CORE-COMPRESSOR ROTATING STALL SIMULATION WITH A MULTI-BLADEROW MODEL

M. Vahdati, AI Sayma, M Imregun

Imperial College London,

Vibration UTC, MED, London, SW7 2BX, UK

m. vahdati@imperial. ac. uk, a. sayma@imperial. ac. uk, m. imregun@imperial. ac. uk

Simpson

Rolls-Royce plc,

Compressor Engineering, Derby DE24 8BJ, UK George-Simpson@rolls-royce. com

Abstract The paper will focus on one specific core-compressor instability, rotating stall, because of the pressing industrial need to improve current design methods. The long-term aim is to minimize the forced response due to rotating stall excita­tion by avoiding potential matches between the vibration modes and the rotat­ing stall pattern characteristics. Using a 3D viscous time-accurate flaw repre­sentation, the front bladerows of a core-compressor were modeled in a whole – annulus fashion. The engine core flow boundary conditions were obtained from a mixing-plane steady-state calculation for which the low pressure compression domain was also modeled. A variable-area nozzle, placed after the last com­pressor bladerow in the model, was used to impose ambient boundary conditions downstream. The rotating stall behavior at two different compressor operating points was studied by considering two different variable-vane scheduling condi­tions for which experimental data were available. In all cases, the rotating stall was initiated by introducing a small amount of geometric mistuning to the rotor blades. Using 3- and 6-bladerow models, the unsteady flaw calculations were conducted on 32-CPUs of a parallel cluster, typical run times being around 3-4 weeks for a grid with about 30 million points. The simulations were conducted over several engine rotations. As observed on the actual development engine, 6 rotating stall cells were obtained for the first scheduling condition while mal – scheduling of the stator vanes increased the number of rotating stall cells to 13. Although there was some discrepancy between predicted and measured speed of the rotating stall pattern, it was concluded that the large-scale modeling method­ology could simulate both the onset of rotating stall and its development as a function of vane scheduling.

Keywords: Rotating stall, core compressor forced response, compressor mal-scheduling

313

K. C. Hall et al. (eds.),

Unsteady Aerodynamics, Aeroacoustics and Aeroelasticity of Turbomachines, 313-329. © 2006 Springer. Printed in the Netherlands.

1. Introduction

Unsteady turbulent high-speed compressible fbws often give rise to com­plex aeroelasticity phenomena by infUencing the dynamic behavior of struc­tures on which they act. Under certain conditions, the energy transfer from the fUid to the structure can cause excessive vibration levels and structural integrity may be compromised. The problem is particularly severe for gas turbines where virtually all bladerows are susceptible to aeroelasticity effects either by inherent self-induced motion (flitter) or by response to aerodynamic fl)w distortions and blade wakes (forced response). Different aeroelasticity phenomena are associated with different components. For instance, fan and core-compressor blades are known to suffer from flutter and rotating stall. Tur­bine blades are subjected to aerodynamic excitation containing both high and low harmonics, the former due to wakes from upstream blades and the latter due to the general unsteadiness of the flow, usually caused by a loss of sym­metry. The most complex and the least understood aeroelasticity phenomena occur in multi-stage core compressors, the subject of this paper, because of their wide operating envelope. During engine development programs, costly structural failures are known to occur because of a mixture of aeroelastic insta­bilities such as acoustic resonance, flutter, forced response, buffeting, vortex shedding, etc. Most such phenomena are believed to be caused by at least one bladerow undergoing severe stall, but the overall compressor still managing to function because of the overall pressure ratio. The numerical modeling of such a situation is a formidable challenge as the analysis must be able to rep­resent accurately not only the aerodynamic and structural properties of a large number of bladerows but also the interactions through these.

Since the avoidance of stall and surge is a major design consideration, a considerable amount of research effort has been devoted to the understanding of the physical mechanisms that give rise to such instabilities. After the initial inception stage, it is not clear which conditions will cause surge or rotating stall, though a simplified non-dimensional parameter, based on basic geometry, the mean fl»w speed and the speed of sound, has been proposed by Greitzner (1978) who provided a means of assessing how design changes were likely to affect the stall/surge behavior of a particular compressor. However, in the general case, there are no rules to determine the speed of rotating stall, its circumferential and radial extent as well as the number of rotating cells. So far, due to modeling difficulties, much of the stall and surge research has been experimental (Longley et al. 1990, Weigl et al. 1998), or based on simplified models (Moor et al. 1986, Paduano et al 1994), though numerical simulations with simplified or partial geometries are beginning to emerge (Niazi 2000, He 1997).

Unlike other instabilities of aeroelastic origin, core-compressor rotating stall behavior needs to be simulated not only at the onset, but also during the event itself. Such a requirement arises from the need to avoid a match between the number and speed of rotating stall cells and the assembly resonances. From an industrial perspective, it is important to understand the factors which inflience changes in stall characteristics as it may result in moving from a safe operating regime to one which causes rotor blade failure. These effects are normally assessed by engine strain gauge tests which are used to map out rotating stall boundaries. However, due to the high cost of these tests and the limited scope for varying engine configuration and operation, the objective of this work is to develop a predictive capability which will minimize rig and engine testing by means of advanced simulations that will cover the fight envelope in an efficient manner.