WM_S2014_09

Very low hydrogen process, which is less than four parts per million. Low hydrogen, medium hydrogen, and high hydrogen potential hydrogen level in the wire. Well, rutile electrodes, which are these covered electrodes but primarily a TiO2 coating, have very high hydrogen. You want to get hydrogen cracking, use a rutile electrode okay. It's like a 7014 I think is rutile. Basic electrodes like the 7018 that have been dried at 100 degrees C, or up here I can get down to 350 to 400, the equivalent moisture in the electrode coating. I can buy electrodes out of a hermetically sealed can, pop them open, and they'll be in the point two to point one, they'll be down in this regime, which would give me four or five ppm in the weld metal. I can use flux-cored electrodes, which are tubular electrodes with the flux on the inside of the tube, and they give me a range of hydrogen. I can get away from electro— these stick electrodes all together and use solid wire with argon or CO2 and get very low hydrogen.

If you had on Groton, Connecticut, most of the hull is welded now with gas metal arc. Back in 1975 it was still basic electrodes, but now they've gotten away from that. This is more productive, more expensive in terms of equipment and stuff, but it's one of the ways they get the lower and lower hydrogen. So we have three ways to control the three circles that we have there, to keep the hydrogen out. Any questions, okay, so far as that goes?

Hydrogen that gets into steel is reversible. If I have a piece of steel and I embrittle it with hydrogen, I stick it in an oven to get the hydrogen out, and there's no cracks that form in the meantime because I had no stress let's say. I get rid of the hydrogen, there's no more embrittlement, it goes back to being a ductile piece of steel. So it is reversible, that's good.

The most rigorous hydrogen control I ever saw was that Bell Helicopter, where they make the mast for the helicopters. You lose your mast and it's not a good day okay. The rotor goes that one way and the fuselage goes the other way, and fuselage tends to go down. The rotor goes up before it goes down, just like those little toys you know. Anyway, so — and it's a high strength steel, it's like 200 ksi steel, which means the stress levels are huge. So you've got to control — you can't control hardness, you want strength, you want light weight in that mast. You want to control the stress if you can, and it's been stress relieved, but it's really a hardness problem. And you got to keep the hydrogen down. They also want to electroplate it for some corrosion resistance, so they have their electroplating bath. And I've been there, and so here's the electroplating bath and here's their oven right here, and it goes from the electroplating bath directly in the oven within less than five minutes. Okay, you got to get the moisture off and other things, it'll dry it off, but they don't waste any time okay. That's what they did at Bell Helicopter.

I will tell you about another company that makes engines, and they have something called a peashooter, which is nothing more than a drive coupling, a spline shaft, that runs from the engine to the transmission of the helicopter. So they were doing a nitrided heat treatment on these splines. You were asking about localized hardening, so one way to harden the surface of a steel on these little spline gears is to nitride them. And so they had a spec, and they had four or five of these peashooters crack, and they were brittle fractures okay. And a couple of them caused the helicopter to crash okay. Wasn't a good day for the people in the helicopter okay.

So I was asked to look at this company, which will remain nameless at this point, but they are a manufacturer of engines, and there's only three or four — there's only three or four of them that we use in helicopter engines in this country anyway. They had a nitriding process, and during the nitriding process you can introduce hydrogen. But they also wanted to make sure that they had the right thickness of the nitride coating. And they had a specification, must be — the hydrogen must be baked out after nitriding within one to four hours. So lo and behold, I get all the specifications, I'm reading this, here's the specification, bake it out within four to four hours [one to four hours]. And then I read the actual process sheet. As soon as they would get it out, they would have one sample, one destructive sample from the lot, they would cut off an end, send it off to the met lab, and two or three days later they would get back the reading of how thick the nitride layer was. And that's when they started counting the one to four hours, when they got the results back two or three days later.

Oh how clever okay. Probably makes sense to some accountant, but you know I'm seeing a few smiles — to an engineer it doesn't make a whole lot of sense. So what do I do? Well, I have a problem. I'm a professional engineer, I have a duty. If someone's doing something that's stupid and could kill someone, I have a duty, a legal responsibility to inform people. Except I've been hired by the people who got hurt in this helicopter crash, but they're already making the same things for other helicopters. I had to inform someone, I couldn't wait for some trial to take place. So I informed the attorney I was working for, and told him he had to tell the attorney working for the engine manufacturer that they had a defective process okay. So I did that, and that sort of relieves me of my responsibility, because I informed people who were — this was the proper chain of command given that there was a lawsuit involved.

