Biofidelity Assessment of the WIAMan Thorax by a Comparative Study With Hybrid III, THOR, and PMHS in Frontal Sled Testing

Anthropomorphic test devices (ATDs), also known as dummies, are mechanical surrogates that approximate the anthropometry and structural response of living humans under comparable loading conditions. Rather than simply being anthropometric manikins, ATDs are complex engineering devices composed of different materials and instrumented with sensors to record the accelerations, forces, and displacements experienced by the body in a simulated crash environment [1,2]. ATDs have been widely used in injury biomechanics to quantify human response and potential trauma, such as occupant injuries in passenger cars and military vehicles. ATDs can be classified according to size, sex, age, and impact direction [3]. In vehicle crash safety tests, the midsize adult male dummy is the most widely used, e.g., Hybrid III 50th Male (HIII-50M), and represents the height and weight of the 50th percentile U.S. adult male in the 1970s. To enhance the biofidelity and achieve higher durability and measurement ability, the test device for human occupant restraint (THOR) was designed using more recent biomechanics research data [4]. For instance, the 50th percentile adult male THOR dummy (THOR-50M) has a more anatomically biofidelic thorax than the HIII-50M, which includes flexible joints in the thoracic spine, clavicles with improved anatomy, more ribs with a downward slant, and less coupling between adjacent and contralateral ribs [1].

In military environments, the use of improvised explosive devices and antivehicular landmines have increased significantly to target land combat and military transport vehicles, especially during the recent military conflicts in Iraq and Afghanistan [5]. Under an antivehicular mine or improvised explosive devices blast, the energy emitted by an explosive shock wave is absorbed by the vehicle hull, which transmits local accelerative loading directly to the occupant and through the seating system via the vehicle structure [6,7]. Following this type of “underbelly blast” or “underbody blast” (UBB) event, the lower extremity of the occupant is the predominantly injured body region [8]. As such, research in military injury biomechanics over the past decades has focused on this region [5,9]. However, occupants in military vehicles could experience various types of injuries, e.g., blast-related traumatic brain injury [10,11], head and neck injuries due to the head impact with the vehicle roof [12], spine injuries [13], and thoracic traumas due to interacting with the primary wave and/or fragments of debris [14]. It is worth emphasizing that advanced restraint systems in civilian vehicles, such as airbags, pretensioners, and load-limiting seat belts, have not been implemented in military vehicles [15].

Like the automotive industry, artificial surrogates (ATDs, manikins) have also been used in the military to assess the risk of injuries to the soldiers and to design more effective countermeasures to prevent and mitigate injury. The frontal impact Hybrid III dummy has been used for evaluating the survivability and protection level of military ground vehicles subjected to blast events [16]. However, the existing automotive ATDs were designed and validated for automotive occupant postures, impact magnitudes (<250 g floor acceleration), and loading durations of 10–50 ms [17]. UBB scenarios are usually high-energy events characterized by high accelerations (>1000 g) and short durations with peak loads occurring within 5 ms [18]. The Hybrid III and THOR lower limbs have been evaluated by controlled explosion simulating antivehicular land mine loading (4.7 and 8.3 m/s vertical loading), which showed poor biofidelity compared to postmortem human surrogate (PMHS) response [19]. By simulating high rate UBB-like loading conditions, Bailey et al. demonstrated that the Hybrid III is insufficient for predicting pelvis and leg injury and replicating the human response to a range of loading rates due to the limited biofidelity and durability, particularly under higher rate impact conditions [9].

