Research on Intelligent Optimization of Shrinkage Porosity and Cavity Defects in Solid–Liquid Composite Pipe Billets
摘要
Bimetallic composite pipes have good corrosion resistance, high strength, and fine mechanical properties. The quality of composite pipe billet preparation directly affects the formability of the finished pipes. In this work, we focused on shrinkage cavity and porosity defects, using a hybrid intelligent algorithm combining Gaussian process regression and random forest to optimize and analyze four key process parameters: pouring temperature, outer mold preheating temperature, pouring time, and outer mold rotation speed. We obtained the optimal parameter combination and also designed a heat preservation device to ensure the performance of 20# steel-316L stainless steel composite pipe billets. After on-site trial production and microstructural analysis, the optimized composite pipe billets showed no internal shrinkage porosity or cavities. The thickness of the bonding layer increased from the original 100μm to 140μm, which improved the billet quality. The results of this study offer an efficient and precise preparation scheme for high-quality composite pipe billets.