Abstract
Purpose This study aims to investigate a three-dimensional computational modelling of free convection of Al2O3 water-based nanofluid in a cylindrical cavity under heterogeneous heat fluxes that can be used as a thermal storage tank. Design/methodology/approach Effects of different heat flux boundary conditions on heat transfer and entropy generation were examined and the optimal configuration was identified. The simulation results for nanoparticle (NP) volume fractions up to 4 per cent, and Rayleigh numbers of 10(4), 10(5) and 10(6) were presented. Findings The results showed that for low Ra (10(4)) the heat transfer and entropy generation patterns were symmetric, whereas with increasing the Rayleigh number these patterns became asymmetric and more complex. Therefore, despite the symmetric boundary conditions imposed on the periphery of the enclosure (uniform in & x19f;), it was necessary to simulate the problem as three-dimensional instead of two-dimensional. The simulation results showed that by selecting the optimal values of heat flux distribution and NP volume fraction for these systems the energy consumption can be reduced, and consequently, the energy efficiency can be ameliorated. Originality/value The results of the present study can be used for the design of energy devices such as thermal storage tanks, as both first and second laws of thermodynamics have been considered. Using the optimal design will reduce energy consumption.