Investigation of Turbulence Modeling and Harmonic Balance Methods Towards Accurately Predicting Compressor Flow Fields

Numerical investigations of the NASA stage 37 compressor case are presented.  Simcenter STAR-CCM+ is used for RANS based aerodynamic quantifications of the transonic compressor, which is operating at near stall conditions and the peak efficiency point. In these axial turbomachines, the unsteady effects are significant, and need to be accounted for in the design methods. Typically, transient simulation of fully realistic engine hardware is unrealistic in terms of the computational expense. However, using a harmonic balance approach in computational fluid dynamic (CFD) simulation has been shown to have a proficiency in capturing the dominant unsteady behaviors. As such, the performance of the NASA stage 37 compressor is characterized with both steady and harmonic balance approaches. Furthermore, a thorough exploration and sensitivity study on the turbulence modeling is conducted.  The lag k-ε turbulence model is considered, due to its capability for improved predictions in highly separated turbomachinery flows.  To the sentiment of reducing computational expense in performing these engineering calculations, an optimization study is performed in HEEDS on the CFD coupled flow solver.  A 14 parameter optimization study is conducted in over 150 designs, towards improved predictions at the peak efficiency point with resulting run times that were 91% faster than the baseline simulation. Further tests conducted with these results, which were blind to the optimization process, affirmed the positive effect on final run times in similar compressor simulations.