Using optimization to auto-correlate suspension characteristics to K&C measurements

Erik Wendeberg
Master’s Thesis in the Automotive Engineering Programme
Department of Applied Mechanics
Division of Vehicle Engineering and Autonomous Systems
Vehicle Dynamics Group

Computer aided engineering is becoming an increasingly important tool in the automotive industry since it can reduce development time of new vehicles. However, in order to draw the same conclusions from test and simulation results it is important that the behaviour and characteristics of a simulation model match test data. Traditionally, to ensure that a suspension simulation model is accurate, it is correlated by a manual adjustment of the parameters in the model. This is time-consuming and error-prone. By automating the correlation process using a suitable optimization technique and a properly defined procedure, the process can be performed faster and the quality of the results can be improved since more parameters and objectives can be included. The aim of this study was to develop a well-defined correlation procedure, with minimal user input, that optimizes parameters in a suspension model so the behaviour of the model matches test data. A design of experiment study was conducted to analyse the influence of suspension parameters on corresponding suspension characteristics, and based on this a suitable correlation method and optimization model setup could be defined. By running the correlation procedure in the optimization software HEEDS MDO, connected with ADAMS/Car, suspension characteristics could be correlated to measurement data. The defined auto-correlation procedure was found to be effective and a front suspension assembly was successfully correlated to physical kinematics and compliance measurement results. However, the baseline suspension model has to be modelled correctly and include all the necessary variables in order to fully correlate the suspension simulation model. Some of the correlated suspension parameters were found to have optimized values outside normal production tolerances, in order to compensate for limitations in the simulation model, such as rigid modelling of components. By using the defined auto-correlation procedure, the correlation time was reduced and it is recommended that HEEDS MDO is used for future correlation of suspension assemblies. If the setup of the optimization model is adjusted, the defined correlation procedure can also be used to create suspension simulation models of competitor vehicles or optimizing suspension design concepts to meet requirements.
 Download (PDF, 1.99GB)