GMSH for compound
Moderator: bernd
Forum rules
and Helpful information for the FEM forum
and Helpful information for the FEM forum
GMSH for compound
Hi, I am trying to create solids with contacting faces. The meshes at the contacts need to line up but the nodes need to be duplicated. It worked Ok for the simple case of 2 stacked blocks combined in a Compound, but not for the example below. I could do with some help from the GMSH gurus.
I tried the following methods. In all cases I use GMSH option Coherence Mesh FALSE:
Type..........................................|Mesh............|Nodes
Compound without gap..........................|Mesh not aligned|Some nodes connected
Compound with gap.............................|Mesh not aligned|Nodes fully disconnected
Boolean Fragments with gap....................|Mesh not aligned|Nodes fully disconnected
Boolean Fragments without gap (mode STANDARD).|Mesh Aligned....|Nodes connected
Boolean Fragments without gap (mode SPLIT)....|Mesh Aligned....|Nodes connected
Boolean Fragments without gap (mode COMPSOLID)|Mesh Aligned....|Nodes connected
None of them give the desired result, i.e. an aligned mesh with duplicated nodes.
Is there a way to achieve this? Perhaps by specifying a physical surface at the interface between the solids of a BooleanFragments object? If so, is there a way to automatically generate physical surfaces at all faces of the solids in a BooleanFragments object?
Exploded view shows meshes are not aligned in a Compound with gap (same in BooleanFragments with gap):
I tried the following methods. In all cases I use GMSH option Coherence Mesh FALSE:
Type..........................................|Mesh............|Nodes
Compound without gap..........................|Mesh not aligned|Some nodes connected
Compound with gap.............................|Mesh not aligned|Nodes fully disconnected
Boolean Fragments with gap....................|Mesh not aligned|Nodes fully disconnected
Boolean Fragments without gap (mode STANDARD).|Mesh Aligned....|Nodes connected
Boolean Fragments without gap (mode SPLIT)....|Mesh Aligned....|Nodes connected
Boolean Fragments without gap (mode COMPSOLID)|Mesh Aligned....|Nodes connected
None of them give the desired result, i.e. an aligned mesh with duplicated nodes.
Is there a way to achieve this? Perhaps by specifying a physical surface at the interface between the solids of a BooleanFragments object? If so, is there a way to automatically generate physical surfaces at all faces of the solids in a BooleanFragments object?
Exploded view shows meshes are not aligned in a Compound with gap (same in BooleanFragments with gap):
Re: GMSH for compound
this is what you would like to have? Am I right?
Re: GMSH for compound
strange, it does not work for your geometry ...
Re: GMSH for compound
Ahh what I did. Use the exported brep, opened gmsh gui just mesh it 1D, 2D, 3D exported as unv and it was not connected.
Bernd
Bernd
Re: GMSH for compound
BTW: His has to work without a gap. A gap is not an option on a complicated geometry IMHO.
Damn suff, two years ago, I spent hours and hours to connect them and now we need tho other one ... Where is this smily which bangs his head on a wall ...
Damn suff, two years ago, I spent hours and hours to connect them and now we need tho other one ... Where is this smily which bangs his head on a wall ...
Re: GMSH for compound
Thanks! I will give it a try. If you explode the view by exaggerating deformation of the decoupled mesh do the element on the decoupled meshes line up?
Re: GMSH for compound
Hi, not sure if this can be of any help, but when I need a point to be present in the generated nodes on a surface I use the Point | Curve { expression-list } In Surface { expression }; command. Maybe in this way, once the first of the two mating surfaces is meshed, GMSH can be forced to insert new nodes on the second mating surface at the same coordinates of the first one.
But AFAIK this won't prevent GMSH to eventually generate also more other nodes, or a different connectivity between nodes, on the second surface. So the second surface may end up with more nodes than the first and most likely elements won't line up.
But AFAIK this won't prevent GMSH to eventually generate also more other nodes, or a different connectivity between nodes, on the second surface. So the second surface may end up with more nodes than the first and most likely elements won't line up.
Re: GMSH for compound
I need to know for sure that duplicate (twin) nodes and element faces line up to make the interface elements work. If this is not guaranteed then I need to duplicate nodes and reconnect volume elements in FreeCAD. I made a start with that, but it means I need to overcome a number of challenges.