Reynolds-Averaged Navier-Stokes Code Research Articles

Wake-blade interaction in steam turbine stages

Abstract The article discusses the phenomenon of stator Wake/Rotor cascade (W/R) interaction in a steam turbine stage, and the ability to capture it in turbine stage design calculations making use of standard numerical codes. Firstly, the W/R interaction is analysed by comparing its real, experimentally recorded course with the numerical results obtained using vortex theory models and methods. This part of the analysis ends with formulating a conclusion about stochastic nature of the W/R interaction and indicating its reason, which is the vortex structure of the stator wake. Next, a question is discussed whether and how this stochastic nature of the examined phenomenon can be taken into account in calculations of Reynolds Averaged Navier-Stokes (RANS) equations. Differences are indicated between the uniform pattern of the stator wake obtained using a RANS code and the vortex structure of the real wake. It is concluded, however, that despite these differences the RANS results correctly reflect the time-averaged course of the real W/R interaction, and the process of averaging the flow parameters on the sliding plane between stator and rotor calculation areas can be treated as sort of “numerical averaging” of different realisations of the W/R interaction.

Numerical Investigations of Lateral Characteristics of an Air-to-Air Missile

Numerical simulations were carried out for investigating the lateral aerodynamic characteristics of an air-to-air missile using an in-house-developed Reynolds-averaged Navier–Stokes code, CERANS. The analyses were carried out for Mach numbers from 0.8 to 2.0 and angles of attack ranging from 10 to 20 deg for roll sweeps from to . The rolling-moment behavior of the missile at Mach 1.2 and an angle of attack of 15 deg was explicitly and exhaustively analyzed because this flight condition is critical from the point of view of severe induced roll. The necessity of studying the rolling-moment behavior at finer roll orientations is demonstrated. The main objectives of the investigations are to explain the governing physical phenomenon and the resultant complex flow behavior that caused excessive induced roll generated during the critical flight condition.

Computational Fluid Dynamics of a Radial Compressor Operating With Supercritical CO2

The merit of using supercritical CO2(scCO2) as the working fluid of a closed Brayton cycle gas turbine is now widely recognized, and the development of this technology is now actively pursued. scCO2 gas turbine power plants are an attractive option for solar, geothermal, and nuclear energy conversion. Among the challenges that must be overcome in order to successfully bring the technology to the market is that the efficiency of the compressor and turbine operating with the supercritical fluid should be increased as much as possible. High efficiency can be reached by means of sophisticated aerodynamic design, which, compared to other overall efficiency improvements, like cycle maximum pressure and temperature increase, or increase of recuperator effectiveness, does not require an increase in equipment cost, but only an additional effort in research and development. This paper reports a three-dimensional computational fluid dynamics (CFD) study of a high-speed centrifugal compressor operating with CO2 in the thermodynamic region slightly above the vapor–liquid critical point. The investigated geometry is the compressor impeller tested in the Sandia scCO2 compression loop facility. The fluid dynamic simulations are performed with a fully implicit parallel Reynolds-averaged Navier–Stokes code based on a finite volume formulation on arbitrary polyhedral mesh elements. In order to account for the strongly nonlinear variation of the thermophysical properties of supercritical CO2, the CFD code is coupled with an extensive library for the computation of properties of fluids and mixtures. A specialized look-up table approach and a meshing technique suited for turbomachinery geometries are also among the novelties introduced in the developed methodology. A detailed evaluation of the CFD results highlights the challenges of numerical studies aimed at the simulation of technically relevant compressible flows occurring close to the liquid–vapor critical point. The data of the obtained flow field are used for a comparison with experiments performed at the Sandia scCO2 compression-loop facility.

Numerical Simulation of Stage Jettisoning of a Two-Stage Rocket with Different Separation Distances

CFD simulations were carried out to evaluate the flow field during the jettisoning process of a two-stage rocket using an inhouse developed RANS code. Two simulations with stage separation distances of 500 mm and 1,000 mm were considered for simulation. For the 500 mm case, there was a severe ‘jet-upstream’ flow interaction which resulted in an upstream flow separation and hence can cause large drag rise. However, there was no such observation of flow separation anywhere on the upper stage for the 1,000 mm case. A study was carried out to evaluate the effect of spatial order of accuracy for modelling the governing Reynolds Averaged Navier–Stokes (RANS) equations on the simulation results.

