You are on page 1of 9

Introduction

Turbulent motion is irregular. The irregularity is manifested through complex variations of


velocity, temperature, etc. with space and time. The irregular motion is generated due to random
fluctuations. At a Reynolds number less than critical, the kinetic energy of flow is not enough to
sustain the random fluctuation against the viscous damping and in such laminar flow continues to
exits. At somewhat higher Reynolds number than the critical Reynolds number, the kinetic
energy of flow supports the growth of fluctuations and transition to turbulence is induced. The
turbulence promotes improved mixing. High transfer rate of momentum, heat, and mass by
fluctuating turbulent motion, are practically most important feature of turbulence.
Turbulent motion carries vorticity which is composed of eddies interacting with each other. At
large Reynolds numbers, there exists a continuous transport of energy from the free stream to
large eddies. From the large eddies a series of increasingly smaller eddies are formed. The
smallest eddies dissipate energy and destroy themselves. The smaller eddies are influenced by
the strain rate imposed by the large eddies and are stretched. The turbulence consists of a wide
spectrum of eddies.
In this text, the commonly used mathematical treatments of turbulence for the analysis
concerning Computational Fluids Dynamics (CFD) are described. We envisage to provide a
comprehensive but brief review of the mathematical treatments used in CFD for applications in
complex flow and heat transfer problems. The results due o the investigations of the author and
his coworkers have been wherever possible. The methods for calculating turbulence can be
grouped into the three broad categories.
 Reynolds averaged Navier-Stokes (RANS) equations of turbulence. The RANS approach
includes eddy-viscosity based models, such as the k- e models and its variants on one
hand and the Reynolds Stress Model (RSM) on the other. Usually the RSM consists of
the second moment turbulence modeling.
 Large Eddy Simulation (LES) techniques.
 Direct Numerical Simulation (DNS) of turbulence.
 
Any flow whether laminar or turbulent, is fully represented by the Navier-Stokes equations. The
Navier-Stokes equations can be solved on a fine enough grid with an exceptionally accurate
discretization method so that both the fine scale and large scale aspects of turbulence can be
calculated. This is termed as the Direct Numerical Simulation (DNS) of turbulence (Rai and
Moin, 1991). However, the length-scale-range of the eddies of varying sizes and the time-scale-
range of the velocity fluctuations due to the eddying motion cannot be economically resolved by
ordinary discretization methods. Therefore, the Engineering Problems may be solved using
Statistical Calculation Methods. Thus it is necessary to use some statistical average and a
measure of the deviation from that average. Rodi (1993), and Wilcox (1993) provide excellent
documentations on statistical approaches of turbulent flows. Notwithstanding the extent of the
coverage of these methods, we shall focus at the equations governing the flow and the heat
transfer in the turbulent regime at the first place. Subsequently, we shall describe various models
and methods to solve the equations.
Reynolds Averaged Navier-Stokes (RANS) Equations
 
In this method, the instantaneous quantities in the governing equations for the
mass, momentum and energy are decomposed into their mean and fluctuating  
components. For an incompressible flow, the instantaneous velocity obeys the  
following equation (in Cartesian tensor form)  
 
(26.1)  
 
The velocity components and scaler quantities such as pressure are decomposed  
following the Reynolds decomposition as  
 
(26.2)  
   
(26.3)  
where and P are the time or ensemble average components and and the
fluctuating components. By substituting and p of Eqs. (26.2) and (26.3) into
Eq. (26.1) and time (or ensemble) averaging, the equations can be written in
terms of the mean quantities as

(26.4)

The continuity equation in terms of mean quantities (for an incompressible flow)


can be written as

(26.5)
Equations (26.4) are called the Reynolds-averaged Navier-Stokes equations. They  
have the same form as the laminar Navier-Stokes equations with the velocities
and other variables representing time-averaged (or ensemble-averaged) values.
However, an additional term appears in Eqn. (26.4) which represents the effect of
turbulence and is called the Reynolds Stress Tensor: . This term
introduces six unknown terms but matching equations are not available to close
the system. Therefore, this term needs to be modeled in order to close the system
of equations. Several approaches have evolved for this purpose. The commonly
followed methodologies include
1. Eddy viscosity models and
2. Reynolds stress transport models.
All these approaches require a special treatment of turbulent flows near the wall.
The special treatment on near wall flows has been described in a separate section.
Eddy Viscosity Models
This is the most well known means to model the Reynolds stresses. The basic concept takes a
recourse to Boussinesq hypothesis that assumes the relationship between the mean velocity
gradients and the Reynolds stresses in the following way

