Structure Optimization of Longitudinal Rectangular Fins to Improve the Melting Performance of Phase Change Materials through Genetic Algorithm

In this paper, the structural parameters of longitudinal rectangular fins used in a horizontal shell-and-tube latent heat storage unit (LHSU) are optimized to increase the melting rate of phase-change materials. The influence of natural convection on the melting process is considered. Due to the extremely nonlinear and expensive computational cost of the phase-change heat-transfer-optimization problem, a new coupling algorithm between genetic algorithm and computational fluid dynamics is developed. The effects of the thermal conductivity of fins; the filling rate of fins; and the number of fins on the optimal structure parameters, including the length, width, and position of each fin, are discussed. The results show that when a single fin is inserted in the half-ring region, the optimal dimensionless fin angle is about 0.2, and the optimal dimensionless fin length is about 0.96. The use of optimal single fin can shorten the dimensionless total melting time by 68% compared with the case of no fin, and 61.3% compared with uniformly arranged single fin. When the number of fins exceeds one, each fin should have a specific length ( L ), thickness ( ∆ ), and position ( ψ ) instead of uniform distribution. The advantage of the optimized fins decreases as the number of fins increases. When the number of fins is four, the optimized fin distribution is almost uniform, and the dimensionless total melting time is only 15.9% less than that of the absolutely uniform fin. The number of fins is a more sensitive parameter affecting the optimal position and structure of fins than the filling rate and thermal conductivity of fins.