socket weld maximum gap 6

[ManuelMarchante]

Dear All,

I need to know the maximum gap between a nipple and a coupling before welding. ASME B 31.3 calls for 1.5 mm minimum, but what about maximum?. Attached, a sketch for clarifications.

Thanks a lot.

 

[metengr]

There is no maximum. The only reason I can think of you asking this question is that your pipe length to make this connection is too short. Fix the problem.

 

[Ron]

A larger gap causes turbulence at that area that can cause additional wear on the fitting/pipe. A larger gap also gives an opportunity to collect debris that could then lead to anaerobic corrosion conditions.

Keep the gap near the specified minimum. Structurally, you want enough pipe penetration in the socket to get stability and maintain alignment, which is probably about 1/2 to 2/3 of the socket length.

The minimum allows for expansion of the pipe from the weld point to the socket end.

 

[stanweld]

I have seen many premature failures in socket welds with insertions less than or equal to 1/4 inch. Ron's estimates are pretty solid.

 

[ManuelMarchante]

My question in regard of socket weld gap was because we have an 5% of Xr inspection on this kind of weld type for to find a deviation on minimum gap, but all of us knows that probably we find more than this minimum gap on Xr films, for instance 3 or 4 mm in many cases.

Thanks

 

[QC1971]

31.3 says APPROX. 1.5mm gap PRIOR to welding. The gap after welding can vary depending on many variables. Welding process, how many tacks were used, size of the weld ect. When socket weld RT is required by our clients I have them specify the acceptance criteria since the code only addresses it with an approximate value prior to welding.

 

[gtaw]

I asked the same question a few years ago when I attended a pipe code seminar. I addressed it to a gentleman that is a member of several B31.X committees.

His response was that as long as the length of pipe inserted into the socket was sufficient to prevent the fillet weld from burning through, it was acceptable.

Good, bad, indifferent, or ugly, that was the response.

Best regards - Al

 

[metengr]

Great response. The code book is not an engineering source book, it provides minimum requirements for safe operation.

 

[racookpe1978]

And how many (tens of thousands, hundred thousands, millions?) socket weld joints out there are "visually sat" externally (because the fillet weld looks right and has the right contour and size) but have only a bare minimum insertion length? We will probably never know until the joints are broken during demolition or replacement of the original pipes.

Be prudent.

 

[gtaw]

If we never know that the insertion length is "too little" until the system is dismantled, it would appear that the insertion length was adequate.

Best regards - Al

 

[boyweld]

I know that the ASME B 31.3 calls for 1.5 mm minimum but i'm not sure about a maximum gap.....wish you luck on finding the answer pal.

BOYWELD

 

[DLite30]

According to what standard are you x-raying these welds.

These types of welds are not generally x-rayed. They are either checked with dye-penetrant or magna-flux.

 

[ManuelMarchante]

The method is according with ASME, but the requirements for to do radiographs on these kind of welds coming from our client.

 

[stanweld]

DLiteE30:
Some Owners and Engineers require random radiography of a percentage of socket welds to confirm the minimum gap. A single shot is used for verification. Some demand that the gap between the pipe and coupling shoulder be measured and be not be less than 1/16"; some demand that a gap simply be detectable. If the pipe end is observed to be in contact with the shoulder or the gap is less than that permitted by the Owners' specifications, the weld is considered defective and must be removed and the joint remade. These requirements are beyond those of the ASME B31 Piping Codes.

 

[moltenmetal]

Whenever these beyond-ASME specs come up, we specify "gap-a-lets" (a wave-shaped washer) which seats in the socket prior to inserting the pipe and thereby ensures the minimum gap required by code prior to welding. RT can then be used to determine absence or presence of the gap-a-let washer.

It always makes me scratch my head when owners spec this stuff, though, as code specifies a minimum BEFORE welding and no minimum after welding. Given the limited size range of SW, doing RT on these joints is silly in my opinion.

The maximum gap after welding is a total judgment call. If turbulence, dirt, debris, crevice corrosion or a need for true free-draining are valid concerns for small lines in a particular service, you shouldn't be using socket welding.

