Measurement Errors Downstream of a Rotor due to Probe–Blade Row Interaction

This paper presents a study of the three-dimensional e owe eld within the blade rows of a single-stage highpressure axial turbine (low-speed, large-scale ). Measurements have been performed in the stationary and rotating frames of reference. Time-mean data have been obtained using e ve-hole pneumatic probes. The transport mechanisms of the stator wake and passage vortices through the rotor blade row have been studied using smoke e ow visualization. Furthermore, unsteady measurements have been carried out using a three-axis hot wire. Steady and unsteady numerical simulations have been performed using a structured three-dimensional Navier ‐Stokes solver to further understand the blade-row interactions. The transport of the stator viscous e ow through the rotor blade-row is described. The rotor passage vortices are affected by the transport of the stator secondary e ow. It is observed that the stator secondary e ow vortices are convected through the downstream rotor blade-row in a similar but not identical way to the wake. At the hub the kinematic interaction between the stator and the rotor passage vorticeshas two effects. First,thesuction sideleg of the statorpassagevortex isdisplaced radially upwards over the developing rotor passage vortex at the hub. Additionally, the pressure side leg of the stator passage vortex is entrained into the rotor passage vortex. The predicted e owe eld was interrogated from the perspective of loss production. The contribution of the unsteady e ow to the stage loss has been evaluated using unsteady numerical simulations. The effect of stator viscous e ow transport on the rotor e ow angles is also discussed in brief. Finally, a kinematic model is proposed for the transport of the secondary-e ow vortices in the downstream blade-row based on the understanding obtained from the measurements and numerical simulations.

This two-part paper presents detailed experimental investigations of unsteady aerodynamic blade row interactions in the four-stage Low-Speed Research Compressor of Dresden. In part I of the paper the unsteady profile pressure distributions for the nominal setup of the compressor are discussed. Furthermore the effect of blade row clocking on the unsteady profile pressures is investigated. Part II deals with the unsteady aerodynamic blade forces, which are calculated from the measured profile pressure distributions. The unsteady pressure distributions were analysed in the first, a middle and the last compressor stage both on the rotor and stator blades. The measurements were carried out on pressure side and suction side at midspan. Several operating points were investigated. A complex behaviour of the unsteady profile pressures can be observed, resulting from the superimposed influences of the wakes and the potential effects of several up- and downstream blade rows of the four-stage compressor. The profile pressure changes nearly simultaneously along the blade chord if a disturbance arrives at the leading edge or the trailing edge of the blade. Thus the unsteady profile pressure distribution is nearly independent of the convective wake propagation within the blade passage. A phase shift of the reaction of the blade to the disturbance on the pressure and suction side is observed. In addition clocking investigations were carried out to distinguish between the different periodic influences from the surrounding blade rows. For this reason the unsteady profile pressure distribution on rotor 3 was measured, while stator 1–4 were separately traversed stepwise in the circumferential direction. Thus the wake and potential effects of the up- and downstream blade rows on the unsteady profile pressure could clearly be distinguished and quantified.

This paper presents a study of the three-dimensional flow field within the blade rows of a high-pressure axial flow steam turbine stage. Compound lean angles have been employed to achieve relatively low blade loading for hub and tip sections and so reduce the secondary losses. The flow field is investigated in a low-speed research turbine using pneumatic and hot-wire probes downstream of the blade row. Steady and unsteady numerical simulations were performed using structured 3-D Navier-Stokes solver to further understand the flow field. Agreement between the simulations and the measurements has been found. The unsteady measurements indicate that there is a significant effect of the stator flow interaction in the downstream rotor blade. The transport of the stator viscous flow through the rotor blade row is described. Unsteady numerical simulations were found to be successful in predicting accurately the flow near the secondary flow interaction regions compared to steady simulations. A method to calculate the unsteady loss generated inside the blade row was developed from the unsteady numerical simulations. The contribution of various regions in the blade to the unsteady loss generation was evaluated. This method can assist the designer in identifying and optimizing the features of the flow that are responsible for the majority of the unsteady loss production. An analytical model was developed to quantify this effect for the vortex transport inside the downstream blade.

