Maximum socket weld gap? 9

[donf]

I'm looking for "rules" or a least guidelines for the maximum allowable gap between the end of a pipe and bottom of the socket in a socket weld connection. Another way of asking the question: What is the required MINIMUM insertion length?

I've looked at ASME B16.11 and B31.3. If it's in there I can't find it.

I'm aware of the requirement for the 1/16" gap. (but I question that. Why wouldn't the same concerns apply to double welded slip-on flanges?)

Thanks for any help.

donf

 

[bernoullies123]

ASME B31.3-2010 Fig. 328.5.2B Typical Details for Double-Welded Slip-On and Socket Welding Flange Attachment Welds No. (3) Socket Welding Flange requires 1.5 mm approximate gap before welding.

ASME B31.3-2010 Fig. 328.5.2C Minimum Welding Dimensions for Socket Welding Components Other Than Flanges - Approximately 1.5 mm gap before welding.

Process piping: the complete guide to ASME B31.3 by C. Becht, ASME Press, 2002 Section 12.12 he discusses the controversy of the 1.5 mm approximate gap before welding. He notes that fatigue testing has shown that socket welds that are welded with 0 mm gap have longer fatigue lives than the ones with a gap.

 

[StevenHPerry]

Don,

I'm not aware of a minimum insertion. However, you might be able to twist ASME B31.1 135.5 to fit your needs:
"Backing off threaded joints to allow for alignment is prohibited."

Regarding double welded slip-on flanges, there is no reason for a backoff. The backoff on SW connections is intended to prevent the end of the inserted pipe from expanding into the "shoulder" (probably a better word for that) and adding stress to the fillet weld. Since a slip on flange has no shoulder to butt up against the pipe's end, there's no way (or need) to give a 1/16" clearance... unless I'm missig your point?

- Steve Perry
This post is designed to provide accurate and authoritative information in regard to the subject matter covered. It is offered with the understanding that the author is not engaged in rendering engineering or other professional service. If you need help, get help, and PAY FOR IT.

 

[racookpe1978]

The gap is - as implied above - to prevent stress from the pipe expanding, but that expansion is due to the heating (cooling) from the socket weld processes, not from the pipe heating from process fluid change-of-temperatures.

If it were bottomed out in the socket before the weld, the weld (as it shrinks down from a solid to room temperature) would crush the bottom of the pipe into the bottom of the socket, which puts a lot of stress on the weld and the pipe. Theoretically, if you could exactly predict how much the pipe would shrink, you could use that measure as the initial gap. Of course, you can't set the pipe that exactly, so a larger gap is required.

 

[Alan1972]

donf

You are right Donf it is 1/16" . Check in ASME b31.3 2010 EDITION PAGE 62 under section:

Fig. 328.5.2B Typical Details for Double-Welded Slip-On and Socket Welding Flange Attachment Welds. I think it is clearely stated in (3) approsimat gap is 1.5MM (1/16")

Give us a feedback if this is not what you are looking for

Regards

 

[gtaw]

Maybe I am reading the post incorrectly, but I believe the question is; "How far into the socket must the pipe extend?"

The ASME codes address how much "pull back" is required before the weld is made, but not how much gap between the end of the pipe and the socket shoulder must be present after welding. It will close as the weld cools, but again, the code is silent on how much it can shrink and still be acceptable. Still, none of that has anything to do with the question.

I put the same question to the chair of one of the ASME piping committees. His response is the strength of the joint has nothing to do with the distance the pipe extends into the socket. The concern is that the welder does not burn through the pipe wall while welding the joint. His position was that as long as the pipe extends into the socket far enough that there is no evidence of burn through or melt through, the intent has been met.


Best regards - Al

 

[phighfill]

Would this possibly go back to B16.11? I have the 1996 edition and Figure 2 on page 5 shows an example and the gap for fit up distance as 0.06". Am I understanding this correctly?

 

[gtaw]

That is the minimum gap in the "as fit-up" condition, i.e., before welding the joint. That gap will get smaller once the welding is completed and cooled to room temperature.

Different welding standards use different gaps between the socket shoulder and the end of the pipe. NAVSEA work usually calls for the gap to be between 1/16 to 1/8 inch. ASME does not have a maximum specified.

It appears that you are reading it correctly.

Best regards - Al

 

[pennpiper]

Go here for a good article on the subject:

http://www.wermac.org/fittings/socket_weld_general.html

I think we all know and understand the 1/16 inch gap we are supposed to allow between the end of a piece of pipe or a fitting and the bottom of the socket on socket weld fittings or flanges. No problem. Right?

But, we do not have a good answer to the original question. I will try asking it in a different way:
1. Can I insert "it" (the connecting object pipe or swage nipple) into the "socket" of a fitting or flange only 3/4 of the way into the socket?
2. If I can insert "it" only 3/4 of way into the socket what about if I insert it only 1/2 of the way?
3. If I can insert "it" only i/2 of way what about inserting "it" only 1/4 of the way?
4. What is the point where the amount of insertion is no longer adequate for the structural integrity of a Socket-Weld joint?

prognosis: Lead or Lag

 

[gtaw]

As per my response, the chair of the piping committee stated the dimension of the insertion is not critical provided the insertion is sufficient to prevent the welder from burning through the joint. The strength of the joint is not dependent on the amount of insertion.

