Socket weld fit up tolerances

Date 10-14-2003 02:17

No other ASME code that I have seen addresses this. At the plants where I work if the pipe breaks the line into the fitting it is acceptable. The socket weld develops it strength from the fillet and not the engagement but being just even (no insertion) could reduce the throat of the fillet if the root is even slightly concave. This is a real problem when reusing fittings in the field repair business. Whats worse, springing the pipe or using a poor fitup? Neither are good but often that is what we face.

I understand B31 committee is looking into this subject now will hand out some guidance in the near future. I suspect that insertion depth will be a rule of thumb and be related to expected depth of weld penetration.

But for right now you do not have any definitive guidance.

By TimGary (****)

Date 10-14-2003 12:44

ASME B31.1 1998 127.3 Preparation for Welding (E) Socket Weld Assembly...

"In assembly of the joint before welding, the pipe or tube shall be inserted into the socket to the maximum depth and then withdrawn approximately 1/16" (2mm) away from contact between the end of the pipe and the shoulder of the socket."

It seems to me that through the use of the word "shall", the Code expects there to be a 1/16" gap. As it is difficult to measure exactly something that cannot be readily seen, they give the fitter a little bit of leeway by saying "approximately", but I wouldn’t take that to mean that 1/16" is the minimum and there is no maximum.
I think it has something to do with the structural integrity of the joint. As you are not achieving the full through thickness properties of the base material with a fillet weld, the joint most likely requires the additional backup strength of the inserted pipe against the inside of the socket in the advent of side loading forces.
I'm no engineer, but that explanation makes sense to me. As a CWI, I would use that argument, if necessary, to backup the rejection of a socket weld fit-up with any gap over 1/4".
Does that sound right to you guys? Any comments welcome.

Tim

By jwright650 (*****)

Date 10-14-2003 13:46

You could insert the pipe as far into the socket as it will go, make a mark on the pipe and that would give you a reference to see how much to pull it out to achieve your minimum 1/16".
Just a thought,
John Wright

By chall

Date 10-14-2003 15:22

Hi Tim,

I appreciate your conservative approach to this issue. However, I have never seen anything related to joint design that considers/includes the strength of the pipe OD, pressed against the socket ID, contributing to the joint strength, regardless of direction of loading.

Having had this question posed to me a number of times, the best answer I have ever come across is the one above by Neal Chapman. The practical advantage of insertion is limited to the weld penetration.

Just my opinion.
Charles Hall

By newbe Date 04-05-2008 20:53

any info on diametrical clearance?

By thirdeye (***)

Date 10-14-2003 14:56

The 1/16" mininum gap is called out in many specifications. Have you guys ever seen socket weld contraction rings? They fit the ID of the joint and have three elevated areas to allow for an exact gap of 1/16".
Not only do you save time in scribing, the connection does not have to be level when using the rings. I'm sure there are several suppliers but the ones I've seen used are called Gap-O-Let and come from GAL Gage Co.

Thirdeye

By JTMcC

Date 10-14-2003 21:57

Your standard issue pipefitter and or welder follows the standard issue technique for socket fittings and it doesn't involve measuring and marking or gap holding rings. He just bottoms out the fitting, rocks it to one side , tacks that side, lets the tack pull the fitting square or close to it, gives it a tap or two if needed to get real square, and tacks the opposite side. Billions upon billions of socket welds have been made in the cosmos this way.

JTMcC.

By cawelder (**)

Date 10-14-2003 23:22

That's the way I've always done em. Drop it down, tack it, then pick it back up until level. The gap is only for expansion, I dont think it has anything to do with structural integrity. I had to replace a lot of pipe once that was just slid all the way in and welded. All of the joints only lasted about 6 months. This was hydraulic oil piping, and when the oil got hot the connection could'nt expand. Their for cracking the welds.

Chuck

By rodofgod (**)

Date 10-14-2003 23:49

Over here in the U.K., the Clients can be quite funny about socket welds, often insisting on a 2mm 'expansion' gap. This is checked by X-ray!
We either use the scribe and with-draw method or use the gap-o-let's previously mentioned.
Also, we must 'show' two weld runs on any socket.That means one down and two over the top!
As has been said before,it's all to do with expansion of the joint during and after welding. I've seen socket welds literally ripped out due to the forces experienced during and after welding.

