CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   Cht tutorial in 15 (

braennstroem July 15, 2008 03:28

Hi, I just downloaded 1.5;

I just downloaded 1.5; nice work and thanks a lot! Though I wonder, if there is a tutorial case for the new cht solver? I could not find.


mattijs July 15, 2008 06:57

Small overview ;-) Attached
Small overview ;-)

Attached a simple testcase. Do not run blockMesh, just chtMultiRegionFoam.

OpenFOAM discussion board administrator July 15, 2008 06:58

three solids, two air doma

three solids, two air domains with flow through the top domain. Bottom of T-shaped heater is heated.

braennstroem July 15, 2008 07:46

No problem, but it seems there
No problem, but it seems there is a new version of the forum-software... I am not able to download the simple testcase ;-)

nishant_hull July 15, 2008 08:04

Even I can not be able to down
Even I can not be able to download the software

nishant_hull July 15, 2008 08:05

sorry .. not software.. but th
sorry .. not software.. but the simple testcase directory!

mattijs July 15, 2008 14:53

Mesh was too big for the forum
Mesh was too big for the forum to upload. I put together a simple blockMesh case. Check the Allrun script on how to run it. Haven't checked the output so please let me know if there are any problems with it. multiRegionHeater.tgz

braennstroem July 16, 2008 07:49

Hi Mattijs, thanks, it's ru
Hi Mattijs,

thanks, it's running, when typing the Allrun script line by line. I couldn't check the output yet, due to the qt-paraview depence (still have the old qt); using 'foamToVTK' with cellSet I was not successful.. I.e. there is just a mesh in the created vtk files. There is probably some other option to get the results!?


jwp July 16, 2008 08:24

Hi! I'm not sure, whether I'm
I'm not sure, whether I'm right here.
I want to use a solver with conjugate heat transfer for an incompressible fluid; but the cht-solver is for compressible flow, isn't it?

I've worked with the conjugateHeatFoam Solver in 1.4.1-dev; but I can't find it in 1.5

Any hints?
Thank you.


alberto July 16, 2008 11:37

Dear Jens, the content of t
Dear Jens,

the content of the -dev tree is not included in the OpenCFD released version.


andrea_barbera July 24, 2008 03:21

Dear Fabian, I had your sam
Dear Fabian,

I had your same problem in using standard parafoam utility, but I have found a way to use VTK: in first time step the mesh of every region is stored. If you copy all that meshes in


and then run many times the assignment

foamToVTK -region [RegionName]

(one time for each region) you should be able to open five different meshes in any version of paraview. Probably better ways exist but it works



braennstroem July 28, 2008 01:35

Hi Andrea, thanks for your
Hi Andrea,

thanks for your help; it works.


shu August 25, 2008 05:44

Hi, I am trying to run the

I am trying to run the chtRegionHeater with the case above according the steps listed in Allrun. Unfortunately, at the step "splitMeshRegions -cellZones" I get a long error message as below. It happens under Suse10.3 and Ubuntun 8.04 too. Could anyone please take a look? I appreciate for any hint.


Bitan Shu

Region 0
Creating mesh for region 0 bottomAir
Testing:"/home/shu/OpenFOAM/shu-1.5/run/projects/multiRegionHeater/system/bottom Air/fvSchemes"
Mapping fields
Mapping field T
#0 Foam::error::printStack(Foam: in "/home/shu/OpenFOAM/OpenFOAM-1.5/lib/linuxGccDPOpt/"
#1 Foam::sigFpe::sigFpeHandler(int) in "/home/shu/OpenFOAM/OpenFOAM-1.5/lib/linuxGccDPOpt/"
#2 Uninterpreted: [0xffffe420]
#3 Foam::calculatedFvPatchField<double>::clone(Foam:: DimensionedField<double,> const&) const in "/home/shu/OpenFOAM/OpenFOAM-1.5/lib/linuxGccDPOpt/"
#4 Foam::GeometricField<double,>::GeometricBoundaryFi eld::GeometricBoundaryField(Fo am::fvBoundaryMesh const&, Foam::DimensionedField<double,> const&, Foam::PtrList<foam::fvpatchfield<double> > const&) in "/home/shu/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/splitMeshRegions "
#5 Foam::GeometricField<double,>::GeometricField(Foam ::IOobject const&, Foam::fvMesh const&, Foam::dimensionSet const&, Foam::Field<double> const&, Foam::PtrList<foam::fvpatchfield<double> > const&) in "/home/shu/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/splitMeshRegions "
#6 Foam::tmp<foam::geometricfield<double,> > Foam::fvMeshSubset::interpolate<double>(Foam::Geom etricField<double,> const&, Foam::fvMesh const&, Foam::List<int> const&, Foam::List<int> const&, Foam::List<int> const&) in "/home/shu/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/splitMeshRegions "
#7 void subsetVolFields<foam::geometricfield<double,> >(Foam::fvMesh const&, Foam::fvMesh const&, Foam::List<int> const&, Foam::List<int> const&) in "/home/shu/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/splitMeshRegions "
#8 createAndWriteRegion(Foam::fvMesh const&, Foam::regionSplit const&, Foam::List<foam::word> const&, Foam::EdgeMap<int> const&, int) in "/home/shu/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/splitMeshRegions "
#9 main in "/home/shu/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/splitMeshRegions "
#10 __libc_start_main in "/lib/tls/"
#11 Foam::regIOobject::readIfModified() in "/home/shu/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/splitMeshRegions "

mabinty October 6, 2008 06:35

dear all!! i run the multiR
dear all!!

