|
[Sponsors] |
September 26, 2020, 11:40 |
Usage of runTimePostProcessing
|
#1 |
Member
Andy
Join Date: Jul 2017
Posts: 62
Rep Power: 9 |
Hi folks,
I would like to make a movie from a DNS-Simulation. Therefore I want to safe fotos(frames) from each time step. My Domain is a simple cube with turbulence inside. I work with OF-1712 sine this special DNS-Solver only exists for this version. The only fuction I know to produce scuch "screenshots" is: runTimePostProcessing Unfortunatelly there is noproper guide for it. The examples in the tutorials and other websites helped only a little. Let's say, I want to make a picture of this cube-domain visualising velocity. You find attached the settings of the runTimePostProcessing as well as the log file with the sim-feedback. Sooo, what I've done wrong? What would be the right syntax here? Maybe somebody has a good example file? Thank you in advance. runtimePostProcessing.txt log.txt |
|
September 28, 2020, 04:13 |
|
#2 |
New Member
Tu Can
Join Date: Jul 2017
Location: China
Posts: 16
Rep Power: 9 |
maybe helpful. I use paraview to make movie. run it parallelly.
save as movies directly or step time picture, both are OK. |
|
September 28, 2020, 04:18 |
|
#3 |
Member
Andy
Join Date: Jul 2017
Posts: 62
Rep Power: 9 |
Thanks bestucan, but I really do not want to safe the results before reading them in paraview. I want to make the movie from the calculation without saving the profiles. Otherwise it would need many terrabytes and a lot of cpu-power would be lost for the saving
|
|
December 2, 2021, 13:58 |
Solution ?
|
#4 | |
New Member
Balaji
Join Date: May 2013
Posts: 21
Rep Power: 13 |
Quote:
Any updates how one can output pictures as jpeg/png and write out every time step or duration? Please post the solution if you have found |
||
December 2, 2021, 16:21 |
|
#5 | |
Senior Member
Mark Olesen
Join Date: Mar 2009
Location: https://olesenm.github.io/
Posts: 1,715
Rep Power: 40 |
Quote:
That is what was used to make the laser melting video. Was a bit of overkill writing png images each time step since it would have made the video far too lengthy, but nonetheless cool. Using runTimePostPro should generally be the fastest, since you can reuse OpenFOAM sampled elements etc and don't have pass things through paraview at all. However, probably only useful for production runs since setting up scenes is a bit of a pain. The Catalyst viz is certainly a bit more wysiwyg Take a look at the video: |
||
December 3, 2021, 04:45 |
|
#6 | |
New Member
Balaji
Join Date: May 2013
Posts: 21
Rep Power: 13 |
Quote:
The problem is we are running transient analysis on cluster, on cluster the png creation is fast using runTimePostPro. Also additional software installation like Catalyst need extra costs. The runTimePostPro tutorial is the best option on cluster. But I am not how to implement it from the tutorials using ("librunTimePostProcessing.so"); libraries. |
||
December 3, 2021, 07:28 |
|
#7 | |
Senior Member
Mark Olesen
Join Date: Mar 2009
Location: https://olesenm.github.io/
Posts: 1,715
Rep Power: 40 |
Quote:
The wind around buildings is perhaps the most extensive example. In general, you would want to take some portion of an existing simulation and then use that for doing the setup (eg, solverName -postProcess) for runTimePostProcessing. There is unfortunately not as much documentation as would be nice (will likely need to explore the source code as well). If you do get something interesting that could either be added to an existing tutorial, or add a new one, this would be a most welcome contribution. That way others can benefit too. Once it gets that far, please open an issue on develop.openfoam.com to get your tutorial added OR perhaps through the tutorials wiki (https://wiki.openfoam.com/index.php?title=Tutorials) |
||
December 27, 2023, 14:44 |
|
#8 | |
Senior Member
ONESP-RO
Join Date: Feb 2021
Location: Somwhere on Planet Earth
Posts: 127
Rep Power: 5 |
Quote:
Code:
Time = 50 smoothSolver: Solving for Ux, Initial residual = 0.00614156, Final residual = 0.000503708, No Iterations 2 smoothSolver: Solving for Uy, Initial residual = 0.0155772, Final residual = 0.00125321, No Iterations 2 smoothSolver: Solving for Uz, Initial residual = 0.0177427, Final residual = 0.00150379, No Iterations 2 GAMG: Solving for p, Initial residual = 0.0704188, Final residual = 0.00267215, No Iterations 2 time step continuity errors : sum local = 8.15277e-05, global = 6.08291e-06, cumulative = -0.000108089 smoothSolver: Solving for epsilon, Initial residual = 0.00588112, Final residual = 0.000463942, No Iterations 1 smoothSolver: Solving for k, Initial residual = 0.0141518, Final residual = 0.000249249, No Iterations 2 ExecutionTime = 31.9 s ClockTime = 57 s ensightWrite ensightWrite write: ( epsilon