Active Turbulence Generation for Film Cooling Investigations

Abstract

In modern gas turbines, heat loads on thermal highly stressed components are reduced by film cooling, where a layer of cold gas is injected for the protection of these components. In order to optimize present cooling designs, experiments under realistic operating conditions have to be performed especially including the effect of turbulence intensity and turbulent length scale. In this work, an active turbulence grid was designed, built and tested in order to increase the turbulence conditions in a closed-loop, thermal wind tunnel facility for future film cooling investigations. The grid design, which is based on designs proposed in literature, and its implementation are described in detail. For the investigation of the resulting flow field without film cooling injection, the measurement techniques hotwire anemometry and high-speed PIV were used, which are described shortly. The measurements were carried out at different axial positions downstream of the turbulence grid, at different main flow velocities and various rotation rates of the grid. The results show that the turbulence intensity decays with increasing distance and stays constant at a distance of X/M = 14 downstream of the grid, which will be the position of film cooling flow injection in future investigations. For the investigated measurement points a decreasing rotation rate of the grid leads to an increase of the turbulence intensity. Increasing the main flow velocity significantly increases the turbulence intensity especially close to the grid. The calculated turbulent length scales for different axial positions downstream of the grid and three different main flow velocities stay within a narrow band between 10 mm and 30 mm, which is below the mesh size of the grid. Furthermore, the calculated data for different rotation rates and main flow velocities at X/M = 14 show a constant turbulent length scale of 20 mm for rotation rates higher than 1200 rpm, independently of the main flow velocity. However, for lower rotation rates a strong dependence of the turbulent length scale on rotation rate and on main flow velocity was seen. The results of both measurement techniques match very well, leading to the conclusion that the presented approach investigating turbulence intensity and turbulent length scale provides a reliable database for future investigations of film cooling configurations.