Virtual Tuning and Optimization of Two-Wheeler Suspension through MBD Simulations 2021-26-0348

The design and tuning of stiffness and damping parameters of the suspension system is an essential step in arriving at the optimal riding comfort for a two wheeler. The conventional method of suspension design involves selecting baseline values of suspension parameters through bench-marking of similar vehicles, followed by extensive real life tests to tune these parameters over different load, speed and track conditions. This conventional approach is often time consuming and heavily dependent on physical prototypes of suspension. With the advent of tools to run optimization on Multi-Body Dynamic (MBD) simulations, a near complete tuning of suspension parameters can be performed entirely in the simulation domain. This paper presents an integrated approach towards the design and tuning of suspension stiffness and damping parameters using MBD simulations, by employing fully parameterized suspension stiffness and damping curves in the optimization process. The optimization is performed on an MBD model representative of the vertical dynamics of the two wheeler. Stiffness and damping curves are first parameterized in terms of common suspension design parameters, for example, one or more stiffnesses, preload, bump-stop and rebound-stop clearances etc. A design of experiment set up before the optimization details out the sensitivity of these design parameters on the ride quality, which tends to vary depending on the amplitude and frequency of the road perturbations. The optimization is then set up on the optimization tool and the working limits of the parameters are specified. The ride quality is evaluated using acceleration amplitudes at tactile points and Vibration Dose Values (VDV) on three different road profiles. The optimized design parameters are mapped out for both short and long amplitude road perturbations. Finally an approach to combine the findings of the three road profiles to arrive at an optimal set of suspension parameters is discussed.