On the Grid Sensitivity of the Wall Boundary Condition of the k-ω Turbulence Model

2004 ◽  
Vol 126 (6) ◽  
pp. 900-910 ◽  
Author(s):  
L. Ec¸a ◽  
M. Hoekstra
Keyword(s):  
Boundary Condition ◽  
Turbulence Model ◽  
Eddy Viscosity ◽  
Solid Wall ◽  
Turbulence Models ◽  
Numerical Implementation ◽  
Two Dimensional ◽  
Grid Refinement ◽  
Wall Boundary Condition ◽  
Order Of Magnitude

This paper presents a study on the k-ω turbulence model with regard to the numerical implementation of the ω boundary condition at a solid wall, where ω tends to infinity. Three different implementations are tested in the calculation of a simple two-dimensional turbulent flow over a flat plate. Grid refinement studies in grids with different near-wall grid line spacings are performed to assess the numerical uncertainty of the predicted drag coefficient CD. The results are compared with the predictions of several alternative algebraic, one-equation, and two-equation eddy-viscosity turbulence models. For the same level of grid refinement, the estimated uncertainty of CD obtained with the k-ω model is one order of magnitude larger than for all the other models.

Turbulent Flow Using a Modified V2f Turbulence Model

2004 ◽  
Author(s):  
Hanif Montazeri ◽  
Siamak Kazemzadeh Hannani ◽  
Bijan Farhanieh
Keyword(s):  
Boundary Condition ◽  
Turbulence Model ◽  
Flow Problem ◽  
Two Dimensional ◽  
Convergence Problem ◽  
New Model ◽  
Step Flow ◽  
Wall Boundary Condition ◽  
Proposed Model ◽  
Accuracy Of Results

An improved version of the V2f turbulence model has been examined in this paper. The objective was to overcome the convergence problem encountered in the original V2f model. The convergence problem is due to the commonly-used wall boundary condition, which therefore has been modified in the proposed model. To test the soundness of the new model, several two-dimensional cases such as Poiseuille flow, channel flow, and backward-step flow has been analyzed and the results are compared with the standard k-ε model, DNS, and in case of the backward flow problem, also with the original V2f model. Based on the comparison, the new model presents a promising approach both with respect to convergence as well as the accuracy of results.

Temperature Corrected Turbulence Model for High Temperature Jet Flow

2004 ◽  
Vol 126 (5) ◽  
pp. 844-850 ◽  
Author(s):  
Khaled S. Abdol-Hamid ◽  
S. Paul Pao ◽  
Steven J. Massey ◽  
Alaa Elmiligui
Keyword(s):  
High Temperature ◽  
Turbulence Model ◽  
Room Temperature ◽  
Eddy Viscosity ◽  
Mixing Layer ◽  
Supersonic Jet ◽  
Turbulence Models ◽  
Jet Flows ◽  
Viscosity Term ◽  
Mixing Layer Flows

It is well known that the two-equation turbulence models under-predict mixing in the shear layer for high temperature jet flows. These turbulence models were developed and calibrated for room temperature, low Mach number, and plane mixing layer flows. In the present study, four existing modifications to the two-equation turbulence model are implemented in PAB3D and their effect is assessed for high temperature jet flows. In addition, a new temperature gradient correction to the eddy viscosity term is tested and calibrated. The new model was found to be in the best agreement with experimental data for subsonic and supersonic jet flows at both low and high temperatures.

Evaluation of Turbulence Models Using Direct Numerical and Large-Eddy Simulation Data

2011 ◽  
Vol 133 (2) ◽  
Author(s):  
Hassan Raiesi ◽  
Ugo Piomelli ◽  
Andrew Pollard
Keyword(s):  
Boundary Layer ◽  
Large Eddy Simulation ◽  
Eddy Viscosity ◽  
Predictive Accuracy ◽  
Turbulence Models ◽  
Special Treatment ◽  
Two Dimensional ◽  
Square Duct ◽  
Eddy Simulation ◽  
Large Eddy

The performance of some commonly used eddy-viscosity turbulence models has been evaluated using direct numerical simulation (DNS) and large-eddy simulation (LES) data. Two configurations have been tested, a two-dimensional boundary layer undergoing pressure-driven separation, and a square duct. The DNS and LES were used to assess the k−ε, ζ−f, k−ω, and Spalart–Allmaras models. For the two-dimensional separated boundary layer, anisotropic effects are not significant and the eddy-viscosity assumption works well. However, the near-wall treatment used in k−ε models was found to have a critical effect on the predictive accuracy of the model (and, in particular, of separation and reattachment points). None of the wall treatments tested resulted in accurate prediction of the flow field. Better results were obtained with models that do not require special treatment in the inner layer (ζ−f, k−ω, and Spalart–Allmaras models). For the square duct calculation, only a nonlinear constitutive relation was found to be able to capture the secondary flow, giving results in agreement with the data. Linear models had significant error.

