Prediction of the Unsteady Flow Field in a Turbine including Air Seal Cavities with Time and Frequency Domain Methods

Abstract

In recent years, the effects of side geometries, such as air seal cavities, have gained increasing attention in the design process of modern turbomachinery applications. This study investigates the hypothesis that radial mixing effects from these geometries can significantly influence performance metrics and radial temperature distributions. This paper explores the use of computational fluid dynamic (CFD) simulations to predict these effects accurately. Our findings suggest that unsteady approaches, such as the Harmonic Balance (HB) method, are essential for capturing the most significant effects while also reducing computational effort, both of which are critical for an efficient design process. The study focuses on a subsonic one and half stage turbine with inner and outer air seal cavities, representative of high and low pressure subsonic turbines in modern aircraft engines. For reference, highly accurate time-resolved results are presented using the conventional dual-time stepping method in the time domain. The flow field within the cavities is analyzed with respect to its frequency content. Cavities are integrated using a sliding mesh approach with a novel, efficient implementation for the frequency domain solver. It is shown that integral performance quantities, such as mass flow and isentropic efficiency, are accurately resolved by the HB method, unlike steady RANS methods. Time-averaged radial profiles of temperature are evaluated and compared to URANS reference results in the time domain. Additionally, the sliding mesh interface demonstrates consistent results across both time and frequency domain methods.