Shell to Solid Coupling-Overconstraint-unconverged residual Forces in connecting nodes

[NRP99]

I am facing problem of non-convergence in my model particularly in shell to solid coupling nodes. The connection is of I beam(no of) cross section(s) welded to flat plate((s)(above and below I section as column)). I beam is shell and plate is solid. Since there are lot of other components and hence contacts on both sides of the flat plate I have modeled it as solid. Simple Representation of model is as shown in the image.

First I kept the solid and shell region same in a part so that i do not need to create coupling separately. But then I got the over constraint/pivot warnings and analysis stopped after 5 cutbacks in first increment. Some lines of message file-

OVERCONSTRAINT CHECKS:
An overconstraint was detected at one of the Lagrange multipliers associated
with element 119084 INSTANCE REACTION_PLATE_R04-1. There are multiple constraints applied directly or chained
constraints that are applied indirectly to this element.

***WARNING: IDCOUP3U ELEMENT 119176 INSTANCE REACTION_PLATE_R04-1 IS
OVERCONSTRAINED

***WARNING: IDCOUP3U ELEMENT 51707 INSTANCE LOWERFRAME-1 IS OVERCONSTRAINED


Since I have tight deadline I separated both shell region and solid region as different parts and given the shell to solid coupling. Here all warnings gone but then the convergence has become very slow and then in some increments the nodes at corner of coupling has residual force which is greater than time avg force and is not reducing. The SDI's are converged but only the nodes at coupling are creating problem in equilibrium iterations and hence solution is not converging after trying some iterations. Any idea on the possible reasons?

 

[FEA way]

How is each shell beam connected with plate - through edge or surface of the shell ? Because if it's shell edge to solid surface connection then shell to solid coupling should work fine. But if it's shell surface to solid surface connection (two parallel surfaces) then there are other options to use instead such as tie constraint.

Try enabling automatic stabilization in the step settings. Also you can run modal analysis to see if the connection works properly.

 

[NRP99]

How is each shell beam connected with plate - through edge or surface of the shell ?

Through edge.

Try enabling automatic stabilization in the step settings.

No Change. Modal analysis shows connections/contact are intact.

Regarding internal shell to solid coupling in the part level, why is the warning of internal coupling elements- IDCOUP3U shown during solving?

By the way, is there any troubleshooting guide/manual on the Warnings/Errors of Abaqus? 6 month of using Abaqus and I am using google as my only resource for the errors and warning explanation and rectification. It is expected that Abqaqus should have provided (at least) pointers to the Warnings and Errors in the their own manual.

 

[FEA way]

You can try different ways of connecting shell to solid. For example using tie constraint.

Indeed errors are rarely discussed in the documentation. You can find the description of some of them in the Knowledge Base but usually external resources are necessary.