k p U ) vtkWrite output Time: 50 Internal : "postProcessing/vtkWrite/windAroundBuildings_00000050/internal.vtu" Boundary : "postProcessing/vtkWrite/windAroundBuildings_00000050/boundary/inlet.vtp" Boundary : "postProcessing/vtkWrite/windAroundBuildings_00000050/boundary/outlet.vtp" Boundary : "postProcessing/vtkWrite/windAroundBuildings_00000050/boundary/ground.vtp" Boundary : "postProcessing/vtkWrite/windAroundBuildings_00000050/boundary/frontAndBack.vtp" Boundary : "postProcessing/vtkWrite/windAroundBuildings_00000050/boundary/buildings.vtp" volScalarField(epsilon k p) volVectorField(U) subset output Time: 50 Internal : "postProcessing/subset/windAroundBuildings_00000050/internal.vtu" Boundary : "postProcessing/subset/windAroundBuildings_00000050/boundary/inlet.vtp" Boundary : "postProcessing/subset/windAroundBuildings_00000050/boundary/outlet.vtp" Boundary : "postProcessing/subset/windAroundBuildings_00000050/boundary/ground.vtp" Boundary : "postProcessing/subset/windAroundBuildings_00000050/boundary/frontAndBack.vtp" Boundary : "postProcessing/subset/windAroundBuildings_00000050/boundary/buildings.vtp" volScalarField(p) volVectorField(U) volScalarField->point(p) volVectorField->point(U) streamLine streamLines write: seeded 40 particles Tracks:40 Total samples:17898 postPro1 render 2023-12-27 12:00:51.230 ( 60.288s) [ B8BA2880]vtkXOpenGLRenderWindow.:464 ERR| vtkXOpenGLRenderWindow (0x11004f30): bad X server connection. DISPLAY=
__________________
Don't keep making the same mistakes. Try to make new mistakes. |
||
December 27, 2023, 15:02 |
|
#9 |
Senior Member
Mark Olesen
Join Date: Mar 2009
Location: https://olesenm.github.io/
Posts: 1,715
Rep Power: 40 |
Did you compile VTK with osmesa?
|
|
December 28, 2023, 07:44 |
|
#10 |
Senior Member
ONESP-RO
Join Date: Feb 2021
Location: Somwhere on Planet Earth
Posts: 127
Rep Power: 5 |
I have no idea how it was compiled since I am using it on the cluster.
How do I check if it is compiled with osmesa? Update: Never mind, I had to check if any libvtk*.so has a dependency on osmesa using ldd command. I think the VTK version used during the OpenFOAM building was not built with osmesa.
__________________
Don't keep making the same mistakes. Try to make new mistakes. Last edited by NotOverUnderated; December 28, 2023 at 09:58. |
|
December 29, 2023, 07:21 |
|
#11 |
Senior Member
ONESP-RO
Join Date: Feb 2021
Location: Somwhere on Planet Earth
Posts: 127
Rep Power: 5 |
I got runTimePostProcessing function object running.
0) Install osmesa, on Ubuntu 22.04: Code:
sudo apt install libosmesa6 libosmesa6-dev 2) cd to VTK directory and create a directory 'build' and cd to it. 3) run: Code:
ccmake .. -DVTK_OPENGL_HAS_OSMESA=ON 4) Now, to find the VTK module required by the runTimePostProcesisng function object, clone the openfoam repository and the visualization module. 5) cd to the runTimePostProcessing directory and find the required modules by running: Code:
grep -r 'VTK::' # For me the output was: VTK::FiltersCore VTK::FiltersGeometry VTK::FiltersSources VTK::IOGeometry VTK::IOImage VTK::IOLegacy VTK::IOXML VTK::ParallelCore VTK::ParallelMPI VTK::RenderingAnnotation VTK::RenderingCore VTK::RenderingParallel 7) Compile vtk and install it Code:
sudo make install #will be installed by default under /usr/local/ Code:
export VTK_DIR=/usr/local 9) Clone OpenFOAM repository (skip if you did in step 4). 10) Build OpenFOAM. 11) Compile the runTimePostProcessing library. 12) If you have tried step (8*) make sure to remove the librunTimePostProcessing.so from $FOAM_USER_LIBBIN and $FOAM_LIBBBIN and replace them with the newly compiled one after building OpenFOAM from source. 13) You must run this to get it to work for this session to load (you need to run this only this time): Code:
sudo ldconfig The steps above are not perfect but they should work.
__________________
Don't keep making the same mistakes. Try to make new mistakes. Last edited by NotOverUnderated; December 29, 2023 at 20:15. |
|
October 17, 2024, 13:16 |
|
#12 |
New Member
Leung Ho Man
Join Date: Aug 2020
Posts: 9
Rep Power: 6 |
Has anyone found better runtimepostpro documentation? it's been a pain trying to get it running. Thanks!
|
|
Tags |
1712, openfoam, postprocessing, runtime, runtimepostprocessing |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Large memory usage for steady simulation | Wenqiang | SU2 | 2 | May 15, 2020 14:40 |
[ANSYS Meshing] is my memory usage normal? | Renold | ANSYS Meshing & Geometry | 2 | April 9, 2019 15:09 |
New workstation for different usage scenarios - CPU and RAM | natem | Hardware | 6 | August 7, 2013 03:47 |
Boosting CPU usage | earlybird | FLUENT | 2 | November 2, 2012 11:32 |
OpenFOAM Solver/BC usage description | murrayjc | OpenFOAM | 3 | August 25, 2009 05:48 |