Influence of Temperature Distribution on Radial Growth of Compressor Discs

Abstract

For the next generation of aero-engines, manufacturers are planning to increase the overall compressor pressure ratio from existing values around 50:1 to values of 70:1. The requirement to control the tight clearances between the blade tips and the casing over all engine-operating conditions is a challenge for the engine designer attempting to minimise tip-clearances losses. Accurate prediction of the tip clearance requires an accurate prediction of the radial growth of the compressor rotor, which depends on the temperature distribution of the disc. The flow in the rotating cavities between adjacent discs is buoyancy-driven, which creates a conjugate heat transfer problem: the disc temperature depends on the radial distribution of Nusselt number, which in turn depends on the radial distribution of disc temperature. This paper focuses on calculating the radial growth of a simplified compressor disc in isolation from the other components. Calculations were performed using one-dimensional (1D) theoretical and two-dimensional finite-element computations (2D FEA) for overall pressure ratios (OPR) of 50:1, 60:1 and 70:1. At each pressure ratio, calculations were conducted for five different temperature distributions; the distribution based on an experimentally-validated buoyancy model was used as the datum case, and results from this were compared with those from linear, quadratic, cubic and quartic power laws. The results show that the assumed distribution of disc temperature has a significant effect on the calculated disc growth, whereas the pressure ratio has only a relatively small effect. The good agreement between the growth calculated by the 1D theoretical model and the FEA suggests that the 1D model should be useful for design purposes.

Ekman Layer Scrubbing and Shroud Heat Transfer in Centrifugal Buoyancy-Driven Convection

“…Long and Childs [22] employed Lloyd and Moran's [27] correlation to scale their heat transfer measurements, using a Nusselt number based on the core-to-shroud temperature difference. The modified 1 Rayleigh number Ra and the modified Nusselt number Nu are given in Eqs. (15) and (16).…”

“…Without corresponding reductions in blade tip-to-casing clearance this will result in a larger blade tip clearance-to-blade span ratio, which will in turn lead to proportionally higher tip clearance losses. The blade tip clearance of the last axial compressor stage at near nominal engine speed is of the same order of magnitude as the thermal and centrifugal growths of the compressor rotor disc [1]. The thermal growth is determined by the thermal stresses on the discs characterised by the disc temperature distribution, which is associated with the flow and heat transfer in the rotating compressor disc cavities.…”

Measurement and Analysis of Buoyancy-Induced Heat Transfer in Aero-Engine Compressor Rotors

“…This review was extended by Luberti et al [2], including discussion of the most recent and relevant experimental work from the universities at Sussex [3], Dresden [4,5] and Beihang [6]. Luberti et al [7] showed that the radial distribution of temperature has a significant effect on the disc growth and consequently on the blade clearance in a compressor. Although the growth due to the thermal stress is small relative to the total growth of the disc, it is the same magnitude as the blade clearance.…”

Ekman Layer Scrubbing and Shroud Heat Transfer in Centrifugal Buoyancy-Driven Convection

“…Long and Childs [22] employed Lloyd and Moran's [27] correlation to scale their heat transfer measurements, using a Nusselt number based on the core-to-shroud temperature difference. The modified 1 Rayleigh number Ra and the modified Nusselt number Nu are given in Eqs. ( 15) and ( 16).…”