Development of a Procedure for the Validation of Statistical Energy Analysis Simulations

Abstract

This paper describes the development of an NVH measurement procedure that can be used for comparisons to Statistical Energy Analysis (SEA). In SEA, the outputs of the simulation are ensemble averaged quantities for each subsystem, which can be obtained in measurements by averaging some measurement point results. For several reasons, the number of measurement points must be as few as possible, but at the same time, they have to provide a well approximated averaged response of the system. The sufficient number of evaluation points and excitation load cases are determined via Finite Element (FE) simulations. It is shown that in case of a simple, flat plate, 17 randomly chosen evaluation points in at least 3 load cases are enough to properly approximate the SEA results.