How can I deal with high temperature duct system(D/t>100) in caesar II? Which one of software answer is true and which one is wrong? Is there any reference for this kind of problems?
The beam element model used by Caesar II is not appropriate to very thin shells where local flexibility domonates. Also, B31 type stress intensification factors cannot be relied upon above D/t=100.
You should consider using FE-Pipe or similar software that uses a shell element model.