[Gmsh] 2 Small characteristic length -> Bad mesh ? (parameter values)

[Christophe Geuzaine]

Taras Danilchenko wrote:
> Hello agan,
>  
>  Parameter values for AxiSym_Cyl_Jet.geo file:
> 
> H_of_pa=    1.00e+001;
> R_of_pa=    5.00e+000;
> R_c=    2.00e-002;
> h_c =    5.00e-001;
> R_I=    1.00e-001;
> Angle_I =    5.24e-001;
> Hnbpt =    1.00e+002;
> Rnbpt =    5.00e+001;
> h =    1.00e-001;
> h1 =    1.05e-003;
> 

It's a limitation in the accuracy of the old 2D isotropic algorithm. 
There are two solutions: in "Tools->Mesh->General", either decrease 
"Random perturbation factor", or set "2D algorithm" to "Triangle".

Best,

Christophe

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