WM_Su2014_22

They realized this was a critical path, and the volume was increasing, the demand on this. And they had another one that was scheduled to come on stream two months after this one actually blew up. So they actually dodged that bullet, okay. How'd you like to have no new aircraft engines for three years? Okay. But anyway, that was partly a hydrogen cracking problem, okay. Um, only costs forty million dollars, okay. So there's lots of stories of hydrogen cracking.

Okay so, um, I think I'm going to skip — well, if I want to look at the heat affected zone or the structure — whoops, probably turn this thing off again, or whatever. [Tom fiddles with the projector.] Bring it back. Yep okay. So this is the microstructure of a weld, uh a steel weld, and you have the weld metal which is sort of columnar, you have a coarse grain region in the heat affected zone, you have a finer grain region. And in fact we can give titles to these: solidified weld metal, solid-liquid transition zone where you actually have the — we call the mushy zone sometimes, it's like the slurpee. Uh grain growth zone, the large grain size, the recrystallized zone where you get far finer grains, partially transformed temperate zone, unaffected base metal. Don't worry about all that. If you can map that onto the phase diagram and the temperatures on this phase diagram — I think this comes out of Easterling's book, so far as that goes.

But you can get lots of different properties in here. And to give you another example of getting different properties — so when I was an engineer at Bethlehem Steel, this is for you Coast Guard folks, they were building LNG tankers right down here in Quincy Massachusetts. It was an old shipyard that Bethlehem Steel had built in World War One I think. And it was — no, it might be a Bethlehem Steel shipyard anyway. They were building the great big spherical LNG tankers, and they had to have a steel skirt. It was basically just a cylinder, about a hundred and fifty feet in diameter. It was just a cylinder, and you put the sphere on top of this skirt. So you have this big aluminum sphere and then you have a skirt that is the pedestal that the sphere sits on, okay, as far as that goes.

And I was trying to develop a very weldable steel for this steel skirt. And it turns out they were using an ASTM A537 steel, which was good to minus 20 degrees. And if you looked in the, um, the Coast Guard regs for this critical application — because you certainly don't want to lose your pedestal that's holding all this liquid natural gas — they said you had to measure the toughness, a Charpy test for impact, right at the center of the weld metal, right at the toe of the weld, and then one millimeter — right, I guess it was one millimeter, three millimeters, and five millimeters away. So you're trying to test each one of these different zones for toughness, because different things can happen and you could lose your toughness in some of those regions.

Uh, well I learned several things as a young engineer. First of all, we would call up — we had to have our machining of these things done at the Bethlehem Steel plant. And basically if they weren't busy and you sent something down there to be machined, you would pay a fortune for it, because they had to fill up their time, right. And so the machinists would take their time. If you called up and said — you'd ask the foreman, are you busy, and he says yeah we're busy, you'd ship your stuff down there when they were busy because you wouldn't get overcharged for it. But if they weren't busy you get overcharged. He never did figure out that when we called up and said hey — I can't remember his name — how busy are you, if he said he was busy he'd get work, and if he said he wasn't busy he wouldn't get work. Completely dysfunctional system.

Uh, but nonetheless — yeah, the same — oh yeah sure, this is always the same company. It's just that's the plant, this is the research department. I mean you can call it the same company but they don't work for the same purposes, right. Just like any other organization right, sub-optimized right.

Anyway, so I would actually have to — we'd have to send them down, get them machined, and then they'd send them back, and a technician would etch them and we'd mark them off and put a little — a punch mark, and so they would then put the little notch. We'd have to send it back down, they put the notch in, and then we test them for toughness. Turns out you had to meet 20 foot-pounds at minus 20 degrees Fahrenheit, because that was the Coast Guard reg.

Well, they were not — they were making the steel at Burns — at Burns Harbor, at Sparrows Point Maryland, Bethlehem Steel shipyard, our steel-making facility. And they were having a hard time getting 20 foot-pounds even in the base plate. They would often have to double normalize, go through two heat treatments to get finer grain, higher toughness steel. Sometimes they'd have to triple normalize. Now if they'd gone to low sulfur steel for an extra twenty bucks a ton, they could have easily gotten the properties. But Bethlehem Steel's management is too cheap to do that. They would much rather spend forty dollars in extra normalizing than to pay twenty dollars and give it away, right.

