Turbomachinery Applications Research Articles

Effect of Side Feed Pressurization on the Dynamic Performance of Gas Foil Bearings: A Model Anchored to Test Data

AbstractComprehensive modeling of gas foil bearings (GFBs) anchored to reliable test data will enable the widespread usage of these bearings into novel high speed turbomachinery applications. GFBs often need a forced cooling gas flow, axially fed through one end of the bearing, for adequate thermal management. This paper presents rotordynamic response measurements on a rigid rotor supported on GFBs during rotor speed run-up and coastdown tests with the GFBs supplied with increasing feed gas pressures to 2.8bars. Rotor speed run-up tests to 35krpm show that the bearing end side feed gas pressurization delays the onset speed of rotor subsynchronous whirl motions. The test data validate closely the predictions of the threshold speed of instability and the whirl frequency ratio derived from a GFB model that implements the axial evolution of gas circumferential flow velocity as a function of the imposed side feed pressure. Rotor speed coastdown tests from 25krpm with a low feed pressure of 0.35bar evidence a nearly linear synchronous rotor response for small and moderately large imbalance mass distributions. A structural finite element rotordynamics model integrates linearized synchronous speed GFB force coefficients and predicts synchronous responses, amplitude, and phase angles, agreeing with the test data. The analysis and measurements demonstrate the profound effect of the end side feed gas pressurization on the rotordynamic performance of GFBs.

Numerical strategies for investigation of gust-airfoil interaction

The noise generated by the interaction between a gust and an airfoil in a uniform flow is investigated. This problematic is of major industrial interest, regarding fans, turbomachinery, or wind turbine applications. A two-dimensional symmetric Joukowski-type airfoil is immersed without incidence in a flow at Mach number 0.5, disturbed by a harmonic gust at 45° of incidence (4th CAA Workshop on Benchmark Problems, 2004). Our methodology is first to perform a high-order direct resolution of Euler's equations of the disturbed flow over the airfoil and the associated acoustic emission, which is taken as a reference simulation. Second, the near aerodynamic field is simulated with Fluent 6.3 solver based on finite volume method with second-order schemes. The aerodynamic data thus obtained are used for far field acoustic prediction, based on Ffowcs Williams and Hawkings analogy. Finally, following another hybrid approach, the noise is predicted by using integral formulations with source field from the DNS. The aim of the study is to provide insight into the efficiency and validity of these numerical strategies commonly used. Comparisons with results of the CAA workshop are given, covering various wavenumbers values.

The vibrational behaviour of bladed disks with multiple coupling devices

AbstractIn turbomachinery applications turbine blades are subjected to high static and dynamic loads. Static loads are due to centrifugal stresses and thermal strains. Especially the dynamic excitation caused by fluctuating gas forces results in high vibration amplitudes which can lead to high cycle fatigue failures (HCF). Therefore, in practical applications, coupling devices like underplatform dampers, lacing wires and tip shrouds are installed to the structure. In case of blade vibrations the relative displacements between these coupling devices and the blades generate friction forces. The resulting energy dissipation provides additional damping to the structure. Furthermore, coupling devices, in particular tip shrouds, snubbers and lacing wires, increase the stiffness of the structure. Hence, they lead to a shift of the resonance frequencies.So far, only effects of single coupling devices and the influencing properties have been examined. Within this paper the effect of multiple couplings is determined and compared with single couplings. The forced response of turbine bladings with multiple couplings is calculated under consideration of geometrical and mechanical parameters of the blading and contacts, respectively. The results are compared with the single coupled blading. Furthermore, a multiple coupled device with under‐platform damper and connecting pin is compared with respect to his effectiveness. Especially the influence on the resonance frequency and the achievable damping is analysed. The results of the simulation are verified by measurements at a two‐blade non‐rotating test rig with an underplatform damper and connecting pin. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

A planar Mie scattering visualization technique for turbo machinery application

A planar Mie scattering technique is described which allows for the systematical analysis of three-dimensional shock configurations without the need for velocity measurements. For that, small particles are added to the flow upstream of the shock and illuminated by a laser light sheet. The scattered light is captured by a CCD camera so that the position of the shock wave can be determined from the increase of the light intensity across the shock. To analyze the three-dimensional structure of a shock wave, the light sheet is moved perpendicular to the flow direction. The measurement technique is applied to both a supersonic wedge flow at Ma = 2.43 and a more complex shock wave configuration in a transonic cascade flow at Ma = 1.09. The gathered results are conclusive with Schlieren photographs, numerical simulations and surface pressure measurements. As the chosen setup allows the application to test sections with restricted optical access and a synchronization of the image capturing process with an external triggering signal, the three-dimensional shock configurations inside a transonic compressor rotor can also be visualized.

