Branch Connection Stress Analysis

[marchieV]

Hello.

I just want to ask regarding the analysis for branch connection composing of flange+valve+blind (see attachment).

Is it necessary to perform the branch connection (with valve and blind) using the platform of CAESAR II (beam)?
or analyze just the local stress using FEA?

Please give me some advice on which is more accurate to consider for analyzing the branch connection.

Thank you in advance.

Sincerely,

marchie

 

[LittleInch]

what is the tee connection? A formed tee to ASME B 16.9?

Or some other concoction.

For a blinded vertical valve like that I can't see why you need to do any calculation providing you're using standard components, but Caesar should have this as a standard component.

Other than extra weight there is no real forces being imposed.

However your post and sketch have no detail so can only guess

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[DSB123]

"Please give me some advice on which is more accurate to consider for analyzing the branch connection."

I cannot understand you asking this question. Obviously the FEA approach is more accurate but as to whether you need to peform FEA on this branch is another matter.

 

[JohnBreen]

It is difficult to understand exactly what information you are requesting from the group. You seem to be considering the structural integrity of a single component that will be a part of a piping system.

To be able to determine the myriad loadings that will be applied to the individual component at issue here, it will be necessary to first perform the structural analysis of the as designed (or as currently existing) piping system (or the portion of the system that lies between two or more bounding, six degrees of fixity, anchors). Obviously, this would best be done with analysis software using beam theory solutions (e.g. Caesar II) and the model must include accurate specifications for all expected boundary intermediate (i.e. “between the anchors”) conditions (e.g. pipe supports, guides, etc.). It would be prudent to model the system with a significant number of “analysis nodes” in the immediate proximity of the branch connection at issue to assure that significant forces and moments are calculated for use as input to the following FEA model. Be sure to include nodes at the point of intersection of the branch pipe and the outlet pipe AND at the assumed weld (at the OD of the run pipe where the branch pipe intersects). It will also be necessary to assure that your beam theory model includes the appropriate Stress Intensification Factors and Flexibility Factors as provided by the Codes and Standards to which the piping system is designed. The valve in your illustration would be modeled as “a rigid beam” (“component”) and as such, the total weight of the valve, blind (and connecting flange) and threaded fasteners would have to be included in the analysis beam model. The “stub” pipe that comprises the outlet from the main run of pipe to the bottom of the valve must also be modeled with accurate dimensions (this assumes that the branch connection is fabricated and welded). You MAY have to consider static and dynamic loading sets in your beam theory analysis.

If you opt for more accuracy, or want more data at various points of interest in the described construct, your 3-D FEA model (ANSYS?) should include a significant length of run pipe on each side of the intersection (consider this when you develop your beam theory model and place node points appropriately in THAT beam model) If the branch connection is fabricated (welded) be sure to ACCURATELY model the weld in the FEA model. Obviously, be sure that the loads applied to the FEA model include the correct pressure-temperature profile (function of time).

The need for a blinded branch connection as described implies that a future piping branch is be constructed. It would be advisable to retain both the beam theory model and the FEA model as a "starting point" (there may be other modifications to the basic system) for a new set of models to be performed as the new (future) branch is designed.

 

[uaepiping]

Dear OP,
As you have 12" line, branch and 12" blinded valve, hence you have to model this along with the main 12" line. This is to take care of any weight of the valve on the main piping. If the supports on the main line are as per standards (and sufficient) nothing will happen to the tee (as said above) provided you are using standard components.
Apart from imparting its weight to the system (a load), this type of connections are harmless.

 

[ADS2K6]

If there is a restraint or physical stopper restricting the flexibility in header pipe centerline axis, then stress analysis need to be done.

 

[JGard1985]

Hi

I'm a little late to the party.

I often perform hand calculations to verify adequacy of branch lines, with 2 caveats:

1) Confirm that you are not adding substantial mass (or eccentricity) to the piping system that could have an aggregate effect on the network stresses.

2)When simplifying branches on piping runs, in addition to supports (load path), you need to be careful of vibration mode (natural frequency) of the branch. Your branch (with large weight and little support) could be more flexible than the piping system, this cause the branch to vibrate at resonance. This could ultimately lead to fatigue failures of your branch-connected piping. I would add the branch into your pipe stress analysis, unless you can document why resonance vibration is not a concern.

Jeff
Pipe Stress/Structural Engineer

http://www.xceed-eng.com