Road safety is a serious problem throughout the world. Every year, more than 1.17 million people die, and over 10 millions are injured in road crashes around the world. Pedestrians constitute 65 percent of the road crash related fatalities worldwide. The lower limb is frequently injured in pedestrian accidents due to the initial impact with the vehicle front. Lower limb injuries, although are non-fatal, they cause long-term impairments and disability, and are therefore associated with high societal cost.

This dissertation presents a body of research trying to enhance the understanding of injury mechanisms and tolerance levels of the lower limb in car-to-pedestrian collisions. For this purpose, a detailed finite element (FE) model of the human lower limb was developed and validated. The geometry of thigh and leg was extracted from the Visible Human Male Project and the geometry of knee was obtained from the 3D-CADBrowser solid models. The two data sets were merged together and scaled to represent the lower limb of a 50th percentile male in a normal standing position. Structured topologies for the lower limb components were created to generate the FE mesh consisting of mostly hexahedral elements and to facilitate the study of mesh convergence. The model consisted of the lower limb long bones, patella, flesh, skin, menisci, the four major knee ligaments, and the patellar tendon. For studying the whole body kinematics the FE model was linked to rigid-body segments of the GEBOD model.

The material properties were selected based on the current knowledge of the constitutive models for each tissue. For flesh, skin, and ligaments, where no reliable data could be found in the literature, new constitutive properties were determined based on FE optimization of cadaveric tests. The validation process of FE lower limb model have started at the component level with several tests specific to the pedestrian accident loadings. The bending and failure properties of long bones were validated against high and low speed tests. To validate the contact interfaces, material interactions, and tolerance limits, three components of the model, namely, the distal thigh, the proximal leg, and the knee were validated against dynamic lateral bending and combined (bending and shear) tests.

The results obtained in the lateral impact leg validation using the proposed flesh and skin models show to be more biofidelic than the data used in the previous FE models. Based on the simulation results, two lower limb tolerances published in the literature were proposed to be reduced. From the simulation results of the leg in three-point bending, it was shown that the mid-leg bending moment tolerance was overestimated by 5-6%. The lateral bending tolerance of the knee recommended by the European Enhanced Vehicle-Safety Committee (EEVC) was found to be overestimated by over 100 %. This finding is in accordance with recent experimental data, suggesting that the lateral bending tolerance of the knee should be reduced.

As computer simulations showed, the FE model developed in this work has the capability to reproduce bone fracture and ligamentous damages in a typical pedestrian impact scenario. Therefore, the model can be used to evaluate the aggressiveness of new front-car designs and pedestrian safety systems.