This paper deals with the parametric shape optimization of a simplified model of brake system under squeal noise criteria. As brake squeal phenomenon induces under-quality perception for industrial structures such as cars and trains, its understanding and management are important challenges for future systems design. Hence, we expose an optimization methodology based on meta-model for a proposed computationally expensive stability criteria representing the squeal noise. Sensitivity analysis is first conducted to assess and validate the chosen geometrical parameters. Then, a Pareto front is obtained through optimization of the system, leading to a set of optimal solutions for theconsidered multi-objective case.