Relating scatter in occupant injury time histories to instability in airbag behaviour

Dealing with natural variation in input parameters and environmental conditions presents the automotive industry with significant challenges. Lack of consideration of variability in results can lead to unpleasant surprises during testing, with a consequent risk of unplanned cost and delay. In a purely virtual product development world, analysis techniques must lead to designs that are robust with respect to external noise sources, in order to minimise test-to- test and test-to-prediction variation. This paper discusses some of the issues faced in dealing with variability in an occupant restraint system, and presents an analysis approach that is helping to provide insight into causes of scatter, leading to potential design improvements to help reduce it. Conventionally the CAE process has used nominal values for input parameters, and has been satisfied with single, deterministic solutions. In many cases this approach is based on unreasonable assumptions, and a structured consideration of variability is vital. In this context we describe an example where Principal Components Analysis has been used to study scatter in an airbag model. Building on previous experience with the application of this technology to deformed geometries, the technique has been extended to allow a consideration of scatter in curves, as exemplified by the set of chest acceleration time-histories shown in figure 1. The mathematical background to the PCA method, as implemented in Diffcrash, is presented, and its extension to curves is explained. It will be shown how scatter in two crash dummy channels can be related to each other and to airbag deformation behaviours, as an aid to developing design improvements. Virtual techniques have much to offer in understanding and managing scatter in physical systems, and the consideration of variability in the CAE process is slowly becoming more common-place. The PCA approach presented here is a useful addition to the toolset available, giving valuable insight into physical phenomena.