Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

model works with identity pairs but not with contact pairs

Please login with a confirmed email address before reporting spam

Any tips on what I am doing wrong in the attached axisymmetric model involving a titanium part with a rounded top that supports a nylon liner that fits inside a titanium "cap"? An approximate image of the structure being modeled is attached as a .jpg. I'm idealizing the top of the rounded center portion as a semi-circle that's swept through the 360 degrees of the axisymmetric model to simulate the top of this part. I'm calling it a "ring" here. The boundary of the cross section of the ring that's not in contact with the nylon is held fixed in my model. Meanwhile, a downward force is applied to the top of the titanium cap. The sketch is not quite representing the exact geometry, but the Comsol geometry captures what I want to explore.

The model with identity pairs works OK, see first attached .mph file. However, when I put in contact pairs in the boundaries shown in these models (i.e., at the interface of the nylon and the Ti cap, and at the interface between the metal "ring" and the nylon), the model does not work; see last attached .mph file. Just FYI, this model works fine when I simply use a "union" command to construct the entire model, e.g., with all interfaces connected. But I would like to test the effects of friction at the interfaces noted, hence the attempts with assembly and contact pairs, etc.

I get various error messages depending on how I tweak the model, including errors such as "segmented solver steps do not include all components", etc. Sometimes there is a convergence error.

Any suggestions?

Thanks in advance.


3 Replies Last Post Oct 28, 2012, 5:44 p.m. EDT
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 6, 2012, 3:51 p.m. EDT
Hi

I'm not by my WS so I cannot open your models just now, but one thing:

the "segmented solver steps do not include all components", often comes from having tweaked (changed) the default solver settings, then changed someting in the model tree, and COSMOL does then not update all solver steps accordingly, because it asumes that you, the user, is taking control, since you have already changed something in the solver nodes.

If you list the dependent variables, and the segregated steps, you will see that one or a few might be missing from the list, for you then to ad them manually in the right segment step.
Another way is to generate a new default study case, and populate it by COMSOL default settings, then compare the new one with your tweaked, one and add the missing variables as the COSMOL default setting.
Or just delete the full solver sequence and hence postprocessing sequence and restart from scratch (note; from v4.3 you have the option to keep the postprocessing nodes, when you elete solver nodes, but you need to check each data-set link thereafter to adapt it to the new ones generated by any new solver sequence, a highly appreciated option, it saves hours of work on complex models)

Then for contact problems, convergence is delicate, so I can only recommend to read a few times the doc, and to use the latest versions (as in fact you are doing ;)

--
Good luck
Ivar
Hi I'm not by my WS so I cannot open your models just now, but one thing: the "segmented solver steps do not include all components", often comes from having tweaked (changed) the default solver settings, then changed someting in the model tree, and COSMOL does then not update all solver steps accordingly, because it asumes that you, the user, is taking control, since you have already changed something in the solver nodes. If you list the dependent variables, and the segregated steps, you will see that one or a few might be missing from the list, for you then to ad them manually in the right segment step. Another way is to generate a new default study case, and populate it by COMSOL default settings, then compare the new one with your tweaked, one and add the missing variables as the COSMOL default setting. Or just delete the full solver sequence and hence postprocessing sequence and restart from scratch (note; from v4.3 you have the option to keep the postprocessing nodes, when you elete solver nodes, but you need to check each data-set link thereafter to adapt it to the new ones generated by any new solver sequence, a highly appreciated option, it saves hours of work on complex models) Then for contact problems, convergence is delicate, so I can only recommend to read a few times the doc, and to use the latest versions (as in fact you are doing ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 8, 2012, 1:03 p.m. EDT
Thanks Ivar:

I'll play around with your suggestions on the solver, but see if you find any obvious glitches when you have a chance to look at the files on your WS. What "doc" were you recommending in your last comment about the contact pairs?

Best, John
Thanks Ivar: I'll play around with your suggestions on the solver, but see if you find any obvious glitches when you have a chance to look at the files on your WS. What "doc" were you recommending in your last comment about the contact pairs? Best, John

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 28, 2012, 5:44 p.m. EDT
Ivar:

Just as a quick update, all is well with my models now. Turns out things worked fine once I adjusted the initial pressure for the contact pairs. (Always something simple, right?)

Thanks.


Thanks Ivar:

I'll play around with your suggestions on the solver, but see if you find any obvious glitches when you have a chance to look at the files on your WS. What "doc" were you recommending in your last comment about the contact pairs?

Best, John


Ivar: Just as a quick update, all is well with my models now. Turns out things worked fine once I adjusted the initial pressure for the contact pairs. (Always something simple, right?) Thanks. [QUOTE] Thanks Ivar: I'll play around with your suggestions on the solver, but see if you find any obvious glitches when you have a chance to look at the files on your WS. What "doc" were you recommending in your last comment about the contact pairs? Best, John [/QUOTE]

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.