Simulating seat discomfort: An experimental design for using digital human models

Abstract

It was the aim of the project UDASim, funded by the German ministry of education and research, to develop a tool for car seat discomfort assessment within the virtual design phase. Seat discomfort depends on a variety of parameters whose correlations are unknown. Therefore, we have chosen an artificial neural network approach using the computational results of three digital human models as input. The training data were computational results based on data from a large-scale seat discomfort study with real test subjects. The digital human models simulated all subjects and experiment conditions, so the experimental design was essential for the project goal. We describe the conducted experiment, its design and the different measurements required by the digital human models. Additionally, we explain the related measurement systems and the choice of test subjects.