Effect of damping in complex eigenvalue analysis of brake noise to control over-prediction of instabilities: An experimental study

Disc brake noise is recognized as a major problem of the automotive industry. Various experimental and numerical techniques have been developed to model the noisy brake and investigate possible solutions. Developing a virtual model of the disc brake which can accurately reproduce the behavior of the brake unit under different conditions is a considerable step forward towards reaching this goal. Among various aspects of the analytical model of a disc brake, application of the correct value of damping based on the material properties and functional frequency range of each component is a significant factor in ensuring correct prediction of the brake system behavior. Complex Eigenvalue Analysis is well established as a tool for predicting brake instabilities which can potentially lead to brake noise. However, it is known to over-predict instabilities i.e. predict instabilities which do not occur in the real brake system. The over-prediction of unstable modes is thought to be as a result of insufficient damping in the model compared with the real brake system. For this reason, the Finite Element Analysis model of the brake unit needs to be tuned in terms of damping characteristics to ensure that the model replicates the system's real behavior. This study aims to tune the damping of different components of the brake unit Finite Element Analysis model using data from an experimental study. The study then compares the instability predictions of the tuned model and the un-damped model, and correlates them with the behavior of the same brake unit when tested on a dynamometer. This is intended to minimize over-predicted instabilities.