CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   Inconsistent Eddy Viscosity (https://www.cfd-online.com/Forums/cfx/242698-inconsistent-eddy-viscosity.html)

jonasa97 May 5, 2022 15:09

Inconsistent Eddy Viscosity
 
Good evening,

I am encountering odd inconsistencies between the inbuilt eddy viscosity in CFD Post and the eddy viscosity, when manually defined according to the underlying equation.

Overview
CFX Pre version: 19.2
CFX Solve version: 20.2
Turbulence model: k-omega SST
Model corrections (production limiter, curvature correction etc.): None

Problem Description
I noticed that in some regions, the inbuilt "Eddy Viscosity" distribution differs significantly from the values I obtain when manually defining the eddy viscosity according to the definition (equation in this resource: https://turbmodels.larc.nasa.gov/sst.html), whereas in other regions the values are a perfect match. The main difference with the 'manual' definition is that high free stream eddy viscosity penetrates further towards the walls of my airfoil.

My expression for the 'manual' eddy viscosity is:

0.31*Density*Turbulence Kinetic Energy/max(0.31*Turbulence Eddy Frequency,Velocity.Curl*Second Blending Function for SST model)

Unless the equation is incorrect, why would the values differ from the inbuilt eddy viscosity?

Many thanks for any suggestions you might have!


EDIT: after some additional research into k-omega SST in ANSYS CFX, I have found that the production limiter is in fact enabled by default. As far as I understand the only way to prevent this is to specify an incredibly high custom production clip factor.

Opaque May 6, 2022 10:37

I think there is an error in your expression. From the link you posted, the expression says to use the "shear strain rate" which is not the "velocity.curl"

Omega*F2 term in the posted link represents the shear strain rate times the 2nd Blending Function

jonasa97 May 6, 2022 10:46

Dear Opaque,


Thank you for your reply. It was indeed my definition which was at fault.


Apparently the original model from Menter's 1993 paper uses the vorticity (i.e. velocity curl), whereas CFX implements the 2003 revision, which uses the strain rate invariant instead of vorticity.


EDIT: for anyone wondering, my expressions are now:


strainInvariant = (2.0*((Velocity u.Gradient X)^2 + (Velocity v.Gradient Y)^2 + (Velocity w.Gradient Z)^2) + (Velocity u.Gradient Y + Velocity v.Gradient X)^2 + (Velocity u.Gradient Z + Velocity w.Gradient X)^2 + (Velocity v.Gradient Z + Velocity w.Gradient Y)^2)^0.5



EddyViscosity = 0.31*Density*Turbulence Kinetic Energy/max(0.31*Turbulence Eddy Frequency,strainInvariant*Second Blending Function for SST model)


which collapses values from the inbuilt and manually defined function.

Opaque May 6, 2022 11:26

Did you try using the built-in "Shear Strain Rate" variable?

jonasa97 May 6, 2022 12:16

Yes, the in-built 'Shear Strain Rate' variable is identical to the 'invariantStrain' expression (in my previous post). I manually defined it from the velocity gradients because I didn't find 'Shear Strain Rate' when I first implemented the SST-2003 model definition for eddy viscosity.

I should have clarified in my previous post that my problem has been resolved. Thank you very much for your help.


All times are GMT -4. The time now is 09:31.