Performance level of protective equipment for soldiers and law enforcement officers is usually evaluated and set against ballistic, stab and shrapnel threats, but not against blast overpressure, which represents a real threat in modern armed conflicts. Air-filled organs such as the lung, ears and gastrointestinal tract are particularly susceptible to primary blast. So far, little is known about the efficiency of protective equipment against blast-induced thoracic damage. Nevertheless, few studies have demonstrated that wearing thoracic protective equipment (TPE) worsens the level of blast-induced body injury, depending on the equipment used (Phillips et al., 1988; Cooper et al., 1991; Gibson, 1994; Cooper, 1996; Prat et al., 2019), although this finding seems to be inconsistent across studies (Long et al., 2009; Wood et al., 2013). Phillips et al. (1988) and Gibson (1994) have shown that multiple plies of ballistic fabric can amplify the peak pressure of the transmitted blast wave to the body. Higher extent of lung injuries was also observed in animals wearing low impedance materials such as ballistic fabric compared with unprotected animals (Phillips et al., 1988; Cooper et al., 1991). However, placing a high density material (such as a ceramic plate) between the low impedance material and the incoming blast wave may help reducing blast-induced lung injury or mortality rate (Cooper et al., 1991; Cooper, 1996; Sekine et al., 2021).

The primary blast threat has not been considered in the development of protection systems to be used by soldiers and law enforcement personnel so far, mainly because no specification exists. In order to correctly evaluate the performance of existing and future TPE against shock-waves produced by detonations of improvised explosive devices (IED), studies on thoracic models, especially mannequins, have recently emerged (Jetté et al., 2004; Bass et al., 2005; Ouellet and Williams, 2008; Carboni et al., 2010; Awoukeng Goumtcha et al., 2014). So far, the aim of these studies has been to demonstrate that the response of thoracic models is influenced by the TPE, although with this approach, one can only test if a protection system is better or worse than a reference system, without getting any information on the severity level of lung injury. The thoracic wall peak acceleration from a specific torso rig or the MABIL (“Mannequin for the Assessment of Blast Incapacitation and Lethality”) (Cooper et al., 1996b; Jetté et al., 2004; Ouellet and Williams, 2008) were found to be suitable for such an evaluation. For the Hybrid III mannequin, it was the thoracic reflected overpressure and the chest acceleration (Jetté et al., 2004; Bass et al., 2005; Carboni et al., 2010).

Comparing the efficacy of different TPE using thoracic surrogates is a real progress in the process of designing optimal protections, but evaluating the level of protection they offer regarding the severity of lung injury would be more appropriate and informative. Indeed, such an evaluation could help find a good compromise between the weight of the systems and their ability to protect. Unfortunately, to evaluate the level of protection of TPE regarding the injury risk, an adapted injury criterion is needed. Different injury criteria have been defined based on blast wave characteristics (Bowen et al., 1968; Bass et al., 2008), maximum thoracic wall velocity (Axelsson and Yelverton, 1996), or the normalized work (Stuhmiller et al., 1996). Incident pressure or impulse-based criteria are limited to unprotected scenarios. Indeed, whatever the equipment worn, the incident blast wave will be the same, and, therefore, criteria based on blast wave characteristics cannot be used to evaluate the impact of wearing TPE on the risk of injury (Boutillier et al., 2016). Regarding the maximum chest wall velocity or the normalized work, pressure measurements from each facet of the Blast Test Devices behind the TPE are needed. Those pressure profiles would then be applied on their respective mathematical thorax model to extract the estimated lung injury severity. In practice, there is no limitation to use those criteria when evaluating the efficacy of a TPE. Nevertheless, some issues with the validity of the criteria, regardless of the use of protection system, have been described (Boutillier et al., 2016). Moreover, no validation of the determined injury risk was performed with experimental data when using a TPE, which would limits the use of those models to unprotected scenarios.

