Comparison of approximation methods forpredicting structural response of a honeycomb panel

Abstract

This paper discusses the findings of a study comparing different approximation models or machine learning models for prediction of structural responses of a honeycomb panel. Three different approximation models have been compared- Response Surface Model (RSM), Radial Basis Functions (RBF) and Universal Kriging method (UK). For each approximation model, average relative error between the predicted response and FEA response is compared for different designs. The findings are summarized and potential next steps to fortify the study are listed.