To address the limitations of automotive ATDs for military use, the U.S. Army initiated the Warrior Injury Assessment Manikin (WIAMan) Program in 2010 to design, develop, and validate a specialized ATD for assessing skeletal injuries and improving the effectiveness of protective technologies for occupants in military vehicles, specifically for vertical loading within UBB environment [16]. Anthropometrically, the WIAMan dummy represents a 50th percentile military male with 178 cm height and 84 kg weight [20], which is larger than the HIII-M50 (175 cm and 78 kg) [21]. In the past decade, research has been devoted to developing and evaluating the WIAMan ATD for use in underbody blast testing. To determine injury thresholds for WIAMan ATD, Salzar developed blast rate-related injury criteria and response corridors for the pelvis, femur, and foot through PMHS testing in a simulated UBB environment [22]. Additionally, Salzar et al. developed a human injury probability curve for unstable pelvic ring sacrum fracture for pelves loaded with a laboratory simulated UBB [5]. Pietsch et al. evaluated the biofidelity of WIAMan by comparing the response with PMHS from three simulated UBB sled test conditions [16]. Baker et al. developed and validated an FE model for the WIAMan ATD lower limb to assess risk of injuries during vertical accelerative loading [23]. Slykhouse et al. defined a skeletal-based local coordinate systems for body regions of PMHS in which kinematic measurements are made using the WIAMan ATD [24]. Barnes et al. developed a methodology to assess the performance of the WIAMan injury assessment reference curves through match-paired tests of whole body PMHS and the WIAMan ATD [25]. Most of the current unclassified studies of the WIAMan ATD biomechanics focused on development and biofidelity assessment under UBB or vertical-loading events, particularly at the regions of pelvis, lower extremities, lumbar spine [26,27], and head-neck [11,28]. On the other hand, the occupant(s) in a ground military vehicle in an uncertain battlefield could experience injuries due to impact with the vehicle interior or high levels of acceleration/deceleration caused by stiff structures. For seat belt restrained occupants in civilian vehicles, the thorax is one of the major body regions injured during vehicle collisions [29]. In the theater of war, probably no event occurs in isolation, pointing toward the need of a more omni-directional dummy response. Although the WIAMan ATD was not originally designed to predict the risk of occupant injuries (e.g., thoracic injuries due to frontal impact), it will be quite useful to investigate its capability of predicting injury in a compound event, particularly its occupant response and applicability under impact crash scenarios. The objective of this study was to determine the biomechanical response of the thoracic region of the WIAMan ATD and evaluate its biofidelity by comparing with Hybrid III and THOR ATDs under the Gold Standard frontal sled test conditions [30], as well as matched PMHS tests. The biomechanical performance of the WIAMan thorax can be identified by the comparative analysis of the kinematics, chest deflection and belt loads during frontal sled tests.

A 40 kph frontal sled tests were conducted to assess the body responses of WIAMan, including the thoracic responses. The tests were performed using a 1.4 MN Servo-Sled™ Catapult Sled System (Seattle Safety). The custom-designed horizontal impact buck was used to approximate a real-world frontal crash of the passenger compartment of a typical midsize U.S. sedan and was named the Gold Standard 1 (GS-1) test condition [31]. The test buck consisted of a rigid planar seat and an adjustable matrix of cables, designed to accommodate the front right passenger occupant. More details about the test buck and fixture can be found in previously studies by our lab [30,31]. The ATD was restrained using a custom three-point, U.S. passenger side belt without retractor or pretensioner (Fig. 1), which was replaced prior to each test. The belt webbing material was made by Narricut (International twill pattern 13,195) with 6–8% elongation and a minimum tensile strength of 26.69 kN [31]. The belt load on the seat belt webbing was measured by the load cell (Make: MESSRING GmbH, Model number: 5BCD1622, Manufacturer: Diversified Technical Systems (DTS), Sampling rate: 200,000 Hz, AA filter: 20,000 Hz). To restrict the forward motion of the surrogates' lower extremities and pelvis, a rigid knee bolster and footrest were used during the entire duration of tests. Therefore, during the impact the surrogates were constrained on the seat by the shoulder and lap belts, knee bolster, and footrest.

The three-dimensional (3D) kinematics of the subjects and test buck were quantified using a 1000 Hz Vicon MX™ 3D motion capture system (Vicon, Los Angeles, CA), which tracked the motion of retroreflective markers through a calibrated 3D space with 16 high-speed cameras. Before testing each subject, the system was calibrated to ensure precise tracking of motion. After testing, the trajectories of the spherical markers placed on the subjects were converted to the reference frame of the buck by a mathematical coordinate transformation [32]. The buck coordinate axes were designated based on the reference point at the center of the bolt head for the right lap belt anchor mount, as shown in Fig. 1. The data were reported [30] in accordance with the SAE coordinate system (J211). For the thoracic response in this study, four markers were adhered to the anterior surface of the thorax (Fig. 1), and another four markers were placed at the corresponding sites on the posterior surface of the thorax, which can characterize the deformation and deflection of the four quadrants of the ribcage: upper left, upper right, lower left, and lower right. The shoulder kinematics and displacement could be measured through markers set on the bilateral acromia. However, the marker on the right acromion region was not able to be tracked by Vicon cameras due to being obstructed by the shoulder belt. As such, the bilateral markers were placed on the lateral deltoids to capture the shoulder kinematics (Fig. 1). The upper shoulder belt was placed on the middle of right shoulder, and the bottom edge of the belt on the centerline of anterior chest was located in the middle between the chin and proximal thighs. Following placement of instrumentation, the belt was then adjusted such that the distance to the upper shoulder from the D-ring was approximately 400 mm. At the conclusion of each test, the seat belts were removed and replaced. Figure 2 provided the comparison of initial belt paths and measurement sites of upper anterior ribcage for WIAMan, PMHS [33], and THOR [33]. The origin of Y and Z coordinates in the figure was defined based on the site on the sternum, which was on the horizontal plane at the eighth thoracic vertebral level (T8) of PMHS. Belt loads were measured by belt tension gauges placed on the upper and lower portions of the shoulder belt, as well as one placed on the lap belt near the belt anchor point. Each belt was pretensioned prior to placing the tension gauges on, with target measurements of 40±4.45 N for the lap belt and 13.3±4.45 N for the shoulder belt.

