Development issueshttps://develop.openfoam.com/groups/Development/-/issues2018-03-21T15:11:12Zhttps://develop.openfoam.com/Development/openfoam/-/issues/773Typo: Smagorinsky in Extended Code Guide2018-03-21T15:11:12ZKutalmış BerçinTypo: Smagorinsky in Extended Code GuideHi,
Within the page: https://www.openfoam.com/documentation/cpp-guide/html/guide-turbulence-les-smagorinsky.html , the free term ***c*** of the quadratic function has the sub-term `(dev(D)⋅D)`.
But, the operation is a **double inner p...Hi,
Within the page: https://www.openfoam.com/documentation/cpp-guide/html/guide-turbulence-les-smagorinsky.html , the free term ***c*** of the quadratic function has the sub-term `(dev(D)⋅D)`.
But, the operation is a **double inner product** rather than an inner product.
Therefore, IMHO, this needs to be changed to `(dev(D):D)`.
Kind regardshttps://develop.openfoam.com/Development/openfoam/-/issues/713Field has no xfer()2018-03-26T20:14:18ZMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comField has no xfer()with
```
scalarField edgeWeight(...);
```
then the .xfer() returns a List<scalar> instead of a Field<scalar>
The workaround is
```
xferMoveTo<scalarField, scalarList>(edgeWeight)
```with
```
scalarField edgeWeight(...);
```
then the .xfer() returns a List<scalar> instead of a Field<scalar>
The workaround is
```
xferMoveTo<scalarField, scalarList>(edgeWeight)
```Mark OLESENMark OLESENhttps://develop.openfoam.com/Development/openfoam/-/issues/706Fix regressions introduced by #6862018-04-06T10:41:16ZMark OLESENFix regressions introduced by #686v1806Mark OLESENMark OLESENhttps://develop.openfoam.com/Development/openfoam/-/issues/800sphereSurfactantFoam and $(FOAM_USER_APPBIN )2018-04-15T20:13:00ZAdminsphereSurfactantFoam and $(FOAM_USER_APPBIN )Hi all,
Inside the Make/files of sphereSurfactantFoam the binary location seems to be set to $(FOAM_USER_APPBIN), not sure if this is really a bug
EXE = $(FOAM_USER_APPBIN)/sphereSurfactantFoam
Cheers
AndreaHi all,
Inside the Make/files of sphereSurfactantFoam the binary location seems to be set to $(FOAM_USER_APPBIN), not sure if this is really a bug
EXE = $(FOAM_USER_APPBIN)/sphereSurfactantFoam
Cheers
Andreahttps://develop.openfoam.com/Development/openfoam/-/issues/802FoamToVTK crashes on cyclic patches2018-04-18T08:11:48ZAdminFoamToVTK crashes on cyclic patchesOpenFOAM plus-20ffdb88930e crashes when I try to create VTK files using 'foamToVTK'. OpenFOAM plus-bb816234a13e runs fine on the same case.
The error is:
`From function virtual Foam::label Foam::cyclicPolyPatch::neighbPatchID() const
...OpenFOAM plus-20ffdb88930e crashes when I try to create VTK files using 'foamToVTK'. OpenFOAM plus-bb816234a13e runs fine on the same case.
The error is:
`From function virtual Foam::label Foam::cyclicPolyPatch::neighbPatchID() const
in file faMesh/faPatches/constraint/cyclic/cyclicFaPatch.C at line 286.`
I've attached my test case where I got the error.
[RotatingWall_64x16x32_WALE.zip](/uploads/c11d77860f50c026e548c03894480553/RotatingWall_64x16x32_WALE.zip)https://develop.openfoam.com/Development/openfoam/-/issues/531Feature: Possibility to suppress slaves from output2018-04-19T08:02:53ZPrashant SonakarFeature: Possibility to suppress slaves from outputWhen using large number of CPU e.g 720+, each execution of utility prints slaves.
-should this printing be optional ?When using large number of CPU e.g 720+, each execution of utility prints slaves.
