Ejectors are passive entrainment and compression devices that have gained particular attention in refrigeration, where they are used as passive compressors. Yet, they present numerous challenges, particularly in transcritical carbon dioxide (CO2) refrigeration systems. These challenges include limited experimental data due to small ejector sizes, transcritical flows inducing flashing, and the complexity of high-cost Computational Fluid Dynamics (CFD) simulations. The purpose of the present research is threefold: (i) identify the main ejector design objectives for optimal system performance, (ii) explore the link between local exergy transport in the ejector and system performance, and (iii) assess the impact of metastability on the performance of both the ejector and refrigeration system. To study ejector operation within the cycle and determine the optimal possible performance, thermodynamic modeling is used to generate performance maps of the ejector cycle. The physically possible operation domain of the Ejector Refrigeration System (ERS) is determined using a state-of-the-art ejector model. Findings suggest that the optimal Coefficient Of Performance (COP) is not achieved when the ejector operates under the on-design regime, i.e., optimal at the component scale, but mainly (depending on operating conditions) under the off-design regime, i.e., not optimal at the ejector scale. Moreover, limiting entropy generation within the device emerges as a highly beneficial approach at the ejector level to enhance cycle performance. To investigate the effects of thermodynamic non-equilibrium, i.e., the presence of metastable states, a Homogeneous Relaxation Model (HRM) is implemented for the first time in the density-based solver SU2. A novel quadtree algorithm is utilized to tabulate thermodynamic relations for the equation of state, which is directly integrated within the solver to ensure fast and accurate predictions. Validation of the simulation tool against experimental nozzle pressure profiles is conducted, comparing the HRM to the commonly used Homogeneous Equilibrium Model (HEM). Notably, the HRM can better fit experimental pressure profiles, particularly during the first part of the expansion. The connection between local exergy transfers within the ejector and overall cycle performance is investigated using a multi-scale numerical approach. Numerical simulation (CFD) accurately predicts ejector operation and provides insights into local flow quantities, while the exergy tube analysis allows to qualify the efficiency of the transfers at the component scale. Calibrating the aforementioned ejector thermodynamic model onto CFD results enables the determination of physically feasible ERS operations. Moreover, the impact of metastability on exergy transfers and system performance is assessed by comparing CFD results of the HEM and HRM. Interestingly, results indicate that the exergy transfers from the primary to the secondary stream is maximal for an off-design operation of the ejector. Additionally, the maximum exergy gain of the secondary stream correlates with the maximum coefficient of performance of the system. Overall, metastability is found to generally deteriorate both the internal performance of the ejector and that of the cycle.
Metsue, A. (2024). Advanced numerical and thermodynamic modeling of CO2 ejectors for their integration in refrigeration systems. https://hdl.handle.net/2078.5/233132