WM_Su2014_20

Preheat yesterday, post-weld heat treatment. If you go back over here to this Venn diagram, you have to have material, stress, in an environment for hydrogen cracking. If you do a post-weld heat treatment, typically for steels it's putting it in a furnace at 1100 to 1200 degrees Fahrenheit, or building a furnace around it, one hour per inch of thickness. So if you're talking a four inch thick pressure vessel, it's going to be in there for, once you get it up to temperature as measured on the outside, you're going to hold it there for four hours if it's four inches thick. That's the typical stress relief time. They have more sophisticated tables, but if you go through the thermal conductivity of the steel in order to get it heated all the way through, that's typically what you would do. And that's the rule of thumb: one hour per inch of thickness at temperature, not starting in the furnace, not starting with the cold furnace, but once you get to temperature, now you start your time.

Uh, you do three things. You basically temper the steel to lower the hardness. So you have this heat affected zone that cooled quickly, high hardenability, so you have a wide hard zone. And if you go up above a thousand degrees — actually go above 300 degrees you're going to temper, but above a thousand degrees you temper and you get a nice soft heat affected zone. Even if it had been Rockwell C55, you'll temper it down to, right after a post-weld heat treatment you'll probably be down Rockwell C30 or something like that. And so that lowers your hardness to a level that is not likely to give you hydrogen cracking. You also relieve the residual stresses — sometimes this is called a stress relief heat treatment, so far as that goes. And you also remove hydrogen.

Now you shouldn't wait too long to do your stress relief heat — well, you can wait as long as you want for stress relief heat treatment, but if you're going to get rid of the hydrogen you can't wait too long. Although if you got to build a furnace around the whole thing, it takes a little while to build the furnace around it after you weld it, and so it might be days so far as that goes. Which is why you need preheat, to try to get rid of some of the hydrogen and stuff. But if you really want to, if you have a very high hardness steel like a 4340 or HY180 — actually HY180 is not a good example, it's a cobalt steel — but anyway, if you have a 4340, a very high strength steel, for example, if you take a, if you go to a well-known helicopter manufacturer in Fort Worth, Texas, okay, which actually does — there are civilian helicopters in Mirabel, Quebec — okay, so it's actually Bell Helicopter folks. And you go, I've actually been there in Fort Worth where they're building the military helicopters. Can't build military helicopters in another country, at least not for the U.S. Department of Defense, so they still build in Fort Worth the military helicopters.

When they — it's not always welding that introduces the hydrogen. I mentioned you can electroplate, the electroplating process can introduce the hydrogen. Well, they actually go from the electroplating bath directly into the furnace that's going to bake the hydrogen out at 375 degrees for 23 hours, okay. It's less than five minutes okay, they don't wait a full hour, because now we're talking about a mast material made out of 4340 steel and a strength level of 240 KSI or something like that. And they are worried enough, because if you lose your mast — you only got one mast on a helicopter and it's not a good day if you lose the rotor. You've kind of dropped like a rock. Can't auto-rotate to ground. So anyway, post-weld heat treatment is basically a high temperature, very expensive process, but necessary if you have very thick steels because of the residual stresses, and also if you have high hardenability steels, okay.

What else do I say? Oh, I did want to say something about joint efficiency and over-matching and under-matching filler metal. People have realized for years that the problem is that, if you have a weld, in general we over-match the weld metal to the — by about, well, it's supposed to be zero to twenty percent, but it's typically ten to twenty percent okay, over-matched, where the weld metal is stronger. And so if this is HY80, we'd be using a stick electrode that would be an E11018. This is years ago, we use gas metal arc now, but we'd be using an E110S. Which 110 means 110 KSI, HY80 is 80 KSI, and you're a little bit more. They might have used a 118 in some cases. If you're doing an HY100, you'd use a 12018 or an E120S electrode, so you'd be nominally twenty percent higher strength in the weld metal. So if you pull on this, it's going to neck down and it's going to fail in the base material.

Well, the problem could be relieved if you could use an under-matching filler metal. If you could use a 7018 electrode, you can't build up residual stress greater than the strength of the weld metal, right, because the weld metal is going to yield and your maximal residual stresses will not be 110 KSI, they'd be 70 KSI, which would be the strength of the under-matching filler metal. And in fact, I told you that's what they did in the early days of fabricating some of the foundations for the Nautilus. They had to go to austenitic filler metal which had half the strength of the HY80 and would absorb a lot of the hydrogen. But they were able to do it until they developed procedures that would not crack with the hydrogen and the higher residual stresses that they were encountering.

