Two layers Welding How to change the CONTACT to TIE constraint during analysis

[KaiChi]

Hi everyone,
I am working on ABAQUS\Standard based mechanical (stress) analysis for simulating two layers welding by a moving laser beam. The deformation on the two layers is caused by the external loading and thermal expansion, given that the temperature field has been obtained by another independent thermal analysis and served as input for the mechanical analysis. I want to look at the stress distribution across two layers.

Currently, I defined TIE constraint for all the welding region and ran the analysis.

However, as I can imagine, the realistic process happening on the welding region is: 1. The two layers should be in frictional contact and can be separated from each other before they are weld together; 2. after the laser beam went through, two layers will be tied together .

So the DIFFICULTY is how to simulate the changing from *contact to *tie during stress analysis!!

Thus, I look up the manual and find some potential way to solve this problem:

1. Use user-defined subrouting Multi-Point Constraint (MPC) to define the constraint relation between two layers. It should be good for mimic the *Tie constraint, but it seems very complicated for doing the frictional contact.

2. Add additional interface element (cohesive element) at the interface of two layers. I am not sure whether it will work and how to read the temperature data from thermal analysis should become tricky since I modify the CAE model by adding interface elements.

3. Use user-defined subrouting UINTER (or VUINTER for Explicit) to modify the properties of contact. But I am not sure wheter this UINTER can achieve similar performce as * Tie constraint.

I want to hear your opinion on this problem, maybe I miss something and there are many better and easier approaches to solve this problem. I really appreciate at eveyone's view. Thank you very much.

Best regards,
Kai-Chi

 

[IceBreakerSours]

For Abaqus/Standard, use the following keyword:

*SURFACE BEHAVIOR, NO SEPARATION

Go through the documentation for more on it.

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[KaiChi]

IceBreakerSours,

Thank you for your opinion. So you think the penalty contact with nonseparation can approximate the TIE constrain. I will try on my model and get back to this Thread later.

Kai-Chi

 

[Mustaine3]

Define a second contact on that region, that is inactive at the beginning. Use No Separation and Friction=Rough as behavior. Later activate that contact and deactivate the other contact.

 

[KaiChi]

To Mustaine3,

Thank you for your suggest. After two days' study, I find the approaches (MPC, cohesive, UINTER ) that I mentioned in first post are not applicable due. And the simple way to achieve what I want is use contact to approximate the TIE once two layers weld together.

Regarding to your suggest, I read the manual but still have some confusions. Can you explain a little bit more?

After preliminary modeling, I find that actually two layers are separating due to the thermal expansion instead of contacting each other within the welding region, where the laser beam hasn't arrived. And after the laser beam passes, two layer should be weld and the "TIE" constraint or "Rough Friction+Non Separation" should be applied.
1. Whether the rough friction still can be applied if the two surfaces don't necessarily contact each other.
2. You mention define two contact on that region, I guess one is TIE, the other one is rough contact, will they conflict with each other?
3. If I want the activate the Rough friction contact in terms of some criteria (such as time, temperature, user-predefined variables), is that any way to do that?

Thank you very much.

Best,
Kai-Chi