If the amount of insertion is not a factor in determining the strength of the joint, 1/16 inch is sufficient provided the welder has sufficient skill to produce a sound weld without burn through or excessive melt-through. If the welder possesses the necessary skill, 1/32 inch is sufficient. If you want to go for the worst case, don't insert the pipe into the socket at all. Just make sure the weld provides the minimum throat dimension needed to transfer the load from the socket to the pipe.


Best regards - Al

 

[stanweld]

I have seen far more in-service failures of socket welds which had very little insertion - usually less than 1/8". I generally recommend specifying 1/4" minimum insertion.

 

[donf]

Thanks to all for your responses. I appreciate you taking the time.

Bernoullies123 – Thanks for the reference to Dr. Becht’s guide. I have it on my list to order.

StevenPerry - My reference to a double welded slip on flange was in response to the theory that when making a socket weld, if the end of the pipe is bottomed out in the socket, the pipe can’t accommodate the thermal expansion, and may cause a fault (crack) in the fillet weld. For a slip on, if the interior weld is made first, then the end of the pipe is definitely “bottomed out”. So why wouldn’t there be more reports of cracking at the exterior weld?

Pennpiper – You provided clarity to my original post “What is the point….” Special thanks for your contributions to this particular post and the forum in general. Your book is on my list to buy as well.

GTAW – I tend to agree with you and the committee chair that’s it’s a welding issue and not a thermal expansion/stress issue.

Some final comments and the reason for my delay in responding. I waited until I had the opportunity to pay a visit to a good friend, “Charlie”, a retired pipefitter, to get his opinion on the topic. He had several interesting observations.

1. He has seen weld failures (cracks) in SW joints that had no gap, 1/16” gap and 1/4” insertion.
2. He has seen SW joints with no failures in joints that had no gap, 1/16” gap and 1/4” insertion.
3. His theory (and mine as well), was that the failures were most likely caused by fatigue. (perhaps another topic for this forum some day).
4. He surmised that the B16.11 dimensions were established based on using the same forging patterns that were used for threaded fittings (which he said came first), and the socket depth was about the same as the thread depth, reduced the amount of material in the finished product, and helped in alignment.
5. Standard fit-up practice in his day (Charlie’s been retired for 25 years) was to insert the pipe all the way, cock the fitting (creating a gap on one side), tack weld that side, then straighten the fitting (creating a gap on the other side), and tack that side. (I need to check this out)

Summary – For now, I plan to follow B16.11 and B31.x as a standard. But based on this discussion, I am more open to the opinion that it’s not a matter of gap or insertion depth – it’s the weld quality.

Thanks again for your time.

donf

 

[DavidOC]

ASME 1 PW-41.5.2. Depth of insertion shall be at least 6 mm.

 

[donf]

DavidOC,

That's it. Thank you for bringing my question to a documentable resolution.

As I said above, I plan to use dimensions given in B16.11 and B31.x as the rule, but now when an exception comes up (which precipitated this issue in the first place), I have an established industry standard to base a decision on.

A related part of this question - based on ASME 1 PW-41.5.2., would you agree that the minimum socket depth is 5/16"? (1/4" + 1/16")

Thanks again for the reference.

donf

 

[DavidOC]

If there is an internal shoulder, then yes, the minimum socket depth would neeed to be 5/16" to meet the PW-41.5.2 requirements.

 

[pennpiper]

DavidOC,
Please define in detail the "PW-41.5.2 requirements"
What is it and what does it say?

prognosis: Lead or Lag

 

[btrueblood]

That's strange. I have manufacturer's drawings from several sources showing typical 1/4" tubing socket weld fittings with socket depths of 0.25". Looks like that ASME spec can't be met with these commerical fittings?

 

[moltenmetal]

You're unlikely to be using tubing on an ASME section 1 boiler, which is what I presume that ASME 1 PW-41.5.2 applies.

There is no requirement in either B31.1 or B31.3 to my knowledge which applies to a minimum insertion of pipe into a socket-welding fitting. If you care about this, and other than burn-through risk I'm not sure why you would other than as an indication of general fit-up competence, then you cannot rely on the code to ensure that it is complied with. Similarly if you want the sockets to show a gap AFTER welding which you will verify by RT, you need to stipulate that requirement contractually- you cannot rely on the code to let fabricators know that you want it.

 

[DavidOC]

Pennpiper,

From ASME 1:

PW-41.5.2 The depth of insertion of a pipe or tube
into a socket shall be at least 1/4 in. (6 mm). There shall be
at least 1/16 in. (1.5 mm) clearance between the end of the
pipe or tube and the internal shoulder of the socket, before
welding.

 

[Kiwi2671]

Sorry to be pedantic guys but I don't think the 1/16" (1.5 mm)comes in to play in this equation.
If the shoulder is 20 mm deep you must insert 6 mm minimum.
If the shoulder is 10 mm deep you must still insert 6 mm minimum.
If the shoulder is 6 mm deep you can comply with half the requirement but not the other half.
Cheers,
Kiwi