Assessment of Various Inviscid-Wall Boundary Conditions: Applications to NACA65 Compressor Blade

2011 ◽  
Author(s):  
Habib Khazaei ◽  
Ali Madadi ◽  
Mohammad Jafar Kermani
Keyword(s):  
Boundary Condition ◽  
Boundary Conditions ◽  
Solid Wall ◽  
Inviscid Flow ◽  
Compressor Blade ◽  
The Body ◽  
Wall Boundary Condition ◽  
Velocity Magnitude ◽  
Wall Boundary ◽  
Wall Boundary Conditions

Boundary condition is one of the major factors to influence the numerical stability and solution accuracy in numerical analysis. One of the most important physical boundary conditions in the flow field analysis is the wall boundary condition imposed on the body surfaces. To solve a three-dimensional compressible Euler equation (with five coupled PDE’s), totally five boundary conditions at the body surfaces should be prescribed. The momentum equation in the direction normal to the inviscid solid wall provides the pressure at the surface of the wall. For the cases with no-heat source or sink, the total temperature at the wall and the incoming flow should remain constant, when the steady condition is prevailed. The no-penetration condition through the solid wall and slip condition provides an equation relating the three velocity components. Assuming identical flow direction at the wall with the adjacent node, the last thing is the velocity magnitude that should be cast in such a way to give accurate, stable and robust solution. In this paper, four different methods for calculation of the wall velocity magnitude are proposed and applied to an identical test case of subsonic and supersonic flows such as: (1) Inviscid flow in a 3D converging-diverging nozzle, (2) Inviscid subsonic flow in a single 90° elbow, (3) Inviscid supersonic flow over a wedge, and (4) Inviscid flow through a compressor blade geometry of NACA 65410. A recently implemented 3D in-house CFD code (based on the flux difference splitting scheme of Roe (1981)) is used to compute compressible flows in generalized coordinates. It is found that the way to specify the additional numerical wall boundary condition strongly affects the overall stability and accuracy of the solution. It is concluded that there is no best boundary condition to cover all of the test cases, but the best wall boundary condition should be introduced very carefully for each type of flow.

Computational Fluid Dynamics Study for Improvement of Prediction of Various Thermally Stratified Turbulent Boundary Layers

2017 ◽  
Vol 139 (5) ◽  
Author(s):  
Hirofumi Hattori ◽  
Tomoya Houra ◽  
Amane Kono ◽  
Shota Yoshikawa
Keyword(s):  
Boundary Layers ◽  
Turbulence Model ◽  
Turbulence Models ◽  
Eddy Diffusivity ◽  
Turbulent Boundary Layers ◽  
Heat Fluxes ◽  
Shear Stresses ◽  
Turbulent Heat ◽  
Two Dimensional ◽  
Turbulent Statistics

The objectives of this study are to reconstruct a turbulence model of both the large Eddy simulation (LES) and the Reynolds-averaged Navier–Stokes simulation (RANS) which can predict wind synopsis in various thermally stratified turbulent boundary layers over any obstacles. Hence, the direct numerical simulation (DNS) of various thermally stratified turbulent boundary layers with/without forward-step, two-dimensional block, or two-dimensional hill is carried out in order to obtain detailed turbulent statistics for the construction of a database for the evaluation of a turbulence model. Also, DNS clearly reveals the characteristics of various thermally stratified turbulent boundary layers with/without forward-step, two-dimensional block, or two-dimensional hill. The turbulence models employed in LES and RANS are evaluated using the DNS database we obtained. In the LES, an evaluated turbulence model gives proper predictions, but the quantitative agreement of Reynolds shear stress with DNS results is difficult to predict. On the other hand, the nonlinear eddy diffusivity turbulence models for Reynolds stress and turbulent heat flux are also evaluated using DNS results of various thermally stratified turbulent boundary layers over a forward-step in which the turbulence models are evaluated using an a priori method. Although the evaluated models do not make it easy to properly predict the Reynolds shear stresses in all cases, the turbulent heat fluxes can be qualitatively predicted by the nonlinear eddy diffusivity for a heat turbulence model. Therefore, the turbulence models of LES and RANS should be improved in order to adequately predict various thermally stratified turbulent boundary layers over an obstacle.

scholarly journals NEW METHODS FOR STABILIZING RANS TURBULENCE MODELS WITH APPLICATION TO LARGE SCALE BREAKING WAVES

2020 ◽  
pp. 19
Author(s):  
Georgios Azorakos ◽  
Bjarke Eltard Larsen ◽  
David R. Fuhrman
Keyword(s):  
Turbulence Model ◽  
Large Scale ◽  
Eddy Viscosity ◽  
Turbulence Models ◽  
Breaking Waves ◽  
Wide Spread ◽  
Cfd Simulations ◽  
New Methods ◽  
Rans Turbulence Models ◽  
Scale Breaking

