Simulation failure with sandwich modele

[floch]

Hi,

I have a problem with simulation, i modelise a sandwich with a solid part for the nida and surfacic part for the skins. I precise that the part is the composite tank.
I made the kind of model because I want to differ the plies in the skins and the existing module in solid simulation can't do it.
But the analysis failure because there are interference between the surface conctact of the skin and the nida.
I tried to change the contact, solidaire, no penetration...
I don't find the solution.

Have you an idea, please?

 

[ktop]

You want to do this in Cosmos/M because it has a laminate shell element. This is not available in CosmosWorks/Simulation AFAIK.

KTOP

 

[floch]

Thanks
for accuratly, I work with solidworks 2009 simulation.
Exaltly, the model composite shell is not available in old version of solidworks.

I phoned at the hotline and they said the problem is onthe contact between the surface shell and the surface solid.
I put the contact manually (solidaire between each face in contact)
But the model don't want converge (the contact that i put didn't worked)

 

[kellnerp]

Connecting shell elements and solid elements is difficult. You are connecting an element with 6dof at each node with an element with 3dof at each node. If you don't constrain the rotational dof from the shell element then you will have an underconstrained model and therefore a singularity.

There are two ways commonly used to attach shells to solids.

1. overlap the shell and solid element where they join and merge nodes. The trick to this is that you have in effect doubled the stiffness of the structure in the overlap zone.

2. use rigid bar elements to transfer the rotational dof to translational dof. You probably can't do this in CosmosWorks/Simulation.

I don't know if you can do the first method in CosmosWorks/Simulation either.

TOP
CSWP
BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."

 

[floch]

Thanks,

I try your solutions, Kellnerp, but the overlaping and using the rigid element don't work in my model.
Solidwords get across the surfacic model through the other part. It's very strange and i don't understand.
I have prefered to re-make the simulation under ANSYS v11 and that's run correctly.

Thank you again for your help

 

[kellnerp]

As I said in the beginning. You need to use Cosmos/M, not Simulation. Simulation is limited in what it can do. Cosmos/M does not have those limitations.

TOP
CSWP
BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."