[snappyHexMesh] Incomplete geometry in snappyHexMesh

anon_a · March 1, 2012, 05:59

Hi there!

I have been trying to use sHM in my geometry for the past couple of days.
The geometry can be considered (in a simplified way) as two large vessels connected through a tiny orifice.

The problem is that the orifice is so small that the geometry is considered "disconnected", so I end up with a mesh for the volume containing the "locationInMesh" point only. I have tried refining the features to level 6 and I also added refinement regions (levels 4-6) around the orifice but the problem remains.

I think that the point where this is done in sHM is the following:

Code:

Removing mesh beyond surface intersections
------------------------------------------

Found point (1.015 0.201 0.001) in cell 50762 in global region 1 out of 4 regions.
Keeping all cells in region 1 containing point (1.015 0.201 0.001)
Selected for keeping : 775563 cells.

Right before that point, there were 1270916 cells.

I would appreciate any help in this. I could try doing it in another program but I would like to have the option of sHM too.

calim_cfd · March 4, 2012, 10:43

hi

when you have mesh problems it's always nice to upload some pics..
not sure i can help you.. i'll just highlight a few points maybe they will help you

-to use shm you need first to have a consistent volume created with blockMesh and a 2d stl(or sth else) boundary. By consistent i mean you need to get the vertices and faces ordering right, else the "location point in mesh" will fail.. there's a topic regarding this issues in these forums.. you may wanna look it up. Else you can check the user's guide either at the website or within OF installation

-try loading your boundary mesh in paraview(or other software) and check for possible errors. look for free edges, they should point you the possible leakage of your mesh!

-and it doesn't matter whether your orifice is 1e-25m² small, smh you "see it" and work it accordingly

in sum: check the volume from where shm will create the wet domain for consistency/dimensions/interferences and the boundary stl mesh for leakage...

can't help you more at this point

good luck

anon_a · March 12, 2012, 05:30

Thanks for your answer and sorry for my delayed post.

I managed to include my complete geometry about a week ago using a feature edge refinement level 7 and modifying my blockMesh. My bounding box has a size of 1.4m, my orifice is around 0.0001m thick and when my blockMesh cells (110 in that direction) are refined 7 times, they reach around the appropriate size and are finally kept by sHM.

March 1, 2012, 05:59	Incomplete geometry in snappyHexMesh	#1
anon_a Disabled Join Date: Mar 2011 Posts: 174 Rep Power: 15	Hi there! I have been trying to use sHM in my geometry for the past couple of days. The geometry can be considered (in a simplified way) as two large vessels connected through a tiny orifice. The problem is that the orifice is so small that the geometry is considered "disconnected", so I end up with a mesh for the volume containing the "locationInMesh" point only. I have tried refining the features to level 6 and I also added refinement regions (levels 4-6) around the orifice but the problem remains. I think that the point where this is done in sHM is the following: Code: Removing mesh beyond surface intersections ------------------------------------------ Found point (1.015 0.201 0.001) in cell 50762 in global region 1 out of 4 regions. Keeping all cells in region 1 containing point (1.015 0.201 0.001) Selected for keeping : 775563 cells. Right before that point, there were 1270916 cells. I would appreciate any help in this. I could try doing it in another program but I would like to have the option of sHM too.

March 4, 2012, 10:43		#2
calim_cfd Senior Member mauricio Join Date: Jun 2011 Posts: 172 Rep Power: 18	hi when you have mesh problems it's always nice to upload some pics.. not sure i can help you.. i'll just highlight a few points maybe they will help you -to use shm you need first to have a consistent volume created with blockMesh and a 2d stl(or sth else) boundary. By consistent i mean you need to get the vertices and faces ordering right, else the "location point in mesh" will fail.. there's a topic regarding this issues in these forums.. you may wanna look it up. Else you can check the user's guide either at the website or within OF installation -try loading your boundary mesh in paraview(or other software) and check for possible errors. look for free edges, they should point you the possible leakage of your mesh! -and it doesn't matter whether your orifice is 1e-25m² small, smh you "see it" and work it accordingly in sum: check the volume from where shm will create the wet domain for consistency/dimensions/interferences and the boundary stl mesh for leakage... can't help you more at this point good luck __________________ Best Regards /calim "Elune will grant us the strength" Last edited by calim_cfd; March 4, 2012 at 10:45. Reason: spelling

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[snappyHexMesh] Problem and doubts with blockMesh, snappyHexMesh and multiple geometry	luca1992	OpenFOAM Meshing & Mesh Conversion	0	August 23, 2017 12:40
[ANSYS Meshing] CATIA Geometry Generation and Meshing	danielsullivan	ANSYS Meshing & Geometry	0	April 3, 2017 22:11
[snappyHexMesh] snappyHexMesh with zero Thickness surface imported from a Geometry	hfs	OpenFOAM Meshing & Mesh Conversion	0	September 4, 2012 12:41
[snappyHexMesh] Parts of Geometry not meshed by snappyHexMesh	Turbulence	OpenFOAM Meshing & Mesh Conversion	5	August 8, 2012 17:27
Problem Importing Geometry ProE to CFX	fatb0y	CFX	3	January 14, 2012 20:42

March 12, 2012, 05:30		#3
anon_a Disabled Join Date: Mar 2011 Posts: 174 Rep Power: 15	Thanks for your answer and sorry for my delayed post. I managed to include my complete geometry about a week ago using a feature edge refinement level 7 and modifying my blockMesh. My bounding box has a size of 1.4m, my orifice is around 0.0001m thick and when my blockMesh cells (110 in that direction) are refined 7 times, they reach around the appropriate size and are finally kept by sHM.