Experimental-based positioning and simulation of Human Body Models in Bicyclist Crashes

As cycling continues to grow as a mode of transportation, an accurate and posture-aware representation of bicyclists in vehicle-impact scenarios becomes increasingly important. This thesis develops an integrated framework that combines realistic bicyclist posture measurement, automated positioning of human body models (HBMs), and finite element (FE) crash simulations to evaluate bicyclists’ injury mechanisms across key body regions. Bicyclist postures were experimentally captured using a photogrammetric motion-tracking approach during active pedaling. Male and female participants were measured while riding city, mountain, and racing bicycles and posture data were collected across multiple pedal angles, revealing substantial inter-subject variability. To translate these data into simulation, a landmark-based positioning methodology was developed. The method establishes direct correspondence between experimentally measured anatomical landmarks and skeletal landmarks defined on the HBM skeleton and applies a predefined kinematic hierarchy to ensure consistent joint articulation. This approach enables fast, reproducible, and user-independent positioning of HBMs into experimentally derived postures, making large posture-driven studies feasible, something not previously achievable in bicyclist safety research. Using this positioning framework, a parametric FE crash simulation campaign (108 simulations) was conducted with VIVA+ HBMs. The simulations examined the influence of vehicle speed, sex, bicycle type, pedal angle, and posture variation on injury outcomes. Injury assessment focused on the head, thorax, and lower extremities, using both global and tissue-based injury criteria. While vehicle speed remained the dominant influencing factor, posture-related parameters significantly affected injury risk, particularly in the lower extremities and head, and revealed sex- and bike-specific injury mechanisms. Overall, the thesis demonstrates that posture-aware modeling is essential for meaningful bicyclist injury assessment and provides a foundation for improved vehicle safety evaluation, the development of posture-aware test procedures, and future countermeasure design.