (given by Equation 10). For irregular fragments, the diameter of an equivalent mass and density sphere is used.

L = layer number of the MDFPIM polythene sheeting.

If the layer thickness for a given batch of foam used in the MDFPIM is measured (and within the allowed limits of 10±1 mm), the actual layer thickness can be used in Equation 6 in place of the constant 10.

The MDFPIM V2.0 calibration curve given by Equation 1 can be used to estimate impact velocities within the validation limits stated in Table 4, where the maximum layer perforated by the projectile and its mass are known. This normally requires removal and weighing of each fragment with its associated deepest perforation in the model.

Mass Diameter Density Geometry Velocity Deformation (9) (mm) (g cm-3) (m s-1) Min 0.004 1 0.95 Cylinder, Fragment Non-deforming, cube, dependent non-sphere fragmenting (irregular) Max 10.5 20 15.6 Table 4: Validation limits for projectiles in the MDFPIM V2.0 Due to the discrete nature of the DoP measurements (related to the nominal mm layer thickness), the predicted impact velocities from the model based on a perforation to layer n (Ln) can be easily bounded by the predicted impact velocity for the layer before and after (Lni-i).

When quoting predicted velocities, it is advised that the bounding velocity for the layer before and after also be given. This can be achieved by rearranging Equation 1 for velocity: 571.43(L -1) v = +74.49 pd Equation 7: Predicted impact velocity to the MDFPIM V2.0, rearranged from Equation 1. Where

When using the MDFPIM V2.0 for trials where the properties of the impacting fragment aren't known prior to the testing, the fragment properties have to be determined by recovering them from the model. In these cases, it is likely to be more practical to use the fragment mass and density. Equation 7 can be modified using the fragment (average) diameter estimated based on the measured fragment mass and (assumed) density, for a spherical geometry fragment, using Equation 8: d = 20 HT31)173 p Equation 8: Estimated diameter of a spherical fragment based on mass and density.

Combining Equation 7 and Equation 8, the mass and (assumed) density of each fragment can be used to predict the impact velocity, using Equation 9: 46.06(L -1) v -+74.49 m1/3p2/3 Equation 9: Predicted impact velocity to the MDFPIM V2.0, based on recovered fragment mass, (assumed) density and maximum layer perforated. Where

Equation 7 or Equation 9 can also be used to estimate the maximum velocity of a given fragment to remain analysable within the MDFPIM V2.0 (the velocity which equates to perforations up to layer 9).

The calibration equation for the MDFPIM V2.0 (Equation 1) using the normalised DoP over density function accounts for projectiles of different diameters and densities and can be used to collapse data onto a single curve with velocity. There is no need to explicitly consider projectile geometry (although will be most accurate for spherical geometries or for fragments with similar length the diameter ratios), as long as projectile diameter is properly accounted for: For cubes, the average projected length in 3D is used: d" = 3d Equation 10: Average diameter (projection length) of a randomly 10 orientated cube in 3D.

Analysis methods After use, the witness pack of the invention may be analysed by physical recovery of fragments. One preferred analysis route is: * For the assessment of corneal abrasion risk, identify any debris adhering to the strike face (sticky polyester film). This indicates the potential for a corneal abrasion injury.

* For assessment of eye penetration risk, count the number of perforations to witness layer 1 (the first witness layer) of the witness pack as embodied by MDFPIM V2.0. The way the MDFPIM V2.0 has been developed means perforation to layer 1 indicates a (50%) risk of eye penetration, independent of projectile mass, density, diameter and geometry.

o The number of perforations per unit area to layer 1 can be scaled from the pack area to the presented area of the eye to estimate the risk of an injurious projectile hitting the eye. This is only valid at the location where the MDFPIM V2.0 was deployed during the test.

* For assessment of skin perforation risk, count the number of perforations to layer 2 of the MDFPIM V2.0. The way the MDFPIM V2.0 has been developed means perforation to layer 2 indicates a (50%) risk of skin perforation, independent of projectile mass, density, diameter and geometry.

o The number of perforations per unit area to layer 2 can be scaled from the pack area to the presented area of a person to estimate the worst case risk (assuming no clothing) of an injurious projectile hitting the body. This is only valid at the location where the MDFPIM V2.0 was deployed during the test.

* For projectile impact velocity estimates: Dismantle the pack, starting from the front layer. For every hole in the polythene sheeting, locate the corresponding hole or fragment in the underlying foam layer.

o Once a fragment is located, recover it, weigh it and recorded the maximum (polythene sheet) layer it perforated.

o Track through holes into deeper layers of the pack, progressing one layer at a time until all fragments have been accounted for. It may be beneficial to record fragment co-ordinates in terms of their impact location on the pack. The need for this will be dependent on the type of output required from the model.

o For each fragment, the impact velocity can be predicted (Equation 7 or Equation 9).

o Each impact velocity that is predicted can be bounded by the predicted impact velocity for the layer before and after (Ln±1). 25 Equation 7 or Equation 9 discounts the need to consider fragment shape. However, estimates using Equation 7 or Equation 9 may under-predict the impact velocity if the fragments have high length: diameter ratios and impact end on, or over-predict the impact velocity if the fragments are irregular or have high length: diameter ratios and impact flat face on.

