Multiscale approach to determine the anisotropic mechanical properties of polyisocyanurate metal panels using FEM simulations

Building on previous publications, in this work an existing mesoscale simulation model was first extended to predict the non-linear anisotropic mechanical behavior of foam structures. Subsequently, a comparison between experimental and simulated data was performed to validate the selected simulation parameters. Here, good agreements were observed, especially for compressive loading. Based on these results, a multiscale approach was described in which the local foam properties were applied to macroscale models of PIR-based metal panels to predict the overall component properties. A further comparison of selected macroscopic experimental and simulated data also showed a good agreement. Due to the high computational and modeling effort, a sensitivity analysis was finally performed on the influence of the resolutions of the local mesoscale foam properties on the macroscopic metal panel properties. It was found that for the metal panels considered, it is sufficient to model the (almost) homogeneous core layer and outer edge areas with one layer each. These findings can also be applied to other metal panels with different thicknesses for future investigations, since it is known from previous investigations that the metal panels have qualitatively comparable property distributions.

Automated summary

Building on previous publications, this work extends an existing mesoscale simulation model to predict the non-linear anisotropic mechanical behavior of foam structures. Comparison of experimental and simulated data validates the selected simulation parameters. A multiscale approach is described to predict the overall component properties of PIR-based metal panels using local foam properties. Sensitivity analysis on the influence of local mesoscale foam properties resolutions on macroscopic metal panel properties is performed.