Mid-Frequency Prediction Accuracy Improvement for Fully Trimmed Vehicle using Hybrid SEA-FEA Technique

The concept of using the capabilities of Statistical Energy Analysis (SEA) as the basis for a hybrid technique together with Finite Element Analysis (FEA) or measurements to make acoustic and vibration predictions for the mid-frequency range has been previously established. With advancements in computer memory and speed, FEA calculations such as drive-point mobility and mode counts can now be obtained for some vehicle components or assemblies up to 1000 Hz. This allows for a larger mid-frequency range to be modeled with a hybrid SEA-FEA model when the system is not suitable for modeling with either pure SEA or pure FEA. This can enable a big improvement of the speed and accuracy of a structural-acoustic prediction in the mid-frequency range and can be the best possible analytical prediction in this frequency range when hardware and measured data are not available or testing is to be reduced.

This paper will present the theory and advantages of using a hybrid SEA-FEA model and will show the mid-frequency model correlation potential for improvement by using component or assembly drive-point conductance information and the resulting stiffness multipliers based on the modal densities calculated from FEA. The structural responses at the instrumented locations are shown to be in good agreement with the predicted responses of the structures in the pure SEA model. It is demonstrated that the mid-frequency range predictions can improved with the least amount of time and modeling effort by selectively making use of FEA calculations for the most important subsystems and applying them to the SEA-FEA hybrid technique for a vehicle.