Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations JAE on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Branch Reinforcement requirement on elbows 1
- Thread starter nipra03
- Start date
nipra03,
My thoughts are (based on B31.3 only, the other B31’s are probably similar):
ELBOWS
Normally ASME B16.9 standard elbows are used. These are B31.3 'listed components', i.e. approved by the code for use in piping systems, ref. table 326.1. If you modify these in any way, they become unlisted components. The code interpretation 18-06 even states that a seemingly inconspicuous modification as trimming the elbow from 90° to say 75° will cause the elbow to become unlisted.
Because the modified elbow is unlisted, you will have to prove the design according to section 307.4.2. If you do it by calculation, this means that you need to do “detailed stress analysis (e.g., finite element method) with results evaluated as described in Section VIII, Division 2, Part 5”.
So for elbows, in my view the answer to your question is NO, you can’t calculate replacement reinforcement for elbows using the rules for straight pipe. Such a calculation would also be complicated by the fact that, except for the welding ends, you don’t know the required wall thickness of the elbow, as the wall thickness is not specified in ASME B16.9, nor do you know the actual, fabricated wall thickness (unless you do a UT measurement).
So if you really need to have a branch on the elbow, the best way forward is probably to buy a branch weld fitting, like e.g. a Elbo Pipet from WFI. WFI claims to have proven all their fittings by proof tests, which is an option given for unlisted components according to B31.3 section 307.4.2.
Regarding the SIF: Modifying the elbow will also tend to change the SIF. However, since the branch is presumably smaller OD than the pipe and is added in-plane, the change may be small. The purpose of SIF is to predict fatigue failure and fatigue cracks on elbows will normally develop on the sides of the elbow, i.e. not in the elbow plane and hence not at the location of the branch. So I think you are justified in leaving the SIF as is when doing the stress analysis of the main line. An analogy is trunnions (dog-leg supports) welded to elbows. I believe that for trunnions it is common practice to leave the SIF for the elbow as is, and - if the trunnion has large forces - to do a separate local stress check for the trunnion-to-elbow connection.
BENDS
B31.3 in 304.2 give rules for calculating the required wall thickness of the extrados of the bend, which is presumably where you wish to place your branch. So – contrary to the B16.5 elbows – you actually know what the required wall thickness is and presumably you also know the actual wall thickness. This makes it possible for you to do a branch reinforcement calculation same as you would for straight pipe. Also bends normally have much larger bend radii than elbows, which means that their geometry doesn’t deviate as much from straight pipe and that it is therefore more justifiable to use straight pipe rules. To my knowledge this approach is not specifically endorsed or prohibited by the code. My estimate is that it is probably OK, in particular if the branch OD is much smaller than bend OD.
That was my considerations; hopefully others will also give their views.
My thoughts are (based on B31.3 only, the other B31’s are probably similar):
ELBOWS
Normally ASME B16.9 standard elbows are used. These are B31.3 'listed components', i.e. approved by the code for use in piping systems, ref. table 326.1. If you modify these in any way, they become unlisted components. The code interpretation 18-06 even states that a seemingly inconspicuous modification as trimming the elbow from 90° to say 75° will cause the elbow to become unlisted.
Because the modified elbow is unlisted, you will have to prove the design according to section 307.4.2. If you do it by calculation, this means that you need to do “detailed stress analysis (e.g., finite element method) with results evaluated as described in Section VIII, Division 2, Part 5”.
So for elbows, in my view the answer to your question is NO, you can’t calculate replacement reinforcement for elbows using the rules for straight pipe. Such a calculation would also be complicated by the fact that, except for the welding ends, you don’t know the required wall thickness of the elbow, as the wall thickness is not specified in ASME B16.9, nor do you know the actual, fabricated wall thickness (unless you do a UT measurement).
So if you really need to have a branch on the elbow, the best way forward is probably to buy a branch weld fitting, like e.g. a Elbo Pipet from WFI. WFI claims to have proven all their fittings by proof tests, which is an option given for unlisted components according to B31.3 section 307.4.2.
