Trouble with rigid body motion in contact problem

[maheshh]

Hi

I have a pretty simple problem but am having lot of trouble modeling it in Ansys 10.0

There are two volumes in my model:
1) is a complex shaped cavity
2) is a complex shaped cam, which moves in this cavity

The side surface of the cam can hit the inner surface of the cavity and there by transmit the forces onto the cavity. Then I need to see the stresses in the cavity material.

To model this, I have fixed one face of the cavity in X,Y,Z direction. So there will be not rigid body motion in this volume.

The cam volume though is free floating in the cavity. It starts being in the center of the cavity and then moves towards one side, eventually hitting the cavity. At least tihs is what I am trying to model. But I keep getting "UY is too high - may suggest rigid body motion" error. How can I tackle this?

The way I have defined the contact/target pair is as follows:
1) select all the inside surface of teh cavity and define this as target
2) select all the outer surface of the cam and define this as contact

If I plot this pair, everything seems fine. Normals are the way I want them to be. But there is rigid body motion.

Earlier I tried to model the same problem as follows:
1) select entire cavity volume and mark it as target
2) select the entire cam volume and mark it as contact
This is very easy to define and less time consuming. But then the solution converged immediately and pos processing suggested there was not contact between cam and cavity. The came pretty much went through the cavity.

Help please!

Thanks,
Mahesh

 

[Stringmaker]

Mahesh, is there anything keeping the cam from moving in the axial direction?

 

[maheshh]

Stringmaker,

The cam has forces only in X and Y direction. So it should not come out of the cavity.

In addition, to force this condition, I have restericted the top surface of the cam in Z direction. So it will not move in the axial direction.

Thanks,
Mahesh

 

[Stringmaker]

Does the Z direction correspond to the axial direction of the cam shaft? And the forces in the X and Y directions come from some sort of lifter which rides on the cam?

Have you considered running a preliminary load step which has no loadings (just BC's) to make sure everything is properly constrained?

Without having the model or atleast an image to work with and better understand things of this nature are tough to diagnose.

-Brian

 

[maheshh]

Brian

http://farm1.static.flickr.com/145/379718692_4795b358d8_b.jpg

This figure shows the how the cam is located in the cavity. The center hatched area is the axial hole.
Fx and Fy are the appied forces on the cam.
The cam is shown in blue.
It is expected to hit the inner surface of the cavity which is shown in black.
The remaining 3D body which is the cavity is not shown. But that cavity body has a face which is completely restricted.
The +Z direction is normal to the picutre coming out towards you.
The cam top surface is restricted in Z direction.

How can this model by experiencing rigid body motion?

Thanks,
Mahesh

 

[Stringmaker]

Mahesh,
Seems like your constraints SHOULD be enought but aren't for some reason. Check your contact pairs with CNCHECK. Are they all indicating that they are closed? One thought is that Ansys might not be detecting your contact.

-Brian

 

[maheshh]

Brian,

I am importing my model from IGES format. IGES import has two options:
1) smooth
2) defeature

Last time I ran my model I had imported my model with defeature option. Do you think this has anything to do with the contact/target not behaving correctly?

Am running the same model with smooth import option right now. Will update once I see some results.

Thanks,
Mahesh

 

[Stringmaker]

Mahesh,
I cannot think of why smooth vs. defeature would change anything. I cannot see what changes there would be to the contact elements you create because of how you import geometry.

 

[cbrn]

Hi,
just a suggestion: ensure that the contacts are "closed" before launching the solution. Even if ANSYS do can treat correctly separation, the solution will not start if the initial status is in "no-contact", because there would be a natural rigid body motion and so the evaluation of initial contact stiffness would be ill. This means that sometimes it is necessary to "artificially" bring the parts into contact.
If you suspect that separation may occur during solution, then set the contact update to "at each equilibrium iteration" and not at "each substep" (or worse "never").

Regards