Challenges in scaling sCO2 compressor speedline to different intake thermodynamic conditions

Compressors operating with carbon dioxide near the critical point experience complex aerothermodynamic phenomena, where deviations from perfect-gas similarity and two-phase flow effects dominate. Existing models inadequately capture the impact of intake thermodynamic conditions on the choked flow rate, leaving a gap in predictive capabilities for these machines. This work addresses this gap by deriving a correlation to predict the choked flow rate as a function of two generalized parameters: the cavitation/condensation parameter and the isentropic pressure-volume coefficient, which describe two-phase and non-ideal effects. A database of 100 speedlines, generated through CFD simulations with varying thermodynamic conditions and fixed peripheral Mach number, was used to train a symbolic regression algorithm based on gene expression programming. This method was chosen to derive an explicit, easy-to-use analytical expression without assuming a priori functional forms. Results showed that the choked flow rate could vary from 90% to 155% of the nominal value depending on thermodynamic conditions, highlighting the dominant role of the two parameters. The derived correlation demonstrated trends consistent with CFD predictions, with an accuracy of ±3 percentage points for most cases. However, an a-posteriori validation against varying peripheral Mach numbers and an alternative impeller geometry revealed significant discrepancies, underscoring the interplay between thermodynamic conditions, geometry, and aerodynamics. This analysis showed that the peripheral Mach number and the geometric features influence choking behavior unpredictably, limiting the correlation's general applicability. While the proposed correlation is not adequate for quantitative scaling across designs, it provides preliminary insights into qualitative trends. For accurate predictions, high-fidelity CFD remains necessary, highlighting the inherent challenges of universal scaling for near-critical operations in sCO₂ compressors.