Paper published in Proceedings of the fourth Bicycle and Motorcycle Dynamics Symposium, 9–11 September 2019, Padova, Italy. Authored by Westerhof, B.E., de Vries, E.J.H., Happee,R., Schwab, A.L.
The use of motorcycle simulators enables manufacturers to develop new motorcycle technologies taking into account rider behaviour. Unfortunately, few motorcycle simulators are used in the development of new motorcycles and motorcycle safety systems. Furthermore, validation of the available motorcycle simulators is lacking and little design knowledge is available. This paper presents and evaluates a new motorcycle simulator and shows that it can be used in motorcycle research. The simulator features six degrees of freedom motion, upper body tracking, and stereo vision with head mounted display. The evaluation consists of three parts.
First, a new motorcycle dynamics model based on literature is developed in Simulink and parameterized to represent a Honda CB750. The motorcycle dynamics model shows realistic “countersteering” or non-minimum phase behavior. The eigenvalues of capsize and weave motion are qualitatively similar to those in the established literature.
Second, the simulator actuator dynamics are evaluated and adjusted.The motorcycle simulator uses a Stewart motion platform and handlebar control loader to provide motion cues and force feedback. It is shown that the adjusted Stewart platform and handlebar control loader have ample bandwidth and sufficiently small phase delays to accurately simulate the motorcycle dynamics.
In the third and final part, a human research approach is used to further evaluate the motorcycle simulator. In a speed perception experiment participants were asked to ride three different speeds on an infinite highway. Participants underestimated their riding speed, where speed perception at higher speeds was relatively more accurate. Both results are in agreement with published speed perception experiments in cars. A curve driving experiment was performed to evaluate steering performance and perceived simulator fidelity and workload, exploring effects of simulator motion and rider body tracking. The steering experiment showed good performance, with positive results for fidelity and modest workloads while motion sickness occurred in only a few participants. Platform motion significantly improved the rider‟s performance and perception of simulator fidelity.No statistically significant effects of upper body tracking could be established.