The incidence pattern of vulnerable road user (VRU) injury is modified by the transport infrastructure on the ground. Recent improvements in modelling capability introduce the possibility of evaluating performance of safety measures by incorporating a distribution instead of a single standardized test. The question addressed in this work is to estimate the probability distribution of crash initiation modes based on post‐crash observations.
A multi‐body simulation based approach of using adaptive Markov Chain Monte Carlo (MCMC) sampling techniques implemented in statistical software “R” has been used. The two input variables under study were location of VRU along the lateral axis in front of vehicle and shape of vehicle front. The observed output was the locations of head hits on vehicle front profile.
Simulation studies on VRU head crash locations from the German In‐depth Accident Study (GIDAS) with the severity of injury and the crash vehicle details have been used. These converge to a distribution with mean around the center plane of the vehicle with spread of around 30 cm in one standard deviation.