CFD simulations have been set-up with an isothermal boundary condition at the casing for running the NASA Rotor 37 axial compressor. The casing temperature was set to the inlet total temperature. The comparison to data was much improved for the efficiency for the 100% speed line relative to the adiabatic simulations. The efficiency difference between the isothermal and adiabatic solutions is about 1%, with the isothermal calculation matching the low flow test condition. The profiles of total temperature with the isothermal boundary condition matched the data near the casing without any overshoot, typical of most compressor calculations. Also the efficiency profile had a similar improvement in matching the data because of its relationship to temperature. A similar comparison between isothermal and adiabatic cases has been carried out for the same geometry with double the design clearance. The working range based on the steady CFD calculations is about half that of the design clearance case which is felt to be realistic. Moreover a detailed analysis based on conservation of Rothalpy has been made and applied to the rotor. Mass averaged Rothalpy is not conserved due to a frictional power term associated with the stationary case as well as heat transfer. The effects of these terms show the extent of the heat transfer is between 10-20% of span away from the casing. The heat transfer effect calculated with the isothermal boundary condition simulation is thought to be real, and accounting for it matches data better than using an adiabatic assumption. However, the real rig would probably not be isothermal at the casing and may require more complex simulations such as a conjugate heat transfer approach.
展开▼