Influence of Flexible Bodies on Subjective Response and Driving Simulator Performance

Abstract

Conventional real-time Driver-in-Loop (DiL) methods use low-fidelity vehicle models in driving simulators.These simplified vehicle models compute the vehicle states from fewer bodies, where their relative motion is mapped in pre-simulated lookup tables. With the improvements of real-time capability in multi-body software, a question arises on how these high-fidelity vehicle models behave in DiL assessments and how they affect the performance of driving simulators. The purpose of this thesis is to study the influence of flexible bodies on driving characteristics and the influence of these flexible bodies on the real-time performance of a simulator. Vehicle variants of different complexity are made by reducing the modal matrix of the flexible bodies. The goal is to develop a method to generate RT-MBD models that captures the effect compliance (through dynamic loading during driving) on the driving characteristics, while maintaining the real-time constraint. Results show that the computational cost is not weighted equally between the eigenmodes of flexible components. The geometric complexity of a flexible body is a significant factor in determining a model’s computational efficiency. It is observed that while some characteristics from subjective analyses can be correlated to objective metrics, vehicle variants with similar objective metrics from certain driving scenarios show significant differences when driven in a simulator.