[Gmsh] Gap between mesh and geometry

[wei]

I think you would better to split the domain into 3 parts. or using a higher order polynomial mesh. eg. 11th.
> On Jan 30, 2015, at 4:00 PM, Bulent Hatipoglu <hatipoglu.bulent at gmail.com> wrote:
> 
> If there is a discontinuity in one of the lines the Gmsh does not count it as a noddal point and a gap occurs between mes boundary and  geometry boundary.  Please see following figures for automatically generated and structured meshes  Is there a way to solve this issue?
> I think I can divide the geometry into separate surfaces or lines and mesh these separate surfaces/lines but this will be a workaround. I am looking for a better solution if available.
> 
> Thanks,
> 
> Bulent
> 
> 
> <image.png>
> Automatic/unstructured
> 
> <image.png>
> Structured
> 
> Codes used to generate these figures are here:
> 
> Automatic /unstructured mesh:
> Point(1) = {100, 100, 0, 1};
> Point(2) = {200, 100, 0, 1};
> Point(3) = {200, 110, 0, 1};
> Point(4) = {100, 115, 0, 1};
> Point(5) = {103.5, 104, 0, 1};
> Point(6) = {150, 115, 0, 1};
> Point(7) = {170, 110, 0, 1};
> Point(8) = {150, 100, 0, 1};
> Point(9) = {170, 100, 0, 1};
> Line(1) = {1,8,9, 2 };
> Line(2) = {2, 3};
> Line(3) = {4,6,7, 3};
> Line(4) = {4, 1};
> Line Loop(6) = {3, -2,-1,-4};
> Plane Surface(6) = {6};
> 
> 
> Structured:
> 
> Point(1) = {100, 100, 0, 1};
> Point(2) = {200, 100, 0, 1};
> Point(3) = {200, 110, 0, 1};
> Point(4) = {100, 115, 0, 1};
> Point(5) = {103.5, 104, 0, 1};
> Point(6) = {150, 115, 0, 1};
> Point(7) = {170, 110, 0, 1};
> Point(8) = {150, 100, 0, 1};
> Point(9) = {170, 100, 0, 1};
> 
> Line(1) = {1,8,9, 2 };
> Transfinite Line {1} = 12 Using Progression 1;
> Line(2) = {2, 3};
> Transfinite Line {2} = 7 Using Progression 1;
> Line(3) = {4,6,7, 3};
> Transfinite Line {3} = 12 Using Progression 1;
> Line(4) = {4, 1};
> Transfinite Line {4} = 7 Using Progression 1;
> Line Loop(6) = {3, -2, -1, -4};
> Plane Surface(6) = {6};
> Transfinite Surface {6};
> Recombine Surface {6};
> 
> 
> 
> 
> _______________________________________________
> gmsh mailing list
> gmsh at geuz.org
> http://www.geuz.org/mailman/listinfo/gmsh