Not being a UT tech...why would you NOT check the corners with UT? Is that because the radius limits contact with the transducer?
If a tube were to break in the weld joint, likely it would happen in the corners where the backing bar is interupted.
My understanding is that the backing is to be continuous, or full pen spliced, before fitting it in yhjr tube. In order to comply with that, we burn out a 1" or thicker plate in the shape of a squared donut to fit the inside profile (QCCWI mentioned the same thing). When the customer is OK with it, we leave the "donut hole" in the plate and use that as is. In doing this, there is never any question as to whether or not the corners should be welded, or if the backing has been CJP spliced beforehand.
Just to clarify my comments above, I don't tell our welders that I don't UT the corners either. I suppose those that are observant enough probably know that although no one has ever voiced that observance to me. Nothing I stated above says we don't make a competent effort to deposit the required CJP in the corners, but only means I don't have an effective, economical way to verify it; nor have I ever been challenged to do so. I have watched some welds being deposited and they have always appeared tied into the backing very well.
The reason I have never UTed the corners is because you can't couple the transducer well on the radius. And with all the different radii for differnet thicknesses of tube, it would require a wide selection of wedges. Nor are standards available even if you could make the wedge radius to match. The UT IIIs I have worked for also never required that area to be UTed so I have never pursued it.
I have never really considered the argument that backing strips need to be CJP (no breaks allowed) since those "joints" go into the corners that are not UTed. I will have to think about that in the future. Thanks. We have placed 4 pieces of backing before, but usually it is only 1 or sometimes 2 pieces. We have also used a solid plate as stated above. This has been the method most often adopted recently so I will try to stick with it as it is the method I prefer.
we use flat bar bent to the id of the ts and then make a cjp splice to make it continuous
Unless time is a real consideration, why not cut a 1" wide piece of the same size tube steel, then cut in half so as to form two ( ) sections and grind the mating surfaces of the backing to fit inside. Although technically it isnt continious, it would be better than leaving the corner radius with no backing. (A possible WPS violation). This doesnt solve the UT problem presented by the geometry of the corners but it would solve the backing problem. You might solve that by using a plexiglass shoe shaped for the radius. With a flat shoe it seems like you would have problems maintaining contact and knowing exactly where your sound is.
if you use the 1" piece of tube cut in half and not cjp welded, then you are violating the code.
You would have to cut the square or rectangular HSS into four pieces not just two to get it to fit inside amd I don't think it would fit tight in the radius anyway because the outside radius is larger than the inside radius by the thickness of the material.
For structural pipe columns, we have used a section of same sized pipe and cut a slit in the 1" section of pipe and squeezed it down to fit inside and full pen the joint. That usually works pretty good.
For us, we find it easier to just form the 1/4 flatbar and full pen the one joint in the backing before inserting it into the joint to be welded. Don't forget to grind the welded seam flush before inserting whatever backing material you use.
Not trying to be disrespectful, but this is how I see it. There are different radii for different wall thicknesses of tubing further complicated by different radii from different manufacturers for the same given thickness. This radius is generally twice the wall thickness, but not exactly. Ever try to splice 2 pieces and see the possible differences?
So if you have a shoe that adequately fits a particular radius, how do you calibrate with that wedge? Do you know of some special IIW type block that you can use for calibrations? And even if the radiused wedge fits the contour pretty well, you should "twist" the transducer to get different angles of sound transmission into the weld as you scan forward and backwards. Can not do that if the wedge fits the radius. Nor can you now move the wedge up onto the tangent are between the flat side and the radius without losing coupling so there is still a potential gap in the inspection area. See the problems? Unless there is someone (client, SER, UT III, etc.) who thinks strongly that these radii need to be UTed and can present a solution including the procedure for implementation, not worth the effort.
Don't wish to beat a dead horse, but that is my $0.02 worth.
Doug,
I feel halfway confident about the corners. Like I had stated earlier, my guys have tried to fool me by placing the splice up in a corner, thinking that I would not find it....they were wrong and had to repair it. You have enough flat surface to sneak up pretty close to the radius and aim sound perpendicular to the long axis of your material and the sound will go around that radius. Then you can wiggle the transducer to get several looks at that corner, then you can couple up on the adjacent face of the tubing and aim back at the joint from that direction. There may be some things hiding in that corner that you might miss...but I've found indications and went after them with the gouger until those indications were gone. Can you apply the rejection chart in D1.1 effectively...probably not, but it is better than not checking the corners at all. Even throw in some MT of those corners if you need some reassurance of something near the surface. It can be done...by code, not really without a Level III's instruction/procedure. So with all that said, I would not go into another fabricator's shop rejecting the corners without being very sure of myself, but at my own shop...I'm checking those pesky corners the best I can and digging out anything that I may find.
John I sent you an e-mail
OK Chris, I'll watch out for it.
Chris,
I haven't seen anything as of yet...
