Transition in deformation behavior of porous components containing optimized pore group network

抄録

Porous materials are anticipated to mitigate stress shielding in implants by providing sufficient deformability to the metallic materials. However, the interaction of stress concentrations may result in pore connections during the fracture process. This study aims to develop a pore group optimization method to remove the interaction of stress concentrations using characteristics determined by Persistent Homology. The critical porosity corresponding to the optimization limit is discussed in terms of the deformation behavior and the transition of fracture mechanisms before and after reaching this critical porosity.

2D complexes determined using pore center distances are key factors in optimization. The interaction of stress concentrations is considered only within each complex, and the pore locations are updated while maintaining the structure of the complexes. The objective function includes the 2D complex diameter, the approximated stress concentration factor, and the penalty term. The pore locations were optimized using a gradient algorithm to minimize the objective function. Consequently, pore groups are dispersed to decrease the interaction of stress concentrations while maintaining the designable region restriction. Additionally, critical porosity was observed, as expected, indicating the limit of dispersion within the restricted region.

The deformation behavior of the optimal pore network structure in tensile tests is discussed. Specimens with initial random pores and optimal network pores were fabricated using the powder bed fusion method of additive manufacturing. The total elongation increased in the optimal pore network structure before the critical porosity was reached. The mechanism of the ductility increase was explained by focusing on the 1D complex distribution and its morphology. Finally, the network characteristics describing the mechanical properties and property improvement by optimization of the network characteristics are discussed.