Abstract: It is a primary topic to choose suitable rubber bushing model for vehicle chassis system dynamics simulation and analysis, taking the model complexity, prediction accuracy and identification difficulty into consideration. In this paper, four semi-empirical dynamic models for rubber bushing, including linear Kelvin-Voigt model and three-parameter Maxwell model, nonlinear BERG model and DZIERZEK model, are investigated in respect of model characterization, parameter identification approach and prediction accuracy. Based on the comparisons, the characteristics and application of each model are described. It is found that the Kelvin-Voigt model and three parameters Maxwell model are simple and easy to get model parameters, but they have a low precision and only can be used for qualitative analysis. The BERG model can simulate the amplitude effect by introducing a friction element, and it has obviously improved accuracy while requiring appropriate identification experiments. The DZIERZEK model consists of both non-linear elastic spring and friction element which can describe nonlinear viscoelastic behavior with amplitude dependency. It has the best precision but needs much more identification tests due to complex model form. Both BERG model and DZIERZEK model are suitable for quantitative simulation analysis。

Key words: Parameter identification, Prediction precision, Rubber bushing, Semi-empirical parameterized model