Multi-objective optimization of cooling of a stack of vertical minichannels and conventional channels subjected to natural convection

Abstract

In this paper, by employing the Computational Fluid Dynamics (CFD) and applying the NSGA II algorithm, cooling of a stack of vertical parallel plates subjected to natural convective heat transfer is multi-objectively optimized. In the optimization process, the plates spacing is changed so as to simultaneously maximize the flow rate of the fluid sucked through the stack of plates (cooling flow rate) and the amount of heat transfer between the plates and ambient air. The plates spacing varies from 200 µm to 100 mm; and thus, the investigated channels can be classified between minichannels and conventional channels. In Section 3, the Pareto front, which simultaneously displays the variations of the cooling suction flow rate and the heat transfer from plates, will be presented, and it will be demonstrated that the Pareto front conveys very important results for the thermal designing of a stack of parallel channels subjected to natural convective heat transfer. At the end, the multi-objective optimization results obtained in this paper will be compared with the related results of the asymptotic method and the analytical approach, which are single-objective optimization methods and aimed at increasing the amount of heat transfer from the plates; and very useful and valuable conclusions will be obtained.