Anthropomorphic Test Device Research Articles

Countermeasures design and analysis for occupant survivability of an armored vehicle subjected to blast load

The demand for blastworthy structures increases due to a large number of casualties in the armored vehicle undergoing improvised explosive device (IED) and landmine attacks. In this research, the numerical studies on the countermeasure analysis to reduce injury biomechanics risk were conducted. The available experiment data of occupant survivability test on a medium size tank was used to validate the numerical model. The subsystem evaluation in this study included the finite element modeling of military personnel, seat system, surrounding interior system, seatbelt, and restraint system with four running conditions. The military personnel inside the armored vehicle was modeled by using Hybrid III 50th percentile anthropomorphic test device (ATD) and its biomechanical response was monitored on the head, neck, thorax, spine, femur, and tibia. The load case for this study referred to NATO STANAG 4569 level 3b with 8 kg TNT explosive load underbelly. The injury assessment reference values (IARV) for the regulation used in this study were based on AEP-55 volume 2. Based on this study, the critical injuries identified on the head injury, neck compression, and tibia axial load. The solid frame as part of seat structure appeared to contribute to an excessive kinematic on the lower extremities. The vehicle acceleration resulted from the load blast was directly transmitted to the lower extremities, resulting in unintended kinematic and interaction on the passenger body. The proposed solutions were to introduce a flexible mounting for the seat system and as well increasing the height of the footrest to avoid direct transmission of vehicle acceleration. The modified countermeasure design reduced significantly head, neck, and tibia injury criteria more than 90 % from the baseline design (existing design). The new anti-mine seat design successfully passed all the standard regulation thresholds of injury criteria.

Biofidelity of THOR 5th Percentile Female ATD in Ankle Eversion and Inversion

<div class="section abstract"><div class="htmlview paragraph">Females have higher frequency and risk of foot and ankle injuries in motor vehicle collisions than similar-sized males. Therefore, lower extremity biofidelity and accurate injury prediction of female ATDs is critical. This paper aims to compare the THOR 5th percentile female (THOR-05F) anthropomorphic test device (ATD) response with male and female PMHS data of various sizes under ankle inversion and eversion. The THOR-05F lower extremity was subjected to dynamic inversion and eversion ankle loading with a constant 2000N axial force applied through the tibia. Twelve THOR-05F tests (3 inversion and 3 eversion on both, left and right legs) were performed with boundary conditions consistent with previous post-mortem human subject (PMHS) lower extremity tests. The biofidelity of THOR-05F ankle stiffness was evaluated via comparison of measured and equal-stress equal-velocity scaled data (using mass-based scale factors) from previous PMHS datasets with mid-size males, small females and larger females. THOR-05F ankle moment-angle response falls within the range of previous mid-sized male and larger female PMHS test data for eversion, when scaled to a small female. However, when compared to PMHS response measured on small female subjects, the THOR-05F response was less stiff in both inversion and eversion. The THOR-05F moments were 65% and 90% less stiff in eversion and inversion respectively, when compared to the average of the measured small female PMHS dataset at 25<sup>°</sup> ankle rotation. Because ATD stiffness differs from measured PMHS ankle stiffness, care should be taken when applying PMHS-based injury risk functions (IRF) to the THOR-05F ankle.</div></div>

The Roles of Vehicle Seat Cushion Stiffness and Length in Child Restraint System (CRS) Performance

