Abstract
During a collision the vehicle occupant is exposed to substantial loads which might be a reason of the occupant injuries. Correct estimation of injury severity demands that the time history of the forces and accelerations acting on the occupant is known. A computer aided traffic accident analysis is introduced, together with human body mathematical models purposed for the occupant injury analysis and prediction. A human body modelling procedure was developed. The occupant body was modelled as a multibody system with rigid body segments connected. The body segment geometrical and inertial properties were estimated using computer tomography technique. Sled tests were conducted on a dedicated test facility. Comparison of experimental data and computer simulation revealed an influence of joint resistive properties on the occupant motion in collisions. The difference between the measured and the simulated response was minimised using optimization methods. Highly individualised human body modelling procedure enabled better prediction of the occupant motion during vehicle collision and thus more precise estimation of possible injuries in real-life situations.