I was assigned to a project for my company, primarily shear-wave ultrasonics and some magnetic particle inspection of structural trusses with partial penetration welds. One of my first questions was, “Where is the acceptance criteria” followed almost immediately by, “What do you mean there isn’t one?“
When I asked how we knew if the welds were good or bad, I was told, “It’ll just look different.” Not a very satisfying answer if you ask me.
Anyone who has ever worked for a company that has been sued knows that from that day onward, all inspection work passes through a “witness stand” mental filter. “Why did you do this, why did you do that.” While this event happened before I joined my first NDE company, Mobile Inspection Service, formerly of Santa Fe Springs, CA, the owner alwaysimpressed this upon his workers. “Never do anything that you wouldn’t be comfortable sitting in front of a jury explaining.”
One day I was looking at a weld with my KrautkrammerUSN-52 and did see something different, 12″ or so of different, so rejected it. When asked how I knew it was bad, all I could say was, “It looks different than everything else I’ve seen.” Turns out that this was only the second time in about 6 months that a partial penetration weld had been rejected ultrasonically. The fact that when tested with a DC yoke, a large fuzzy line appeared on the surface, only helped to reinforce my belief that the weld was rejectable.
I kept picturing myself sitting in front of a jury and being asked, “So how did you determine that the weld was good?’ or worse yet, “Why didn’t you see the flaw in the weld?”, and feeling stupid for not having a really good answer. I called senior people I knew from the local Phoenix Chapter of the American Society for Nondestructive Testing (ASNT) andnobody could point me to an acceptance standard for partial pen welds. I called ASNT and was told that nobody had evercome up with an acceptance standard for partial penetration welds. I’m one of those people that when something bothers me, my mind keeps going over the situation over and over, even if it’s on the back burner. One day the answer occurred to me.
I thought back to some of my first days at MIS, the first day I met Bill Combs the owner. Bill didn’t know me from Adam and I was fresh out of the College of Oceaneering where I completed the Spectech program for diving inspectors. Much of MIS work was on power generation turbines so Bill brought me a thrust bearing, sat me down in front of a USL-38 , wow dit it have a lot of knobs, and told me to inspect the bearing. I calibrated the unit and started…and was really, really confused but said nothing. He’d given me a red china marker to mark and delaminations I found.
Finally after going over the bearing again and again, I called Bill to come check my work, there was a dime sized area I’d marked in the center of the bearing. “What’s this?” he asked pointing to the red circle. “The good part”, he shook his head in disappointment and said he was going to show me what I’d done wrong. After a minute of scanning, he called one of the other Level III inspectors and asked him to come take a look. It turned out that nobody had inspected this bearing in yearsand I had been correct, only a dime sized area was still bonded! I suppose I passed his test.
With this memory in mind, it occurred to me that if I looked at the partial penetration weld as an angular bond test instead, it should be pretty easy to establish an accurate acceptance criteria. Ignore the fact that it was shear wave and treat it as you would a longitudinal beam test. Any relevant indications would appear in the third leg and rather than being at a depth below the surface, would be above within the height of the weld. What were the chances of there being flaws in the parent material at exactly this point below the weld?
I knew what angle the sound was passing through the part at, I had a pretty good idea of what the beam spread would be at a given distance. I knew that the area of interest was between the horizontal parent material and the vertical parent material so after researching what their requirement was for depth of the root pass drew a line at the appropriate angle straight back toward the transducer area I would map on the surface. Anything that reflected energy inside the box was evaluated, anything greater than 50% screen height on one of the boundary lines was rejected, everything else was good!
Given the knowns’, transducer angle, beam spread, material thickness and required weld profile, I made an Excelspreadsheet for all the plate thicknesses we encountered that told me how far out from the face of the horizontal piece I should see my root pass and where I should start seeing the surface of the weld. I drew out these boundaries with my dial micrometers on the horizontal surface and looked at a weld. It seemed to work!
I showed my idea to one of my coworkers who had been an ASNT UT Level III elsewhere and he didn’t see any reason it wouldn’t work. I showed it to the customers Level III who I knew from ASNT previously and he agreed that it would work. I showed it to our corporate Level III and got agreement. Nobody understood why nobody had ever tackled this question this way before, it seemed obvious. It was agreed that this would be our new procedure.
The following morning I was called to inspect a partial penetration weld and with new procedure accepted and in hand, I headed out. I inspected a weld 12′ (144″) in length and rejected all but about 18″ of it for inadequate root penetration. Boy it felt better to be so specific rather than just saying, “It just looks different.” Suddenly the shop foreman was interested in why we’d passed so many other welds and I was now rejecting this one. I explained that for the first time ever, we had anactual written acceptance criteria and demonstrated it to him. I also explained that their Level III had signed off on the process, he arrived shortly, looked at the screen and agreed it was rejectable. The shop foreman was not happy, phone calls were made, people met and the result, several hours later was that we no longer did any UT on partial penetration welds! They would undergo magnetic particle inspection only. We also would not go back and reinspect any previous welds with the new acceptance criteria.
This reduced our workload by about 70% and reduced the companies liability due a poorly defined initial acceptance criteria. I still rejected some full penetration welds and was always relieved when the welders said they had found what I marked at the correct depth and length, but I never had to UT another partial penetration weld ever again.