Abstract
The continuous temperature and pressure increase to target thermodynamic efficiency and power density in modern gas turbines (GT) stretches the design space and calls for high-pressure turbines (HPT) designed with very accurate design tools. Cooling, both internal and external, is necessary as the evolving fluid temperature is beyond the material’s capability. Therefore, the HPT design process is growingly assisted by computational fluid dynamics (CFD) that due to computers and numerical methods improved efficiency and accuracy respectively allows running CFD during design iterations. The design evolves from the conceptual phase, usually completed with the help of simple correlations and proprietary technology curves, not covered here, and proceeds with the preliminary and detailed phases where aero, thermal, mechanical, and geometrical details are progressively defined. It is paramount that design choices made in the preliminary phase are refined in the detailed phase without requiring excessive rework.
This paper concentrates on the CFD approach typically used in a detailed design phase, where the flow path and airfoils defined in a preliminary phase are completed with the geometrical and operational details of disk cavities and their purge, shrouds or squealer tips, seals, cooling and all the details that completely define the HPT. The paper describes the challenges of incorporating all the geometrical details and compares two approaches: steady-state with mixing plane and unsteady with a specific technique to reduce the computational effort compared to the full-wheel simulation (Time Transformation method). The differences in terms of aerodynamics and adiabatic wall temperatures are highlighted and assessed in the perspective of the computational effort associated with the different modelling approaches. The analysis reveals that unsteady effects and cavity purge flows and cavity geometry may have an important effect on the predictions of the thermal load in specific areas of the hot gas path.