<div class="section abstract"><div class="htmlview paragraph">The objective is to determine whether responses and injury risks for pediatric occupants in child restraint systems (CRS) are affected by vehicle seat cushion stiffness and fore/aft cushion length. Eighteen sled tests were conducted using the Federal Motor Vehicles Safety Standard (FMVSS) 213 frontal pulse (48 km/h). Seats from a recent model year vehicle were customized by the manufacturer with three different levels of cushion stiffness: compliant, mid-range, and stiff. Each stiffness level was quantified using ASTM D 3574-08 and all were within the realistic range of modern production seats. The usable length of each seat cushion was manipulated using foam spacers provided by the manufacturer. Two different seat lengths were examined: short (34.0 cm) and long (43.5 cm). Three different types of CRS were tested with size-appropriate anthropomorphic test devices (ATDs): rear-facing (RF) CRS with 12-month-old CRABI, forward-facing (FF) CRS with Hybrid III 3-year-old, and high-back booster with Hybrid III 6-year-old. Each CRS, vehicle seat (including cushion and frame), seat belt webbing and buckle were replaced after every test. ATD kinematic and kinetic data were compared across seat cushion lengths and cushion stiffness levels to determine which seat configurations were the most beneficial for each type of CRS. For RF CRS, short vehicle seats allowed more y-axis rotation (SAE J211) but reduced several injury metrics including HIC36. For FF CRS, long and short seats resulted in similar injury metrics across matched conditions. For boosters, short seats increased chest resultant acceleration but did not have a noticeable effect on other injury metrics. The range of cushion stiffness examined in this study did not have a consistent or relevant effect on any of the CRS or occupant responses.</div></div>

MERA (motorcycle emergency ride assistance): plug-in for accident detection and notification

The condition to trigger the Automatic crash detection for motorcycles is more complex, since the collisions in real life are usually different than those in the lab. so in order to boost the accuracy and affectivity, the researchers must reduce inconsequential deployments, the occupant weight, and riding condition, as well as re-calculate the speed to crash conditions. In this paper was present a continuously evaluates and innovate for crash detection with support automated notification system. The accident detection, communicates the accelerometer, vibration, speed and lean values to the determine accident status and level. In this concept, three different variable measurements are attached to head of the motorist and motorcycle body, Crash dummy tests are done for trial particular systems in helmet by throwing the helmet with different altitudes to simulate the effect of crash to the motorist, and for capture the location of accident, real data is collected by driving the motorcycle. The implemented prototype system showed promising results for automatic crash detection, and the system also showed its success for deliver real-time notifications when an accident was happening.

Kinematic and Biomechanical Response of Post-Mortem Human Subjects Under Various Pre-Impact Postures to High-Rate Vertical Loading Conditions.

Limited data exist on the injury tolerance and biomechanical response of humans to high-rate, under-body blast (UBB) loading conditions that are commonly seen in current military operations, and there are no data examining the influence of occupant posture on response. Additionally, no anthropomorphic test device (ATD) currently exists that can properly assess the response of humans to high-rate UBB loading. Therefore, the purpose of this research was to examine the response of post-mortem human surrogates (PMHS) in various seated postures to high-rate, vertical loading representative of those conditions seen in theater. In total, six PMHS tests were conducted using loading pulses applied directly to the pelvis and feet of the PMHS: three in an acute posture (foot, knee, and pelvis angles of 75°, 75°, and 36°, respectively), and three in an obtuse posture (15° reclined torso, and foot, knee, and pelvis angles of 105°, 105°, and 49.5°, respectively). Tests were conducted with a seat velocity pulse that peaked at ~4 m/s with a 30-40 ms time to peak velocity (TTP) and a floor velocity that peaked at 6.9-8.0 m/s (2-2.75 ms TTP). Posture condition had no influence on skeletal injuries sustained, but did result in altered leg kinematics, with leg entrapment under the seat occurring in the acute posture, and significant forward leg rotations occurring in the obtuse posture. These data will be used to validate a prototype ATD meant for use in high-rate UBB loading scenarios.

Biofidelic Evaluation of the Large Omni-Directional Child Anthropomorphic Test Device in Low Speed Loading Conditions.

