The consumption of NdFeB sintered permanent magnets has been increasing drastically with the increase in the number of electric vehicles (EV) and hybrid electric vehicles (HEV). The addition of rare earth (RE) metals to magnets improves its properties and has a direct impact on its costs. Therefore, there is a high demand for permanent magnets with less RE content or inexpensive RE metals.
Therefore, search for novel magnetic phases requires efficient quantitative microstructure analysis to extract microstructural information and correlate it with its intrinsic magnetic parameters. This helps in obtaining the optimized microstructures in magnets with good intrinsic magnetic properties. Hence, quantitative microstructure analysis and controlling process parameters are vital for new material development and improving quality.
In this paper, we use the classical machine learning approach and advanced deep learning algorithm for the extraction of microstructural information such as grain size distribution and micro magnetic domain patterns from the NdFeB sintered permanent magnets from Kerr microscopy images. Due to the complex microstructural features, it is not a feasible option to use traditional approaches of image analysis for extracting quantitative information from these micrographs. The performance of the trained models is compared to EBSD data and a manually hand-labelled dataset prepared by a subject expert. The model has proved to be accurate and robust to different magnet samples with accuracy in range of 91-93% for different magnet alloys and is nearly 20 times faster than existing approaches.