Turbulence free-stream boundary conditions

From CFD-Wiki

(Difference between revisions)

Latest revision as of 11:59, 15 January 2014

In most CFD simulations it is necessary to specify values of the turbulence variables at the inlets. For example, if you are using a $k-\epsilon$ model you have to specify values of $k$ and $\epsilon$ at the inlets. This is often difficult and a source of uncertainty since the incoming turbulence is rarely known exactly. Most often you are forced to make a more or less educated guess of the incoming turbulence.

Estimating the turbulence model variables, like turbulent energy, dissipation or Reynolds stresses, directly is often difficult. Instead it is easier to think in terms of variables like the incoming turbulence intensity and turbulent length scale or eddy viscosity ratio. These properties are more intuitive to understand and can more easily be related to physical characteristics of the problem. For some guidelines on how these variables can be estimated see the respective CFD-Wiki page describing them.

Formulas for computing the turbulence model variables

Once an appropriate turbulence intensity and turbulence length scale or eddy viscosity ratio have been estimated or measured, the primitive turbulence model variables can be computed from the following formulas. Some codes may not require the $C_\mu$ constant so it is a good idea to check the user manual if possible.

Modified turbulent viscosity

The modified turbulent viscosity, $\tilde{\nu}$ , can be computed using the following formulas:

From the turbulence intensity and length scale

$\tilde{\nu} = \sqrt{\frac{3}{2}} \; (U \, I \, l)$

Where $U$ is the mean flow velocity, $I$ is the turbulence intensity and $l$ is the turbulent length scale.

Ideally with the Spalart-Allmaras model $\tilde{\nu}=0$ , but some solvers can have problem with that so $\tilde{\nu}<=\frac{\nu}{2}$ can be used. This is if the trip term is used to "start up" the model. A convenient option is to set $\tilde{\nu}=5{\nu}$ in the freestream. The model then provides fully turbulent results and any regions like boundary layers that contain shear become fully turbulent.

Turbulent energy

The turbulent energy, $k$ , can be computed as:

$k = \frac{3}{2} \; (U \, I)^2$

Where $U$ is the mean flow velocity and $I$ is the turbulence intensity.

Dissipation rate

The turbulent dissipation rate, $\epsilon$ , can be computed using the following formulas:

From the turbulence length scale

$\epsilon = C_\mu \, \frac{k^\frac{3}{2}}{l}$

$k$ is the turbulent energy and $l$ is the turbulent length scale

Please note that some CFD codes, Fluent, Phoenics and CFD-ACE for example, uses a different length-scale definition based on the mixing-length, and therefore the following formula should be used:

$\epsilon = C_\mu^\frac{3}{4} \, \frac{k^\frac{3}{2}}{l}$

Where $C_\mu$ is a turbulence model constant which usually has a value of $0.09$

From the eddy viscosity ratio

$\epsilon = C_\mu \, \frac{\rho \, k^2}{\mu} \, (\frac{\mu_t}{\mu})^{-1}$

Where $C_\mu$ is a turbulence model constant which usually has a value of $0.09$ , $k$ is the turbulent energy, $\rho$ is the density, $\mu$ is the molecular dynamic viscosity and $\frac{\mu_t}{\mu}$ is the eddy viscosity ratio.

Specific dissipation rate

The specific turbulent dissipation rate, $\omega$ , can be computed using the following formulas:

From the turbulence length scale

$\omega = \frac{\sqrt{k}}{l}$

Please note that some CFD codes, Fluent, Phoenics and CFD-ACE for example, uses a different length-scale definition based on the mixing-length, and therefore the following formula should be used:

$\omega = C_\mu^{-\frac{1}{4}} \, \frac{\sqrt{k}}{l}$

Where $C_\mu$ is a turbulence model constant which usually has a value of $0.09$ , $k$ is the turbulent energy and $l$ is the turbulent length scale.

From the eddy viscosity ratio

$\omega = \frac{\rho \, k}{\mu} \, (\frac{\mu_t}{\mu})^{-1}$

Where $k$ is the turbulent energy, $\rho$ is the density, $\mu$ is the molecular dynamic viscosity and $\frac{\mu_t}{\mu}$ is the eddy viscosity ratio.