So under-matching has been a holy grail and people have tried to do it. The Navy has spent on several different occasions millions of dollars doing, making great big foundations and testing folks near full-scale type things. When I say near full scale, something the size of a small car okay, and doing mechanical tests and seeing how it fractures. Because remember, in the submarine business you got to set off an explosion next to it, because that's what's going to happen if someone hits you with a depth charge, okay. So they've tried it and they always have found that it doesn't work. And the reason it doesn't work has nothing to do with welding, it has to do with fracture mechanics. Just think about it: if I had two very strong things surrounded by, or held together by, something very soft, and I now deform this in an underwater explosion or anything else, just in the laboratory I'm going to concentrate all my strain in the weld metal. These other two things are like grips and they don't deform at all. If you're severely under-matching, so if you're under-matching by more than five percent, you're going to find that you will fail. You'll concentrate all your strain in the weld metal, and if you've got fifty times as much base metal as weld metal, you now have fifty times as much strain concentrated in the weld metal, and you're just going to break the weld metal. And that's what happens.

So the fracture properties when you start plastically deforming in an explosion are lousy. People have tried this in civilian business because they can save money on stress relief and things, and they've proved that they don't like the impact properties. When you concentrate all your strain, all of a sudden your impact properties fall apart. And so they did a study and spent millions of dollars in the 60s, and then they started about twenty-five years later and did millions of dollars again, and they concluded the same thing again. So about every twenty-five years you get a new generation of engineers that someone has a bright idea, oh, we can use under-matching filler metal, okay. You can't, forget it. Don't go spend millions of dollars trying to use under-matching filler metal, okay. It's been tried, doesn't work. Okay. Any questions?

We haven't finished, but I was going to take some of this time and start giving you some examples of sort of classical failures of some of these things. And I, we've already sort of gone through, not all at once, but I've talked about the Seawolf. And the Seawolf problem was they ended up having a very high — I can actually just put this up — and the Seawolf, they end up having a very high material strength in the weld metal, too high. It was actually over-matching by about thirty or forty percent, and that's too high. And you ended up with not necessarily an inherent problem in the weld metal, but you ended up with higher stresses, and higher hardenability, and they didn't have enough preheat for this higher hardenability to get rid of the hydrogen, or to get the hydrogen down to a low enough level.

Um, I don't know that I want to keep going through the Seawolf, but there are some other things to show that this still occurs on a fairly regular basis. So one of the stories which I was involved in the early 1990s is Pacific Gas and Electric decided to build a pumped water storage project, because the nuclear reactors like to run flat out. They don't like to cycle up and down, high speed, low speed, okay, power. They like to just run continuously twenty-four hours a day, nice and steady. But that's not the way people use their electricity, okay. They use electricity — the peak electricity time is between five and seven PM, when people get home from work, it's still hot in the day, they want their air conditioner running during the summer.

And so they decided they could go to the High Sierras, and they had a lake up at about 8,000 feet and they had one about 6,000 feet. And they decided all they had to do was build a pipe between the two, and during the night the nuclear reactor could pump the water up to the upper lake, and during the day they could let it run down this pipe and they could run it through turbines. They'd have a hydroelectric plant. So it's called a pumped water storage project, and they've built a number of these around the world but this was one of the larger ones. They had a 22-foot diameter pipe, big pipe okay, and it was miles long. Not very round pipe, maybe that's why it failed, anyway. 22 feet in diameter. It was 188 PSI water at the failure location, which is up near the upper lake okay. That's just a static pressure head okay, that they had at this particular location.

And they had a failure. Almost everybody left the construction site, and some guys were just left there to clean up. And they had some little porta-potties, and one of them had gone into the porta-potty — was nearly five o'clock in the day — and he comes out and he hears this crack, okay, as he's exiting the door. And he looks about fifty yards from him, he sees water spraying out of this, out of the manhole that had a man-way, just a little flange with a cap bolted on here. And this was a man-way if they ever had to get in here. This was a way for someone to get in. You empty the whole pipe and someone could crawl in here. This was a little crossing. It was 180 feet across the canyon. Everything else was buried in the rock. They just bored a tunnel in the rock, put the pipe in there, grouted it with cement on the outside, and so it was supported everywhere else. This was actually just an open 180-foot — this was a 20-foot section, and they had three, maybe had four 40-foot sections, okay, going across this canyon. And they had — in the little one they had this little man-way down about the five-thirty position or so.

