Turbulence Models

2024

DEVRİM GÜRKAN ÇELİK – SKY

EXPERIENCE INTERN

Türk Havacılık ve Uzay ’nin bağlı ’nin

Sanayii A.Ş., ortaklığı ve iştirakidir.
Descriptors of Turbulent Flow
Reynolds Decomposition
 The Reynolds decomposition defines flow property ϕ at a point as
the sum of a steady mean component Φ and a time varying
fluctuating component ϕ′(t) with zero mean value.

 ϕ(t) =Φ+ ϕ′(t).

 For turbulent modelling, flow properties can be written in this form.

Descriptors of Turbulent Flow
Time Average or Mean
Δ𝑡
1
The mean Փ of flow property ϕ is defined as: Φ= ∫
Δ𝑡 0
❑𝜑 (𝑡)d 𝑡

Δ𝑡
1
The time average of the fluctuations ϕ ′ is, by definition, zero:
′
𝜑= ∫
Δ𝑡 0
❑ 𝜑 ′ (𝑡 )d 𝑡 ≡ 0

From now on we shall not write down the time-dependence of ϕ and ϕ ′ explicitly, so we write ϕ =Φ+ ϕ ′ .
Descriptors of Turbulent Flow
Variance, r.m.s. and turbulence kinetic energy

The descriptors used to indicate the spread of the fluctuations ϕ ′ about the mean value Φ are the variance and root
mean square (r.m.s.):

[ ]
Δ𝑡 1 /2
Δ𝑡
( 𝜑 ) = 1 ∫ ❑( 𝜑′ ) d 𝑡
′ 2
Δ𝑡 0
2
√ ′ 2
𝜑 𝑟𝑚𝑠 = ( 𝜑 ) =
1
∫
Δ𝑡 0
❑ ( 𝜑
′ 2
) d𝑡

The r.m.s. values of the velocity components are of particular importance since they are generally most easily
measured and express the average magnitude of velocity fluctuations.
Descriptors of Turbulent Flow
Turbulent Kinetic Energy and Turbulence Intensity

The variences of velocity fluctuations , , and cause additional stresses experienced by fluid elements in turbulent flow.

One-half times these variances has a further interpretation as the mean 1

kinetic energy per unit mass contained in the respective velocity 𝑘= ( 𝑢′ 2 +𝑣 ′ 2+ 𝑤′ 2 )
2
fluctuations. The total kinetic energy per unit mass of the turbulence at a
given location can be found as follows:

( )
1/ 2
2
The turbulence intensity is the average r.m.s. velocity divided by a reference mean 𝑘
flow velocity and is linked to the turbulence kinetic energy as follows: 3
𝑇 𝑖=
𝑈 𝑟𝑒𝑓
Descriptors of Turbulent Flow
Moments of different fluctuating variables

The variance is also called the second moment of the fluctuations. Important details of the structure of the fluctuations
are contained in moments constructed from pairs of different variables. For example, consider properties ϕ =Φ+ ϕ ′ and
ψ =Ψ+ ψ′ with = 0. Their second moment is defined as :

If velocity fluctuations in different directions were independent random

Δ𝑡 fluctuations, then the values of the second moments of the velocity
1
′
𝜑𝜓= ∫
′
Δ𝑡 0
❑𝜑 ′ 𝜓′ d 𝑡 components , and would be equal to zero. However, as we have seen,
turbulence is associated with the appearance of vortical flow structures
and the induced velocity components are chaotic, but not independent, so
in turn the second moments are non-zero.
REYNOLDS AVERAGED NAVIER STOKES EQUATIONS FOR INCOMPRESSIBLE
FLOW

A summary of the topics covered in the previous slides:

‾
𝜕𝜑 𝜕Φ
′ ′
𝜑 =𝜓 =0 Φ =Φ =
𝜕𝑠 𝜕𝑠
∫ 𝜑 d 𝑠=∫ Φd 𝑠

′ ′ ′
𝜑 +𝜓=Φ+Ψ 𝜑𝜓 =ΦΨ +𝜑 𝜓 𝜑 Ψ =ΦΨ 𝜑 Ψ =0

Further rules can be applied:

