Analysis and Measurements of Vehicle Door Structural Dynamic Response

Abstract: In order to reduce lead time and cost in the product development of vehicles more development will be made virtually. However, the predictability capability of simulation models is questioned and the simulation models need to be correlated versus hardware measurements and modeling techniques improved. As part of the process of vehicle system model capability improvements the main objective of this project is to improve the structural dynamic response prediction capability of a vehicle door simulation model in a free-free configuration under steady-state conditions. The actions performed can then be rolled down to: simulate eigenmodes and frequency response, perform hardware measurements, make correlations of simulations versus measurements, using modal assurance criterion, frequency response assurance criterion and sum-blocks, and update simulation model. These actions are performed for four successively more complex door structures starting from a door in white and ending at a trimmed door. The correlation status of the original model was only reasonably good for the door in white configuration. All other configurations displayed serious correlation mismatch. By replacing the existing antiflutter models (connecting the side impact rail to outer panel) in the door in white configuration with simple spring elements the correlation for the door in white configuration was improved. With the window and seals attached the correlation problems was solved by introducing stiffness in the plane of the window of the springs acting as seals. The idea was to take friction into account. Also, by adjusting the spring stiffness of the seals, fair correlation could be achieved. The most important issue is to relate these results to component properties known before building simulation models. The following two configurations need more attention for better correlation. By using more detailed models the correlation could be improved, which shows the obvious trade-off between accuracy and computational effort. However, improving the model detail level fall outside the limitations of this project.