View Issue Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0003603 | OpenFOAM | Feature | public | 2020-11-27 14:19 | 2020-11-27 15:03 |
Reporter | thw1021 | Assigned To | henry | ||
Priority | none | Severity | feature | Reproducibility | always |
Status | closed | Resolution | no change required | ||
Platform | GNU/Linux | OS | Ubuntu | OS Version | 18.04 LTS |
Summary | 0003603: reconstructPar utility cannot correctly reconstruct velocity distribution on inlet | ||||
Description | I followed this tutorial [1] to implement a new parabolic velocity profile on inlet boundary. If run the simulation in serial, the results are OK. However, if run in parallel, the reconstructed velocity distribution on inlet is wrong. I found this problem can be avoided if I decompose the domain in the direction which the whole inlet patch is assigned to only one processor, otherwise the reconstructed inlet velocity will be wrong. I have also reported this problem on paraview [2] | ||||
Steps To Reproduce | 1. cd into parabolicVelocity directory and run wclean && wmake to compile the libparabolicVelocity.so lib. 2. cd into simpleFoamParabolicInlet, run decomposePar to decompose the domain, and then run mpirun -np 12 simpleFoam -parallel > log.simpleFoam 2>&1 3. run reconstructPar to reconstruct the domain 4. open case.foam in paraview, you can see the wrong inlet velocity as shown in the attached figure. | ||||
Additional Information | Reference: [1] http://www.tfd.chalmers.se/~hani/kurser/OS_CFD_2019/lectureNotes/ImplementParabolicVelocityFixedValueBoundaryCondition.pdf [2] https://gitlab.kitware.com/paraview/paraview/-/issues/20346 | ||||
Tags | No tags attached. | ||||
|
|
|
The parabolicVelocity code is not part of OpenFOAM and hence not supported by the OpenFOAM Foundation. We could implement it correctly for you or help you correct it but this user support would need to be funded. |
|
Do you mean this is a problem related with the parabolicVelocity code but not the reconstructPar utility ? The parabolicVelocity code should be right since I have tested without error as long as the inlet patch is assigned to only one processor. |
|
> Do you mean this is a problem related with the parabolicVelocity code Yes |
|
Emm..., OK, thank you. I am learning OpenFOAM. I'd better try figure out this problem by myself at first. |
|
User support request. |
Date Modified | Username | Field | Change |
---|---|---|---|
2020-11-27 14:19 | thw1021 | New Issue | |
2020-11-27 14:28 | thw1021 | File Added: case.tar.xz | |
2020-11-27 14:33 | henry | Note Added: 0011775 | |
2020-11-27 14:33 | henry | Priority | normal => none |
2020-11-27 14:33 | henry | Severity | major => feature |
2020-11-27 14:33 | henry | Category | Bug => Feature |
2020-11-27 14:42 | thw1021 | Note Added: 0011778 | |
2020-11-27 14:47 | henry | Note Added: 0011779 | |
2020-11-27 14:55 | thw1021 | Note Added: 0011780 | |
2020-11-27 15:03 | henry | Assigned To | => henry |
2020-11-27 15:03 | henry | Status | new => closed |
2020-11-27 15:03 | henry | Resolution | open => no change required |
2020-11-27 15:03 | henry | Note Added: 0011781 |