Orifice tap max loads?

[JoeSeag]

I’m curious if there’s any suggestions for coming up with maximum loads that can be applied to 1/2” nipples on the taps on an orifice flange pair?

 

[KevinNZ]

You could treat the tapping as piping and look at the pipe loads/stress with appropriate SIF for the treaded connection and reduced wall thickness.

 

[JoeSeag]

KevinNZ- thank you. That's where I started with this. I modeled them in caesar and applied a Socketweld SIF as these are welded taps, not threaded. Just was curious if there were any other obvious routes for analyzing.

 

[mk3223]

Typically for the process piping, the 1/2 inch tap piping of the orifice flange may be bridge welded with the short nipple to increase the piping stength of the joints.

 

[JoeSeag]

Thanks for the feedback mk3223- do you mind schooling me on what you mean by "bridge welded"? I am not familiar with this term.

 

[mk3223]

Below is a simple sketch of the “bridge weld” between two socket fitings:

1. fittings with socket joint
2. short nipple with a minimum distance between two socket fittings, may be 1/2“
3. ”bridge weld“ to cover the open gap between two socket fittings

 

[Krausen]

Orifice flange nipples can also be braced with gussets if loading is a concern. Typically with a 1-1/2" x 1-1/2" x 3/16"-thk angle member that is fillet welded to the orifice flange & to the 1/2" block valve connected to 1/2" nipple

 

[KernOily]

Modeling this in CAESAR would give you a defensible estimate but it will be far from an accurate answer. The load calculated by CAESAR will be the reaction needed at the nipple-to-flange node to create a stress level equal to the max allowable developed in the pipe for whatever load case you are modeling. Note this does not really account for the stress riser at the threads, which is critical, except for the reduced wall thickness. The SIF for a threaded connection is a fatigue-based derating factor that will, as I said, get you a defensible estimate. Also, if this is in vibrating service, there are additional considerations. If you need a better number than CAESAR, it should be modeled in FEA. Bottom line, best to decide how accurate of an answer you need, then proceed. If all you need is an approximate number you can defend with documentation, I'd use CAESAR, then derate that value by 20% or by some other appropriate factor. This will protect you if, for example, some dude decides to stand on the flange taps, and it snaps off. Ask me how I know about that one sometime. LOLz