Self-recirculating injection, which bleeds air from the downstream duct of the last blade row and injects air as a wall jet upstream of the first rotor blade row, is experimentally investigated after the design of its structure in single- and three-stage axial flow compressors. External injection and outlet bleed air are selected for comparison. Results show that self-recirculating injection can improve the stall margin by 13.67% and 13% on the premise of no efficiency penalty in single- and three-stage axial flow compressors with only 0.7% and 4.2% of the total injected momentum ratio recirculated near stall, respectively. The self-recirculating injection is the best among all the three cases if the influence on pressure rise coefficient and efficiency is comprehensively considered. Moreover, findings indicate that the upstream injection plays an important role in terms of stability-enhancement. The details of the flow field are captured using a collection of pressure transducers on the casing with circumferential and chordwise spatial resolution. A detailed comparative analysis of the endwall flow indicates that the self-recirculating injection can postpone the occurrence of stalling in the proposed compressor by delaying the forward movement of the interface between the tip leakage flow (TLF) and main stream flow (MF), weakening the unsteadiness of TLF (UTLF), and sharply decreasing the circumferentially propagating speed dominated by the UTLF that triggers the spike-type stall inception. Finally, the stall control concept on the stage that first generates stall inception using self-recirculating injection is proposed. This study helps to guide the design of self-recirculating injection in actual application.

This paper is conceived as an extension in the analysis of the periodic stator-rotor interaction in a low-speed axial fan with inlet guide vanes. Here, the present work focuses on the deterministic fluctuations that occur in the axial gap between the blade rows. In particular, we present experimental data on the phase averaged velocity in the stator frame of reference. Detailed measurements of the flow field were obtained using hot-wire anemometry in a traverse sector that covered the whole span of the stage for a complete stator pitch. The blade-to-blade velocity gradient in the rotor passage is observed as an unsteady flow in the stator frame of reference due to the relative motion of the blades. As a consequence, this periodic fluctuation is superimposed on the steady vane-to-vane velocity distribution, resulting in a nonuniform unsteadiness with additional phase-dependent wake-blockage interaction. This phase-dependent interaction is determined by the difference between the overall deterministic fluctuation and the rotating spatial blade-to-blade distribution, when the latter is accurately transformed to the stator reference frame. The results revealed that high unsteadiness in the tip region is mainly derived from the radial increase of the blade-to-blade nonuniformities in the rotor, whereas the wake-blockage interaction exhibits a spanwise uniform distribution. Hence, the hub-to-tip torsion of the blades is responsible for setting off a major spatial distortion between the rows. Complementarily, we observed that a reduction in the rows’ spacing or an increase in the blade loadings leads to a higher wake-blockage interaction, modifying the impingement of the incoming stator wakes. In addition, the deterministic stresses were calculated by time averaging the phase-dependent velocity correlations and compared to the turbulent stresses. The deterministic stresses were also segregated in temporal, spatial, and spatial-temporal correlations in order to analyze the dominant mechanisms involved in their generation. At this point, the deterministic kinetic energy levels were found to be generally lower than turbulent kinetic energy levels, with both temporal and spatial correlations being the most significant terms of the tensor.

A new stall model to describe rotating stall in axial-flow compressors is established, where blade rows are replaced by semiactuator disks, flow fluctuations are permitted to be finite and nonsteady, and blade row characteristics are taken into consideration in nonlinear and nonsteady forms. Through numerical analysis using the finite difference method, it is attempted to make clear the aspects of rotating stall—such as number of stall cells, stall propagation velocity, wave shape and magnitude of disturbance and their variations with flow rate through the compressor, and the mechanisms which control them. Most of the aims are achieved by taking into consideration the effects of blade row interference and the nonlinearities of blade row characteristics.

