Assessment of polyurethane spheres as surrogates for military ballistic head injury

The aim of this project was to assess if SYNBONE® spheres (SYNBONE AG, Neugutstrasse 4, 7208 Malans, Switzerland) are suitable for simulating military ballistic head injury at engagement distances of 50 to 100 m.

Much of the ground work in simulating cranial gunshot injury with synthetic models has been done by Thali and colleagues [1‐3]. In his initial paper [1], Thali expresses concern that the physical mechanisms behind ballistic trauma are poorly understood. To address this, the group built a synthetic head model using a layered polyurethane sphere (to simulate bone structure), a latex periosteum and a silicone cap to substitute for the scalp. Ten percent gelatine at 4 °C was used as a ‘brain’ fill. The model was shot with a broad range of ammunition (including 7.62 × 51 mm and 7.62 × 39 mm, but mainly 9 × 19 mm full metal jacket, FMJ, Luger) from 10 m and the authors concluded that the injuries created in the model were fully comparable to those seen in real incidents.

Further work with this model included impacting it with 9 mm Luger bullets to explore the underlying mechanisms for entrance wound characteristics [2] and a study of tangential gunshot head injury [3]. In the latter study, the bullets were fired directly at the synthetic skull with the latex periosteum layer (but not the silicone scalp) and found to produce realistic tangential injury and fracture patterns.

More recent work by Taylor and Kranioti has used SYNBONE® spheres to investigate execution style gunshot injuries [4]. The gelatine-filled models were shot at a range of 30 cm with seven different handgun ammunitions with the aim of detecting similarities and differences in wound characteristics for use in future investigations. The authors provide examples of two clinical cases (shot with 0.22LR and 0.45 ACP ammunition) that closely match the corresponding models.

Smith et al. [5] carried out a detailed analysis of the differences between injuries inflicted on a real bone compared to polyurethane bone substitutes. They used both flat plates of synthetic bone (5 mm thick) and spheres (5 and 7 mm wall thickness), and impacted them with a crossbow bolt, a ball fired from a black powder musket, and modern rifle ammunition (0.243″ Winchester Soft Point, velocity 905 m/s, and 7.62 × 51 mm NATO FMJ, velocity 853 m/s). The weapons were fired from a 2 m distance at the targets. They initially compared impacts on flat plates and empty spheres to see if the different shapes affected the response to impact, and compared these with shots into cattle scapulae. There were no gross differences between flat plates and spheres; both showed internal bevelling at the entry site. Differences between the synthetic and real bone are considered later in this paper.

Subsequent work involved spheres filled with 10% gelatine at 4 °C. The secondary and tertiary fracture patterns produced by modern firearms were generally consistent with those seen in published examples of real cranial trauma [5].

Nine SYNBONE® spheres (190 mm diameter, 6 mm wall thickness, thin rubber skin covering outer surface) were filled with ballistic gelatine of either 5,7, or 10% by mass (Fig. 1a). The gelatine was allowed to set at around 17 °C for 24 h. Other work by our group has shown no difference in fracture patterns in a skull model when impacted at a series of temperatures from 4 to 25 °C [6] and no difference with the above gelatine % fills.

The models were taken to the Small Arms Experimental Range, Cranfield University, Defence Academy of the UK (Fig. 1b), and placed 10 m from a no 3 Enfield proof mount fitted with an accurate barrel and shot with 7.62 × 39 mm Ukrainian mild steel core (MSC) ammunition (Soviet State Factory, Lugansk, manufactured 1967; mean impact velocity 654 m/s, SD 7 m/s). The ammunition was downloaded to achieve these velocities simulating engagement distances of around 50–100 m) [7].

Bullet velocity was tracked using a Weibel Doppler (Fig. 1c) and impacts filmed using two Phantom high-speed cameras (V12 from above, sample rate 20,000 frames per second, exposure time 5 μs, resolution 512 × 480; V1212 from the side, sample rate 34,000 frames per second, exposure time 10 μs, resolution 640 × 480).

Models 1–3 were each shot twice to assess their suitability for simulating more than one gunshot injury and assessing if the order of shot impact could be determined as described by Thali [1]. Models 4–9 were each shot once to assess entry and exit fracture patterns from one impact sequence.

The condition of the models in situ post-impact was captured using a Nikon D3200 DSLR camera fitted with an AF-S NIKKOR 18–55 mm lens.

The nine models were then examined by two Home Office forensic pathologists with extensive experience of assessing ballistic injury. The pathologists were invited to score the entry wound, exit wound, and overall fracture pattern using a 4-point Likert-type scale [8] (where 4 = exactly like a real injury, 3 = a lot like a real injury, 2 = a bit like a real injury and 1 = nothing like a real injury) and provide comment as needed. The scores are summarised in Table 1.

SYNBONE® spheres have been successfully used to simulate ballistic injury by a number of authors. Smith [5] found that the models produced different fracture patterns when impacted by the three projectile types described above, and the black powder carbine did produce a realistic key hole defect from a tangential impact, similar to that described by Thali [3]. Taylor and Kranioti [4] noted differences in the entry wound characteristics between the ammunition types tested with ‘entrance wound radius ...positively correlated with the caliber dimension’ and ‘the number of radiating and concentric fractures is also increasing with the caliber dimension’ [4].

There have been a number of observations regarding how the models differ from real injury. Smith [5] noted that the exit fracture patterns were different from real bone and described ‘stepped fractures where the radius of defect varied widely forming jagged corners around margins, unlike usually rounded/ovoid shapes in real bone’.

Taylor and Kranioti [4] also noted that the exit wounds in their model were larger than real injury.

The eviscerating injury seen in Fig. 4 was first described by the Swiss Surgeon Rudolf Ulrich Krönlein in relation to close range gunshots with the 1889 Swiss repeating rifle [9]. This effect was also seen by Thali et al. [1].

Our experience using 7.62 × 39 mm ammunition at simulated engagement ranges is that the bony injuries produced in our models were too comminuted and fractured in comparison with contemporary military bony injuries reviewed by the pathologists. This suggests that the model has less utility for this purpose than when used in the tests described by others [1‐5]. Of note, two of the 10% gelatine fill spheres had marginally higher scores when compared with the other fills, although the number of replicates for each experiment is small.