It looks like the issue I raised was already discussed in 2008. See:
http://www.geuz.org/pipermail/gmsh/2008/003152.html
Regards, Peter.
Quoted text:
Hi.
Troubled by some results on a more complex geo file (I noticed that using mesh
element sizes of 2.5 and 3 generated identically sized .msh files) I did a test
of meshing on a single line with a length of 10 (ie. 1D meshing only). In both
cases (actual and test cases), I specify "desired mesh element sizes at the
geometrical points of the model" using the point command.
Here's my test geo file:
length=10.0;
disc=1.0;
Point(1)={length/2.0,0.0,0.0,disc};
Point(2)={-length/2.0,0.0,0.0,disc};
Line(1)={1,2};
I expected that using a mesh element size of 1 (specified using "disc") would
give me 10 elements along my line. It however only gives me 9 (8 nodes along
the line). The results are shown below:
disc=1 9 elements
disc=0.9 11 elements
disc=0.9999 10 elements
The same general type of effect seems to hold for other desired element sizes:
disc=2.5 yields 3 elements (rather than the expected 4 elements which can be
obtained using disc=2.4).
I couldn't find information on such effects - apologies if it has been
discussed previously though. Is this maybe a known effect of rounding,
precision or the meshing algorithm itself?
Thanks in advance.
Best regards, Peter.
_______________________________________________
gmsh mailing list
[email protected]
http://www.geuz.org/mailman/listinfo/gmsh