Motor vehicle crashes remain the leading cause of death for children. Traditionally, restraint design has focused on the crash phase of the impact with an optimally seated occupant. In order to optimize restrain design for real-world scenarios, research has recently expanded its focus to non-traditional loading conditions including pre-crash positioning and lower speed impacts. The goal of this study was to evaluate the biofidelity of the large omni-directional child (LODC) ATD in non-traditional loading conditions by comparing its response to pediatric volunteer data in low-speed sled tests. Low-speed (2-4 g, 1.9-3.0 m/s) frontal (0°), far-side oblique (60°), and far-side lateral (90º) sled tests, as well as lateral swerving (0.72 g, 0.5 Hz) tests, were conducted using the LODC. The LODC was restrained using a 3-point-belt with an electromechanical motorized seat belt retractor, or pre-pretensioner. Motion capture markers were placed on the head, torso, and belt. The LODC was compared to previously collected pediatric volunteer data as well as the HIII 10 and Q10. Significant difference between the pediatric volunteers and ATDs were identified by comparing the mean ATD response to the pediatric volunteer 95% CI. The LODC exhibited lower forward head excursion (262 mm) compared to pediatric volunteers (263 - 333 mm) in low-speed frontal sled tests (p<0.05), but was closer to the pediatric volunteers than the HIII 10 (179 mm) and Q10 (171 mm). In lateral swerving, the LODC (429 mm) exhibited greater lateral head excursion (p<0.05) compared to pediatric volunteers (115 - 171 mm). The LODC exhibited a greater reduction in kinematics compared to the pediatric volunteers in all loading conditions with a pre-pretensioner. These data provide valuable insight into the biofidelity of the LODC in non-traditional loading conditions, such as evaluating pre-crash maneuvers on occupant response.

A Base Study to Investigate MASH Conservativeness of Occupant Risk Evaluation

Abstract The manual for assessing safety hardware (MASH) defines crash tests to assess the impact performance of highway safety features in frontal and oblique impact events. Within MASH, the risk of injury to the occupant is assessed based on a “flail-space” model that estimates the average deceleration that an unrestrained occupant would experience when contacting the vehicle interior in a MASH crash test and uses the parameter as a surrogate for injury risk. MASH occupant risk criteria, however, are considered conservative in their nature, due to the fact that they are based on unrestrained occupant accelerations. Therefore, there is potential for increasing the maximum limits dictated in MASH for occupant risk evaluation. A frontal full-scale vehicle impact was performed with inclusion of an instrumented anthropomorphic test device (ATD). The scope of this study was to investigate the performance of the flail space model (FSM) in a full-scale crash test compared to the instrumented ATD recorded forces which can more accurately predict the occupant response during a collision event. Additionally, a finite element (FE) model was developed and calibrated against the full-scale crash test. The calibrated model can be used to perform parametric simulations with different testing conditions. Results obtained through this research will be considered for better correlation between vehicle accelerations and occupant injury. This becomes extremely important for designing and evaluating barrier systems that must fit within geometrical site constraints, which do not provide adequate length to redirect test vehicles according to MASH conservative evaluation criteria.

Effect of anthropometry scaling on the response of the piper child scalable human body model subject to pelvic impact

The Open Source PIPER child scalable human body model was publicly released in April 2017 (www.piper-project.org) along with frontal and side impact validation conditions. The objective of this paper is to investigate the effect of anthropometry scaling on the response of the model in side pelvic impact.Three setups from two published studies were used: (1) a lateral drop test (2) a greater trochanter impact with a rigid pendulum (3) a pelvis side impact with a flat surface. The first study used scaling assumption developed for crash test dummy design (setups 1 and 2) and the second performed tests on post mortem human surrogates. The baseline 6 years old child model was scaled using a model morphing methodology to match the stature and weight of the surrogates used in the two published studies.Overall, the main trends observed in the three setups can be approached using the baseline model. Although the model morphing did not account for specific skeletal dimensions, it reduced some of the discrepancies between model response and reference for the drop test and flat plate impact. However, it had little effect on the pendulum test. In that case, the model response was in the corridor at low speed but above at higher speeds. Possible reasons for this difference should be further investigated.

