Calibration strategy of a 1D model for supersonic ejectors

Ejectors are passive devices used for the purpose of compressing or entraining a fluid. They can be used in a wide range of domains such as aerospace, desalination and, more importantly, in refrigeration cycles. Their main benefit is the fact that their operation requires no moving pieces. However, the design of these devices remains a challenge given the particularly complex phenomenon occurring inside. A way to predict simply and quickly the performance of a given ejector is by the use of so-called thermodynamic models. These models need to be calibrated because they use efficiencies to characterize losses inside the ejector. In this thesis, various strategies to calibrate a thermodynamic model are adopted. To do so, the relationships between all the parameters of the model are explored. These calibration strategies are planned to make the thermodynamic model usable so as to be able to further calibrate the efficiencies with a machine learning model.