Kinematic modeling using FEA 1

[csk]

Hello!

Is there anyone out these who is familiar with cam-follower mechanism. If so, did anyone try to model the motion of the cam-follower using FEA? If yes, would like to get some pointers...

The challenging part of is accurate modeling. There is a cam, a follower and a preloaded spring that keeps the cam and follower in contact and in turn achieves the desired motion tranlation.. I am trying to use connector elements in ABAQUS, but so far have not been successful..

If anyone has any experience/ideas on this please let me know,

Thanks,
csk.

 

[RiBeneke]

I have not done FEA on solid bodies so am not an expert.
However I would suggest that you may be able to reduce your problem initially to a plane strain type of problem to make the initial analysis easier.
Also I think that as you will be dealing with contact stresses that are infinite unless you have elastic deformation, you will need to do a second-order analysis.

 

[csk]

Thank you RiBeneke. It's a good suggestion to start in the 2d. I didnot even think of this, and this is what happens if you are too much caught up with the problem on hand.

-csk.

 

[brad]

Please reclarify exactly what your approach is:
It sounds like your intent wass to use connectors within ABAQUS, but now it sounds as if you are considering RiBeneke's suggestion with contact and second-order elements.

I'm quite sure that the connector approach will work; I just have not yet thought through the problem to give you the detailed approach.

I understand ReBeneke's suggestion to be utilizing contact with 2nd order elements. If my understanding is incorrect my apologies; however if this is what is being suggested, I would STRONGLY discourage this. Most 2nd order elements can be very bad in contact, resulting in grossly incorrect contact forces. Furthermore, if you are really interested in the kinematics only, this approach strikes me as computational overkill.

My impression of your initial post is that you are interested in the kinematics only. Is this indeed correct?

I do not completely know the kinematics description of a cam-follower mechanism. If you can describe it better to me, I may be able to give you some advice as to how to model this with ABAQUS connectors.

Brad

 

[csk]

Bradh,

It is correct that second order elements donot give accurate results when used with contact in ABAQUS. Right noe my problem is using modified second order tetrahedral elements (c3d10m). I chose these elements as the cam geometry is quite complicated and it would take me considerable effor to mesh it with hex elements. I thougt that I can first try out a 2d problem with connectors. You are correct in interpreting that I am interested in only the kinematics of the problem and I still want to use connector elements. But to give you some more details..

there are 3 components in the cam-follower mechanism.
1. Cam 2. Follower and 3. Pre-compressed Spring

Spring is reponsible for keeping the cam and follower in contact through out the motion. This spring is what i am trying to model with a connector. Force is applied on one of the faces of the cam, this in turn causes it to translate axially compressing the spring by a certain extent and rotates along the axis of the follower. Follower is fixed throughout the mechanism. The purpose of the follower is to guide the cam. Once he cam rotates due to the force, it would have rotated all the components that come into contact with it. The cam's inside surface resides on the follower's shaft's outside surface. And the spring is compressed along the shaft of the follower. I hope this gave you some understanding.

Thanks, -csk