We present an electron backscattered diffraction (EBSD)-trained deep learning (DL) method integrating traditional material characterization informatics and artificial intelligence for a more accurate classification and quantification of complex microstructures using only regular scanning electron microscope (SEM) images. In this method, EBSD analysis is applied to produce accurate ground truth data for guiding the DL model training. An U-Net architecture is used to establish the correlation between SEM input images and EBSD ground truth data using only small experimental datasets. The proposed method is successfully applied to two engineering steels with complex microstructures, i.e., a dual-phase (DP) steel and a quenching and partitioning (Q&P) steel, to segment different phases and quantify phase content and grain size. Alternatively, once properly trained the method can also produce quasi-EBSD maps by inputting regular SEM images. The good generality of the trained models is demonstrated by using DP and Q&P steels not associated with the model training. Finally, the method is applied to SEM images with various states, i.e., different imaging modes, image qualities and magnifications, demonstrating its good robustness and strong application ability. Furthermore, the visualization of feature maps during the segmenting process is utilised to explain the mechanism of this method's good performance.

With the development of the materials genome philosophy and data mining methodologies, machine learning (ML) has been widely applied for discovering new materials in various systems including high-end steels with improved performance. Although recently, some attempts have been made to incorporate physical features in the ML process, its effects have not been demonstrated and systematically analysed nor experimentally validated with prototype alloys. To address this issue, a physical metallurgy (PM) -guided ML model was developed, wherein intermediate parameters were generated based on original inputs and PM principles, e.g., equilibrium volume fraction (V_f) and driving force (D_f) for precipitation, and these were added to the original dataset vectors as extra dimensions to participate in and guide the ML process. As a result, the ML process becomes more robust when dealing with small datasets by improving the data quality and enriching data information. Therefore, a new material design method is proposed combining PM-guided ML regression, ML classifier and a genetic algorithm (GA). The model was successfully applied to the design of advanced ultrahigh-strength stainless steels using only a small database extracted from the literature. The proposed prototype alloy with a leaner chemistry but better mechanical properties has been produced experimentally and an excellent agreement was obtained for the predicted optimal parameter settings and the final properties. In addition, the present work also clearly demonstrated that implementation of PM parameters can improve the design accuracy and efficiency by eliminating intermediate solutions not obeying PM principles in the ML process. Furthermore, various important factors influencing the generalizability of the ML model are discussed in detail.