Multi-Objective optimization of suspension kinematics

Abstract

Suspension kinematics design can be a challenging process. In racing and motorsports development processes, where the design schedule is tighter and the quality of kinematics are critical, the job requires even more maturity. The suspension designer must conciliate vehicle packaging and driver perception requirements whilst trying to make the best use of the tires at all possible times. Computational resources are constantly getting cheaper and can be used to solve the kinematics issue aided by optimization techniques. The optimization process can generate new engineering solutions that are not obvious to the human understanding and would be hardly achieved by the engineering team. Given the multi-objective nature of this type of problem and the lack of convexity between the objective functions, this work proposes the integration of an Evolutionary Multi-Objective Optimization (EMOO) to a suspension kinematics solver. The optimization yields a set of sub-optimal solutions that are ranked by a set of weighting functions and scaling factors. This process offers an excellent time-to-benefit ratio, once it can potentially reduce weeks of workload to a few hours of computational effort. This work illustrates the power of such optimization process with a case study where a Double A-Arm suspension system is completely synthesized by the optimization algorithm, which is composed by 19 objective functions split in 4 different movements.