Abstract
Micro-milling is increasingly recognized as a crucial technique for machining intricate and miniature 3D aerospace components, particularly those fabricated from difficult-to-cut Ti-6Al-4V alloys. However, its practical applications are hindered by significant challenges, particularly the unavoidable generation of burrs, which complicate subsequent finishing processes and adversely affect overall part quality. To optimize the burr formation in the micro-milling of Ti-6Al-4V alloys, this study proposes a novel hybrid-ranking optimization algorithm that integrates Grey Relational Analysis (GRA) with the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). This approach innovatively combines GRA and TOPSIS with a random forest regression (RFR) model, facilitating the exploration of nonlinear and complex relationships between input parameters and machining outcomes. Specifically, the effects of spindle speed, depth of cut, and feed rate per tooth on surface roughness and burr width generated during both down-milling and up-milling processes were systematically investigated using the proposed methodology. The results reveal that the depth of cut is the most influential factor affecting surface roughness, while feed rate per tooth plays a critical role in controlling burr formation. Moreover, the GRA-TOPSIS-RFR method significantly outperforms existing optimization and prediction models, with the integration of the RFR model enhancing prediction accuracy by 42.6% compared to traditional linear regression approaches. The validation experimental results agree well with the GRA-TOPSIS-RFR-optimized outcomes. This research provides valuable insights into optimizing the micro-milling process of titanium components, ultimately contributing to improved quality, performance, and service life across various aerospace applications.