Re: [Getfem-users] 2d plane strain/stress analysis

[Renard Yves]

Dear Umut,

The problem you have seems to come from the fact that you use a  
vectorial fem to describes the Lame coefficients lambda and mu. You  
have to declare a scalar mesh_fem to describe them.

Concerning your second question, the assembly of the 2d elasticity  
problem uses Lame coefficients. So you have to set the right Lame  
coefficients depending if you consider a plane stress model or a plane  
strain model from the Young modulus and the Poisson ratio.

Yves.


Umut Tabak <address@hidden> a écrit :

> Dear all,
>
> I was wondering when using the  
> asm_stiffness_matrix_for_linear_elasticity function for a 2d  
> problem(on a mesh that is imported from gmsh), I would like to  
> define a 2d vectorial problem(a plane strain or stress one  
> actually). So I set the dimension to 2 since this is a 2s mesh. But  
> when using this method in the format given in the format of the manual
>
> invalid data mesh fem (Qdim=1 required)
>
> It has been sometime I have not been using the library, this can be  
> a simple mistake. The second question is that the thickness of the  
> domain in say z direction, is that taken as unit when computing the  
> stiffness matrices? And if I would like to define a different  
> thickness, should I use generic assembly procedure for the beam  
> bending problems. I guess there I can supply this as a parameter to  
> the integration process.
>
> Simple code is attached.
>
> Best regards,
> Umut