i run the multiRegionHeater tutorial case; the calculation finishes successfully. the problem i face now is the visualization. when opening the case with paraFoam i can import all the regions/meshes (left/rightsSolid, heater, top/bottomAir) stored in the VTK directory, but cannot display any quantity. No physical quantity can be chosen on "Vol Field Status" of multiRegionHeater.OpenFOAM. the property menu of the different regions is empty and no quantity can be chosen from the "color by" tab.

does anybody have an idea? appreciate any comment!

thx in advance!

sylvester October 6, 2008 07:54

Hi all, When running the tu
Hi all,

When running the tutorial provided with chtMultiregionFoam, the temperature is not affected by the velocity. I increased the inlet velocity to 0.1, but the temperature field does not change. Although the velocity is calculated correctly, the temperature seems to be affected by conduction only, as shown in the picture.

Any help is appreciated.


mabinty October 8, 2008 03:18

nobody an idea? aram
nobody an idea?


mabinty October 16, 2008 10:12

hi!! tried the method sugge

tried the method suggested by Andrea and it works well! its a bit complicated though, so does somebody know another way?

thx to all!

mabinty October 21, 2008 09:39

hi!! i m about to study and

i m about to study and modify the multiRegionHeater tutorial case. there i get some results which seem not to be realistic; e.g. is the range of density roh in the fluid region (air) calculated with [9.30455, 13.0787] kg/m^3.

@Sylvester: have you made further experience with the chtMultiRegionFoam solver?

appreciate any comment!! thanks in advance!

mabinty October 23, 2008 10:19

hi all!! dissolved my confu
hi all!!

dissolved my confusion concerning the density (one should have a better look on the BCs!!); it was calculated correctly as the pressure p (specified in system/topAir/changeDictionaryDict) was set to 1000000 Pa (= 10 bar, instead of the expected 1 bar).

Furthermore did I have a look on the BCs coupling the T-field of fluid and solid regions. I would like to pose a question: why is the flux not divided by the patch surface for the calculation of the gradient (see solidWallHeatFluxTemperatureCoupledFvPatchScalarFi eld.C), as the flux, calculated in solidWallTemperatureCoupledFvPatchScalarField.C is determined in the units of Watt?

_solidWallHeatFluxTemperatureCoupledFvPatchScalarF ield.C:

gradient() = refCast<const>(neighbourField).flux()/K;


Foam::solidWallTemperatureCoupledFvPatchScalarFiel d::flux() const
const fvPatchScalarField& Kw =

const fvPatchScalarField& Tw = *this;

return Tw.snGrad()*patch().magSf()*Kw;

thx in advance for any comments!!

woody November 7, 2008 05:37

Dear All, I have a problem,
Dear All,

I have a problem, impressing a finer meshRegion on top of the left&right Solid in the multiRegionHeater. I already managed to refine the hole topAir region by applying refineMesh on the cellSet.
My next idea was to create a layer cellSet fineMesh:
cellSet topAir new boxToCell (-100 0.02 -100 )(100 100 100)
cellSet fineMesh new boxToCell (-100 0.01 -100 )(100 0.02 100)
cellSet fineMesh delete cellToCell topAir
), refine this area and restart chtMultiregionFoam.
But there are further Problems coming up with this Idea:
1. spliting the fluidregion creates a fluid-fluid Interface --> what kind of BC do i have to set?
2. stitchMesh does not work:

stitchMesh fineMesh_to_topAir topAir_to_fineMes

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create mesh for time = 0.001

Coupling patches fineMesh_to_topAir and topAir_to_fineMesh
Resulting (internal) faces will be in faceZone fineMesh_to_topAirtopAir_to_fineMeshCutFaceZone

Note: the overall area covered by both patches should be identical ("integral" interface).
If this is not the case use the -partial option

Cannot find patch fineMesh_to_topAir
It should be present and of non-zero size
Valid patches are
3. using refineMesh in combination with subsetMesh makes the other cellZones inaccessible for setsToZones and splitMeshRegions

Did I miss a command or how can I remerge the two regions topAir and fineMesh?

Any Ideas?

THX Tobias

All times are GMT -4. The time now is 01:00.