And of course the attorney working for the company, he just laughed, you know. Okay, they've had a few more failures, but — all I could do is hope that he laughed because he didn't want to pay millions of dollars in the lawsuit because it would have been essentially admitting their liability. But I could hope that he actually went back and said, hey, this guy at MIT says we're doing something stupid okay. And in fact the metallurgist at this company, if they actually saw what the process people had ended up doing to their one to four hours, they would have had a fit okay. But in any case, you do have a responsibility as a professional engineer to try to save lives, even if you're involved in something else. Even to — your primary responsibility is the public safety. Okay.

So this is — I was going to give you a graph but I probably won't. How big are the residual stresses if I'm welding? And you — we have that one inch piece back there okay. Yeah, that's — those are pretty good size welds. I'll pass this one around. Mike's talking tomorrow, but well I can put it up here for now. Okay, this is a weld. [Tom passes around a weld sample.]

This was part of a forging press. It's built originally in 1949, but it developed a crack. And it was about a 10-inch — big cast steel, this is cast steel along here, this is the weld zone. They had a crack. This was a 7,500 ton forging press. They were making aluminum wheels for trucks, they're forging aluminum wheels for trucks. And you don't have a lot of spare 7,500 ton presses. This was about a 60 ton cap to the forging press. This is the top of the forging press, it was about 10 inches thick. The crack ran through a section, it was about 14 inches in the direction that the crack ran. So they gouged it out.

And the company that came in on an emergency basis to do the welding — this thing has about three or four hundred weld passes. You can see each little pass of a stick electrode. You can see the heat affected zone is fairly narrow. We took a section here that didn't show well. There obviously was no crack here. This was actually their second or third attempt. The first attempt, it cracked before they ever got it on a crane to put it back up. The second time, they did a better job of preheating to get rid of the hydrogen and post heating, and they got it up onto the press, and the first time they hit for a forging crack, it cracked again.

And it's because they didn't really get rid of the hydrogen, and when you have something this thick you bury the hydrogen underneath. And that little video, they have a nice little computer diagram showing the effect of preheating to drive the hydrogen off, with no preheat, quarter inch thick material, 5/8 inch thick material. Well, when you have material this thick, when you go above one or two inches thick, you can bury the hydrogen down here with multiple weld passes and you'll never get it out. You go through the diffusivity, 10 to the minus 6 centimeters square per second, and see how long it takes to go through 2 or 3 inches of steel, much less 14 inches of steel. You have to get the hydrogen out as you go. And that means they actually have to stop and do a post heat. You might weld for an hour, stop for two hours, and keep it hot, drive the hydrogen off, then you can go back and weld some more. Because if you keep on putting passes on top, all you're doing is burying the hydrogen. So welding very thick sections and avoiding hydrogen cracking is a real problem.

In fact, it's such a real problem that back 20, 25 years ago when we still had battleships, there was a — some sailor down in Puerto Rico or whatever on the battleship had another seaman who was his lover jilt him, and so to spite him he set off one of the 16-inch shells inside the turret okay. Boom. You know, the turret goes jumping up in the air, and the Navy wanted to repair it. The armor plate is 14 inches thick, and back around 1980, no one knew how to repair our 14 inch thick armor plate. The last 14 inch thick armor plate we welded was like 1947. That was the last battleship we built with heavy armor okay. Because today shaped charge go right through 14 inches, boom, all right, go through 36 inches, I've seen it okay. Just shaped charge would cut right through 36 inches of solid steel. But back in World War II they had 14 inches, 16 inch thick armor plate, all welded with stick electrodes okay.

But we lost the technology because everyone died who knew how to do it. We could figure it out again, we'd have to develop a procedure. But developing a procedure for 10 inch thick material would probably cost a half a million dollars okay. Well, when they're building, you know, multiple battleships a year, they could do that, but they can't do it today, because we don't weld steel that thick very often. And so — and we also don't weld it with stick electrodes with 400 passes. You can imagine why you don't want to have to weld it with 400 passes — not very efficient. But they didn't have another way to do it in World War II, so they did it the inefficient way okay.

I'll — Mike — I'll be here tomorrow, but Mike will be lecturing because I'll be in — maybe a little few minutes late, and then I'll be here Thursday and Friday. Yep, weld for thick metals together, we use gas metal arc, and we control — that's a very low hydrogen process to begin with. So we basically keep the hydrogen out to begin with. We shrink this circle, we do some things to control the hardness, we will stress relieve afterwards. I'm going to talk about preheating and post heating a lot more on Thursday.

Okay, actually what I'm going to do is, I'm going to take this Venn diagram, I'm going to show you sometimes we shrink this circle, sometimes we're shrinking that, sometimes we shrink two of them. There's various things that we do, but you can think of all of them as just variants on the Venn diagram. Okay, thanks.