Correlation of CAE Predictions to Actual Physical Tests with Production Variability Taken into Account

One of the critical items for a good matching between CAE predictions and actual hardware data is the input parameters variability. A vehicle component, for example, after manufactured shows imperfections that can be translated in small variation on nominal values of geometries and material properties. These variations of physical and geometrical parameters, fundamental data to construct a numerical model, cause impact on dynamic response. Analysis of this impact requires the use of appropriate stochastic methods for structural problems, mainly in correlation analysis between numerical and experimental results. This work assesses Body-in-White Point Mobility and Vibration Transfer Function correlation between CAE predictions and actual vehicle performance, including production variation and test uncertainty data.