Comparison of Turbulent Models: Applications in CFD
Comparison of Turbulent Models: Applications in CFD
Comparison of
Turbulent Models
Objectives:
Introduction to Turbulence
Why we need Turbulence Modelling
Types of Turbulent Models
Application of Turbulent Models in CFD
Comparison of different Turbulent model
The Knudsen number is the ration of free mean path to length of interest.
Introduction to Turbulence
A Dissipative Flow state characterized by nonlinear fluctuating threedimensional vorticity.
Randomness of transport variable with respect to time and space
Strong mixing
Wide range of time scales and length scales.
Eddies or vorticity are the basic entities of Turbulence flow.
Internal flows
Where:
https://www.racfd.com/en/articles/ansys-fluentconcept/turbulence-modeling.html
Energy cascading
By Mathematical Derivation:
Kolmogorov
Length scale
Visco
us
Term
Pressur
e
Gradie
nt
Where,
Total Derivative Source of
Vorticity
Diffusion
Term
Consider
a Large Eddy
Scale
In les is negligible
Then is negative.
That is vortex element has
less moment of inertia.
1.
3.
This term should have
similar physical meaning.
Notes:
1.Interaction will increase
with other vortex element
2.Vortex stretching
Intensification of vorticity
with stretching of vortex
element
3.Source term appears in
vorticity transport equation is
due to stretching of vortex
element.
Space Averaging:
Types of Turbulence
Homogenous Turbulence:
Turbulent statistics are independent of co-ordinate translation.
Isotropic Turbulence:
Turbulence statistics are independent of rotation, reflections and translation of co-ordinates .
For
any turbulent quantity f:
(1)
Substitute , ,
And taking the average of the whole equation we get the following equation:
RANS
The process of averaging has introduced new term( Term A)
Physical significance of Term A, Let us consider Velocity fluctuation along x-direction that can interact with
fluctuation along y-direction therefore there can be a momentum transfer. So there is a extra component of
stress.
Total stress is given by
The viscous shear stress dominates near the surface while the turbulent (Reynolds) shear stress is significant
some distance away from the surface.
Reynolds
stress or Turbulent
stress. It is also represented
by . This is a second order
Tensor
RANS
In the RANS equations, there are six additional unknowns:
,,, , ,
http://www.southampton.ac.uk/~nwb/lectures/Goo
dPracticeCFD/Articles/Turbulence_Notes_Fluentv6.3.06.pdf
Note:
is the turbulent viscosity
is the turbulent kinetic energy
* So now the problem is to model the
Turbulent viscosity
Based on dimensional analysis, can be calculated from a turbulence time scale( or velocity scale) and a length scale.
Turbulent kinetic Energy(
Turbulence dissipation rate +
Specific Dissipation rate /
Each Turbulence models has different approach to calculate .
Spalart-Allmaras:
Solve the transport equation for a modified turbulent viscosity.
Standard , RNG , Realizable
These models solve the transport equation for both
)
Standard , SST
Solves the transport Equation for
)
Spalart Allmaras
The transport variable in Spalart Allmaras model, ,is similar to the turbulent kinematic viscosity except in the
near-wall(viscosity affected) region. The transport for is
Where,
is the production of turbulent viscosity and is the destruction of turbulent viscosity that occurs nearwall region due to wall blocking and viscous damping.
and are the constants
In the Spalart Allmaras model kinetic energy is not calculated therefore the term in the turbulent
viscosity is neglected.
Models
Standard
RNG
Realizable
All the above three models have similar forms with two
transport equation of
The main difference is method of
calculating turbulent kinetic energy.
The turbulent Prandtl numbers governing the turbulent
diffusion of
The generation and destruction term in equation
Transport Equation for
Transport Equation for :
Transport Equation
(
RNG K The RNG model has an additional in its equations that improves the accuracy for rapidly strained flows.
The effect of swirl on turbulence is included in the RNG model, enhancing accuracy for swirling flows.
The RNG model provides the analytical formula for turbulent Prandtl numbers, while the standard model uses
user-specified, constant values.
The standard K- is only for high Reynolds number, the RNG theory provides an analytically-derived
differential for effective viscosity that accounts for low Reynolds number effects.
*These feature make the RNG model more accurate than standard K- model and reliable for a wide range of
flows.
The model do not contain terms which are undefined at the wall. They can be integrated to the wall without
using wall functions.
They are accurate and robust for a wide range of boundary layer flows with pressure gradient.
Fluent has two models in K Standard K1. Most widely used in aerospace and turbomachinery.
2. Several sub models are present to include the effects of compressibility, transitional flows.
. SST K- (Menter, 1994)
1. The SST K- model uses blending function to gradually transition from the standard K- model near the wall
to high Reynolds number version of k- in the outer portion of the boundary layer.
2. Contains a modified turbulent viscosity formulation to account for the transport effects of the principal
turbulent shear stress.
Thank you