d iv 𝐚 = di v 𝐀 ;¿ ¿

d i v g rad 𝜑=di v g rad Φ

REYNOLDS AVERAGED NAVIER STOKES EQUATIONS FOR INCOMPRESSIBLE
FLOW

Continuity and Navier Stokes equations in Cartesian

coordinates:
d i v 𝐮 =0
𝜕𝑢 1 𝜕𝑝
+ ¿(𝑢 𝐮)=− + 𝑣÷(grad (𝑢)) This system of equations governs every turbulent flow, but we
𝜕𝑡 𝜌 𝜕𝑥
investigate the effects of fluctuations on the mean flow using the
𝜕𝑣 1 𝜕𝑝 Reynolds decomposition in equations, thus it is appropriate to
+¿ (𝑣 𝐮 )=− + 𝑣÷(grad (𝑣 )) replace the flow variables u (also and) and by the sum of mean
𝜕𝑡 𝜌 𝜕𝑦
and fluctuating component.
𝜕𝑤 1 𝜕𝑝
+ ¿(𝑚 𝐮)=− + 𝑣÷(grad (𝑤)) 𝐮 =𝐔 + 𝐮′ 𝑢=𝑈 +𝑢′ 𝑣=𝑉 +𝑣 ′ 𝑚=𝑊 + 𝑤′ 𝑝 =𝑃 +𝑝 ′
𝜕𝑡 𝜌 𝜕𝑧
REYNOLDS AVERAGED NAVIER STOKES EQUATIONS FOR INCOMPRESSIBLE
FLOW

Note that div u= div U. This yields the continuity equation to d i v 𝐔 =0

Similar process can be applied to x-momentum equation

𝜕 𝑢 𝜕𝑈
= ¿ ⁡(𝑢 𝐮 )=¿ ⁡( 𝑈 𝐔 )+ ¿ ⁡( 𝑢 ′ 𝐮′ )
𝜕𝑡 𝜕𝑡
1 1 𝜕𝑃
− =−
𝜕𝑝 𝜌 𝜕𝑥 𝑣 ÷⁡(grad ⁡(𝑢))=𝑣 ÷⁡( grad ⁡(𝑈 ))
𝜌
𝜕𝑥
Substituting the above results gives time-average x-momentum equation.

𝜕𝑈 1 𝜕𝑃
+¿ ⁡(𝑈 𝐔 )+¿ ⁡( 𝑢 𝐮 ) =−
′ ′
+𝑣÷⁡(grad ⁡(𝑈 ))
𝜕𝑡 𝜌 𝜕𝑥
(1) (2) (3) (4) (5)
 REYNOLDS AVERAGED NAVIER STOKES EQUATIONS FOR INCOMPRESSIBLE
FLOW
Repetition of this process for y and z-momentum equations
Final form of the momentum equations (RANS):
𝜕𝑉 1 𝜕𝑃
( ′ ′
)
+¿ ⁡(𝑉 𝐔 )+¿ ⁡ 𝑣 𝐮 =− +𝑣÷⁡(grad ⁡(𝑉 ))
𝜕𝑡 𝜌 𝜕𝑦 𝜕𝑈 1 𝜕𝑃
+¿ ⁡(𝑈 𝐔 )=− +𝑣÷⁡(grad ⁡(𝑈 ))
(1) (2) (3) (4) (5) 𝜕𝑡 𝜌 𝜕𝑥

𝜕𝑊 1 𝜕𝑃
+¿ ⁡(𝑊 𝐔 )+¿ ⁡( 𝑤 𝐮 ) =−
′ ′
+𝑣÷⁡(grad ⁡(𝑊 ))
𝜕𝑡 𝜌 𝜕𝑧
(1) (2) (3) (4) (5) 𝜕𝑉 1 𝜕𝑃
+¿ ⁡(𝑉 𝐔 )= ¿ − +𝑣 ÷ ⁡(grad ⁡(𝑉 ))
𝜕𝑡 𝜌 𝜕𝑦

[
+ 1 𝜕 (− 𝜌 𝑢 𝑣 ) 𝜕 ( − 𝜌 𝑣 ) 𝜕 (− 𝜌 𝑣 𝑤 )
]
′ ′ ′2 ′ ′
It is important to note that the terms (1), (2), (4) and (5) also + +
appear in the instantaneous equations, but the process of time 𝜌 𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝑊 1 𝜕𝑃
averaging has introduced new terms (3) in the resulting time- + ¿ ⁡(𝑊 𝐔 )=¿ − + 𝑣÷ ⁡( grad ⁡(𝑊 ))
𝜕𝑡 𝜌 𝜕𝑧
average momentum equations. The terms involve products of
fluctuating velocities and are associated with convective
momentum transfer due to turbulent eddies.
+1
𝜌 [ 𝜕 ( − 𝜌 𝑢′ 𝑤 ′ ) 𝜕 ( − 𝜌 𝑣 ′ 𝑤′ ) 𝜕 ( − 𝜌 𝑤 ′ 2 )
𝜕𝑥
+
𝜕𝑦
+
𝜕𝑧 ]
REYNOLDS AVERAGED NAVIER STOKES EQUATIONS FOR INCOMPRESSIBLE
FLOW

Reynolds Stresses

Time averaged momentum equations produces six additional stresses:

Normal Stresses Shear Stresses

These extra turbulent stresses are called the Reynolds stresses. The normal stresses involve the respective variances of the
x-, y- and z-velocity fluctuations. They are always non-zero because they contain squared velocity fluctuations.
Eddy Viscosity Models

