Vehicle Model Simplification for Efficient Pedestrian Impact Assessment with Human Body Models

Abstract

Finite element simulations with Human Body Models provide detailed biomechanical information for pedestrian safety assessment, but their high computational cost limits their use in iterative vehicle studies. This work investigates whether this cost can be reduced by simplifying the vehicle model without significantly altering the pedestrian response. A 2020 Nissan Rogue finite element model was coupled with THUMS Pedestrian version 4.1 and simulated using LS-DYNA. Several reduced vehicle configurations were generated by replacing selected deformable components with nodal rigid bodies, while keeping the human body model unchanged. The reduced models were compared with the full vehicle baseline using runtime, vehicle internal energy, biomechanical outputs, and ISO/CORA-based correlation scores. Simulation 3 provided the best compromise between efficiency and accuracy, reducing runtime by 35.64% while achieving the highest agreement with the baseline.