Turbulent Boundary Conditions Research Articles

New inflow boundary conditions for modelling the neutral equilibrium atmospheric boundary layer in computational wind engineering

Modelling neutral equilibrium atmospheric boundary layers (ABL) in CFD is an important aspect in computational wind engineering (CWE) applications. In this paper, new inflow boundary conditions are introduced from the viewpoint that these boundary conditions should satisfy the turbulence model employed. The new set of inflow turbulence boundary conditions is an approximate solution to the standard k– ε model transport equations. The capability of these boundary conditions to produce an equilibrium ABL is demonstrated by performing numerical simulations in an empty domain. The new inflow turbulence boundary conditions in this paper support future practical applications in CWE and future research in modelling equilibrium ABLs.

CFD modelling of unsteady cross ventilation flows using LES

Traditional methods using steady-state simulations with RANS models may predict adequate flow characteristics of cross ventilation, provided the selected turbulence model, boundary conditions and level of mesh resolution are appropriate. However, these methods are less satisfactory if we are more interested in the statistical description of fluctuating flow rate. This study used a LES method to investigate the fluctuating flow rate induced by wind with a CFD code capable of generating time-dependent and unsteady inflows. Simulations were carried out for two wind directions: normal to the openings (case 1) and parallel to the openings (case 2). The results were in good agreements with experimental data. For case 1, the observed flow rate was relatively steady, whereas for case 2, it exhibited very unsteady phenomena. Observations indicated that the flow pattern around the building significantly affected the ventilation flow rate. For case 1, mean flows were dominant in the air exchange process, whilst for case 2 the fluctuating flows appeared to have more influence on the ventilation mechanism.

Influences of equilibrium atmosphere boundary layer and turbulence parameter on wind loads of low-rise buildings

In this paper, the influences of two important problems in computational wind engineering (CWE), which are the modeling of equilibrium atmosphere boundary layer (ABL) and the specification of turbulence model parameters, on the numerical simulations of wind pressure distributions on a typical low-rise building are investigated sequentially through detailed comparisons between the numerical results and the wind tunnel test data. The capability of the proposed inflow turbulence boundary conditions in constructing equilibrium ABL is verified, and the effect of the turbulence parameters on the numerical results is illustrated. The combination of the carefully considered inflow boundary conditions, and the turbulence parameters can improve the numerical simulation accuracy of the wind pressures of the low-rise building. The present work seems to be helpful to further realize the importance of these two problems as well as to provide a referential study to the accuracy improvement of CWE.

Large eddy simulation of a square cylinder flow: Modelling of inflow turbulence

The present study aims to generate turbulent inflow data to more accurately represent the turbulent flow around a square cylinder when the inflow turbulence level is significant. The modified random flow generation (RFG) technique in conjunction with a previously developed LES code is successfully adopted into a finite element based fluid flow solver to generate the required inflow turbulence boundary conditions for the three-dimensional (3-D) LES computations of transitional turbulent flow around a square cylinder at Reynolds number of 22,000. The near wall region is modelled without using wall approximate conditions and a wall damping coefficient is introduced into the calculation of sub-grid length scale in the boundary layer of the cylinder wall. The numerical results obtained from simulations are compared with each other and with the experimental data for different inflow turbulence boundary conditions in order to discuss the issues such as the synthetic inflow turbulence effects on the 3-D transitional flow behaviour in the near wake and the free shear layer, the basic mechanism by which stream turbulence interacts with the mean flow over the cylinder body and the prediction of integral flow parameters. The comparison among the LES results with and without inflow turbulence and the experimental data emphasizes that the turbulent inflow data generated by the present RFG technique for the LES computation can be a viable approach in accurately predicting the effects of inflow turbulence on the near wake turbulent flow characteristics around a bluff body.

