Development of a Numerical Based Correlation for Performance Losses due to Surface Roughness in Axial Turbines

Abstract

In the present work, a recently developed in-house 2D CFD code is used to study the effect of gas turbine stator blade roughness on various performance parameters of a two-dimensional blade cascade. The 2D CFD model is based on a high resolution flux difference splitting scheme of Roe (1981). The Reynolds Averaged Navier-Stokes (RANS) equations are closed using the zero-equation turbulence model of Baldwin-Lomax (1978) and two-equation Shear Stress Transport (SST) turbulence model. For the smooth blade, results are compared with experimental data to validate the model. Finally, a correlation between roughness Reynolds number and loss coefficient for both turbulence models is presented and tested for three other roughness heights. The results of 2D turbine blade cascades can be used for one-dimensional models such as mean line analysis or quasi-three-dimensional models e.g. streamline curvature method.

Dynamic Stall Analysis of S809 Pitching Airfoil in Unsteady Free Stream Velocity

“…Nevertheless, having a y  lower than 1 leads to more accurate results. In the present study, the maximum value for y  is equal to 3.5 which gives acceptable results while reducing the computational overhead (compared to the y  case) of the simulation [18,19]. First, the numerical accuracy of  SST model has to be reliable if y  of the adjacent node to the blade surface would be lower than 5.…”