In total, six tests were conducted to assess the WIAMan response, including three tests without wearing a helmet, and three tests wearing a U.S. Army advanced combat helmet (ACH, 1.76 kg), as shown in Fig. 1. The effects of the helmet on head kinematics and neck loads during simulated impact was examined and will be documented in a future publication. To assess the WIAMan thoracic response, the kinematics, chest deflection, and belt loads were measured and evaluated in this study. To evaluate biofidelity, these responses were compared to other surrogates, including PMHS (eight subjects), HIII-50M ATD (three tests, termed HIII hereinafter), and THOR-50M Mod Kit ATD with SD-3 Shoulder (three tests, termed THOR hereinafter), which were reported in previous studies [30,33]. The assessment of thoracic kinematics was specifically focused on the displacements of bilateral shoulders and the first thoracic vertebra (T1) with respect to the coordinate system of the sled buck. No scaling of the displacement data was implemented since the sizes of the ATDs employed were close (midsize male or military male), and the PMHS subjects were approximately midsize male [33]. The chest deflection was characterized by the points (markers) on the four quadrants, i.e., the difference between the marker on the anterior surface of the thorax and the corresponding marker on the back of the thorax in each quadrant. The thorax was compressed by the seat belt during the impact process [34], thus it is significant to report the response of the belt loads.

The similarities of thoracic response in time histories between ATDs and PMHS were evaluated by a widely used objective rating tool known as correlation and analysis or CORA [35–37]. The CORA rating can evaluate the similarity between curves using the cross-correlation method, which evaluates error according to phase shift, magnitude, and curve shape to produce a relative score ranging from 0 (no correlation) to 1 (perfect match) [38]. In this study, the CORA scores of 0–0.26, 0.26–0.44, 0.44–0.65, 0.65–0.86, and 0.86–1.0 were evaluated as poor, marginal, fair, good, and excellent, respectively [39].

After the whole body WIAMan was tested in the GS-1 test configuration, the responses of the WIAMan ATD thorax were compared to the responses of the HIII, THOR, and PHMS in matched frontal sled tests.

The positions of the bilateral shoulders and T1 were quantified to determine the occupant kinematics relative to the sled buck coordinate system. Figure 3 compared the overhead view of the kinematics of four surrogates in the X–Y plane of the sled buck, e.g., the average positions of test results for each surrogate at 20 ms intervals during the impact. It was noted that the positions for WIAMan testing results were separated by tests without wearing a helmet (unhelmeted) and tests with a helmet (helmeted). The overlay of positions of all surrogates at 120 ms were also provided (Fig. 3(f)). Figure 4 compared the displacement-time histories of the right shoulder and T1 for all surrogates at 120 ms. The peak right shoulder displacements of WIAMan were 75.0 mm (unhelmeted) and 77.9 mm (helmeted), which were lower than the average PHMS response (110.4 mm). In Table 1, the CORA scores for the right shoulder displacement of the WIAMan tests were lower than both HIII and THOR (WIAMan average score for all tests: 0.435; HIII: 0.893; THOR: 0.693). Although the T1 displacements of WIAMan were found to be generally smaller than the PMHS, the curves of the T1 displacement-time histories of WIAMan were almost identical with the Hybrid-III, and only slightly smaller than the THOR. The CORA scores of the T1 displacement histories for each ATD compared to PMHS were similar (WIAMan average: 0.839; HIII: 0.851; THOR: 0.877).

