Influence of L-PBF additive manufacturing parameters on the residual stresses and thermal distortions in AISI 316L stainless steel parts

The work aimed to numerically model through the Finite Element Method (FEM) the distribution of residual stresses and thermal distortions in parts generated by Laser Powder Bed Fusion (L-PBF) in stainless steel AISI 316L and validate the results obtained through experimental measurements on previously manufactured parts.The design methodology followed a numerical approach through the Finite Element Method (FEM), the distribution of residual stresses and thermal distortions in parts generated by Selective Laser Powder Bed Fusion (L-PBF) in stainless steel AISI 316L and the FEM approach was validated trough the results obtained through experimental measurements on previously manufactured parts. The influence on three levels was verified through complete factorial planning of some manufacturing parameters, such as laser power, speed, and distance between scans (hatch), on the stress and distortion results of the samples and also on the samples simulated by FEM.When results were compared about the average diameters, a relative error of less than 2.5% was observed. The average diameter was influenced by power and speed. Increasing power decreased the average diameter of the samples, while increasing speed and hatch increased the average diameter. When results are compared to measure the residual stresses, it is observed that the relative error was less than 1%. Power, speed, and the hatch itself influenced the residual stress. Increasing power increases residual stress while increasing speed and hatch decreases residual stress. The cooling rate and the transient thermal history are the control factors that influence the residual stresses and are directly related to the process parameters. The computational modelling followed by measurements and calibrations carried out in the experimental stages proved to be efficient and enabled the reproduction of thermal distortion and residual stresses with statistical confidence.Following the research, the aim is to evaluate the prediction of thermal distortions and residual stresses using the machine learning approach. Future research will study heating the building platform, which should also impact residual stresses.Based on the results obtained in this research, it will be possible to select better additive manufacturing parameters for manufacturing 316L stainless steel parts. The parameters evaluated in the work were laser power, scanning speed, and hatch.The innovation of the work lies in the robust simulation of the thermo-elastic behaviour of samples subjected to the additive manufacturing process, where it was possible to accurately relate the thermal distortions and residual stresses that appeared in the samples printed with the parts modelled by the FEM. The numerical-experimental validation makes it possible to extrapolate the studies to several other manufacturing parameters using only computational simulation and work with a more significant amount of data for a prediction study.