Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / ASME Codes / repair attempts
- - By WELDICCAWSCWI (**) Date 11-10-2013 16:42
Working under B31.3.  a weld fails xray for IP  . prior to making the repair the welder
drags up. The second welder makes the repair but it fails x ray for the same reason IP
The second welder  attempts a second repair which also fails . I cant find anything in B31.3
that limits the number of repair attempts or in the job specs between customer and contractor.
Were shooting two tracers on the original welder. are we required to shoot tracers on the second
welder who failed the two repair attempts?  how many attempts do they get to make this repair?
12" sa 106 sch 80   is this the county fair pitch till u win .trying to help a friend  not sure what
to tell him
Parent - By 46.00 (****) Date 11-10-2013 19:58
ASME B31.3 makes no mention of how many repair attempts, only that the repair is made acceptable to criteria. Usually the amount of tries is covered in contract specs and two times seem the industry norm.
I think the Prof (G.S.Crisi) did a paper on this subject a couple or three years ago that caused some controversy. You should be able to do a search on this forum for it.
You do not have to chaser tracer's on repair welds as per interpretation below:

Interpretation: 16-01
Subject:ASME B31.3-1996 Edition.Para.341.3.4,Progressive Examination
Date Issued: May 20, 1997
File:B31-96-053

Question:In accordance with ASME B31.3-1996 Edition,para.341.3.4,Progressive Examination, if a
defective weld is repaired, found defective again, repaired a second time and again found to be defective,
is it necessary to examine two additional items for each failed repair?

Reply: No. See para.341.3.3.

Hope this helps.
Parent - - By dbigkahunna (****) Date 11-10-2013 22:11
Unless it is prohibited in the contract, then it is pitch till you win.
It seems to me the contractor would be so red faced they would bring in a real live golden arm to make the next repair.
Is this a highly restricted weld?
Parent - - By WELDICCAWSCWI (**) Date 11-10-2013 23:26
This weld is on a friend of mines job hes a fairly new CWI I had told him to check the job
specs between owner and contractor and he said it states welding to follow guidelines
of asme b31.3 it says nothing about repairs. The weld is right out in the open . He thinks
welder may be screwing it up on purpose since it wasnt his repair in the first place.  All
the golden arms have moved to greener pastures y. This is the last welder on the job.
seems like ASME should provide more detail . They give us countless parameters  for making
a weld but tuern a blind eye concerning repairs. I know in Southen cal    SO CALIF gas Allows two
repair attempts then its a cut out.  I'll take the BIG PINK TEDDY BEAR
Parent - - By welderbrent (*****) Date 11-11-2013 00:11
One other consideration, but it is based upon D1.1 where my main interest and experience is...

There may be a clause, as in D1.1 Clause 5 about sub part 23, where repairs are dealt with.  It is stated that the contractor has the option of repairing according to certain types of methods but a couple of things must come first. 

1) Repairs must be to an approved WPS;
2) Repairs must be approved by the engineer.

You might try locating such a clause within B31.3.  That would mean the number of times would be up to the engineer on the job who would determine the usage and material ability to stand up to multiple repairs.

Have a Great Day,  Brent
Parent - - By 46.00 (****) Date 11-11-2013 02:18
There is no such clause in B31.3.
Parent - By welderbrent (*****) Date 11-11-2013 03:47
So the applicable code makes no provision for laying the responsibility for the decision onto the engineer....

He claims to find nothing in the job specs to indicate that a procedure has been defined for repairs.  But, to the OP, did he just check the General Notes on the drawings or go to the actual 500 or so page Job Specifications book?  There may be something he is missing that will help with this decision.

In his place, I believe I would still have to request an RFI at this point and thus bring it to the attention of the engineer and get his official participation in answering the question. 

Afterall, it is his job.  With no guidelines it is not the inspectors responsibility to mandate how many repairs, what methods of repair, or any other decision affecting product integrity. 

Have a Great Day,  Brent
Parent - By jarsanb (***) Date 11-12-2013 15:47
If D.O.T. or State regulations are involved, then the Code of Federal Regulations may come into play. There are restrictions to repair attempts in C.F.R. then  qualified repair procedures are required. Again, this may not apply. Given your background - it may.
Parent - - By CWI7611 (**) Date 11-16-2013 15:07
Strictly under B31.3, yes, "pitch till you win" unless regulated by some state or federal code or the owner's specification. For A106 material, relatively soft and lower yield, probably not too much damage done if any, but you do have to draw a line somewhere. Depending on how much slack in the line you have, maybe it's time to cut the weld out and make a new one. If there's no slack and you have to put in a pup, you may now have two failures. Next time keep the "golden arm" and let every one else go.

