Hello all:

I have a question that I could not find addressed in the forum. I am curious what other people are doing in industry.

How are you guys modeling tie points (new pipe tieing to an existing pipe)?

Q1) Is anyone modeling this connection as a flexibile anchor?
Q2) Is anyone modeling the entire existing pipe back to anchor points on each side?
Q3) Is anyone modeling the existing pipe to a substantial distance away?

A difficult question which will garner a wide variety of view points I am sure....

I try to go back as far as needed to define fixity in X,Y, and Z usually this means that I end up accounting for more existing pipe then new pipie in the model..... One could also do something similiar by taking the tie in point and imposing displacements on it (while modeling the actual geoemetry of the tie in so you have properly accounted for the SIF)


Make sure that you have accounted for thermal reversal if your tie in is a hot tap!

How critical job, how much time, how much money...
- I would go to next fixed points
- I would take the old lines up to points that error in connection is acceptable
- I would manually calculate connection point movements and enter them
- I would just use general pipe design rules/experience and forget CAESAR.

For me no general rule (except that I would always start from the last item). During the last 2 months I have used all options in my calcs. Each job is different.

Hello AugustaEng,

There has been some discussion of this topic ("uncoupling") and you might want to look at this previously posted question and the susequent answers in this thread:

http://www.coade.com/ubbthreads/ubbthrea...h=true#Post5713

Regards, John.

dear sir
it depend on the size of pipes

for example consider we have a 40 " existing header and new branch with size below than 1/3 of existing header size.so you can read the displacement at the node of connection in 40" header and then enter this displacement at the branch node .if the new branch is bigger than 1/3 of header size you must modeling the existing pipe back to anchor point
(and if you have semianchor point near the branch point you must use virtual support for preventing rotation.for example consider a pipe in x direction you must add rx at the semi anchor point)

If you are hot tapping and plan to analyze this in Caesar II then please read the newsletter article on this website entitled "A Technique for the Analysis of Non-Simultaneously Thermally Displaced Piping Systems (Hot Tapping)" in the October, 1996 edition.

Dear sir,

i have read the paper concerning "modeling large D/d Tees"in the COADE MECHANICAL ENGINEERING NEWS-december 1998, i am not very clear about one
things: the branch SIF should be applied at the surface of header(node 2025)-as the paper recommended in Figure4, but is it needed to apply SIF
at node 2020 at the same time,which mean to apply sif not only at node 2025 and but also at node 2020? i think it is needed ,but i am not very clear about it? can anyone give me some advice.

the second question, usually the tee wall thickness is thicker than the run pipe of header and the branch pipe, is it reasonable to specify the thickness of the tee(which are thicker than run pipe of header and branch)on the element 2025-2020,2020-2030,and 2025-2026( i add node 2026 between node 2025 and 2200,the length of element 2020-2025+element 2025-2026= Tee hight)
then add SIFs at node 2010,2020,2030,2025,2026.
the SIFS added at node 2010,2030 and 2026 are for wall thickness change.

THANK YOU FOR YOUR TIME!