Particle deposition in turbulent duct flows—comparisons of different model predictions

Numerical studies of transport and deposition of nano- and micro-particles in turbulence flow field have been studied in the past few decades. In most current industrial applications, Reynolds averaged turbulence models were used due to its relative simplicity and computational efficiency. In this work, a series of numerical simulations were conducted to study the transport and deposition of nano- and micro-particles in a turbulent duct flow using different turbulence models. Commercial software (FLUENT TM 6.1.22) was used for turbulence mean flow simulation. Simulations of the instantaneous turbulence fluctuation with and without turbulence near wall correction, and particle trajectory analysis were performed with the in-house PARTICLE (object-oriented C++) code, as well as with FLUENT TM code with and the use of user's defined subroutines. The simulation results for different cases were compared with the available experimental data, and the accuracy of various approaches was evaluated. In addition, the importance of turbulence model, boundary conditions, and turbulence fluctuation particularly near wall on particle transport and deposition were carefully evaluated. It was shown that when sufficient care was given to the modeling effort, the particle deposition rates could be predicted with reasonable accuracy. The presented results could provide guidelines for selecting appropriate procedure for simulating nano- and micro-particle transport and deposition in various applications.

A database of PIV measurements within a turbomachinery stage and sample comparisons with unsteady RANS

This paper describes an experimental database obtained using two-dimensional Particle Image Velocimetry (PIV) measurements within an axial turbomachinery stage, and presents sample unsteady Reynolds Averaged Navier–Stokes (RANS) simulations to illustrate its applicability for turbomachinery model validation. The experiments are performed in a refractive-index-matched facility that provides unobstructed view, and cover the entire second stage of a two-stage axial pump. The data were obtained at ten different rotor phases covering one rotor blade-passing period, and at mid-span. Several features of the data at selected phases have already been presented and discussed in prior publications. Here we present the complete database together with sample CFD results. Two-dimensional unsteady RANS simulations are performed using the commercial flow solver FLUENT™, with two standard turbulence models, i.e. Renormalization Group (RNG) k-ϵ and Reynolds Stress Transport Model (RSM). The spatially non-uniform inlet velocity and turbulence boundary conditions are provided from the experimental data. Detailed side-by-side comparisons of measured and computed velocity as well as turbulence fields within the entire stage are presented using line distributions within the rotor–stator gap and the stator wake regions, as well as whole-field animations. The results show that, although there is reasonable agreement, in general, between the experimental results and the computational simulations, some critical flow features are not correctly predicted. The turbulent kinetic energy levels are generally too high in the simulations, with substantial amount of unphysical turbulence generation near the blade leading edges, especially in the case of RNG k-ϵ model. Also, wake diffusion is underestimated. The results highlight the usefulness of comparisons that cover the entire unsteady flow in a passage, as afforded by the present database and side-by-side animations.

Modelling of Effect of Inflow Turbulence on Large Eddy Simulation of Bluff Body Flows

A random flow generation (RFG) algorithm for a previously established large eddy simulation (LES) code is successfully incorporated into a finite element fluid flow solver to generate the required inflow/initial turbulence boundary conditions for the LES computations of viscous incompressible turbulent flow over a nominally twodimensional circular cylinder at Reynolds number of 140,000. The effect of generated turbulent inflow boundary conditions on the near wake flow and the shear layer and on the prediction of integral flow parameters is studied based on long time average results. No-slip velocity boundary function is used but wall effects are taken into consideration with a near wall modelling methodology based on van Driest Damping approach. The numerical results obtained from simulations are compared with each other and with the experimental data for different turbulent inflow boundary conditions to assess the functionality of the RFG algorithm for the present LES code and hence its influence on the vortex shedding mechanism and the resulting flow field predictions.

Modeling of Effect of Inflow Turbulence Data on Large Eddy Simulation of Circular Cylinder Flows

A random flow generation (RFG) algorithm for a previously established large eddy simulation (LES) code is successfully incorporated into a finite element fluid flow solver to generate the required inflow/initial turbulence boundary conditions for the three-dimensional (3D) LES computations of viscous incompressible turbulent flow over a nominally two-dimensional (2D) circular cylinder at Reynolds number of 140,000. The effect of generated turbulent inflow boundary conditions on the near wake flow and the shear layer and on the prediction of integral flow parameters is studied based on long time average results. Because the near-wall region cannot be resolved for high Reynolds number flows, no-slip velocity boundary function is used, but wall effects are taken into consideration with a near-wall modeling methodology that comprises the no-slip function with a modified form of van Driest damping approach to reduce the subgrid length scale in the vicinity of the cylinder wall. Simulations are performed for a 2D and a 3D configuration, and the simulation results are compared to each other and to the experimental data for different turbulent inflow boundary conditions with varying degree of inflow turbulence to assess the functionality of the RFG algorithm for the present LES code and, hence, its influence on the vortex shedding mechanism and the resulting flow field predictions.

