Laser powder-bed fusion (LPBF) process, as one of the most widely used technologies of additive manufacturing, enables fabrication of parts with intricate geometries. The choice of process parameters in this technology plays a major role in defining the microstructural, mechanical and surface properties of the fabricated parts. In this study, the effects of LPBF process parameters on static tensile properties (including yield strength and ultimate tensile strength and elongation) of Ti-6Al-4V samples were investigated using artificial intelligence methods. Deep learning approach was employed by using neural networks for prediction, optimization and parametric and sensitivity analyses. Relevant experimental data available in the literature were collected to feed the network. Stacked auto-encoder was assigned to the networks for high accuracy pre-training. LPBF process parameters including laser power, scanning speed, hatch spacing, layer thickness and sample direction were regarded as inputs while yield strength, ultimate strength and elongation were considered as outputs of the neural networks. The obtained results indicate the high potential of neural networks to be used as a powerful tool for process parameter optimization for enhanced mechanical performance of additive manufactured parts.
Application of artificial intelligence to optimize the process parameters effects on tensile properties of Ti-6Al-4V fabricated by laser powder-bed fusion
Maleki E.;Bagherifard S.;Guagliano M.
2022-01-01
Abstract
Laser powder-bed fusion (LPBF) process, as one of the most widely used technologies of additive manufacturing, enables fabrication of parts with intricate geometries. The choice of process parameters in this technology plays a major role in defining the microstructural, mechanical and surface properties of the fabricated parts. In this study, the effects of LPBF process parameters on static tensile properties (including yield strength and ultimate tensile strength and elongation) of Ti-6Al-4V samples were investigated using artificial intelligence methods. Deep learning approach was employed by using neural networks for prediction, optimization and parametric and sensitivity analyses. Relevant experimental data available in the literature were collected to feed the network. Stacked auto-encoder was assigned to the networks for high accuracy pre-training. LPBF process parameters including laser power, scanning speed, hatch spacing, layer thickness and sample direction were regarded as inputs while yield strength, ultimate strength and elongation were considered as outputs of the neural networks. The obtained results indicate the high potential of neural networks to be used as a powerful tool for process parameter optimization for enhanced mechanical performance of additive manufactured parts.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.