-should this printing be optional ?https://develop.openfoam.com/Development/openfoam/-/issues/803Overset bugs2018-04-25T09:13:01ZAdminOverset bugsI'm using 68185c76 on Debian GNU/Linux 8 (jessie). I believe I found some problems with the overset mesh implementation.
1. dynamicOversetFvMesh.C:96
Pstream::exchange<labelList, label>(remoteFaceCells, sendCells);
Neither remoteFa...I'm using 68185c76 on Debian GNU/Linux 8 (jessie). I believe I found some problems with the overset mesh implementation.
1. dynamicOversetFvMesh.C:96
Pstream::exchange<labelList, label>(remoteFaceCells, sendCells);
Neither remoteFaceCells nor sendCells are used after this exchange. I can't see its purpose. Results are not affected when I comment it out.
2. Parallelisation problem. The overSimpleFoam tutorial (aerofoil) converges in serial, but fails to converge in parallel. May be related to above. Mind you, this may be a red herring, because I get this warning:
--> FOAM Warning :
From function bool Foam::oversetPolyPatch::master() const
in file oversetPolyPatch/oversetPolyPatch.C at line 149
The master overset patch is not the first patch. Generally the first patch should be an overset patch to guarantee consistent operation.
-Davidhttps://develop.openfoam.com/Development/openfoam/-/issues/811annularThermalMixer tutorial does not run2018-04-27T06:18:15ZMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comannularThermalMixer tutorial does not runshm does not set the cellZones -> both rotor and stator side rotate.shm does not set the cellZones -> both rotor and stator side rotate.Mattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comhttps://develop.openfoam.com/Development/openfoam/-/issues/496Lagrangian case injector bug2018-05-01T08:50:52ZAdminLagrangian case injector bugI am performing a lagrangian simulation with sprayFoam solver on OpenFOAM-v1612+. I am writting you down to report a bug that I have noticed when initializing a buzzard with coneNozzleInjection option. If the injection method is a disc a...I am performing a lagrangian simulation with sprayFoam solver on OpenFOAM-v1612+. I am writting you down to report a bug that I have noticed when initializing a buzzard with coneNozzleInjection option. If the injection method is a disc and particles size is constant, particles injection is no longer done on a surface but a line from the innerDiameter to the outerDiameter.
I have work around with it and the bug only appears when running the simulation in parallel with metis decomposition. I tried a simple and hierarchical decomposition but the case can't run with such methods. A simple solution to avoid the problem is to use a normal distribution with a small variance.https://develop.openfoam.com/Development/openfoam/-/issues/556feature: Fan BC with swirl2018-05-01T08:54:48ZPrashant Sonakarfeature: Fan BC with swirlRefer EP#467Refer EP#467v1712AdminAdminhttps://develop.openfoam.com/Development/openfoam/-/issues/426Smoother printStack Error Caused by Already Working Mesh After Another Export2018-05-01T08:59:09ZAdminSmoother printStack Error Caused by Already Working Mesh After Another ExportHello all,
I am facing a strange problem. I am using motorbike tutorial's settings. I have an pipe geometry which I have created in Pointwise a year ago and exported the mesh a year ago. The mesh is working fine. Last month I have ope...Hello all,
I am facing a strange problem. I am using motorbike tutorial's settings. I have an pipe geometry which I have created in Pointwise a year ago and exported the mesh a year ago. The mesh is working fine. Last month I have opened the same pointwise file with pipe geometry and didnt make any changes and exported the mesh. Now I am facing a printstack error with the same mesh. The error is:
```
#0 Foam::error:rintStack(Foam::Ostream&) at ??:?
#1 Foam::sigFpe::sigHandler(int) at ??:?
#2 ? in "/lib64/libc.so.6"
#3 Foam::symGaussSeidelSmoother::smooth(Foam::word const&, Foam::Field<double>&, Foam::lduMatrix const&, Foam::Field<double> const&, Foam::FieldField<Foam::Field, double> const&, Foam::UPtrList<Foam::lduInterfaceField const> const&, unsigned char, int) at ??:?
#4 Foam::symGaussSeidelSmoother::smooth(Foam::Field<d ouble>&, Foam::Field<double> const&, unsigned char, int) const at ??:?