Secondary Flow Near an Undulatory Surface Induced by Wall Blocking Effect

Prandtl's secondary mean motions of the second kind near an undulating surface were explained in terms of turbulent blocking effect and kinematic boundary conditions at the surface, and its order of magnitude was estimated. Isotropic turbulence is distorted by the undulating surface of wavelength λ and amplitude h with a low slope, so that h<<λ. The prime mechanism for generating the mean flow is that the far-field isotropic turbulence is distorted by the non-local blocking effect of the surface to become anisotropic axisymmetric turbulence near the surface with principal axis that is not aligned with the local curvature of the undulation. Then the local analysis can be applied and the mechanism is similar to the mean flow generation mechanism for homogeneous axisymmetric turbulence over a planer surface, i.e. gradients of the Reynolds stress caused by the turbulent blocking effect generate the mean motions. The results from this simple analysis are consistent with previous exact analysis in which the effects of curvature are strictly taken into account. The results also qualitatively agree with flow visualization over an undulating surface in a mixing-box.

Comparison between Two Turbulence Models and Analysis of the Effect of the Substrate Movement on the Flow Field of a Plasma Jet

In the first part of the present study, an appropriate inflow turbulent boundary condition is chosen. Then, a comparison is made between two turbulence models for a plasma jet discharged into air atmosphere. The plasma jet gas phase flow is predicted with the standard k–ɛ model and the RNG model of turbulence. Particles behavior is modeled using stochastic particles trajectories. A validation of the plasma jet model is made by comparison with experimental data. This part of the study shows that the flow features are better predicted with the RNG model. The choice of appropriate boundary conditions seems to be crucial for a better simulation of plasma thermal spraying. Afterwards, computations are performed for projection of Ni particles. It is found that the computed particles velocities and temperatures are also better predicted with the RNG model compared with the k–ɛ model. The second part of this study is concerned with the effect of the substrate movement on the gas flow field. This is performed in order to simulate a realistic coatings process where a relative movement between the torch and the substrate always exists. Three substrate velocities have been used and it is found that the flow fields are affected only very near the substrate wall.

Evaluation of boundary conditions used to model dilute, turbulent gas/solids flows in a pipe

A turbulent gas/solids model, based on the work of Simonin [1] [Simonin, O., 1996. Continuum modeling of dispersed two-phase flows, in Combustion and Turbulence in Two-Phase Flows, Von Karman Institute of Fluid Dynamics Lecture Series 1996-2], has been recently implemented in the MFIX computational fluid dynamic (CFD) code. This theory includes the effects of turbulence in the gas phase as well as inter-particle collisions. The extension of this theory [2] [Balzer, G., Simonin, O., Boelle, A., Lavieville, J., 1996. A unifying modelling approach for the numerical prediction of dilute and dense gas-solid two phase flow, CFB5, 5th Int. Conf. on Circulating Fluidized Beds, Beijing, China] to dense gas/solids systems was made possible by including the kinetic theory of granular material to describe the solids stresses. The turbulence model and boundary conditions were evaluated by simulating the gas/solids flow experiments of Jones [3] [N.E. Jones, An experimental investigation of particle size distribution effect in dilute phase gas–solid flow, Ph.D. thesis, Purdue University (2001)]. Their experimental results included velocity and turbulence measurements for fully developed flows for a range of particle loading from very dilute to relatively dense. Our numerical calculations were conducted by imposing periodic boundary conditions as well as in a long pipe with different length-to-diameter ratios to achieve a fully developed condition. We propose modifications to the single-phase wall functions, to include the effect of the particulate phase. However, these modifications had only a minor effect on the predictions of gas turbulent kinetic energy due to the dilute nature of the flow considered in this study. The turbulent gas/solids flow model based on the work of Simonin [1] [Simonin, O., 1996. Continuum modeling of dispersed two-phase flows, in Combustion and Turbulence in Two-Phase Flows, Von Karman Institute of Fluid Dynamics Lecture Series 1996-2] is able to predict reasonably well dilute gas/solids flows with appropriate boundary conditions (BC). Four different types of boundary conditions were investigated to assess their sensitivity. The experimental data fall between the large-friction/no-sliding and small-friction/all-sliding limits of Jenkins and Louge [4] [J.T. Jenkins, M.Y. Louge, On the flux of fluctuating energy in a collisional grain flow at a flat frictional wall, Phys. Fluids 9 (10), (1997) 2835–2840] BC. However, the physical behavior of the particle–wall interactions is close to the small-friction/all-sliding limit of Jenkins and Louge BC or the Johnson and Jackson [5] [P.C. Johnson, R. Jackson, Frictional-collisional constitutive relations for granular materials, with application to plane shearing. J. Fluid Mech. 176 (1987) 67–93]BC with a small specularity coefficient or simply the free-slip BC.

