Prediction of surface structure and optimization of corrosion resistance in magnesium alloys via machine learning for nanosecond laser processing

Magnesium alloys, valued for their low density and high specific strength, are widely used in aerospace, automotive, and biomedical industries. However, their poor corrosion resistance significantly limits applications. This study proposes a machine learning-guided approach to optimize laser-induced micro-groove structures on AZ31B magnesium alloy to enhance corrosion resistance. Three regression models—Support Vector Regression, Random Forest, and Light Gradient Boosting Machine—were developed and integrated into an ensemble learning framework to predict the groove aspect ratio (depth-to-width). The model demonstrated superior accuracy, achieving the lowest RMSE and highest R² among all models. To identify optimal laser processing parameters, the Non-dominated Sorting Genetic Algorithm II (NSGA-II) was employed, successfully maximizing the aspect ratio. Experimental validation involved fabricating optimized grid microstructures, followed by stearic acid surface modification. The treated surface transitioned from superhydrophilic to superhydrophobic, achieving a water contact angle of 166°. SEM, EDS and XPS analyses confirmed that the hierarchical micro/nano structure, combined with the low-surface-energy coating, created an effective air cushion layer that minimized interaction with corrosive media, significantly enhancing corrosion resistance. This work demonstrates the potential of machine learning for precise control and optimization of laser-induced surface structures on magnesium alloys, offering a robust and scalable strategy for functional surface engineering.