로터 익형 KU109C 풍동시험 및 천이유동 해석결과의 검증

Transition prediction results are validated with experimental data obtained from a transonic wind tunnel for the KU109C airfoil. A Reynolds-Averaged Navier-Stokes code is simultaneously coupled with the transition transport model of Langtry and Menter and applied to the numerical prediction of aerodynamic performance of the KU109C airfoil. Drag coefficients from the experiment are better correlated to the numerical prediction results using a transition transport model rather than the fully turbulent simulation results. Maximum lift coefficient and drag divergence at the zero-lift condition with Mach number are investigated. Through the present validation procedure, the accuracy and usefulness of both the experiment and the numerical prediction are assessed.

Numerical study comparing RANS and LES approaches on a circulation control airfoil

A numerical study over a nominally two-dimensional circulation control airfoil is performed using a large-eddy simulation code and two Reynolds-averaged Navier–Stokes codes. Different Coanda jet blowing conditions are investigated. In addition to investigating the influence of grid density, a comparison is made between incompressible and compressible flow solvers. The incompressible equations are found to yield negligible differences from the compressible equations up to at least a jet exit Mach number of 0.64. The effects of different turbulence models are also studied. Models that do not account for streamline curvature effects tend to predict jet separation from the Coanda surface too late, and can produce non-physical solutions at high blowing rates. Three different turbulence models that account for streamline curvature are compared with each other and with large eddy simulation solutions. All three models are found to predict the Coanda jet separation location reasonably well, but one of the models predicts specific flow field details near the Coanda surface prior to separation much better than the other two. All Reynolds-averaged Navier–Stokes computations produce higher circulation than large eddy simulation computations, with different stagnation point location and greater flow acceleration around the nose onto the upper surface. The precise reasons for the higher circulation are not clear, although it is not solely a function of predicting the jet separation location correctly.

Analysis of Unsteady Tip and Endwall Heat Transfer in a Highly Loaded Transonic Turbine Stage

In a previous study, vane-rotor shock interactions and heat transfer on the rotor blade of a highly loaded transonic turbine stage were simulated. The geometry consists of a high pressure turbine vane and a downstream rotor blade. This study focuses on the physics of flow and heat transfer in the rotor tip, casing, and hub regions. The simulation was performed using the unsteady Reynolds-averaged Navier–Stokes code MSU-TURBO. A low Reynolds number k-ε model was utilized to model turbulence. The rotor blade in question has a tip gap height of 2.1% of the blade height. The Reynolds number of the flow is approximately 3×106/m. Unsteadiness was observed at the tip surface that results in intermittent “hot spots.” It is demonstrated that unsteadiness in the tip gap is governed by inviscid effects due to high speed flow and is not strongly dependent on pressure ratio across the tip gap contrary to published observations that have primarily dealt with subsonic tip flows. The high relative Mach numbers in the tip gap lead to a choking of the leakage flow that translates to a relative attenuation of losses at higher loading. The efficacy of new tip geometry is discussed to minimize heat flux at the tip while maintaining choked conditions. In addition, an explanation is provided that shows the mechanism behind the rise in stagnation temperature on the casing to values above the absolute total temperature at the inlet. It is concluded that even in steady (in a computational sense) mode, work transfer to the near tip fluid occurs due to relative shearing by the casing. This is believed to be the first such explanation of the work transfer phenomenon in the open literature. The difference in pattern between steady and time-averaged heat fluxes at the hub is also explained.

