UT Reject or not

Date 02-27-2005 01:56

I will try my best to address the questions.

Direct answer to the question is yes and no.
Physically seeing an rejectable indications and how small can physically be seen depend on many factors such as:
a) utlrasoic sensitivity level used - equipment calibrated how small imperfections to detect
b) accceptance/rejection criteria - what type of imperfection & how small imperfections are rejectable
c) type of defect - either volumetric defects such as slag, porosity or planar defects such as crack, lack of fusion etc
d) orientation defect -orientated parallel or normal to the excavation
e) excavation method used - grinding, machining or gouging etc.,
f) how much metal removed between each excavation followed by inspection etc.,

I agree the notion that machine does not lie, but it should be noted that some most important parts the ultrasonic examinations are calibration of equipment and interpreatations/evaluation of indications which most cases are performed by ultrasonic inspector.

I hope this helps.
Muthu

By dschlotz (***)

Date 02-27-2005 13:16

Thank you for your reply. This weld was under D1.1 and was a CP—single bevel groove with backup. The thickness varied between .5" and .75" depending on the beam involved. The back up bars were 3/8 by 1". The beams were from 6"—10" flange width and the configuration of the beams was a "T". There were no special rules for the UT other than the general requirement for UT inspection for a building of three stories. The excavation was done with a die grinder and was observed by a CWI that was also a UT technician. The indications were in the direction of the welding. This welding was done with FCAW, E71T-8.
Both the welder and the UT level II were very intent on finding the indicated defect, so the excavation was done very carefully beyond and below the width and depth of the indication. There was nothing found.
My general questions are should you usually be able to find a rejectable defect with careful excavation? In general what is the size of a defect that will be rejectable in the weld I described?

By thirdeye (***)

Date 02-27-2005 16:46

I agree with the issues addressed by Muthu. In addition to what he mentioned, backing bars can be notorious for causing so called “spurious” reflectors. Under some conditions, you may get a reflection from a bad fit-up, the edge of the root, or sound can even enter the backing bar and reflect off of a corner. The trig calculations used to plot location and depth of a reflector (using known angles and sound path distances) must be accurate, and the operator must be sure where the sound is in the part, or the calculation can give an incorrect surface distance (ahead of the index point on the transducer) and depth of a reflector. Most UT technicians use a series of clear overlays which show sound path and surface distances. These overlays, used in conjunction with a full scale cross section detail of the weld joint, help plot locations of reflectors. Here is a sample of a joint sketch and overlay.

http://img.photobucket.com/albums/v377/thirdeye2/welding01/df71624a.jpg

EDIT: That was a sound plot chart, below is a sketch with an overlay:

http://img.photobucket.com/albums/v377/thirdeye2/welding01/2513e75c.jpg

EDIT 2: To put the whole thing in perspective, here is an end view of a 1" thick joint with a 3/8" thick backing bar. The cap been flush ground. I marked the approximate bevel. Note the gaps and the root tie in areas between the plate and the bar.

http://img.photobucket.com/albums/v377/thirdeye2/welding01/0442867e.jpg

There is a section in the commentary of D1.1 that addresses the interpretation issues with respect to backing bars, along with some sketches and techniques used to prove out some reflectors.

That being said, I have been told hundreds of times that the welder did not see anything during the excavation process, but lo and behold, the repair UT came up clean.

~thirdeye~

By g32141

Date 07-19-2005 23:52

A good idea is to mathematically figure out where a defect should come up and gate that area. If nothing comes up in that area or if it is late in time you can disregard it. The problem is that there are not cal blocks for this and they have to be made and be certified.

Gates are important.

If you are scanning moment connections the gate doesn't care if the backing bar is not seated right and the sound goes up to the cap and comes back. If it goes up to the cap as a long wave it will still be outside of your gated area of inspection and should be disregarded. You should not have to dampen the cap with your finger to make sure that you are seing the cap. If you have fusion the reflections after your fusion gate are either the backing plate or weld geometry and they need to be ignored. If you get no reflections in your gated area then the weld is ok. You need to do this for every angle you are working with.

By metalcare (*)

Date 02-27-2005 17:31

With scenario you have outlined, in most cases you could able to see the defect while carefully grinding out, but it may not be the case 100% of the time as third eye pointed out.

How small defects can be seen while excavation? Let us say, you wanted to find 1/32"size (depth of defect), you may need to grind and inspect atleast every 1/32" deepth. In case you grind too much, the defect might get washed away. Even some tight cracks or lack fusion may not be visible under naked eye even by grinding (with coarse grinding wheel; fine grinding wheel might yield better results), but could be seen by magnetic particle or dye check.