Reduced-basis modeling of turbulent plane channel flow

Very-large-eddy simulations of turbulent plane channel flow of various Reynolds numbers have been performed. The simulations are based on relatively small sets of modes from a proper orthogonal decomposition (POD) of a direct numerical simulation or large-eddy simulation of the same flow. The intention is to investigate how these reduced basis simulations respond to change in Reynolds number and geometry. We find that the dynamical systems are able to reproduce the mean velocity and Reynolds stress profiles as long as the aspect ratio of the channel or the Reynolds number does not increase or decrease from the original setting with more than 50% or 30%, respectively. When the objective is a coarse description of the flow dynamics, significant savings in terms of computational cost are achieved by deriving the dynamical systems from the POD-basis. One possible application of such reduced basis simulations can be to efficiently generate turbulent inlet boundary conditions for direct numerical simulations.

CFD Study of the Thermal Environment around a Human Body: A Review

Computational fluid dynamics (CFD) simulations of thermal manikins can provide detailed information which it is impossible to obtain from experiments. In this review several issues from CFD studies of thermal manikins are discussed, such as geometric complexity of computational thermal manikins (CTMs), selection of turbulence models, grid generation and boundary conditions. Calculation of CTMs with simple geometry requires less computational resource but a more anthropomorphic geometry can provide more accurate results. In the simulations using complex geometric representations of the human body the computational domain is discretised by body-fitted co-ordinates or unstructured grids. The total number of cells can reach the quantitative magnitude of 1 million. The simulation results about airflow field, radiative and convective heat transfer and contaminant distribution from different researchers are compared. Further developments in the use of these simulations are proposed.

Reduced basis simulations as a tool for generating turbulent inlet‐data for two opposing jets

AbstractReduced basis simulations are used in order to generate turbulent inlet boundary conditions for a direct numerical simulation of two opposing wall jets. The two entering jets are fully turbulent channel flows of Reynolds number 180, based on friction velocity and half channel height. The inflow turbulence is generated by solving dynamical equations for the large scales only, while less energetic small scales are added randomly. The proper orthogonal decomposition method is used to identify the large scale modes used in the basis set. Mean velocity and turbulence statistics have been reported in various planes of the interaction region and the outwash jet. The increase in maximum turbulence energy from the inlet jet to the stagnation region is about 500%. The spreading rate of the merged jet was found to be 0.125. Copyright © 2005 John Wiley &amp; Sons, Ltd.

Analysis of the Immediate Boundary Conditions of an Axial Flow Impeller

The design and analysis of impellers for mixing applications such as stirred tank reactors (STR) have significant application in many engineering processes. In particular, the evaluation of the turbulent flow boundary conditions immediately around a model impeller has been examined using flow visualization, laser Doppler anemometry (LDA) and particle image velocimetry (PIV). Mean and fluctuating velocities are presented for all three velocity components. Good agreement between LDA and PIV measurements was obtained once an optimized PIV arrangement was defined. Turbulence parameters such as kinetic energy, dissipation, and length scales are estimated from the measurements. Several approaches to evaluating dissipation were used. Triggered PIV measurements were used to evaluate the variation in radial and tangential velocities in the entire planes above and below the impeller at constant axial distances. The size and orientation of the vortices shed from the impeller blade tips and the corresponding regions of high kinetic energy and dissipation indicate the importance of these structures in mixing.

