Cosmos Designstar 2006 - Contacts

[sbowman]

I have cosmos designstar 2006. I am trying to use the contacts to analize stress on an assembly of parts. I believe I need to set the global contact as "Free". Then create surface to surface contact sets between components. This just fails.

I thought maybe my parts are to complex, let me try simple parts. so I made two blocks, one goes in the other. Think of a tongue and groove connection. I set up the same contacts. Creating surface to surface contact sets on the top of the groove in block1 and top surface on block2. The same was done on the bottom of the groove to the bottom surface of block2.

I restrain block 1 on the side opposite of the groove, allowing the whole block to flex under stress. The force is applied to a load pad on the top of block2. These two simple blocks also fail.

Does this software package do this analysis? Do I need an upgrade? Or am I doing something obviously wrong? Thanks for any help.

 

[gfbotha]

I mostly still use DS4.5 because of certain limitations/issues with DS2006 as far as my typical requirements are concerned. However, I belief DS2006 should be able to do your contact problem. Did you have a look/a go at the online tutorials under Help?

I do not clearly understand your test model, restraints and loading...
Is the t-g-joint loaded in bending or simply pressed (deeper) into each other?
How does the analysis fail? With any (solver) error message?
What type of materials are involved?
Are the surface pairs touching each other initially on your CAD model? (If so, edit geometry to introduce a small initial gap.)

Regards

 

[sbowman]

The surface pairs were touching, so I added a little gap, unfortunatly the same results. Let me see if I can describe my loads and restraints a little better.

Block one is 2"x2"x2". In one side it has a .550" high by 1.75" wide x .75" deep slot running horizontal-parallel to the floor. Block two slides into this slot and hangs over by about 2". At the end of block two I have 1 load on a load pad pushing towards the floor, so the block should be bending. All the restraints are on block one. The global contact is set as free. Then all matching surfaces have a surface to surface - no penetration contact.

The materials are 380 aluminum for both parts. The error message reads PCG Iterative solver stopped status code 9 matrix_singular_or_indefinate no results saved.

 

[gfbotha]

So, nothing prevents block 2 from sliding sideways out of the slot? Add a restraint to block 2 that makes displacement in the lateral direction zero (you could anyway have this when considering symmetry).

I am assuming you did already fix block 1 sufficiently. First keep the small gaps you introduced...

Further/alternatively:
You could first try solving the problem using linear ("draft" quality) elements.
You could try and see that your contact pair definitions are consistent (say, always first pick the surface belonging to block 2, then block 1). Or, change the definitions around.

Hope this works.

 

[badboggs]

Before you even go that far just do two blocks face to face. Make sure they are propery restrainted and that the contact faces to not match line to line. Make on face larger than the other. Try that and see if you still have failure.