Modeling and optimizing of thermal conductivity of composite phase change materials based on multivariate statistical analysis

Low thermal conductivity is a common drawback for phase change materials (PCMs). The addition of nanomaterials, expanded graphite, or embedded foam metals can improve the thermal conductivity but this also leads to an increase in cost or energy consumption, which limits their application. In this study, a low-cost PCM was developed using sodium acetate trihydrate (SAT) as the base material, and graphite powder (GP) as the thermal conductivity enhancer. The effects of additives and their interactions on the thermal properties of PCMs were analyzed, and the mathematical model of thermal conductivity was developed based on multivariate statistical analysis. The optimal solution of the thermal conductivity model is GP size: 80 mu m, GP potency: 10 wt%, xanthan gum (XG): 0.75 wt%, diatomaceous earth (DE): 7 wt%, dodecahydrate sodium hydrogen phosphate (DHPD): 2.0 wt%, and the optimum value of the thermal conductivity model is 1.37 W/(m<middle dot>K), which is increased to 228 %. From the orthogonal experiments, at low GP concentrations and low mesh number, the thermal conductivity increased by 11.1 % and 5.0 % for each 1 wt% increase in GP and DHPD concentrations, respectively; however, the thermal conductivity decreased by 11.3 % and 1.0 % for each 1 wt% increase in XG and DE concentrations, respectively. At high GP concentrations and high mesh number, the thermal conductivity increased by 3.8 %, 4.3 %, and 1.3 % for each 1 wt% increase in GP, DHPD, and DE concentrations, respectively; the thermal conductivity decreased by 8.8 % for each 1 wt% increase in XG concentration. The modeling method is a new and universally applicable method for fabricating and optimizing PCMs.