Cooled Turbine Model

Abstract

A significant trend in aero engine design has been the rise in turbine inlet temperatures, as well as the drive to produce raise efficiency. Since the 1960s, turbine inlet temperatures have exceeded turbine material limits, with turbine cooling systems being used to bridge the gap. Modern engines require substantial amounts of cooling air, prompting a need to understand further the impact of turbine cooling on turbine and engine performance in cycle calculations.

Current models employed for performance analysis of cooled turbines are based on technical assumptions that are several decades old. This raises questions about the possibility of adapting models to more accurately represent modern engine technology. As such, this thesis aims to develop a cooled turbine model (CTM) for use in PyCycle, an open-source engine cycle analysis platform. The CTM is based on the cooled turbine blade row model defined by Young and Wilcox, which employs empirical constants to estimate cooling mass flow rates. The CTM can estimate the cooling flow requirements and the associated entropy rise for the turbine, accounting for the irreversibility in the turbine cooling process.

The implemented CTM models a complete turbine stage and multi-stage turbines based on three key aspects: the thermodynamics of a cooled turbine row, the work extraction in an equivalent uncooled turbine stage, and the conversion of thermodynamic properties between an absolute and rotating frame of reference. The CTM was verified and validated using three other cases, and it was found to accurately capture the effects of turbine cooling on bulk flow properties.

Following the implementation of the CTM, a study into the (semi-)empirical parameters and constants was performed to update existing parameters for modern engines. The limited availability of data relating to flow velocities and Mach numbers in aero-engine turbines forms a significant obstacle to the accurate specification of some empirical parameters. An estimation for the average Stanton number over turbine blades based on the gas temperature and Reynold’s number was derived and validated with the Von Karman Institute’s LS-89 turbine cascade results. The impact of updating the empirical parameters used in the CTM
has been assessed by formulating the cooling flowestimation as an optimization problem.
Using the updated ranges for the empirical parameters, the optimization study showed the potential for a significant reduction in the estimated cooling fraction.