[Gmsh] surface normals

[Christophe Geuzaine]

Jack Chessa wrote:
> I am sorry to bother you I am sure you are swamped with questions about
> gmsh.  But I have on that I cannot figure out.  I am useign 1.46 on RH
> and I cannot get the surface normals to flip.  I tried changing the sign
> of the Physical Surface definition but this has no effect. Is this a bug
> oor just my mstake.   Here are two example *.geo files which produce the
> same surface normals (-z direction)
> 
> // file 1
> L = 1;
> h = L/3;
> Point(1) = {0,0,0,h};
> Point(2) = {L,0,0,h};
> Point(3) = {L,L,0,h};
> Point(4) = {0,L,0,h};
> Line(1) = {1,2};
> Line(2) = {2,3};
> Line(3) = {3,4};
> Line(4) = {4,1};
> Line Loop(5) = {4,3,2,1};
> Plane Surface(6) = {5};
> 
> Physical Surface(10) = {6};
> 
> 
> and
> // file 2
> L = 1;
> h = L/3;
> Point(1) = {0,0,0,h};
> Point(2) = {L,0,0,h};
> Point(3) = {L,L,0,h};
> Point(4) = {0,L,0,h};
> Line(1) = {1,2};
> Line(2) = {2,3};
> Line(3) = {3,4};
> Line(4) = {4,1};
> Line Loop(5) = {4,3,2,1};
> Plane Surface(6) = {5};
> 
> Physical Surface(10) = {-6};
> 
> 
> Only the last line has changed! Any Ideas?
> 

Actually, the normals are correct--have a look at the mesh files.

Gmsh processes the "Physical" commands when it writes the mesh to disk.
So, to check the orientation, just open the mesh file (File->Open/Merge
file.msh): this will display exactly what Gmsh has saved.

Cheers,

Christophe

-- 
Christophe A. Geuzaine
Applied and Computational Mathematics, Caltech
geuzaine at acm.caltech.edu - http://geuz.org