And the guy saw spray of water, 188 PSI water, coming out of this thing. He started heading for high ground and he made it to high ground. Nobody got hurt here. There was one other guy and they got in their pickup truck to go up to the upper lake to shut the 22-foot diameter valve. Now, turns out 22-foot diameter valves are not the easiest thing in the world to move, and this one hadn't really been exercised a whole lot even though it had probably been put in place a year before during construction. It took them a half an hour to close it. In the meantime, when this thing completely broke apart in a brittle fracture, you had 4 million horsepower heading up against the canyon wall. That 4 million horsepower of water ate away a hundred feet of the canyon wall. It was no longer 180 foot wide canyon, it was a 280 foot wide canyon when this was all done. The upper lake dropped by about two-thirds in volume down to the lower lake. Some of it of course came out down the ravine that went beside the mountain, and it carried all the stuff. I mean, a number of bears and rabbits and raccoons and others were running for their lives, and some of them probably died and drowned in the flood. Built about a 300-foot long sand bar at the other lake, just wiping out the ravine.

Well, it also did 180 million dollars worth of damage or something like that. And so, as a public utility will typically do, they go to the rate commission in California, they say oh, we had an accident, we'd like to put it on the rate base, we're going to spread this like peanut butter over all the rate payers in California. And the Public Utilities Commission said oh no, you can't do that unless you sue American Bridge of U.S. Steel, because your expert said this was a manufacturing defect. And American Bridge was the guys that manufactured this. So it turns out that it was one of the world's largest engineering consulting firms in California — the guy was one of the founders of this company — was out there three days later. He gives a press release and said it was American Bridge's fault, okay. Thing goes on and they now have to go to trial. I mean, they were told by the Public Utilities, if you don't go to trial we're not going to let you put the money in the rate base. And it turns out it wasn't the — the actual rebuilding was like 180 million or whatever — the actual loss was 800 million, okay, in terms of lost power and production and everything else. And they wanted to put all that in the rate base, so it had to go to trial.

They already had two metallurgists and they hired me as a third metallurgist. And I said what's going on here? And they said well, the judge has said he can't have more than two, we're just going to let you talk about the welding. We want you to talk about how did this crack occur, because there was — if you looked in here at the fillet weld that held this man-way to the pipe, the little fillet weld right here, there was what looked to me like a hydrogen-induced crack. It was all oxidized — actually the crack was, it was hydrogen-induced but it was oxidized. And you can see from the stuff — well, it turns out, as people will do according to the boiler and pressure vessel code, this was an inch and a quarter thick material, and an inch and a quarter thick material with a fillet weld does not have to be stress relieved. If it had been an inch and a half thick, the code says you had to stress relieve it. If it's an inch and a quarter, under certain exceptions in a certain paragraph of the code, you don't have to stress relieve it. Well, it costs a lot of money when you're 6,000 feet in the Sierras to take something that's 22 feet in diameter and stress relieve it, so they decided we won't stress relieve it.

And so I had to look and figure out why this thing failed. Well, there's my stress — they didn't stress relieve it, okay, no post-weld heat treatment. I had a material that was susceptible but not terribly susceptible, and they had welding procedures that should have prevented the hydrogen getting in there. But in fact, when they're building this in the High Sierras — and you can't just truck up 22-foot diameter pipes okay, they don't fit on the trucks or the roads — so they had to fabricate it on site. And they had a big camp, okay, and bunkhouses and stuff where these people would work. And it was getting to be close to Thanksgiving and they were going to shut down after Thanksgiving, because the snows were — it started snowing in October, beginning of October okay. And so these people were sort of working in the snow up at 6,000 feet in the Sierras. Um, actually it's not Sierras, it's the Sierra Nevada. Sierra Sierra Nevada. I was corrected many times for making it plural. Anyway.

So they were welding on this thing and they had supervisors and control people, except all those people decided they wanted to go home early for Thanksgiving. And so far as we can tell, this was welded on the Tuesday or Wednesday before Thanksgiving. We have a progress picture showing six inches of snow on the inside, okay, on the day that they were supposedly welding. They had to preheat, and they probably did preheat the outside, but how well do you think that lasted if you had six inches of snow an inch and a quarter away? I mean, so they probably got their torch and they used their little template stick to see, oh, we got the outside to 150 degrees, which is what their preheat temperature was, but that probably lasted all of about four or five minutes once you took the torch away, okay. So it's not too surprising they basically were welding without preheat, even though they probably were doing it. There was no supervisor to tell them, hey, go shovel the snow off the inside of the pipe okay. The pipe was lying horizontal and everything. You know, you can see in the picture, snow on the inside of the pipe.

