Flow Separation Modelling Dowstream of Military Low Pressure Compressors

Abstract

Abstract In military turbofan architectures, a flow separation is operated downstream the Low Pressure Compressor (LPC), mainly to increase the propulsive efficiency and to cool down and provide oxygen to the reheat channel. The flow is separated into a primary flow that goes through the high-pressure components and combustion chamber and a secondary flow which remains cold and is mixed with the primary flow before the reheat. The Bypass Ratio (BPR) represents the mass flow rate ratio between secondary and primary flow. During engine operation from high to low power, the BPR varies drastically. Moreover, current and future military engines present many variable geometries to improve their maneuverability, and the BPR can strongly depend on the settings of these geometries. BPR variations lead to a shift in the LPC exit separation streamline and therefore to modifications on the thermodynamic conditions in both primary and secondary flows. Existing methods usually perform the separation on the overall LPC map using corrective coefficients. The aim of this work is to provide an alternative method to model the effects of the shifting of the separation line on the thermodynamic parameters. Instead of using an overall compressor map, this method uses two compressor maps, one for the inner and one for the outer stream, at a reference BPR. When the BPR is different from this reference value, corrections are applied to the reference maps. The main hypothesis of this method is that the aerodynamics of the LPC is uncoupled from the splitter, meaning that the total temperature and total pressure profiles do not depend on the BPR. Analytical profiles are used from hub to tip for work and losses coefficients parameters. The analytical profiles are chosen to better fit the actual profiles in the working BPR range of interest. This method is designed to be included in a performance model. The method presented is validated on LPC Rig test and CFD data.