contact of elements not initially touching

[Mech151]

I have a simple part I am trying to analyze. The problem would be similar to a u shaped part with opossing forces at the legs (final deformed part would look more like an o). I would prefer to use workbench for this analysis, but could switch to Classic Ansys if it is required. The 'legs' are not in initial contact. What contact settings can I set that will allow the gap to close before a reaction occurs at the interface. The actual geometry is a single solid body. When I try compression only contact elements the gap between the legs will not close.
Thanks

 

[FArias]

What contact settings can I set that will allow the gap to close before a reaction occurs at the interface

You mean you want the gap to be closed even though the parts still don´t touch each other? Then use the "Adjust to touch" option in Interface treatment. If you want them always to stay in touch, use the type Bonded. (Both of them for Ansys WB)

Regards
Fernando

 

[Mech151]

Maybe my question was vague, but I did figure it all out, so here is what I came up with.

This is similar to a press fit dowel scenario.

If bonded contact is used, the initial gap is ignored, and a press fit dowel remains unstressed.

If augmented or normal lagrange formulation of frictionless contact is used, the surfaces will not close and the parts remain unstreassed (i.e., initial penetration or gap is ignored; regardless of add offset or adjust to touch setting, acts like a Constraint Equation would).

If frictionless (or rough) contact with pure penalty formulation is used, any initial penetration is closed resulting in stresses in the adjacent parts. Any gaps will close only if external forces cause them to, but no reaction will occur on the contact surface until the geometry actually comes into contact. The point is initial gaps are not 'forced' to close, and that reactions on the contact area only occur if penetration is calculated.

Just to reiterate the answer, for a press fit dowel or initially gap that may close on load, use frictionless (or rough) contact, with pure penalty formulation.

 

[alphaden]

Hello Mech151,

I'm doing kinda similar thing and my interference or initial penetration was completely ignored by the software. I was analysing 0.98 ID bore & 1.00 OD pin assembly, and it gave me zero stresses after such a long time of calculation. I'm using Ansys 9.0 Multiphysics. Is it ok ... or i need to go with the Workbench ??

Thanks

 

[Mech151]

Ansys is fine, just need to work out the contact options. The workbench equivalent is 'use frictionless (or rough) contact, with pure penalty formulation'. Not sure off of the top of my head what the settings are in Ansys, but you will need to set up contact elements on the adjacent surfaces.

 

[alphaden]

I tried ... nothing happened. I had 8 node 185 brick elements and used contact wizard to set up contact. included initial penetration, checked penalty approach. friction 0.2, penalty stiffness 0.2 (factor). but still it ignored the penetration. is there anything other than that ???

Thanks