Matthew A. Bennington

The Influence of Tip Clearance Momentum Flux on Stall Inception in a High-Speed Axial Compressor

Experimental and numerical studies were conducted to investigate tip-leakage flow and its relationship to stall in a transonic axial compressor. The computational fluid dynamics (CFD) results were used to identify the existence of an interface between the approach flow and the tip-leakage flow. The experiments used a surface-streaking visualization method to identify the time-averaged location of this interface as a line of zero axial shear stress at the casing. The axial position of this line, denoted x(zs), moved upstream with decreasing flow coefficient in both the experiments and computations. The line was consistently located at the rotor leading edge plane at the stalling flow coefficient, regardless of inflow boundary condition. These results were successfully modeled using a control volume approach that balanced the reverse axial momentum flux of the tip-leakage flow with the momentum flux of the approach fluid. Nonuniform tip clearance measurements demonstrated that movement of the interface upstream of the rotor leading edge plane leads to the generation of short length scale rotating disturbances. Therefore, stall was interpreted as a critical point in the momentum flux balance of the approach flow and the reverse axial momentum flux of the tip-leakage flow.

A Computational Fluid Dynamics Study of Circumferential Groove Casing Treatment in a Transonic Axial Compressor

Numerical investigations were conducted to predict the performance of a transonic axial compressor rotor with circumferential groove casing treatment. The Notre Dame Transonic Axial Compressor (ND-TAC) was simulated at Tsinghua University with an in-house computational fluid dynamics (CFD) code (NSAWET) for this work. Experimental data from the ND-TAC were used to define the geometry, boundary conditions, and data sampling method for the numerical simulation. These efforts, combined with several unique simulation approaches, such as nonmatched grid boundary technology to treat the periodic boundaries and interfaces between groove grids and the passage grid, resulted in good agreement between the numerical and experimental results for overall compressor performance and radial profiles of exit total pressure. Efforts were made to study blade level flow mechanisms to determine how the casing treatment impacts the compressors stall margin and performance. The flow structures in the passage, the tip gap, and the grooves as well as their mutual interactions were plotted and analyzed. The flow and momentum transport across the tip gap in the smooth wall and the casing treatment configurations were quantitatively compared.

A CFD Study of Circumferential Groove Casing Treatments in a Transonic Axial Compressor

Numerical investigations were conducted to predict the performance of a transonic axial compressor rotor with circumferential groove casing treatment. The Notre Dame Transonic Axial Compressor (ND-TAC) was simulated by Tsinghua University with an in-house CFD code (NSAWET) for this work. Experimental data from the ND-TAC were used to define the geometry, boundary conditions and data sampling method for the numerical simulation. These efforts, combined with several unique simulation approaches, such as non-matched grid boundary technology to treat the periodic boundaries and interfaces between groove grids and the passage grid, resulted in good agreement between the numerical and experimental results for overall compressor performance and radial profiles of exit total pressure. Efforts were made to study blade level flow mechanisms to determine how the casing treatment impacts the compressor’s stall margin and performance. The flow structures in the passage, the tip gap and the grooves as well as their mutual interactions were plotted and analyzed. The flow and momentum transport across the tip gap in the smooth wall and the casing treatment configurations were quantitatively compared.Copyright

An Experimental and Computational Investigation of Tip Clearance Flow and Its Impact on Stall Inception

A numerical and experimental study was conducted to investigate the tip clearance flow and its relationship to stall in a transonic axial compressor. The CFD results were used to identify the existence of an interface between incoming axial flow and the reverse tip clearance flow. A surface streaking method was used to experimentally identify this interface as a line of zero axial shear stress at the casing. The position of this line, denoted xzs , moved upstream with decreasing flow coefficient in both the experiments and computations. The line was found to be at the rotor leading edge plane when the compressor stalled. Further measurements using rotor offset and inlet distortion further corroborated these results, and demonstrated that the movement of the interface upstream of the leading edge leads to the generation of rotating (“spike”) disturbances. Stall was therefore interpreted to occur as a result of a critical momentum balance between the approach fluid and the tip-leakage flow.Copyright

Effects of Steady Tip Clearance Asymmetry and Rotor Whirl on Stall Inception in an Axial Compressor

Effects of rotor centerline offset and whirl on the pre-stall and stall inception behavior of a high-speed tip-critical axial compressor were investigated. The observations were made using a circumferential array of unsteady pressure transducers. The maximum amount of rotor offset and whirl used in this investigation was 26% and 13% of the design axisymmetric tip clearance respectively. Measurements were conducted using transient throttle movements which quickly decreased the mass flow in the compressor until the onset of rotating stall. A second set of measurements used quasi-transient throttling starting from a mass flow about 0.5% larger than the stalling mass flow. These data were analyzed with the traveling wave energy method, visual inspection of the filtered pressure traces, and a two-point spatial correlation technique. For the uniform tip clearance case rotating stall occurred while the slope of the pressure rise characteristic was negative. As expected, the flow breakdown exhibited “spike” inception with no observable rotating disturbances in the pre-stall time period. The introduction of small levels of steady and unsteady tip clearance asymmetry did not significantly alter the time average performance of the stage; circumferential variations in pressure rise and flow coefficient were minimal and the stalling flow coefficient remained unchanged. However, significant short length-scale rotating disturbances were observed in both of these cases prior to stall inception. As in the symmetric tip clearance case, short length-scale disturbances initiated rotating stall in the non-uniform tip clearance experiments. The location of the generation of the incipient stall cells with respect to the non-uniform tip clearance was strongly effected by the rotor offset/whirl.Copyright

