A Study of Unsteady Secondary Flow in a Water Flow Axial Turbine Model

The influence of secondary flows on the performance of turbines has been investigated in great detail in the last decades. The interaction of vortices with following blade rows has been identified to be one of the loss mechanisms within a turbo-machine. This paper presents for the first time detailed flow visualization photographs of the interaction of the vane passage vortex with the rotor. The appearance vortex breakdown could be identified before and within the rotating passage of the turbine. The measurements were taken in a vertical water channel. Water is used instead of air because the flow visualization can be realised very easily with injected ink. For different relative positions of rotor to stator a series of photographs were taken. With an image editing process the average and the pixel RMS were calculated for each relative position. The pixel RMS is a useful indicator to identify highly turbulent regions in the flow field. The photographs of the vortex breakdown show spots of high pixel RMS which are associated with very high turbulence and therefore can be regarded as sources of loss. Insight is gained into the nature of the passage vortex breakdown mechanisms as follows: first the pressure wave of the rotor stretches the vortex causing a spiral vortex instability, then the vortex interacts with the leading edge as it attempts to cut the vortex. In the stagnation region of the blade a bubble type instability forms, expands and then convects through the rotor. The absolute trajectory of the vortex fluid reveals that it exchanges no work with the rotor.

Numerical Simulations of Onset of Volute Stall Inside a Centrifugal Compressor

In a previous publication (Guo & Chen et al., 2007), the authors solved the unsteady, 3-D Navier-Stokes equations with the k-ε turbulence model using CFX software to show that there is a volute stall coincided with the stage stall of a turbocharger centrifugal compressor operated at 423m/s tip speed and the stage stall frequency is dictated by a volute standing wave. This paper presents the flow condition at the vaneless diffuser and volute from the same simulation at various mass flow rates from stage peak efficiency to deep stage stall. Time averaged flow conditions show that (1) the influence of exducer blade passing at the volute inlet rapidly diminishes at the compressor peak pressure ratio point and the influence vanishes when the stage is in stall; (2) only at the peak pressure ratio point, circumferentially averaged, spanwise distribution of radial velocity at the volute inlet has an inflection point and the distribution meets the requirement of the Fjo̸rtoft instability theorem; (3) in the volute discharge section, the flow stalls after the stage stalls and the vortex core at the cross sectional center of the section breaks down; (4) impeller total pressure rise curve has a flat region in the middle before the stage stalls and (5) diffuser stall triggers the stage stall and drives the volute into stall.

Experimental Investigations and Numerical Validation of an Outlet Guide Vane With an Engine Mount Recess

This paper presents experiments and CFD calculations of a Low Pressure Turbine/Outlet Guide Vane (LPT/OGV) equipped with an engine mount recess (a bump) tested in the Chalmers linear LPT/OGV cascade. The investigated characteristics include performance for the design point in terms of total pressure loss and turning as well as a detailed description of the downstream development of the secondary flow field. The numerical simulations are performed for the same inlet conditions as in the test-facility with engine-like properties in terms of Reynolds number, boundary-layer thickness and inlet flow angle. The objective is to validate how accurately and reliably the secondary flow field and losses can be predicted for an LPT/OGV equipped with a bump. Three different turbulent models as implemented in FLUENT, the k-ε realizable model, the kω-SST model and the RSM are validated against detailed measurements. From these results it can be concluded that the kω-SST model predicts both the secondary flow field and the losses most accurately.

Shroud Leakage Modeling of the Flow in a Two Stage Axial Test Turbine

Three dimensional steady multistage calculations, using mixing plane approach, are presented for two different blade geometries in a two stage axial test turbine with shrouded blades. A 3D multiblock Navier-Stokes finite volume solver (TBLOCK) has been used in all the simulations. In order to study shroud leakage flow effects the whole shroud cavity geometry has been modeled, overcoming most of the limitations of simple shroud leakage model in calculating fluid flow over complex geometries. Numerical investigations are mainly focused on assessing the ability of the solver to be used as multistage design tool for modeling leakage-mainstream flow interaction. Several calculations are compared. The first computes the main blade flow path with no modeling of the shroud cavities. The second includes the modeling of the shroud cavities for a zero leakage mass flow rate. Finally a multiblock calculation which models all the leakage flow paths and shroud cavities has been carried out for two different levels of shroud seal clearance. It is found that neglecting shroud leakage significantly alters the computed velocity profiles and loss distributions, for both the computed blade geometries. A numerically predicted shroud leakage offset loss is presented for the two considered blade geometries, focusing on the relative importance of the leakage flow, re-entry mixing losses, and inlet and exit shroud cavity effect. Results demonstrates that full calculation of leakage flow paths and cavities is required to obtain reliable results, indicating the different effects of the leakage-to-mainstream flow interaction on the blade geometries computed. Despite a slight increase in the computational time, multiblock approach in handling leakage flow problem can now-days be used as a practical tool in the blade design process and routine shroud leakage calculations.