An alternative analysis method is image analysis, if there are very large numbers of fragment impacts and/or the mass of the impacting fragments is known without having to physically recover each individually, image analysis can be used to identify and characterise the holes in the polythene witness sheet for each layer. To summarise the image analysis method: * A scaled photograph of the polythene sheet is taken where any perforations can be easily differentiated in terms of the pixel intensity.

* This is repeated for each layer of the MDFPIM perforated.

* Image processing software can then be used to count the number of perforations, as well as measure the average diameter and x-y impact co-ordinates.

* The data for the perforation in the shallower layers need to be discounted so that only the data for the maximum perforation for each fragment remain. The x-y impact co-ordinates (in addition to the measured diameter) can be used to locate the corresponding hole in a different layer.

* The density of the fragment related to each hole needs to be estimated from knowledge of the threat and scenario. This along with the measured average diameter from the image analysis and the corresponding maximum layer perorated can be used as inputs to Equation 7 or Equation 9.

* As for the physical fragment recovery method, each impact velocity that is predicted using Equation 7 or Equation 9 can be bounded by the predicted impact velocity for the layer before and after (Ln+i).

MDFPIM V2.1 or V2.2 can be used to provide assessment of eye penetration risk and skin perforation risk in the same manner as V2.0 (either by physical inspection and hole counting or by image analysis).

MDFPIM V2.2 is optimised for corneal abrasion assessment, in addition to assessment of eye penetration risk and skin perforation risk, with the transparent polythene for layer la (the foremost of the double thickness polythene sheet comprising layer 1).

It will be understood that the present invention has been described above purely by way of example, and modification of detail can be made within the scope of the invention. Each feature disclosed in the description, and (where appropriate) the claims and drawings may be provided independently or in any appropriate combination.

Moreover, the invention has been described with specific reference injury modelling the fields of munitions and explosives. It will be understood that this is not intended to be limiting and the invention may be used more generally. For example, the invention may be used more generally in the security and military fields, and may be used in civil applications such as structural analysis, oil and gas exploration or civil engineering. Additional applications of the invention will occur to the skilled person.

Claims (19)

1. CLAIMS 2. 3. 4. 5.A witness pack for detecting low velocity and/or energy and/or nonmetallic projectile fragments, the witness pack comprising a plurality of alternating witness layers and impact absorbing layers, wherein the first layer of the plurality of alternating witness layers and impact absorbing layers is a witness layer and wherein the material composition of the witness layers and impact absorbing layer(s) are individually selected to perform their respective functions.
2. A witness pack according to claim 1, wherein the witness layers are formed from a flexible polymer, preferably a flexible polymer selected from polyethylene, polyvinyl chloride and polypropylene.
3. A witness pack according to claim 1 or claim 2, wherein the witness layers each have a thickness up to 125 micron.
4. A witness pack according to any one of claim 1 to 3, wherein the impact absorbing layers are formed from a material having a density in the range 50 to 2500 kg/m3, preferably in the range 100 to 1500 kg/m3, more preferably in the range 150 to 500 kg/m3.
5. A witness pack according to any preceding claim, wherein the impact absorbing layers are formed from a rubber or foam rubber material, preferably a material selected from the group consisting of neoprene foam, silicone foam, silicone rubber and butyl rubber.
6. A witness pack according to any preceding claim, wherein the impact absorbing layers have a thickness in the range 5 to 25 mm.
7. 7. A witness pack according to any preceding claim, wherein the number of witness layers lies in the range 3 to 10.
8. 8. A witness pack according to claim 7, wherein the pack is for general use, and wherein the number of impact absorbing layers is 9.
9. 9. A witness pack according to claim 7, wherein the pack is adapted to model skin and eye penetration, and wherein the number of impact absorbing layers is 3.
10. 10. A witness pack according to any preceding claim, wherein the witness layers are formed from a single flexible sheet of polymer and wherein said single sheet forms windings around the impact absorbing layers and, optionally, an external winding around the witness pack.
11. 11. A witness pack according to any preceding claim, wherein the pack comprises a front face and wherein an additional strike layer is applied to the front face to model corneal penetration.
12. 12. A witness pack according to claim 11, wherein the strike layer comprises a polymer film, for example polyester film.
13. 13. A witness pack according to claim 12, wherein the strike layer has a thickness in the range 15 to 30 micron, preferably 20 to 26 micron.
14. 14. A witness pack according to any preceding claim additionally comprising a strawboard layer towards the rear face of the witness pack to retain any residual (fast) projectile fragments.
15. 15. A witness system comprising two or more witness packs according to any preceding claim positioned side by side in an array.
16. 16. Use of a witness pack according to any preceding claim for injury prediction.
17. 17. Use according to claim 16, wherein the witness pack comprises a strike layer applied to the front face of said pack.
18. 18. Use of a witness pack according to any preceding claim for testing the effectiveness of personal protective armour and/or personal protective equipment.
19. 19. A system or method substantially as described herein, with reference to the accompanying drawings.