The present thesis documents experimental measurements of a low-speed linear turbine cascade. The overall objective of the thesis was to enhance the understanding of turbine near endwall flows, termed secondary flows. Of particular interest were the loss generation and energy dissipation mechanisms present within the turbine blade row. Point measurements of the downstream pressure fields for different cascade configurations were made using a seven-hole pneumatic pressure probe. The generation of total pressure losses were examined in relation to the inferred secondary flow structures. Also, the corresponding turbulence field was measured for one of the cascade configurations using a hot-wire probe thereby providing insights into the role of the turbulence field in the generation of total pressure losses. The interpretation of the flow fields were aided with experimental oil film flow visualization and corresponding CFD simulations. In a real turbine engine there is an axial gap between the stationary and rotating bladerows. The term rim-seal refers to the geometry of the axial gap and the cooling flow, termed purge flow, which is typically injected into the mainstream from within the gap to prevent ingress of the hot mainstream gases. The present work examines the effects of the rim-seal geometry on the secondary flow and associated losses for different levels of purge flow. It was found that the presence of the rim-seal geometry significantly alters the secondary flow compared to that of a traditional flat endwalled turbine cascade. The addition of purge flow further enhances the secondary flow and the bladerow losses increase with increasing purge flow. It was therefore recommended that modern turbine designs consider both the rim-seal geometry and the purge flow during the design process. Previous studies have shown that a secondary loss mitigation technology, termed non-axisymmetric endwall contouring, significantly reduces the endwall losses generated through a turbine bladerow. The present work investigated a computationally optimized endwall design for the present turbine cascade for varying levels of purge injection. The experimental measurements found that the endwall contouring, with the upstream rim-seal geometry, was effective at reducing the bladerow losses for the range of positive net purge flows investigated.

A computationally efficient time-accurate vortex method for unsteady incompressible flows through multiple blade row systems is presented. The method represents the boundary surfaces using vortex systems. A local coordinate system is assigned to each independently moving blade row. Blade shed vorticity is determined from two generating mechanisms and convected using the Euler equation. The first mechanism of vorticity generation is a potential mechanism from a nonlinear unsteady pressure-type Kutta condition applied at the blade trailing edges. The second mechanism is a viscous mechanism from a viscous wake vorticity (VWV) model implemented to simulate the viscous shear layers on the blade pressure and suction sides. Two different two-blade-row compressor systems, a rotor/stator (R/S) system and a stator/rotor (S/R) system, were used to investigate the interaction forces on each blade row. Computational results of the potential and viscous interaction forces are presented and compared to measurements. The comparison suggests that the viscous wake interaction accounts for 25–30% of the peak loading for an axial spacing of 10% chord length between the blade rows. The efficient computational method is particularly attractive for blade indexing study. Therefore a three-blade-row rotor/stator/rotor (R1/S/R2) compressor system is used to demonstrate the indexing calculations between the two rotor positions. Resultant forces on each blade row are presented for ten rotor indexing positions and three axial gap sizes for the gaps between R1 and S and between S and R2. The unsteady peak-to-peak force can reach 10–15% of inflow dynamic head for the gap spacing investigated. The minimum-to-maximum variation of the unsteady force can account for 40–50% of averaged unsteady force.

