Problems with IDDES in openfoam2.1.1
 

Dear Foamers:
I am now trying to use IDDES to simulate the aerodynamic performance of high-speed train, but i got the errors shows below:

request for RASModel RASProperties from objectRegistry region0 failed
available objects of type RASModel are
0
(
)

Can any one help me with with.

Greetings Joanna Huang,

It would help if you could describe the steps you've taken to reach that error message. ;)

From what I can figure out, there is an "IDDESDelta" class in OpenFOAM's source code at "src/turbulenceModels/incompressible/LES/SpalartAllmarasIDDES/IDDESDelta/", which is part of the incompressible LES turbulence library. The error you're getting is referring to the RAS model... so I can only guess that you'll need to switch to the LES model instead.

Best regards,
Bruno

Hi Bruno,
Thank you for your answer. So for IDDES simulation, I don't need RASProperties file in constant directory, the IDDES LES turbulence modle will activate RASmodel automaticlly to sovle the near wall region. Am I right?

Best Regards,
Joanna

Hi Joanna,

You can only use LES or RAS, not both at the same time... or at least I don't think so.

Have a look at the tutorial "incompressible/pisoFoam/les/pitzDaily" to see how to set-up a case to simulate with LES.

Best regards,
Bruno

Hi Bruno,
Thank you for your kindly answering. I am a little confused, the IDDES is a hybrid mehod of RANS and LES, how can I control it and separate it. Below I attach the Case I am doing(contains the 0,constant,system), Can you kindly tell me what is the problem. Thank you so much.

Problems with IDDES in openfoam2.1.1
 

Dear Foamers:
I am now trying to use IDDES to simulate the aerodynamic performance of high-speed train, but i got the error showed below:

request for RASModel RASProperties from objectRegistry region0 failed
available objects of type RASModel are
0
(
)

Is anyone familar with this error, Can anyone help me with this.
Below I attach the file case(0, constant, sysytem),it will be really greatful if somebody could figue this out.

Best regards,
Joanna

Hi Joanna,

I've moved the other thread into this new thread you created, so that the whole story was easily readable by others. ;)

OK, I've given it a try and here's what I've managed to figure out:

1. Just for reference, there is a thread on the topic of RANS and LES: http://www.cfd-online.com/Forums/mai...s-vs-rans.html
2. It looks like the "SpalartAllmaras" model is present on both the RAS and LES side.
3. But "SpalartAllmarasDDES" and "SpalartAllmarasIDDES" are only present on the LES side. This is possibly because the OpenFOAM authors did not have yet a request to make these two models work on the RAS side.
4. You can look at the source code in the following folders and try to implement the "SpalartAllmarasIDDES" model for the RAS side; run in the terminal the following commands, to see where the "SpalartAllmaras" models are located:
   
   Code:

   ---

   echo $FOAM_SRC/turbulenceModels/incompressible/RAS/SpalartAllmaras
echo $FOAM_SRC/turbulenceModels/incompressible/LES/SpalartAllmaras

   ---

   The IDDES one can found by running:
   
   Code:

   ---

   echo $FOAM_SRC/turbulenceModels/incompressible/LES/SpalartAllmarasIDDES

   ---

5. Nonetheless, if the case can be executed with LES, I've changed inside the file "constant/LESProperties" the following entry:
   
   Code:

   ---

   SpalartAllmarasIDDESCoeffs
{
    delta  IDDESDelta;

    IDDESDeltaCoeffs
    {
        deltaCoeff      1;

        //according to src/turbulenceModels/incompressible/LES/SpalartAllmarasIDDES/IDDESDelta/IDDESDelta.C
        cw              0.15;
    }

    alphaNut        1.5;
    kappa            0.4187;
    Cb1              0.1355;
    Cb2              0.622;
    Cw2              0.3;
    Cw3              2.0;
    Cv1              7.1;
    Cv2              5.0;
    CDES            0.65;
    ck              0.07;

//These are the default values
//    sigmaNut        0.66666;
//    kappa          0.41;
//    Cb1            0.1355;
//    Cb2            0.622;
//    Cv1            7.1;
//    Cv2            5;
//    CDES            0.65;
//    ck              0.07;
//    Cw2            0.3;
//    Cw3            2;
//    fwStar          0.424;
//    cl              3.55;
//    ct              1.63;

}

   ---

   The only difference is that I added the references to the default values that OpenFOAM uses.

Best regards,
Bruno

Hi Joanna,

I've found out a couple more helpful details:

1. It looks like the DES and IDDES models are on the LES side, but it might look like they can work on both LES and RAS, simply by using them on the LES side.
2. There is someone else also working with DES, but that forum member is using OpenFOAM 2.2: http://www.cfd-online.com/Forums/ope...on-object.html - instructions are provided there that help get the indication of how much is LES and RAS being used!

Best regards,
Bruno

Hi, Bruno
Yes, the IDDES should work both LES and RAS side, but I still didn't figue out how this error happens:

request for RASModel RASProperties from objectRegistry region0 failed
available objects of type RASModel are
0
(
)

it seems the RAS can't be activated for IDDES. Have you checked my case, is anywhere I set in a wrong way?

Thank you for the information above, it also helpful

Best regard,
Joanna.

Hi Joanna,

Which solver did you use?



Either way, I've now proven that the trick is to set-up things to work in LES mode, using in "constant/turbulenceProperties":
And in "constant/LESProperties" use:
I've used the solver pisoFoam, based on the latest case you attached. Attached to this post you can find 2 test cases:

• "case.211.tar.gz" - The case is prepared to run with OpenFOAM 2.1.1.
• "case.22x.tar.gz" - The case is prepared to run with OpenFOAM 2.2.x. The detail on this case is that it is using the "DESModelRegions" which gives valuable information about how much of the domain is running in RAS and how much in LES, as explained here: http://www.cfd-online.com/Forums/ope...tml#post424652 - post #2

Both cases can be executed by running the following commands:
Or for running in parallel:
Best regards,
Bruno

Hi Bruno,

Thank you so much for your usefule information, these really help me lots. I am now testing the case you posted (still queuing in the cluster), I hope it works.