Investigation of Tip-Flow Based Stall Criteria Using Rotor Casing Visualization

Recent stall inception investigations have indicated that short-length scale stall initiates when the interface between the tip gap flow and the approach flow spills forward of the leading edge of the adjacent blade. This hypothesis was investigated in the present work using both numerical and experimental results from a range of compressor geometries and speed. First, full annulus unsteady computations of R35 were used to generate contours of entropy at the casing. It was found that a large gradient in entropy, which marked the leakage fluid, was aligned with the leading edge plane at the stalling mass flow. It was also observed that the flow direction in the region of increased entropy was in the reverse axial direction. The interface between the approach fluid and the reverse-direction leakage flow was related to a region in which the axial component of the wall shear stress was zero. The axial location of this line was measured experimentally using a surface streaking method using two separate facilities. It was found that the location of this line is determined by a momentum balance between the approach fluid and the tip leakage fluid. Measurements were acquired with varied tip clearance, radial distortion, and centerline offset to support these conclusions. In all cases the zero axial shear line was found to move upstream with decreased flow coefficient, and was in close proximity to the rotor leading edge at the stalling mass flow.Copyright

Over Rotor Casing Surface Streak Measurements in a High Speed Axial Compressor

Time averaged wall shear stress patterns were recorded during quasi-steady throttling to stall in a high speed compressor. The technique utilized a CCD camera to capture digital images of oil streaks on a transparent casing section located over the rotor. The most notable feature of the surface streaking was a bifurcation line of zero time average axial shear stress. The location of this feature was found to represent the location where the approach fluid and the reverse flow from the tip gap meet and separate from the casing surface. The location of this line with respect to the rotor leading edge was denoted as xzs . The values of xzs were found to be positive (downstream of the leading edge) at high flow coefficients, and moved upstream as the compressor mass flow was reduced. Compressor stall was observed to occur when xzs was negative, with magnitude of order 6% of the axial blade tip chord. In other words, the zero axial shear line crossed the leading edge plane at a flow coefficient slightly higher than the stall point. The present paper describes the location of xzs as a function of both the flow coefficient and the local blade tip clearance. Both of these independent variables were found to have a substantial impact on the endwall flow near the leading edge, with little variation downstream. A simplified model was used to better understand the flow mechanisms associated with changes in xzs . An interpretation of these results will be given in terms of experimental and computational efforts related to blade tip flows that are described in the recent literature.Copyright

Blade image velocimetry: development and uncertainty analysis

This paper describes the development of a non-contact structural vibration measurement technique referred to as blade image velocimetry (BIV). The technique utilizes a commercially available single camera PIV system and, when used with PIV, can measure fluid and structure velocity simultaneously. The utility of this measurement technique was demonstrated on a flat plate undergoing flutter. The structural velocity measurements were experimentally demonstrated to be of comparable accuracy to standard PIV measurements. The theory relating the measurements obtained by BIV of the structure is presented. An uncertainty analysis was developed by considering two sources of error associated with the modal amplitude estimates. Through benchtop tests, it was found that these two sources of error predict the experimentally observed bias errors well.

Surge Margin Extension and Circumferential Groove Interaction with the Rotor Tip Clearance Flow Field

Experimental and computational studies were conducted to study the role of the tip leakage flow in axial compressor stall and the relationship between the tip clearance flow field and surge margin extension from circumferential groove casing treatment (CGCT). Experimental measurements of surge margin extension from seven CGCT configurations with a fixed groove geometry demonstrated that the contribution of individual grooves in a multi-groove casing to surge margin extension is an (a) additive and (b) linear function of the smooth wall tip clearance axial momentum flux at the location of a each groove. Cameron et al.’s 2013 axial momentum model of the rotor tip clearance was extended to include the influence of a CGCT. Control volume analysis of this model showed that circumferential grooves reduce the tip leakage flow axial momentum through radial transport. The equivalent force due to a circumferential groove was demonstrated to be related to the smooth wall tip clearance axial momentum flux through a coefficient of drag that had a log-linear dependence on groove aspect ratio. The paper closes with a demonstration of how findings from the experiments described above can combine with the one-dimensional momentum balance to make first-order estimates of surge margin extension from a given circumferential groove casing treatment.

Performance and Short Length-Scale Disturbance Generation in an Axial Compressor With Non-Uniform Tip Clearance

The performance, efficiency, and stall inception of an axial compressor was investigated experimentally with small levels of rotor centerline offset. The measurements were acquired using a high-speed, single-stage compressor. The rotor was levitated magnetically during operation which allowed precise positioning of the rotor centerline within the circular casing. The offset magnitude used in this study was 0.23% of the rotor tip chord, equivalent to approximately 24% of the nominal gap value. The resulting asymmetry in the tip gap resulted in circumferential and radial variations in the measured stagnation pressure and stagnation temperature downstream of the stage. However, the spatially averaged performance of the compressor was not measurably different from that obtained with a concentric rotor. An array of unsteady (Kulite) pressure transducers was used to investigate the flow field during stall inception. These measurements were recorded during transient throttle movements which quickly decreased the mass flow in the compressor until the onset of rotating stall. A second set of measurements was acquired during quasi-transient throttling starting from a mass flow about 1% larger than the stalling mass flow. In both the symmetric and offset cases the flow breakdown was consistent with spike type inception. The measurements with offset indicated that the asymmetries in the local compressor flow field produced significant changes in the number of short-length scale rotating disturbances observed during throttling to stall. These disturbances appeared in the region of the annulus where the local flow coefficient was lowest and usually decayed upon rotating to the higher flow region. In this way, the addition of very small amounts of rotor offset tended to fix the disturbance generation location in the stationary reference frame. This was in contrast to the symmetric tip clearance case where the location of spike generation appeared stochastic.Copyright