Frequency-based substructuring for virtual prediction and uncertainty quantification of thin-walled vehicle seat structures 2024-01-2946

Finite element simulation (FE) makes it possible to analyze the structural dynamic behavior of vehicle seat structures in early design phases to meet Noise-Vibration-Harshness (NVH) requirements. For this purpose, linear simulations are usually used, which neglect many nonlinear mechanical properties of the real structure. These models are trimmed to fit global vibration behavior based on the complex description of contact or jointed definitions. Targeted design is therefore only possible to a limited extent. The aim of this work is to characterize the entire seat structure and its sub-components in order to identify the main contributors using experimental and simulative data. The Lagrange Multiplier Frequency Based Substructuring (LM-FBS) method is used for this purpose. Therefore, the individual subsystems of seat frame, seat backrest and headrest are characterized under different conditions. Furthermore, the complex connection interface is discretized by means of a high-resolution description with virtual points, which consider translational and rotational degrees of freedom. In addition, measurement uncertainties, including nonlinearity, are quantified to achieve greatest reproducibility. The vehicle seat structure, validated using the experimental FBS method, is then used for FE model comparison. The main contributor and vibration parameters are then identified. One of the conclusions confirms the FBS recommendation to characterize complex interface situations in thin-walled vehicle seat structures using experimental and simulative data.