Dear all!

I see in B31.1 Table D.1 that for an intersection there are two different SIFs: one for the run pipe and another for the branch. The run pipe SIF depends only on the kind of intersection (weldolet, welding, tee, etc) but not on the size or schedule of the branch. Isn't it a bit strange? The SIF of a 18" run pipe with a weldolet is the same with a 1" branch than with a 10" branch?

Best regards,

Hi

At an intersection like stub-in, is it required to specify both SIF (inplane & outplane) 3 times, which are required to model the stub-in..
Pl. clarify.

Thanks & regards

Carletes,

Go through WRC 329 for some interesting observations pertaining to your question. Also you can go through ASME SEC III and "Schneider" option available CAESAR II Configuration set up.


Regards

Carletes

First of all, the Code tells you that if you have better SIF data than that provided in Appendix "D", you should use it. Do you have better data?

"The run pipe SIF depends only on the kind of intersection (weldolet, welding, tee, etc) but not on the size or schedule of the branch. Isn't it a bit strange?"

Well, yes it would seem strange if you do not know the background. The SIF's are based upon testing done with size-on-size branch connections - they will be "conservative" for branches less than that (conservatism is needed in this area). As Anindya suggests, there is background information published and you should read it. Start with the Markl papers. Then find articles by E. C.Rodabaugh to bring you up to date. And, as Anindya mentions read the paper by Bob Schneider regarding the importance of the branch/run ratio on determining the REAL SIF.

The 18 x 1 example you give would be trivial (assuming all welds are good) for the NPS 18 pipe and the stress in the branch would be important. Do you place a node at the OD of the pipe to locate the SIF for the branch connection?

I am glad that you are thinking about this as it is an important (and often misunderstood) aspect of pipe stress analysis.

MK, read your manuals.

Regards, John.