Abstract
Squeal noise is one of the on-going problems of braking systems that directly relate to customer requirements, and various studies have been conducted in automobile/components manufacturing companies to solve that NVH issue. Complex Eigenvalue Analysis (CEA) is widely used in the automotive industry to reduce squeal noise because it has advantages of fast calculation and easy application. However, the methodology of squeal noise reduction using the complex eigenvalue analysis, especially about the caliper assembly system, shows different tendency depending on the vehicle type and the squeal generation mode. Therefore, most engineers tend to increase the system stiffness and weight without the knowledge base in order to solve the squeal problem. This kind of trend is not an appropriate way related to the activity of vehicle weight reduction to increase fuel efficiency, and it is time to continuously study deep into weight reduction as well as refinement of NVH comfort. In this study, a shape optimization method which can decrease part weight and squeal instability (analytical indicator of squeal noise level) is adopted to develop a robust system for squeal. This parametric shape optimization is driven by a wedge-constrained liner algorithm, which is a derivative free mode. Also, the effect of shape changes in squeal instability can be directly detected by using shape basis vectors. Through a stability map, the caliper bracket which is the most effective one for reducing its weight and squeal instability is selected as a target component. Consequently, the shape optimization leads to the final design of the caliper bracket for systematic improvements. To verify the feasibility of the proposed design, statistical analysis of CEA results is performed by applying a methodology in terms of DOE (Design of Experiments) based sampling method. Total 243 full matrix combinations of simulation condition are constructed considering not only the dispersion of material properties about major parts among brake corner module but also the rotational speed of disc when braking. Statistical analysis of the squeal reduction effect is conducted to examine the usefulness of the optimization result. Selecting a vehicle that is a squeal noise issue at 4 kHz or less low frequency band, optimization analysis and statistical verification are performed using techniques of complex eigenvalue analysis. Based on the result of sampling analysis, a stability map can be generated to select the most effective parameter for decreasing squeal noise. After that, the final design which can meet the target of weight and squeal instability reduction is derived. Through this study, it is possible to suggest a new analysis process that can overcome the limits of complex eigenvalue analysis and derive a realistic improvement of squeal noise. In the future, more advanced statistical analysis such as Latin hypercube sampling method will be discussed.