This paper presents results of numerical investigations carried out to explore the benefit of end wall boundary layer removal from critical regions of highly loaded axial compressor blade rows. At the loading level of modern aero engine compressors, the performance is primarily determined by three-dimensional (3D) flow phenomena occurring in the end wall regions. Three-dimensional Navier–Stokes simulations were conducted on both a rotor and a stator test case in order to evaluate the basic effects and the practical value of bleeding air from specific locations at the casing end wall. The results of the numerical survey demonstrated substantial benefits of relatively small bleed rates to the local flow field and to the performance of the two blade rows. On the rotor, the boundary layer fluid was removed from the main flow path through an axisymmetric slot in the casing over the rotor tip. This proved to give some control over the tip leakage vortex flow and the associated loss generation. On the stator, the boundary layer fluid was taken from the flow path through a single bleed hole within the passage. Two alternative off-take configurations were evaluated, revealing a large impact of the bleed hole shape and the location on the cross-passage flow and the suction side corner separation. On both blade rows investigated, rotor and stator, the boundary layer removal resulted in a reduction of the local reverse flow, blockage, and losses in the respective near-casing region. This paper gives insight into changes occurring in the 3D passage flow field near the casing and summarizes the effects on the radial matching and pitchwise-averaged performance parameters, namely loss and deviation of the rotor and stator when suction is active. Primary focus is put on the aerodynamics in the blade rows in the main flow path; details of the internal flow structure within the bleed off-take cavities/ports are not discussed here.

The paper describes an advanced three-dimensional blading concept for highly loaded transonic compressor stators. The concept takes advantage of the aerodynamic effects of sweep and dihedral. To the knowledge of the authors this is the first approach reported in the open literature that combines those two basic types of lean in an engine-worthy aerofoil design. The paper makes a contribution to the understanding of the endwall effect of both features with special emphasis put on sweep. The advanced three-dimensional blading concept was applied to an Engine Section Stator (ESS) of an aero-engine fan. In order to demonstrate how three-dimensional flow can be controlled, numerical analysis of the flow structure in a conventional and an advanced stator configuration was performed using a three-dimensional Navier–Stokes solver. The numerical analysis showed the advanced blade improving both radial loading distribution and the three-dimensional endwall boundary layer development. In particular, a strong hub corner stall could be largely alleviated. High-speed rig testing of the advanced ESS confirmed the concept and showed good qualitative agreement between measurement and prediction. The work presented was closely linked to the development of the BR710 engine on which the advanced ESS is in service today.

This paper presents experimental investigations on a large-scale low-speed compressor facility with four repeating stages equipped with CDA-profiles (Controlled Diffusion Airfoils). Two different builds were investigated. Both builds used identical rotors, but had stators configured either in cantilevered or in shrouded form. Traverse measurements of total pressure and flow angle at six axial locations (IGV, two rotor and three stator exit planes) were performed between 1% and 99% annulus height and across two blade pitches. Circumferentially mass-averaged radial profiles were used in a through-flow code for reconstruction analysis of the measurements. In addition to the traverse measurements surface static pressures on stage 3 rotor and stator were measured. The effect of the “free-end” configuration on an embedded stage of this multistage compressor is described and compared to the shrouded configuration. The objective of this study was to investigate the differences of these two configurations and especially the effects caused by the hub clearance vortex in the cantilevered case. The entire set of measurements and through-flow analysis was performed at two operating points of the compressor i.e. at peak efficiency and near stall condition. Thus also the effects of the hub clearance vortex which influences the stall margin of the compressor are described. The analysis of the results shows slightly higher pressure rise coefficients for shrouded stators, but slightly higher stall margin in the cantilevered case. This is due to a stabilizing effect of the hub clearance vortex (cleans up separation on hub) in the cantilevered configuration because its direction is opposite to the secondary flow in the passage.

Different active flow control techniques have been investigated in a 1.5-stage axial-flow compressor. Looking at a low-speed single-stage environment, many researchers have shown that highly loaded compressors are tip critical, showing stall inception caused by short length scale disturbances (spikes). It has been shown by several authors that these disturbances are related to the spillage of endwall flow ahead of the blading (spill forward). For the present work, different tip injection configurations were investigated in order to stabilize the near casing flow, increasing the operating range of the compressor. Stall margin improvement and the impact on stage efficiency are compared and discussed. Oil flow pictures of the casing wall above the rotor and of the stator blades as well as traverse data from pneumatic 5-hole probes show the impact of flow control on rotor and stator performance. Another method of energizing the casing wall boundary layer is the removal of low energy fluid by a circumferential slot above the rotor, which was also studied experimentally. Again, the impact on compressor operating range and efficiency, as well as flow field information collected by oil flow visualization and traverse data are discussed. Comparing the different flow control techniques, it is shown that increasing stall margin is not directly linked to stage efficiency. As described in various publications, discrete tip injection is a very powerful technique as far as range extension is concerned, but it also has substantial drawbacks such as the circumferential inhomogeneity of the rotor exit flow. These inhomogeneities may result in poor stator performance, overall resulting in a drop of stage efficiency. This problem does not occur if circumferential boundary layer removal above the rotor is used. This method however shows much less potential for increasing the operating range.

This paper introduces an approach for considering manufacturing variability leading to a nonaxisymmetric blading in the computational fluid dynamics simulation of a high-pressure compressor stage. A set of 150 rotor blades from a high-pressure compressor stage was 3D scanned in order to obtain the manufacturing variability. The obtained point clouds were parameterized using a parametric blade model, which uses typical profile parameters to translate the geometric variability into a numerical model. Probabilistic simulation methods allow for the generation of a sampled set of blades that statistically corresponds to the measured one. This technique was applied to generate 4000 sampled blades in order to investigate the influence of a nonaxisymmetric blading. It was found that the aerodynamic performance is considerably influenced by a variation of the passage cross section. Nevertheless, this influence decreases with an increasing number of independently sampled blades and, thus, independently shaped passage cross sections. Due to its more accurate consideration of the geometric variability, the presented methodology allows for a more realistic performance analysis of a high-pressure compressor stage.