Re: [Getfem-users] Rigid body elements

[Renard Yves]

Dear Kostas,

It is of course potentially usefull to have a constraint brick which
is able to link some dofs of two different variables. This could be done
extending the existing one in order it accepts an optional second variable
or by creating a new brick.

If you intend to create a brick which automatically compute the constraint
corresponding to your modelization, it is better to create a separate brick
since a brick normally have to react when a finite element method is modified
(refinement, ...). Of course, this brick may derive from the constraint brick
 (similarly to the contacts brick which derive from the basic contact brick).

Yves.






Konstantinos Poulios <address@hidden> a écrit :

> Hi,
>
> I am trying to model a torsion moment applied onto a shaft. A common
> practice in a commercial FE- Solver would be to define a single node
> with 6 dofs lying on the axis of the shaft. Then connect this node via
> rigid body elements with the nodes lying on the cross-section of the
> shaft where the torsion moment is to be introduced. Finally, constrain
> 5 dofs of the previously defined single node and apply the torsion
> moment as a neumann condition on its 6th rotational dof.
>
> In order to transfer this technique to getfem I thought to define a
> fixed size variable of size 6 to represent the single node and add
> constraints between this variable and the shafts dofs lying on the
> considered cross-section to represent the so called "rigid body
> elements". Though, I realized that the constraint brick can be defined
> only on a single variable, so it is not possible to use it in order to
> interconnect dofs from two different variables.
>
> I wonder if it is a good idea to extend the constraint brick in a way
> that it can accept two variables or if there is a simpler technique to
> solve my problem. Does it make sense to introduce some kind of "rigid
> body elements" into getfem based on such an extended two variables
> constraint brick?
>
> Kostas
>
> _______________________________________________
> Getfem-users mailing list
> address@hidden
> https://mail.gna.org/listinfo/getfem-users