from D1.1 2004: 2.2.3 " End returns and hold-backs for fillet welds, if required by design, shall be indicated on the contract documents". Back charge the Engineer for the work.
Welders will typically wrap the welds rather than stopping at the end of the joint. I guess this is because the welder feels the weld looks better and there are fewer problems with the weld profile. (A welder's point of view - not necessarily my point of view.)
However, I clearly remember when I worked for a different company and we got "bitten" really bad for wrapping welds. This was before I knew how to spell "QC". We were nearly finished with the job (3 coats of paint to boot) when the engineer visited the shop and saw we had wrapped welds around HSS column through plates. He went ballistic and said if the welds were supposed to be wrapped he would have made sure the drawings said so. It cost us a lot to fix the problem - not only the columns but the beams also had to be reworked.
Depending on the job, I do not automatically reject a weld that has been wrapped. But I do point out that unless the welding information says otherwise, the welds should not be wrapped. You never go wrong if you follow the written information (general notes, welding symbols, special instructions, etc.). Having said that though, I feel we have a responsibility to seek clarifications if something "doesn't look right".
Verbal information can bite you. On one hand, if you trust the source, verbal info saves time. One the other hand, people's memories are "selective" and it's amazing how stories can change later on.
Some have said that you should capture a conversation in a email and send it to all parties, with the invitation to challenge the email if they do not agree with it. If you do that, be careful. I have done that, per instructions, (including read receipts, voting buttons and all that nice stuff) only to have the person(s) say that although they got the email, they never read it over, so it doesn't count. The best approach is to send an RFI (Request For Informtion) and insist on a response. If you don't get one in a timely manner, you have to be prepared to force the issue (there are lots of ways to do that but I won't mention them so as not to get anyone into trouble).
I don't know the name but some wise person said "If it isn't written down, it never happened".
Unless the parts and get galvanised, and the drawings say to "seal all welds" I have instructed our welders not to wrap the ends. In the beginning it was diffucult to get them to stop doing it becasue a lot of our welders used to work at ELectric Boat and when welding there they always wrap all there welds for submarines.
Our building don't go under water so I had to get them to stop that. Look in A2.4 Standard Welding Symbols, there'a an illustration that show when to wrap and when not to.
I agree with the other responces, unless the shop drawing says it, they should not be wrapping, and its up to the engineer or detailer to tell them otherwise. Chris
By 803056 
Date 02-03-2009 02:26
Edited 02-03-2009 02:29
A2.4 addresses welding, brazing, and NDT symbols. It is not a fabrication standard and it does not address workmanship or design issues. You would have to refer to the applicable fabrication or welding code such as D1.1, D1.2, AISC Steel Construction Manual, etc. for design requirements.
The following photograph in bozaktwo1 changes things. He has a tubular connection that includes a PJP groove weld. He is not talking about a framed connection.
Best regards - Al
I deal with a lot of structural tubular joints here, and that's one of my biggest pet peeves. If the corners aren't wrapped up correctly, then this is what you get:
This all should have been established during the submittals. The engineer should of noted the return or not, if it was the detailers design. but if its the engineers design then it sounds like this could possibly be alleviated by using a heavier connection piece,ie angle clip. and using a larger fillet. It doesnt appear hes worried about flexibility as much as the accumulated amount of weld. so if you have a larger fillet you could use less weld length and not require a return. or am I confused? regardless always get something in writing, email or something I make it a habbit to never inspect to verbals unless its stated in my report and the individual who told me is signing it.
Based on the photograph, you have other problems besides the "wraps". As indicated by the mark "LF", it looks like you have incomplete fusion not only in the corners, but between the weld bead and the members. Are you using GMAW-S (short circuiting transfer) by any chance? I would be willing to put money on the fact that there is incomplete fusion in the area of the joint root.
I believe the terms wraps, boxing, end returns are the same thing. The length of the end return has to be controlled to prevent a simple framed connection from becoming too rigid and subject to potential "unzippering" failures. See Figure 2.7 of D1.1-2008. The same figure also appears in several earlier editions of D1.1.
Assuming you are working to D1.1, the engineer is likely to be hanging his hat on the welds being continuous around the corner of PJP groove welds, i.e., the vertical weld in your photograph. The requirement is contained in note 7 of figure 3.5 in D1.1-2008. This is not a new requirement. It has been in the last several editions of D1.1. In this case, the term "wraps" is not the proper terminology. There is no need to specify the requirement to carry the weld continuously around the corner, it is a requirement for all PJP (flare groove weld as shown in figure 3.5 lower right "side match". It should or can be depicted and specified as part of the WPS. That being said, this is a case where the fabrication drawings (not the design drawings) could include a note to reiterate the need to carry the weld continuously around the corners. Redundancy is not always a bad thing.
Best regards - Al
I agree with the incomplete fusion at the root comment. That looks like a 5/16" or 3/8" fillet, made with a weave. It's a good bet that the root is bridged over and the cavity fis ull of slag.
Good eye as always, gents.
