Abaqus Explicit. Ignored Boundary Condition

[uwcomposites]

Dear All,

First of all, thank you for reading this. I am new to Abaqus and especially Abaqus Explicit but have a lot of experience with FEM, implicit and explicit.
I am investigating using Abaqus explicit for the following application and I am running into a singular error.

I have two hexagonal ducts with non constant cross-section. Please refer to the attached file. One duct is pinned at the base (throught a reference point) and also constrained from X-Y translations in a region near the center (as shown in the first picture).
The second duct is not constrained at one end and is being pulled with a smooth prescribed displacement at the other end.
The ducts are specified a Surface to Surface contact with friction and linear pressure-Overclosure normal behavior with a contact stiffness of 1000 (N/mm).

No problem running this with Abaqus Implicit. I am investigating using Abaqus Explicit (I made of the necessary changes, like element types, step, etc..) but results are mind-boggling.

Abaqus Explicit results: the duct with prescribed pulling simply rolls away from the constrained ducts as shown in the attached file.

Can anybody with experience with Abaqus explicit give advices / opinions?

Thank you for your time!

-Francesco

 

[Mustaine3]

"The second duct is not constrained at one end and is being pulled with a smooth prescribed displacement at the other end."

Not constrained?

 

[uwcomposites]

Hello Mustaine3,

the duct being pulled (the front duct in the images on the attached file) is not constrained at the base (the right end side). The only BC is therefore the prescribed displacement.

This model obviously doesn't seems very useful as it. I am building it step by step, so once I get it to work with two components, I'll add more.

Would you have any idea of why when running it explicitly (rather than standard) it ignores the applied displacement and rather rolls away from the static duct? I am thinking it has something to do with the contact algorithm.

Thanks again for your time.

 

[Mustaine3]

Then remove your current contact definition and use the general contact with the default settings and the default contact behavior. Usually they work pretty good.
Your linear pressure-overclosure curve is unusual and could be the problem.

If that's not helping, zip and attach your .inp.