Welds fail in part for no obvious reason

[karthur1]

Attached here is a weldment that doesnt want to cooperate.  All is well with the assembly when I open it (no red cross).  When I change the length of extrusion 2 in -077-ta from 20.25in to 15in, all the welds in the weldment fails for no obvious reason.

I tried to edit the welds and fix them, but nothing that I try works.  Next thing I tried was to delete all the welds and just recreate them.  This doesnt work either.  I get the error "weld compute faild due to a modeling error".

Am I missing the obvious?

[johnsonshiue]

Kirk,

This is quite strange. I cannot seem to reproduce the problem on R2011 or R2012 by using the dataset you attached and the steps you described. Did I miss anything? Could you attach a snapshot of the failed state?

Thanks!

[karthur1]

Simply amazing... I scratched my head over this for an hour this morning.  Now, I was going to record a video for you to show you my steps and I cant get it tofail to save my life....

Thanks

[johnsonshiue]

Kirk,

I suspect maybe the solver was in a bad state. It might have something to do with prior operations causing the solver to confuse. If you can reproduce it from scratch, please let me know asap.

Thanks!