#5 Foam::smoothSolver::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const at ??:?
#6 ? at ??:?
#7 ? at ??:?
#8 ? at ??:?
#9 ? at ??:?
#10 __libc_start_main in "/lib64/libc.so.6"
#11 ? at ??:?
Floating point exception
```
I have checked two geometry and everything is same except the faces file is written in different order in polyMesh folder but number of cells are exactly same. CheckMesh results are exactly same. I tried to create a simple cube and pipe geometry from scratch and now they create the same print stack error. I couldnt understand the reason and I am becoming insane. I will appreciate any help.
In the attachment I have put one running pipe case and one not running case with same geometry and also my simple cube case.
[problem.rar](/uploads/13553099f6ea3c1a0fce1189f008f3ee/problem.rar)https://develop.openfoam.com/Development/openfoam/-/issues/812case fails in 17122018-05-02T11:26:40ZAdmincase fails in 1712motorbike case fails with wallBoundedstreamline FO in 1712.
Reference case is attached herewith[motorBike-wallBoundedStreamLineFO.tgz](/uploads/e5ce2ed5152fe2708e970c01260ce354/motorBike-wallBoundedStreamLineFO.tgz)motorbike case fails with wallBoundedstreamline FO in 1712.
Reference case is attached herewith[motorBike-wallBoundedStreamLineFO.tgz](/uploads/e5ce2ed5152fe2708e970c01260ce354/motorBike-wallBoundedStreamLineFO.tgz)Mattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comhttps://develop.openfoam.com/Development/openfoam/-/issues/441Minor improvements to profiling2018-05-03T18:08:12ZAdminMinor improvements to profilingFirst patch has two minor improvements:
* Make it possible to switch on profiling via `site/vXXX+/controlDict` (I like to always have profiling)
* For multi-region cases add the region-name to the name of the profiling-information
[pr...First patch has two minor improvements:
* Make it possible to switch on profiling via `site/vXXX+/controlDict` (I like to always have profiling)
* For multi-region cases add the region-name to the name of the profiling-information
[profilingImprovements.patch](/uploads/d0ccbbef5254485e88c758521f550b60/profilingImprovements.patch)
The second patch adds profiling to the time loop by deleting/creating a new profiling::Trigger-object at every call to run(). That way in the profiling information it can now be seen how much time was spent inside and outside the time-loop (usually should be in the order of 99% to 1). The ProfilingTrigger-class was necessary as apparently C++ does not allow the forward-declaration of interior classes like profiling::Trigger
[profilingProfileTimeLoop.patch](/uploads/cde3f5058ae4e86fc90357f680fc22f1/profilingProfileTimeLoop.patch)
I split this into two patches because obviously fiddling with the Time-class is always controverialMark OLESENMark OLESENhttps://develop.openfoam.com/Development/openfoam/-/issues/440Profiling does not correctly store data if two instances of profiling::Inform...2018-05-03T18:08:12ZAdminProfiling does not correctly store data if two instances of profiling::Information have the same nameIn cases where two profiling::Information-instances were created with the same name but different parents (for Example: energy-equation in a chtCase) only one was created and used for accounting the wrong parent. Which leads to strange/u...In cases where two profiling::Information-instances were created with the same name but different parents (for Example: energy-equation in a chtCase) only one was created and used for accounting the wrong parent. Which leads to strange/unusable profiling results because the parent might "believe" that more time was spent in its children (leading to negative "self"-times).
This patch is a port of the way it was implemented in the original profiling-patch (but there a STL-datastructure was used to make sure that profiling::Information ls looked for by name AND parent-ID. Now it uses a Foam-container)
[profilingFixSameNameDifferentParent.patch](/uploads/48de75cccf6e59d7822924bf674fd393/profilingFixSameNameDifferentParent.patch)Mark OLESENMark OLESENhttps://develop.openfoam.com/Development/openfoam/-/issues/434Fails to build scotch with Intel MPI and Intel Compiler2018-05-03T18:08:12ZAdminFails to build scotch with Intel MPI and Intel CompilerThe Makefile in ThirdParty sets CC to "icc" for scotch but "mpicc" is still called for MPI builds. Intel MPI can use either compiler and has wrappers specifically for Intel Compiler, e.g. mpiicc. The current build script fails on "Cent...The Makefile in ThirdParty sets CC to "icc" for scotch but "mpicc" is still called for MPI builds. Intel MPI can use either compiler and has wrappers specifically for Intel Compiler, e.g. mpiicc. The current build script fails on "CentOS Linux release 7.3.1611 (Core)" with the following link errors:
...
