Aerodynamic Behaviour of a Novel Three-Dimensional Shaped Transonic Compressor Rotor Blade

This paper compares the aerodynamic behaviour of baseline and redesigned versions of the well-known NASA Rotor 37. The aerodynamic behaviour of the two rotors was evaluated using an accurate and validated 3-D CFD RANS model. The redesigned version showed both higher efficiency and wider stall margin. The new rotor was modeled by giving to the radial stacking line of baseline blade a three-dimensional shape. The blade was curved mainly toward the direction of rotor rotation. The applied blade curvature comes from previous personal investigations on the effects of stacking line shape in transonic bladings. Steady-state simulations were carried out to calculate the flow field inside the two rotors. The numerical model was developed using a commercial CFD code. The code was validated by simulating the Rotor 37 and comparing the computed results to the experimental data available in the open literature. The validation process gave a satisfactory agreement between predictions and measurements, showing that the overall features of the three-dimensional shock structure, shock-boundary layer interaction and tip clearance flows are calculated well in the numerical solution. With respect to the baseline rotor, the redesigned version gave a higher efficiency in a large part of the operating range, with a maximum increment of about 1.2 percentage points around the peak efficiency condition. The improvements in efficiency can be associated with a less detrimental shock-blade boundary layer interaction at the outer span, probably due to the different three-dimensional shock structure developed. At the outer span, in fact, the new blade showed a blade-to-blade shock front more oblique than in the baseline rotor. Further, the new blade positively impacted the flow field near the casing at low flow operating conditions. A less detrimental shock-tip clearance vortex-boundary layer interaction, along with a considerable reduction of the low momentum fluid region after the shock, was observed. CFD flow visualizations showed clearly a higher stall margin. The last stable operating point provided by the numerical model implemented gave a normalized mass flow of about 92% in the case of Rotor 37 (in accordance with experimental data) and about 90% in the case of redesigned version. The two rotors showed a similar choking mass flow rate.