Numerical parametric optimization of fluid flow profiles in membranes using the Taguchi method

Membrane modeling and flow through it are important studies, and their applications are for gas and water separation. The examples of membrane modeling applications are artificial kidneys and reactors. Some studies are available on membrane modeling for different applications; however, optimization studies are rarely available. In the current study, an initial membrane model is simulated by COMSOL 6.3 and then optimized for input parameters to decide the optimal set at which the performance and flow profiles are better. For this study, the design of experiments table through Minitab software using the Taguchi method is generated. Then it tested for an optimal parametric selection. In this study, the focus is on the effect of dialysate concentration (D_c), partition coefficient (K), dialysate velocity (D_v), and permeate velocity (P_v) on concentration individually and then in combination. The individual effect of parameters with different values is first studied. Then an array of L₁₆ (4×4) using the design of experiments method (DOE) concept is finalized for variance analysis (ANOVA) to study the combined effect. The result of this analysis concludes that the major impact on concentration and velocity profile is due to K. The combined effect results conclude that the effect of different parameters is in the order of K > P_v > C_d > D_v for concentration and velocity profiles. This means that the major effect is due to the partition coefficient, and the minor effect is due to the permeating velocity. Before testing new results, the current COMSOL model is validated against published experimental work to ensure its accuracy.