Large-Eddy Simulation of Film Cooling in an Adverse Pressure Gradient Flow

Abstract

In order to analyze the interaction of multiple rows of film cooling holes in flows at adverse pressure gradients, large-eddy simulations (LESs) are performed. The considered three-row cooling configuration consists of inclined cooling holes at an angle of 30 deg with a lateral pitch of p/D=3 and a streamwise spacing of l/D=6. The cooling holes possess a fan-shaped exit geometry with lateral and streamwise expansions. For each cooling row the complete internal flow is computed. Air and CO2 are injected in order to investigate the influence of an increased density ratio on the film cooling physics at adverse pressure gradients. The CO2 injected at the same blowing rate as air shows a higher magnitude of the Reynolds shear stress component and, thus, an enhanced mixing downstream of the cooling holes. The LES results of the air and CO2 configurations are compared to the corresponding particle-image velocimetry (PIV) measurements and show a convincing agreement in terms of the averaged streamwise velocity and streamwise velocity fluctuations. Furthermore, the cooling effectiveness is investigated for a zero and an adverse pressure gradient configuration with a temperature ratio at gas turbine conditions. For the adverse pressure gradient case, reduced temperature levels off the wall are observed. However, the cooling effectiveness shows only minor differences compared to the zero pressure gradient flow. The turbulent Schmidt number at CO2 injection shows large variations. Just downstream of the injection it attains low values, whereas high values are detected in the upper mixing zone of the cooling flow and the freestream at each film cooling row.

Effect of Blade Row Interaction on Rotor Film Cooling

“…Detached-eddy simulation (DES), which operates as URANS in boundary layers and LES away from walls, is used by Kim and Hassan (2010) and Foroutan and Yavuzkurt (2015). Implicit sub-grid scale modelling, where numerical dissipation is assumed to account for unresolved turbulent dissipation, is also used by some authors (Guo et al, 2006;Konopka et al, 2013;Muldoon and Acharya, 2006;Oliver et al, 2017;Renze et al, 2008).…”

“…required number of grid points for a successful simulation. However, recent studies with second-order codes have used approximately 15 million points (Konopka et al, 2013;Zhong et al, 2016).…”

Hybrid LES-RANS study of an effusion cooling array with circular holes

“…Their main focus was on the interaction of an acoustic wave with the coolant jets rather than the cooling effectiveness. Konopka et al [13] analysed the interaction of multiple rows of coolant jets using LES.…”

Large eddy simulation of film cooling on turbine vane

“…Zhong et al (2016) indicated that a steady core with low-temperature variance appears near the exit of the round hole, and high-temperature variance appears around this steady core. However, in large eddy simulation (LES) results from Renze et al (2008), Guo et al (2006) and Konopka et al (2013), hairpin vortexes are replaced by a large number of broken vortexes, and horseshoe vortexes also become unobvious. It might be because that, in these cases, the mainstream turbulent boundary layer is fully-developed before reaching the discharged hole.…”