Quantifying EMC Measurement Accuracy Using Feature Selective Validation

Abstract

It is commonplace in the field of Computational Electromagnetics (CEM) for engineers to validate models against experimental results. In some cases, this is performed with little understanding about the accuracy of the experimental data used to validate the underlying calculations from which Electromagnetic models are formed. This paper therefore explores the accuracy and more importantly the areas of inaccuracy and variability that may be associated with experimental data. The Feature Selective Validation (FSV) method is used to assess each area of variability, and thus quantify the quality of test configurations and test samples. In examining experimental repeatability rather than comparison to electromagnetic analysis results, this paper concludes that, while substantial variation between experimental results can exist, the use of FSV provides considerable assistance in quantifying repeatability and therefore assigning confidence to measurements against which CEM results can be compared. While this paper is based on experience in the automotive sector, it is anticipated that these findings are more widely applicable.