RANS Simulations of Supercavity Flows

This work aims at evaluating the capacity and limitations of conventional Reynolds-averaged Navier-Stokes (RANS) techniques to numerically simulate supercavity flows. The configuration is that of a two-dimensional (2D) symmetrical supercavitating wedge investigated experimentally by Michel (1974). Mesh effect is studied in detail under noncavitating conditions. The computational grid is refined in the region where cavitation develops in order to accurately track the supercavity. The effect of tunnel height is also analyzed, and the height finally chosen was large enough to simulate an infinite flow-field. The cavitating flow is treated as a homogeneous mixture of variable density. To account for vaporization and condensation, an additional continuity equation for the vapor (or the liquid) is solved with an appropriate source term expressing mass transfer between the two phases. The effect of nuclei concentration on the vaporization rate and then on the development of the supercavity is investigated. Results obtained using two different RANS codes are compared. They are also compared with experimental data and with inviscid solutions including a nonlinear boundary element method. They concern in particular cavity length and shape, pressure distribution, and drag coefficient. Under unsteady conditions, a special attention is paid to the characteristic growth time of the supercavity following a sudden pressure drop, from noncavitating conditions.

High-Fidelity Numerical Analysis of Per-Rev-Type Inlet Distortion Transfer in Multistage Fans—Part II: Entire Component Simulation and Investigation

Part I of this paper validated the ability of the unsteady Reynolds-Averaged Navier-Stokes (RANS) solver PTURBO to accurately simulate distortion transfer and generation through selected blade rows of two multistage fans. In this part, unsteady RANS calculations were successfully applied to predict the 1/rev inlet total pressure distortion transfer in the entirety of two differently designed multistage fans. This paper demonstrates that high-fidelity computational fluid dynamics (CFD) can be used early in the design process for verification purposes before hardware is built and can be used to reduce the number of distortion tests, hence reducing engine development cost. The unsteady RANS code PTURBO demonstrated remarkable agreement with the data, accurately capturing both the magnitude and the profile of total pressure and total temperature measurements. Detailed analysis of the flow physics identified from the CFD results has led to a thorough understanding of the total temperature distortion generation and transfer mechanism, especially for the spatial phase difference of total pressure and total temperature profiles. The analysis illustrates that the static parameters are more revealing than their stagnation counterpart and that pressure and temperature rise are more revealing while the pressure and temperature ratio could be misleading. The last stage is effectively throttled by the inlet distortion even though the overall engine throttle remains unchanged. The total temperature distortion generally grows as flow passes through the fan stages.

Comparison of Several Optimization Strategies for Robust Turbine Blade Design

This paper addresses the problem of turbine blade shape optimization in the presence of geometric uncertainties. Several strategies are tested and compared on a two-dimensional compressor blade optimization process for which performance is assessed using a commercial Reynolds-averaged Navier―Stokes computational fluid dynamics code. In each case, a range of shape errors are considered that attempt to simulate foreign object damage, erosion damage, and manufacturing errors. These lead to stochastic performance measures that, in turn, are considered in a multi-objective optimization framework. Because of the long run times associated with Reynolds-averaged Navier―Stokes codes, use is also made of surrogate or response-surface-based optimization methods to speed up the search processes. The paper shows that a range of techniques can be used to tackle this problem, but that no one method is clearly best overall. The practitioner is therefore cautioned against favoring a single approach for such design problems. Further research may help clarify these issues.

Experimental and numerical investigation of the flow field in a high-pressure centrifugal compressor impeller near surge

Numerical and experimental investigations were conducted in a transonic centrifugal compressor stage composed of a backswept splittered unshrouded impeller and a vaned diffuser. The present article focuses on the results obtained within the impeller, at an operating condition close to the surge of the compressor. The experimental results were obtained from a laser Doppler anemometry investigation. Unsteady numerical simulations of the compressor stage were performed using a three-dimensional Reynolds-averaged Navier—Stokes code with a phase-lagged technique, at both peak efficiency and close to surge operating conditions. A good agreement between the experiments and simulations were obtained, which justifies the use of the computational fluid dynamics results for the comparison of the flow field at both operating conditions (peak efficiency and near surge). Even if the change in flow field within the impeller from peak efficiency to near surge looked to be gradual, an overall rotation of the whole flow in the blade passages led to a non-homogeneous flow at the impeller exit in terms of angle and velocity level. Therefore, the vaned diffuser has to tolerate upstream flows, which are all the more distorted as the operating point moves towards surge.