My personal experience: When I mark defects by UT, many occasions welders able to see the defects even while gouging and few occasions the welders could not see anything but the defects were removed and confirmed while retest. Hence, it does not mean that when the welder could not see the defect, there is no defect. Anyone or more of the factors which I have outlined in my previous post are not favorable to physicall see the defect.

As third eye pointed out, it is always challenging with UT of welding with backing bar, especially when defects lie closer to the root. Transducer angle and beam exit point must be measured and kept as accurate as possible to exactly locate the indication.

Hopefully this helps in what you are looking for.

Muthu

By dschlotz (***)

Date 02-28-2005 00:07

I don't want to beat this horse to death. I'm still looking for this answer: in size how big is a rejectable defect in a UT examination? Please try to answer from the wiew point of someone without your vast knowledge and experience. Please use the criteria discussed above. This is for personal information and not to win anargument. I'm a welder and I study all aspects of inspection. I'm also a CWI. I get involved in these kinds of questions all of the time. I'm in between welders and management trying to placate both sides and remain credable and keep my integrety at the same time. Without good information I can only go with my gut. I don'think gut will ever replace UT. Help me out here. I have been able to find the indications plotted by the UT tech.

By dlmann

Date 02-28-2005 01:07

dschlotz: I could not take the plunge give a straight forward number (size, how big is a rejectable defect) to a client or inspector because of all the variables that are listed in the above posts. What turns out to be a 1/8 inch round indication that gives a good signal on one weld but not on another weld because of orientation etc…

With that being said, consider the following: the technician calibrates his sensitivity on the IIW block to a 0.060 in diameter hole. Sensitivity is further adjusted according to the thickness of the material. When pressed for a number that is how I present it.

Here’s another thought from some of the things mentioned in the other posts related to your question. It’s a good feeling when you mark a reject and the welder finds it with the grinder. Helps validate that I did it right and get less grief from the welders.

Regards, Donnie Mann

By dlmann

Date 02-28-2005 01:14

Something else came to mind. I once told a welder that he had to grind down 0.125 inch according to what the flaw detector calculated. He reminded me that the grinder was'nt calibrated so he would just grind down about an eighth inch or so.

Regards, Donnie Mann

By jwright650 (*****)

Date 02-28-2005 13:38

I have seen a big ole chunk of slag give about the same indication as a lack of fusion before. The big ole chunck of slag was very easy for the welder with careful gouging to find, however that lack of fusion was near impossible to ever see with a naked eye.

Without going back over all the ground covered already, you have to picture a motorcycle heading straight at a police officer with radar, as opposed to a motorcycle traveling by the officer at a right angle to him. Which motorcyle will give a larger indication?, both are exactly the same size bike/rider yet the one travelling 90 degrees to the policeman will give a better reflector for his radar to bounce back, rather than the narrow profile that the motorcycle shows heading straight at him.

Without giving an UT class and going into this in depth, I'm not sure anyone can answer "definitively" your question. I hope this helps, look at at the IIW block and take note of the .060" side drilled hole, this is the refrence reflector the UT operator is using for the in the laboratory perfect situations, but we all know it isn't this cut and dry as to the "size" as far as reflectors go. AWS D1.1 has us scan at a minimum of 14db (depends on the length of the sound path distance, there is a chart under Table 6.2 for this) above the reference line (that's 14db above the level of sound it took for that side drilled hole to produce a reflection to reach 40% screen height) and evaluate reflectors from there. "Most"(not all) indications that fly off the screen after taking out several db, will be readily seen by the eye when carefully gouging. There is a chart in D1.1 in Section 6 (Table 6.2 is most often used) that we must use to evaluate these indications to see whether they are acceptable or rejectable. If you look at that chart you will soon see that there are several things that are taken into account before determining acceptance or rejectance(note there are formulas for figuring attenuation and final indication ratings and all the columns on the chart the inspected part may fall into due to thicknesses and angles and such). I said all that to say this, there is no real cut and dry answer to the size, because of all these variables that must be taken into account on each and every indication(they are like snow flakes in that no two are alike).
John Wright

By thcqci

Date 02-28-2005 15:49

I would say most discontinuities that are rejected, should be able to be seen when excavating, depending upon the process used to remove metal (grinder, die grinder, arc gouging) and the skill and desire of the excavator. If the location is accurately marked and care is taken when excavating, then it will usually be found and completely removed the first time. In my experience, this is most often the case. If you have confidence in the UT inspector, it is possible that the welder will just gouge/grind away to remove metal around the area and reweld it. It would be better if the discontinuity was seen and removed, but not always necessary based upon time permitted to make the repair.

Confidence in the UT inspector grows due to a rejected discontinuity being found when a welder excavates for it. Some welders do not want to find the rejected discontinuity. They would rather blame the UT inspector for being too critical and just wanting to reject stuff. Got one of those here. You know, his stuff doesn't stink. You can't change that. A conscientious welder will want to do better. Here I have a pretty good reputation since most welders find most marked discontinuities.

