Optimization of pore window size of replicated open-cell porous aluminium for heat transfer application

Pore window optimization has not been thoroughly investigated as a means of improving the potentials of replicated porous aluminium in thermal management systems. The study focuses on the optimization of the factors affecting the pore window diameter of porous aluminium structures fabricated by replication casting technique. The pore window diameter of replicated porous aluminium was optimized by controlling the following parameters: infiltration pressure; infiltration temperature; perform particle size and perform packing density. Each is evaluated at four levels. Using a 4 × 4 matrix in the experimental design, an L16 orthogonal array was employed to generate sixteen sets of parameter combinations. Porous aluminium samples were fabricated based on these combinations, microstructurally analyzed using a JEOL 6400 scanning electron microscope and scanco micro computed tomography scanner while measurement of pore window diameter was performed using a Nikon optical microscope which is equipped with a CMEX digital camera for the capturing of images. Statistical analysis, including Taguchi method and analysis of variance (ANOVA) was conducted using Minitab 19.1 software and employed to identify the most influential factors. The Taguchi main effect plot identified the optimum parameter levels for maximizing pore window diameter. The results indicated that infiltration pressure has the most significant impact on window diameter, followed by preform packing density. The ANOVA analysis shows that infiltration pressure is the only input parameter that has significant effect on window size. Taguchi analysis further showed that the optimum parameter for effective window diameter can be obtained at infiltration pressure of 10 kPa, packing density of 0.58, preform particle size of 2.25 mm and infiltration temperature of 700 °C. Confirmation tests on the optimal parameters reveal improved thermal response for porous sample fabricated at the chosen parameters when compared with samples produced using combination of other input parameters.