Aerodynamics of Compressor Casing Treatment: Part I—Experiment and Time-Accurate Numerical Simulation

This paper presents both experimental and unsteady RANS investigations of a slot-type casing treatment at a transonic axial flow compressor rotor. Experimental results show that at 60% and 98% of rotor design wheel speeds, approximately 100% and 200% extra extensions of the rotor operation ranges are achieved, respectively. On the other hand, there are about 3.6% and 2.0% drops of efficiencies at 60% and 98% speeds respectively if comparisons are made at the same peak-efficiency mass flow rates of the solid casing case. If comparing the respective peak efficiencies for the solid casing case with those for the treated casing case, there are still about 3.4% and 0.7% drops at 60% and 98% speeds, respectively. As for the unsteady RANS study, an in-house unsteady RANS code has been used to study the casing treatment flow at several operating points, i.e., the peak efficiency and the near stall with regard to the solid casing case at 60% speed and 98% speed, respectively. It is shown that the interactions between the blade passage flow and the casing treatment flow exhibit different manner at two rotating speeds. The flow condition in which the rotor operates, i.e., either the subsonic condition at the 60% speed or the transonic condition with passage shock presented at the 98% speed, is one of the determinate factors that are responsible for the manner the casing treatment works. The loss production due to casing treatment is also particularly discussed.

Transition Modeling Effects on the Simulation of a Stator Cascade With Active Flow Control

An investigation of a stator cascade is undertaken by means of steady 3D RANS simulations, the focus of which is on two computational setups. The first takes transition effects into account using a correlation-based transition model as suggested by Abu-Ghannam and Shaw, while the second is considered to be fully turbulent. In a first step the base flow is validated by experimental measurements, followed by configurations employing active flow control by means of steady jets with varying mass flow. By investigating the differences arising due to the varying level of modeling complexity the necessity of using a transition model can be illustrated.

Improving Efficiency of a High Work Turbine Using Non-Axisymmetric Endwalls: Part I—Endwall Design and Performance

This paper is the first part of a two part paper reporting the improvement of efficiency of a one-and-half stage high work axial flow turbine by non-axisymmetric endwall contouring. In this first paper the design of the endwall contours is described and the CFD flow predictions are compared to five-hole-probe measurements. The endwalls have been designed using automatic numerical optimization by means of an Sequential Quadratic Programming (SQP) algorithm, the flow being computed with the 3D RANS solver TRACE. The aim of the design was to reduce the secondary kinetic energy and secondary losses. The experimental results confirm the improvement of turbine efficiency, showing a stage efficiency benefit of 1%±0.4%, revealing that the improvement is underestimated by CFD. The secondary flow and loss have been significantly reduced in the vane, but improvement of the midspan flow is also observed. Mainly this loss reduction in the first row and the more homogeneous flow is responsible for the overall improvement. Numerical investigations indicate that the transition modeling on the airfoil strongly influences the secondary loss predictions. The results confirm that non-axisymmetric endwall profiling is an effective method to improve turbine efficiency, but that further modeling work is needed to achieve a good predictability.

Second Vane Total Pressure Loss Due to Endwall Iceform Contouring

The iceformation design method generates an endwall contour, altering the secondary flows that produce elevated endwall heat transfer load and total pressure losses. Iceformation is an analog to regions of metal melting where a hot fluid alters the isothermal surface shape of a part as it is maintained by a cooling fluid. The passage flow, heat transfer and geometry evolve together under the constraints of flow and thermal boundary conditions. The iceformation concept is not media dependent and can be used in analogous flows and materials to evolve novel boundary shapes. In the past, this method has been shown to reduce aerodynamic drag and total pressure loss in flows such as diffusers and cylinder/endwall junctures. A prior paper [1] showed that the Reynolds number matched iceform geometry had a 24% lower average endwall heat transfer than the rotationally symmetric endwall geometry of the Energy Efficiency Engine (E3). Comparisons were made between three endwall geometries: the ‘iceform’, the ‘E3’ and the ‘flat’ as a limiting case of the endwall design space. This paper adds to the iceformation design record by reporting the endwall aerodynamic performances. Second vane exit flow velocities and pressures were measured using an automated 2-D traverse of a 1.2 mm diameter five-hole probe. Exit plane maps for the three endwall geometries are presented showing the details of the total pressure coefficient contours and the velocity vectors. The formation of secondary flow vortices is shown in the exit plane and this results in an impact on exit plane total pressure loss distribution, off-design over- and under-turning of the exit flow. The exit plane contours are integrated to form overall measures of the total pressure loss. Relative to the E3 endwall, the iceform endwall has a slightly higher total pressure loss attributed to higher dissipation of the secondary flow within the passage. The iceform endwall has a closer-to-design exit flow pattern than the E3 endwall.

