Experimental and numerical investigation of jet impingement cooling onto a rib roughened concave internal passage for leading edge cooling of a gas turbine blade

Abstract

Considered is thermal protection of the concave surface within an internal passage, which models the leading edge cavity of vanes or blades of stationary or aero engine gas turbines. Included along the internal surface are different arrangements of V-shaped ribs to further augment surface cooling heat transfer rates. Impingement jets are employed for the cooling, which are configured so that associated cooling air exits the cavity through a crossflow outlet and an array of staggered film cooling holes. The curved part of the asymmetric, concave target plate features a constant radius and connects to straight lines on the pressure and suction sides. Addressed are H/Djet values of 2.7 and 4.0, and Reynolds number Re values of 18,000, 24,000, and 30,000. A maximum crossflow configuration is utilized wherein the coolant flow exits the cooling passage through the film cooling holes and from one end of the passage, as the opposite end is blocked. Different surface configurations are investigated, including 4 V-ribs in four different positions, 8 V-ribs in one position, and a smooth surface of the leading edge cavity. Each V-rib geometry consists of a regular V-rib pointing upstream. A transient thermochromic liquid crystal (TLC) technique is employed to obtain spatially-resolved surface heat transfer data on the internal wall of the leading edge cavity. Resulting experimental heat transfer data are used to validate and verify CFD simulations. Employed for this purpose are steady-state 3D RANS simulations obtained using the OpenFOAM Version 7 numerical code with a kω-shear-stress-transport (SST) turbulence model. Experimentally-measured surface Nusselt number results indicate significant local enhancements adjacent to rib wakes, provided local crossflow velocity values are substantial. Local surface Nusselt number enhancements thus depend upon rib position, but local surface heat transfer increases are most strongly tied to magnitudes of local crossflow velocity. In contrast to these experimental results, CFD simulation results do not show significant variations as rib position is altered. The best thermal performance is associated with 4 V-rib arrangements in two positions, and 8 V-ribs in one position.