Computation of Subsonic and Transonic Compressor Rotor Flow Taking Account of Reynolds Stress Anisotropy

A two-layer k-ε/algebraic Reynolds stress model (ARSM) has been adopted to the three-dimensional, Reynolds-averaged, Navier-Stokes code to include explicitly the Reynolds stress anisotropy. The code has been used to study the complex flow fields of a transonic axial compressor rotor (i.e., NASA Rotor 37) and a subsonic centrifugal compressor impeller (i.e., the backswept impeller of Krain, first reported in 1988). The computed results have been compared with those from a Baldwin-Lomax model, a low-Reynolds number k-ε turbulence model and actual experimental data. Calculated results for the axial compressor are compared with data reported by Suder in 1994. The suitability of the turbulence model to predict accurately the overall performance of the rotor, spanwise distributions of aerodynamic characteristics, and the wake flow profiles is assessed. Calculations for the centrifugal compressor impeller are compared with the experimental data reported by Hah and Krain in 1989. The usefulness of the turbulence models to predict accurately the overall performance of the impeller, the impeller-exit-velocity profile, and the meridional velocity and flow angle profiles at the cross-channel planes (via L2F measurements) has also been investigated. For modeling the turbulence of both the rotor and the impeller, reasonably good predictions have been obtained with the ARSM and the low-Reynolds number k-ε models, but have not been attainable using the Baldwin-Lomax model. The solutions obtained with the ARSM show better agreement with experimental data than those obtained with the other models. However, in some cases, the predicted differences between the ARSM and the low-Reynolds number k-ε models are not significant. The computed secondary flow and the relative helicity have also been used to investigate the effect of wall curvature and frame rotation on the flow field inside the centrifugal impeller for three operating conditions (i.e., design point, choke, and near surge) and the results are discussed.