So, the interesting thing was, they had just pressurized the pipe up nine months later in the spring. The snows had melted, it was spring, they're about to start this project up. And when the — the other guys were up there cleaning up and most of the people had left again, and the crack didn't occur for 18 hours. The delayed cracking. And people said, well — turns out the guy who three days afterwards had said it was not hydrogen cracking, he said because he knew hydrogen cracking always occurs in the first couple of weeks. That's true if you don't refrigerate it, okay. You can trap the hydrogen in there, and if it's frozen for eight and a half months of the year you can still get a hydrogen crack. And so it turns out it had been pressurized 18 hours before at 188 PSI. They had all the pressure logs that had — there was no wave of stress going through there. It just let go 18 hours later. So I said it was a hydrogen crack, but I wasn't allowed to say it was a hydrogen crack because this other guy, who, you know, his wife drove around in a Rolls-Royce — and a big Rolls-Royce, not a small one — he had said three days afterwards.

So we go to trial. And I'm the last expert at trial, and I talked about the lousy welding procedures and the lack of process, you know, controls and supervision and stuff. That's fine. Well, we end up losing the case. U.S. Bridges' — uh, American Bridge's defense was, the foundation wasn't steady enough and it had cracked 18 hours later because the foundation had slipped. I thought that was a stupid thing, but nonetheless the jury found in their favor. We've done like the Public Utility Commission said — we take them to trial. The interesting thing was, because this guy from the big consulting firm had blamed American Bridge in the press three days after the event, not only did we not collect the 100 million dollars they were asking for, they had to pay 17 million punitive for defaming American Bridge, okay. So now we did — we lose 100 million, not get 100 million — we lost 17 million. But they had now done what the Public Utilities Commission said, so the California rate payers got to pay for all of that, right, in this in the world. Great.

Um, the more — actually there's another interesting thing on, I might as well put this on YouTube. My bill at the time, this is early 90s, for the previous two months and several trips out to California, everything, was 55,000. Largest bill I had ever sent anybody for, by a factor of two and a half. I thought I'd take forever to get paid, I got paid in less than three weeks, okay. That's because the big consulting firm, their bill for the same period was three million dollars, okay. And it turns out the utility litigated against them because they had so many people working on this thing. They were having the younger lower-priced people bill at the rate as if they were the higher-priced person, and those higher-priced people were being — putting in for 120 hours a week of their time, and some accountant noticed this and said, this doesn't make sense. So anyway, that was one of the reasons I got paid quickly, because the other side — work — anyway, I shouldn't say what they were, because this could go public, right okay.

So we've talked about Seawolf, we talked about the Helms project. This is one of the few cases I know where the hydrogen did remain some hydrogen for nine months, but it's sort of a special case. You don't usually freeze your weld for nine months. Um, another thing — actually it was like 19, must have been about 1978, I joined the faculty in '76, and couldn't really quite feed my family. But I got a call one afternoon from what used to be a shipyard, we had a Navy shipyard we had in Boston, since been closed, but they had a Navy shipyard, and they had a destroyer that had been dry docked for a number of weeks. It was coming up on the time to get out in another week or two, and it wasn't going to make it, because they were trying to weld chrome-moly steel pipes from the boiler.

And the chrome-moly — who actually, was it — was like a one or one and a quarter inch diameter pipe, it's fairly thick wall. It was stainless steel in the boiler because the higher temperatures, and then they had to make a transition weld to two and a quarter chrome, one moly steel, which is very high hardenability, has very good high temperature strength. So the very high temperatures in this heating system had to be stainless steel, but then stainless steel is expensive so they had to make a transition weld. And they had to repair some of these pipes in the boiler. And they had a procedure from Philadelphia Naval Shipyard that developed for making this weld. It was a gas tungsten arc weld, just a little pipe like this. And they'd been trying for six weeks to make this weld. They'd had experts from, uh, we would, you call them area experts right, they had people from Philadelphia Shipyard come up to show them how to make this weld. And the weld was supposed to be made, if I remember, with um, Inconel — I can't remember if it was Inconel filler metal, but uh, yeah it was Inconel. They wanted to use the NAVSEA procedure, said use Inconel filler metal, which is a nickel-chrome alloy. If I, I might have this backwards, it's been thirty, almost forty years.

