Multi-response Optimization of Thermal Cycling Process for Al6092/SiC/ZrW2O8 Composites Using RSM-MOGA

• Published in: Arabian journal for science and engineering (2011) 2022, Vol.47 (6), p.7669-7682
• Main Authors: Zhang, Shihao, Hou, Qinglin, Fu, Zhixiang, Jiang, Haiyun
• Format: Article
• Language: English
• Subjects: Aluminum; Aluminum base alloys; Aluminum matrix composites; Dimensional stability; Engineering; Genetic algorithms; Heat conductivity; Heat transfer; Heating; Humanities and Social Sciences; Hyperspaces; multidisciplinary; Multiple objective analysis; Optimization; Particulate composites; Phase transitions; Research Article-Physics; Residual stress; Response surface methodology; Science; Silicon carbide; Sorting algorithms; Strain; Thermal conductivity; Thermal cycling; Thermal expansion; Thermal mismatch; Thermal stress; Thermophysical properties; Variance analysis; Zirconium carbide
• ISSN: 2193-567X; 1319-8025; 2191-4281
• DOI: 10.1007/s13369-021-06536-0

Thermal cycling can reduce or eliminate thermal stresses in aluminum matrix composites, which is essential to improve the dimensional stability of the composites. In this work, the effect of thermal cycling process on the thermophysical properties of 6092 aluminum alloy/silicon carbide/zirconium tungstate composites was investigated by combining response surface optimization and multi-objective genetic algorithm. The purpose is to elucidate the effects of the process of heating temperature, heating time and cycles on the coefficient of thermal expansion (CTE), micro-strain and thermal conductivity of the composites. A series of robust response models were proposed and verified using analysis of variance, externally studentized residuals distribution and the predicted residuals distribution. According to these models, the heating temperature, heating time and cycles interact in a nonlinear manner concerning on thermophysical properties. The experimental results confirmed that thermal cycling can effectively reduce or eliminate thermal residual stresses in the composites and induce the phase transformation process of zirconium tungstate. The heating temperature plays a dominant role in promoting the thermal mismatch stress release between the metal matrix and the reinforcing particles. The Pareto-optimal hypersurface results on the response surface were obtained according to the non-dominated sorting genetic algorithms (NSGA-II), indicating that the optimal thermal cycling processes were 200 °C, 2 h and 3 cycles with corresponding CTE, micro-strain and thermal conductivity of 20.55 × 10 –6 /K, 1.16 × 10 –3 and 116.68 W/(m K), respectively.