Development and performance assessment of different squeal index metrics based on experimental data

As an NVH problem observed in automotive brake systems, the significance of brake squeal phenomenon still persists; and the estimation of squeal propensity is a crucial operation. Thus, developing different means to define squeal propensity is an important task. As being one metric, squeal index is widely used for this purpose; and there already exists several different squeal index formulations in the available literature. The main objective of this study is to develop new squeal index formulations and assess their performances. Though, unlike the existing formulations, not only the measured data but also the operating and design parameters of the system of interest are also incorporated in the proposed formulations. The experimental data used for squeal index formulations are obtained on a mass-sliding belt experiment, where squeal-like behavior can successfully be observed. Experiments are conducted at different operating and design parameter conditions. First, cases with presence and absence of squeal-like behavior are observed on the measured data in time and frequency domains. Then, eight different squeal index formulations are developed, where different set of parameters are utilized in each formulation. Some of these parameters are extracted from the measured data (time and frequency domains), and the rest are adopted from the key operating and design parameters. Finally, the performances of proposed squeal indexes are assessed through their relative amplitude variations. In conclusion, the second squeal index formulation is found to have superior performance for distinguishing the effects of operating and design parameters on the occurrence of squeal-like behavior. Furthermore, the sixth squeal index formulation is found to have the best performance for differentiating the effect of spring stiffness on the initiation of squeal-like behavior.