(26.6)

where
 

         (26.7)
 
And k is the turbulent kinetic energy. Equation (26.6) is valid for incompressible flows. The
symbol v t is the turbulent eddy viscosity which is not a fluid property but depends strongly on
the state of turbulence. Hence, vt may vay significantly from one point in the flow to another and
also from flow to flow.
zero-Equation Models
In this model, no additional differential equation to needed to obtain the turbulent viscosity vt ,
which is defined as a function of the mean flow. The Baldwin and Lomax (1978) is one such
model based on the prandtl mixing-length theory. In this model, two different expressions are
given for the turbulent viscosity.
 For the inner zone ,
(26.8)
where the mixing length l and the vorticity | ω | are given by

(26.9)

(26.10)

with A + = 26, and the von Kármán constant K=0.42.


 For the outer region
(26.11)
The factor Fω decides the length scale depending upon whether the field point lies in the bounded
zone or in the wake
(26.12)
where,
(26.13)

 
ymax is the value of y at which F(y) achieves its maximum value. The quantity Udif is given by
(26.14)
 
The Klebanoff's Intermittency Function is expressed as

(26.15)

Here y is the distance normal to the wall, yc the inner zone (viscous sublayer) thickness, and d the
boundary layer thickness. The values of α , Ccp , CKleb , and Cwk are 0.0168, 1.6, 0.3 and 1
respectively. In many flows of industrial importance the zero-equation model as stated above or
its variants find useful application. Maji and Biswas (1999) used a simpler zero equation model
in order to predict complex flow in the spiral casing of a hydraulic turbine (Figure 26.1).

Figure 26.1: Secondary flow at different cross sections of a spiral for Re = 106 (a)  
θ = 0° , (b) θ = 90° , (c) θ = 180°
 
One-Equation Model
This is a very popular model used widely for external aerodynamic flows. Spalart and Allmaras
(1992) have developed one equation model, which is designed specifically for aerospace
applications involving wall-bounded flows. The Reynolds stresses are given by
(27.1)
The eddy viscosity vt is given by

(27.2)

is the molecular viscosity, obeys the transport equation

(27.3)
     
 
Here

(27.4)

where S is the magnitude of the vorticity, and d is the distance to the closest wall. The function fw
is

(27.5)

For large r, fw reaches a constant, so large values of r can be truncated to 10 or so. The wall
boundary condition is . In the freestream 0 is best, provided numerical errors do not push
to negative values near the edge of the boundary layer (the exact solution cannot go negative).
Values below /10 will be acceptable; the same applies to the initial condition.
 
In some codes a portion of the solid surface, typically the fuselage, is treated with a free-slip
condition while another portion, typically the wing, is treated with a no-slip condition. For ,
the appropriate condition on the free-slip surface is a Neumann condition (zero normal
derivative). In addition, the free-slip wall points are not included in the search when d is
computed for the field points.
The function ft2 is
(27.6)
 
The trip function ft1 is as follows. The quantity dt is the distance from the field point to the trip,
which is on a wall, wt is the wall vorticity at the trip, and ΔU is the difference between the
velocity at the field point and that at the trip. Then gt = min (0.1, ΔU / wtδ xt where δxt is the grid
spacing along the wall at trip, and

(27.7)
The constants are cb1 = 0.1355, σ = 2/3, cb2 = 0.622, = 0.41, cw1 = cb1/k2 + (1 + cb2 )/σ , cw2 = 0.3,
cw3 = 2, cv1 = 7.1, ct1 = 1, ct2 = 2, ct3 = 1.2 and ct4 = 0.5.
 
The model has demonstrated good results for boundary layer flows subject to adverse pressure
gradients. It performs well in the near wake and appears to be a good candidate for more
complex flows.
 
