Axial compressor

About: Axial compressor is a research topic. Over the lifetime, 12035 publications have been published within this topic receiving 127766 citations.

Numerical Investigation of Three-Dimensional Separation in an Axial Flow Compressor: The Influence of Free-Stream Turbulence Intensity and Endwall Boundary Layer State

Abstract: Regions of three-dimensional separations are an inherent flow feature of the suction surface endwall corner in axial compressors. These corner separations can cause a significant total pressure loss and reduce the compressor’s efficiency. This paper uses wall-resolved LES to investigate the loss sources in a corner separation, and examines the influence of the inflow turbulence on these sources. Different subgrid scale (SGS) models are tested and the choice of model is found to be important. The σ SGS model, which performed well, is then used to perform LES of a compressor endwall flow. The time-averaged data is in good agreement with measurements. The viscous and turbulent dissipation are used to highlight the sources of loss, with the latter being dominant. The key loss sources are seen to be the 2D laminar separation bubble and trailing edge wake, and the 3D flow region near the endwall. Increasing the free-stream turbulence intensity (FST) changes the suction surface boundary layer transition mode from separation induced to bypass. However, it doesn’t significantly alter the transition location and therefore the corner separation size. Additionally, the FST doesn’t noticeably interact with the corner separation itself, meaning that in this case the corner separation is relatively insensitive to the FST. The endwall boundary layer state is found to be significant. A laminar endwall boundary layer separates much earlier leading to a larger passage vortex. This significantly alters the endwall flow and loss. Hence, the need for accurate boundary measurements is clear. ∗Address all correspondence to this author (as2341@cam.ac.uk). INTRODUCTION Over-turning of the endwall boundary layer causes a threedimensional separation to form in the corner formed by the suction surface and endwall of axial compressors. A number of studies have discussed the importance of these corner separations in both stator and rotor blades [1,2]. They can cause passage blockage and effectively limit the loading and static pressure rise achievable by the compressor. Additionally, they may cause a significant total pressure and a reduction in the compressor’s efficiency. Traditionally, the size of three-dimensional separations has been correlated to global parameters such as inlet and exit flow angles, and pitch to chord ratio. An example of this is the endwall diffusion parameter and corner stall metric proposed by Lei et al. [3]. These are useful in the early design stage, but to maximise the compressor efficiency it is important to consider the 3D separation in more detail. For example: Goodhand and Miller [4] examine the sensitivity of the 3D separation to small leading edge geometry features, while Gbadebo et al. [5] study the influence of surface roughness on the 3D separation. Both of these studies conclude that any process leading to premature boundary layer transition on the early suction surface, near the endwalls, will dramatically increase the size of the 3D separation. This is due to the suction surface boundary layer being excessively thickened. Premature suction surface boundary layer transition may be caused by leading edge geometry or surface roughness, however it could also be caused by incoming free-stream turbulence (FST). Zaki et al. [6] use direct numerical simulation (DNS) to study the influence of FST intensity on the transitional processes on a compressor blade. They find that the mode and location of the boundary layer 1 Copyright c © 2016 by Rolls-Royce plc transition is very sensitive to the FST intensity. This was at a Reynolds number of Rec=0.14×10, and this sensitivity would be expected to decrease at higher Reynolds numbers, with the boundary layers eventually becoming fully turbulent. However, Steinert and Starken [7] showed experimentally that at Rec = 0.84 × 10 and Ti = 2.5% the suction surface boundary layer stayed laminar to peak suction over a wide range of incidences. At cruise, Reynolds numbers between 0.4×106 and 1.6×106 are seen in aero-gas turbine compressors [4]. Therefore the FST may have an important effect on the 3D corner separation in compressors. With the above in mind, this paper presents a series of numerical investigations intended to address the following: a) The effect of the FST intensity on the suction surface boundary layer, and therefore the 3D corner separation, is examined. The existence of any direct interactions between the FST and corner separation is also considered. b) Gbadebo [8] found that the 3D separation is also sensitive to the thickness of the incoming endwall boundary layer. Taking this further this paper examines whether the endwall boundary layer state (i.e. laminar vs turbulent) is important. c) Denton [9] investigated the loss sources in the 3D separation region using the entropy generation rate. In a similar manner this paper uses the viscous and turbulent dissipation to identify the loss sources. In particular the effect of the inflow conditions on these loss sources will be studied. Large eddy simulation (LES) will be used for the numerical investigations outlined above. Lardeau et al. [10] found that LES can successfully predict the transitional processes occurring in a compressor flow, at a fraction of the cost of a DNS. However, the sub-grid scale (SGS) models used (MTS and dynamic Smagorinsky) require additional filtering which is problematic for the 3D endwall geometry considered here. To investigate whether such advanced SGS models are necessary a number of purely local SGS models are first tested on a simpler quasi-2D blade geometry. FLOW CONFIGURATIONS The two linear compressor blade cascades detailed in Tab. 1 are simulated in this paper. Both are representative of highly loaded compressor stator blades found in a modern gas-turbine compressor. Cascade 1 consists of NACA-65 aerofoils and was tested experimentally by Hilgenfeld and Pfitzner [11]. This cascade was also simulated at a lower Rec using DNS by Zaki et al. [6] and using LES by Lardeau et al. [10]. Cascade 2 is a linear CDA (Controlled Diffusion Aerofoil) cascade investigated experimentally by Gbadebo et al. [8,12]. The computational grid for cascade 2 is displayed in Fig. 1. A similar H-O-H topology is used for cascade 1. Downstream of the blade a sponge zone is used to prevent reflections from the outflow boundary. Pitchwise periodicity is enforced with periodic boundaries at mid-pitch. Some of the cases run are spanwise periodic (i.e. no TABLE 1: GEOMETRICAL AND INFLOW PARAMETERS FOR THE TWO COMPRESSOR CASCADES.

