OptiStruct CONTACT Setting – FREEZE + ADJUST=AUTO not closing gap

julia4567

I have a question regarding the proper CONTACT setup for modeling surfaces that are fully bonded with no initial gap.

In previous cases, I was advised to include a small clearance (e.g., 0.01 mm) between components to avoid convergence issues due to nodes being too close. However, in my current model, the physical assembly involves surfaces that are perfectly in contact, so introducing such a gap does not accurately reflect the actual condition.

Currently, I am using CONTACT with the type set to FREEZE, and the ADJUST parameter set to AUTO.
However, the 0.01 mm gap remains throughout the analysis and is not automatically closed, even with the ADJUST=AUTO setting enabled.

Could you please advise on the correct way to model this situation in OptiStruct?
I would like the surfaces to behave as if they are fully bonded or welded — with no gap and no relative motion — while also ensuring numerical stability during the simulation.

Rajashri_Saha

Hi Julia,

As you are looking for fully bonded contact, I will suggest to use TIE contact instead of FREEZE.

With this, there won't be any relative motion between the parts even if the gap exists.

Thanks

Rajashri

julia4567

Hi Rajashri,

Thank you for your suggestion.

I tried using TIE contact as you recommended, but unfortunately the analysis is not converging when I apply it.
Would you happen to have any alternative approaches or tips to help with convergence in this case?

Any additional guidance would be greatly appreciated.

Best regards, Julia

Rajashri_Saha

please share the .out and nl.out file of the run.

You can refer the below link for troubleshooting of NL convergence issue.T

https://community.altair.com/community?id=kb_article&sys_id=a6b1eafedbfd4d10cfd5f6a4e2961906
https://community.altair.com/community?id=kb_article&sys_id=a6b1eafedbfd4d10cfd5f6a4e2961906

Thanks

Rajashri

julia4567

250319_1channel_gap test.out

Thank you very much for your kind response.
I’ve attached the .out file for your review.
I would greatly appreciate it if you could take a look and provide any feedback or suggestions.

Rajashri_Saha

Can you decrease the DTMIN value in NLADAPT card to 1e-7 and rerun?

You can try another run with NLCTRL instead of NLPARM+NLADAPT.Below is the link for this card.

https://help.altair.com/hwsolvers/os/topics/solvers/os/nlctrl_bulk_r.htm?zoom_highlight=nlctrl

Also check the nl.out file for more information on non-convergence like element distortion.