Investigation of the inverse crashing behavior of thin-walled shells filled with Flamelet-Generated Manifold foam and prediction of energy absorption utilizing artificial neural networks

In this study, the energy absorption capability of thin-walled shells filled with foam with Functional properties with inverse deformation mode has been numerically evaluated. The relationship between force and shell displacement has been acquired using nonlinear finite element analysis and artificial neural networks. Three types of properties’ distributions have been studied for aluminum porous core foam under quasi-static vertical loading. Furthermore, by studying the geometrical parameters of the shell and the mechanical properties of the core foam, the impact parameters, including the specific energy absorption, have been investigated as the test objectives. The results demonstrate that the use of foam improves the displacement mode favorably. In such a way that in the shells without foam, the folds are irregular with a larger length, whereas, with the use of Foam, the Folds become more regular. Furthermore, the average 'crushing force and absorbed energy per unit mass increase with increasing foam density and the Shell’s Wall Thickness. The optimal geometric parameters obtained using a genetic algorithm and neural network demonstrated that the specific energy absorption capacity of the Shell einforced with foam can rise by 120% compared to a similar hollow specimen. In addition, the distribution of foam properties has a significant effect and increases the absorption of specific energy up to 66.5%. • Investigation of the inverse crashing behavior of thin-walled shells filled with FGM foam and prediction of energy absorption utilizing artificial neural networks. • Investigation of Meshing, boundary conditions and analysis method. • Investigation of the Grid independence prediction of energy absorption utilizing artificial neural networks.