Prediction of Velocities and Heat Transfer Coefficients in a Rotor-Stator Cavity

The calculation of swirl velocities and convective heat transfer coefficients in a rotor-stator cavity has been mostly based on equations taken from empirical data. However, the validity of these empirical relations is questionable in geometries and environments other than the specific ones for which they were derived. A commercial CFD code, Fluent, has been used to predict the swirl velocities and rotor disk convective heat transfer coefficient distribution for a rig at Arizona State University. The rig was run at several rotational Reynolds numbers (Re_φ) varying from 4.6×10⁵ to 8.6×10⁵ and for various mass secondary flows. Several different turbulence models were used and the resulting predictions were compared with data obtained from the rig. Fluent was able to predict the swirl velocities, on average, within 30% and the convective heat transfer coefficients, on average, within 30% and often within 20%. The degree of agreement with the measured data was found to depend on which turbulence model that was used, mesh resolution, as well as the secondary flow and Re_φ.