
[Sponsors] 
March 3, 2017, 09:42 
k and omega at inlet for komega SST

#1 
New Member
Audrius
Join Date: Mar 2015
Posts: 13
Rep Power: 5 
Dear Foamers,
I have recently started using OpenFOAM and gradually ran into some convergence issues that I cannot tackle myself. What I am trying to do is to compare my OpenFOAM simulation with a simulation done in ANSYS CFX. The case is a swirl induced by turning vanes. Please find attached the geometry domain screenshot. I am running the case with simpleFoam using kOmegaSST turbulence model. Please see the case files attached or let me know if you require any specific details. Now the problem is convergence in my OpenFOAM simulation. As soon as I try using second order divergence schemes, I get problems with convergence. I am quite lost where the problem is. Please see the residual plot attached. For comparison, the residuals in CFX go down below 1e6 (RMS) or 1e3 (max). I am using ‘High resolution’ Advection Scheme setting there. I would really appreciate any guidance on this and apologies if it’s just another beginner level question. Is there anything obvious I should check first or anything obviously wrong in the setup? Many thanks for your time! Kind regards, Audrius 

March 17, 2017, 10:51 

#2 
New Member
Audrius
Join Date: Mar 2015
Posts: 13
Rep Power: 5 
Hello Everyone,
Recently started learning OpenFOAM. I am trying to define the inlet turbulence exactly how I have it in ANSYS CFX. In CFX I have used Medium, which means 5% intensity and a viscosity ratio μ_t/μ equal to 10. In OpenFOAM I have used Intensity and Length Scale: For k inlet { type turbulentIntensityKineticEnergyInlet; intensity 0.05; value $internalField; } For omega inlet { type turbulentMixingLengthFrequencyInlet; mixingLength 0.021; //0.07 x Hydraulic Diameter value $internalField; } It's kOmegaSST turbulence model BTW. I know the problem is in the k and omega specification because if I specify them as fixed values, OpenFOAM and CFX give identical results. So my questions are: 1) Is it possible in OpenFOAM to specify the intensity and the viscosity ratio instead of length scale? 2) If not, can I calculate turbulent length scale from the viscosity ratio? 3) Maybe I should'n stick to the CFX's default turbulence setting and there is a better way of estimating the turbulence values for the inlet? Many thanks, Audrius  Hello Everyone, I have just started using OpenFOAM and got into some issues trying to specify identical inlet conditions to ANSYS CFX. In CFX I was using 'Medium' turbulence, which is 5% intensity and viscosity ratio μ_t/μ equal to 10. In OpenFOAM I have found a way to specify the Intensity in k using turbulentIntensityKineticEnergyInlet and length scale in omega using turbulentMixingLengthFrequencyInlet. For the length scale I have specified 0.07 x Hydraulic Diameter. Now this does not give me identical results to CFX. I know this discrepancy is because of k and omega values at the inlet. When I specify k and omega as fixed values, I get identical results in OpenFOAM and in CFX. So my questions are: 1) Is there a way to specify viscosity ration in OpenFOAM? 2) Is there a way of estimating the lengthscale from the viscosity ratio? Many thanks, Audrius Last edited by wyldckat; April 30, 2017 at 11:10. Reason: merged posts that were done ~2h apart 

March 20, 2017, 12:24 

#3 
New Member
Matteo Biazotti
Join Date: Feb 2017
Posts: 4
Rep Power: 3 
hello foamers, I'm a bit of a newbie to the software. ive been trying to use the komega SST model without a wall function to model the effect of a ducted tip hydrofoil, and it effects on the CL/CD. i have my Y+~ 1, but i am still wondering what values for K and Omega to use in the initial conditions.
i have found several different sources to giving different equations to calculate the these values. currently using: k=10e6*U^2 Omega = 5U/L l=chord length this is to model the tipvortex of a hydrofoil. Reynolds number ~ 1.5e6 anyadvice is much appreciated Last edited by wyldckat; April 30, 2017 at 11:16. Reason: merged posts an hour apart from different threads 

