WINDAGE LOSSES IN CO2 TAYLOR-COUETTE-POISEUILLE FLOW UNDER WIDE RANGES OF PHYSICAL PROPERTIES

Supercritical carbon dioxide (sCO2) Taylor-Couette-Poiseuille (TCP) flow in high-speed rotating turbomachinery generates significant windage losses. Current research on windage losses is limited to specific pressure ranges and lacks a unified understanding of their variation across broad operating conditions. This study analyzes windage loss variations in sCO2 TCP flow and develops a high-precision prediction model for the surface friction coefficient (Cf) - a critical parameter for windage loss estimation under diverse thermophysical conditions. Numerical simulations employing the shear stress transport (SST) turbulence model characterized TCP flow dynamics. The effects of pressure (1-15 MPa) and temperature (273.15-413.15 K) on flow stability, vorticity, and wall shear stress were systematically investigated. The results indicated that near-critical conditions (293.15 K, 7.38 MPa) enhance vortex formation and shear stress fluctuations, significantly increasing windage losses. Traditional empirical correlations for Cf, constrained by narrow operational ranges, were compared with a backpropagation (BP) neural network model. The neural network exhibited superior accuracy and generalization by capturing nonlinear relationships between Cf and thermophysical parameters, axial/rotational Reynolds numbers, and geometric factors, achieving relative validation errors below 4.48%. This study proposes a computational model for windage loss prediction in sCO2 TCP flow under wide-ranging thermophysical property variations，providing a robust tool for optimizing turbomachinery design and accelerating the technological maturation of sCO2 power cycles in next-generation thermal energy conversion systems.