This research presents a comprehensive analysis of deep neural network models (DNNs) for the precise prediction of Vickers hardness (HV) in nitrided and carburized M50NiL steel samples, with hardness values spanning from 400 to 1000 HV. By conducting rigorous experimentation and obtaining corresponding nanoindentation data, we evaluated the performance of four distinct neural network architectures: Multilayer Perceptron (MLP), Convolutional Neural Network (CNN), Long Short-Term Memory network (LSTM), and Transformer. Our findings reveal that MLP and LSTM models excel in predictive accuracy and efficiency, with MLP showing exceptional iteration efficiency and predictive precision. The study validates models for broad application in various steel types and confirms nanoindentation as an effective direct measure for HV hardness in thin films and gradient-variable regions. This work contributes a validated and versatile approach to the hardness assessment of thin-film materials and those with intricate microstructures, enhancing material characterization and potential application in advanced material engineering.

Recrystallization in the carburized case of M50NiL steel was achieved without external deformation through optimized carburizing heat treatment, utilizing the particle-stimulated nucleation (PSN) mechanism of the large size M7C3 carbides. This process refined the grain structure to the sub-micron scale. A cellular automaton (CA) model was developed to predict recrystallization during carburizing and quenching, incorporating PSN and thermal deformation effects to quantify the recrystallized fraction in the carburized layer. The influence of thermal deformation on the microstructure of the sub-micron crystalline region (SMCR) was systematically investigated. Focused ion beam-transmission electron microscopy (FIB-TEM) revealed that the SMCR experienced significantly less deformation than the non-sub-micron crystalline region (NSMCR). Thermal deformation also promoted austenite nucleation at carbide interfaces, establishing a cube-on-cube orientation relationship between carbides and adjacent austenite. This semi-coherent interface reduced interfacial energy, thereby suppressing the stress-induced austenite-to-martensite transformation during stress relaxation.

To mitigate the severe corrosion experienced by M50 bearing steel during service—particularly the accelerated degradation induced by chloride ions—Cr/N multilayer composite films were fabricated on its surface using plasma-based ion implantation and deposition (PBIID). By coupling magnetron sputtering with high-energy nitrogen ion implantation, the films developed a mixed microstructure consisting of nanocrystalline CrN, Cr, and Cr2N phases together with partially amorphous regions. XRD, TEM, electrochemical measurements and surface analytical techniques were employed to systematically evaluate how the multilayer modulation structure and implantation-induced amorphization influence the corrosion resistance of the films. The results show that increasing the ion implantation dose reduces the Cr₂N fraction while promoting the appearance of CrN(220) and Cr(110) texture orientations. The optimal specimen exhibits an ultralow corrosion current density of 3.93 × 10-8 A·cm-2 and demonstrates the most stable passivation behavior in 3.5 wt.% NaCl solution. Micro-scale analysis reveals that corrosion preferentially initiates at the martensite–carbide interfaces, where localized electric-field intensification enhances the migration rate and local enrichment of Cl⁻, thereby accelerating subsurface dissolution. During corrosion, negatively charged Cr₂O₃ nanoparticles undergo electrophoretic migration and subsequent redeposition driven by the local electric field, forming a secondary passivation layer on the martensitic side and effectively suppressing further corrosion propagation.