Identifying the key injury regions of vehicle occupants is an important step to further improve the safety of future vehicles. Their estimation for occupants of future vehicle concepts with novel restraint strategies poses a particular challenge. While accident databases can offer valuable insights, they are limited to the current vehicle fleet. Human body models (HBMs) are an interesting alternative, since they enable the investigation of injury mechanisms and the estimation of effects of changing boundary conditions and protective countermeasures. To produce reliable results though, these models need to be set up with representative boundary conditions and reliable injury predictors have to be identified.

As a first step, the present study tried to match the relative injury frequency across all body regions observed in a real-world accident database sample to simulation results obtained with a detailed HBM. The occupant model was seated in a simplified vehicle model and variability in the vehicle fleet was accounted for by varying loads, boundary conditions and restraint system parameters.

Differences between the predicted frequencies of injury were identified – most prominently for the head. Since this method can be seen as a plausibility check for the human body model, thresholds for the strain-based assessment are discussed.