Seatbelt and seatback control for occupant protection in frontal automotive collisions

This paper investigates the potential benefits of an imminent collision prediction system for improving occupant protection in a frontal automotive crash. Knowledge of an impending unavoidable crash is assumed to be known 100 ms before the crash occurs. A three dof human occupant model is developed using a Lagrangian approach to represent occupant translation with respect to seat, torso rotation and neck rotation. The performance of traditional elastic seat belts is compared with that of an analytically calculated seat belt law in which the force values are calculated in real-time so as to just prevent collision with car interior. Simulations verify that the analytical seat belt force calculation results in less force on occupant for the same level of safety. Furthermore, results show that knowledge of a future collision can be used to pre-tension seat belts but can provide no additional benefits, if seat belts are the only means for active occupant protection. However, if seat tilt-back can be deployed using an on–off mechanism, such predictive knowledge of a future collision can provide significantly improved occupant protection in terms of preventing occupant collision with car interior.