Plug Weld Question

[azcats]

We've run situation where I'd like some opinions. In a detail like the attached, the height of the plug weld slots are sized per Sec J2.3a which specs a min height of the thickness of the outer pipe plus 5/16". So for a 1/2" pipe those slots should be 13/16" minimum. A thicker wall pipe (5/8" instead of 1/2") was substituted and the slots were not adjusted accordingly.

The field welding inspector noticed the discrepancy and wrote an NCR. The welds were completed and inspected with the smaller slots.

Clearly the welds don't meet the specification. The commentary essentially says that the minimum sizes are to provide geometry that is conducive to good fusion and to avoid deep, narrow holes.

How much (if any) of these plug welds is usable? Should they be disregarded completely? Should the effective width be reduced to 9/16" (1/8" subtracted from each side)? Should the welds be UT and if that passes call them good? Should I call them good because they were inspected during installation and there were no issues with access?

Any opinions are welcome.

 

[qcrobert]

The field inpector has filed an NCR which should be brought to the attention of the Owner/Client at this point who will decide the disposition of the NCR.

 

[dhengr]

Why would the field weld inspector write a National Conservation Review (NCR?), or maybe a National Cash Register Corp.?, on a weld?

You must understand the reasoning and logic behind the slot width vs. plate thickness issue, for a plug or slot weld. Then, the exact value, (plate thickness + 5/16") is somewhat arbitrary and just assures proper weld fusion at the root pass on the two sides of the slot. (plate thickness + 3/16") might work too, if the welder took the right procedures for the root pass. The welding rod or wire stick-out and welding head must reach down into the root at 45° for the first welding pass. As you move up in the welding passes less slot width is needed to accomplish this. If everything else on that detail is designed properly, the slot welds are probably not too highly stressed. The most difficult part of that detail is perfectly matching the slot locations in the outer pipe with the thin inner plate, and getting a good root condition for that weld.

I am not trying to overrule AISC or AWS, but there are other ways to skin that cat. They just don’t try to teach welders to weld or engineers to design welds, in the codes. They finally come up with a number (5/16ths) which should be O.K. in most situations.

 

[Ron]

dhengr makes some good points. Capping a plug weld with reinforcement is exactly as dhengr noted...to compensate for potential root discontinuities.

Another point to consider is that you can still go in and add weld to a plug weld. I would clean the existing weld so that it is free of slag and any other contaminant, then pre-heat the weld area, add the weld and inspect.

UT would not typically offer any help and you would probably find either confusing or irrelevant indicators that would open another can of worms. Not that you shouldn't find weld defects, its just that UT of a plug weld on pipe is difficult to interpret and could give misleading results. If you want to do nondestructive testing, use RT.

 

[rscassar]

I believe the acronym NCR in some construction industries refers to a Non-Conformance Report.