A Physics-Informed Convolutional Neural Network with Custom Loss Functions for Porosity Prediction in Laser Metal Deposition

Physics-informed machine learning is emerging through vast methodologies and in various applications. This paper discovers physics-based custom loss functions as an implementable solution to additive manufacturing (AM). Specifically, laser metal deposition (LMD) is an AM process where a laser beam melts deposited powder, and the dissolved particles fuse to produce metal components. Porosity, or small cavities that form in this printed structure, is generally considered one of the most destructive defects in metal AM. Traditionally, computer tomography scans measure porosity. While this is useful for understanding the nature of pore formation and its characteristics, purely physics-driven models lack real-time prediction ability. Meanwhile, a purely deep learning approach to porosity prediction leaves valuable physics knowledge behind. In this paper, a hybrid model that uses both empirical and simulated LMD data is created to show how various physics-informed loss functions impact the accuracy, precision, and recall of a baseline deep learning model for porosity prediction. In particular, some versions of the physics-informed model can improve the precision of the baseline deep learning-only model (albeit at the expense of overall accuracy).

Investigation of the Influence of Direct Metal Deposition Modes on Microstructure and Formation of Defects in Samples of Heat Resistant Alloy

The article discusses the influence of technological modes of the DMD method on the macro- and microstructure of a heat-resistant nickel-based alloy to use this technology for heat-resistant materials in the manufacture of parts for combustion chambers in gas turbine plants.

Development of an Experimental Robotic Complex for Direct Metal Deposition and Testing of Deposition Modes for Heat-Resistant Powder Material

The paper considers the issue of optimizing the movement of an industrial robot used in additive manufacturing in the technology of direct metal deposition of parts. The developed mathematical model that takes into account the joint work of a six-axis robot manipulator and a two-axis positioner is described. The algorithm for calculating the motion based on the relative position of two adjacent points of the working tool trajectory relative to the rotary axis of the positioner with a given accuracy is described. The simulation of processing is carried out both when working only with the manipulator, and when working together with a two-axis positioner, and control programs with recalculated coordinates and rotation angles of the positioner are obtained.