Towards Overcoming the Challenges of the Prediction of Brake Squeal Propensity

Abstract

Despite significant efforts made in the last two decades for the analysis and prediction of brake squeal propensity, brake squeal remains a major source of customer dissatisfaction and warranty-related costs. Brake squeal is a fugitive nonlinear self-excitation phenomenon induced by friction. Traditional linear complex eigenvalue analysis (CEA) combined with noise dynamometer tests has achieved some success in the analysis of brake squeal and in designing countermeasures in suppressing brake squeal. However, prediction of brake squeal propensity without experimental tests is generally unreliable because of three main challenges. Firstly, most conventional analysis methods used are linear but brake squeal could be caused by nonlinearities. Secondly, while nonlinear time-domain simulations have achieved some success, they require substantial high-performance computing resources and are too time-consuming for practical applications. Thirdly, even if the second challenge can be overcome, there are many determining and interacting factors that are not known exactly, such as material properties of pad and disc, operating conditions (brake pad pressure, temperature, speed), contact conditions between pad and disc, nonlinear boundary conditions and modelling of friction. In this chapter, examples are given to illustrate the success and limitations of the CEA. To address difficulties in predicting brake squeal posed by nonlinearity, uncertain contact conditions and friction modelling, examples will be given to illustrate how a stochastic approach using the traditional linear eigenvalue analysis could be applied to improve the prediction of brake squeal by identifying unstable vibration modes that are not too sensitive to these uncertainties.