mpicc -m64 -fPIC -O3 -DCOMMON_FILE_COMPRESS_GZ -DCOMMON_RANDOM_FIXED_SEED -DSCOTCH_RENAME -Drestrict=__restrict -DINTSIZE64 -I../../include -I../libscotch dgtst.c -o dgtst -L../../lib -lptscotch -lscotch -lptscotcherrexit -Xlinker --no-as-needed -m64 -lz -lm -lrt../../lib/libscotch.so: undefined reference to `__intel_sse2_strcpy'
../../lib/libscotch.so: undefined reference to `_intel_fast_memset'
../../lib/libscotch.so: undefined reference to `_intel_fast_memcpy'
../../lib/libscotch.so: undefined reference to `__intel_sse2_strlen'
../../lib/libscotch.so: undefined reference to `_intel_fast_memmove'
../../lib/libscotch.so: undefined reference to `__intel_sse2_strchr'
../../lib/libscotch.so: undefined reference to `__intel_sse2_strncmp'
Mark OLESENMark OLESENhttps://develop.openfoam.com/Development/openfoam/-/issues/431Use logical instead of physical path for locating WM_PROJECT_DIR2018-05-03T18:08:12ZMark OLESENUse logical instead of physical path for locating WM_PROJECT_DIRAs experienced by @ivanspisso during installation, the current use of `pwd -P` makes the installation in some cluster environments more difficult since the physical path may correspond to some arbitrary disk and not to the canonical name...As experienced by @ivanspisso during installation, the current use of `pwd -P` makes the installation in some cluster environments more difficult since the physical path may correspond to some arbitrary disk and not to the canonical name.
For example,
OK: /cineca/prod/opt/applications/
NOK: /marconi/prod/opt/applications/
NOK: /galileo/cineca/prod/opt/applications/Version v1706Mark OLESENMark OLESENhttps://develop.openfoam.com/Development/openfoam/-/issues/427topoSet: nearestToCell, nearestToPoint return nearest on each processor inste...2018-05-03T18:08:12ZMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comtopoSet: nearestToCell, nearestToPoint return nearest on each processor instead of overall nearestMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comhttps://develop.openfoam.com/Development/openfoam/-/issues/423setTimeStep additional functionality2018-05-03T18:08:12ZMattijs Janssens4-Mattijs@users.noreply.develop.openfoam.comsetTimeStep additional functionalityGoal: at certain time switch from adjustableRunTime, adjustTimeStep to full control:
- ramp time step
- suppress writing
Can this be done through setTimeStep FO?
```
E.g.
up to time 1: adjustableRunTime, write interval 4.
fr...Goal: at certain time switch from adjustableRunTime, adjustTimeStep to full control:
- ramp time step
- suppress writing
Can this be done through setTimeStep FO?
```
E.g.
up to time 1: adjustableRunTime, write interval 4.
from 1 to 2: ramp to e.g 0.1s
from 2 to ..: 0.1s
Time DeltaT
---- ------
adjustable
1 ramping to user defined
2 user defined
3 user defined
4 user defined + write as specified by writeInterval
(so only write at end of simulation)
```https://develop.openfoam.com/Development/openfoam/-/issues/422STL ascii being detected as binary2018-05-03T18:08:12ZMark OLESENSTL ascii being detected as binaryMark OLESENMark OLESENhttps://develop.openfoam.com/Development/openfoam/-/issues/421BUG merge interfaces etc in ccm conversion2018-05-03T18:08:12ZMark OLESENBUG merge interfaces etc in ccm conversion- seems not to be as robust as thought. Needs re-examination- seems not to be as robust as thought. Needs re-examinationMark OLESENMark OLESEN