Thermal performance optimization of a sinusoidal wavy channel with different phase shifts using artificial bee colony algorithm

Abstract

In this article, heat transfer optimization of forced convection in a wavy channel with different phase shifts between the upper and lower wavy walls is represented. The flow is laminar in the range of (200 ≤ Re ≤ 800) and a uniform temperature of 350 K is considered in the wavy sections. The governing mass, momentum and energy equations are solved using the finite volume method. Different design parameters such as the channel height (h = 10, 15 and 20 mm), the amplitude of the wavy wall (A = 1.5, 2, 2.5 and 3 mm) and the phase shift of the upper wavy wall $$ (0^\circ \le \gamma \le 360^\circ ) $$ are investigated. For optimization process, a recent method, named artificial bee colony (ABC) algorithm, is applied and compared with two other meta-heuristic algorithms, called particle swarm optimization (PSO) and differential evolution (DE). An “in house” code is developed which simultaneously uses the meta-heuristic algorithms and the computational fluid dynamics solver. The results indicate that ABC algorithm has higher accuracy and faster convergence rate than PSO and DE. The parameter considered for optimizing the average Nusselt number as the objective function was the phase shift. But, for optimizing the thermal performance factor, selected parameters were the wavy wall amplitude and the phase shift. The results showed that the maximum average Nusselt number is attained at $$ \gamma = 250.2 $$ , A = 3 mm, h = 10 mm and Re = 800, in which the heat transfer rate has 96.6% enhancement rather than the parallel-plate channel. Also, it is found that $$ \gamma = 283.3 $$ , A = 2.65 mm, h = 10 mm and Re = 800 are the optimized solutions to obtain the maximum thermal performance factor.