And they — every time they welded it, maybe I got it backwards — uh no, I think I've got it right. Every time they welded it, they'd have to do a post-weld heat treatment for an hour. That was part of the procedure. And they'd take it off, they'd do a dye penetrant check because it's stainless steel rather than a magnetic particle, and there'd be a crack. So they kept on getting cracks, and they weren't gonna — if they couldn't weld these things, the ship couldn't go out of dry dock. So they called me in and they said, we've been able to weld it, but if we use stainless steel filler metal — and we want you to write a letter to the Navy and tell them it's okay to use stainless steel filler metal. Well, I was 28 years old, okay. I didn't have a clue. I was supposedly a welding engineer, just left Bethlehem Steel a couple years before. And I went home late, I didn't go catch the bus in Central Square to go home, I went to the library and I looked it up in the welding journal. And found out from ten years before that if you weld it with stainless steel it'll be just fine for the first ten years, then it will crack, okay, about ten years later, because you'll get carbides that go into the stainless steel. And that's why NAVSEA had this procedure of welding with Inconel filler metal, because carbon couldn't diffuse through in some of these regions and stuff.

So I thought, well, I can't tell them that, okay. So I went up there, I went down to the shipyard. And I thought I was just going to talk to one or two people. I was still naive about consulting. And they put me in this conference room right off the shop, and they have about ten engineers come in, some from the contractor and some from the Navy. And they said, well — the contractor, says who hired me, said, we want you to tell them it's okay to use stainless steel. And I said, well, I can't, and here's the paper that says why you can't. I said, but tell me a little bit more about your problem. And they were making welds — well, the weld prep looked like this, because there's a pipe, you had to weld from one side. And they were putting the root pass in with gas tungsten arc and then they try to put in fill passes, but the root crack would always crack. And so they said, if we make it with Inconel we get — or we make it with — if we make it with stainless steel we get no cracks, we make it with Inconel we always get a crack, and we want to be able to do it with stainless steel.

And I said, well, you can't. And they said, well, what can we do? And I only — all I knew was this kind of Venn diagram. I couldn't change the material, that's what they made the ship out of. I couldn't change the environment, they're using gas tungsten arc, and I talked to them about, was everything clean and stuff, do they have any sources of hydrogen? No, they're paying attention to all that. So the only thing I had left was stress. If I could relieve the stress on the weld, I might be able to make this joint. And frankly, they've been making this joint in other shipyards for years, okay. So I said, well, why don't you just machine the joint prep a little differently and make a big long, make a U-joint or a J-joint, whatever you want to call it, and make it with a big long land. So this might be 3/16 of an inch, so you'd have 3/8 of an inch, and then put your root pass in there, because now it's like a flexible diving board, right, I can get rid of some of the stress. This is fairly highly restrained, nowhere for it to go. And in fact, the reason other people have probably been successful, I don't really know, is that these guys were probably butting things up tight, and other people may not have had things quite so tight or whatever, or they were putting in a bigger weld, whatever it was, I don't really know.

And I said, so that's the only thing I can think of. And see, remember I'm a 28-year-old among all these forty, fifty, or sixty-year-old people. And I was about ready to leave realizing that I had gotten in over my head and I didn't know what I was doing. And they said, oh no wait, we'll do it — you know, we'll make the weld right here. So they went and they got a couple of stubs, and the guy machined it on the lathe. And the whole time the foreman is saying, this will not work. This is the 60-year-old foreman of welding, and he's, this is not going to work, this is not going to work. So they machine it and they asked me, was this how you want it, and I said, oh yeah, that's about right okay. And so they welded it, and then they had to wrap it in a fire blanket in order to slow the cooling rate, so it had the slow cooling of the post-weld heat treatment. That took 45 minutes. And then they had to do the dye penetrant inspection. Anybody knows anything about dye penetrant, you first spray it with a red penetrant lacquer-type stuff out of a spray can. And so this foreman is now spraying it and he's still saying, it's not going to work. And he — you have to wait five minutes for penetrant — and then he cleans it off, and then he goes to put the developer on, and he's saying as he's saying this is not going to work, and it worked. No cracks. Okay.