The objective of this study was to demonstrate the ability of the new mannequin called BOPMAN (for “Blast OverPressure MANnequin”), specifically designed to model primary blast exposure, to estimate the risk for lung injury in protected, and unprotected soldiers. In this study, we correlated data from BOPMAN to the severity of lung injury measured in blast-exposed live animals. Anesthetized swine wearing soft, hard or no ballistic protection were subjected to an open-field, right sided, high-intensity blast, while the homemade anthropomorphic mannequin BOPMAN was exposed in a front standing position to the same threat. Effect of TPE on animal’s injury severity, thoracic wall motion and on BOPMAN thoracic response was investigated.

Contemporary ballistic protections are designed to protect against projectiles. Given the high rate of blast exposure in modern military conflicts, the current study aimed to demonstrate the ability of the anthropomorphic mannequin BOPMAN to evaluate the efficiency of body protections against blast-induced lung injuries.

To achieve our goal, both anesthetized swine and BOPMAN were exposed face-on to a high-intensity Friedlander blast wave (mean characteristics: 434 kPa, 2.3 ms, 236 kPa ms). The present setting was designed to simulate, as close as possible, the primary blast wave generated by an IED exploding at the level of the ground. The recorded peak overpressure and phase duration situated the threat just near the “50% mortality” rate described by Bass et al. (2008) after weight adjustment. Interestingly, this did not result in any death at 1 h post-exposure. This raises the possibility that the assumption made by Bass et al. that mortality rates after blast exposure in an open-field and near a wall are similar may be erroneous. By contrast, our scenario leads to 99% survivability on Bowen’s curves, which is much closer to what was observed in our experimental setting. It is somewhat difficult to compare our threat data to Bowen and Bass’s curves, mainly because these curves were based on death rates measured over a 24 h-period, whereas we only measured death rates in our experiments 1 h post-blast. Because blast-induced thoracic alterations may be associated with pulmonary hemoptysis, which can lead to delayed mortality due to progressive asphyxiation, we may have missed few deaths in our animal cohorts. Moreover, bowel injuries can occur and lead rapidly to death such as peritonitis case.

Thoracic injuries measured in our study were consistent with the pattern previously described (Mayorga, 1997), although the right lung only was affected in our experimental conditions. This can be easily explained by the side facing position of the animal in our experimental setting. The lung weight ratio RL/LL was chosen to describe injury severity as it is known to be fairly constant in healthy swine independently of the body weight (e.g., the control group results), and more representative of the extent of lung contusion, given the animals are exsanguinated prior to measurement. Indeed, exsanguination enables to measure intra-alveolar blood accumulation only. Therefore, RL/LL is a much more precise tool for measuring the extent of lung contusion than total lung weight to animal weight ratio. It is noteworthy that blast-induced injuries were restricted to the lung side facing the blast wave in our experimental setting, therefore, abnormal RL/LL data were exclusively due to right lung contusion. We show, here, that NIJ level IV ballistic protection (Ballistic Resistance of Body Armor National Institute of Justice, Standard-0101.06), composed of soft protection and additional ceramic hard plate (P3), provided better lung protection against high-intensity primary blast overpressure than soft protection (NIJ level IIIa, P2). In fact, the level IV protection tested here offered a very good protection against primary blast injuries, as shown by the almost total lack of lung injuries in the P3 group, assessed by ASS and lung weight ratio. By contrast, animals protected by a soft ballistic armor presented lung injuries that were somewhat more severe than those measured in animals with no protection. Previous studies already suggested a possible amplification of the transmitted pressure and an increase in primary blast injuries behind soft body protections (Phillips et al., 1988; Cooper et al., 1991; Gibson, 1994; Cooper, 1996). Indeed, it was demonstrated that multiple plies in ballistic fabric can amplify the peak pressure of the transmitted blast wave, when exposed to short-duration low-intensity blast (Phillips et al., 1988; Cooper et al., 1991; Gibson, 1994) or long-duration high-intensity blast (Cooper et al., 1991). Philips et al. (1988) compared the effects of blast exposure on sheep wearing a soft ballistic vest or no protection and exposed to long-duration low-to high-intensity shock waves. A similar study was conducted by Cooper et al. (1991) on rats and swine which were exposed to short-duration high-intensity blast waves while covered with foam. Overall, these studies demonstrated that lung injury (lung weight/body weight ratio) was increased in animals wearing a low-impedance material, as compared to those that were not wearing any protection. Interestingly, in Cooper and coll. studies, it was demonstrated that, by placing a high-density material between the foam and the incoming blast wave (which is similar to our P3 level protection), a reduction of the level of blast-induced body injury could be observed. It was hypothesized that such an attenuation could result from a “stress-wave decoupling”, which can be achieved by mounting material with a high acoustic impedance (for example the ceramic) onto a material with low acoustic impedance (for example, foam or rubber). It is noteworthy that not all studies to date noticed this amplification phenomenon behind fabric-based armors. Long et al. (2009) exposed anesthetized rats outside a shock tube and concluded that Kevlar vest reduces the mortality rate of the animals. However, rats thorax greatly differs from human’s one, and no indication on the dimensioning of the vest as compared with the size of the animal were given. Wood et al. (2013) also noticed pressure attenuation on a rigid cylinder representing the human thorax when wearing a NIJ II (woven Kelvar vest) and IV protection system (NIJ II + ceramic plate): NIJ IV < NIJ II < unprotected. Reflected pressure from BOPMAN demonstrated a similar evolution (except the NIJ II was here a NIJ IIIa), while swine injury did not follow this trend. This raises the question of the validity of using of the rigid cylinder for evaluation of the efficacy of TPE against blast loading.

