Research in Engineering and Aviation

Computational Study of High Lift Low Pressure Turbine Cascade Aerodynamics at Low Reynolds Number

March/April 2013

Author(s): Arko, B., and McQuilling, M.

Journal: Journal of Propulsion and Power, 29(2), 446-459, March-April 2013. DOI: 10.2514/1.B34576


This study uses a Reynolds-averaged Navier–Stokes finite volume flow solver to simulate the flowfields around a two-dimensional linear turbine cascade model at a Reynolds number of 25,000. Three blade profiles have been simulated, including the aft-loaded Pack B, which has a nominal Zweifel loading coefficient Zwequal to 1.15, the midloaded L1M (Zw=1.33), and the front-loaded L2F (Zw=1.59). All three blade profiles are known to be susceptible to varying degrees of laminar flow separation along the suction surface. Turbulence models used, which to the authors’ knowledge have been applied for the first time here, are the Abe–Kondoh–Nagano linear low-Rek-εas well as the Kato–Launder modification. Time-accurate simulations, including fully laminar computations, are compared with experimental data and higher-order computations to judge the accuracy of the results, where it is shown that Reynolds-averaged Navier–Stokes simulations with appropriate turbulence modeling can produce both quantitatively and qualitatively similar separation behavior as seen in experiments. However, large-scale vortical motions predicted by the turbulence models and laminar solver advect downstream and instigate flow reattachment, a result the authors believe would be attenuated to some degree by three-dimensional vortex breakdown. Results collectively show that characterization of transition to turbulence as judged by analysis of the Reynolds shear stress is not sufficient in two dimensions, and further analysis including spectral methods may be necessary to better predict transition at low Reynolds number.