I am working on making a DNS mesh for simulating the break up of a liquid sheet. Initially i had problem in generating the meshes on a 32 bit machine using blockMesh, so i moved to 64 bit. Then my blockMesh for running, but it takes unusually large time to make the mesh. But finally I had a mesh written and when i wanted to check the mesh it again takes unusually long time to run the checkMesh and it is stuck at this point
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create polyMesh for time = constant
Time = constant
internal points: 0
internal faces: 16292950
boundary patches: 6
point zones: 0
face zones: 0
cell zones: 0
Number of cells of each type:
tet wedges: 0
Boundary definition OK.
Point usage OK.
Upper triangular ordering OK.
Face vertices OK.
Number of regions: 1 (OK).
Checking patch topology for multiply connected surfaces ...
Patch Faces Points Surface topology
inlet 50 102 ok (non-closed singly connected)
outlet 4050 8102 ok (non-closed singly connected)
channelWalls 2000 4004 ok (non-closed singly connected)
walls 4000 8004 ok (non-closed singly connected)
sides 4000 8004 ok (non-closed singly connected)
I think there are large number of default cells which are set to empty and it is checking those meshes. But i want visualize and check if my mesh is ok or not.
But i also tried to visualise the mesh in ParaFoam but i could not load the file itself, paraFoam just hangs after i ask it to display the mesh.
I have approxiamtely 8 million cells in my mesh. So i want to know if anybody has faced similar kinds of problems while trying to visualize meshes with large number of cells. Is there any specific way to visualize meshes with large number of cells.
Are you sure about your machine? is the memory large enough? have you coarsen your mesh so that see whether in this way paraFoam can handle it or not, so as to exclude the problem from the man-made reasons?
I use 6G memory to handle my 2 million mesh, and it is really a hard time, but it won't stuck but did take a long time.
CheckMesh some times will report errors, but even then it is okay for my case, I can still run my case.
So my little advice is to try a coarsened one, and see what happens?
BlockMesh maximum number cells
I'm trying to generate a 100 Mln cells grid for a DNS application. I can't do that yet. Is there anybody who has made it? How to do? How to avoid the lack of memory? Can blockMesh run in parallel?
|All times are GMT -4. The time now is 12:59.|