So anyway, every now — the average toughness on the plates going to the shipyard had to meet a 20-foot-pound minimum, and the average was just under 21, okay. So not a lot of leeway on what we were shipping to them. Most of them were down around 15 or 20, after — 15 to 19 on the first normalized, and they'd have to second normalize some of them. Got a triple normalized. But every now and then one of the plates would come in at 30 foot-pounds, okay.

And what the shipyard would do is they'd go out and find that plate and they would cut it up into test plates. Because every weld that went on the ship, you had a runoff tab. So if I was welding a seam right here, you had to have a runoff tab to make an extra 12 inches of weld so you could cut it up for the — you know, just like the prolongation I was talking about, you had to have a piece you're going to cut off that you could test. And so they were testing 30 foot-pound plate every time. And the weld might degrade something from 21 to 15, but the Coast Guard got a report that they were all above 20, because it started out at 30 and might drop down to 24. Isn't that great. Hey, we're only talking safety here.

It turns out it's not that big a deal, because what's happened over the years is we put safety factors on top of safety factors. That report I showed you from 1946 showed that the problem for the Liberty ships were Liberty ships that had less than 10 foot-pounds of toughness, okay. So when they came out in the 1950s with requirements they said, we'll add another 5 foot-pounds, so you have to have 15 foot-pounds minimum. And that was through the 50s and 60s. Then starting in the 70s the Coast Guard decided, well, some people are playing games, we'll go to 20 as the minimum. And then when they started building pipelines in Alaska and places they said, let's go to 30. And at one point they were talking about going to 80 foot-pounds minimum, okay, which is just way beyond anything anyone would ever need. Hey, but if you're not paying for it and you're just regulating someone, you just tell them what to do, right. And you don't care, you know, it's their nickel not yours.

Um, and so it turns out, you know, they were probably safe because we had an extra 10 foot-pounds in there to begin with, okay. But in fact people will shave things and play games, uh, to get around various requirements, okay.

So we're about to finish up steels, but not quite. I want to talk about a little bit about fatigue design of — actually before I talk about fatigue design I have some other things on productivity.

You can make welds in steel at — this is pounds per hour, this is kilograms per hour, whichever numbers you like. If you're welding with stick electrodes, and these are stick electrode numbers for shielded metal arc welding which is stick electrode welding, you are probably welding — unless you have an iron powder electrode like 7024 or 7018 — you're probably welding in the seven pounds per hour range. Okay, so for persons using one of these they can legally lay down about seven pounds an hour of electrode an hour.

If you go to, um, gas tungsten arc welding, you're probably only welding at about the same rate, even maybe even less. If you go to gas metal arc welding, you can get significantly higher productivity, you can lay down — and that's because up here the welder's only welding, the arc is only on 25 percent of the time, maybe 30 percent. The rest of time he's chipping slag, he's throwing away his welding stubs, he's grinding or whatever. But the manual welder has only got the arc on 25 to 30 percent of the time max, and a lot of times it's only five or ten percent. Gas metal arc welding, where you don't have any slag that you're going to have to chip off and you have a continuous feed electrode, you can get 10 to 15 pounds an hour, and so you can double your productivity, okay.

There's something called flux cored arc welding which is similar to gas metal arc weld[ing]. There's submerged arc welding, which can only be done in the flat position typically in like a panel line or something. You have a like a — it could be a eighth inch or even five sixteenths inch diameter great big electrode. Instead of a hundred or two hundred amps up here, you might be using five or six hundred amps down here. Put down great big, uh, welds. And you can get to 40 or 50 pounds an hour. If you add iron powder to your flux, you can basically melt in the iron powder. We actually use iron powder on some of these electrodes, the 7018, 7024 have some iron powder in them, so you're not just melting in the arc, you're actually incorporating iron powder that melts into the liquid metal pool.

And so you can get above 160 pounds an hour in submerged arc welding if I use — I can go with one electrode, two electrodes, three, four, or five electrodes. And with five electrodes, this is kind of being like a pipe panel, a pipe shop where I'm welding miles and miles of pipe, taking plate and turning it into a circle and welding the seam. Could be a panel line. Most panel lines don't go more than two or three electrodes, okay, to get higher productivity. But you can get — you know you can get 80 pounds an hour, which is 10 times what you're going to get out there in the ways, okay, in a panel line. But you can only do it in the flat position, great big puddles, uh, but very high productivity and good high quality welds in general, okay. So you should be a little bit aware of some of the productivity aspects.