Regarding the SIF: Modifying the elbow will also tend to change the SIF. However, since the branch is presumably smaller OD than the pipe and is added in-plane, the change may be small. The purpose of SIF is to predict fatigue failure and fatigue cracks on elbows will normally develop on the sides of the elbow, i.e. not in the elbow plane and hence not at the location of the branch. So I think you are justified in leaving the SIF as is when doing the stress analysis of the main line. An analogy is trunnions (dog-leg supports) welded to elbows. I believe that for trunnions it is common practice to leave the SIF for the elbow as is, and - if the trunnion has large forces - to do a separate local stress check for the trunnion-to-elbow connection.
BENDS
B31.3 in 304.2 give rules for calculating the required wall thickness of the extrados of the bend, which is presumably where you wish to place your branch. So – contrary to the B16.5 elbows – you actually know what the required wall thickness is and presumably you also know the actual wall thickness. This makes it possible for you to do a branch reinforcement calculation same as you would for straight pipe. Also bends normally have much larger bend radii than elbows, which means that their geometry doesn’t deviate as much from straight pipe and that it is therefore more justifiable to use straight pipe rules. To my knowledge this approach is not specifically endorsed or prohibited by the code. My estimate is that it is probably OK, in particular if the branch OD is much smaller than bend OD.
That was my considerations; hopefully others will also give their views.
It is a bad idea to put branches on elbows for many analytical, practical and fabrication reasons.
Under almost all circumstances, a branch can be moved to an adjoining section of pipe.
Unless there is an extreme circumstance,(that the piping designer can explain to me) I will not permit it.
"We have always done it this way" is not good enough....
MJCronin
Sr. Process Engineer
Under almost all circumstances, a branch can be moved to an adjoining section of pipe.
Unless there is an extreme circumstance,(that the piping designer can explain to me) I will not permit it.
"We have always done it this way" is not good enough....
MJCronin
Sr. Process Engineer
MJC is totally justified in his opinion.
Just a bit more advice, don't go assuming that the codes are similar in anything before you read them. They are similar in some circumstances, vastly different in others. Even where similar, many critical details often vary. You will get into plenty of trouble not reading the codes.
Learn from the mistakes of others. You don't have time to make them all yourself.
Just a bit more advice, don't go assuming that the codes are similar in anything before you read them. They are similar in some circumstances, vastly different in others. Even where similar, many critical details often vary. You will get into plenty of trouble not reading the codes.
Learn from the mistakes of others. You don't have time to make them all yourself.
How are "elbolets" treated?
I've seen hundreds (maybe thousands) of NPS 3/4 elbolets or contoured couplings welded on elbows to accomodate thermowells.
I've always assumed that the elbolets were integrally reinforced and if welded properly, additional reinforcement was not required.
donf
I've seen hundreds (maybe thousands) of NPS 3/4 elbolets or contoured couplings welded on elbows to accomodate thermowells.
I've always assumed that the elbolets were integrally reinforced and if welded properly, additional reinforcement was not required.
donf
At XL83NL's: even though elbolets are incorporated in the latest edition of MSS SP-97, hasn't the elbow becomes an unlisted component once we drill the hole for it? Shouldn't the assembled elbow + elbolet be considered as unlisted components requiring a proof test according to B31.3 section 304.7.2 (c) ?
May I missed something in B31.3 "unlisted component" definition? Isn’t it regards to any operation been made outside of fitting's butt welding ends (including trim exceed 3 deg, hole drilling, weld of external support, etc.) ?
May I missed something in B31.3 "unlisted component" definition? Isn’t it regards to any operation been made outside of fitting's butt welding ends (including trim exceed 3 deg, hole drilling, weld of external support, etc.) ?
At XL83NL's: even though elbolets are incorporated in the latest edition of MSS SP-97, hasn't the elbow becomes an unlisted component once we drill the hole for it? Shouldn't the assembled elbow + elbolet be considered as unlisted components requiring a proof test according to B31.3 section 304.7.2 (c) ?
May I missed something in B31.3 "unlisted component" definition? Isn’t it regards to any operation been made outside of fitting's butt welding ends (including trim exceed 3 deg, hole drilling, weld of external support, etc.) ?
May I missed something in B31.3 "unlisted component" definition? Isn’t it regards to any operation been made outside of fitting's butt welding ends (including trim exceed 3 deg, hole drilling, weld of external support, etc.) ?
