|
[Sponsors] |
Which turbulence models in openFOAM can be solved up to the viscous sub-layer ? |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
September 24, 2017, 10:06 |
Which turbulence models in openFOAM can be solved up to the viscous sub-layer ?
|
#1 |
Member
Paul Palladium
Join Date: Jan 2016
Posts: 93
Rep Power: 10 |
Dear Foamers,
I am quite lost with wall treatment in openFOAM. As far as I know Omega based models can be solved in the viscous sublayer. For exemple : standard k-omega and k-omega SST. With these turbulence models it's possible to use low and high reynolds boundary conditions for wall : type omegaWallFunction; (valid for both low and high reynolds. The condition based the blanding of viscous and log layer) type kqRWallFunction; (For high Re, zero gradient condition) type kLowReWallFunction; (for high and low Re, the condition based on the position of y+) But when using k-epsilon model, for instance Standard, RNG or Realizable, is it possible to go down in to the viscous sublayer ? By using this BC : type epsilonLowReWallFunction (for High and Low Re, the condition based on the position of y+) I am more than doubtful that it's possible to use this BC with the standard k-epsilon turbulence model. If you have any ideas on the subject I would be very grateful. Regards |
|
September 25, 2017, 00:59 |
|
#2 |
Senior Member
Uwe Pilz
Join Date: Feb 2017
Location: Leipzig, Germany
Posts: 744
Rep Power: 15 |
If you mesh you geometry in a way that you resolve the Prandtl layer you don't need any turbulence model: You don't model the turbulence in that layer in an indirect way, but cover it by the NS equations.
If you use a model anyway the coefficients get very small and switch of the modelling this way.
__________________
Uwe Pilz -- Die der Hauptbewegung überlagerte Schwankungsbewegung ist in ihren Einzelheiten so hoffnungslos kompliziert, daß ihre theoretische Berechnung aussichtslos erscheint. (Hermann Schlichting, 1950) |
|
September 25, 2017, 05:29 |
|
#3 |
Senior Member
Joachim Herb
Join Date: Sep 2010
Posts: 650
Rep Power: 21 |
As piu58 wrote depending on your y+ values you might need wall functions or not (there are several forum threads about this topic). But nevertheless, you will need a turbulence model (if you don't resolve all of your domain with such a fine mesh that you can do DNS).
|
|
September 25, 2017, 06:24 |
|
#4 |
Member
Paul Palladium
Join Date: Jan 2016
Posts: 93
Rep Power: 10 |
I don't understand your answers. Are you sure is it legit to calculate the viscous sub layer (go to y + ~1) with a k-epsilon type model when using the epsilonLowReWallFunction BC ?
I mean there is no damping function in the k-epsilon models available in OpenFOAM. As far as I know there is no low reynolds formulation for k-espilon model implemented in OpenFOAM. Theoretically the turbulent dissipation goes to infinity when approaching the wall. That's why k-epsilon models need wall functions, or a two layers wall treatment with Wolfstein model for example, or damping functions. BUT in OpenFOAM there is a boundary condition epsilonLowReWallFunction. The code is here : https://cpp.openfoam.org/v4/epsilonL...8C_source.html This BC set two different values of the turbulent dissipation depending on the position of the first cell (below or upper y+~11) So :when using this BC what is the recommendations for y+ ? And moreover, Is it ok to use this BC with (for exemple) the standard k-epsilon model ? For exemple : In commercial code CFX, when using k-epsilon model it's only possible to use scalable wall function. This wall treatment avoid the deterioration of standard wall functions under grid refinement below y+ = 11. "CFX force the usage of the log law in conjunction with the standard wall functions approach by limiting the minimal y+ computed at the wall. So in CFX it's not possible to go below y+~11 wen using a k-epsilon type model" Regards Last edited by Fauster; September 26, 2017 at 03:19. |
|
September 26, 2017, 03:32 |
|
#5 |
Member
Paul Palladium
Join Date: Jan 2016
Posts: 93
Rep Power: 10 |
I guess I found the answer. I met some difficulties because I thought that the boundary condition epsilonLowReWallFunction was based on the standard wall treatment which is pretty bad when going below y+~11.
In fact the boundary condition epsilonLowReWallFunction is more like the Non-Equilibrium Wall Functions treatment available in FLUENT which is design to avoid the main inconvenient of standard wall functions. Do you agree with me ? Regards |
|
June 15, 2019, 18:09 |
|
#6 | |
Member
CFD USER
Join Date: May 2019
Posts: 40
Rep Power: 6 |
Quote:
|
||
February 15, 2021, 11:49 |
|
#7 | |
New Member
Josef Dobeš
Join Date: Mar 2012
Location: Czech republic
Posts: 12
Rep Power: 14 |
Quote:
Did you tested the setup for LowRe kepsilon group of models for some basic cases? Because I will try it, simple straight circular tube (ID 33 mm) and results are not good, even if I have yplus below 1... Thanks for your comment. |
||
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
OpenFOAM Training Jan-Jul 2017, Virtual, London, Houston, Berlin | CFDFoundation | OpenFOAM Announcements from Other Sources | 0 | January 4, 2017 06:15 |
OpenFOAM v3.0+ ?? | SBusch | OpenFOAM | 22 | December 26, 2016 14:24 |
Adding a custom turbulence model in OpenFOAM 3.0 | Sylvain | OpenFOAM Programming & Development | 3 | July 18, 2016 08:15 |
Questions about Boundary Layer Thickness and Turbulence Models | famerfamer | STAR-CCM+ | 3 | July 12, 2012 09:47 |
Discussion: Reason of Turbulence!! | Wen Long | Main CFD Forum | 3 | May 15, 2009 09:52 |