Shock/Boundary-Layer/Tip-Clearance Interaction in a Transonic Rotor Blade

A newly designed rotor was modeled from the well-known radially stacked NASA rotor 37 by applying a three-dimensional shape to the original blade stacking line. A considerable curvature toward the direction of rotor rotation was given to the new blade. A three-dimensional numerical model, developed and validated using a commercial computational fluid dynamics Reynolds-averaged Navier―Stokes code, was adopted to predict the flowfield inside the new rotor. Steady-viscous-flow calculations were run at the design speed of the baseline configuration. Compared with rotor 37, the new rotor showed a higher efficiency, mainly due to a three-dimensional modification of the shock structure. At the outer span, the new rotor developed a blade-to-blade shock front located more downstream than in the baseline rotor, with a considerable impact on the flowfield near the casing. Computational fluid dynamics flow visualizations showed a less detrimental shock/boundary-layer/tip-clearance interaction at low-flow operating conditions, with a considerable reduction of the low-momentum-fluid region after the shock.

Experimental and computational investigation of an ‘open’ transonic cavity flow

This paper presents an investigation of a transonic flow ( M∞=0.85) over a rectangular cavity having a length-to-depth ratio of 5. Velocities were measured inside the cavity on the central plane and two off-centre planes using a two-component particle image velocimetry system. These measurements were supported by surface flow visualization, and mean and time-varying surface pressure measurements. The flow was also simulated using an unsteady Reynolds-averaged Navier—Stokes code, with a realizable k — ε turbulence model. It is shown that this CFD model does not capture all the characteristics of the flowfield correctly. However, by using this integrated experimental and computational approach we have been able to identify three-dimensional flowfield structures within the cavity. The influence of the thickness of the approaching boundary layer is discussed.

Computational Fluid Dynamics Validation for a Fin/Bulb/Winglet Keel Configuration

A wind tunnel test of an America's Cup keel model is used for validation of one Reynolds-averaged Navier Stokes (RANS) code and one potential flow/boundary layer code. The effects of grid size, stagnation point anomaly, and turbulence model on the RANS results are discussed. Various setups of the potential code are compared. The ability of both methods to predict forces and trends are shown. The errors of the RANS code are slightly larger than the experimental error, whereas the potential flow/boundary layer results are within the experimental uncertainty, provided that a correct panelization is used. A comparison of the experimental wake flow pattern to the one computed with RANS is presented. The k-w turbulence model is shown to give the best predictions of the wake.

Numerical simulation of vortex-induced vibration of a circular cylinder at low mass-damping using RANS code

Fundamental research on vortex-induced vibration (VIV) of a circular cylinder is still needed to build more rational VIV analysis tools for slender marine structures. Numerical results are presented for the response of an elastically mounted rigid cylinder at low mass damping constrained to oscillate transversely to a free stream. A two-dimensional Reynolds-averaged Navier–Stokes (RANS) code equipped with the SST k−ω turbulence model is applied for the numerical calculations. The numerical results are compared in detail with recent experimental and computational work. The Reynolds-averaging procedure erases the random disturbances in the vortex shedding process, so that the comparison between experimental data and the numerical results obtained by RANS codes may reveal some random characteristics of the VIV response. How random disturbance affects the observation in the experiments is discussed in this paper and the issues influencing the appearance of the upper branch in experiments are especially investigated. The absence of the upper branch in RANS simulations is explained in depth on account of discrepancies, which exist between experiments and RANS simulations. In addition, the formation of the 2P vortex shedding mode and its transition through the lock-in region are well reproduced in this investigation.

Computational study of effect of tabs on a jet in a cross flow

A three-dimensional viscous flow analysis is performed using a time-marching Reynolds-averaged Navier–Stokes code for the case of a jet in a cross flow with a delta tab placed on the windward side of the jet to inhibit mixing for film cooling applications. The flow configuration which was previously studied experimentally, involved a jet discharging normally from the floor of the wind tunnel test section into the cross flow with a momentum ratio (jet/cross-flow) of 36. The computed results are compared with the experimental data which include streamwise velocity and vorticity distributions at various axial locations downstream of the jet. The computational results show reasonably good agreement with the experimental data.

