Hello NDT III,
It is a very good question but I think it is a mistake in the code saying the inspection method is radiography only.
I think it should be radiography and/or visual inspection.
The acceptance criteria for H is less than or equal to 1mm (1/32"), how can you possibly tell the exact depth with radiography. It would have to be measured manually (visually). I presume if the density is darker in the undercut area than the parent metal it would be reason to then visually inspect and measure.
Hope that makes sense,
Regards,
Shane
I've wondered the same thing, and have assumed it's something along the lines of Shanes post. I assume if the image shows in the film, then a physical measure is called for.
" A check mark indicates examination method 'generally' used for evaluating this kind of weld imperfection."
It doesn't say 'the' method, it says 'generally'. so there is room to utilize other methods namely visual. I believe Shane is spot on to intent.
generally, may, should, can, all squirm words. Shall, will, must, those words are literal must do's.
In order to measure it, there has to be some length to it to get the gauge into it to start with. At 1/32" the head of the typical gage will be at least that wide.
Therefore if it's greater, it's out. If you can't measure it, it's in.
As for measuring it with RT. I know of no method that can do that accurately and economically at the same time. I seriously doubt it was the intent of the committee to profile them with RT, and since they've left the door open for other methods, that what I use.
My two cents worth,
Gerald
I agree with both of you. I do believe visual examination was inadvertently deleted when they updated the code to the 2004 Edition. The 2004 Edition does not call out Visual for Undercutting. I know previous Editions did. However, you can't assume intent when using codes without an interpretation or a code case from the code commitee to back it up? Otherwise you could get yourself into trouble.
What if you have root undercut? How would you measure that? You can't so you have to rely on RT. I would assume that if you can see it on the film it is probably greater than 1/64", but that is only an assumption. I guess for now it's just a judgement call. But if you as a client rep sees root unercut, how do you justify making the contractor repair it without a backcharge? What basis would you have? If I were a contractor and you rejected undercut by RT that was let's say 1/4" long on the root, I would say "prove it or pay for it".
The same thing goes for Burn Through. It is not addressed in B31.3 so you would have to evaluate it as Root Concavity. That's clear. By the way, Gerald, it does say Examination Method and not General.
I think it's time for a new Code Interpetation to let the code committee explain themselves. I think I will send one in unless someone else already has.
NDT III,
I only have a 2002 copy of B31.3,
Under General Notes: (d) it states that a tick designates examination method generally used for evaluating this kind of weld imperfection.
Under General Notes: (e) it states that a dotted line designates examination method not generally used for evaluating this kind of weld imperfection.
I did not read this when I initially responded, is it the same in the 2004 version or is it changed ?
With regard to root undercut it comes down to an intepreters knowledge and experience.
External undercut will show up on the radiograph darker than the parent metal (wall thickness) because it has lessened the wall thickness.
Internal (root) undercut will generally not show up as darker than the parent metal (wall thickness) because it is compensated by the thickness of the weld cap. (Same theory that is used for root concavity) However if it shows up as darker than the parent metal you can guarantee that it is deeper than 1 mm (1/32") and there is not a lot the contractor could say in arguement.
Hope that makes sense,
Regards,
Shane
Thanks Shane,
Makes perfect sense and I agree wholeheartedly. I agree the density wuld be compensated for on root undercut be the cap, however you also cannot take credit for reinforcment. That is excess metal. You're right about the General Notes. I missed that. However, it does say thay Visual is not Generally used for undercutting. I think the Code Committee should take responsibility and make a decision about undercutting. The subject is very unclear in the code.
I also agree that it is a judgement call and is not usually argued over, but could be if it came down to it. As I said, the interpretations say you cannot use density as a basis for determining depth of undercutting.
If you happen to have very deep undercut in the root nd your cap is not very high, you know the density will be higher. On what basis can you reject it if it is only 1/4" long, as an example?
It is a subject the code needs to address. That is all I am really trying to say here, but thanks for your input.