Computation time reduction of PCM melting process by changing modeling parameters

Abstract

This study can be considered as a helpful reference for whom endeavor to boost the computation efficiency of the PCM melting process. Researchers sacrifice accuracy to decrease computation time since computational fluid dynamics (CFD) solutions of PCM melting processes require comparatively very long time, i.e., from hours to days or weeks, depending on the system geometry. The present study compares the approaches recommended in the literature in terms of their influence on computation time reduction and accuracy. A horizontally finned tube LHTES unit is modeled in 2-D space using ANSYS Fluent, the most common commercial CFD software for the considered problem in the literature. The outcomes obtained from the attempts to boost the computation efficiency are as follows: adaptive time step size approach causes 72% enhancement in computation time (from 90 hours to 25 hours), frozen flux algorithm and constant thermophysical properties have almost no influence on computation time. Even though low convergence criteria and neglecting natural convection reduces computation time drastically, the errors in accuracy are not in acceptable level.