Response of hybrid III and Mil-Lx ATD legs to explosively driven vehicle floor loading

Results are reported of small-scale explosive experiments with Hybrid III and Mil-Lx anthropomorphic test device (ATD) legs. The legs were subject to loadings from deforming metallic plates, driven by explosive loadings to replicate the movement of the floor of a protected vehicle subject to a land-mine strike. The forces measured by the legs are reported and compared between the two different leg types. The benefits of protective measures, including false-floors and commercially available footpads, are compared for their ability to reduce the forces measured in each leg type. It is concluded that the two leg types respond differently to different protective measures and hence cannot be used interchangeably.

Anthropomorphic Blast Test Device for Primary Blast Injury Risk Assessment.

Blast overpressure health hazard assessment is required prior to fielding of weapon systems that produce blast overpressures that pose risk of auditory and nonauditory blast lung injuries. The anthropomorphic blast test device (ABTD) offers a single device solution for collection of both auditory and nonauditory data from a single blast at anthropometrically correct locations for injury risk assessment. It also allows for better replication of personnel positioning during weapons firings. The ABTD is an update of the blast test device (BTD), the current Army standard for collection of thoracic blast loading data. Validation testing of the ABTD is required to ensure that lung injury model validated using BTD collected test data and sheep subjects is still applicable when the ABTD is used. Open field validation blast tests were conducted with BTD and ABTD placed at matching locations. Tests at seven blast strength levels were completed spanning the range of overpressures for occupational testing. The two devices produced very similar values for lung injury dose over all blast levels and orientations. The ABTD was validated successfully for open field tests. For occupational blast injury assessments, ABTD can be used in place of the BTD and provide enhanced capabilities.

Ukemi Technique Prevents the Elevation of Head Acceleration of a Person Thrown by the Judo Technique 'Osoto-gari'.

Biomechanical analysis was performed to evaluate the effectiveness of mastering ukemi in preventing severe head injury in judo. One judo expert (tori) threw another judo expert (uke) with a skilled break-fall (ukemi) four times. We obtained kinematic data of uke with a digital video camera. Both translational and rotational accelerations were measured with a six-degree-of-freedom sensor affixed to uke’s forehead. When Osoto-gari was performed, uke fell backward and his arm made contact with the tatami; the translational and rotational accelerations rose to peak values. The peak resultant translational and rotational accelerations were respectively 10.3 ± 1.6 G and 679.4 ± 173.6 rad/s2 (mean ± standard deviation). Furthermore, when comparing the values obtained for the judo experts with those obtained using an anthropomorphic test device (ATD: the POLAR dummy) that did not perform ukemi, both the peak resultant translational (P = 0.021) and rotational (P = 0.021) accelerations of uke were significantly lower than those for the ATD, whose head struck the tatami. Additionally, there was no significant difference among the three axis directions for either translational (ax: 7.4 ± 0.2, ay: 8.5 ± 2.1, az: 7.2 ± 0.8 G) or rotational (αx: 576.7 ± 132.7, αy: 401.0 ± 101.6, αz: 487.8 ± 66.6 rad/s2) acceleration. We confirmed that performing correct ukemi prevented the elevation of head acceleration by avoiding head contact with the tatami when a judoka is thrown by Osoto-gari. Judoka should therefore undertake intensive practice after they have acquired ukemi skills.

A recent investigation on design aspects of vehicle crash analysis in India

In recent years, the overriding purpose of Automobile Industries has transformed into altering the design features of an automobile regularly, for instance, the chassis, braking system, Air-bag Control Unit (ACU), and others. The predominant motive is to mitigate the probability of susceptible injuries during varying impacts of collision ascribed to increasing average Global Road Crash Rate of 3700 deaths per day. To initially alleviate the fatality rate, experimental and virtual analysis study the car dynamics, velocity, repercussions of a vehicle collision, crash sensors, and personal injuries. The empirical crash analysis utilizes the Anthropomorphic Test Device (ATD), a.k.a “dummies” to ensure the profound protection of design aspects in crashworthiness and crash compatibility for auto-mobiles. The dynamic analysis employs progressive soft-wares viz. Hyper-crash, LS-DYNA, ANSYS Workbench, and Radioss for simulation to contemplate real-world evaluation. This Literature review study serves as the guidance for future generations to scrutinize the significance of crash analysis on diverse testing circumstances and studies.

