Multi-objective optimization design of the heating/cooling channels of the steam-heating rapid thermal response mold using particle swarm optimization


Abstract

The layout of the heating/cooling channels is of great significance for rapid heat cycle molding (RHCM) mold with steam heating and coolant cooling because it not only affects heating/cooling efficiency, temperature uniformity of mold cavity surface, but also has a great influence on mold strengt. Thermal, structural and fatigue analysis based on finite element method (FEM) is performed to investigate the effect of the heating/cooling channels layout on thermal response, structural strength and fatigue life of the RHCM mold. In order to obtain a reasonable layout of heating/cooling channels, this study focuses on the development of an effective methodology for the layout optimization of the heating/cooling channels by integrating response surface methodology (RSM) and multi-objective particle swarm optimization (MOPSO) algorithm. Three design variables describing the layout and scale of the heating channels are selected to do the design of Box-Behnken experiment with three factors and three levels. Thermal/structural analysis is carried out using ANSYS to obtain the corresponding values of three objective variables, including required heating time, maximum cavity surface temperature difference and maximum von-Mises stress due to thermal expansion, which are used to describe heating efficiency, temperature uniformity and mold strength, respectively, for different sets of design variables. RSM is utilized to analyze the effect of the design parameters and further construct mathematical models to quantitatively describe the relationship between design variables and objective variables via regression analysis. Analysis of variance (ANOVA) demonstrates that the developed quadratic models are highly effective and significant. Confirmation experiment is also conducted to verify the effectiveness and accuracy of the developed quadratic polynomial models. Based on these mathematical models, a MOPSO algorithm is then introduced to optimize the design variables by comprehensively considering thermal efficiency, temperature uniformity and structural strength of the RHCM mold. The optimum results show that thermal efficiency and temperature uniformity of the RHCM mold can be greatly improved with the optimum design variables for the layout of the heating/cooling channels. The following verification experiment demonstrates the validity of the optimum results.