Single-phase ejector geometry optimisation by means of a multi-objective evolutionary algorithm and a surrogate CFD model

The effort made by researchers and companies to reduce the expected increase in the energy demand for cooling and air conditioning has generated an interest in ejector cooling technologies. Researchers have proposed several typologies of ejector cooling cycles. However, the Coefficient of Performance achieved by this type of cycles is still poor, mainly due to the poor performance of the ejector. Several authors have dealt with the low performance of ejector cooling cycles by suggesting modifications to the cycle combining the ejector with other components. However, few approaches have attempted to optimise the geometrical parameters defining the ejectors, and even fewer to optimise the back pressure and the entrainment ratio, which are key parameters in the ejector efficiency. The main aim of this article is to optimise the performance of a single-phase ejector by means of a Multi-Objective Evolutionary Algorithm coupled with a surrogate model based on CFD simulations. The methodology has been implemented for air as an ideal gas and for CO2 with a real gas model. The results showed a potential increase of 55% and 110% in the back pressure and the entrainment ratio for air and 10% and 35% respectively for CO2. •Multi-Objective optimisation problem for air and CO2 single-phase ejector geometries.•A surrogate model based on CFD simulations to calculate back pressure and entrainment ratio.•110% of improvement in the entrainment ratio for air and 35% for CO2.•55% of improvement in the back pressure for air and 10% for CO2.•Promising results were found by a robust and reliable methodology.