A computationally efficient time-accurate vortex method for unsteady incompressible flows through multiple blade row systems is presented. The method represents the boundary surfaces using vortex systems. A local coordinate system is assigned to each independently moving blade row. Blade shed vorticity is determined from two generating mechanisms and convected using the Euler equation. The first mechanism of vorticity generation is a potential mechanism from a nonlinear unsteady pressure-type Kutta condition applied at the blade trailing edges. The second mechanism is a viscous mechanism from a viscous wake vorticity (VWV) model implemented to simulate the viscous shear layers on the blade pressure and suction sides. Two different two-blade-row compressor systems, a rotor/stator (R/S) system and a stator/rotor (S/R) system, were used to investigate the interaction forces on each blade row. Computational results of the potential and viscous interaction forces are presented and compared to measurements. The comparison suggests that the viscous wake interaction accounts for 25–30% of the peak loading for an axial spacing of 10% chord length between the blade rows. The efficient computational method is particularly attractive for blade indexing study. Therefore a three-blade-row rotor/stator/rotor (R1/S/R2) compressor system is used to demonstrate the indexing calculations between the two rotor positions. Resultant forces on each blade row are presented for ten rotor indexing positions and three axial gap sizes for the gaps between R1 and S and between S and R2. The unsteady peak-to-peak force can reach 10–15% of inflow dynamic head for the gap spacing investigated. The minimum-to-maximum variation of the unsteady force can account for 40–50% of averaged unsteady force.

The current paper describes the experimental determination of a complete four-quadrant map for a three stage low-speed axial flow compressor. Non-dimensional pressure rise and torque coefficients and efficiency characteristics were determined. Six possible modes of operation were identified in the four quadrants. The characteristic for a compressor with a rotational speed of zero rpm, in terms of pressure rise coefficient and flow coefficient, is S-shaped and forms the dividing line between two alternate modes of operation which exist in the second and fourth quadrants. First quadrant operation is only possible during positive rotation, and third quadrant operation during negative rotation, as the zero-speed S-curve passes through the origin and so does not pass through these quadrants. Consequently, only one mode of operation is possible in each of these two quadrants. A system of notation for describing the various modes of operation has also been developed. Second quadrant operation for positive compressor rotation, and fourth quadrant operation with negative rotation are highly dissipative conditions of operation. First and third quadrant operation are compressor-like modes. Second quadrant negative rotation and fourth quadrant positive rotation operation are turbine-like modes.

This paper focuses on optimizing the blade design of miniature centrifugal flow fans for application in processor cooling solutions of portable power electronics. The design parameter varied is the blade chord length and the resulting fan performance is gauged by examining flow rate, pressure rise and power consumption characteristics. The former two of these are measured using a BS848 fan characterization rig and the latter by directly measuring the power consumed. These characteristics are studied for three sets of scaled fans with diameters of 15mm, 24mm and 30mm, and each set considers six individual blade chord lengths. A novel theory is put forward to explain the anticipated effect of changing this parameter and the results are analyzed in terms of the relevant dimensionless parameters: Reynolds number; chord length to diameter of fan ratio; flow coefficient; pressure coefficient and power coefficient. When these characteristic parameters are plotted against Reynolds number, similar trends are observed as the chord length is varied in all sets of scaled fans. The results show that the flow coefficient for all the miniature fans degrade at low Re values but the onset of this degradation was observed at higher Re values for longer blade chord designs. Conversely, it was found that the pressure coefficient is elevated at low Re and the onset Re for this phenomenon correlates well with the drop off in flow coefficient. Finally, the trend in power coefficient data appears to be identical to that for the flow coefficient. The derived theory is used to correlate this data for which all data points fall within 6% of the correlation. Overall, the findings reported herein provide a good understanding of how changing the blade chord length affects the performance of miniature centrifugal fans and provides guidelines for designers to aid in selecting the optimum fan design for a specific application.

Detailed numerical comparisons of pressure and flow angle measurements made in stage three of a four-stage, large scale, low speed, axial compressor are presented. The measurements are in both the rotating and stationary frame and were obtained as part of a BRITE/EURAM collaborative study of cantilevered and shrouded-stator compressor configurations. The numerical analysis is 3D, considers three blade rows simultaneously and incorporates multiple row effects by use of a conservative mixing-plane model allowing circumferential variation at the mixing plane.The paper discusses the early results of a study sponsored by Alstom Gas Turbines to examine steady-state, multiple blade-row modelling techniques. Growth in the endwall flow region due to multi-row effects is revealed from both the numerical and experimental results. The numerical simulation is conducted without altering blade gap spacings to assist numerical stability; the axial gap is increasingly being seen as a critical performance parameter for multiple row analysis. The limitations inherent in an approach using mixing-planes are presented and a review of alternative, more rigourous, treatments of these effects is then discussed. These treatments attempt to retain the unsteady flow structure in a steady-state model by the derivation of so-called deterministic stresses.

