|
[Sponsors] |
[GAMBIT] Reversing surface normals for a face mesh? |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
March 2, 2010, 16:17 |
Reversing surface normals for a face mesh?
|
#1 |
New Member
GA
Join Date: Feb 2010
Posts: 7
Rep Power: 16 |
Hello, I am having some trouble reversing the direction of the normal to a surface face mesh.
I have a geometry volume which I am meshing all the faces with an unstructured Tri/Pave face mesh. I need the surface normals to point OUT of the volume for all of the faces. By default when I just go from face to face meshing, I get some normals that point out of the volume, and some that point into the volume. Is there a simple way to fix this and choose which direction the normal to a face mesh is? Alternatively, if there is a way to do this if I import the mesh into ANSYS, that would be fine as well. Last edited by enfinity; March 2, 2010 at 17:10. Reason: adding request for ANSYS info |
|
March 3, 2010, 02:10 |
|
#2 |
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,297
Rep Power: 41 |
Are you trying to attach a boundary layer? In this cas I could understand what you you mean with surface's normal.
Else surface normal shouldn't be a thema. If I am wrong, you may post a picture
__________________
In memory of my friend Hervé: CFD engineer & freerider |
|
March 3, 2010, 02:52 |
|
#3 | |
New Member
GA
Join Date: Feb 2010
Posts: 7
Rep Power: 16 |
Quote:
I am not super familiar with Gambit and how surfaces are treated, but i'll try to simplify my question. Let's say I have a round circular pipe. Then I mesh the FACE of the length of the pipe only, with triangular/pave elements. Thus, I will have a face mesh only of triangular elements. Now with these triangular elements, a "surface normal" as I would define it would be if I applied the simple right-hand rule of physics by following the 3 nodal points of a triangular element. So if a triangular element has nodes 1, 2, 3... right hand rule to 1,2,3 would result in a certain normal and a right hand rule to 3,2,1 would result in a normal in the opposite direction. I'm wondering, and this might not exist at all for Gambit but I'm just curious, is there a way that a mesh of surface elements is listed with its nodes in such an order that all their "normals" would point into the pipe? Or will I have to resort to post-meshing scripts? |
||
March 3, 2010, 04:26 |
|
#4 |
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,297
Rep Power: 41 |
In fluent, if you check your mesh, there is a check about left handed cells.
__________________
In memory of my friend Hervé: CFD engineer & freerider |
|
March 3, 2010, 13:49 |
|
#5 |
New Member
GA
Join Date: Feb 2010
Posts: 7
Rep Power: 16 |
ok i will try that... i actually know for sure that some of my normals are messed up when i've visualized the mesh, so i was more looking for a way to flip the normals around and make them correct rather than to "check" if they are correct, which i've already done. It sounds like the option you gave me is just a check, but i'll try it out nonetheless.
can fluent or gambit or any program do this flipping? |
|
March 4, 2010, 07:18 |
|
#6 |
Member
Join Date: Mar 2009
Posts: 32
Rep Power: 17 |
Basically GAMBIT is geometry based meshing software. In sense to say the mesh refers to the underlying geometry in other words a geometrical topology is present. The geometrical topology is responsible for quality of mesh and various other meshing scheme attributes.
Example To explain with a example you create a simple volume by standard box command it is volume so it signifies that the space confide by the surfaces is the domain by default so the surface normals will always be pointing inwards. Now in trying to answer your question. Case1 -- You create the geometry in GAMBIT. In this case the normals of the surface are already defined when a surface is created in gambit i.e. prior to meshing it. Now when you mesh it, Mesh will have a normals in the same format as of the surface normals were while geometry creation. Now there might a complicity if you had used a Boolean operation or a split operation.These operation might change the geometry definitions in background which might alter normals of the surfaces. Case2--- When you import a CAD file into GAMBIT its going to be very tricky the CAD interpreter in Gambit has procedure to convert file this engine will certainly try to make normals wherever it find a closed volume. However for stand alone surface it might be different. So looking Into case1 and case 2 mostly for case1 you might have mesh normals points into the volume, It might have issues with Boolean operations. And for case2 it would by no no situation other than the volumes detected while importing. Explained above would more perfectly explained by the ANSYS support. And next regarding you are right about the right hand rule for face normals. And if you you want to check the face normals you can certainly do it in T-grid In display mesh by just displaying the face normals. Hope it suffices-es your quest .. Uday |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[Commercial meshers] TGridFluent mesh with internal by prism layer and internal face for diagnostic | sponiar | OpenFOAM Meshing & Mesh Conversion | 2 | March 30, 2009 16:02 |
virtual geometry | richard | FLUENT | 4 | January 7, 2009 06:16 |
Recover surface from surface mesh in IcemCFD | Jerry Tanner | CFX | 0 | August 20, 2008 13:48 |
BL from surface mesh on ICEM | Davide | CFX | 0 | August 14, 2008 07:45 |
[blockMesh] Axisymmetrical mesh | Rasmus Gjesing (Gjesing) | OpenFOAM Meshing & Mesh Conversion | 10 | April 2, 2007 15:00 |