By insp76

Date 10-15-2003 02:36

Thanks for all the replies , somebody dug this thread up from Aug. 28, and I think it`s a good question, It seems the biggest problem is that only ASME Sec. I has a definitive rule on minimum insertion as I mentioned above, why?? only Sec. I, it seems all ASME codes should have this statement. But anyway, the problem is that when a fabricator makes a mistake and cuts the pipe a little short they want to make it up by pulling out the pipe from the fitting, sometimes all the way to the edge. If you perform an in process exam per B31.3 or RT the joint for gap and discover this you as an inspector donot have the right to reject the fit up because you have no statements in the code to back it up with. The problem I always run into is , if I make a call to cut something out and refit ,I better have a statement somewhere in the codes or owner specs to back it up, or I`m in big trouble!! I never try to win a argument unless I`m holding a highlighted and underlined statement from the code book or spec, then usually there is no argument. I just say , do you want to read it or do you want me to read it to you? I kind of think ASME should help is out what do yall think? Thanks, and Have a good one!

By TimGary (****)

Date 10-15-2003 15:19

You're right.
The code should call out a max dimension as well as the min.
I hate being put in the position of having to "interpret" what the code is trying to achieve, as an interpretation is basically an opinion, which most always leads to useless argument...

Tim

By JTMcC

Date 10-15-2003 20:36

The method I described, gives you the proper gap. It's not some hokey redneck method, it's a proper technique.

JTMcC.

By pipewelder_1999 (****)

Date 10-16-2003 13:12

I may be a hockey redneck myself but the method you describe is the one I have always used on "real jobs".

On Nuc work I can see a need for closely monitoring this to MINIMIZE how much gap is left AFTER welding which would reduce the size of a crevice that could hold contaminated corrosion and wear products. This is based on something I read years ago and may not reflect the practices of the civilian nuclear power industry.

Have a good one

Gerald

By pipewelder_1999 (****)

Date 10-16-2003 13:00

You check the socket bottoming clearance by RT AFTER welding ? That seems useless. If you do it before welding. that seems like an VERY excessive cost.

I have used the scribe method before but the most efficient method is to rock the fitting iin the socket. Can anyone provide ANY documentation that indicates the negative effects from having too much bottoming clearance ? If this is so "critical" then shouldn't the size of the 1st pass, number of passes, amount of tacking be equally controlled. All of those factors could cause the pipe to draw in more than the 1/16".

I have NEVER seen sockets come apart because of this. I have seen some cut out by untrained individuals that noticed that after welding the bottoming clearance no longer existed. No swellinmg of the pipe end or weld cracking was observed.

I think this is foolish and clearly indicates a lack of understanding on the part of the engineer/inspector that made this call. Of course I made my statement regarding what the "code says" and why and then just went on playing "Dumb ole welder". But I do think its pretty silly how much cost is added to things that involve people that add to the code.

Thanks

Gerald

By insp76

Date 10-17-2003 02:09

Gerald, I`m not a red neck I`m a "coon ass" but that`s a whole nother story ,but anyway the reqirements of the ASME codes are met when the 1/16 gap (BEFORE WELDING) are met, the codes don`t say anything about what the joint should look like after welding is completed. I agree 100% that it is rediculous to go through the time and expense to RT socket welds for the gap requirements when accually the codes do not state any such requirements. Like you say if you fit up with a 1/16 gap depending on the heat involved the joint will close up leaving no gap. The only way to insure compliance with the codes is to preform an inprocess exam as described in B31.3. The ole hold it down tack it rock it up while it`s still hot with a level or a square on it works every time and it gives you about a 1/8 gap that will allow for any expansion of the joint. I have never seen a socket weld crack after welding but I have seen many crack after being in service but I`m not sure if the lack of a gap caused it but it could have. Thermal expansion can cause a weld already under stess to crack in a heart beat ,can`t it. Have a good one!

By pipewelder_1999 (****)

Date 10-17-2003 20:04

I had some 1/2" Sch 80 SS butt welds RT'd years ago in which the adjacent socket welds were observed and REJECTED. The organization that had them cut out was the Navy. This was one of thsoe cases in which a little knowledge caused alot of rework.