The performance of a single-stage, axial-flow turbomachine was studied experimentally at low Reynolds numbers. The study was made with a turbomachine modeled from a large jet-engine type of axial-flow compressor. Low Reynolds numbers were obtained by using a mixture of glycerine and water as the working fluid. The overall performance was determined over a range of Reynolds numbers RT (based on rotor-tip speed and rotor chord) from 2000 to 150,000. The flow rate at each Reynolds number was varied from near shutoff to the maximum permitted by the turbomachine-tunnel systems. Blade-row characteristics were studied by means of quantitative flow surveys before and after each blade row, and by means of extensive flow-visualization experiments within each blade row. The investigation established that sudden or critical changes in performance do not occur in the type of machine tested, between RT of 150,000 and 20,000. Below 20,000 the performance deteriorated more rapidly. A relatively sharp change in performance occurred between RT of 20,000 and 10,000. The results clarified many of the viscous flow details in each blade row which are associated with the deterioration of performance. These effects were very pronounced at RT of 4000 and below. Consequently, a considerable part of the paper is concerned with results obtained at these lower Reynolds numbers. From the point of view of a designer, information is presented in regard to overall performance, guide-vane turning, and guide-vane and stator total-pressure losses, all as functions of Reynolds number. These results are expected to be indicative of performance in turbomachines similar to the one tested here. Other details are concerned with problems such as wall boundary layers, flow reversal at low flow coefficients, lip-clearance flow, flow patterns near shutoff, and flow comparisons in stators with rotating and stationary hubs.

A 3D numerical simulation was conducted to study the effect of inlet boundary layer thickness on rotating stall in an axial compressor. The inlet boundary layer thickness had significant effects on the hub-corner-separation at the corner of hub and suction surfaces. The hub-corner-separation grew considerably for a thick inlet boundary layer as the load increased, while it diminished to become indistinguishable from the rotor wake for a thin inlet boundary layer and another corner-separation originated near the casing. This difference in the internal flow near stall also had a large effect on characteristics of the rotating stall, especially the initial asymmetric disturbance and the size of stall cells. While a pre-stall disturbance arises firstly in the hub-corner-separation for the thick inlet boundary layer, an asymmetric disturbance was initially generated in the tip region because of the corner-separation for the thin inlet boundary layer. This disturbance was transferred to the tip leakage flow and grew to become an attached stall cell, which adheres to the blade passage and rotates at the same speed as the rotor. When this attached stall cell reached a critical size, it started moving along the blade row and became a short-length-scale rotating stall. The size of the stall cell for the thick inlet boundary layer was larger than for the thin inlet boundary layer. Due to the bigger size of the stall cell, the performance of the single rotor for the former case dropped more significantly than for the latter case.

A numerical study of 3D unsteady flows in centrifugal compressor stages solving the Navier-Stokes equations is presented. The emphasis is on the effect of the radial gap between blade rows on the aerodynamic performance. In the numerical tests, Krain’s centrifugal impeller was combined with a DCA (Double Circular Arc) type radial vaned diffuser. The compressor stages with three settings of radial gap ranging from 5 to 15 percent of the impeller trailing edge radius are configured and unsteady flow simulations are carried out to compare the time-averaged efficiencies. The performance predictions show that the efficiency is deteriorated if the radial gap between blade rows is reduced with intensified blade row interaction, which is in contradiction to the general trend for axial compressor stages. In the centrifugal compressors tested, wake chopping by diffuser vanes, which usually benefits efficiency in axial compressor stages, causes unfavourable wake compression through the diffuser passages to deteriorate the efficiency.