Surface Properties Enhancement of SLMed AlSi10Mg via Ultrasonic Surface Rolling: Experiments and Data-Driven Optimization

Selective laser melting (SLM) shows great potential for fabricating lightweight and complex components in aerospace and biomedical industries. However, the process-induced defects, such as high surface roughness and porosity, severely deteriorate the surface integrity and mechanical properties of SLMed parts, limiting their application in critical scenarios. To mitigate these issues, the Ultrasonic surface rolling process (USRP) was applied to SLMed AlSi10Mg as a post-processing method. The effects of key USRP parameters, including static load, feed rate and number of passes, on surface roughness, microhardness and microstructural evolution were systematically investigated. Results indicate that a combination of low feed rate, moderate static load and an appropriate number of passes reduces the surface roughness by approximately 56.8% (achieving a minimum Ra of 1.08[Formula: see text][Formula: see text]m), increases the surface hardness by about 22.4% (reaching a peak value of 165.2 HV), and facilitates grain refinement through severe plastic deformation. Furthermore, this study integrates machine learning with NSGA-III for multi-objective optimization of surface properties, yielding an optimal parameter set: [Formula: see text][Formula: see text]mm/s, [Formula: see text][Formula: see text]N and [Formula: see text], which achieves a balance between low roughness and high hardness. This work not only elucidates the strengthening mechanisms of USRP on SLMed AlSi10Mg, but also establishes a robust data-driven framework for optimizing surface enhancement processes in additive manufacturing.