CFD‐based analysis of entropy generation due to mean and fluctuating fields during turbulent natural convection in a square enclosure

Abstract

The current study aims to numerically investigate the entropy generation during the natural convection flow of air in a square cavity. The governing equations for the conservation of mass, momentum, energy, and turbulence are solved using a control volume‐based technique employing the commercial code Fluent. Runs have been performed for both laminar and turbulent flow regimes by varying the Rayleigh number (Ra) from 103 to 1010. On the other hand, various viscous distribution coefficients (ϕ = 10−4, 10−3, and 10−2) and constant Prandtl number (Pr = 0.71) were considered. Given the conflicting perspectives in the literature regarding the entropy generation under turbulent regimes, more research is needed to better understand the impact that the fluctuating flow has on entropy production. The four terms of entropy generation inherent to turbulent natural convection (entropy generation due to dissipation in the mean and the fluctuating velocity fields in addition to the heat flux due to the mean and the fluctuating temperature) are computed in the present work and compared to calculations based on only mean values of temperature and velocity gradients. It was found that taking into account the fluctuating terms of temperatures and velocities augment the total entropy generation by 10.10%, 14.43%, and, 17.70%, up to 32.60%, respectively, for Ra = 5 × 108, Ra = 109, Ra = 1.58 × 109, and Ra = 1010. The gain shows the tendency to increase with the Rayleigh number. Thus, the fluctuating terms cannot be neglected particularly for high Rayleigh numbers. Furthermore, unlike entropy production due to the mean flow field, numerical outcomes reveal that the generated irreversibilities due to fluctuating flow are located around the upper hot and the lower cold corners of the heated walls. In addition, a numerical relationship between the first and the second laws of thermodynamics has been derived. A promising result that emerged from this study has shown that the Nusselt number and therefore the first law of thermodynamics is sufficient to estimate the heat part of entropy generation without the necessity of using the second law.

Predicting heat transfer rate and system entropy based on combining artificial neural network with numerical simulation

“…They summarized that two main excellent controlling parameters for entropy of the system are the Darcy parameter, corresponding to porosity of the medium, as well as power-law index. In the same context, the following references give a detail investigation for the entropy generation in nanofluid (Oliveski et al, 2009;Tasnim et al, 2023;Fattahi et al, 2023;Said et al, 2023;Murali et al, 2023;Alqaed et al, 2023;Ahmed et al, 2022). The integration of artificial neural network (ANN) in the context of heat transfer involves using these computational models to analyze, predict and optimize the transfer of thermal energy within a given system.…”