Abstract
In gas turbines, the stator wells play a key role in the efficiency of the turbomachine. The research for performance gains requires a good understanding and an accurate modeling of the flows and heat transfers occurring in these areas. Within the framework of the European program main annulus gas path interaction (MAGPI) WP1, a two-stage axial turbine test rig provided an experimental database used to validate the computational fluid dynamics (CFD) models. The aim of this study is to setup a numerical methodology using the CFD solver ANSYSFluent to accurately predict the conjugate heat transfer in the stator well area. The validation of the methodology relies on thorough comparison of the results with the MAGPI WP1 experimental temperature/pressure measurements. A geometry with axial cooling injection through lock plate slot was chosen. A Reynolds-averaged Navier–Stokes (RANS) three-dimensional sectorized CFD model of the turbine with conjugate heat transfer was used. It includes main gas path, cavities with labyrinths, disks rotor, the casing, and the nozzle guide vanes (NGV). Mixing planes are placed between the static and rotating frames. Different influences (mesh, turbulence model, thermal boundary conditions, radial labyrinths clearances) were studied and compared with experimental data. As a baseline, the first calculations were performed with a cooling flowrate chosen so that hot gas ingresses from the main stream into the stator well cavity. Good agreements between predicted and measured temperatures/pressures were observed, especially in the vicinity of the stator well. Discrepancies were spotted at the first rotor hub endwall and at the upstream wheelspace and will be discussed. Two other cooling configurations were conducted, one with cooling air exiting from the disk rim cavity to the main gas path and the other with the lowest cooling flowrate and so the highest ingress. Finally, the turbine performance under nonadiabatic conditions has been evaluated with an appropriate efficiency definition.