Modeling of 3-D losses and deviations in a throughflow analysis tool

This contribution is dedicated to the modeling of the end-wall flows in a throughflow model for turbomachinery applications. The throughflow model is based on the Euler or Navier-Stokes equations solved by a Finite-Volume technique. Two approaches are presented for the end-wall modeling. The first one is based on an inviscid formulation with dedicated 3-D distributions of loss coefficient and deviation in the end-wall regions. The second approach is directly based on a viscous formulation with no-slip boundary condition along the annular end-walls and blade force modification in the region close to the end-walls. The throughflow results are compared to a series of 3-D Navier-Stokes calculations averaged in the circumferential direction. These 3-D calculations were performed on the three rotors of a low pressure axial compressor, for a series of tip clearance values.

Application of particle image velocimetry to a transonic centrifugal compressor

As part of an ongoing research project the performance and internal flow field of a high-pressure ratio centrifugal compressor is being investigated. Based on previous, primarily, point-wise laser-optical measurements the compressor was redesigned and resulted in an improved impeller and diffuser with a single-stage pressure ratio of 6:1 at 50,000 rpm. Current research activities involve the use of particle image velocimetry (PIV) to analyze and further improve the understanding of the complex flow phenomena inside the vaned diffuser given the capability of PIV of capturing spatial structures. The study includes phase-resolved measurements of the flow inside a diffuser vane passage with respect to the impeller blade position. Both, instantaneous and phase-averaged velocity fields are presented. The flow field results obtained by PIV are to be used for future validation of the related CFD calculations, which in turn are expected to lead to further improvements in compressor performance. In addition, the potential of stereo PIV for this type of turbomachinery application could be successfully demonstrated.

Fast response aerodynamic probes for measurements in turbomachines

The present paper concentrates on the use of single- and multi-sensor pressure probes designed for the accurate measurement of unsteady total pressure and flow angle for turbo-machinery applications. Fast response pressure probes have the advantage of offering a high analogue bandwidth, a high signal-to-noise ratio, and a good reliability and robustness as compared to hot-wire techniques for instance. In addition to the magnitude and direction of the flow velocity vector, aerodynamic probes also indicate the total and static pressure as time-resolved quantities. Appropriate data processing then allows to retrieve not only the steady, but also the periodic and random components of the measured variables. The design of several single or multi-sensor probes is briefly described in this paper. These probes have been developed, manufactured and fully calibrated at the von Karman Institute to be used in laboratory and in industry environments over the last 8 years. The main objective of this contribution is to address some examples of applications but focus on the capabilities of the technique rather than on flow field analysis.

Symmetric and Asymmetric Underplatform Dampers for Turbine Blades

AbstractFriction damping devices like underplatform dampers are widely used in turbomachinery applications to reduce vibration amplitudes and to increase lifetime and reliability of the bladed disk. Nowadays, in practical applications, a variety of different underplatform damper geometries is applied. Nevertheless, a detailed study of the in.uence of the geometric and dynamic properties of the damper is still not available. Within this paper the most frequently applied damper types like cylindrical and wedge as well as asymmetrical dampers are investigated and compared to each other with respect to their effectiveness. Especially the in.uence of the damper geometry on the resonance frequency and vibration amplitude is pointed out. (© 2006 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Active Flow Control in Circular and Transitioning S-duct Diffusers

