Machine Learning-Based Multi-Objective Optimization for Enhancing the Performance of Block Support Structures for Electron Beam Additive Manufacturing

Electron beam melting (EBM) technology has gained prominence owing to its ability to enhance production efficiency and meet green manufacturing standards. However, overhang structures are a significant issue for additive manufacturing due to their need for supporting structures during printing. This increases manufacturing time, requiring more material, extra effort, and a more complex engineering procedure. Therefore, this research aims to develop an intelligent optimization method based on AI-ANFIS/Al-ANN and improved NSGA-III, integrating the AM design, 3D printing, and post-processing phases to enhance the performance of block support structures and the quality of the EBM parts produced. To achieve this, statistical analysis was performed to detail the simultaneous influence of block support type, block support structure design, and EBM parameters on fabricating performance, warping deformation, support removal time, and support volume. After that, intelligent models based on ANFIS/ANN and the advanced NSGA-III method were developed for monitoring and optimizing the performance of specified block support structures. The results reveal that the block support type, block support structure design, and EBM parameters simultaneously significantly affect block support structures’ performance. This study illustrated that the AI models based on ANFIS might provide more accurate and reliable estimation models for monitoring and predicting support volume, support removal time, and warping deformation, exhibiting reduced errors of 0.992%, 1.2%, 1.28%, and 1.06%, respectively, in comparison to empirical measurements, ANN models, and regression models. Finally, the developed intelligent method obtains the optimal block support type, block support design, and EBM parameters to enhance the quality of produced parts, reduce material wastage, and reduce the post-processing time of fabricated EBM Ti6Al4V. Henceforth, smart systems may be employed to create innovative solutions that integrate the AM design, 3D printing, and post-processing stages. This will allow for the monitoring and improvement of AM process performance, as well as the fulfillment of Industry 4.0 requirements.