Data-Driven Optimization of Surface Roughness in Post-Processing of SLM-produced Ti-۶Al-۴V Parts using Python-based Machine Learning Models

Subject: Data-Driven Optimization of Surface Roughness in Post-Processing of SLM-produced Ti-۶Al-۴V Parts using Python-based Machine Learning Models. Selective Laser Melting (SLM) has emerged as a transformative additive manufacturing technology for aerospace and medical applications. However, the high surface roughness and hard martensitic (a') microstructure of as-built Ti-۶Al-۴V components necessitate precision post-processing through machining. Traditional machining parameters often fail to account for the unique material properties and anisotropy inherent in ۳D-printed parts. This research proposes a data-driven framework for the optimization of surface roughness (Ra) using Python-based machine learning models and Genetic Algorithms (GA). Experimental trials were conducted on SLM-produced Ti-۶Al-۴V samples at varying build orientations (۰°, ۴۵°, and ۹۰°). Machine learning models, including Support Vector Regression (SVR) and Random Forest (RF), were developed to predict Ra based on cutting speed, feed rate, depth of cut, and build orientation. The Random Forest model demonstrated the highest predictive accuracy, achieving an R² score of ۰.۹۷ and a Mean Absolute Error (MAE) of ۰.۰۲۸ μm. Sensitivity analysis revealed that build orientation significantly impacts machinability, contributing ۱۲% to the variance in surface quality due to varied micro-hardness levels. Finally, a Genetic Algorithm (GA) was employed to identify the global optimal cutting parameters. The optimized conditions (V = ۱۰۵ m/min, f = ۰.۰۶mm/tooth) resulted in a surface roughness of ۰.۳۴ μm, representing a ۶۳% improvement over standard industrial machining practices. The results validate that integrating additive manufacturing history with ensemble learning provides a robust solution for enhancing the surface integrity of hybrid components, ensuring they meet stringent aerospace requirements.

کلیدواژه ها:

Selective Laser Melting (SLM) ، Ti-۶Al-۴V ، Surface Roughness ، Machine Learning ، Random Forest ، Genetic Algorithm ، Hybrid Manufacturing