AN INTEGRATED NONLINEAR APPROACH FOR TURBOMACHINERY FORCED RESPONSE PREDICTION. PART I: FORMULATION

This paper describes the formulation of an advanced numerical model for the simulation of high- and low-engine-order forced response for turbomachinery applications. The various forced response mechanisms are explained in some detail and a specification for an accurate prediction system is discussed with emphasis on both fluid and structural modelling aspects. The Favre-averaged Navier–Stokes equations are used to represent the unsteady flow in a nonlinear time-accurate fashion. Features such as turbulence modelling, boundary conditions, meshing strategies and numerical treatments are discussed in detail. The structural model is based on a linear modal model, though local nonlinearities due friction dampers can be accommodated using an iterative scheme. The fluid mesh is moved at each-time step according to the structural motion, so that changes in blade aerodynamic damping and flow unsteadiness can be accommodated. It is concluded that the model can be used for large simulations involving multi-bladerow whole-annulus calculations.

Treatment of turbulent wall boundary conditions using linear-logarithmic elements

In the present paper special wall elements are introduced in order to implement near-wall boundary conditions accurately. These linear-logarithmic wall elements allow a no-slip wall condition with less resolution requirements than would else be the case. This approach is combined with a modified two-layer turbulence model, and applications to some two-dimensional cases are shown to illustrate the potential of the method.

A two‐dimensional numerical model which simulates the temperature, salinity and velocity fields in knight inlet, British Columbia

A two‐dimensional laterally integrated model of Knight Inlet, British Columbia is used to simulate the time‐varying temperature, salinity and velocity fields. The model includes the influence of the wind, tides, solar heating and freshwater inflow. The level 2.5 turbulent closure scheme of Mellor and Yamada is used to parametrize the diffusion terms in the model. In this paper, two turbulent kinetic energy boundary conditions at the air‐water interface are compared. One boundary condition is for an enhanced surface turbulent layer due to the flux of wind energy while the other is based on the Law of the Wall approach. The model is compared to data collected in the summer of 1989 from four moored vertical arrays. Salinity, temperature, horizontal current velocity and wind velocity were measured for about thirty days. Measurements were made throughout the water column so that the thin surface layer and deep water flows could be resolved. A moored array near the mouth of the inlet was used to provide open boundary data for the model, while the remaining data were compared to the simulated data. The modelled salinity fields were very realistic particularly when the enhanced turbulent surface layer boundary condition was used. The temperature fields were well modelled as long as the surface temperature was prescribed. The simulated velocities were very similar to those produced by a numerical model which simulated density (with the pressure effect removed) directly, instead of solving for temperature and salinity and then determining density via an equation of state.

Anchored CCD for Gas Turbine Combustor Design and Data Correlation

Correlations based on design database, combined with multidimensional computational combustion dynamics (CCD) models are used in the combustion design process. However, because of limitations in the current turbulent combustion models, numerics, and boundary conditions, CCD has provided mainly qualitative trends for aerothermal performance, emissions, and liner wall temperature levels and gradients. To overcome these deficiencies, hybrid modeling approaches have been proposed to analyze existing combustors. A typical hybrid modeling approach combines empirical and semianalytical correlations with CCD to give quantitatively accurate predictions of NOx, CO, HC, smoke, lean blowout, ignition, pattern factor, and liner wall temperatures. An alternate approach, anchored CCD, is described in this paper. First, the models were anchored with one of the five modern turbopropulsion engine combustors. The anchored models were then run for the other four combustors. The predicted results correlated well with measured NOx, CO, HC, LEO, and exit temperature quality data, demonstrating a broader applicability of the anchored method. The models were also used for designing a new combustion concept. The pretest prediction agreed well with sector rig data from development hardware, showing the feasibility of using the anchored methodology as a design tool.

Flow separation and hydraulic transitions over hills modelled by the Reynolds equations

A model based upon Reynolds equations with a standard (K, e)-turbulence closure and boundary conditions may predict important features of flows with separation, attachment, hydraulic transitions and internal wave breaking quite robustly and realistically.