Joint efficiency - UG-27 ASME BPVC

[MartinVT]

Hi everyone!

This is my first thread here, and I’m sure it won’t be the last!
I’ve searched the forum for an answer, but I haven’t found one—so I’m relying on your expertise.

I’m designing a vessel according to ASME BPVC Section VIII, Division 1 (2023), and I’m calculating the pipe per UG-27(b).
My question concerns the joint efficiency to use. I came across the 4th answer in this thread : https://www.nationalboard.org/PrintPage.aspx?NewsPageID=144. I have a similar vessel setup: a pipe, two nozzles spaced 1500 mm apart, and two flanges.

No ligament efficiency — got it;
Cylinder joint — okay;
Nozzle joint — okay;
But why are the circumferential joints multiplied by two? Is this related to converting circumferential to longitudinal joints? If so, how is that done?

I’ve seen rules for converting ligament efficiencies (UG-52/53), but not for joint efficiencies.
Also, where exactly in UG-27 does it say that I should use longitudinal joint efficiencies only ?

Thanks in advance, and have a great weekend!

Martin

 

[Geoff13]

But why are the circumferential joints multiplied by two?

Not sure where you're reading this. Let me take a guess, and if I'm on the wrong track ask a follow-up question with the specific sentence or formula.

With a simple free-body diagram you can figure out the basic thin-plate membrane forces. The load in the hoop direction of a cylinder, which will be resisted by the longitudinal weld, is PR per unit length. The load in the axial direction of a cylinder, which will be resisted by the circumferential welds, is PR/2 per unit length. Other shapes (spherical, conical, ellipsoidal, etc) will have different formulas.

Also, where exactly in UG-27 does it say that I should use longitudinal joint efficiencies only ?

Joint efficiency is only related to the weld type and level of inspection as described in Table UW-12, not the direction of the weld / stress.

 

[MartinVT]

Good morning and first of all, thanks for your answer !

Not sure where you're reading this.

I'm reading this on the webpage i shared, it says
"Circumferential joints
– Table UW-12 column (c) = 0.70 x 2 = 1.4"
So yes, there is probably a link with the coefficient 2 you mentionned between loads in hoop and axial directions of a cylinder, but I can't find where it's mentionned in the ASME.

Joint efficiency is only related to the weld type and level of inspection as described in Table UW-12, not the direction of the weld / stress.

I totally agree, that's what i understood, but once again, the Q&A link I shared mentions "Expressed in terms of equivalent longitudinal efficiency:"

But my main concerns are the origins of the coefficient 2.
Thank you again and have a nice day.

Martin

 

[Geoff13]

I've been designing pressure vessels for 30+ years and have seen the name Robert Schueler many times, so I know he's very knowledgeable, but I can't make much sense of his answer. I've never heard of "equivalent longitudinal efficiency'. Thus I only tried to give information I know, and ignored his response.

Good luck.

 

[MartinVT]

I've been designing pressure vessels for 30+ years and have seen the name Robert Schueler many times, so I know he's very knowledgeable, but I can't make much sense of his answer. I've never heard of "equivalent longitudinal efficiency'. Thus I only tried to give information I know, and ignored his response.

Good luck.

Thank you ! It has been only a year for me, so I don't know his name and often have questions.

So, according to you, the coefficient 2 shouldn't appear in these formulas ? Then the value of E to consider for the calculation of the thickness of the shell would be 0.7 ?

Thank you once again and have a nice day.

Martin

 

[David3399]

E would not be .7 for pipe. Seamless or welded (no filler metal) = 0.85
Could be 1.0 if slip on flanges are used.

 

[Geoff13]

So, according to you, the coefficient 2 shouldn't appear in these formulas ? Then the value of E to consider for the calculation of the thickness of the shell would be 0.7 ?

NO. I said I've never heard of "equivalent longitudinal efficiency". Thus I have no opinion on how, or even why, anyone would calculate it.

My only advice would be to follow the Code rather than a 20 year old Q&A posting. If the Code is not clear to you I'd suggest asking your supervisor how to do this.

 

[Morcuse]

You generally use the longitudinal joint efficiency for the hoop stress calculation, and the circumferential joint efficiency for the longitudinal stress calculation.

Also, using 2 x 0.7 as joint efficiency would provide an allowable general membrane stress of 40% above what the code allows.