[Amrita Lonkar]

SU2 now adds halo cells to only one side of the mesh. We do not require two sets of halo cells, and the information is exchanged between the set of halo cells, and a set of interior points on the other side of the mesh.

[shirazbj]

Originally Posted by Amrita Lonkar

SU2 now adds halo cells to only one side of the mesh. We do not require two sets of halo cells, and the information is exchanged between the set of halo cells, and a set of interior points on the other side of the mesh.

In a cfg file, the sender and receiver is define twice and the PBC program calculates twice.

Is the added halo cells a copy of the set of interior points on the other side?

[ITA91]

*bump*

It has been a while since the last post in this thread, but for me it seemed suitable to post it here since this thread is the only one I could find on the SU2 part of this forum considering periodic boundary conditions.

I'm trying to do a relatively simple flow analysis of a Caradonna & Tung rotor for compressible, inviscid rotating flow at 2500 rpm. Hence, this case is almost similar to the case provided in the Test Cases, I'm just using my own mesh instead of the one provided in the Test Cases. The main difference is that I try to run this case using a periodic boundary condition, which should be possible given that the situation is axisymmetric. Unfortunately, I'm experiencing some issues with this. First of all, the SU2_PBC module seems to have disappeared from the 3.2 version, is this correct? I couldn't find it on the Github page. Luckily my supervisor still has the 3.1 version installed next to the newest one, this version still includes SU2_PBC. I have adapted the mesh using the SU2_PBC executable from his V3.1 and then I used my own V3.2 to run the CFD simulation. For the first ~30 iterations everything seems to go well and it seems to converge nicely, but afterwards the residuals start to increase and the simulation diverges. I have already had to add a shaft at the rotation axis to prevent mismatches in the mesh after running SU2_PBC, such that my mesh looks similar to the case presented here: http://www.task.gda.pl/files/quart/T...4/tq312j-e.pdf. Also see the screenshot that I've attached (from Pointwise)

So, does anybody know what I can try to do in order to solve this? I've tried copying this mesh and rotating it 180 degrees about the x-axis to obtain the "full" grid in order to check if this might yield some new information, but the CFD simulation using this mesh converges normally. I've also attached a view from the periodic_halo.plt file that SU2_PBC gives as output. It only shows halo cells on one of the periodic boundaries, is this correct? In the configuration file I have defined the MARKER_PERIODIC both ways so this shouldn't be the problem:

MARKER_PERIODIC= ( periodic1, periodic2, 0.0, 0.0, 0.0, -180.0, 0.0, 0.0, 0.0, 0.0, 0.0, periodic2, periodic1, 0.0, 0.0, 0.0, +180.0, 0.0, 0.0, 0.0, 0.0, 0.0 )

Running the simulation without adapting the mesh doesn't help either; the simulation starts but it already has non-physical points from the beginning and the calculation quickly diverges.

Thanks in advance for your help!

[Lokanayaki]

Hi Tom,
I am looking for tutorial cases with periodic boundary condition. I tried to download the files from the link provided but it seems that the link does not exist. Could you please provide the files. It will be of great help.

[jomunkas]

Hi,

I have a problem with periodic boundary conditions. I am generating the mesh with Pointwise which should ensure the periodic matching but I am getting "segmentation fault" error message:

