[Jane L]

Hello!

For modeling an unsteady stall phenomenom, I'd like to use DES which uses k-omega in the near-wall region as RAS model. I just found the original DES with Spalart-Allmaras in OpenFOAM 2.1.0.
Before I start programming: Is there a method in OpenFOAM (maybe in .x or .dev) for DES with k-omega or did anyone of you implement this before?

Thanks in advance!

[Jane L]

Hi Foamers!

Finally I finished the programming of the turbulence models which you can find in the attached archive.

This file contains the source code for the tree turbulence models DES, DDES and IDDES with k-w as SGS model (instead od Spalart Allmaras).

The turbulence models where tested with OpenFOAM 2.1.1 and gave fair results for airfoils with very high angles of attack. Lift and drag was compared to experimental values given in the paper "Detached-Eddy Simulation of an airfoil at high angles of attack" by Garbaruk, Shur, Strelets and Travin. The averaged aberration in lift and drag for angles of attack of 45 and 60 degree was always below 10%.

All turbulence models include an extra function "RANSregion" which gives an indication of the RANS- and LES-content of each cell. This function can be visualized with paraview and has positive values in cells where the turbulence is fully modelled by the RANS model. The RANS model works in SGS mode in cells with values lower 0.

I hope that this will be helpful. In case you use it, I would be glad if you give some feedback.

best regards!

[mrs206]

Hi Jane,

Thank you very much for writing these. They have been very useful in helping me to create a k-epsilon based DES model. I used the modifications proposed by Strelets for the turbulent length scale (i think this is the same method you have used?) but i have been unable to find a reference for the method you have used for the DDES model you have implemented.

Please could you point me in the right direction?

Many thanks,

Michael

[Jane L]

Hi Michael.

The DDES model was developed and first described by: P. Spalart, S. Deck, M. Shur, K. Squires, M. Strelets, A. Travin.
in: A new version of detached-eddy simulation, resistant to ambiguous grid densities.
You can find the paper about IDDES via google:
M. Shur, P. Spalart, M. Strelets, A. Travin. A hybrid RANS-LES approach
with delayed-DES and wall-modelled LES capabilities

For might also have a look at:
http://www.sharcnet.ca/Software/Flue.../th/node91.htm
http://www.sharcnet.ca/Software/Flue...sec_iddes.html

I hope this helps! Kind regards

[killsecond]

Hi Jane,
Have you tried to creat kOmegaSST based DES?

[saiguruprasad]

I am trying to use k-Omega SST DES model in OpenFOAM 5.x. I am trying to simulate turbulent Couette Flow at Re_w=8600.

I have tried SA DES model and was able to get satisfactory results(used perturbUChannel to set initial condition). But when I use the kOmega SST DES model implemented in OF-5.x, there is no difference between the instantaneous and mean velcoity magnitude. They both become the same after few iterations. This results in the wrong calculation of Reynolds stress.

Am I doing something wrong in setting up the simulation?

[Sakun]

Originally Posted by saiguruprasad

I am trying to use k-Omega SST DES model in OpenFOAM 5.x. I am trying to simulate turbulent Couette Flow at Re_w=8600.

I have tried SA DES model and was able to get satisfactory results(used perturbUChannel to set initial condition). But when I use the kOmega SST DES model implemented in OF-5.x, there is no difference between the instantaneous and mean velcoity magnitude. They both become the same after few iterations. This results in the wrong calculation of Reynolds stress.

Am I doing something wrong in setting up the simulation?

Hi,

i am trying to run a simulation using SA-DES. i really really appreciate if you can guide me or the way you learnt.

[B6_lacen]

Originally Posted by saiguruprasad

I am trying to use k-Omega SST DES model in OpenFOAM 5.x. I am trying to simulate turbulent Couette Flow at Re_w=8600.

I have tried SA DES model and was able to get satisfactory results(used perturbUChannel to set initial condition). But when I use the kOmega SST DES model implemented in OF-5.x, there is no difference between the instantaneous and mean velcoity magnitude. They both become the same after few iterations. This results in the wrong calculation of Reynolds stress.

Am I doing something wrong in setting up the simulation?

Hi

I'm really lost I'm looking for the implementation of the model kOmega-SST DDES in the versions of openFOAM V6.
Can you show me how you did it please