Evaluation of dynamic response variability in aluminum honeycomb sandwich panels using PCE and Kriging-based metamodel

The aim of this study is to model the dynamic response variability in commercial aluminum honeycomb sandwich panels. An experimental modal analysis is performed on 35 identical commercial AHSPs. Based on 10,000 samples, a computational model is constructed according to the estimated weight of the panel. The modal frequencies of the 35 samples for the first 10 flexible modes are compared with the computational results and deviations are referred to as errors. The thickness of the facing sheets and thickness of the cell wall of the core are considered as sources of uncertainty. A data-driven meta-model called PCE-Kriging is created.