SBIR/STTR Award attributes
Additive manufacturing technology is becoming more popular for the fabrication of 3D metal products as it offers rapid prototyping and large design freedom. However, with more complex geometric features due to topology optimization, it becomes infeasible to carry out traditional surface machining process to improve surface roughness for fatigue performance. In this effort, we will develop a novel electropolishing process which is capable of accessing all open surfaces for complex part and critically examine its improvement on surface roughness and fatigue performance by comparing with traditional mechanical mass finishing. In order to simulate and optimize the electropolishing process, Integrated Computational Materials Engineering and machine learning approach will be utilized to link the process parameters with surface properties and fatigue performance. An artificial neural network will be implemented to relate electropolishing process parameters with polished surface roughness, while the Integrated Computational Materials Engineering toolset provides mechanistic modeling of surface roughness effects on fatigue life. Furthermore, a state-of-the-art hybrid optimization approach, combining sensitivity analysis, response surface method and genetic algorithm, is proposed to optimize process parameters to minimize surface roughness and maximize fatigue performance.