Thanks again and have a nice weekend then.

Best regards.
Joanna

Dear Bruno
Thank you for your helpful comments in this post and also in How to use DESModelRegions function object
I have a more fundamental question:
is the IDDES model that is implemented in OpenFOAM a low-Re or High-Re model?
I mean what is the best range of the first point Yplus for the walls?
can it be used with wall functions for High-Re problems?or it needs very fine mesh near the wall?

Hi Mohsen,

Prepare to be amazed as I do the simplest of tricks! I shall look at the source code! https://github.com/OpenFOAM/OpenFOAM...DESDelta.H#L28
"And how did you do this?" I hear you ask!
I went to the online code documentation: http://www.openfoam.org/docs/cpp/ - which is accessible from the main OpenFOAM web site - http://www.openfoam.org - I looked at the menu, the "Documentation" entry and flew my mouse cursor over it and it showed me the "C++ Source Guide". I clicked on it and it took me to the page I speak of.

From there, I searched for "IDDES" and so it gave me this page: http://foam.sourceforge.net/docs/cpp/a00881.html (actually because I accessed from http://foam.sourceforge.net/docs/cpp/ )


But since this link will be outdated as soon as another OpenFOAM release is made, I looked at the "Detailed Description" section on that page and there were links to the source code (which don't work very well). I then went to the online repository at https://github.com/OpenFOAM/OpenFOAM-2.2.x/ and did a bit of magic (copy-paste-adapted the link given before) et voilà, the answer is shown again at the link I gave at first :D

Best regards,
Bruno

Thanks Bruno
But ANSYS FLUENT 12.0 Theory Guide section 4.10 (which I quote here) Says :
or:
Well, as you see ANSYS FLUENT has more options for DES. the realizable k-e model is a High-Re model which requires larger Y+ near the wall for using wall functions.
I think we do not have such options in OpenFOAM.
Please let me know if you find more about this matter.

Hi Mohsen,

I did a quick search with Google and had another look into the code documentation and here is what I found:

1. http://www.openfoam.com/features/turbulence.php
   
   Quote:

   Detached eddy simulation (DES): Hybrid method that treats near-wall regions with a RAS approach and the bulk flow with an LES approach. DES models are found in the list of LES models.

2. http://www.openfoam.com/features/LES.php
   
   Quote:

   SpalartAllmarasDDES - Spalart-Allmaras delayed detached eddy simulation (DDES) model SpalartAllmarasIDDES - Spalart-Allmaras improved DDES (IDDES) model

3. https://github.com/OpenFOAM/OpenFOAM...AllmarasDDES.H
   
   Quote:

   Code: SpalartAllmaras DDES LES turbulence model for incompressible flows Reference: P.R. Spalart, S. Deck, S., M.L.Shur, K.D. Squires, M.Kh Strelets, and A. Travin. `A new version of detached-eddy simulation, resistant to ambiguous grid densities'. Theor. Comp. Fluid Dyn., 20:181-195, 2006.

4. https://github.com/OpenFOAM/OpenFOAM...llmarasIDDES.H
   
   Quote:

   Code: SpalartAllmarasIDDES LES turbulence model for incompressible flows

For more information I suggest that you ask the people at OpenCFD directly, through their support page.

Best regards,
Bruno

Thank you so much Bruno
Yeah, I've seen the sources of DDES & IDDES in OpenFOAM.
O.K. I will contact them.
Thanks

Hi Mohsen,

Please do share the answer with us, if you get one from them!

Best regards,
Bruno

Hi, I have some questions about SA - DDES model. Since your question is on SA IDDES which is somewhat related to mine, I feel it is quite close to the question that I wanted to ask.

My question is : Is the SA - DDES model implemented in OpenFoam 2.1.x a low-Re or high-Re model? what is the best range of the first point Yplus for the walls?
can it be used with wall functions for High-Re problems?or it needs very fine mesh near the wall?

Greetings Alvin,

From what I can see, it's for High Reynolds:

• SpalartAllmaras in RAS indicates this: https://github.com/OpenFOAM/OpenFOAM...lartAllmaras.H
• If you look at SpalartAllmaras in LES, you'll see that they do not mention any extended features: https://github.com/OpenFOAM/OpenFOAM...lartAllmaras.H
• The SA DDES is derived from SpalartAllmaras in LES (SA-DES) and seems to only be slightly changed. There is no clear reference to being applicable in low Reynolds.

As mentioned in post #13, only IDDES seems to be applicable to low-Reynolds.


As for the other questions, I don't know. But I do believe that you could try running a few simulations to diagnose this.

Best regards,
Bruno

LES volume less than RAS volume IDDES case
 

Dear All,

I am trying to run a IDDES simulation, well it seems to be running without any issues but the only problem I can find is the

That LES percentage is so less compared to the RANS, how can I resolve this issue ? any ideas ?

- I ran a URANS simulation and used one of the very well last converged time step to start this simulation !!! say 0.3576 time step

Would it be starting the simulation with the first time step of URANS making it more of a RANS simulation or something ?

What could be happening here, from what I know the RANS is suposed to be only close to the walls of the simulation and not the entire domain for IDDES,

This is my first IDDES simulation so please explain well,

Thanks for your time and effort,

Edit:
If the post here hasnt given enough information, I have also posted a similar question here http://www.cfd-online.com/Forums/ope...tml#post577568

Kind Regards,
Hasan K.J