Optimisation of the design parameters of a domestic refrigerator using CFD and artificial neural networks

• The parameter's values of the domestic refrigerator are optimised by CFD and ANN.
• The numerical results were validated by comparing with the experimental results.
• The performance of the refrigerator was improved by approximately 7.7%.
• The annual energy consumption will be decreased to 17.52 kWh.

This paper reports the use of computational fluid dynamics (CFD) and artificial neural networks (ANN) for optimising and improving the performance of a static type domestic refrigerator. The effects of the hydrodynamic and thermal fields on the air flow inside of the refrigerator compartment were considered by exact modelling of the fan. The interior volume of air of the refrigerator was modelled using a CFD method, and analyses were performed. The numerical results were validated by comparing experimental results with the calculated interior design parameters. An optimisation study of the obtained design parameters was performed using a parametric method. The optimum design case was predicted by the ANN. The performance of the refrigerator improved by approximately 7.7% based on the amount of heat taken from the evaporator surface assuming thermal uniformity of the interior volume of air and the ISO 15502 standards. As a result, the annual energy consumption will decrease by 17.52 kWh.