Suspension System Design for Revolve NTNU

Abstract

This thesis involves modeling and simulation of the dynamic behavior of Revolve NTNU's race car. The purpose of vehicle dynamics simulations is to predict race car behavior during practical testing. Being able to apply informed setups to the race car prior to driving, as well as based on observations at the track, is crucial for achieving maximum performance in the limited test time. A broad approach to simulations and surrounding aspects is taken in an effort to cover all the important aspects of vehicle behavior. Simulations range from simplified models like the quarter-car and half-car model used for transmissibility analysis, to full car simulations in CarMaker. Modelling of the suspension and aerodynamics in CarMaker's interface with Simulink is covered. How the "third spring" concept, which has been implemented into the suspension, effects vehicle behavior and how it should be set up to utilize the aerodynamics is explored. The characteristics of an air shock with exceptionally low weight, originally design for mountain bike use, are uncovered through testing. Most of the adjustment range is found to be outside the range that is suitable for Formula Student use. A comprehensive model for estimation of tire normal forces based on linear displacement sensors mounted to the dampers has been created to aid evaluation of vehicle performance. The model includes geometric effects of the suspension. Initial evaluation indicates viability, but inaccuracies of up to 20\% must be expected. Limit and transient lateral vehicle simulations are conducted in CarMaker. The results are seen in light of a quasi-static method to indicate correlation. The quasi-static method provides an approach to quickly evaluate sensitivities of multiple suspension setups. The torque vectoring algorithm is seen to have most say in transient conditions, reducing time lags and increasing response.