Performance enhancement of three-dimensional S-duct diffusers by secondary flow and separation control using vortex generator jets is the objective of the current experimental investigation. Two different diffuser geometries namely, a circular diffuser and a rectangular—to—circular transitioning diffuser were studied. The experiments were performed in uniform inflow conditions at a Reynolds number of 7.8×105 and the performance evaluation of the diffusers was carried out in terms of static pressure recovery and quality (flow uniformity) of the exit flow. Detailed measurements that included total pressure, velocity distribution, surface static pressure, skin friction, and boundary layer measurements were taken and these results are presented here in terms of static pressure rise, distortion coefficient, total pressure loss coefficient, and the transverse velocity vectors at the duct exit. The use of vortex generator jets resulted in around 26% in total pressure loss and about 22% decrease in flow distortion coefficients in the circular and transitioning diffusers. The mass flow rate of the air injected through the VGJ was about 0.1% of the mass flow rate of the main flow for secondary flow control and about 0.06% of the main flow for separation control. The physical mechanism of the flow control devices used has been explored. The structure of the vortices generated by the control methods are presented in the form of smoke visualization images. The method of flow control used here is perceived to have applications in turbomachinery like turbines and compressors.

Numerical study of the flow field through a transonic linear turbine cascade at design and off-design conditions

The flow field through a transonic linear turbine cascade is studied in this paper at design and off-design conditions. The compressible flow field is obtained by solving the equations governing the fluid flow and heat transfer. Two eddy-viscosity turbulence models are used to simulate the turbulence in turbine cascades: the standard k-ϵ model and the Spalart–Allmaras model. The standard k-ϵ model is the most universal and popular model for industrial flow and heat transfer simulations. It has the shortcoming of accurately predicting the profile loss in turbomachinery applications. The Spalart–Allmaras model is a relatively recent and simple one-equation model. In this paper, the model is tested for turbomachinery applications. The ability of the model to accurately predict the flow field in turbine cascade is tested. Blade loading, downstream wake distribution, total pressure loss coefficient, and exit flow angle are used in this study with comparisons to the standard k-ϵ model and experimental data. The design condition and the off-design conditions are considered.

A Correlation-Based Transition Model Using Local Variables—Part I: Model Formulation

A new correlation-based transition model has been developed, which is based strictly on local variables. As a result, the transition model is compatible with modern computational fluid dynamics (CFD) approaches, such as unstructured grids and massive parallel execution. The model is based on two transport equations, one for intermittency and one for the transition onset criteria in terms of momentum thickness Reynolds number. The proposed transport equations do not attempt to model the physics of the transition process (unlike, e.g., turbulence models) but form a framework for the implementation of correlation-based models into general-purpose CFD methods. Part I (this part) of this paper gives a detailed description of the mathematical formulation of the model and some of the basic test cases used for model validation, including a two-dimensional turbine blade. Part II (Langtry, R. B., Menter, F. R., Likki, S. R., Suzen, Y. B., Huang, P. G., and Volker, S., 2006, ASME J. Turbomach., 128(3), pp. 423–434) of the paper details a significant number of test cases that have been used to validate the transition model for turbomachinery and aerodynamic applications. The authors believe that the current formulation is a significant step forward in engineering transition modeling, as it allows the combination of correlation-based transition models with general purpose CFD codes.

The Vibrational Behavior of Bladed Disks in Consideration of Friction Damping and Contact Elasticity

AbstractRotating turbine blading is subjected to fluctuating gas forces during operation that cause blade vibrations. One of the main tasks in the design of turbomachinery blading is the reduction of the vibration amplitudes of the blades to avoid high resonance stresses that could damage the blading. The vibration amplitudes of the blades can be reduced significantly to a reasonable amount by means of friction damping devices such as underplatform dampers. In the case of blade vibrations, relative displacements between the friction damping devices and the neighboring blades occur and friction forces are generated that provide additional damping to the structure due to the dry friction energy dissipation. In real turbomachinery applications, spatial blade vibrations caused by a complex blade geometry and distributed excitation forces acting on the airfoil accur. Therefore, a three dimensional model including an appropriate spatial contact model to predict the generalized contact forces is necessary to describe the vibrational behavior of the blading with sufficient accuracy, see [1] and [2]. In this paper the contact model presented in [2] is extended to include also local deformations in the contacts between underplatform dampers and the contact surfaces of the adjacent blades. The additional elasticity in the contact influences the resonance frequency of the coupled bladed disk assembly. (© 2005 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Reynolds number effects on the non-nulling calibration of a cone-type five-hole probe for turbomachinery applications