March 23, 2017, 16:14 

#4 
Senior Member
Khamlaj
Join Date: Nov 2010
Location: United States
Posts: 241
Rep Power: 10 
Hi mash,
If you are not using wall function i.e y+<6, the following initial condition for k and omega should be used. k@omega.png Note: for omega use 1e+10 instead of minus (close to the infinity) but you have the freedom to try them both. Best, 

March 28, 2017, 15:59 

#5 
Member
Fredi Cenci
Join Date: Dec 2016
Posts: 33
Rep Power: 3 
I am trying to reproduce a transient flow around a circular cylinder. I am setting the values of k, omega and nut as the equations on this link https://www.cfdonline.com/Tools/turbulence.php, it might help you.
I normally give the following entries: Turbulent intensity, eddy viscosity ratio, free stream velocity and laminar viscosity. I have to point out that, in my case, i am using nut/nu = 0.01, however my simulation doesnt give this value back in my mesh. I am trying to investigate why. Bests Regards. 

April 7, 2017, 12:35 

#6 
New Member
Matteo Biazotti
Join Date: Feb 2017
Posts: 4
Rep Power: 3 
im having a similar issue, i have seen a ridiculous number of different methods for estimating these values, all of which get different answers,
im working with a 3d hydrofoil with chord 0.1m and inlet velocity of 16m/s also finding many different formula for K and Omega. if anyone has a method they are sure of it would be much appreciated cheers matteo 

April 30, 2017, 13:48 

#7 
Super Moderator
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 10,158
Blog Entries: 39
Rep Power: 112 
Greetings to all!
I've merged a few threads that were all on the same topic into this single thread, given that pretty much the same answers can be given here. Therefore, for those who asked the questions, please reread the whole thread to see if there is already an answer for your question. As for the questions about "eddy viscosity ratio", I can't find any public implementation for OpenFOAM of a boundary condition for the "omega" field that takes as a parameter the "eddy viscosity ratio". I wasn't even aware that this was sometimes a better alternative to using mixing length. I've done a bit of research and the following links are relevant to this topic:
Right now I'm unable to try and code this myself, but if anyone can provide a simple test case in which this can be verified that it's easier to use this type of boundary condition instead of using mixing length, I can try and code the boundary conditions some weekend in May 2017... Best regards, Bruno
__________________


May 1, 2017, 10:02 
Turbulent Inlet & Boundary Calculator

#8 
Member
Arvind Jay
Join Date: Sep 2012
Posts: 71
Rep Power: 8 
Attached is the calculator, I have used for my cases.
Cheers, J 

August 7, 2017, 14:33 

#9 
New Member
Martin
Join Date: Oct 2013
Location: Newcastle
Posts: 19
Rep Power: 6 
Hello,
I am trying different settings for the kw SST model in OpenFOAM 2.3.0 at the moment. I have come across a number of different settings, espcially the internalField set up and started to wonder what is the best/right way? are there any differences between using internalField uniform $turbulentOmega; and internalField uniform [value]; does the second mean that omega is fixed? Does anyone have any experience with this? 

August 7, 2017, 14:46 

#10  
Member
Arvind Jay
Join Date: Sep 2012
Posts: 71
Rep Power: 8 
Quote:
Cheers, Jay 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
At high Y+ values does the K Omega SST model just behave like the K Epsilon model?  JuPa  CFX  0  December 22, 2015 07:44 
how to calculate the omega at inlet boundary in k omega sst  Scabbard  OpenFOAM Running, Solving & CFD  2  September 30, 2014 13:06 
Nonphysical flow: omega is fixed at inlet  aljazari  OpenFOAM Running, Solving & CFD  0  November 15, 2012 12:18 
overshooting of Omega in SST komega using simpleFoam  cm_jubayer  OpenFOAM  1  August 26, 2011 13:18 
Boundary Conditions for k omega SST  dancfd  OpenFOAM PreProcessing  0  June 9, 2011 23:25 