Doug,
The requirements of D1.1 5.10.2 are clear - "steel backing shall be made continuous for the full length of the weld". The length of the weld in a butt or tee-joint in square tubulars includes the corners. Check out the Commentary also (C5.10.2). One technique that I have seen that seems to work fairly well is to cut a piece of 1" plate to fit the shape of the ID of the square tubular, tack it in place on one member, and fit it to the other (this would obviously not work if RT were required, though I have never personally seen RT specified for square tubular butts). The code doesn't specify a minimum width for the backing, and it certainly meets the minimum thickness requirements of 5.10.3.
Regarding the UT, it is indeed difficult. An additional consideration is that length sizing of reflectors is difficult (though it can be done) in these situations. The usual technique (using 6dB drop) of measuring the distance between the centerpoints of the ducer at the two extremities at the steel surface will indicate a length that is greater than actual due to the geometric configuration of the joints. The same thing can happen on heavy tubular girth welds.
Regards,
Mankenberg
as a side note kipman, 5.10.3 has only one requirement that is, that the backing shall be of sufficient thickness to prevent melt-through. the thicknesses listed are only recommendations.
So I have learned from the above comments that I have not been competently UTing the tubes by ignoring the corners. I will be more diligent in trying to include them when it comes up again. Also, even though we have sometimes used the backing strip and sometimes used the solid plate technique for backing, I will solidly fall on the side of ¾" or 1" plates in the future. I am aware of the CJP requirement for backing, but did not put 2+2 together in this application. My mistake and thanks to you guys for correcting me. Guys in the shop are sure gonna cry when I throw that blasted code book at them again.
There are several companies out there that make "exact fit" backing rings for round, square, and round tubing. The only one I can recall is Robvon Backing Rings. www.robvon.com.
These come with consumable spacer pins attached. Not sure as to the price part, but, if you can cut your fitup time and the time spent tying in the corners, then this would be money well spent, both from the fabricators prospective and the amount of time the UT guy spends looking at the corners. I would be interested in hearing if anyone has used these inserts before?
Wayne,
I tried to use those backing rings, they are only 1/8" thick. I called them and talked with them about the thickness and they will not do a 1/4" backing ring. Our 3/32" FCAW did not like 1/8" backing rings.
if you have access to an asme (mini) type transducer and shoe, they have a smaller contact surface that will allow a higher percentage of the shoe to be in contact with the base metal. it allows for better volumetric coverage of weld at corners
True...but D1.1 tells you what size transducer to use.
straightbeam transducer 6.22.6
angle beam transducer 6.22.7.1, 6.22.7.2, 6.22.7.3
John it appears the e-mail address listed in the bio- section isn't correct. I used that as a link to e-mail you through. Chris
Chris I just checked my profile and that is the correct address... "jwright650@aol.com" ....without the quote marks
I'm going out of town to a NACE training class so I won't be able to look in on the forum this coming week.
after rereading this thread i'm wondering if AWS has ever addressed this issue? for as long as i've been inspecting, the ut of tube steel corners, has fallen into two categories. first as some have stated they "just don't do it" or second " it's a best guess". codes and specs require the 100% ut inspection of certain welds. why has the issue of radius corners never been addressed? on the ut report is aws listed as the procedure and evaluation spec? regardless of whether you add a comment that no ut of corners was performed can it be listed as performed to aws and evaluated to aws? also if you give it your best effort, can it be listed as complying to aws? has anyone used annex s to prepare a procedure for inspection and an evaluation procedure, and have shoes made. all of this is required for tky connections. why not tube steel connections? i've heard all of the excuses like it just can't be done, who is going to pay for it, this way is the industry standard. i also believe that when ut reports are reviewed that the people looking at them don't have very much experience in ut, mostly looking at pass or fail and code it was performed to, not putting the additional comments about corners in correct context. i was just wondering why everyone seems so nonchalent about this. when everything else is by the book
jwright - it was already stated in this thread that the inspection was not to code. i was making a suggestion based on this. allowing for better coverage.
Gentlemen, I would like to thank all of you for your time and input on this matter.
Clearly what we have done on occasion in the past was not the correct way to accomplish this joint configuration, I am currently working on getting myself and everyone in the shop up to speed with what is needed and expected when and if this comes up again.
Like someone mentioned earlier in this post, I didn't put the pieces together and see the big picture about how the splices of the backing bars are expected to be as stated in the text of the code when it comes to square or rectangular shapes. It's interesting ( and scary sometimes) how many times you can read something in the code, and not completely picture every application it can be used for, then when it's explained in a forum like this, the light goes on and it's an "Oh Sh!t" moment when you realize you made ( or allowed to happen) a mistake. Have a Good Day, Chris
I agree with thcqci. I have never ut tested the corners of tubing by scanning over the corner. If you wanted to, you could produce a procedure using miniture transducers, I don't think it neccessary.
In all of the tubes I have tested, I fill confidant that I had a good inspection of the corners by rotating the tx into the corners. No- it's not 100%, but it gives you an indication if there is a gross reflecter there.