Two-dimensional Degenerated Model of Next-generation Crash Test Dummy THOR 5F

Two-dimensional Degenerated Model of Next-generation Crash Test Dummy THOR 5F

Numerical assessment of occupant responses during the bus rollover test: A finite element parametric study

Rollover crashes of buses are usually associated with multiple impacts that can result in complex interactions between passengers and a bus superstructure. Although there have been a few field data studies that provide some insights into occupant injuries (e.g. severity and distribution of injuries) during the real-world bus rollover crash, because they had used post-crash data, the occupant kinematics and injury mechanisms were not completely detailed in their results. Based on a literature review, available numerical and experimental studies on a bus rollover safety have mainly focused on structural integrity rather than considering occupant responses in their assessment. In addition, their results about occupant responses in bus rollover crashes show some discrepancies in terms of the estimated injury distribution, severity, and causes. Therefore, the main objective of this study was to provide a more detailed understanding of the occupant kinematics and associated injury risk during the ECE R66 tilt table bus rollover test using validated finite element (FE) models. The ECE R66 tilt table rollover was simulated using a full finite element model of the bus. A 50th percentile male Hybrid III Anthropomorphic test device (ATD) and EuroSID-2re FE models were selected to simulate the occupant’s motion. Each ATD was seated adjacent to the impacted side wall and restrained with a 2-point seatbelt. Simulation parameters included two impact surface friction values and different side window conditions. The results indicated that both ATD estimated the highest injury risk when the partial ejection occurred. They predicted a similar injury risk for the head and thorax. The ES-2re estimated a very low risk of neck injury in all simulations, whereas the Hybrid III estimated the high risk of a neck injury. Finally, recommendations to potentially reduce the injuries were provided and possible future works were suggested.

Biomechanical Analysis of the Head Movements of a Person Thrown by the Judo Technique 'Seoi-nage'.

The present study examined the kinematics and biomechanical parameters of the head of a person thrown forward by the judo technique ‘Seoi-nage’. A judo expert threw an anthropomorphic test device (the POLAR dummy) five times. Kinematics data were obtained with a high-speed digital video camera. Linear and angular accelerations of the head were measured by accelerometers mounted at the center of gravity of the dummy’s head. When Seoi-nage was performed, the dummy fell forward accompanied by contacting the anterior parietal regions of the head to the tatami, and the linear and angular accelerations of most axes reached peak values when the head contacted the tatami. Peak resultant linear and angular accelerations were 20.3 ± 9.8 G and 1890.1 ± 1151.9 rad/s2, respectively (means ± standard deviation). Peak values in linear and angular acceleration did not significantly differ between the three directional axes. Absolute angular accelerations in all axes observed in Seoi-nage were high and the resultant value was approximately equal to the already reported in Ouchi-gari, one of the predominant techniques causing judo-related acute subdural hematoma. However, the remarkable increase of linear acceleration in the longitudinal direction and/or angular acceleration in the sagittal plane, as previously reported in techniques being thrown backward (i.e., Ouchi-gari and Osoto-gari), was not detected. The likely mechanism of acute subdural hematoma caused by Seoi-nage is that a large angular acceleration causes large strains and deformations of the brain surface and subsequent rupture of cortical vessels.