Recently, Larsen and Fuhrman (2018) have shown that seemingly all commonly used (both k-omega and k-epsilon variants) two-equation RANS turbulence closure models are unconditionally unstable in the potential flow beneath surface waves, helping to explain the wide-spread over-production of turbulent kinetic energy in CFD simulations, relative to measurements. They devised and tested a new formally stabilized formulation of the widely used k-omega turbulence model, making use of a modified eddy viscosity. In the present work, three new formally-stable k-omega turbulence model formulations are derived and tested in CFD simulations involving the flow and dynamics beneath large-scale plunging breaking waves.Recorded Presentation from the vICCE (YouTube Link): https://youtu.be/T2fFRgq3I8E

Measurement of a Recirculating, Two-Dimensional, Turbulent Flow and Comparison to Turbulence Model Predictions. I: Steady State Case

1983 ◽  
Vol 105 (4) ◽  
pp. 439-446 ◽  
Author(s):  
D. R. Boyle ◽  
M. W. Golay
Keyword(s):  
Steady State ◽  
Turbulent Flow ◽  
Turbulence Model ◽  
Turbulence Models ◽  
Scale Model ◽  
Test Facility ◽  
Two Dimensional ◽  
Mean Velocity ◽  
Flow Area ◽  
Flow Measurements

Turbulent flow measurements have been performed in a two-dimensional flow cell which is a 1/15-scale model of the Fast Flux Test Facility nuclear reactor outlet plenum. In a steady water flow, maps of the mean velocity field, turbulence kinetic energy, and Reynolds stress have been obtained using a laser doppler anemometer. The measurements are compared to numerical simulations using both the K–ε and K–σ two-equation turbulence models. A relationship between K–σ and K–ε turbulence models is derived, and the two models are found to be nearly equivalent. The steady-state mean velocity data are predicted well through-out most of the test cell. Calculated spatial distributions of the scalar turbulence quantities are qualitatively similar for both models; however, the predicted distributions do not match the data over major portions of the flow area. The K–σ model provides better estimates of the turbulence quantity magnitudes. The predicted results are highly sensitive to small changes in the turbulence model constants and depend heavily on the levels of inlet turbulence. However, important differences between prediction and measurement cannot be significantly reduced by simple changes to the model’s constants.

Turbulence Input Parameters Correction Methodology in Water Lubricated Thrust Bearings

2018 ◽  
Author(s):  
Xin Deng ◽  
Harrison Gates ◽  
Brian Weaver ◽  
Houston Wood ◽  
Roger Fittro
Keyword(s):  
Shear Stress ◽  
Wall Shear Stress ◽  
Film Thickness ◽  
Turbulence Model ◽  
Thermal Deformation ◽  
Eddy Viscosity ◽  
Turbulence Models ◽  
Water Lubrication ◽  
Low Viscosity ◽  
Eddy Viscosity Model

Oil-lubricated bearings are widely used in high speed rotating machines such as those found in the aerospace and automotive industries. However, environmental issues and risk-averse operations are resulting in the removal of oil and the replacement of all sealed oil bearings with reliable water-lubricated bearings. Due to the different fluid properties between oil and water, the low viscosity of water increases Reynolds numbers drastically and therefore makes water-lubricated bearings prone to turbulence effects. This requires finer meshes when compared to oil-lubricated bearings as the low-viscosity fluid produces a very thin lubricant film. Analyzing water-lubricated bearings can also produce convergence and accuracy issues in traditional oil-based analysis codes. Thermal deformation largely affects oil-lubricated bearings, while having limited effects on water lubrication; mechanical deformation largely affects water lubrication, while its effects are typically lower than thermal deformation with oil. One common turbulence model used in these analysis tools is the eddy-viscosity model. Eddy-viscosity depends on the wall shear stress, therefore effective wall shear stress modeling is necessary in determining an appropriate turbulence model. Improving the accuracy and efficiency of modeling approaches for eddy-viscosity in turbulence models is of great importance. Therefore, the purpose of this study is to perform mesh refinement for water-lubricated bearings based on methodologies of eddy-viscosity modeling to improve their accuracy. According to Szeri [1], εm/v for the Boussinesq hypothesis is given by Reichardt’s formula. Fitting the velocity profile with experiments having a y+ in the range of 0–1,000 results in Ng-optimized Reichardt’s constants k = 0.4 and δ+ = 10.7. He clearly states that for y+ > 1000 theoretical predictions and experiments have a greater variance. Armentrout and others [2] developed an equation for δ+ as a function of the pivot Reynolds number, which they validated with CFD simulations. The definition of y+ can be used to approximate the first layer thickness calculated for a uniform mesh. Together with Armentrout’s equation, the number of required elements across the film thickness can be obtained. For typical turbulence models, the y+ must be within a certain range to be accurate. On the condition that the y+ is fixed to that of a standard oil bearing for which an oil bearing code was validated, the number of elements across the film thickness and coefficients used in the eddy-viscosity equation can be adjusted to allow for convergence with other fluids other than that which the traditional oil bearing code was designed for. In this study, the number of required elements across the film for improved prediction quality was calculated based on the proposed eddy-viscosity model mesh correction from the known literature. A comparison between water lubrication using the parameter correction and oil lubrication was also made. The results of this study could aid in improving future designs and models of water-lubricated bearings.

