Showing results for 
Search instead for 
Do you mean 
Reply
Solved! Go to solution

SOL101 with contact: Convergence ratio

To all,

 

I am "playing" with sol101 and a simple contact. I an looking at the 'Contact Force Convergence Ratio'  output in the f06 file and it seems that the ratio is less the (default) tolerance (CTOL=0.01) and yet Nastran carries on iterating, see attached snaphot.

 

Has anyone any idea why nastran does that when it should stop?

 

Thanks

Regards

JXB

Production: NX9.0.3.4, NX10.0.2.6
Development: VB.NET (amateur level !)
2 REPLIES
Solution
Solution
Accepted by Ganesh_S (Community Manager)
‎08-26-2015 04:32 AM

Re: SOL101 with contact: Convergence ratio

The SOL 101 contact algorithm consists of 2 loops. The inner (contact force) loop is considered converged when the contact force convergence ratio is less than the specified CTOL. The outer (contact status) loop is considered converged when the number of status changes is less than NCHG. Both must be converged at the same time for the overall solution to be considered converged

 

For the first iteration shown in your .png image (contact iteration 4), the contact force convergence ratio is less than CTOL, but the number of status changes is not less than NCHG. In contact iteration 5, both the number of changes and the contact force have converged.

 

So, why is there a 6th contact iteration? If there is friction specified, there will always be a minimum of 6 contact iterations performed.

Re: SOL101 with contact: Convergence ratio

Thanks for the insight JimB. I have learned something today.

 

Where in the doc is that described, especially the min. 6th iteration if friction is involved? I have indded put a friction for my test as I am trying to see/test if I can only rely on friction to stop the part "flywing away" (+ the preload)  (http://community.plm.automation.siemens.com/t5/NX-Nastran-Forum/Non-linear-static-capability-preload... have always been told that one should never rely on contact/friction to prevent rigid body movement in an NL analysis but that was a few years ago. Things must have "moved" on by now

 

I'll do a test in a few minutes (maybe tonight) by changing NCHG default (0.02). I don't like changing default values in NL analysis without a clear understanding as to why I am doing it but I'd like to see if can get better "behaviour" in my test. The part is "clamped" by preload only to a rigid surface and under pure (bolt) preload the part seems to twist when one would expect no out-of-plane movement

 

Thanks

Regards

Production: NX9.0.3.4, NX10.0.2.6
Development: VB.NET (amateur level !)