Numerical Simulation of Tip Clearance Control in Axial Turbine Rotor: Part 2—Passive Control of Five Different Tip Platforms

The tip leakage flow has significant effects on turbine in loss production, aerodynamic efficiency, etc. Then it’s important to minimize these effects for a better performance by adopting corresponding flow control. The active turbine tip clearance flow control with injection from the tip platform is given in Part-1 of this paper. This paper is Part-2 of the two-part papers focusing on the effect of five different passive turbine tip clearance flow control methods on the tip clearance flow physics, which consists of a partial suction side squealer tip (Partial SS Squealer), a double squealer tip (Double Side Squealer), a pressure side tip shelf with inclined squealer tip on a double squealer tip (Improved PS Squealer), a tip platform extension edge in pressure side (PS Extension) and in suction side (SS Extension) respectively. Combined with the turbine rotor and the numerical method mentioned in Part 1, the effects of passive turbine tip clearance flow controls on the tip clearance flow were sequentially simulated. The detailed tip clearance flow fields with different squealer rims were described with the streamline and the velocity vector in various planes parallel to the tip platform or normal to the tip leakage vortex core. Accordingly, the mechanisms of five passive controls were put in evidence; the effects of the passive controls on the turbine efficiency and the tip clearance flow field were highlighted. The results show that the secondary flow loss near the outer casing including the tip leakage flow and the casing boundary layer can be reduced in all the five passive control methods. Comparing the active control with the passive control, the effect brought by the active injection control on the tip leakage flow is evident. The turbine rotor efficiency could be increased via the rational passive turbine tip clearance flow control. The Improved PS Squealer had the best effect on turbine rotor efficiency, and it increased by 0.215%.

Effects of Rotating Blade Wakes on Separation and Pressure Recovery in Turbine Exhaust Diffusers

Highly efficient turbine exhaust diffusers can only be designed by taking into account the unsteady interactions with the last rotating row of the turbine. Therefore, a scale model of a typical gas turbine exhaust diffuser consisting of an annular and a conical diffuser is investigated experimentally. To investigate the influence of rotating wakes, a variable-speed rotating spoke wheel with cylindrical spokes as well as with NACA bladed spokes generates high-energy turbulent wakes simulating turbine rotor wakes. For the rotor with the NACA blades, the drive of the wheel is run in motor as well as in generator mode. Additional measurements in a reference configuration without a spoke wheel allow the detailed analysis of changes in the flow pattern. 3-hole pneumatic probes, static pressure taps, as well as a 2D-Laser-Doppler-Velocimeter (LDV) are used to investigate velocity profiles and turbulent characteristics. Without the wakes generated by a spoke wheel, the annular diffuser (with a 20° half cone opening angle) separates at the shroud for all swirl configurations. Increasing the swirl results in increasing pressure recovery at the shroud whereas the hub boundary is destabilized. For a non-rotating spoke rotor and low swirl numbers, the 20° annular diffuser separates at the shroud. Increasing the swirl number, a strong deceleration of the axial velocity at the shroud is generated without separation and a higher pressure recovery is achieved. The boundary layer at the shroud of the 20° annular diffuser separates for all operating points with the bladed rotor. A partly stabilized 20° annular diffuser can only be achieved for much higher values of the flow coefficient than that for the design point. At this high mass flow, the NACA-bladed rotor operates as a turbine, resulting in the generator mode of the electric drive. Contrary to the numerical design calculations, the flow at the shroud of a 15° annular diffuser does not separate for all swirl configurations in the experiment. Pressure recovery of the 15° annular diffuser can be increased by increasing the inlet swirl whereas the hub boundary layer is destabilized. For the NACA bladed rotor, the flow in the 15° annular diffuser as well as the pressure recovery strongly depend on the flow coefficient. For flow coefficients lower than the design value, the flow partly separates at the shroud whereas large flow coefficients result in increased pressure recovery. The pressure recovery also depends on the direction of swirl and thus the swirl number.