Improved Prediction of Turbomachinery Flows Using Reynolds Stress Model with Helicity Modification and Modified 𝜸 Transition Model

Turbomachinery flows, characterized by strong turbulent non-equilibrium and anisotropy, pose challenges for Reynolds-Averaged Navier-Stokes predictions, especially in cases involving corner separation and boundary layer transition. To improve accuracy, a blending of the Speziale-Sarkar-Gatski/Launder-Reece-Rodi Reynolds stress model (RSM) and the intermittency (𝛾) transition model (SSG/LRR-𝜔-𝛾 model) is implemented in OpenFOAM. A modification of 𝛾 transition model is proposed to address the delay in the transition onset in the T3 series of flat plates under adverse pressure gradients. Additionally, by incorporating the mechanism of energy backscatter, a helicity-modified production term is introduced into the scale determination equation to improve the prediction accuracy of large-scale vortices, especially in scenarios like compressor corner separation, resulting in the SSG/LRR- 𝜔 -H- 𝛾𝑚 model. Validation is conducted across the T3 flat plates and the linear highly loaded prescribed velocity distribution (PVD) compressor cascade. Results indicate improvements in the transition onset under adverse pressure gradients for the T3 flat plates. In the PVD compressor cascade, the proposed SSG/LRR-𝜔-H-𝛾𝑚 model exhibits prediction accuracy in transition onset comparable to that of the SSG/LRR-𝜔-𝛾 model and superior to the shear stress transport model with 𝛾 transition model (SST-𝛾 model). Notably, the proposed model demonstrates accurate predictive performance for complex three-dimensional corner separation involving boundary layer transition in the PVD compressor cascade.