Gentlemen;

The reason for the space between the end of the pipe and the shoulder of the socket is to permit the pipe within the socket to expand while it is being welded and to allow the fillet weld to contract as it cools without inducing a tensile load on the root of the weld.

The problem is more acute in some materials than others. Austenitic stainless steels would most likely survive such an event because it has good ductility over a wide range of temperatures. However, that isn't the case with carbon steels (SA53) or carbon manganese steels (SA106).

As is the case with welding, you may not experience root cracking in 90 welded joints, but the next 10 that do crack will offset any profits you had hoped to make. The other concern is that the root crack may not propagate to the weld face until several thermal cycles or until "water hammer" or a shock load event occurs. The results of such an incident is unpredictable and easily avoidable if the gap is maintained at the time of fitup.

The attached sketch is one that I use when training welders new to welding pipe and ASME or military requirements. The minimum space between the end of the pipe and the socket shoulder is the same for all ASME piping codes and NAVSEA TP278. While a single pass fillet weld is acceptable to ASME, it is not acceptable for work meeting NAVSEA TP278. The multilayer weld is required for military work and it has recently been shown to improve the fatigue life of a system subject to cyclic loads.

The minimum leg dimension of the fillet weld is dependent on the piping code used. In the case of B31.3, the minimum size is 1.4 times the thickness of the pipe wall, but not less than 1/8 inch.

Getting back to the need for the gap, consider that the pipe is expanding while it is being welded, but as soon as the fillet weld is initiated, it fixes the pipe so that is can not move freely. If the pipe is up tight against the shoulder when the fillet weld is initially started, it wants to expand (in length as well as diameter), but the fillet weld prevents the pipe from lengthening (within the  socket) as the pipe is heated. Due to the fixity offered by the fillet weld, the pipe (within the socket) pushes up against the socket shoulder, the hotter it gets, the more it tries to expand thus exerts a greater the force against the socket shoulder. When the delta T, i.e., change in temperature, reaches about 212 - 230 degrees F, the magnitude of the force is equal to the yield strength of the pipe. That essentially is the most force it can exert because any increase in loading will result in plastic deformation of the pipe end. The same force is exerted against the fillet weld that is not yet completed, thus there is an opportunity to initiate a root crack, and again, don't forget that the weld can undergo a brittle transition phase until it cools sufficiently to regain some ductility. Once the weld is completed and the pipe begins to cool, everything begins to contract, including the length of the weld leg. The contraction of the fillet weld tends to pull the pipe further into the pipe, but it is already up tight against the shoulder, thus either the weld will yield if there is sufficient ductility or the forces initiate a root crack if the force is greater than the tensile strength of the fillet weld as it passes through the brittle stage during cooling. One might jump to the conclusion that the pipe within the socket is going to contract away from the socket shoulder as it cools. It might do just that if it were not for the fact the delta T of the pipe between the ambient temperature and its peak temperature is less than the delta T of the weld metal, i.e., the peak temperature being the liquid temperature of the weld, and ambient temperature.

The derivation of the temperature change required to reach the yield point of the material was discussed a couple of months ago in another post so there's no reason to bore everyone with it once again.

Just to add a little something, I have attached a photomacrograph of a case where the failure to observe code requirements to maintain at least a 1/16 space between the socket shoulder and end of the pipe resulted in a root crack. Note that it doesn't extend to the face of the weld. Thus, it was not detected when penetrant testing was performed in the "shop". This is just one example of a number of such cracks on this particular project. Several fillet welds in this particular "lot" were clearly cracked by the time they reached the site. The remainder of the welded joints showed evidence (as shown in this photo) the cracks had already initiated when the welds were sectioned.

As for terminology; it would so very nice if the proper terminology was used so that all us poor slovenly under educated inspectors without years of field experience would not have to wade through the flotsam  to decipher the replies we read. I always thought a "stab-in" was a branch fabricated without the use of a fitting. See, I hear something new everyday.

Best regards - Al

Superflux,
  Whoa there big boy, let me fill you in on something.

If it is not backed up by a code and procedure you can do whatever you want. Otherwise, QA/QC must "thunder down" if you do not follow code and procedure.

This is NOT a "self appointed pedestal" but rather a requirement set out by contractual agreement between a customer and provider. We as QA/QC do not establish the requirements, we just inforce them. We are the middle men in this battle of the "minds" so to speak. The EOR's set the standards with procedures and code adherment, production does not always want to follow, and the customer knows this. This is why there is a need for QA/QC. THAT is why we "thunder down" as you say. WE DO NOT MAKE THE RULES!! We just enforce them.

So I say to you sir, "self appointed" MY A**!!!

John

EDIT: BTW try to "self appoint" yourself as a CWI, good luck!! I see by your profile that you list CWI, did you appoint yourself? And do you inforce requirements that are not neccessary?

You and I have similar opinions in that regards. QA/QC working with production can yield better results that an adversarial relationship.

However, I've had projects where the contractor was totally adversarial from the get go. The results is usually less than desirable, but you live with what you have, not what you'd like to have in the words of Rumsfeld (or words to that affect).

I've had a few projects where, after seeing it was to their benefit, the contractor decided to have a meeting of the minds and asked, "What do we have to do to get this project back on the right track?" They quickly discovered that they had nothing to gain and everything to lose if they didn't change their attitude and methodology.

Whenever possible, I like to get in on the ground floor with the architect and engineer as the project specifications are developed. I usually request the boiler plate be modified to indicate the owner pays for the initial inspection, but the contractor pays for any reinspection. It is amazing to see the change in attitude and workmanship when the contractor discovers he is paying the inspector out of his pocket. The reject rate drops exponentially and when I do get called back for reinspection, the punch list is completed, unlike other projects were the contractor completes one item and waits to see if I'm going to reinspect every item on the punch list or just the one he's completed. I got called back 8 times on one project that had a list of 21 items to be repaired. Finally the owner asked me what it was going to take for me to give him a "clean bill of health". I told him the argument wasn't with me, it was with his steel erector that was playing games and not completing the work. Instead the contractor was completing just one item (of the punch list) at a time. The next time I got called back for a reinspection the entire list was completed.

There are some good contractors and some great contractors. One that I encounter on a regular basis usually will give me a call and asks me to have a meeting with all the crews so they know exactly what to expect from me. We review the bolting requirements and the welding requirements. The plumbing-up crew knows that I will be checking their work as well. The projects usually progress with only minor problems.

Then there's always the problem child, the contractor that tries to get away with everything he can. As one said to me, the general contractor, and the owner, "Al, at any one time I have about ten jobs going. You can only be on one and I can afford that. Getting caught is simply a cost of doing business." I'm happy to say the owner fired him on the spot and that contractor has gone out of business long ago.

So, I couldn't agree with you more. A prejob meeting with the workers pays dividends to the owner in the form of getting top quality work that is on schedule and the contractor makes more money because he doesn't have to go back and rework or repair their work. Its done correctly the first time. Every body wins.

Best regards - Al

I appreciate your time to look that up. I don't know where I think I read otherwise. As other posters have noted, it wouldn't make sense to 'bottom out' anyway.