The effect of geometric parameters of the zigzag, rectangular, and serpentine channels on convective heat transfer coefficient and pressure drop was investigated using computational fluid dynamics (CFD). In all channels, the same boundary conditions were considered, and the number of steps was equal to 10. The simulations were performed for turbulent flows (liquid water as the operating fluid), and Reynolds number (Re) range between 20000 and 60000 was selected. The zigzag channel showed a best thermal performance and the serpentine channel showed the best hydraulic performance. The thermal-hydraulic performance (THP) factor was employed for comparing the channels. As the complexity of the channels surfaces increased, the two parameters of convective heat transfer coefficient (positive factor) and pressure drop (negative factor) increased simultaneously. Therefore, predictive correlations for friction factor and Nusselt number were presented using genetic algorithm (GA), and the multi-objective optimization was performed to obtain the most appropriate Nusselt number and minimum friction factor as the two basic objective functions. The resulting Pareto set, which includes the optimum geometric dimensions of the heat exchangers, allows a designer to choice the geometries based on higher heat transfer or lower pumping power.
