[Gmsh] Possible bug with surface extrusions

[Christophe Geuzaine]

> Layne Wright wrote:
> 
> Greetings All!
> 
> Got some weird results when I attempted to contruct a torus by
> rotation of a disk; it appears
> one cannot extrude a surface beyond an angle of Pi. Is this a bug, or
> a feature?

Yep, it's a feature ;-)

Indeed, you should not expect any sensible results when playing with
arcs >= Pi (the '=' is important here...)

Christophe

> See the program, 'torus.geo', listed below
> 
> Best Regards
> Layne L. Wright
> 
> 
> /*********************************************************************
> 
>  *
>  * Torus.geo Gmsh Program to Construct a Torus
>  *
>  *********************************************************************/
> 
> cm  = 1e-02;          // Define our scale
> Lc1 = 0.5 * cm;       // Mesh size
> r1  =   4 * cm;       // Origin->Disk center radius
> r2  =   1 * cm;       // Disk radius
> 
> r2c = r2 * Cos( Pi/4 );  // Cos/Sine terms to define disk
> circumference points
> r2s = r2 * Sin( Pi/4 );
> 
> deg90 = Pi / 2;          // Radian representation of 90 Degs
> 
> Point(1) = { 0.0,  0.0,   0.0,  Lc1 };   // Origin
> 
> Point(2)  = { 0.0,  r1    ,  0.0 ,  Lc1 };   // Circle center
> Point(3)  = { 0.0,  r1+r2 ,  0.0 ,  Lc1 };   // 8 points along the
> circumference
> Point(4)  = { 0.0,  r1+r2c,  r2s ,  Lc1 };
> Point(5)  = { 0.0,  r1    ,  r2  ,  Lc1 };
> Point(6)  = { 0.0,  r1-r2c,  r2s ,  Lc1 };
> Point(7)  = { 0.0,  r1-r2 ,  0.0 ,  Lc1 };
> Point(8)  = { 0.0,  r1-r2c,  -r2s,  Lc1 };
> Point(9)  = { 0.0,  r1    ,  -r2 ,  Lc1 };
> Point(10) = { 0.0,  r1+r2c,  -r2s,  Lc1 };
> 
> Circle(1) = {  3, 2,  4 } ;   // Arcsegments along the circumference
> Circle(2) = {  4, 2,  5 } ;
> Circle(3) = {  5, 2,  6 } ;
> Circle(4) = {  6, 2,  7 } ;
> Circle(5) = {  7, 2,  8 } ;
> Circle(6) = {  8, 2,  9 } ;
> Circle(7) = {  9, 2, 10 } ;
> Circle(8) = { 10, 2,  3 } ;
> 
> // Merge arcsegments into a disk
> Line Loop (9) = { 1,2,3,4,5,6,7,8 };
> Plane Surface(10) = {9};
> 
> // Extrude the disk to form a partial torus
> 
> // Things are fine as long as extrusion angle is <= Pi
> /* Extrude Surface {10, {0,0,1}, {0,0,0},  1* deg90 }; */
> 
> // Things get somewhat odd for angles > Pi
> Extrude Surface {10, {0,0,1}, {0,0,0},    3* deg90 };
> 
> // Volume assigment done interactively in 'Gmsh'
> Surface Loop(53) = {10,23,27,31,35,39,43,47,51,52};
> Volume(54) = {53};
> Physical Volume(55) = {54};
> 
> 
> 

-- 
Christophe Geuzaine

Tel: (626) 395-4552    http://www.geuz.org
Fax: (626) 578-0124    mailto:geuzaine at acm.caltech.edu