Prediction of Part Quality in Laser Powder Bed Fusion Metal Additive Manufacturing
MetadataShow full item record
This work proposes physics-based analytical modeling methods for the prediction of part quality in laser powder bed fusion (LPBF) metal additive manufacturing, including lack-of-fusion porosity, keyhole-induced porosity, balling phenomenon, and surface roughness. The temperature distributions in the products in the conduction and keyhole melting modes are predicted by analytical thermal models with closed-form solutions, with the consideration of laser power input, melting mode-dependent laser absorptivity, powder bed material properties, and heat loss through part boundaries. The geometric characteristics of vapor depression and molten pool are obtained by comparing the calculated temperature profiles with the boiling and melting points of the material, respectively. The keyhole porosity, lack-of-fusion porosity, balling defect, and surface roughness are predicted by physics-based analytical models based on the information of process conditions (i.e., laser power, scan speed, hatch space, layer thickness, scan strategy, laser spot size etc.), molten pool dimensions (i.e., width, depth, length, height, contact angle), vapor depression depth, powder packing densification, characteristics of pore formation, and powder material properties. The occurrences of these defects and the process map of LPBF are also predicted by physics-based analytical models. All the proposed models do not rely on any finite element-based numerical iterations, which ensures their high computational efficiency. To validate the proposed models, the predicted results under different combinations of process conditions are compared with the experimental data of Ti6Al4V, Inconel 718, Inconel 625, and SS 316L. All the predictions show good agreement with the experimental results, which demonstrates the acceptable and encouraging predictive accuracy of the proposed models. The sensitivity analyses of part quality to process parameters are also studied. The proposed analytical models can help the researchers understand the underlying physics in LPBF process and they can work as efficient tools to optimize the process conditions of LPBF so that to fabricate products with high quality.