Five — no, five different power supplies, and the first one might be DC, the second one might be AC, or second two might be AC. And they have to have — the electrical engineers have a field day here, because you have to be careful about AC and DC and different phases of the AC, okay. You can't just use regular old three-phase AC on these things, you actually have to have things that are 180 degrees out of phase, because these arcs can start interacting with each other. But it can be all one liquid puddle, and you could have six — five or six electrodes and five or six different power — five, like five power supplies all in a puddle that could be ten inches long.

I mean the weld puddle is not three quarters of an inch long, you actually end up going very fast and getting this long puddle with five electro— five independent electrodes in there. Why do they need to be independent? Because of the arc characteristics. You can't run arcs in parallel, but from basic physics an arc is not a resistive element, it actually is a negative resistance element. And so if you have to try to run two arcs in parallel, from a physics point of view one will snuff out the other and you just carry all the current in one arc, okay. Because the conductivity of the plasma becomes much greater with more current. And so if you have fifty percent of your current here and fifty percent here, this — one of them will take a hundred percent of the current and starve the other one of current. So you actually have to have a separate power supply for each electrode so they don't see each other electrically. But then if you have AC or DC you have to worry about magnetic interactions between them. So it gets to be sort of messy from an electrical engineering point of view, but from a physics point of view you can't run arcs in parallel.

Student: Okay yeah, what does the equipment look like? I mean is that a robotic weld, or is that a manual?

Yeah, it's a — what robotic, I mean it's a great big machine, it's a wire fed — yeah it's wire fed. The machine might be a million dollar welder, I mean with five — it's actually in a panel line, it's a big gantry crane. I've been talking about panel lines, I assume some of you have been through shipyards, you've seen a panel line. How many people have seen a panel line? Maybe I should ask that question. No one's seen a panel line. You've seen a panel?

Student: You're just talking about the automated system.

Yeah, but basically, you know, the plate comes in from the steel mill on some ship or on a railroad car, and it's picked up by great big magnets and it's brought over to a lay-down yard, okay.

Somewhere near that lay-down yard is a big building that's about 20 or 30 feet tall, and it's got a welding panel line. It's going to be about at least half the size of a football field, if not the size of two football fields. If you go to something like the Hyundai shipyard in, you know, Osan Korea or something, it'll be two football fields, okay. And they bring the plate in, they have a gantry crane that is the full width of that plate, and they have the welding equipment on this gantry crane. The gantry crane is only about 10 or 15 feet tall, but it's got, you know, 20 tons of welding equipment on it.

And they'll bring the plate in, they'll bring another plate, and they'll weld the two plates together to make a big plate, okay. Now weld another plate. They continuously weld these plates, and then further down they cut out all the weird shapes. You want to make a real ship — you don't make ships out of rectangular plates, you make it out of curved plates and things like that. So in a panel line you want to take the biggest plates you can buy from the ship — from the steel mill — and weld them together into bigger plates that you can cut big sections out of, and make these huge, you know, things that you're going to form and shape and turn into the sides of the ship, okay.

And so in a panel line, when you just got flat plates and you're just welding square plates together, you want the highest productivity you can. Once you start cutting them into weird shapes and taking them out and putting them up against some egg crate construction, now you got some guy with, you know, stick electrode, and it's much harder to automate. But yeah, the panel line — you should have asked me about what's a panel line, okay. I thought — actually almost none of you have been to the sub— super soup ship, right. Usually half the people in the class have been to a sup ship, but here no one's been to a sup ship. A couple of you haven't seen a panel on a sup chip.

Oh yeah, you don't have panel lines with two inch plate or three inch plate or whatever — I'm not supposed to say with thickness plate. Yeah, they might have a small panel line or something, but in general you take the big heavy plate and you take it to this forward [form] shop and they form it like a press brake that just sort of takes a flat plate and turns it into a curved plate, okay. Which is an art in itself. And they have these big plywood forms — actually they may do it with lasers now, but when I was there when they were forming the plate into the curved sections, they would have a master craftsman who had been working at this for 35 years. And he would take this big plywood form that had the shape, and they — he, you know — they would form it and then he'd take the plywood form up and see if it fits, okay. And if it didn't, you know, they'd go and he'd hit it again, you know, to shape it back to. And they — you know you'd have to have tolerances like an eighth of an inch on some of, you know, the outer surface of the sub has got to be pretty precise for hydrodynamic reasons, right.

So in any case, well, you ought to get a tour of a shipyard someday, both a surface shipyard and a sub shipyard, because they're very different, okay. Extremely different, okay. I can't give you a tour of the shipyard, sorry about that. In fact I can't even get into a shipyard anymore.