CFD-based shape optimisation of a CO₂ two-phase ejector mixing section

In this study, the geometry of a CO₂ ejector mixing section was optimised. Two optimisation algorithms, a genetic algorithm (GA) and an evolutionary algorithm (EA), were used with a validated CFD model based on the homogeneous equilibrium model (HEM). The ejector was designed to increase the Coefficient of Performance (COP) of CO₂ refrigeration systems used in supermarkets. Hence, the objective function (OF) for the optimisation was defined as the ejector efficiency for various refrigeration system loads, which determined the ejector operating conditions (OCs). The optimisation results showed a very strong relation between ejector performance and the mixer diameter. Similar results were obtained with both optimisation algorithms. In each case, the mixer diameter was nearly equal to that of the baseline; however, the length of the mixing section was greater in the optimal design. The ejector efficiency of the optimised design was as much as 2% higher than that of the baseline design.