Comparison of Dummy and Human Body Models in Automotive Side Impact Collisions According to the Regulatory Standards
Abstract Side-Impact car accidents are the second leading cause of fatalities in the United States. Regulatory standards have been developed for occupant protection in side impact car accidents using dummies or Anthropomorphic Test Devices (ATDs). Although the regulations are based on the use of ATDs, there might be differences between an actual human crash performance and that of a dummy crash performance. In recent years, technology has improved in such a way that crash scenarios can be modeled in various computational software. The human dynamic responses can be examined using active human body models including a combination of rigid bodies, finite elements, and kinematic joints, thus making them versatile to use in all crash test scenarios. In this study, the nearside occupants are considered as per regulatory standards set by National Highway Traffic Safety Administration (NHTSA). Vehicle side-impact crash simulations are carried out using LS-DYNA finite element (FE) software, and the occupant response simulations are obtained using MADYMO. Because the simulation of an entire FE model of a car and occupant is quite time-consuming and computationally expensive, a prescribed structural motion (PSM) technique has been utilized in this study and applied to the side-door panel with an occupant positioned in the driver seat of the car in MADYMO. Regular side-impact deformable barrier and pole test simulations are performed with belted and unbelted occupant models considering two different target vehicles namely — a mid-size sedan and a small compact car. Responses from the dummy and the human body models are compared in order to quantify differences between the two in side impacts. The results from this study indicate that human body model behavior is generally similar to that of dummy model in terms of kinematic responses. However, the corresponding injury parameters of the human model are typically higher than that of the dummy model.