The effects of Reynolds number on the non-nulling calibration of a typical cone-type fivehole probe have been investigated for the representative Reynolds numbers in turbomachinery. The pitch and yaw angles are changed from −35 degrees to 35 degrees with an angle interval of 5 degrees at six probe Reynolds numbers in range between 6.60 × 103 and 3.17× 104. The result shows that not only each calibration coefficient itself but also its Reynolds number dependency is affected significantly by the pitch and yaw angles. The Reynolds-number effects on the pitchand yaw-angle coefficients are noticeable when the absolute values of the pitch and yaw angles are smaller than 20 degrees. The static-pressure coefficient is sensitive to the Reynolds number nearly all over the pitch- and yaw-angle range. The Reynolds-number effect on the totalpressure coefficient is found remarkable when the absolute values of the pitch and yaw angles are larger than 20 degrees. Through a typical non-nulling reduction procedure, actual reduced values of the pitch and yaw angles, static and total pressures, and velocity magnitude at each Reynolds number are obtained by employing the calibration coefficients at the highest Reynolds number (Re=3.17×104) as input reference calibration data. As a result, it is found that each reduced value has its own unique trend depending on the pitch and yaw angles. Its general tendency is related closely to the variation of the corresponding calibration coefficient with the Reynolds number. Among the reduced values, the reduced total pressure suffers the most considerable deviation from the measured one and its dependency upon the pitch and yaw angles is most noticeable. In this study, the root-mean-square data as well as the upper and lower bounds of the reduced values are reported as a function of the Reynolds number. These data would be very useful in the estimation of the Reynolds-number effects on the non-nulling calibration.

Design and analysis of new concept fast-response pressure probes

Two cylindrical fast-response pressure probes for unsteady flow measurements in turbomachinery have been designed and tested. Commercially available miniaturized pressure sensors have been used to ensure high reliability and low manufacturing costs. Analytical and numerical models have been applied to the design of the line–cavity system connecting the pressure tap and the (encapsulated) sensor, to improve the dynamic behaviour of the probes. Dynamic calibrations in a low-pressure shock tube have also been carried out to determine the probe transfer function and to extend the probe operating range beyond the resonance frequency of the line–cavity system. One of the new probes exhibits a frequency response of about 80 kHz in the range 0–35 kPa, a suitable value for most turbomachinery applications.

A Framework for Coupling Reynolds-Averaged With Large-Eddy Simulations for Gas Turbine Applications

Abstract Full-scale numerical prediction of the aerothermal flow in gas turbine engines are currently limited by high computational costs. The approach presented here intends the use of different specialized flow solvers based on the Reynolds-averaged Navier-Stokes equations as well as large-eddy simulations for different parts of the flow domain, running simultaneously and exchanging information at the interfaces. This study documents the development of the interface and proves its accuracy and efficiency with simple test cases. Furthermore, its application to a turbomachinery application is demonstrated.

Fan Interaction Noise Predictions Using RANS-BEM Coupling

A source-to-far-field computation procedure aiming at predicting the noise generated by rotor-stator fan interactions and inlet noise radiation is presented here. It is applied to the NLR turbofan model, tested in DNW-LLF anechoic chamber under the framework of DUCAT project. The unsteady aerodynamic input for the acoustic computation is obtained using a 3D RANS code, CANARI, developed at ONERA for turbomachinery applications. The CFD solutions are coupled, using a modal expansion approach, to a BEM code, solving the Helmholtz equation in an arbitrary bounded space. A single interacting cut-on mode amplitude is either directly deduced from experiment or provided by post-processing the CFD data. With this approach, the predicted noise radiation can be related to the directivity patterns measured at several axial positions upstream of the turbofan inlet. A fairly good agreement is found using both experimental and CFD input data.

