Finite Element Investigation of Head and Neck Injury Mechanism during Rollover Crashes

A set of state-of-the-art FE vehicle, restraint system, dummy, and whole body human models was constructed and validated against several sets of quasi-static and dynamic rollover tests. HIII dummy responses during three selected trip-over scenarios were simulated using an FE vehicle model equipped with four different roof stiffness values. It was found that the high dummy head accelerations and neck loads were mainly caused by the inertia of the occupant’s torso compressing the head into the roof/ground before any significant roof crush occurred. Therefore, roof crush is not causally related to the head and neck injuries during the simulated rollover scenarios. However, the roof stiffness of the near-side roof did affect the duration of the head-to-roof impact of the far-side occupants, if two consecutive roof-to-ground impacts occurred in a single roll. It was observed that the stiffer the roof, the lower the head and neck injury risks for the far-side dummies in this scenario. However, no trend between roof stiffness and injury risk was observed when the same simulation was run with the THUMS human body model. This difference in response is most likely due to the ability of the compliant human neck to change impact orientations with different vehicle models, significantly changing the head and neck responses. During all the simulated rollover scenarios, the effect of roof stiffness on the injury risks of near-side occupants was not consistent as no clear trend was apparent. More rollover tests using both the dummy and cadaver are needed to further validate the hypothesis proposed in this study. INTRODUCTION T he finite element (FE) method, the most up-to-date advanced technology available to aid all types of automotive safety research, has been routinely used in frontal, side, and rear impact analyses. However, it was rarely used for rollover simulations due to the relative long simulation time and complicated vehicleto-ground and occupant-to-vehicle interactions.