Application of Design for Six Sigma methods in the development and validation of chassis control systems

Due to the increasing number of body styles, engine/gear combinations and the variety of modern control systems, the use of simulation techniques becomes more and more important in the development process of classical chassis as well as chassis control systems. Following GM’s Road-Lab-Math (R-L-M) strategy, the number of prototype cars in ongoing and future vehicle programs will be significantly reduced. The goal is to cancel one hardware step for all future vehicle platform derivatives [1, 4]. Therefore, this paper describes the success of CAE utilization within chassis development and – based on that – in virtual application and release of chassis control systems. Since not all upcoming vehicle variants can be simulated, even with the inhouse developed Test-Automation [2] (which includes automated post-processing) it is necessary to reduce the complexity of this task. An ongoing Design for Six Sigma project (DFSS) investigates various methods to study the influences of different vehicle parameters on the control systems’ performance of the overall vehicle based on vehicle dynamics simulation. After constructing a kriging surface, among other methods, ANOVA (Analysis of variants) is employed to analyze the influences of (deterministic) vehicle parameters like e.g. inertia, CG height, axle load distribution etc. as well as their interaction (see Fig. 1). Fig. 1: Example of the influences of vehicle parameter variations on one dynamic vehicle state variable of interest To create the necessary data, an inhouse developed tool (“Kriging Wizard” developed by GM R&D Warren, [6]) has been combined with a Software-in-the-Loop simulation (SiL) environment currently being used at the GME Engineering Center. This SiL/HiL environment (Fig. 2), consisting of the IPG CarMaker for the vehicle dynamics simulation and the respective controller codes of several suppliers combined in a Simulink environment, has already proven its validity in previous vehicle projects [3, 5]. The described simulation environment allows for fast data generation and analysis. Fig. 2: HiL/SiL – simulation environment for chassis control system development and validation 1 1 1 1