[Gmsh] morphing capacity in GMSH?

[Pierre JUILLARD]

Hi!

I was thinking back to morphing, and actually, I realized that there is
nearly everything already in GMSH.
The only thing I do not clearly see (because I'm actually not a GMSH expert
I guess, but also because it may not be implemented in the GUI) is how
setting up a morphing problem in GMSH.

Basically, achieving morphing needs a finite element solver which is already
implemented in GMSH.
If I guess correctly, it is like solving a linear elastic problem in a
specific region of a mesh defined by boxes.
Out of the box, all nodes are constrained.
In the box, all the elements are perfectly elastic.
When a handle is moved, it is like imposing a displacement to... mmhh...
well, something... the closest node in the box? a set of closest nodes in
the box?
And the position of all the nodes in the box is re-computed assuming linear
elasticity.
One could define an elastic modulus for the elastic material (that would
give different shapes).

Then, from the deformed shape obtained, a new mesh would have to be
redefined to comply with specific criteria, what GMSH is able to carry out,
isn't it?

Could some GMSH experts or people who have tried out some similar studies
may give their opinions?
I would be very grateful if you could give any advices?

Thanks in advance!


Pierre


2009/5/4 Pierre JUILLARD <pierre.juillard at gmail.com>

>
> Hi!
>
>
> I would like to know if there are some morphing capacities in GMSH for
> surface meshes (tria or quad)?
>
> I have understood that GMSH is able to re-mesh part of an exisiting mesh.
> This is one needed feature for the morphing.
> However, are there the needed tools to setup the morphing problem,
> i.e. possibility to create
>  - handles
>  - boxes defining areas where the mesh is impacted by handles displacement
>  - to setup the "stiffness" of the mesh inside a given box of influence?
>
> If no, do you think nonetheless that there could be a way to achieve
> morphing using GMSH with some additionnal tools that may have some features
> GMSH could be missing (maybe there could be some tools allowing to morph
> somehow a BREP geometry for instance that would then only need a
> remeshing with GMSH).
>
> I thank you in advance for your help and advices.
> Best regards,
>
>
> Pierre
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20090504/cc101f49/attachment.html>