-------------------------------------------------------------------------
|    ___ _   _ ___                                                      |
|   / __| | | |_  )   Release 6.1.0  "Falcon"                           |
|   \__ \ |_| |/ /                                                      |
|   |___/\___//___|   Suite (Mesh Adaptation Code)                      |
|                                                                       |
-------------------------------------------------------------------------
| The current SU2 release has been coordinated by the                   |
| SU2 International Developers Society <www.su2devsociety.org>          |
| with selected contributions from the open-source community.           |
-------------------------------------------------------------------------
| The main research teams contributing to the current release are:      |
| - Prof. Juan J. Alonso's group at Stanford University.                |
| - Prof. Piero Colonna's group at Delft University of Technology.      |
| - Prof. Nicolas R. Gauger's group at Kaiserslautern U. of Technology. |
| - Prof. Alberto Guardone's group at Polytechnic University of Milan.  |
| - Prof. Rafael Palacios' group at Imperial College London.            |
| - Prof. Vincent Terrapon's group at the University of Liege.          |
| - Prof. Edwin van der Weide's group at the University of Twente.      |
| - Lab. of New Concepts in Aeronautics at Tech. Inst. of Aeronautics.  |
-------------------------------------------------------------------------
| Copyright 2012-2018, Francisco D. Palacios, Thomas D. Economon,       |
|                      Tim Albring, and the SU2 contributors.           |
|                                                                       |
| SU2 is free software; you can redistribute it and/or                  |
| modify it under the terms of the GNU Lesser General Public            |
| License as published by the Free Software Foundation; either          |
| version 2.1 of the License, or (at your option) any later version.    |
|                                                                       |
| SU2 is distributed in the hope that it will be useful,                |
| but WITHOUT ANY WARRANTY; without even the implied warranty of        |
| MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU      |
| Lesser General Public License for more details.                       |
|                                                                       |
| You should have received a copy of the GNU Lesser General Public      |
| License along with SU2. If not, see <http://www.gnu.org/licenses/>.   |
-------------------------------------------------------------------------

------------------------ Physical Case Definition -----------------------
Input mesh file name: mesh_simple.su2

----------------------- Grid adaptation strategy ------------------------
Grid modification to run periodic bc problems.

-------------------------- Output Information ---------------------------
Output mesh file name: mesh_out.su2. 
Output mesh file name: mesh_out.su2. 
Restart flow file name: restart_flow.dat.

------------------- Config File Boundary Information --------------------
Euler wall boundary marker(s): HUB, FF.
Far-field boundary marker(s): IN.
Periodic boundary marker(s): SIDE-, SIDE+.
Outlet boundary marker(s): OUT.

---------------------- Read Grid File Information -----------------------
Three dimensional problem.
303 points.
6 surface markers.
126 boundary elements in index 0 (Marker = FF).
66 boundary elements in index 1 (Marker = HUB).
50 boundary elements in index 2 (Marker = IN).
50 boundary elements in index 3 (Marker = OUT).
44 boundary elements in index 4 (Marker = SIDE+).
44 boundary elements in index 5 (Marker = SIDE-).
1158 tetrahedra.

----------------------- Preprocessing computations ----------------------
Setting local point and element connectivity.
Check numerical grid orientation.
Identify faces, edges and vertices.
Set control volume structure.
Volume of the computational grid: 0.589273.

-------------------- Setting the periodic boundaries --------------------
Setting the periodic boundary conditions.
Checking SIDE+ boundary against periodic donor, SIDE-. 
Checking SIDE- boundary against periodic donor, SIDE+. 
Number of new points: 39.
Number of new interior elements: 145.
Number of new boundary elements added to preexisting markers: 37.

------------------------- Solver Postprocessing -------------------------
[langos-lnx:12882] *** Process received signal ***
[langos-lnx:12882] Signal: Segmentation fault (11)
[langos-lnx:12882] Signal code: Address not mapped (1)
[langos-lnx:12882] Failing at address: 0x80
[langos-lnx:12882] [ 0] /lib/x86_64-linux-gnu/libpthread.so.0(+0x11390)[0x7f36aed7c390]
[langos-lnx:12882] [ 1] SU2_MSH[0x547e6a]
[langos-lnx:12882] [ 2] SU2_MSH[0x54b269]
[langos-lnx:12882] [ 3] SU2_MSH[0x41cf66]
[langos-lnx:12882] [ 4] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7f36ae9c1830]
[langos-lnx:12882] [ 5] SU2_MSH[0x47ab79]
[langos-lnx:12882] *** End of error message ***
Segmentation fault (core dumped)

Do you know what I am doing wrong?


Kind Regards,
Zeno

[pcg]

Can you use version 7? This step through SU2_MSH is no longer required.

[jomunkas]

Originally Posted by pcg

Can you use version 7? This step through SU2_MSH is no longer required.

Indeed, version 5 and 7 does the job without a problem.