Two-Equation Models
The family of two equation models are usually implemented as k- ε , and k- ω
models. First let us introduce k- ε model.
 
k- ε Model
The turbulent viscosity , vt in Equation (26.6), is computed from a velocity scale (
k 1/2 ) and a length scale ( k3/2 / ε ) which are predicted at each point in the flow via
solution of the following transport equations for turbulent kinetic energy ( k ) and
its dissipation rate ( ε ):

(27.8)

(27.9)

where P is the generation of k and is given by

(27.10)

The turbulent viscosity is then related to k and ε by the expression

(27.11)
The coefficients Cμ , C1ε , C2ε , σk and σε are constants which have the following
empirically derived values
 
Cμ = 0.09, C 1ε= 1.44, C 2ε= 1.92, σk = 1.0, σε= 1.3
 
This is the central concept for a family of two equation of models (Jones and
Launder, 1972; Launder and Spalding, 1974) where the equation for turbulent
kinetic energy determines the velocity scale. The two equation models are quite
successful and have become very popular for engineering applications. With only
little modification, they are able to simulate a large variety of flows with
reasonably good degree of accuracy.
 
Following the philosophy of momentum equation, the thermal energy or species
concentration conservation equation can be written as
(27.12)

where T is temperature and S is the source term. Reynolds decomposition


suggests:
 
 
and the time averaged equation becomes

(27.13)

In direct analogy to the turbulent momentum transport, the turbulent heat or mass
transport is often assumed to be related to the gradient of the transported quantity

(27.14)

k - ω Model
Historically, Kolmogorov (1942) proposed the first two equation model. Kolmogorov chose the
kinetic energy of turbulence as one of the parameters, while the other parameter as the
dissipation per unit turbulence kinetic energy, ω . Wilcox (1988) and Speziale et al. (1990) also
regard ω as the ratio of ε and k . Here we present the model due to Wilcox as
The turbulent viscosity is related to k and ω by the expression

(28.1)

The transport equations for turbulent kinetic energy (k) and its dissipation rate per unit
turbulence kinetic energy (ω) are
(28.2)

(28.3
)
 
The coefficients have the following empirically derived values
α = 5/9, β = 3/40, β* = 9/100, σk = 2.0, σω = 2.0  
The equation for ω may also be derived from the ε equation using variable transformation

This tranformation relates α and β coefficients to Cε1 and Cε2

 
         and        
Thus the turbulent diffusion term from the transformed k-ε model will, however, contain
additional terms

(28.4)
     
which is the major difference between k-ε and k-ω model. Also σk=2.0 compared with σk=1.0 in
the k-ε model
SST (Shear Stress Transport) Turbulence Model
A major concern in flow simulation is the separation from a surface under adverse pressure
gradient. Separation has a strong effect on the near-wall turbulence and therefore on the turbulent
heat transfer. The SST model, has demonstrated the capability of accurate separation predictions
in numerous cases, and it is used as the basis for heat transfer predictions by many authors. The
idea behind the SST model is to combine the best elements of the k - ε and the k - ω model with
the help of a blending function F1. F1 is one near the surface and zero in the outer part and for
free shear flows. It activates the Wilcox model (k-ω)in the near-wall region and the k - ε model
for the rest of the flow. By this approach, the attractive near-wall performance of the Wilcox
model can be utilized without the potential errors resulting from the free stream sensitivity of
that model.
The formation of the SST model is as follows:

(28.5)

(28.6)

with
(28.7)

The coefficients, φ , of the model are functions F1 : φ = F1φ1 + (1 - F1 )φ2 , where φ1 , φ2 stand
for the coefficients of the k- ω and the k- ε model respectively:
σ k1 = 2.0, σω1 = 2.0, = 0.41, γ1 = 0.5532, β1 = 0.0750, β* = 0.09, c1 = 10  

σ k2 = 1.0, σω2 = 1.168, = 0.41, γ2 = 0.4403, β2 = 0.0828, β* = 0.09,  


with

(28.8)
    

(28.9)

(28.10)
    

(28.11)

 
An additional feature of the SST model is the introduction of an upper limit for the turbulent
shear stress in boundary layers in order to avoid excessive shear-stress levels typically predicted
with Boussinesq eddy-viscosity models. The eddy viscosity is defined as:

(28.12)

With a1 = 0.31. Again F2 is a blending function (described in Eq. (28.9)) similar to F1 , which
restricts the limiter to the wall boundary layer, as the underlying assumptions are not correct for
free shear flows.
 

You might also like