Problem with using icoFoam for flow in a straight tube

edificelp · February 10, 2011, 06:01

Hi:

I am trying to use OpenFoam to simulate incompressible, laminar flow of Newtonian fluids in a straight tube. And the 3-d mesh is converted from .msh file of Gmsh. After setting up boundary condition and running the simulation, the courant number mean increases so fast and becomes nearly infinite. Then the simulation stops with the information below:

Code:

Courant Number mean: 4.47866e+122 max: 1.17861e+124
#0  Foam::error::printStack(Foam::Ostream&) in "/opt/openfoam171/lib/linux64GccDPOpt/libOpenFOAM.so"
#1  Foam::sigFpe::sigFpeHandler(int) in "/opt/openfoam171/lib/linux64GccDPOpt/libOpenFOAM.so"
#2   in "/lib/libc.so.6"
#3  Foam::PBiCG::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const in "/opt/openfoam171/lib/linux64GccDPOpt/libOpenFOAM.so"
#4  
 in "/opt/openfoam171/applications/bin/linux64GccDPOpt/icoFoam"
#5  
 in "/opt/openfoam171/applications/bin/linux64GccDPOpt/icoFoam"
#6  __libc_start_main in "/lib/libc.so.6"
#7  
 in "/opt/openfoam171/applications/bin/linux64GccDPOpt/icoFoam"
[1]+  Done                    kate ~/.bashrc
Floating point exception

The file of this simulation has been attached. Could someone help me solve this problem?
Thanks a lot!

jhoepken · February 10, 2011, 09:27

A blind guess: Try to reduce your time step to suffice the CFL condition.

Jean El-Hajal · February 10, 2011, 14:07

Hi,

it seems your problem came from the fvSolution and fvSchemes file. Replace them with the original ones from the cavity tutorial.

However the mesh is very bad !

Regards,

Jean

edificelp · February 10, 2011, 17:23

Thanks a lot! After reading user guide, the incompressible steady state flow just can be simulated by simpleFoam. If icoFoam is used with steadyState, the solution will not converge.

By the way, Jean EI-Hajal, could you tell me how do you determine the mesh is bad or not? This mesh is generated by Gmsh automatically with the characteristic length. I feel headache to check this 3-d mesh.........

jhoepken · February 11, 2011, 03:27

Run checkMesh and post the output. That tool checks the quality of the mesh and prints some statistics to your console.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
analytical solution of 1D shock tube problem	ma	Main CFD Forum	10	May 11, 2015 05:20
laplacian1%7cAUp is undefined in dictionary problem with icoFoam	derath	OpenFOAM Pre-Processing	3	June 14, 2013 06:24
problem when running icoFoam on a complex shape flow field	wendywu	OpenFOAM	1	May 20, 2009 23:40
VOF problem, liquid film thickness	Pedro24	FLUENT	0	May 13, 2009 05:57
Shock tube problem	ger	Main CFD Forum	1	September 24, 2003 07:41

February 10, 2011, 09:27		#2
jhoepken Senior Member Jens Höpken Join Date: Apr 2009 Location: Duisburg, Germany Posts: 159 Rep Power: 17	A blind guess: Try to reduce your time step to suffice the CFL condition.

February 10, 2011, 14:07		#3
Jean El-Hajal New Member Jean El-Hajal Join Date: Jun 2010 Location: Ulm Posts: 16 Rep Power: 15	Hi, it seems your problem came from the fvSolution and fvSchemes file. Replace them with the original ones from the cavity tutorial. However the mesh is very bad ! Regards, Jean

February 10, 2011, 17:23		#4
edificelp New Member Join Date: Feb 2011 Posts: 2 Rep Power: 0	Thanks a lot! After reading user guide, the incompressible steady state flow just can be simulated by simpleFoam. If icoFoam is used with steadyState, the solution will not converge. By the way, Jean EI-Hajal, could you tell me how do you determine the mesh is bad or not? This mesh is generated by Gmsh automatically with the characteristic length. I feel headache to check this 3-d mesh.........

February 11, 2011, 03:27		#5
jhoepken Senior Member Jens Höpken Join Date: Apr 2009 Location: Duisburg, Germany Posts: 159 Rep Power: 17	Run checkMesh and post the output. That tool checks the quality of the mesh and prints some statistics to your console.