As has been discussed above, discontinuities can be different sizes and be acceptable or rejectable based upon the characteristic of the discontinuity. Shorter discontinuities can be larger in size. Longer ones must present a smaller sound return. Location in the weld can be cause for non-acceptability. A grouping of porosity may be rejectable where single pores of the same size found in the cluster may be accepted. Something may be rejectable in a "tension" weld that may be acceptable otherwise. Something may be acceptable in a thicker section that is rejectable in a thinner section. In other words, lots of variables. Have your UT inspector take some time to go over the acceptance-rejection criteria to explain how it works. Have him demonstrate setup and calibration. Have him explain the acceptance/rejection criteria. Have him explain his findings after a UT inspection when rejectable discontinuities are found. It will be time well spent.

Hope this helps.

By thirdeye (***)

Date 02-28-2005 17:04

In short, I think what everyone is saying is: the indication on the screen is based on the reflectability of a discontinuity. That is why calibration blocks or mock ups are so important. It is possible to change probe sizes, MHz and angles etc. to help in grading or sizing.

I have compared discontinuities on x-ray film to UT displays of the same item and have seen the accept / reject criteria go both ways, that is, rejectable using UT and acceptable using RT and vice versa. Here is a photo of an indication in a 1 1/4" single vee grove procedure qualification test which would be graded as an "A" indication (UT D1.1) and rejectable. The area of the joint was inspected with RT and the indication was acceptable.

http://img.photobucket.com/albums/v377/thirdeye2/welding01/be7e53cc.jpg

~thirdeye~

By dschlotz (***)

Date 03-01-2005 02:28

I would like to thank all of you. I have come to know a lot about some of you by the answers you have taken the time to give in this forum. I live in an area where there is no input for most inspection questions. Being able to post quiestions and get answers, almost instantainiously is more than I ever expected.

I used to call AWS and talk to their engineers and try to get answers. They were very noncommital and I supposed were worried about being quoted,or sued.

I need to understand the parts of the inspection field that impact me and my employer. It's my job to to see that we comply with the codes and specifications, and to QC my welders and protect my employer and our customers all at the same time.

I have had numerous questions answered on this forum. They have helped me to perform an informed service.

Thanks Dennis

By KeithCWI Date 03-11-2005 03:50

The best way i have found to clean out a defect is with an Arc Gouger. If your welding SS this will not be a good method. For carbon steel this is what i use. Just gouge out a little at a time. I have found some very small defects this way. Plasma gouging will work well also if your gouging SS. For plasma gouging on SS i use a Argon/Hydrogen Mix of 65/35 for my plasma gas and 100% argon for my secondary gas. Just make sure no one has a problem using these processes on the project.

By BankerQC (*)

Date 05-02-2005 19:54

Well it looks like this discussion is pretty well over but I thought this might help add some fuel to the fire.The question about minimum size detectable defect or "indication" has been addressed and all of the answers presented are indeed factors. I think though that what the original question was seeking was the smallest detectable size of indication. I could be wrong but it seems that I recall it being something on the order of one half of the wavelength produced by the calibrated setup of the flaw detector taking into account the frequency, transducer dia. , materials sound velocity etc.

By jwright650 (*****)

Date 05-03-2005 11:42

Good Morning BankerQC,
I'm not disagreeing with you but AWS D1.1:2004 nails you down as a UT operator (without the Level III writing special UT procedures) as far as the equipment allowed.
for example...
6.22.7.1 Frequency. The transducer frequency "SHALL BE" between 2 and 2.25MHz, inclusive.
6.22.7.2 Transducer dimensions. The transducer crystal "SHALL BE" square or rectangular in shape and may vary from 5/8" to 1" in width and from 5/8" to 13/16" in height. The maximum width to height ratio "SHALL BE" 1.2 to 1.0, and the minimum width to height ratio "SHALL BE" 1.0 to 1.0.
6.22.7.3 Angles. The search unit shall produce a sound beam in the material tested within a +/- 2 degrees of "ONE" of the following proper angles: 70,60, or 45 degrees as described in 6.29.2.2.

We are basically told what to use rather than allowing us to use something more appropriate for the situation. The Level III that we were using here brought some sample plates with built in flaws for me to evaluate. Some of these were too near to the surface of this relatively thin plate for good resolution with the AWS mandated equipment. There are other sizes and frequencies and angles that would have worked much better, but we are locked in.

So I'm thinking those variables are not relavant to the original question, however if I'm off base, please correct me as I am still learning as I go.
John Wright

By BankerQC (*)

Date 05-10-2005 18:25

Exactly ! I was trying to get that point across and adress the basic question of how small a defect can a UT examination discern. guess I didnt do a very good job...always happens when we get in a hurry..LOL !