Tracers (or info shots) are always good for repairs on welds connecting pipe to fittings or valves where a cut out might destroy the end prep and make a replacement weld very difficult. Of course if you do have to cut the weld out, you could do this the way I have heard some company's field personnel do it, completely remove a portion of the weld by grinding it all the way out, then re-weld it, move to another section of the weld and repeat till you have all the weld replaced. Not sure this was approved by the company (I know it wasn't). But I guess in the mind of the field supervisor this worked out just fine. Wonder how he sleeps at night. Can't blame the welders, just doing what they are told.

An info shot on this one, IP, is kind of a no brainer. You have to go all the way through. Of course you could fill the bevel all the way up, leave the cap off, shoot the tracer and if the defect is still there call it an in-process information shot and grind back in and repair it again, then repeat as often as necessary. Personally, I wouldn't allow the weld to be filled al the way up prior to the info shot. In someone's small mind they feel "If you don't put the cap on the repair is not completed so the repair is not final", right? Again stretching the imagination a little bit, but people (and companies) do it all the time. Perfectly acceptable if you don't think about it too much.

Hope this doesn't add to the confusion too much.
Parent - - By 46.00 (****) Date 11-16-2013 18:09
CWI17611 I think we have an issue with terminology. Tracers are not info shots, we are, strictly talking about progressive sampling, in the OP case using radiography as the test method, presumably in a 10% or xx% radiograph situation.

Extract from B31.3 2011
341.3.4 Progressive Sampling for Examination.

When required spot or random examination reveals a
defect, then
(a) two additional samples of the same kind (if
welded or bonded joints, by the same welder, bonder,
or operator) shall be given the same type of examination
(b) if the items examined as required by (a) above
are acceptable, the defective item shall be repaired or
replaced and reexamined as specified in para. 341.3.3,
and all items represented by these two additional samples shall be accepted, but
(c) if any of the items examined as required by (a)
above reveals a defect, two further samples of the same
kind shall be examined for each defective item found
by that sampling
(d) if all the items examined as required by (c) above
are acceptable, the defective item(s) shall be repaired or
replaced and reexamined as specified in para. 341.3.3,
and all items represented by the additional sampling
shall be accepted, but
(e) if any of the items examined as required by (c)
above reveals a defect, all items represented by the progressive sampling shall be either
(1) repaired or replaced and reexamined as
required, or
(2) fully examined and repaired or replaced as necessary, and reexamined as necessary to meet the requirements of this Code
(f) If any of the defective items are repaired or
replaced, reexamined, and a defect is again detected in
the repaired or replaced item, continued progressive
sampling in accordance with (a), (c), and (e) is not
required based on the defects found in the repair. The
defective item(s) shall be repaired or replaced and reexamined until acceptance as specified in para. 341.3.3.
Spot or random examination (whichever is applicable)
is then performed on the remaining unexamined joints.

I think paragraph (f) is the answer to what the OP was asking in that part of his question.
Parent - By CWI7611 (**) Date 11-19-2013 03:28 Edited 11-20-2013 01:06
I too believe there is a terminology problem. Possibly I didn't understand the difference between progressive examination and tracers. There is no difference and info shots don't have anything to do with either. I went back and reread the question and I see where the OP asks if tracers are required for the original welder.

Thanks for pointing that out to me.

Different areas of the welding industry use different terminology that ultimately mean the same thing, wagon tracks (parallel slag lines), suck back or cows xxxk which also can be interpreted as a crater crack or incomplete fill on a stop/start, it could be interpreted as incomplete fusion too if you want to go to the extreme.
Parent - By Hollaway Date 12-13-2013 21:13
Very informative post. My question is, was the weld actually rejectable? I had one fail the other day and when I asked the RT tech if he measured the IP, his response was "no need, there's none allowed", which is not correct. When y'all made the first couple repairs and it was reshot, did any of it go away? You're allowed 1.5" of IP in any 6" of weld in B31.3 normal service.
- - By 803056 (*****) Date 11-11-2013 17:45
It is never the inspector's job to dictate how or how many times a repair can be made.

In the wonderful world of ASME, the contractor has full responsibility to determine how things are accomplished. The contractor is the expert, as such the contractor is responsible to provide the expertise needed to make decisions based on code requirements or good engineering judgment. In short, the contractor is the Engineer. That isn't to say the Owner has no say, but if there is a concern, it must be addressed in the project specifications.

