This study presents a novel methodology for investigating and analyzing the aerothermal dynamics of high-loaded gas turbine vanes. The approach integrates an advanced Wall-Modeled Large Eddy Simulation (WMLES) technique with an Immersed Boundary Method (IBM), enabling the treatment of complex geometries on Cartesian grids, while accurately handling high Mach and Reynolds number flows. A notable strength of the method, as compared to others discussed in the literature, is its compatibility with high-order numerical schemes and explicit algorithms. This enables efficient computation on modern GPU-based architectures, allowing for scale-resolved simulations to be conducted within a time frame comparable to that of standard unsteady Reynolds-Averaged Navier-Stokes (RANS) simulations. The effectiveness and accuracy of the methodology are validated in canonical configurations, including channel flow, spatially developing boundary layers, and shock-boundary layer interactions. The present discussion introduced the method's ability to replicate renown experimental data associated with a gas turbine nozzle in the transonic regime [1] demonstrating alignment between the numerical and the experimental results.
Scale-Resolved Modeling of Gas Turbine Vanes: A Wall-Modeled LES and Immersed Boundary Approach
De Vanna F.
;Benini E.
2025
Abstract
This study presents a novel methodology for investigating and analyzing the aerothermal dynamics of high-loaded gas turbine vanes. The approach integrates an advanced Wall-Modeled Large Eddy Simulation (WMLES) technique with an Immersed Boundary Method (IBM), enabling the treatment of complex geometries on Cartesian grids, while accurately handling high Mach and Reynolds number flows. A notable strength of the method, as compared to others discussed in the literature, is its compatibility with high-order numerical schemes and explicit algorithms. This enables efficient computation on modern GPU-based architectures, allowing for scale-resolved simulations to be conducted within a time frame comparable to that of standard unsteady Reynolds-Averaged Navier-Stokes (RANS) simulations. The effectiveness and accuracy of the methodology are validated in canonical configurations, including channel flow, spatially developing boundary layers, and shock-boundary layer interactions. The present discussion introduced the method's ability to replicate renown experimental data associated with a gas turbine nozzle in the transonic regime [1] demonstrating alignment between the numerical and the experimental results.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
