a question about: Convergence criterion is not fulfilled, distrust the solution accuracy

posted by endii on Wed, 27/07/2011 - 16:16

Hello, there
I have a question about the message I get during the analysis. hope you would have a solution for this. 
I'm running non-linear push-over analysis of 3-D structure which is composed of curved shell elements (CQ40S and CT30S) and beam elements (L13BE). when the analysis is running, the below message comes up almost at every step: 
SEVERITY : WARNING
ERROR CODE: /DIANA/SO/SO41/0048
ERRORMSG.W: Convergence criterion is not fulfilled, distrust the solution accuracy.
Seemingly, it does not affect on the result. but ss that right?
Could you also let us know what it tries to tell us and How can this issue can be sloved? 
Thank you for your help! 
 

Forums: 
DIANA

On Wed, 27/07/2011 - 21:34

What boundary constraints are you using?
Just as a thought, apply a fixed boundary constraint to one of your nodes and check if the error message returns. Your soultion may be related to this.
T

By tscullion

On Fri, 29/07/2011 - 13:54

thank you for a quick reply. yes, I think so too. 
the beam runs between two facing walls and no constraint is put at the both ends where the beam and wall are connected. 
Thus, They just share the same node at that point. That node is shared by CT30S, CQ40S (wall) and L13BE (beam) element. 
In fact, on trial i fixed the rotations of that node (translations are still allowed) and ran analysis. Then, the message didn't occur. 
So I get the impression that there is some sort of constraint issue coming from the element compatilibity. 
Is that right? 
wait on your repsonse. 
Thank you! 

By endii