Instability prediction of a fully updated brake system with uncertainty in contact conditions

Brake squeal, as friction-induced audible noise above 1 kHz to 20 kHz, is a major concern to automotive manufacturers because of noise, vibration and harshness performance and warranty-claim related costs. The prediction of brake squeal is as difficult as ever due to nonlinearities, uncertainty boundary and operating conditions. The exact contact conditions between pad linings with rotor are often not known and difficult to model. The friction contact has been found to be multi-scaled, inhomogeneous, and highly dependent on operating conditions. In this study, the surface roughness of a pad lining is measured using a Nanovea profilometer with a P1 – OP3500 measurement pen. Statistical distributions are generated for describing the surface roughness. A FE full brake model is updated by modal testing of individual components, subassemblies and full assembled brake. The variation in roughness of pad lining is implemented into the FE model for analysing its effect on brake instability prediction using the conventional complex eigenvalue analysis. The instability predictions are compared with noise dynamometer tests. Results are discussed with a view to applying the methodology to quantify the reliability of squeal prediction against the uncertainty in contact conditions.