AWS welding codes are cook books that if followed should result in acceptable welds. ASME operates under a different philosophy and set of assumptions. ASME doesn't tell the contractor how to do anything. The contactor is expected to have the expertise or to get the expertise needed to comply with the minimum requirements included in the applicable code sections.

Best regards - Al
Parent - - By welderbrent (*****) Date 11-12-2013 03:09
Interesting.  One of these days I'm going to have to buy the ASME codes just so I can read them and see what is in those deep dark recesses of 'Almost, Sometimes, Maybe, and Except'.

Have a Great Day,  Brent
Parent - - By Jovi Zhu (**) Date 11-13-2013 12:19 Edited 11-13-2013 14:16
I can't agree more with Al's post.:cool:

The "answer" can be found in B31.3 book, however, too often overlooked. See below~

B31.3-2012

INTRODUCTION

The Code sets forth engineering requirements deemed necessary for safe design and construction
of pressure piping. While safety is the basic consideration, this factor alone will not necessarily
govern the final specifications for any piping installation. The designer is cautioned that the
Code is not a design handbook; it does not eliminate the need for the designer or for competent
engineering judgment.

Understanding in a simple way is that the ASME safety Code is followed to prevent the built items from killing people but doesn't guarantee that the contractor following the Code will be a world-class company in the industry.

Engineering judgement is always necessary.

To my knowledge and experience, e.g., for elevated service temperature, repeated welding repair for several times of P No.1 steel doesn't harm too much as long as qualified WPS are followed. Creep failures can occur in carbon steel materials when subjected to longterm overheating while under stress. However, carbon steel materials are generally not used under conditions where creep is expected and the heat input of repair welding is far from making the steel and its joint longterm overheated. Same repeated welding repair to P No.8 stainless steels will result in excessive heat input that weaken the anti-corrosion ability of the steel. Common WPS's are not required by IX to be qualified by creep strength test or ASTM A262 test. That is out of the the safety Code, e.g. B31.3's scope of consideration.

For both the above steels, when dealing with low temperature service condition where impact requirement is specified, heat input is important and IX lists "supplementary essential variable" QW-409.1 to control the impact toughness of the welded joint. If repair was made too many times the owner may question the qualification of the WPS to make such repair since it was not qualified to weld with that level of heat input and the contractor can't guarantee the repeatedly welded joint still possesses the desired impact toughness.

I think ASME Code expects that an incompetent welder fails the performance qualification test by IX before (s)he is sent to the shop floor or to the site to repeatedly fail the RT. Have to confess that the actual condition of shop floor and site is not as ideal as that of the performance qualification test. But there are circumstances that some contractors fail to let the welders fail earlier.:grin:

Just some thoughts.

JZ
Parent - - By 803056 (*****) Date 11-13-2013 15:03
I am going to have to remember this post. It is clear, concise, and emphicises the right points.

Best regards - Al
Parent - - By petty4345 (**) Date 12-18-2013 14:08
I realise that thisis late and somewhat off base but, alot of threads talk about number of repair cycles.
I see alot responses that say "no code gives amount of repair cycles".
Well, that's not totally true.
NAVSEA TP 248 limits repair cycles on S-10H materials. You must gouge out and reweld your test plate as many times as you plan to do in production.
Also, for RT done after PWHT, repairs cycles are limited by PQR time at temp. (80% min.). If you don't plan for this, you may get zero attempts to repair. (Without requialification).
Parent - - By 803056 (*****) Date 12-18-2013 14:48
True, but a repair to S10H is a special case for duplex stainless steel. It is not the rule.

Best regards - A
Parent - - By petty4345 (**) Date 12-18-2013 15:12
Correct, it is a special case for S-10H duplex but, a case none the less. Just pointing out that there is no across the board answer.
Parent - - By welderbrent (*****) Date 12-18-2013 15:30
And that is exactly how 'MOST' of us 'ALMOST' 'ALWAYS' answer the question; that there are few codes that limit the number of repairs that may be made on a project.  And we continue, that is an issue for the contract documents to have stated or be clarified by the engineer of record on the job. 

As noted by you and Al, it is largely dependent upon the materials in use, the application of the usage, and the engineer's requirements for the job.  There are few restrictions within any of the codes but they are there and must be watched for. 

Have a Great Day,  Brent
Parent - By 803056 (*****) Date 12-18-2013 15:42
ASME says it best; Always, Sometimes, Maybe, or Except.

The problem comes about when some people try to make the "Sometimes" an "Always."

Al :cool:
Up Topic Welding Industry / ASME Codes / repair attempts

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill