|
[Sponsors] |
November 23, 2023, 13:52 |
Turbulence model help
|
#1 |
New Member
Mathew
Join Date: Jan 2023
Posts: 5
Rep Power: 3 |
After reviewing the topic continuously, I am left with the same question regarding turbulence models on CFX. Most literature related to my topic - surge in a compressors plenum (Pipe with reverse flow) - use the k-e model. However, due to computational limitations, due to the scale of the domain, I can not use a low y+ value. Therefore, I am using the scalable wall function for this problem. Basically, im looking for advice on what y+ to use for my first element size (should it be between 30 and 300?). Also, do you think this model will accurately predict the pressure drop in the pipe?
|
|
November 23, 2023, 18:51 |
|
#2 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,703
Rep Power: 143 |
Some general comments:
* Most researchers use k-e as it is the default turbulence model in most cases. It has well known advantages and disadvantages, so you should make sure you are aware of them so you can judge whether is it appropriate in your case. * k-e uses wall functions on the wall, and these are not applicable for y+<11. So for k-e you need to make sure your mesh is not too fine on the walls. * In CFX the SST model is the one recommended for general use. It can handle any y+ on the wall with the automatic wall functions. To determine what y+ you need you should do a sensitivity study - which is a mesh sensitivity study as it will involve refining the mesh. For a pipe flow with few separations you might well be able to use quite a large y+. But you cannot be sure unless you do a sensitivity study and confirm it for your case.
__________________
Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum. |
|
November 25, 2023, 13:34 |
|
#3 | |
Senior Member
Join Date: Jun 2009
Posts: 1,804
Rep Power: 32 |
Quote:
I rarely see the use of k-e anymore in the industrial turbomachinery community. Keep in mind that SST, as well as BSL should produce nearly the same solution as the k-e model in a coarse mesh. A one-sentence summary for SST (sure it is more complex, but in short): SST/BSL = blend * k-w + (1-blend) * k-e + heuristics
__________________
Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum. |
||
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[GUIDE] Switching turbulence model to SpalartAllmaras | gabrielfelix | OpenFOAM Running, Solving & CFD | 1 | March 24, 2022 20:19 |
Question about adaptive timestepping | Guille1811 | CFX | 25 | November 12, 2017 17:38 |
Overflow Error in Multiphase Modelling with Two Continuous Fluids | ashtonJ | CFX | 6 | August 11, 2014 14:32 |
An error has occurred in cfx5solve: | volo87 | CFX | 5 | June 14, 2013 17:44 |
Wrong calculation of nut in the kOmegaSST turbulence model | FelixL | OpenFOAM Bugs | 27 | March 27, 2012 09:02 |