The chest deflection-time histories of the WIAMan ATD along the anterior–posterior direction were quantified through the four markers on the anterior surface of the thorax, shown in Fig. 5. The average X-axis chest deflection of PMHS was shown by thick gray lines, while the individual response of each cadaver subject was displayed as thin light gray lines to show subject-to-subject variability. In Fig. 5, the negative displacement indicated that the measured site moved toward the posterior direction, and the chest was compressed at that site. Similar to the sternum response approximation used for THOR ATD [33], the deflection of the WIAMan sternum was averaged from the markers. The deflection-time histories of the three tests of each WIAMan ATD configuration (i.e., unhelmeted and helmeted configurations) exhibited good repeatability, thus only the averaged response of each configuration was provided in each plot. The deflection-time histories of the unhelmeted and helmeted tests were also very similar. Accordingly, the differences of CORA scores for deflection at each quadrant between two configurations were all less than 0.03 in Table 1. In Fig. 5, the peak deflection appeared between 120 ms and 140 ms in the WIAMan thorax, and also in the upper thorax of the HIII. Regarding upper chest deflection, those produced by WIAMan were generally larger than both HIII and THOR ATDs, as well as the average PMHS response. The deflection-time histories of the WIAMan upper left, upper right, and sternum regions tended to be more similar to the PMHS histories in terms of magnitude. Accordingly, the WIAMan produced higher average scores (0.907 for upper left and 0.811 for upper right) than the HIII (0.799 and 0.742) and THOR (0.707 and 0.322). For sternum deflection, the WIAMan also had a much higher average CORA (0.971) than the HIII (0.728) and THOR (0.713). However, the lower left quadrants of all ATDs were compressed more severely than the PMHS. Consequently, the CORA scores were all low for deflections of lower left quadrant yielded by all ATDs (<0.56 for HIII and THOR, and 0.284 for average WIAMan). For the lower right quadrant, the PMHS experienced expansion (positive deflection in Fig. 5(d)), i.e., the measurement site moved away from the back of the subjects, and the THOR also had a slight expansion of lower right ribcage before 85 ms. It should be noted however both WIAMan and HIII did not display expansion at the lower right region of the chest.

Figure 6 provided the comparison of the average belt loads, including the lap belt load, upper shoulder belt load, and lower shoulder belt load measured by a lower anterolateral belt load cell. It was found that the maximum WIAMan lap belt load (0.810 kN) was closer to the PMHS (0.826 kN), compared to other ATDs (Fig. 6(a)). The results of the upper shoulder belt load response in Fig. 6(b) indicated that both the WIAMan and HIII had higher peak belt load than that of the average PMHS response (WIAMan: 7.29 kN; HIII: 7.11 kN; PMHS: 6.32 kN), and the THOR was an excellent match with the average PMHS curve, compared to those of the HIII and WIAMan. For maximum lower shoulder belt load, all ATDs exhibited lower magnitude than the PMHS (5.35 kN), and the WIAMan has the lowest peak magnitude (4.15 kN) in Fig. 6(c). Comparing the shape of the belt load-time histories between the ATDs and PMHSs, it was evident that the WIAMan shoulder belt load-time history curves reached their first peaks before 60 ms, which were earlier than the PMHS and other ATDs. Overall, both the upper and lower shoulder belt load time histories of WIAMan were less similar to the PMHS in shape. Notably, when comparing the CORA scores of belts loads, the WIAMan was less similar to the PMHS compared to the HIII or THOR. However, all ATDs had CORA scores in the “good” or “excellent” ranges for all measured belt loads.

To our knowledge, this is the first study to characterize the thoracic response of the WIAMan ATD to a three-point seatbelt, which can offer valuable insight into expanding its use in occupant crash protection in a military vehicle under frontal impact scenarios. The biomechanical responses of the WIAMan thorax and shoulder were evaluated using the Gold Standard frontal sled test condition. Comparative biofidelity of the WIAMan thorax and other surrogates, including HIII, THOR, and PMHS, were assessed under similar restraint and testing conditions.

The comparison of belt paths and measurement sites of upper anterior ribcage in Fig. 2 showed some minor differences among surrogates, although the testing of WIAMan ATD aimed to match the previous GS-1 sled test condition as close as possible. These differences in initial configurations could be caused by the differences of design and geometry of the ATDs, as well as discrepancies of stature and mass between ATDs and PMHS in some extent. The comparison of the kinematics of the upper thorax in Figs. 3 and 4 revealed substantial differences in shoulder motion between the WIAMan and other surrogates. This overhead view illustrated that WIAMan has a larger shoulder width and experienced a greater clockwise rotation of the shoulders about the vertical direction (Z-axis of the sled buck) than other ATDs, as well as PMHS. When examining the overlay of all subjects at 120 ms, the displacement of WIAMan T1 along the X-axis (forward excursion) was very close to the HIII and less than THOR and PMHS. However, the forward excursion of the WIAMan left shoulder was much greater than the HIII, but only slightly larger than the THOR and PMHS. High speed video also confirmed that the WIAMan exhibited considerable rotation outside of the sagittal plane, compared to the other ATDs. The left shoulder experienced substantially more forward and upward excursions than the HIII, and the right shoulder had less displacement than the THOR, but was closer to the PHMS, as shown in Fig. 4(a). The HIII showed little thoracic rotation outside of the sagittal plane and less asymmetric excursion between the bilateral shoulders among all surrogates. Compared to HIII, the left shoulder of WIAMan and THOR experienced more rotation outside of the sagittal plane, which showed more similarity to the PMHS in Fig. 3. Similar observation between HIII and THOR was reported in previous sled tests of HIII, THOR, and PMHS at higher speeds [40]. In Fig. 4(b), all three ATDs exhibited similar levels of T1 forward excursion, which were about 20–30% lower than the PMHS response. However, when the difference of the peak forward excursions was compared between T1 (Fig. 4(b)) and the right shoulder (Fig. 4(a)), the WIAMan response was found to be close to the PMHS (WIAMan: 150 mm; PMHS: 157 mm), while in Fig. 3(d) the THOR experienced nearly equal excursions of right shoulder and T1 [33]. The differences in thorax and shoulder kinematics between surrogates illustrated that each surrogate interacted differently with the belt system [40] throughout the duration of the sled tests. Notably, these differences could be an indicator of the discrepancies between the design of the surrogates, since the belt load was mostly sustained by the shoulder complex and the anterior ribcage [41]. The thoraces and shoulders of WIAMan and THOR were clearly designed with a lower flexural rigidity under anterior–posterior bending than the HIII under sled compression.

Regarding the upper chest deflection, the WIAMan generally experienced larger deformation than other surrogates during belt-induced compression, while both the HIII and THOR exhibited shallower deflection compared to the average PMHS response. This implied that the WIAMan thorax is more compliant in this mode of loading compared to the HIII and PMHS. For the WIAMan chest, the maximum deflection occurred in the compression side of the lower chest, similar to HIII and THOR. This was consistent with previous studies of HIII and THOR [42,43]. However, the sled tests using PMHS showed that the maximum chest deflection of PMHS occurred in the upper thorax at 40 kph GS-1 sled impact [33], lower speed (29 kph), belted, noninjurious sled tests [44], and 56 kph sled tests [1]. These discrepancies of chest deflection may be a result of differences in ribcage design of ATDs. The WIAMan thorax was designed using steel ribs to approximate human rib geometry and an abdomen insert made of foam-filled urethane, covered by an out flesh jacket made flexible urethane [16]. However, its midanterior chest was designed without comparable structure to produce a humanlike behavior such as the sternum or costal cartilages with appropriate stiffness, while the HIII included a rigid sternum plate and the THOR improved the design of the plate to allow more anterior chest flexibility [45]. By contrast, the WIAMan anterior thorax behaved more compliant when the middle region of the anterior chest surface was compressed by the belt loading under frontal sled tests. This difference could be addressed in future design improvements of WIAMan thorax region. It is very interesting to note that the WIAMan and HIII did not exhibit expansion of the lower chest, but the phenomenon of “bulging out” was found on the side opposite the belt in PMHS and THOR lower chest, e.g., the right side of the lower chest of a belted occupant in frontal impact tests in a study using a right-front passenger configuration. This discrepancy of lower chest deformation pattern between HIII and THOR/PMHS was also observed in previous comparative experiments on PMHS, THOR, and HIII, such as Refs. [1,43], and [44].

As compared in the Results section, there were differences in the shape and peak magnitudes of the belt load-time histories between the ATDs and PMHS. Compared to the PMHS, HIII, and THOR, the WIAMan had a higher peak magnitude of upper shoulder belt load and lower peak magnitude of lower shoulder belt load. This indicated that a larger portion of shoulder belt load was shared by the upper WIAMan thorax compared to other surrogates. The WIAMan experienced the greatest thoracic rotation outside of the sagittal plane and associated excessive upper thorax defection under belt loading, possibly due to lower stiffness of midthorax; therefore, the WIAMan yielded lower CORA scores of belt loads than other ATDs in Table 1. The lap belt tension could be affected by both thorax movement and the interaction of the subject's pelvis and lower extremity with the rigid knee bolster and footrest. In this study, the experiment videos showed that the lap belt provided adequate pelvis restraint to prevent the pelvis from moving forward and downward significantly (i.e., the phenomenon of submarining). In general, it is worth emphasizing that the thoracic responses of an ATD were not only determined by its thorax design, but also influenced by other associated body parts. For instance, in this study the WIAMan head experienced more forward and downward excursions and took longer time to rebound back to its original position, likely influenced by the very different neck designs [46]. A further investigation is needed to determine which factors or major factor produced the double peaks of the shoulder belt loads during the sled testing. Finally, the WIAMan showed good repeatability between the unhelmeted tests and helmeted tests according to the comparison of all thoracic responses, including the shape, phase, and peak magnitude. The mass of the WIAMan head region and associated moment of inertia were increased in the tests with the helmeted WIAMan ATD. Despite the similarity between tests with and without wearing a helmet, slight lagging of phases of the belt load histories was found in the average results of helmeted tests.

A detailed investigation of the impact biomechanics of the WIAMan thorax is beyond the scope of this study, but a comparative study of selected responses was conducted. There are some limitations to this study. First, the thoracic responses were evaluated only using a specific frontal sled test with designed restraint conditions, which limited a thorough comparison of thoracic stiffness under various crash environments [47] between WIAMan ATD and other surrogates. For future work, extensive impact testing, similar to the tabletop tests [34] using belt-like loading and THOR certification tests [48], can be performed to evaluate the crash performance of the WIAMan ATD comprehensively. Second, there exists some variability in measurement sites between different surrogates (Fig. 2), which may influence the accuracy of measurement to a certain extent. For PMHS testing, there were some differences of the marker locations on the thoraces between subjects [30] due to inherent anatomical subject-to-subject variation. Meanwhile, the kinematic data of WIAMan ATD was measured by a 1000 Hz Vicon MX™ motion capture system, which is newer than the Vicon system used for PMHS testing [30]. In addition, the small number of tests used in this study limited the statistical analysis for the results of each surrogate, although ATDs produced highly repeatable responses. This study was also limited by ignoring some other potential factors, such as the differences in pelvis and abdomen design of different ATD's, seatbelt-ATD torso friction, and the CORA rating metric has its own limitations [49].

In conclusion, this study evaluated the WIAMan thoracic response in frontal sled tests. The biofidelity of the WIAMan thorax was analyzed in a comparative study with PMHS, along with the HIII and THOR ATDs. The comparison of the kinematics of the upper thorax showed that the WIAMan left shoulder experienced more rotation outside of the sagittal plane and exhibited a similar level of forward excursions as the PMHS and THOR ATD, which were significantly larger than the HIII. The comparison of the chest deflection illustrated that the WIAMan experienced larger deflections during belt-induced compression than the other two ATDs. Along the belt path, the WIAMan had a higher peak magnitude of upper shoulder belt load and a slightly lower level of lower shoulder belt load, compared to the other ATDs. The overall quantitative performance of the WIAMan thorax was more compliant than HIII, but reasonable compared with the PMHS responses. Considering the WIAMan ATD was not originally designed for the use of predicting occupant injury under frontal impact conditions, its overall thoracic responses were deemed relatively reasonable and biofidelic according to the correlations with the cadaver responses. The current assessment of the WIAMan thoracic biomechanics was focused on its global response and differences with other surrogates under the designed frontal sled test condition. For future studies, it is expected to comprehensively evaluate the comparative biofidelity of WIAMan and other surrogates under different impact and restraint conditions, especially belted occupants under simulated battlefield environments.

This work was sponsored by the U.S. Army Research Lab in support of the Warrior Injury Assessment Manikin Program. The authors gratefully acknowledge the contributions of the WIAMan Engineering Office. The content presented herein does not necessarily reflect the position or policies of the sponsor.

The authors declared that they have no known competing interests in this paper.

• U.S. Army Research Laboratory (Funder ID: 10.13039/100006754).

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.