Disappearing contacts

[346gnu]

Hello,

We have a moderately large model with many contacts.

The contacts are required to get the correct force system into some of the components.

Simpler component models just don't get close to a real answer.

We build the models with no mesh refinments and get them to run. They run ok (in a short-ish time), behave intuitively and freebody loads add up.

Then, to get better stress estimates in certain components we add mesh refinements to the component(s), wait for a free weekend and run a big model. (we don't have submodelling in Mechanica/Creo)

Components that we originally constrained by contacts and light springs seem to 'lose' the contact and pass through adjacent parts in a manner which completely defies the contact definition.

Any thoughts?

Thanks

This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.

[ChrisKaswer]

Did you make any changes other than refining the mesh? Have you divided the load into smaller steps? When you have contact, you can ask the solver to take smaller bites of of the problem by dividing the load into 4 or 5 or more steps. This approach may help it converge. You can try it out by simply reducing the total load to 1/10 or smaller and run it with the normal single load step. If it solves, then you know this approach will work.

Good Luck

[346gnu]

All we did was refine the mesh. The refinements were in areas such as rounds so that we have good info for a fatigue calculation. I will check to see if any of these refinements were on the failed contact surfaces.

We had some initial interference at a 'failed' contact. These are resolved at load step 0 which does show an initial non-zero contact area.

Perhaps things went wrong on the application of the load. I will refine the timesteps.

Thanks

[jsarkar]

if possible can u share some screenshot?

[346gnu]

A bit more (vague) info regarding possible cause of disappearing contacts.

The mating components with contacts between had been broken into volume regions (so that pressure loads could be applied to different areas, surface regions were not possible).

What was originally 1 or 2 contacts between the touching components without surface regions became many contacts.

We changed the method of loading, thus reducing the number of contacts at the offending interfaces and now these components have a working contact region between them again.

(this doesn't explain some others where volume regions surfaces did not form part of the contact)

Could there be a model accuracy consideration?