Category B, Type 1, Sport RT....E = ??? 5

[dtn6770]

I hate to resurrect a topic that’s been utterly beaten to death…but I will. Actually, I more so hate being "that guy” who brings the subject up AGAIN.

It’s the dreaded UW-11(a)(5)(b) paragraph relating to the joint efficiency. Here’s a very specific situation; Non-lethal, non-steam service, NPS 12 XH seamless shell section (SA-106-B) attached to a seamless formed, 0.500 in nominal, 2:1 ellipsoidal head (SA-516-70) with 2 inch straight flange. RT-2 is the claim. The circumferential joint will be of Type 1 and will be spot radiographed.

What joint efficiencies should be used to calculate the longitudinal stress calculations? My interpretation pushes me towards E = 0.85 but I suspect that a lot of folks would contest that for E = 1.0.

My rational is as follows.

UW-11(a)(5) applies to the full radiography requirements of Category A and D butt welds whereas the circumferential joint in question is a Category B butt weld.
UW-11(a)(5)(a) criteria is met because the Category B weld in question will be of Type 1.
UW-11(a)(5)(b) criteria is met because the Category B weld in question connects seamless vessel sections or heads and will be spot radiographed.

UW-12(d) throws out joint efficiencies (E) for calculations involving circumferential stress (longitudinal seams) NOT for longitudinal stress calculations (circumferential seams).

FIG. L-1.4-1 leads me to E = 0.85 for Category B, Type 1 butt welds because of the spot radiography.

FIG. L-1.4-2 implies E = 1.0 for a seamless head when UW-11(a)(5)(b) is met but may only apply to seamless hemispherical heads joined to main shell (Cat. A welds).

FIG. L-1.4-4 leads me to E = 0.85 because the joint is a Type 1 and will be spot radiographed.

Example L-1.5.2(b) supports the use of an E not equal to 1.0 for a Category B, Type 2, spot radiographed joint meeting UW-11(a)(5)(b).

Example L-1.5.5(b) supports the use of an E not equal to 1.0 for a Category B, Type 2, spot radiographed joint meeting UW-11(a)(5)(b).

Example L-1.5.5(f) supports the use of an E not equal to 1.0. However for a Category B, Type 2, spot radiographed joint meeting UW-11(a)(5)(b) I would expect E = 0.80 instead of the given solution of E = 0.65 which implies no radiography…a possible mistake in the solution?!

Lastly but not imply that Code software packages are always correct...APV defaults to the same longitudinal stress E (0.85) for RT-2.

 

[vesselfab]

in my humble opinion

if you put in a good type 1 seam

e=1.0 because seam is per ug-11(a)5(b) and not
the normal spot rt

 

[vesselfab]

i should have said if UG-11(a)5(b) is being used and met

and the vessel is being stamped rt-2 but could also be using rt-4 if the head is the only thing that needs E=1.0

and the other category seams are also being spot x-rayed per UG-11(a)5(b) or enough x-rays being done to compensate

APV software should have a button for ug-11(a)5(b)

 

[dtn6770]

APV does alow the user to change the default joint efficiencies. There should be an image of their help section where RT-2 is described. Note that I have pointed out to them that they have "circumferential" and "longitudinal" swapped in the last sentence of the RT2 write up.

 

[TomBarsh]

I didn't read in detail all of the first post. The key is that you want to determine the joint efficiency for the weld based on longitudinal stress...not circumferential stress or for thickness of the head.

Because the head-to-shell seam is spot RT then E=0.85 is applicable to this joint when considering the required thickness of the cylinder based on longitudinal stress. This is assuming that this seam has spot RT in addition to the particular spot required to meet UW-11(a)(5)(b), but then this is implied by RT2.

 

[jamesl]

I think it depends on how many spot RT you did on the circumferential joints. If you just did one spot, that is just enough for UW-11(a)(5)(b); there is nothing left for circumferential joints. We will have E=1 for hoop stress, E=0.7 for longitudinal stress.

I know E=0.7 will be controversial. However read that article published in 2000 by Tom Pastor of HSB.

 

[dtn6770]

jamesl...by grace of the Google Gods I was able to acquire a copy of the article you referenced (as well as the other HSB Pressure Points publications since 1998, for future reading). The 70% joint efficiency that is mentioned does offer food for thought. More importantly the article nails the point I’m trying emphasize which is the difference between E for hoop/circumferential stress (long. seam) and E for long./axial stress (circumferential/girth seam).

TomBarsh…yes!

UW-11(a)(5)(a) and UW-11(a)(5)(b) identify criteria that must be met to allow “…Category A and D butt welds…joint efficiency permitted by UW-12(a).” I get the impression that some take the association of Type 1 and 2 Category B welds with UW-11(a) Full Radiography as allowing for E = 1.0 even though UW-11(a)(5)(b) states that those welds are (minimally) spot radiographed.

 

[CodeRef]

Tom was correct to point out the additional spot shot..... because Spot RT of weld increments performed to UW-11(a)(5)(b) MUST be separate shots. Paragraph UW-11(a)(5)(b) states the following:
Spot radiographs required by this paragraph shall not be used to satisfy the spot radiography rules as applied to any other weld increment.
See UW-52(b)(1) for the definition of a weld increment.

Note: The definition of a weld increment is not specific to any category of weld. A weld increment needs only to have a joint efficiency assigned per Table UW-12 and shall not exceed 50 ft in length and includes welders/operators as applicable…

FYI.......There are blocks in the Manufacturer's Data Report (Form) for the joint efficiencies applicable to the longitudinal and circumfirential welds....The AI should be able to clear up any confusion onemay have..



FAQ731-376

 

[vesselfab]

about that .70 joint efficiency

uw-12(d) explains it quite nicely for type 1 welds

for circumferential stress (long seams)
e = 1.0 when using uw-11(a)5(b)
e = .85 when uw-11(a)5(b) is not met

each circuferential seam (longitudinal stress)
is treated on it's own

e=1.0 when uw-11 is met
e=1.0 when uw-11(a)5(b)
e=.85 when uw-11(b)is used

that guy from HSB may be the greatest thing since sliced bread....but i will continue what is written in the code book

 

[jamesl]

vesselfab,

UW-11(a)(5)(b) is meant to be used for Category A and D joints, not for Category B.

For Circumferential joints, if full RT, E=1; if spot RT (in addition to spot RT used for UW-11(a)(5)(b)), E=0.85; if no RT or spot RT only to satisfy UW-11(a)(5)(b), E=0.7.

 

[vesselfab]

read UW-11(a)5(B)

listed under full radiography

category b or c welds yada yada

I don't see how you can say does not apply to category b welds.

it's written for category b and c welds under full radiography.

it allows E=1.0 for non special service vessel as long as you follow it and stamp vessel RT-2

 

[jamesl]

Vesselfab,

There are two different joint efficiency here, one for long seam, the other for circum seam.

When you have full RT on long seam, no RT on circum seam, joint efficiency for long seam is E=0.85 (E=0.7 for circum seam). You then do UW-(a)(5)(b), E=1.0 for long seam (you can stamp RT-2), E=0.7 for circum seam. If you do additional spot RT on circum seam, E=0.85 for circum seam. Only after you have full RT on circum seam can you use E=1.0 for circum seam.

 

[vesselfab]

that's not the way we read it...but you go ahead and do it your way

no problem

 

[weldtek]

jamesl,
I've read the article you referenced, but, don't see it as you do.

 

[jamesl]

Weldtek,

Can you comment on my last post and point out which part you see differently? Thanks.

 

[dtn6770]

The QC Manager and AI I'm dealing with conceded to E=0.85, not 1.0, for the situation described initially. I had to emphasize E for longitudinal stress at the circumferential joint. Ironically, this stress usually doesn’t govern the design, the circumferential stress of a longitudinal joint does.

As debatable as this topic is I put a lot of stock in the Appendix L examples...L-1.5.2, L-1.5.3, L-1.5.5, and L-1.5.6.

Good discussion and I hope no one goes home mad.

 

[eliebl]

Don't confuse the head efficiency with the joint eff. If you use the spot from UW-11(a)5(b) for an E of 1 for the head, you cannot use it for the circ seam. It shouldn't matter because that E would only covern if there are loads which would increase the longitudinal stress.

There should be differentiation between the head "quality factor" E=1 for UW-11(a)5(b) and E=1 of the circ seam which is efficiency for long stress.

There is also an interpretation with respect to something like this but I do not know the number off hand. It is likely around 2004.

My 2 cents.

EJL

 

[CodeRef]

jamesl is correct, Tom Barsh is correct, I am correct,......hmmmmmmmm lets see......oh....your AI is always correct..... so just go ask him/her. After all he/she is the one who will sign the Data Report. If it is wrong, you will know about it come Joint Review time...or whenever the AI supervisor comes around to check up on the AI...

FAQ731-376

 

[vesselfab]

dtn

get your qc manager and AI to read this old National board article. it pretty well sums up what i have been trying to say, not too eloquently, i should add

https://www.nationalboard.org/SiteDocuments/Bulletins/SU06.pdf

 

[dtn6770]

vesselfab, thanks for the reference material.

Still, the predominant topic in the examples (pages 3-5) involved efficiencies of seamless shell sections and seamless heads…not the welds joining them. The governing equations for thickness determination are UG-27(c)(1) for (circumferential stress) shell sections and UG-32(d) for 2:1 ellipsoidal heads, as an example. I don’t dispute anything mentioned.

For shell sections there’s still the need to evaluate, using a potentially different efficiency, the longitudinal stress, UG-27(c)(2) even though it doesn’t usually influence the design (page 19, footnote 16).

Part of the answer to the article’s 4th question completely blows my mind. I’m good with the 1.0 for the longitudinal cylinder joint and 0.85 for the nozzle joint but what he did with the circumferential (head-to-shell) joints has me stumped. They’re Category B, Type 1, spot examined, butt welds so I would have selected 0.85 from UW-12 column (b). How he arrived at 0.70 from column (c) and why he multiplied it by 2 are complete mysteries to me.