So uh, what can you say when you don't know anything about it? Well, you can go back to the fundamentals, okay. The fundamentals — I couldn't change material, I couldn't change — they've already had a very good low hydrogen environment, the only thing I could do is relieve the stress. And I thought of a way to relieve the stress, and I walked out of there a hero. I meant to look up and see what I had made. I have records, I think I probably charged 180 bucks for that, and I'd been there a couple hours. I walked out the hero and I thought, well, I'll get some more work from these people. Well, then they closed down the shipyard. But they did get the ship out. So that's a hydrogen cracking problem, and it kind of illustrates: you go back to the fundamentals and figure out what can you change, okay.

Another hydrogen — well, I showed you the 17-inch thick weld from the forging press. This press had been — it had gone into service in like 1948 or something, and it had been — it's a 7,000-ton press that had been making forgings for over 50 years, close to 60 years. And one day it just cracked into the top head of this forging press. So it's like, the piece is about eight feet tall, and it had some hydraulic ports and stuff, and you have this great big hydraulic chamber, and you have posts that come down from here, and the whole thing stands about six or seven stories tall. But it was this top piece, this top casting, which when I was standing next to the casting to get up to the top of the casting, you had to get on a ladder, because the whole casting here was about eight feet tall.

And they had to weld something about 17 inches up here near a man-way where the crack had occurred — or not a man-way, but a hydraulic port for the press. In any case, so they hired this company from down in Georgia, very accomplished, very sophisticated welders. These guys probably — well, they probably graduated from high school, and they may have had some welding training, but they probably never — they weren't engineers, but they were very good welders. In fact, that's an excellent weld, but I still have it here, yeah, I do still have it here. [Tom locates a weld sample.] Um, and they knew it should be preheated if it's that thick, okay. This, if you remember, is this guy that I passed around. I could pass it around again, but actually it was excellent welding, okay. Technically it was excellent. And we had pictures, they preheated it for about an hour. Except it was 17 inches thick there, and it's supposed to be one hour per inch of thickness.

Now if I had a roast beef that was 17 inches thick, great big steamship round of roast beef on a cruise ship, and they put it in the oven for one hour, do you think you would want to eat a nice slice through the middle of that? Probably be a little rare, don't you think? Well, it turns out steel has maybe a little bit better thermal conductivity, but they never got anywhere close to getting a real preheat in there, okay. For them that's all they had ever had to do. They had never welded 17-inch thick material before. They did know how to make a good weld. They were able to put down somewhere between 400 and 600 passes, and they did it in like a day and a half's time. They were working 24 hours a day on this. But when they finished the first time — well, they finished one day, and that evening someone heard a big bang. That was the crack running. And it was just from residual stresses. Hadn't been stress relieved, had been preheated. They didn't think they were going to have to do a post-weld heat treatment 17 inches thick, or they didn't do peening in between or whatever we used to do on armor steels that were 17 inches thick. We don't even know, uh, to relieve the residual stresses.

So they decided, well, we need more preheat. So they came in with heating blankets and they heated up just locally. I mean, this thing is big, I mean it's like 15 feet across and eight feet tall. And they just heated up in a local region, and this time they successfully made the weld, put it back on its post, and on the first shot to forge something — big bang and it cracked again. At that point they decided — well, they decided it was scrap, but they had a spare turns out from 60 years before, some other press piece. And they were able to fix the machine and get working again. But in the meantime, this would cost about five million dollars. And the owner wasn't happy with their welding. There was not sufficient engineering oversight on what would be a little bit beyond the normal welding that these excellent welders knew how to do, okay. So it's not just a certified welder who can put down a good bead. Sometimes you need some real engineering expertise, okay. And more, I see that we don't have the real engineering expertise. I think yesterday's presentation on what happened at Newport News — which one of you was talking about, yeah okay — I think they just didn't have the engineering oversight there, you know, hey by the way guys, there are different types of materials out there.

Okay, so that's one story. Another story, a little more disconcerting, was a pea shooter. A pea shooter is the drive shaft coupling on a helicopter between the engine and the transmission. Now you have to — in a helicopter they sort of flex, and so you need to have something that can articulate a little bit between the engine and the transmission. And so they have this hardened steel tube, it has internal splines that fit in at each end. And they call it a pea shooter because it's just a tube, but it's high hardness, it's nitrided. And, you know, it's kind of a big pea shooter — it's like three-quarter inch black iron pipe, except it's probably a five thousand dollar pipe okay, it's machined very precisely. And they had a procedure, and there was another metallurgist — just from, actually he was from Vancouver, Canada — working on this. And all —