[RESOLVED] Exception en virgule flottante
Hello everybody,
I'm proud to post my first comment in this forum :). I don't speak English so good so I apologize for every English mistakes :D. And, to show how much I am motivated, I will put pictures in my post : I set a good example, isn't it :D ? Ok, I'm a false beginner in OpenFOAM that I study for my Final Project and it happens something that I don't understand and that I very would like to solve. My problem is that when I make a coarse blockMesh geometry for a own icoFoam case, it works, but when I use a finer mesh, it makes an "Exception en virgule flottante" during calculation after very few first iterations. Here is my coarse case (a 2D case with 2 "empty" boundary conditions and only one cell in the z direction) : http://www.cfd-online.com/Forums/%3C...coarseMesh.jpghttp://img4.hostingpics.net/pics/860420coarseMesh.jpg with, in my transportProperties file : Code:
transportModel Newtonian; Code:
application icoFoam; and the results : http://img4.hostingpics.net/pics/255946firstResult.jpg Pretty good, isn't it ? Then, I tried to use a finer mesh by only modifying the simpleGrading values in the blockMeshDict: http://img4.hostingpics.net/pics/634018fineMesh.jpg but this happens when I launch icoFoam after deleting the old time directories (0 0.1 0.2 etc.) : Code:
Time = 0.011 Code:
deltaT 0.0000001; My extraordinary happiness when I computed my first case totally collapsed when I realized I wasn't able to redo it with fine mesh ! Does somebody have an idea ? ;) |
hello,
Check your Co number:1e34 max ! My guess i a bad boundary condition: give boundary + fvSchemes for more info. regards, olivier |
Hello,
Yes, the curent number is awesome ! But I think it is due to a solver problem because my theoretical curent number should be very low, as the max velocity is around 0.025 m/s and the size of my cells is about 0.025 m, so : C0 = dt * |U| / dx => C0 = 0.001 * 0.025 / 0.025 = 0.001 = C0 Here are my files : constant/polyMesh/blockMeshDict : Code:
/*--------------------------------*- C++ -*----------------------------------*\ Code:
/*---------------------------------------------------------------------------*\ Code:
/*---------------------------------------------------------------------------*\ Code:
/*---------------------------------------------------------------------------*\ constant/TransportProperties: Code:
/*---------------------------------------------------------------------------*\ Code:
/*---------------------------------------------------------------------------*\ Code:
/*---------------------------------------------------------------------------*\ |
Hi,
you use a transient solver with a steady state ddt scheme. Either use a different ddt scheme (e.g. Euler) or a steadyState solver (e.g. simpleFoam) Regards, Christian |
Hello,
Thank you very much, guys ! It works perfectly ! :) |
All times are GMT -4. The time now is 12:42. |