Thesis for the Degree of Doctor of Philosophy

In vehicle–pedestrian crashes, head injuries account for an overwhelming percentage of all severe and fatal injuries. These injuries are caused by the linear acceleration and rotation of the head. To mitigate head injuries, tools such as Human Body Models (HBMs) are used in the development and evaluation of pedestrian safety systems. The tools need to be compared with experimental data to evaluate their biofidelity. Previous studies regarding full-scale pedestrian experiments with post-mortem human subjects (PMHSs) have mainly provided two-dimensional linear trajectories and injuries. Six-dimensional linear and angular whole-body kinematics from full-scale pedestrian experiments are scarce. Detailed data on the subject’s anthropometry and initial body posture would increase the quality of simulations but are rarely published. The main aim of this thesis is to quantify six-dimensional head translational and rotational kinematics in car–pedestrian crashes prior to head impact against the vehicle. This aim is pursued by means of PMHS testing and Finite Element (FE) simulations with the Total Human Model for Safety (THUMS) version 4.0. The PMHS data are generated to provide HBM evaluation data and to investigate how pedestrian anthropometry and minor differences in initial stance influence head and upper body kinematics in car–pedestrian crashes. Additional aims are to evaluate THUMS in pure shoulder impacts and on a full-scale level, and to provide full-scale experimental data and pragmatic HBM scaling methods to industry and academia. Six-dimensional kinematics of the head, spine, pelvis and shoulders were quantified in five new full-scale pedestrian PMHS experiments with a small sedan. Varying anthropometry and minor variations in initial posture influenced pelvic sliding over the bonnet and ipsilateral upper arm responses, which in turn influenced head kinematics. THUMS was generally biofidelic although the arm abduction and the neck stiffness should be improved. In full-scale simulations, the best pragmatic scaling method was to use two scaling factors to adjust height and weight, and to translate THUMS to adjust pelvis height. Overall, the findings in this thesis increase the knowledge on how pedestrian upper-body 6DOF kinematics influence head kinematics. They highlight the importance of elbow and shoulder impacts and will thereby contribute to increase the quality of testing and simulating. Adding new inspiration for novel pedestrian safety systems, this work will contribute to decreasing pedestrian fatalities and mitigating pedestrian injuries.