Computation of rotor wake turbulence noise

A major source of fan broadband noise is the interaction of rotor wake turbulence with the fan outlet guide vanes. A broadband noise model that utilizes computed rotor flow turbulence from a Reynolds averaged Navier–Stokes code is used to predict fan broadband noise spectra. The noise model is employed to examine the broadband noise characteristics of the 22-in source diagnostic test fan rig for which broadband noise data were obtained in wind tunnel tests at the NASA Glenn Research Center. A 9-case matrix of three outlet guide vane configurations at three representative fan tip speeds are considered. For all cases inlet and exhaust spectra of acoustic power are computed and compared with the measured spectra where possible. The acoustic power levels and shape of the predicted spectra are in good agreement with the measured data for the fan exhaust duct radiation at approach condition where direct comparisons are possible. The predicted spectra show the experimentally observed trends with fan tip speed, vane count and vane sweep. The results also demonstrate the validity of using computational fluid dynamics based turbulence information for fan broadband noise calculations.

Prediction of tip vortex cavitation inception using coupled spherical and nonspherical bubble models and Navier?Stokes computations

A spherical and a nonspherical bubble dynamics models were developed to study cavitation inception, scaling, and dynamics in a vortex flow. The spherical model is a modified Rayleigh–Plesset model to account for bubble slip velocity and for nonuniform pressures around the bubble. The nonspherical model is embedded in an unsteady Reynolds-averaged Navier–Stokes code with appropriate free-surface boundary conditions and a moving chimera grid scheme around the bubble. The effect of nonspherical deformation and bubble/flow interaction on bubble dynamics is illustrated by comparing spherical and nonspherical models. It is shown that nonspherical deformations and bubble/flow interactions are important for an accurate prediction of cavitation inception. The surface-averaged pressure-modified Rayleigh–Plesset scheme is a significant improvement over the conventional spherical model, and is able to capture the volume changes of a bubble during its capture. It is also a fast scheme for studying scaling. In a preliminary study, the scaling effects on cavitation inception were examined using two different Reynolds numbers owing to two different chord lengths. The nuclei-size effect on the prediction of cavitation inception was also studied, and its important effects are highlighted.

Computational aeroacoustic simulations of leading-edge slat flow

High-lift devices on modern airliners are a major contributor to overall airframe noise. In this paper the aeroacoustics of leading-edge slat devices in a high-lift configuration are investigated computationally. A hierarchical methodology is used to enable the rapid evaluation of different slat configurations. The overall goal is to gain a deeper understanding of the noise generation and amplification mechanisms in and around the slat, and the effects of slat system geometry. In order to perform parametric studies of the aeroacoustics, a simplified 2-D model of the slat is used. The flow and aeroacoustics are computed using a compressible, unsteady, Reynolds-Averaged Navier-Stokes code. A robust buffer zone boundary condition is used to prevent the reflection of outgoing acoustic waves from contaminating the long-time solution. A Ffowcs Williams–Hawkings solver is used to compute the far field acoustic field from the unsteady flow solution and determine the directivity. The spanwise correlation length used is derived from experimental data of this high-lift configuration. The effect of spanwise correlation length on the acoustic far field is examined. The aeroacoustics of the slat system are largely governed by the geometry, especially in terms of slat overlap. We perform a study of the effects of trailing edge thickness, horizontal and vertical overlap settings for the slat on near field wave propagation and far field directivity. The implications for low-noise leading edge slat design are discussed.

Navier-Stokes and Comprehensive Analysis Performance Predictions of the NREL Phase VI Experiment

A lifting-line code, CAMRAD II, and a Reynolds-Averaged Navier-Stokes code, OVERFLOW-D, were used to predict the aerodynamic performance of a two-bladed horizontal axis wind turbine. All computations were compared with experimental data that was collected at the NASA Ames Research Center 80-by-120-foot Wind Tunnel. Lifting-line computations were performed for both axial and yawed operating conditions while the Navier-Stokes computations were performed for only the axial conditions. Various stall delay models and dynamic stall models were used by the CAMRAD II code. For axial operating conditions, the predicted rotor performance varied significantly, particularly for stalled wind speeds. The lifting-line required the use of stall delay models to obtain the proper stall behavior, yet it still has difficulty in predicting the proper power magnitude in stall. The Navier-Stokes method captures the stall behavior and gives a detailed insight into the fluid mechanics of the stall behavior.