Separate step to establish contact

[g.alshamsi]

Hello all, I have a general question regarding general contact using ABAQUS/Explicit

I'm modeling a shear wall assembly made from several shell parts (S4R; beams/columns/steel sheet). I've positioned the parts in the assembly with an initial gap between each part equal to half the shell thickness. After submitting the job, the sta and msg files reported the following errors:

218 node-face overclosure(s) left unresolved and will be stored as offsets.
To help identify the nodes involved in the overclosures a node set
InfoNodeUnresolvInitOver has been created. Check the message file for more
detailed information.

Abaqus/Explicit will attempt to resolve 442 initial node-face overclosure(s).

Abaqus/Explicit will attempt to resolve 12 initial edge-edge overclosure(s).

From my understanding, the general contact algorithm solves initial overclosures using strain-free adjustments. However I'm not confident about the quality of the simulation with these errors. From the literature and the manual, the general recommendation is to define a step after the initial step to establish contact. I'm not quite sure how to achieve that. Do I position parts with a significant gap between them then define a displacement BC to ensure that contact is established (by pushing them towards each other)? Keep in mind that contact is achieved mainly through connector elements (to simulate screws).
Any help would be be appreciated, I've been stuck on this problem for sometime and I'm running out of ideas. If anymore info is needed I'll gladly provide it.

Thanks!

 

[FEA way]

These are just warnings from the solver but it's good to verify them. Check the node set created by Abaqus. Contact modeling is much easier in Abaqus/Explicit since you don't have to worry about rigid body motions due to gaps in contact (thus there's no need to reposition parts like you would do in Standard). When it comes to overclosures, most of them should be resolved by Abaqus automatically but if some of them aren't then you can adjust the tolerance of strain-free adjustment. Check the "Controlling initial contact status for general contact in Abaqus/Explicit" documentation chapter for more details.

 

[oldNail]

Is it possible you have miscalculated your initial gap?

I assume you meant to say that your nodes are on the mid plane of each shell and that the distance between nodes is half of the shell thickness. However, the distance between nodes should be half of the shell thickness1 + half of shell thickness2 in order to avoid over penetration.

 

[g.alshamsi]

Thanks for the reply, however I did put each shell thickness into consideration when modeling the initial gap (should have worded it better).

For anyone who is wondering, I was able to solve the problem by using the *Contact clearance command using the keyword editor.

Cheers,