Eddy Viscosity and Eddy Diffusivity

Newton’s Law of Viscosity for an incompressible flow:

This is the formula for Reynolds stresses:

For explanation for this formula, Versteeg and

Malalasekera (1995) pg 67-68 should be examined.
is calculated by the chosen turbulence
model to close out the RANS equation.
Eddy Viscosity Models

Turbulence Model
Multiplying both Reynolds equations and each of the instantenous Navier-Stokes equation by appropriate fluctuating
velocity components, then subtracting these two equation, and rearranging the resulting equation yields to the
governing equation for turbulent kinetic energy :

If we want to describe the equation above in words:

Eddy Viscosity Models

Turbulence Model

The viscous dissipation term (VI)

Contributes negatively to the governing equation for turbulent kinetic energy due to the fluctuating deformation rates
. The rate of dissipation per unit volume is normally written as the product of the density and the rate of dissipation
of turbulent kinetic energy per unit mass :

This is the destruction term on the turbulent kinetic energy equation.

Eddy Viscosity Models

Turbulence Model

turbulence model uses and to define velocity scale ϑ and length scale as:

If dimensional analysis is applied, eddy viscosity can be specified as:

is a dimensionless constant.
Eddy Viscosity Models

Turbulence Model

Transport equation for turbulent kinetic energy :

Transport equation for

We can describe the terms on the equations like this:

The equations contain five constants that can be adjusted:

Eddy Viscosity Models

Turbulence Model

To compute the Reynolds stresses in model, we use this

equation:

In any flow the sum of the normal stresses is equal to minus twice the turbulent kinetic energy per unit volume
Eddy Viscosity Models

Turbulence Model Boundary Conditions

The governing equations for and are elliptic due to the gradient diffusion term. These equations needs certain
boundary conditions listed below:

 inlet: distributions of and must be given

 outlet or symmetry axis: and
 free stream: and
Eddy Viscosity Models

Turbulence Model Boundary Conditions

At high Reynolds number the standart model avoids the need to integrate the model equations right through
to the wall by making use of the universal behaviour of near wall flows. If is co-ordinate direction normal to a
solid wall, the mean velocity at a point with satisfies the log-law and measurements of turbulent kinetic energy
budgets indicate that the rate of turbulence production equals to dissipation. Using these assumptions and the
eddy viscosity formula, it is possible to develop the following wall functions:

𝜅 =0.41 , 𝐸=9.8
Eddy Viscosity Models

Turbulence Model Boundary Conditions

For heat transfer, the universal near wall temperature distribution valid at high Reynolds number:

is the ‘pee-function’, a correction function dependent on the ratio of laminar to turbulent Prandtl numbers.
Eddy Viscosity Models

Turbulence Model Boundary Conditions

At low Reynolds number the log-law is not valid so the above-mentioned boundary conditions cannot be used.
Wall damping needs to be applied to ensure that viscous stresses taka over from turbulent Reynolds stresses at
low Reynolds numbers and in the viscous sub-layer adjacent to solid walls.
Eddy Viscosity Models

Turbulence Model

Transport equations for and :

𝜕(𝜌 𝑘) 𝜕 ( 𝜌 𝑢 𝑗 𝑘 )
𝜕𝑡
+
𝜕𝑥𝑗
=𝑃 − 𝛽 ∗ 𝜌𝜔𝑘+
𝜕
𝜕𝑥 𝑗 [
( 𝜇+𝜎 𝑘 𝜇𝑡 )
𝜕𝑘
𝜕𝑥𝑗 ] Where is: 𝑃=𝜏𝑖𝑗
𝜕 𝑢𝑖
𝜕𝑥𝑗

𝜕(𝜌𝜔) 𝜕 ( 𝜌 𝑢 𝑗 𝜔 ) 𝛾
𝜕𝑡
+
𝜕𝑥𝑗
2
= 𝑃 − 𝛽𝜌 𝜔 +
𝜈𝑡
𝜕
𝜕𝑥𝑗 [(
𝜇+𝜎 𝜔 𝜇𝑡 )
𝜕𝜔
𝜕𝑥𝑗
+2 ( 1− 𝐹 1 )
]
𝜌 𝜎 𝜔 2 𝜕𝑘 𝜕𝜔
𝜔 𝜕𝑥𝑗 𝜕𝑥𝑗

Ω=√ 2𝑊 𝑖𝑗 𝑊 𝑖𝑗
𝜌 𝑎1 𝑘
The eddy viscosity is computed from: 𝜇 𝑡 = Vorticity magnitude:
max ( 𝑎 1 𝜔 , Ω 𝐹 2 )
Eddy Viscosity Models

Turbulence Model