In order to improve the pedestrian safety during an impact with a vehicle, subsystem tests have been defined to evaluate the aggressiveness of the frontend of cars. These subsystems tests have to be reproducible and are representative of the three decomposed impacts of the pedestrian with the car:​ lower leg on the bumper, upper leg on the hood, head on the hood or the windscreen. The velocity, angle and mass of the adult headform impactor and its impact area are invariable parameters. Upper legform impactor parameters are determined by vehicle characteristics. Lower legform impactor parameters are invariable (velocity and positioning). Nevertheless, these decoupled tests do not take into account the influence of the whole body on impacts. Therefore, it appears important to compare these subsystem tests with global conditions observed in real accidents. The objective of this paper is to perform this work on two French vehicles. Concerning the global conditions, four full-scale experimental tests with PMHS and the associating multibody numerical simulations were performed in classical (lateral impact for the pedestrian, centred for the vehicle) and real configurations.

In that way, two real accidents have been chosen in this impact configuration with a velocity value close to 40 km/h. Each reconstruction of accidents is based on In-Depth Accident Investigation first. Then, a parametric study using multibody models, validated with an experimental test, gives a hypothetic initial configuration of the accident. This configuration is used to put on an experimental reconstruction. Then, results from numerical and experimental studies are compared for the adult headform, the upper legform and the lower legform impacts. Finally, a global comparison is analysed more specifically on injuries not include on the subsystem approach. These injuries are also compared to Accidental Database to know whether their proportion is important or not.