Redneck, coon ass, etc. Isn't it all another word we can use to describe "Different"

Have a good day.

Gerald Austin

By Michael Sherman (***)

Date 10-15-2003 20:12

By jwright650 (*****)

Date 10-15-2003 20:49

I agree with you all on the min gap, but I think the original question also addressed "What is the Max?". I think this is what was not very clear to Inspect76. I'm not getting too involved with this one as I don't have the code to look at for myself. What ever method you use to prove you have complied with the 1/16" with-drawl, in my opinion doesn't matter as long as you can prove what you did during fitup satisfies that requirement. Now what is the "Max allowed" by the code?
John Wright

By jwright650 (*****)

Date 10-15-2003 20:51

Maybe it was Tim that concluded there was no Maximum.
John Wright

By insp76

Date 10-16-2003 03:03

Mike , you have a good point but when you take a look at FIG. 127.4.4(C) "....(Minimum) welding dimentions required for socket welding components other than flanges.....This statement seems to only give (minimum) gap not maximum gap or minimum insertion. This is the closest statement to a minimum insertion rule but it just doesn`t` quite make it in my opinion. Like you said " all comments are welcome " Thanks for your interest, Have a good one!

By Niekie3 (***)

Date 10-18-2003 15:33

Hi Mike

ASME B31.3, para 328.5.2 also addresses this, but by the use of pictures, (figures) very similar to those in B31.1. It would also seem to me that the 1/16" distance is neither a minimum or maximum. It is an "approximate" dimension. This means that something twice this value would not qualify as being approximately 1/16".

Lets look at this another way. If the collar of the fitting served no purpose, why do the manufacturers put it there? If I was making a fitting with a collar, into which a pipe needs to fit, I would make it the minimum size to meet code design requirements, with a small factor of safety. So, if you only need to have a 1mm insertion depth, (to allow proper weld penetration) why make it 20mm long? I would be loosing money!

Just as a matter of interest, we in South Africa also often radiograph these socket welds for quality control purposes. I have heard all the arguments about doing it right to begin with, then the radiography becomes superfluous, (Using the scribe method etc.) but if we take this line of reasoning, very little inspection would be required because we can just trust the welders and boilermakers to do everything right. Where I come from we have a saying: "Trust is good, check is better."

Regards
Niekie Jooste
Fabristruct Solutions

By SA Inspector Date 01-11-2007 09:54

Well said Niekie!

RT is the only practical non-destructive method to check on the size of socket gaps (and weld quality). If no gap was used before fitting the pipe into the socket the resultant X-Ray image is the same as a socket that has closed up after welding shrinkage - therefore an after welding gap is needed in the codes so that the existence of a gap before welding is proven!

Regards
Mark Rudman
Olivier Survey Group
South Africa

By js55

Date 01-11-2007 19:05

Not that the code bodies necessarily always get it right, but perhaps the silence on this one is indicative of a lack of evidence (or non existent evidence) that a max matters so much. Perhaps Neal was the closest. This sounds a lot like terf wars to me. Like arguing over how many angels can dance on the head of a pin. Maybe the code writers are aware that there is no evidence a max compromises the viability of the joint. I think before anybody representuing inspection or production goes 'bowing up' and insisting upon an interpretation in the face of code silence perhaps some research as to how much it matters is in order. Until then the issue remains ambiguous. Even 'approximately' is connotatively ambiguous.

By js55

Date 01-11-2007 19:11

As a follow up, a code document is a minimum safety standard. It is not a cookbook. Minimalist wisdon folks. Minimalist wisdom. So I would ask, if one is concerned about maximum gaps in socket welds, where are the failures to substantiate this concern? If one is to argue that it is the codes responsibility to prevent failures (which is certainly true to an extent), then how far shall we go. How big do the codes need to be before this, once again, ambiguous, criteria is satisfied.

By jon20013 (*****)

Date 01-11-2007 19:19

Niekie! Man, I haven't seen you posting in here for AGES!!! Have you been away? Good to see you back anyway!

By jwright650 (*****)

Date 01-11-2007 21:44

Uh, Jon....his post is dated back to 2003....in fact this whole thread dates way back.
I agree that I haven't seen him post in here in ages either, and would like to see more of his postings.