Shmulik, youve got a good point.
From what Ive noticed in the past on this subject of 304,7,2 here at eng-tips is a lot of controversy and vagueness. (I remember one topic where it was mentioned that cutting a 90 deg B16.9 elbow in 2, therefore making it two 45 deg elbows, also makes them fit in 304.7.2).
The naswer may well be yes, but I believe common practice is just to put the darn thing on, weld it, and use MSS SP-97 as your argument of it not being unlisted.
PS: not sure if weld of external support is really something for 304.7.2, if you mean e.g. a trunnion of dummy leg.
From what Ive noticed in the past on this subject of 304,7,2 here at eng-tips is a lot of controversy and vagueness. (I remember one topic where it was mentioned that cutting a 90 deg B16.9 elbow in 2, therefore making it two 45 deg elbows, also makes them fit in 304.7.2).
The naswer may well be yes, but I believe common practice is just to put the darn thing on, weld it, and use MSS SP-97 as your argument of it not being unlisted.
PS: not sure if weld of external support is really something for 304.7.2, if you mean e.g. a trunnion of dummy leg.
XL83NL, I'm not sure whether we fall into the "Sometimes" or into the "Maybe" of the ASME…….
As I may understand, the fittings defined by ASME B16.9 are buttwelding fittings, and this is the only purpose you are allowed to use them since you never going to have their manufacturer's design basis. You may hold 2 identical elbows from 2 different manufacturers, and each would react differently for a same hole you drill. The reason that it becomes "unlisted" is that the result of any operation you are going to make is unexpected.
As I may understand, the fittings defined by ASME B16.9 are buttwelding fittings, and this is the only purpose you are allowed to use them since you never going to have their manufacturer's design basis. You may hold 2 identical elbows from 2 different manufacturers, and each would react differently for a same hole you drill. The reason that it becomes "unlisted" is that the result of any operation you are going to make is unexpected.
RoboCop16,
"An analogy is trunnions (dog-leg supports) welded to elbows. I believe that for trunnions it is common practice to leave the SIF for the elbow as is, and - if the trunnion has large forces - to do a separate local stress check for the trunnion-to-elbow connection."
This is not correct. The SIF's for the bend are altered by the addition of the Trunnion or in this case a branch. You cannot just use the Code calculated SIF's for bends which have welded attachments. The Flexibility of the bend is also compromised.
"An analogy is trunnions (dog-leg supports) welded to elbows. I believe that for trunnions it is common practice to leave the SIF for the elbow as is, and - if the trunnion has large forces - to do a separate local stress check for the trunnion-to-elbow connection."
This is not correct. The SIF's for the bend are altered by the addition of the Trunnion or in this case a branch. You cannot just use the Code calculated SIF's for bends which have welded attachments. The Flexibility of the bend is also compromised.
". . .even though elbolets are incorporated in the latest edition of MSS SP-97, hasn't the elbow becomes an unlisted component once we drill the hole for it."
Nope - by definition, any O'let requires a hole in the main run/header. They are a Branch Weld Fitting. Without a hole, there is no branch connection. And without a hole, there is no need for any reinforcement -- reinforcement/repad is put on a pipe/vessel to mitigate the hole that just got cut in the pressure boundary. No hole, no need for reinforcement.
Nope - by definition, any O'let requires a hole in the main run/header. They are a Branch Weld Fitting. Without a hole, there is no branch connection. And without a hole, there is no need for any reinforcement -- reinforcement/repad is put on a pipe/vessel to mitigate the hole that just got cut in the pressure boundary. No hole, no need for reinforcement.
-
1
- #14
I suggest writing an interpretation on this is the best approach, maybe attend an B31.3 meeting.
Unfortunately, since MSS SP-97-2006 is referenced in te latest edition of B31.3, we'll have to wait (I guess) for the new B31.3, which hopefully references MSS SP-97-2012, therefore allowing the interpretation to be made up
Unfortunately, since MSS SP-97-2006 is referenced in te latest edition of B31.3, we'll have to wait (I guess) for the new B31.3, which hopefully references MSS SP-97-2012, therefore allowing the interpretation to be made up
- Status
Similar threads
- Locked
- Question
- Question
- Locked
- Question