Our study shows that under short-duration high-intensity blast exposure (ΔI_I≈230 kPa ms), TPE can influence the level of lung injury in swine as follow: P3 < P0 ≤ P2. Chest response to blast was also investigated in swine wearing P0, P2, and P3 level protections. As for lung injury, our results suggest that chest response (costal acceleration and maximum velocity) was distributed as follows: P3<P0≤P2 (Figure 9A). An acceptable correlation fit was found between Vmax and RL/LL (R² = 0.47) in contrast with the Γmax vs. RL/LL correlation (R² = 0.26). Chest velocity on swine had previously been correlated with the probability of death in 30 min, but for behind armor blunt trauma (Hinsley et al., 2002). Our result on the chest wall acceleration seems contradictory to Cooper and al. findings (1996) who retained the parameter as injury criterion, with a lung injury threshold defined at 10,000 m/s². Nevertheless, Cooper’s study failed to give details regarding the animal weight, the chest instrumentation, the location of the sensor on the thorax or the sensor fixation procedure, which makes the comparison with our study difficult.

We then investigated whether the mannequin BOPMAN can record the same severity tendencies (P3 < P0 < P2) using its sensors, and evaluate if any parameter measured with BOPMAN is correlated to the severity of lung injury in swine. Data from unprotected (P0) and protected (P2, P3) BOPMAN were analyzed, which led to the conclusion that the force and the force impulse (ΔI_FORCE) measured at the mannequin thoracic level may be used to discriminate the ability of TPE to reduce or enhance lung injuries (Figure 9B) Indeed, our correlation analyses clearly demonstrated that the force and ΔI_FORCE measured on BOPMAN followed the trend P3 < P0 < P2 and that they are highly correlated to the lung injury severity RL/LL in swine (R² = 0.74 and 0.77, p < 0.05). Force and force impulse are simply correlated to the observed injury, without any causal relation. This evaluation can only be made for face-on exposure and for a single exposure without any reflection. More data will be needed to validate the use of BOPMAN for multiple exposures or complex blast waves. Existing mannequins have only proven their ability to qualitatively discriminate the level of efficacy of several TPE (better or worse than), but none of them are able to estimate the risk of lung injury, by contrast with BOPMAN. Jetté et al. (2004) studied the response of a torso rig to mild and high-intensity blast without thoracic protection, with low and high impedance materials. Measured maximum thoracic wall acceleration was lower for fragment protective vest and ceramic chest plate (equivalent to our P3 protection) than without protection, while values with the fragment protective vest (equivalent to our P2 protection) were slightly higher than when unprotected. The Hybrid III 50th percentile male mannequin was also used for that purpose (Bass et al., 2005; Carboni et al., 2010). Studies with this mannequin demonstrated that the external thoracic pressure can be reduced by wearing bomb suits, as compared to the unprotected condition (Bass et al., 2005). In addition, these studies found that maximum thoracic wall acceleration was reduced when both thoracic protections (aramid-based fabric with or without hard plate) were used (Carboni et al., 2010). Finally, for short-duration, mild- and high-intensity blast waves, the MABIL mannequin managed to measure higher maximum thoracic wall accelerations with class III protection while lower accelerations were obtained with class IV protection, as compared with data without protection (Ouellet and Williams, 2008). This study showed the evolution of MABIL thoracic wall acceleration as a function of the threat incident overpressure. However, for short-duration waves, risk of lung injury is related to the maximum incident impulse (Boutillier et al., 2019). Therefore, estimating the risk of injury with the MABIL in protected and unprotected settings may not be possible, without further evaluations. Only Cooper et al. (1996b) managed to evaluate the risk for lung injury for high acoustic impedance materials using a rig that simulates the peak of chest acceleration of a swine thorax under blast loading. Nevertheless, no further studies with this rig and no additional data with low impedance materials are available in the literature. Moreover, it used Cooper (1996) chest acceleration criterion with the limitations given in the previous paragraph.