 

[unclesyd]

We have as many sockets welds as the anyone used in our vaporized heating media service.
Our requirements based on experience are one wall thickness for both the minimum and maximum with a two pass weld.
Our requirements are based on the fact that one cup of heat transfer fluid vaporized, or ignitied can level a large building. With the majority of our system being in a building with people and essentially a confined space where have required 100% RT since 1990.
Based on a incident where we vented 22 gals of medium as a vapor due to a bad weld we went back and did 100% RT on every weld in each system in a 18 month long project. I forget the total number of welds but I remember that we made about 200 repairs and re-piped 4 systems.

 

[moltenmetal]

unclesyd, can you explain what you were gaining by RT of the socket welds? If you were spec'ing and paying for 100% RT already, it sounds like you should probably have been using butt welding where the RT could have told you something meaningful. Another alternative would be to inspect the socket weld fillets for flaws by other means (MP, LP etc).

 

[unclesyd]

The process for making Nylon polymer is like a solid rocket, once you hit start there is no pause button. We have about a thirty minute window if we loose any part of a polymer system. An overhaul cost five days of production and between $150,000 and $200,000, this is whether the line ran 1 hour or 1 year and we have 17 production lines.
Another reason for the 100% inspection was that two independent experts on explosions and detonations concluded that we sitting on a time bomb as the majority of our piping was in confined space. Therminol and Dowtherm will detonate at 927F and there is the big possibilty of BLEV. Then alone cam OSHA and lower the limit of Therminol vapor release to about 1 tablespoon full. Also during this peroid we have the big leak leak mention in the last post, Fortunately this was before the lowering release limits. Another problem we encounter was welder apathy, they either didn't like making socket welds or just couldn't.
Normally Leaks caused by porosity are small enough to be penned shut.
Our process areas are very confined spaces with very little room to manipulate pipe, 3/4" to 2", so the use of socket welds were deemed the most expedite way to accomplish the piping installation. A secondary factor not normally considered is that during the initial construction the majority of small CS pipe was welded by acetylene torch. It was a lot quicker to make a fillet weld on a socket than using the existing electrodes or acetylene torch of the day to make a butt weld.
We look at the gap. If the pipe was cut with a pipe cutter or portaband. The portaband cut are normally the ones that caused problems as a lot of cuts are biased by as much as 1/4" so one side has the proper insertion and the other is flush with the end of the fitting. We require that the inside edge of the fitting and the pipe be fused, no bridging of the fillet weld. No discernible porosity. Two passes on the fillet weld. Fitters using the fitting as a crutch, no insertion of the pipe in the fitting.
We have both a primary vapor heating system system, 70 psig and 650F, and a secondary system, 30 psig @ 600F. Each line has 6 reboilers that are feed primary system condensate and the primary vapor is used as the heating media to make the secondary vapor use to heat our process vesssels. We only use the latent heat of vaporization of the heating media.

 

[sjrfc2]

The fillet weld is eccentrically loaded and if the pipe/coupling is subjected to a tensile load (axially) then the fillet will have a tendency to rotate as the pipe tries to pull out of the socket. If you assign a moment per Roark & Young 7th ed., Table 13.2, case 4 & 11 (thin wall pipe) you'll get different deflections as you vary the distance from the end of the pipe. The pipe inserted into the socket is providing stiffness and keeping the weld from rotating.

The proper pipe insertion will keep weld stresses down but its a matter of degrees. I looked at this a few years ago and its not linear. As the insertion gets really poor, weld quality will be impacted. I remember reading work done as part of an EPRI effort that showed really poor insertion would result in sugaring at the ID.

 

[Duwe6]

As an ex-nuke inspector who has had many of the fit-ups & weldouts of sockets I visually inspected & accepted X-rayed:

Preferred gap/withdrawl is 1/8" to 3/16", because weld shrinkage will decrease the final gap. If a 1.5mm / 1/16" gap is used, X-ray will seldom show a remaining gap. Minimum insertion was 2 x wall thickness of pipe.