An improved turbulence model for predicting unsteady cavitating flows in centrifugal pump

2015 ◽  
Vol 25 (5) ◽  
pp. 1198-1213 ◽  
Author(s):  
Jian Wang ◽  
Yong Wang ◽  
Houlin Liu ◽  
Haoqin Huang ◽  
Linglin Jiang
Keyword(s):  
Centrifugal Pump ◽  
Turbulence Model ◽  
Eddy Viscosity ◽  
Turbulent Viscosity ◽  
Flow Characteristics ◽  
Turbulence Models ◽  
Content Type ◽  
Cavitation Inception ◽  
Cavitation Performance ◽  
Unsteady Rans

Purpose – The purpose of this paper is to study the unsteady caivitating flows in centrifugal pump, especially for improving the turbulence model to obtain highly resolution results-capable of predicting the cavitation inception, shedding off and collapse procedures. Design/methodology/approach – Both numerical simulations and experimental visualizations were performed in the present paper. An improved RCD turbulence models was proposed by considering three corrected methods: the rotating corrected method, the compressible corrected method and the turbulent viscosity corrected method. Unsteady RANS computations were conducted to compare with the experiments. Findings – The comparison of pump cavitation performance showed that the RCD turbulence model obtained better performance both in non-cavitation and cavitation conditions. The visualization of the cavitation evolution was recorded to validate the unsteady simulations. Good agreement was noticed between calculations and visualizations. It is indicated the RCD model can successfully capture the bubbles detachment and collapse at the rear of the cavity region, since it effectively reduces the eddy viscosity in the multiphase region of liquid and vapor. Furthermore, the eddy viscosity, the instantaneous pressure and density distribution were investigated. The effectiveness of the compressibility was found. Meanwhile, the influence of the rotating corrected method on prediction was explored. It is found that the RCD model solved more unsteady flow characteristics. Originality/value – The current work presented a turbulence model which was much more suitable for predicting the cavitating flow in centrifugal pump.

scholarly journals Prediction of Heat and Mass Transfer in a Rotating Ribbed Coolant Passage With a 180 Degree Turn

1998 ◽  
Author(s):  
David L. Rigby
Keyword(s):  
Heat Transfer ◽  
Mass Transfer ◽  
Boundary Condition ◽  
Turbulence Model ◽  
Heat Mass Transfer ◽  
Stokes Equations ◽  
Three Dimensional ◽  
Reynolds Numbers ◽  
Wall Boundary Condition ◽  
Wall Boundary

Numerical results are presented for flow in a rotating internal passage with a 180 degree turn and ribbed walls. Reynolds numbers ranging from 5200 to 7900, and Rotation numbers of 0.0 and 0.24 were considered. The straight sections of the channel have a square cross section, with square ribs spaced one hydraulic diameter (D) apart on two opposite sides. The ribs have a height of 0.1D and are not staggered from one side to the other. The full three dimensional Reynolds Averaged Navier-Stokes equations are solved combined with the Wilcox k-ω turbulence model. By solving an additional equation for mass transfer, it is possible to isolate the effect of buoyancy in the presence of rotation. That is, heat transfer induced buoyancy effects can be eliminated as in naphthalene sublimation experiments. Heat transfer, mass transfer and flow field results are presented with favorable agreement with available experimental data. It is shown that numerically predicting the reattachment between ribs is essential to achieving an accurate prediction of heat/mass transfer. For the low Reynolds numbers considered, the standard turbulence model did not produce reattachment between ribs. By modifying the wall boundary condition on ω, the turbulent specific dissipation rate, much better agreement with the flow structure and heat/mass transfer was achieved. It is beyond the scope of the present work to make a general recommendation on the ω wall boundary condition. However, the present results suggest that the ω boundary condition should take into account the proximity to abrupt changes in geometry.

Sign in / Sign up

Export Citation Format

Share Document