Dr.-Ing. Michael Klein, INTES GmbH, GERMANY

Virtual product development for brakes systems are widely accepted and used in industry. Using simulations in an early design stage reduces time to market, saves costs and improves the physical behavior of the brake system. But it is often neglected that many parameters are subject to uncertainties. Identification of parameters with uncertainties and an efficient integrated method to take them into account by brake system simulation are necessary.

Parameters of brake systems could be classified in two main categories. Application parameters and design parameters. The application parameters are e.g. brake pressure and rotation speed. The influence of the designer on these parameters is very limited, because they vary during application in a wide range. These parameters must be examined by sampling in their entire value range by simulation. Design parameters, like the shape, materials and the realization of all connections are based on decisions during the development of a brake system. Although precise decisions have been made, the parameters are uncertain. E.g. the component geometries vary in the production process and the material, especially the brake pad material, is not always exactly the same.

Measurements during production and during application deliver information about the distribution of the values of uncertain parameters. The task of simulation is to take into account those uncertainties for the calculation of sensitivities with respect to NVH behavior. However, the computational effort for such additional information should be kept as small as possible. The solution for maximum efficiency is the integration of the uncertainty analysis into the FEA software.

Solver integrated methods for this challenging task are available in the high-performance solver PERMAS. The evaluation is based on stability maps that contain the uncertainties. Advanced methods with control of the covered analyses reduce the effort to the possible minimum.

An example of CEA with several uncertain parameters shows the practical application of the advanced options of loop control. The mathematical benefits are used for the industry oriented numerical example to show the effectiveness of the approach.

Brake Squeal is an important and challenging phenomena to capture from an OEM's perspective, to avoid expensive warranty issues and provide squeal proof driving experience to its customers. It becomes critical to identify and take corrective measures in the initial design phase, to avoid multiple and expensive test on prototypes. Complex eigenvalue analysis (CEA) to predict the squeal frequencies is a well-established approach since last two decades. However, CEA has always been a domain for the experienced analyst, unexplored by the designers. In this study, an attempt has been made to automate complete CEA procedure to evaluate brake squeal characteristics which guides a novice designer to begin with a 3D CAD model and end up with the squeal propensity evaluation. Above procedure includes complete automation of brake model setup, including material properties, contact creation, boundary conditions and the result generation using a template-based definition. It also allows more robust and concrete evaluation of contact characteristics which is quintessential for mode coupling phenomena that ultimately leads to brake squeal. Traditional techniques would take an experienced engineer, anywhere from few days to few weeks in performing CEA. Using this automation, the complete CEA analysis time can be reduced to just 3 to 4 hours, enabling designers to have deeper insights at a very early stage of brake design and development.

"