HOW TO DEFINE SIFS & TEE IN CAESAR II FOR BIG DIA

Hi experts,

I need your help. I have a pipe of 1000mm diameter & branch having 800mm diameter. Please guide me in CAESAR II SIF & tees where I need define in the branch. small pipes always we will consider for center of pipe & branch but this is big diameter.

in the attached fig whether I have to define SIF & tees for Node 10 or Node 20. or I need to define for both nodes.

please reply experts

Dear Nag,
what code are you using? ASME B31.3?
First of all check if you are inside the code scope because CAESAR II applies the code rules; if you are outside you could perform some FEA to calculate sifs and stiffnesses.
For large piping, sometime we use rigid element between the centerline and pipe walls. There are pros and cons of this solution. If you search in this forum, there are some threads about this topic.
Finally I strongly advise you to give your model to a senior stress engineer for his evaluations.

Best regards
Corrado Vannella

Hi Vannella,

Thanks for your reply,

Yes I am using ASME B31.3 code & D/t<100 only.

only my concern is whether defining SIF at the intersection point is correct or at the surface of the header pipe.

Even I checked in the forum & not found any topic related to this.

Having the intersection apply at the center of the main line is always an approximation of what happens with the intersection at the wall.

Do you have (approximately) lots of pipe? Then you have a good approximation.
Do you have (approximately) size on size? Then you have a good approximation.

Do you have short runs of pipe and large disparity in sizes? Then it's not as good of an approximation.

I recommend you click on the "offsets" check button and hit F1 and read up on it.

SIFs at the intersection with offsets is equivalent to SIFs at the wall.

Hi Michael,

Thanks for your reply.

Keep in mind though that if you use offset method, the expansion of the branch will be calculated based on the length of the branch minus the offset (1/2 of header diameter). So the radial expansion of the header is not taken into account. It may or may not matter much in your calculation. When in doubt, I'd suggest that you model a rigid element from the header center to the branch connection.

In thread http://forums.coade.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=68106 , Mr. Ay posted some instructions with this subject.
Also, I attached a pdf from PRG in that thread. The title is "Using SIFs and Flexibilities in CAESAR II".