As stated above, animals being hit by the blast wave on their right side displayed ipsilateral, but no contralateral, lung injuries. Bilateral pulmonary contusion was never observed in our experimental conditions. These data, along with those showing that TPE density plays a major role in subsequent injury level, suggest that primary blast injuries to the torso could be regarded as large area direct blunt force traumas, without transmission of a blast wave along tissues. Of course, this hypothesis needs further evaluation.

our study has several limitations. First, the explosive charge used was either a 4 kg spherical charge of Hexomax or C4, which have different TNT equivalence. Nevertheless, change in explosive charge did not impact the overpressure signature curve measured near both models. Then, there are limited data on BOPMAN so far, and it is therefore still uncertain as to whether it could be truthfully used to evaluate the efficacy of TPE for other blast scenarios. Experiments testing different levels of incident impulse should be performed to increase the confidence on BOPMAN’s ability to predict the risk for lung injury in unprotected and protected configurations. Moreover, part of the loading could be missed when equipped with the hard ceramic plate due to the curvature of the plate not perfectly fitting the mannequin shape. If the back face of the protection was in contact with the gelatin part, measurement should include a part due to the wave propagation and a second part due to the impact of the protection on the silicone part. The contribution of each part is unknown and should be studied to evaluate the potential loss of transfer due to the plate curvature. Even if the ceramic hard plate probably does not perfectly fit the swine chest shape (and also the soldier’s), which could lead to different chest motion and lung injury outcomes, the part of the load due to the plate impact will be transferred to the thorax at certain points, contrary to BOPMAN. The fact that a part of the loading could be missed due to the slight spacing between the ceramic plate and BOPMAN silicone block could indeed have an effect on the fitting results (and the relation P3 < P0 < P2) and so the conclusion. If further studies conclude on the importance of the plate impact contribution, the gelatin block of BOPMAN should be modified so that the force recorded at the gelatin block is representative to the mean force applied on the whole plate/thorax area. Furthermore, results obtained on animals may not be readily translated to what is observed in human subjects, because of obvious inter-species morphological differences. A study (Prat et al., 2010) comparing the thoracic wall behavior in large animals and human cadavers submitted to an identical ballistic blunt thoracic trauma showed that the younger pig’s bones were less brittle and more flexible than older PMHS bones, leading to a higher thoracic wall motion under thoracic ballistic impacts. However, comparison to young human subjects is still unknown.

In conclusion, this is the first time, to the authors’ knowledge, that parameters measured on an anthropomorphic mannequin have been correlated to the level of lung injury induced by blast exposure when unprotected or equipped with low and high acoustic impedance body protections. This study provides an initial investigation of using BOPMAN to evaluate the efficacy of TPE against blast loading based on lung injury risk. The results demonstrate that BOPMAN has potential to be used as a test mannequin for evaluating blast lung injury. Additional work to investigate the sensitivity and reproducibility of BOPMAN to other blast exposures (e.g., severity levels, environments) and TPE is necessary to further validate this test device. Validating a test device for evaluating blast lung injury, like BOPMAN, is important to properly evaluate protection systems for soldiers and law enforcement officers.