Shreyes N. Melkote’s research while affiliated with Georgia Institute of Technology and other places

Wire-Arc Directed Energy Deposition (Wire-Arc DED) is a promising metal additive manufacturing process due to its high deposition rate and ability to produce large parts. However, residual stress and geometric accuracy challenges persist. While interlayer machining in Wire-Arc DED has shown potential to improve geometric accuracy and mechanical properties, its impact on residual stress in the hybrid process remains unexplored. In this regard, developing accurate models is crucial for understanding and optimizing the residual stress in Hybrid Wire-Arc DED. This paper investigates the challenge of predicting the residual stress in Hybrid Wire-Arc DED using the Finite Element Method. Interlayer milling interventions are simulated by modelling material removal as a predominantly geometric effect through element deactivation, excluding the thermo-mechanical effects of cutting. We demonstrate the limitations of this approach through simulations and experiments, highlighting the need for improvements in modelling the residual stress induced by interlayer machining in Hybrid Wire-Arc DED.

Automated manufacturing feature recognition is a crucial link between computer-aided design and manufacturing, facilitating process selection and other downstream tasks in computer-aided process planning. While various methods such as graph-based, rule-based, and neural networks have been proposed for automatic feature recognition, they suffer from poor scalability or computational inefficiency. Recently, voxel-based convolutional neural networks have shown promise in solving these challenges but incur a tradeoff between computational cost and feature resolution. This paper investigates a computationally efficient sparse voxel-based convolutional neural network for manufacturing feature recognition, specifically, an octree-based sparse voxel convolutional neural network. This model is trained on a large-scale manufacturing feature dataset, and its performance is compared to a voxel-based feature recognition model (FeatureNet). The results indicate that the octree-based model yields higher feature recognition accuracy (99.5% on the test dataset) with 44% lower GPU memory consumption than a voxel-based model of comparable resolution. In addition, increasing the resolution of the octree-based model enables recognition of finer manufacturing features. These results indicate that a sparse voxel-based convolutional neural network is a computationally efficient deep learning model for manufacturing feature recognition to enable process planning automation. Moreover, the sparse voxel-based neural network demonstrated comparable performance to a boundary representation-based feature recognition neural network, achieving similar accuracy in single feature recognition without having access to the exact 3D shape descriptors

Wire-Arc Directed Energy Deposition (Wire-Arc DED) has emerged as a promising additive manufacturing technique known for its high deposition rates. However, the variability in microstructure and mechanical properties (e.g., hardness) of the manufactured components poses significant challenges. This study delves into these issues, focusing on the influence of interlayer machining on the microstructural evolution and mechanical properties of thin-wall Wire-Arc DED structures. It is shown that as-built Wire-Arc DED structures exhibit a pronounced microstructure variation between different regions along the build direction, primarily governed by the differences in thermal history. In contrast, a Hybrid Wire-Arc DED process that integrates interlayer machining into the build process to induce severe plastic deformation leads to a microstructure characterized by refinement and homogenization, compared to a Wire-Arc DED process. This study provides insights into the impacts of plastic deformation due to machining and thermal cycling due to subsequent layer depositions on the microstructure and hardness obtained in Wire-Arc DED and Hybrid Wire-Arc DED processes, highlighting the potential of hybrid manufacturing to generate tailored microstructures to enhance the mechanical performance of functional components.

Electric motors transform the electrochemical energy into mechanical energy needed for the vehicles’ motion. Given that the transition from internal combustion engines to electric vehicles is likely to be the main driver of sustainable transportation, high‐performing electric motors will be increasingly important for this transition. Herein, the effect of the grinding process on the surface integrity and magnetic properties of NdFeB permanent magnets are investigated. NdFeB permanent magnets are most commonly used in the fabrication of permanent magnet electric motors and they are usually ground during their fabrication process chain in order to create a high‐quality cylindrical rotor for subsequent mechanical assembly. In particular, the effects of wet and dry grinding processes on magnetic properties and surface integrity are studied and compared. Guided by a thermal model, the grinding process parameters are varied highlighting their effects on the underlying thermomechanical phenomena responsible for the changes in surface integrity and magnetic properties observed. Wet grinding proves to be a viable process for NdFeB magnets. Additionally, acceptable results are achieved with dry grinding, which is particularly appealing due to its sustainability. It is noteworthy that the range of process parameters yielding acceptable results is narrower.