A Correlation-Based Transition Model Using Local Variables—Part II: Test Cases and Industrial Applications

A new correlation-based transition model has been developed, which is built strictly on local variables. As a result, the transition model is compatible with modern computational fluid dynamics (CFD) methods using unstructured grids and massive parallel execution. The model is based on two transport equations, one for the intermittency and one for the transition onset criteria in terms of momentum thickness Reynolds number. The proposed transport equations do not attempt to model the physics of the transition process (unlike, e.g., turbulence models), but form a framework for the implementation of correlation-based models into general-purpose CFD methods. Part I of this paper (Menter, F. R., Langtry, R. B., Likki, S. R., Suzen, Y. B., Huang, P. G., and Völker, S., 2006, ASME J. Turbomach., 128(3), pp. 413–422) gives a detailed description of the mathematical formulation of the model and some of the basic test cases used for model validation. Part II (this part) details a significant number of test cases that have been used to validate the transition model for turbomachinery and aerodynamic applications, including the drag crisis of a cylinder, separation-induced transition on a circular leading edge, and natural transition on a wind turbine airfoil. Turbomachinery test cases include a highly loaded compressor cascade, a low-pressure turbine blade, a transonic turbine guide vane, a 3D annular compressor cascade, and unsteady transition due to wake impingement. In addition, predictions are shown for an actual industrial application, namely, a GE low-pressure turbine vane. In all cases, good agreement with the experiments could be achieved and the authors believe that the current model is a significant step forward in engineering transition modeling.

Incorporation of Cabot's wall functions in RANS with applications in turbomachinery

Cabot's wall functions model (CWM) was originally introduced in conjunction with Large Eddy Simulations (LES), for modelling the turbulent near-wall flow. The present paper presents a novel application of CWM, namely its use within the framework of Reynolds Averaged Navier-Stokes (RANS) modelling. In the present work a q-ω turbulence model has been used. The implementation of the model has been validated considering five two-dimensional test cases including turbomachinery applications. The results confirm that CWM provides an advantageous approach also for RANS modelling, providing an optimal balance between accuracy and economy for modelling the near-wall turbulent flow. Compared to the low Reynolds number models (LRN), the CWM allows the use of a coarser grid near the wall, while providing better accuracy compared to the standard logarithmic wall-function methods. The applicability of a coarser near-wall resolution speeds up the simulation not only due to the reduced number of grid nodes, but also due to the fact that larger time-steps can be used without endangering the stability of the numerical solution. The computational results produced by CWM have been shown to either improve or retain the accuracy of the LRN model. Even in adverse pressure gradients, the model is observed to provide still a relatively good solution. Hence, CWM, with proper coding and suitable grids, has the potential providing solutions with comparable accuracy to LRN, while being much more economical.

Effect of Flow-Induced Radial Load on Brush Seal/Rotor Contact Mechanics

Brush seals comprised of special-alloy wire bristles are currently being used in lieu of traditional labyrinth seals for turbomachinery applications. This advancement in seal technology utilizes close-packed bristles that readily undergo lateral deformation arising from aerodynamic loads as well as loads imparted by the rotor surface. Thus, during operation, filament tips remain in contact with the rotor surface, which, in turn, inhibits leakage between successive stages of the turbine, and increases engine efficiency. However, contact forces generated at the interface of the rotor and fiber tips can lead to eventual bristle fatigue and wear of the seal/rotor system. Therefore, it is important that reliable modeling techniques be developed that can help identify complex relationships among brush seal design parameters, in-service loads, and contact forces that arise during the operation of turbomachinery. This paper is concerned with modeling and evaluating bristle deformation, bending stress, and bristle/rotor contact forces that are generated at the interface of the fiber and rotor surface due to radial fluid flow, and augments previous work reported by the author’s, which assessed filament tip forces that arise solely due to interference between the bristle/rotor. The current problem derives its importance from aerodynamic forces that are termed “blow-down,” that is, the inward radial flow of gas in close proximity to the face of the seal. Thus, bristle deformation, bristle tip reaction force, and bristle bending stress is computed on the basis of an in-plane, large-displacement mechanics analysis of a cantilever beam that is subjected to a uniform radial load. Solutions to the problem are obtained for which the filament tip is constrained to lie on the rotor surface, and includes the effect of Coulombic friction at the interface of the fiber tip and rotor. Contact forces are obtained for a range of brush seal design parameters including fiber lay angle, flexural rigidity, and length. In addition, the governing equation is cast in non-dimensional form, which extends the range of applicability of solutions to brush seals having a more general geometry and material composition.

Development of a Three-Dimensional Geometry Optimization Method for Turbomachinery Applications

Development of a Three-Dimensional Geometry Optimization Method for Turbomachinery Applications