@@ Line 1: / Line 1: @@
-In most CFD simulations it is necessary to specify values of the turbulence variables at the inlets. For example, if you are using a <math>k-\epsilon</math> you have to specify values of <math>k</math> and <math>\epsilon</math> on the inlets. This is often difficult and a source of uncertainty, since the incoming turbulence is rarely known exactly. Most often you are forced to make a more or less educated guess of the incoming turbulence.
+In most CFD simulations it is necessary to specify values of the turbulence variables at the inlets. For example, if you are using a <math>k-\epsilon</math> model you have to specify values of <math>k</math> and <math>\epsilon</math> at the inlets. This is often difficult and a source of uncertainty since the incoming turbulence is rarely known exactly. Most often you are forced to make a more or less educated guess of the incoming turbulence.
-Estimating the incoming turbulence model variables, like turbulent energy, dissipation or even Reynolds stresses, directly is often difficult. Instead it is easier to think of the incoming turbulence in terms of variables like [[turbulence intensity]] and [[turbulent length-scale]] or [[eddy-viscosity ratio]]. These properties are more intuitive and can more easily be related to physical characteristics of the problem.
+Estimating the turbulence model variables, like turbulent energy, dissipation or Reynolds stresses, directly is often difficult. Instead it is easier to think in terms of variables like the incoming [[turbulence intensity]] and [[turbulent length scale]] or [[eddy viscosity ratio]]. These properties are more intuitive to understand and can more easily be related to physical characteristics of the problem. For some guidelines on how these variables can be estimated see the respective CFD-Wiki page describing them.
-{{stub}}
+==Formulas for computing the turbulence model variables==
+Once an appropriate turbulence intensity and turbulence length scale or eddy viscosity ratio have been estimated or measured, the primitive turbulence model variables can be computed from the following formulas. Some codes may not require the <math>C_\mu</math> constant so it is a good idea to check the user manual if possible.
+===Modified turbulent viscosity===
+The modified turbulent viscosity, <math>\tilde{\nu}</math>, can be computed using the following formulas:
+====From the turbulence intensity and length scale====
+:<math>\tilde{\nu} = \sqrt{\frac{3}{2}} \; (U \, I \, l)</math>
+Where <math>U</math> is the mean flow velocity, <math>I</math> is the [[turbulence intensity]] and <math>l</math> is the [[turbulent length scale]].
+Ideally with the [[Spalart-Allmaras_model|Spalart-Allmaras model]] <math>\tilde{\nu}=0</math>, but some solvers can have problem with that so <math>\tilde{\nu}<=\frac{\nu}{2}</math> can be used. This is if the trip term is used to "start up" the model. A convenient option is to set <math>\tilde{\nu}=5{\nu}</math> in the freestream. The model then provides fully turbulent results and any regions like boundary layers that contain shear become fully turbulent.
+===Turbulent energy===
+The turbulent energy, <math>k</math>, can be computed as:
+:<math>k = \frac{3}{2} \; (U \, I)^2</math>
+Where <math>U</math> is the mean flow velocity and <math>I</math> is the [[turbulence intensity]].
+===Dissipation rate===
+The turbulent dissipation rate, <math>\epsilon</math>, can be computed using the following formulas:
+====From the turbulence length scale====
+:<math>\epsilon = C_\mu \, \frac{k^\frac{3}{2}}{l}</math>
+<math>k</math> is the turbulent energy and <math>l</math> is the [[turbulent length scale]]
+Please note that some CFD codes, Fluent, Phoenics and CFD-ACE for example, uses a different length-scale definition based on the mixing-length, and therefore the following formula should be used:
+:<math>\epsilon = C_\mu^\frac{3}{4} \, \frac{k^\frac{3}{2}}{l}</math>
+Where <math>C_\mu</math> is a turbulence model constant which usually has a value of <math>0.09</math>
+====From the eddy viscosity ratio====
+:<math>\epsilon = C_\mu \, \frac{\rho \, k^2}{\mu} \, (\frac{\mu_t}{\mu})^{-1}</math>
+Where <math>C_\mu</math> is a turbulence model constant which usually has a value of <math>0.09</math>, <math>k</math> is the turbulent energy, <math>\rho</math> is the density, <math>\mu</math>is the molecular dynamic viscosity and <math>\frac{\mu_t}{\mu}</math> is the [[eddy viscosity ratio]].
+===Specific dissipation rate===
+The specific turbulent dissipation rate, <math>\omega</math>, can be computed using the following formulas:
+====From the turbulence length scale====
+:<math>\omega = \frac{\sqrt{k}}{l}</math>
+Please note that some CFD codes, Fluent, Phoenics and CFD-ACE for example, uses a different length-scale definition based on the mixing-length, and therefore the following formula should be used:
+:<math>\omega = C_\mu^{-\frac{1}{4}} \, \frac{\sqrt{k}}{l}</math>
+Where <math>C_\mu</math> is a turbulence model constant which usually has a value of <math>0.09</math>, <math>k</math> is the turbulent energy and <math>l</math> is the [[turbulent length scale]].
+====From the eddy viscosity ratio====
+:<math>\omega = \frac{\rho \, k}{\mu} \, (\frac{\mu_t}{\mu})^{-1}</math>
+Where <math>k</math> is the turbulent energy, <math>\rho</math> is the density, <math>\mu</math>is the molecular dynamic viscosity and <math>\frac{\mu_t}{\mu}</math> is the [[eddy viscosity ratio]].

Turbulence free-stream boundary conditions

From CFD-Wiki

Latest revision as of 11:59, 15 January 2014

Contents

Formulas for computing the turbulence model variables

Modified turbulent viscosity

From the turbulence intensity and length scale

Turbulent energy

Dissipation rate

From the turbulence length scale

From the eddy viscosity ratio

Specific dissipation rate

From the turbulence length scale

From the eddy viscosity ratio

Views

My wiki

wiki navigation

wiki search

wiki toolbox