I have taken several of these joints apart myself over the last couple years and believe it or not, found no IP at the root in the fillets (don't ask about the grooves on the other 2 faces). This is GMAW, 92/8, 22-25v, ER70S-6 .035", and for this 1F it's hauling out at 425-450 ipm. Yeah, the toes leave a little something to be desired, but profiles are good. I have a high confidence in these fillets, much better than my trials last year with the aluminum shop. We haven't dropped an engine on anyone yet, by golly! Now, I could show you guys some real doosies out of that aluminum (edited) shop. I better stop there, don't get me going!
I think it was actually 1F(UBAR) :)
Can we have a separate topic page for the Ugliest Welds? I have some photos that would blow your mind.
LOL, I see now. 1F fubar. Just wondered because it looks to be a 2F position. :-)
At least the welder had enough pride to clean off the high spots. If they had spent that much effort to clean first, they might not had to have so much post-weld beautification work.
"Post-weld beautification work" - I like that term. If it isn't in the Farm Code, it should be.
I've always separated "wrap" from boxing (a Canadian term) and End Return (AWS term)
Wrap being just enough to get around a corner or seal the end of something or the wrapping the radius of a tube corner and welded in a continuous motion with the rest of the required fillet (fig 3.5 note 7 & fig 3.6 note 2)
Boxing being "The continuation of a fillet weld around a corner of a member, as in and extension of the principal weld" fig E14E (CSA W59 with no specific length limit I've seen noted) (CSA W59 - "End Return, a non-standard term for boxing")
AWS - End Return being at least 2 times but not longer than 4 times the fillet leg size (D1.1 - 2.8.3.3 & fig 2.7) and welded in a continuous motion with the rest of the required fillet (fig 3.5 note 7 & fig 3.6 note 2)
(And fillet minimum weld length being 4 times the leg size or more or ............ D1.1 - 2.3.2.3)
Al, good to see someone else has noticed fig 3.5 note 7 ....I've personally ran into only 2 others.
And YES I do enjoy your insight. Thank you
A compliment from someone I consider to have an outstanding knowledge of welding is appreciated. Thank you.
Best regards - Al
James,
Thanks for that excellent clairification!
There are some AWS doccuments for workmanship samples and I think some Skills USA text that state "Employ the boxing technique where applicable" Maybe that language should be cleaned up in future revisions if "Boxing" is a non standard term in the language of AWS
Edit: Boxing is a standard term in A3.0.
Boxing: The continuation of a fillet weld around a corner of a member as an extension of the principal weld See Fig 23(F)
end return: a nonstandard term for boxing.
So I guess it's D1.1 that might be using slightly nonstandard language.
I think I'm starting to understand.
It does sound like the A3.0 committee and the D1.1 committee has a communication problem, no real surprise. As its been said before, CWI test questions should be answered with AWS terms and not API or ASME, the same words may not have the same meaning for each org. Now you can't trust even that.
I'll have to admit, I don't like gray areas.
D1.1 does use the non-standard term "end return". But if you notice at Annex K in D1.1 for the definition of "end return" there is an asterisk (*) in front of it. If you go to the beginning of Annex K it explains that where an asterisk appears it denotes "...terms, preceded by asterisks, which are defined as they relate to this code." So from a standpoint of consistency between documents, there is no conflict between the two documents because D1.1 states that it is only for D1.1.
If you go through Annex K you'll notice that there are many definitions with asterisks, and many of these are terms that can be found in A3.0 but are used differently in D1.1. The rationale for this is:
- A3.0 attempts to standardize terminology across numerous industries and publications,
- the US structural steel industry and the D1.1 code has been around far longer than A3.0, and as such it seems to be the general belief of those on the D1 committee that those terms have been used in the US structural steel industry for so long it would be confusing to adopt the A3.0 terms.
It could be that in the very long term, D1.1 will eventually use no non-standard terminology. For now, clearly it is important to understand the terminology used in the code that you are working to, and for D1.1 Annex K clearly indicates where terminology used is not in accordance with A3.0.
I find that sometimes this makes report writing a bit of a challenge, especially when different disciplines may use similar terms to mean different thing. A good example are the various terms to describe the throat of a fillet weld (theoretical throat, actual throat, effective throat). These may mean different things to structural designers and to fabricators or inspectors. In my experience, it is important to understand who will be reading your report, and to take great care with terminology and not assume that everyone reading it is familiar with all of the terminology (i.e. I will sometimes define in the report certain terms).
Mankenberg
In a similar note, you should never "tie-in welds on alternate sides of a common plane" if that sentence doesn't make sense I've included a helpful picture.
we dealt with this a lot in the sheet metal and sheet metal-structural world. Often times we are making containers that must remain air or water tight and there has to be an evaluation of how a fully boxed weld will affect strength.
The left "do not weld here" I agree with; the weld is leaving one plane and heading off into another.
But in the one on the right side, the way I see that picture, they're showing three welds all in the same plane, and calling for 3 separate welds instead of continuing around the corner. I disagree with breaking the weld at those corners. Unless I'm not looking at the picture right.
Hg