A numerical investigation of factors affecting lumbar spine injuries in frontal crashes

Recent field data analyses have shown that lumbar spine fractures occurred more frequently in late model vehicles than the early ones in frontal crashes. Therefore, the objective of this study was to investigate risk factors associated with lumbar spine fractures in frontal crashes.Parametric simulations were conducted using a set of validated vehicle driver compartment model, restraint system model, and a HIII mid-size male crash test dummy model. Risk factors considered in the study included occupant seating posture, crash pulse, vehicle pitch angle, seat design, anchor pre-tensioner, dynamic locking tongue, and shoulder belt load limiter. ANOVA and ANCOVA were used to test the statistical significance (p < 0.05).Simulation results showed that all the factors that reduced the risk of submarining increased the lumbar spine forces, indicating a direct conflict between submarining and lumbar spine fractures. Among all the factors selected, seat structure is the most significant factor in determining the lumbar spine force (p < 0.001). Crash pulse severity, time at which the peak crash deceleration reached, and pitch angle are also crucial for lumbar spine force. Specifically, increase in vehicle pitch angle increased lumbar spine force, but reduced injury measures to other body regions; while a crash pulse with early peak produced greater lumbar spine force than that with a late peak. On average, more reclined posture increased the lumbar spine force compared to upright posture, and decreases in the coefficient of friction between the pelvis and the seat cushion reduced the lumbar spine force. However, they are not statistically significant.This study provided better understanding of effects from design countermeasures to reduce occupant lumbar spine injuries in new generation of vehicle models.

The biomechanics of concussive helmet-to-ground impacts in the National Football league

This paper presents a detailed characterization of helmet-to-ground impacts in the National Football League. Video analysis was performed for 16 head-to-ground impacts that caused concussions. Average resultant closing velocity was 8.3 m/s at an angle nearly 45° to the surface. Preimpact rotational velocity of the helmet ranged from negligible to as high as 54.1 rad/s. Helmet impacts were concentrated on the posterior and lateral aspects.To study the interaction in greater detail, a helmeted anthropomorphic test device (ATD) was launched over a football field and fell to the ground in various impact conditions. Substantial decoupling between the helmet and the head was observed, such that the head rebounded within the helmet and underwent changes in linear and rotational motion greater than those of the helmet. Vertical helmet rebound was also observed; the helmet underwent a change in vertical velocity on average 24% greater than the vertical component of its closing velocity. Frictional interaction between the helmet and the ground surface caused the helmet to undergo an average horizontal change in velocity of 57% of the horizontal component of its closing velocity. Finally, the duration of a helmet-to-ground impact was generally in the range of 15 – 30 ms, suggesting that the impact surface provides little ride-down. Lengthening this duration could be beneficial both by reducing the peak linear and rotational acceleration and by shifting the impact toward a time regime where brain strain is related to rotational acceleration rather than rotational velocity.

Simulated single-bicycle crashes in the VTI crash safety laboratory

Objective: The objective of this study was to examine the influence of bicycle design and speed on the head impact when suffering from a single-bicycle crash, and the possibility to study this using crash tests.Methods: Simulations of single-bicycle crashes were performed in the VTI crash safety laboratory. Two bicycle crash scenarios were simulated: “a sudden stop” and “sideways dislocation of the front wheel”; using four different bicycle types: a “lady’s bicycle”, a commuter bicycle, a recumbent bicycle and a pedelec; at two speeds: 15 and 25 km/h. In addition, sideway falls were performed with the bicycles standing still. All tests were done with a Hybrid II 50th percentile crash test dummy placed in the saddle of the bicycles, with acceleration measurements in the head.Results: The crash tests showed that a sudden stop, e.g. a stick or bag in the front wheel, will result in a falling motion over the handle bars causing a forceful head impact while a sideways dislocation of the front wheel will result in a falling motion to the side causing a more moderate head impact. The falling motion varies between the different bicycle types depending on crash test scenario and speed. The pedelec had a clearly different falling motion from the other bicycle types, especially at a sudden stop.Conclusions: The study implies that it is possible to examine single-bicycle crashes using crash tests, even though the setup is sensitive to minor input differences and the random variation in the resulting head impact values can be large. Sideway falls with the bicycles standing still were easier to perform with a good repeatability and indicated an influence of seating height on the head impact.

Collision between a Pedestrian and Tram - Pilot Experiment

This article intends to present the first results of a long-term research project, which will result in developing a validated model of a pedestrian for the simulation of crash tests involving tram fronts and, where applicable, the fronts of other urban rail vehicles. The current phase of research includes results of the pilot experiment with a crash-test dummy, and these results supplement the results from simulations, thus demonstrating how important it is to pay special attention to the individual stages of a collision event and how important the localisation of and moulding by an individual tram's front panels are for the nature of the monitored stages. In the first stage, inertia of individual body segments plays a significant role, with the primary contact taking place between the tram's bumper and dummy's thigh. The dummy subsequently takes the shape of the tram's front with progressive bumps to shoulders and head following. At that moment, the tram brakes, and the dummy begins to disentangle from the front panel. The friction force between the dummy's soles and the surface of the rail track is very significant for the nature of this second stage of the collision event .The dummy then hits the ground. It is an accelerated fall, and under the given conditions, it is the stage that has the most devastating impact on the dummy. The simulation made shows the way to modify this dangerous stage to be less harmful to a pedestrian involved in a tram collision.

Head injury metric response in finite element ATDs and a human body model in multidirectional loading regimes

Objective: The objective was to quantify head injury metric sensitivity of the 50th percentile male Hybrid III, THOR, and Global Human Body Models Consortium simplified occupant (GHBMC M50-OS) to changes in loading conditions in loading regimes that may be experienced by occupants of spaceflight vehicles or highly autonomous vehicles (HAVs) with nontraditional seating configurations.Methods: A Latin hypercube (LHD) design of experiments (DOE) was employed to develop boundary conditions for 455 unique acceleration profiles. Three previously validated finite element (FE) models of the Hybrid III anthropomorphic test device (ATD), THOR ATD, and GHBMC M50-OS were positioned in an upright 90°-90°-90° seat and with a 5-point belt. Acceleration pulses were applied to each of the three occupants in the ± X, +Y, and ± Z directions, with peak resultant acceleration magnitudes ranging from 5 to 20 G and times to peak ranging from 32.5 to 120.8 ms with duration 250 ms, resulting in 1,248 simulations. Head injury metrics included peak linear head acceleration, peak rotational head acceleration, head injury criteria (HIC15), and brain injury criteria (BrIC). Injury metrics were regressed against boundary condition parameters using 2nd order multiple polynomial regression, and compared between occupants using matched pairs Wilcoxon signed rank analysis.Results: Across the 416 matched-simulations that reached normal termination with all three models, HIC15 values ranged from 1.0–396.5 (Hybrid III), 1.2–327.9 (THOR), and 0.6–585.6 (GHBMC). BrIC ranged from 0.03–0.95 (Hybrid III), 0.03–1.21 (THOR), and 0.04–0.84 (GHBMC). Wilcoxon signed rank analysis demonstrated significant pairwise differences between each of the three occupant models for head injury metrics. For HIC15, the largest divergence between GHBMC and the ATDs was observed in simulations with components of combined underbody and rear impact loading. The three models performed most similarly with respect to BrIC output when loaded in a frontal direction. Both the GHBMC and the Hybrid III produced lower values of BrIC than the THOR on average, with the differences most pronounced in rear impact loading.Conclusion: In conclusion, observed differences between the occupant models’ head injury metric output were quantified. Loading direction had a large effect on metric outcome and metric comparability across models, with frontal and rear impacts with low vertical acceleration tending to be the most similar. One explanation for these differences could be the differences in neck stiffness between the models that allowed more rotation in the GHBMC and THOR. Care should be taken when using ATDs as human volunteer surrogates in these low energy events.