[Gmsh] About second order mesh on 3D hemisphere cylinder.

Xiaodong Liu xliu29 at ncsu.edu
Thu Jul 2 17:47:43 CEST 2015


Hello,

I am trying to generate 2nd order mesh on 3D hemisphere cylinder, which is
used to calculate the turbulence flow. I found when the 1st layer is very
thin, 1e-4, the 2nd order mesh always intersect with each other near the
boundary.  I also used the high order tools to try to smooth the mesh. But
it didn't work.
The .geo file is attached as follows. Thank you for your help.
And could you recommend some papers about the high order tools, such as the
Fast curving , Elasticity.

lr = 1;
Point(1) = {0, 0, 0, lr};
Point(2) = {20, 0, 0, lr};
Point(3) = {.5, 0, 0, lr};

Line(1) = {3, 2};
 Transfinite Line{1} = 30 Using Progression 1.3;
Extrude {{0, 0, 1}, {0, 0, 0}, Pi/2} {
Line{1};Layers{16};Recombine;
}
Extrude {{0, 0, 1}, {0, 0, 0}, Pi/2} {
Line{2};Layers{16};Recombine;
}

Extrude {{0, 1, 0}, {0, 0, 0}, Pi/2} {
Surface{5};Layers{16};Recombine;
}

Extrude {{0, 1, 0}, {0, 0, 0}, -Pi/2} {
Surface{9};Layers{16};Recombine;
}

Extrude {0,-10,0} {
  Surface{18}; Layers{ 30 };Recombine;
}

Extrude {0,-10,0} {
  Surface{39}; Layers{ 30 };Recombine;
}

Best regards,
Liu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20150702/a326575f/attachment.html>