Validation and optimization of the leaf spring multibody numerical model

Abstract

In this paper, a multibody approach for leaf springs modelling that may be used for full vehicle simulations is presented. Two different configurations of the leaf spring are investigated. The first one reproduces an actual mounting configuration of a leaf spring suspension in a vehicle, including the shackle. In the second configuration, the shackle is eliminated by mounting the leaf spring on linear bearings that are able to move on a guide rail. This allows for better spring rate evaluation, without the influence of shackle’s length and position. Both configurations were modelled in commercially available software LMS Virtual.Lab Motion and tested experimentally. Results of investigations have proved that developed model correctly reproduces behaviour of the leaf spring for dynamic simulations, with better convergence achieved for configuration without shackle. Initial results for static tests have shown significant error in behaviour of the numerical model and experimental test rig. In order to minimize observed errors, the optimization process of the model parameters was performed in Noesis Optimus by means of single-objective optimization. It was proved that presented approach may be used for proper modelling of the leaf spring behaviour, and this type of models may be effectively used in full vehicle simulations.