Floating point exception with simpleFoam

andrei.cimpoeru · January 25, 2016, 09:35

Hi All,

I am trying to run simpleFoam on the highlift geometry (high lift prediction workshop 1). I've created a mesh myself in Pointwise, I set up the test case and simpleFoam gives floating point on the very first cycle. I cannot upload the grids since it is very big (35M cells). The problem is that I was able to run other solvers on the same mesh and it seems OK.
I have used fixedValues at the inflow and zeroGradient at the outflow. The problem is that simpleFoam fails when is trying to solve pressure equation. I tried to drop the order of the discretisation from 2nd to first and no chance to get it to run.
I would like to know if anyone experienced the same problem and how they fixed it.

I would appreciate your help and many thanks.

Cheers,

Andrei

Odranoel · January 26, 2016, 04:51

Hi,

could you post your system/fvSchemes and system/fvSolution? Usually the problem is caused by GAMG solver for pressure; I will suggest you to switch to PCG.
Alternatively you could launch the command 'renumberMesh -overwrite' (remeber to make a copy of your constant folder before) and keep on using GAMG.

andrei.cimpoeru · January 26, 2016, 06:50

Quote:

Originally Posted by Odranoel

Hi,

could you post your system/fvSchemes and system/fvSolution? Usually the problem is caused by GAMG solver for pressure; I will suggest you to switch to PCG.
Alternatively you could launch the command 'renumberMesh -overwrite' (remeber to make a copy of your constant folder before) and keep on using GAMG.

Thanks for the reply Odranoel. The first thing I did when the test case failed I switched to a conjugate gradient method for the pressure equation and it didn't work. Also I also run renumberMesh -overwrite and it failed with thing as well.
Please see the attached system file and I would be grateful if you find any mistakes. Much appreciated for your suggestions .

Cheers,

Andrei

Odranoel · January 26, 2016, 10:21

Hi Andrei,

it seems to be everything ok.

I would just remove the library included in controlDict (I think they are useless in this case), but I don't think it will solve the problem.
Did you try to run potentialFoam? Did checkMesh show errors?

Sorry for not being so helpful.

andrei.cimpoeru · January 26, 2016, 12:03

Quote:

Originally Posted by Odranoel

Hi Andrei,

it seems to be everything ok.

I would just remove the library included in controlDict (I think they are useless in this case), but I don't think it will solve the problem.
Did you try to run potentialFoam? Did checkMesh show errors?

Sorry for not being so helpful.

Hi Odranoel,

No worries. I tried all sorts of things. The thing is that potentialFoam is failing as well. checkMesh gave have some related to some high aspect ratio cells. The thing is that in the past I was able to run even on very bad meshes compared with what I have now. The thing is that I have slip conditions on my farfield regions. I am just thinking if this can cause problems. Have you had any problems with the slip conditions? Thanks for you help this is the error.

Create time

Create mesh for time = 0

Reading field p

Reading field U

Reading/calculating face flux field phi

Floating point exception (core dumped)

Odranoel · January 27, 2016, 09:44

Hi Andrei,

I never tried slip conditions at farfield but I don't think it's the problem because the check of U and p (before core dumped) is ok. By the way you can try to switch to zeroGradient or freestream just to be sure.
You can check the file polymesh/boundary if you have the same patch of BC; I am still thinking it's a mesh problem. Did you change the controlDict?

Let me know if something changes.

January 25, 2016, 09:35	Floating point exception with simpleFoam	#1
andrei.cimpoeru Member AndreiCFD Join Date: Nov 2012 Posts: 47 Rep Power: 14	Hi All, I am trying to run simpleFoam on the highlift geometry (high lift prediction workshop 1). I've created a mesh myself in Pointwise, I set up the test case and simpleFoam gives floating point on the very first cycle. I cannot upload the grids since it is very big (35M cells). The problem is that I was able to run other solvers on the same mesh and it seems OK. I have used fixedValues at the inflow and zeroGradient at the outflow. The problem is that simpleFoam fails when is trying to solve pressure equation. I tried to drop the order of the discretisation from 2nd to first and no chance to get it to run. I would like to know if anyone experienced the same problem and how they fixed it. I would appreciate your help and many thanks. Cheers, Andrei

January 26, 2016, 04:51	fvSchemes and fvSolution	#2
Odranoel New Member Join Date: May 2015 Location: United Kingdom Posts: 5 Rep Power: 11	Hi, could you post your system/fvSchemes and system/fvSolution? Usually the problem is caused by GAMG solver for pressure; I will suggest you to switch to PCG. Alternatively you could launch the command 'renumberMesh -overwrite' (remeber to make a copy of your constant folder before) and keep on using GAMG.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
icoFoam floating point exception (8)	leizhao512	OpenFOAM Running, Solving & CFD	7	November 1, 2018 12:43
Floating point exception (core dumped), running a new solver	Mahyar Javidi	OpenFOAM Running, Solving & CFD	6	April 7, 2018 13:43
Floating point exception (core dumped) for GAMG solver	yuhou1989	OpenFOAM Running, Solving & CFD	2	March 24, 2015 20:28
simpleFoam Floating point exception error -help	sudhasran	OpenFOAM Running, Solving & CFD	3	March 12, 2012 17:23
Pipe flow in settlingFoam floating point exception	jochemvandenbosch	OpenFOAM Running, Solving & CFD	4	February 16, 2012 04:24

January 26, 2016, 10:21		#4
Odranoel New Member Join Date: May 2015 Location: United Kingdom Posts: 5 Rep Power: 11	Hi Andrei, it seems to be everything ok. I would just remove the library included in controlDict (I think they are useless in this case), but I don't think it will solve the problem. Did you try to run potentialFoam? Did checkMesh show errors? Sorry for not being so helpful.

January 27, 2016, 09:44		#6
Odranoel New Member Join Date: May 2015 Location: United Kingdom Posts: 5 Rep Power: 11	Hi Andrei, I never tried slip conditions at farfield but I don't think it's the problem because the check of U and p (before core dumped) is ok. By the way you can try to switch to zeroGradient or freestream just to be sure. You can check the file polymesh/boundary if you have the same patch of BC; I am still thinking it's a mesh problem. Did you change the controlDict? Let me know if something changes.