Plateau stress prediction of roll formed A1050 porous aluminum by machine learning using X-ray CT images

Porous aluminum is expected to be used in various industrial applications due to its low specific density and excellent shock absorption properties. Recently, roll forming has been investigated as a method of shaping porous aluminum. It has been found that the pore structures of porous aluminum are maintained by these processes on softened porous aluminum immediately after foaming. However, it is not clear whether the mechanical properties of the porous aluminum are retained after roll forming. In this study, we attempted to predict the compressive strength of roll formed porous aluminum X-ray CT images using a machine learning model trained on unformed porous aluminum. By this prediction, we examined whether roll formed porous aluminum and unformed porous aluminum exhibit similar compressive properties. As a result, it was found that the machine learning model trained on unformed porous aluminum could be used to predict the compressive strength of roll formed porous aluminum. The results also suggested that roll formed porous aluminum indicate the similar compressive properties to unformed porous aluminum.