A combined CFD and FE method, which could be applied to a wide range of internal air system rotor-stator cavities and which overcomes the disadvantages of many non-coupled approaches, is presented. It is used to predict windage and heat transfer in the HP compressor rear cone outer cavity of a service aeroengine. From the CFD it is shown that rotor wall torque, and hence windage, decreases as cavity throughflow increases, and that the data from several engine cavities can be reduced to a single characteristic of windage versus mass flow. Stator wall torque is also presented. A comparison with engine thermocouple data shows that further development of the modelling is required before engine testing could be replaced.

In an engine design process, thermo-mechanical analyses of compressor drums and casings are undertaken, to predict component temperatures and displacements, which are ultimately used for material selection, blade clearance control and lifing of components. The thermal boundary conditions are sourced from a small number of standard flow field and heat transfer solutions, leading to a reliance on engine thermocouple tests to provide calibration factors on the boundary conditions, which with changes in inlet flows and cavity geometry from the tested arrangements are unproven, limiting the ability to readacross the test information into new designs. Given that the thermal boundary conditions in compressor drum and casing components are largely driven by complex flow physics, in the absence of suitable test information, CFD methods can be used to provide boundary specification of the thermo-mechanical problem, incorporating the complex physics involved. Without the insight of the flow field solution in complex flow regions, specification of the boundary conditions is rather subjective and mostly based on intuition. This study shows the use of CFD to provide the boundary conditions for the rotor-stator cavity at the front of an IP compressor drum. The CFD is run adiabatically and through a set of unit heat transfer cases on separate sections of the cavity wall, at key points in the flight cycle. The analyses provide appropriately characterized thermal boundary conditions (specifically heat transfer coefficients and adiabatic wall temperatures) that are transferred into the thermo-mechanical model, which can then be run through a wide range of cycles without the need for further CFD calculations.

Surface-mounted thermocouples were located on the annulus walls of every stator stage in both core compressors of a large civil aeroengine. The relationships of the steady state wall temperatures to the compressor inlet and exit air temperatures are shown, as is their variation with engine rotational speed. The transient response of the thermocouples to fast engine throttle manoeuvres was measured, for acceleration to full power, deceleration to idle and for both acceleration and deceleration to two intermediate power levels. Heat transfer coefficients are deduced from the matching of a thermal finite element model to the measured transients and are expressed as factors with respect to the standard free-disc formula. The variations in the factors through the compressor and through the speed range are shown. The advantages and disadvantages of using an alternative standard formula, based on duct flow, are discussed. Finally, the importance of understanding the transient local air temperature response arriving at a compressor stage due to upstream slugging is shown.