|
[Sponsors] |
Convergence issues with RANS, Axisymmetric: SU2 has diverged (NaN detected). |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
June 12, 2024, 12:37 |
Convergence issues with RANS, Axisymmetric: SU2 has diverged (NaN detected).
|
#1 |
New Member
Join Date: Apr 2024
Posts: 3
Rep Power: 2 |
Hello,
I am little new to open source codes, I would really appreaciate some input on this issue, So, I have been trying to simulate a RANS 2d external turbojet engine flow, using an axisymmetric domain, of 923k elements for a domain spread across 5m by 3m, with SST turbulence model, I have not implemented any wall functions, as it keeps throwing me warnings of y+<5 or y+<1 points. but when I turn of the wall functions, after a while the solution solves and diverges. or is it because of the CFL number being 15 being a bit higher....not sure what exactly is causing the divergence. As you can see from the txt file, it runs upto 800 iterations and diverges, I dont know if my mesh is not fine enough or if i should use a coarser mesh, or is there some configuration issue within the convergence parameters or boundary conditions, I have also attached a velcity contour that I visualized before it diverged, the jet velocity and mach number are within the anticipated range (which gives me a hope that the boundary conditions are right..), I dont know why the simulation diverges still. I have also attached my .cfg file under the name of "forum.txt". I am running on 6 cores in mpi on my laptop. Thankyou!! |
|
June 12, 2024, 14:31 |
|
#2 |
Senior Member
bigfoot
Join Date: Dec 2011
Location: Netherlands
Posts: 670
Rep Power: 21 |
Hi, can you share the mesh as well?
|
|
June 12, 2024, 14:57 |
|
#3 |
New Member
Join Date: Apr 2024
Posts: 3
Rep Power: 2 |
Here is the mesh, around the farfield and near the jet, I am able to attach only attach images, since the forum has only certain file types allowed for uploading.
|
|
June 12, 2024, 18:50 |
|
#4 |
Senior Member
bigfoot
Join Date: Dec 2011
Location: Netherlands
Posts: 670
Rep Power: 21 |
I mean as a file, so I can run it...
By the way, it looks really really fine, you can probably reduce the number of cells by a lot. That is also what the warning message is telling you: the mesh is so fine that the wall model is switched off because you are fully resolving the boundary layer. |
|
June 13, 2024, 10:08 |
|
#5 |
New Member
Join Date: Apr 2024
Posts: 3
Rep Power: 2 |
I think you maybe right, so I tried it with a coarser mesh, and I activated Standard Wall Functions, and the solution still seems to be diverging because it says something like..
| 81| -3.278009| -0.867095| -0.862354| 2.054079| Warning: Computation of wall coefficients (y+) did not converge in 1 points. Warning: y+ < 5 in 126 points, for which the wall model is not active. Warning: Computation of wall coefficients (y+) did not converge in 1 points. Warning: y+ < 5 in 128 points, for which the wall model is not active. Warning: T_Wall < 0 I have attached the link to the coarser mesh and the .txt file of the new error herewith, https://drive.google.com/file/d/1Zz1...ew?usp=sharing with the coarser mesh, it crashes (diverges) sooner... Thankyou for your time!! Last edited by Vaths; June 13, 2024 at 11:11. |
|
Tags |
axisymmetric 2d, external flow, rans, su2 |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Divergence detected in AMG solver. VOF. | Mr.Mister | Fluent Multiphase | 5 | November 22, 2024 07:32 |
Error SIGSEGV using VOF and UDF | JERC_UTFSM | Fluent UDF and Scheme Programming | 14 | November 8, 2021 00:17 |
Axisymmetric simpleFoam convergence issues | benham | OpenFOAM Running, Solving & CFD | 2 | August 15, 2016 07:47 |
Why simulation results showing "nan" values? | Sargam05 | OpenFOAM | 4 | September 10, 2012 13:37 |
Could anybody help me see this error and give help | liugx212 | OpenFOAM Running, Solving & CFD | 3 | January 4, 2006 19:07 |