Measurements of the Tip Clearance Flow for a High-Reynolds-Number Axial-Flow Rotor

Abstract: A high-Reynolds-number pump (HIREP) facility has been used to acquire flow measurements in the rotor blade tip clearance region, with blade chord Reynolds numbers of 3,900,000 and 5,500,000. The initial experiment involved rotor blades with varying tip clearances, while a second experiment involved a more detailed investigation of a rotor blade row with a single tip clearance. The flow visualization on the blade surface and within the flow field indicate the existence of a trailing-edge separation vortex, a vortex that migrates radially upward along the trailing edge and then turns in the circumferential direction near the casing, moving in the opposite direction of blade rotation. Flow visualization also helps in establishing the trajectory of the tip leakage vortex core and shows the unsteadiness of the vortex. Detailed measurements show the effects of tip clearance size and downstream distance on the structure of the rotor tip leakage vortex. The character of the velocity profile along the vortex core changes from a jetlike profile to a wakelike profile as the tip clearance becomes smaller. Also, for small clearances, the presence and proximity of the casing endwall affects the roll-up, shape, dissipation, and unsteadiness of the tip leakage vortex. Measurements also show how much circulation is retained by the blade tip and how much is shed into the vortex, a vortex associated with high losses.

Three Dimensional Characteristics of Turbulent Wakes Behind Rotors of Axial Flow Turbomachinery

Abstract: Analytical and experimental investigations of the characteristics of three-dimensional turbulent wakes downstream of a turbomachinery rotor are reported in this paper. An approximate quasi-three-dimensional turbulent wake model for turbomachinery rotor is developed and compared with the cascade and isolated airfoil wake models. The rotor wake model is capable of predicting the decay of mean component of radial and streamwise velocities as a function of rotor geometry, speed of rotation, and the turbulence properties of the flow field. A rotation parameter based on similarity analysis is derived. The velocity profiles in both the radial and cylindrical planes are coupled together. Measurement of mean velocities (Un , Us , Ur ), turbulent intensities and stresses (un2, us2, ur2, usun, unur, usur) is carried out using a triple sensor hot wire probe in a stationary system at various axial and radial locations downstream of the rotor. Profiles of mean and turbulent quantities are obtained. Semi-theoretical expressions for the decay rates of the defect in mean velocity, turbulence intensity, and Reynolds stress (maximum values) with distance downstream of the rotor are derived. The experimental data on the rotor wake are compared with that of an isolated airfoil and cascade of airfoils. The investigation suggests that rotor wake decays much faster than the cascade or an isolated airfoil wake.

Reduced drag airfoil platforms

Abstract: A flow directing apparatus for use in an axial flow turbomachine is disclosed. Techniques for reducing aerodynamic drag along the walls of the flow directing apparatus are developed. In one embodiment, rotor blades have multiplanar platform surfaces which reduce aerodynamic drag pressure losses at the interface between each blade platform and the adjacent structure.