§1. Bessemer and the problem of melting steel [00:03]
If no one has any questions today, I want to talk about how we melt metals before we cast them. We may get a little bit into some of the casting today, too. One of the things I'd like to point out is it's not that easy to melt steel. In fact, I always say if it was so easy to melt steel, we didn't need Henry Bessemer in 1856 to teach us how to melt steel and bring us into the steel age. And it really was Bessemer who did that. We had steel going back thousands of years, but most of it, virtually all of it, was either meteoric iron — someone found a big meteor somewhere and they beat it into swords typically —
or it was solid-state diffusion iron, where they essentially took iron ore and buried it in the sand kind of like a clam bake, only this was an iron ore bake. You'd throw it in with some charcoal and heat it up in the absence of oxygen. The iron oxide in the presence of carbon would turn into sponge iron through solid-state diffusion. You never melted the steel, and when you're all done you get this spongy mass of iron which had all kinds of other impurities in it, and then they would just beat that and forge it. A very inefficient process, but that's how they got swords a thousand years ago.
It wasn't really until Henry Bessemer taught people how to melt steel. The problem is that a typical fuel-air flame — and the fuel can be natural gas, wood, charcoal, coke, or whatever — is only about 2,000° F. And I'm going to prove that to you in a second. Steel melts at around 2,500° Fahrenheit. Obviously you can't melt something at 2,500 with a 2,000-degree flame. A Bessemer converter is just a vessel turned on its side, typically with an asymmetric spout, and it blows the air in. It starts out as cast iron, which you get out of a blast furnace, and they blow air in through a tube. The hot gas goes out the top, the cool air comes in, and the cool air gets preheated by the gas going out. There's a lot of mixing, but if you do the Bessemer converter properly, you preheat the air. If you preheat the air to 1,000° and now burn it, you get enough flame temperature to turn the cast iron into steel.
I can prove this to you fairly simply, as long as the safety officer doesn't know what we're doing in the classroom. [Tom produces a piece of steel and a propane torch.] This is a piece of steel and I can put it in a propane flame. It should start to glow red. Okay, so it's glowing red and I can hold it there all day long. And that's probably only about 1,600°. Anybody want to feel it?
So that's a pure hydrocarbon flame, premixed, burning in air. What we actually do today in most steel mills is we use pure oxygen and you can easily get up to the temperatures. But back in 1856, they didn't have any source of pure oxygen. So they had to come up with the Bessemer converter — a vessel that, when they blew air in just like the blacksmith with his forge, would preheat the incoming air with the outgoing gas. That allowed them to get temperatures sufficient to melt the steel.
§2. Survey of melting technologies [05:15]
Today we have lots of different ways to melt metals. You can almost categorize them by mechanism: radiation and conduction in a reverberatory furnace — that's basically the open hearth. A cupola, which I mentioned before in the Great Leap Forward in China in the 1960s, when they started building cupolas at every little commune so they could all make little cast iron parts. The blast furnace is just a big version of a cupola. Electric arc — you're using electric energy. Induction — electric energy, but through electromagnetic fields rather than current.
Vacuum induction is a variation on that. Plasma, which is again electricity but turning it into an arc plasma. Crucible — we don't use crucibles for melting steel; crucible melting is typically for coppers or aluminums or other things that melt at lower temperatures. Electroslag — electricity. Vacuum arc — electricity. Skull melting can be electron beams, arcs, induction. In general, we're using radiation, conduction, or electric energy to do the melting. But there's lots of variations, and I want to talk about some of the industries. The big industry for metals is steel, so we're going to talk some about steel.
Once you get your metal by some melting technique, you've now got to solidify it. You're going to have to extract the heat — diffusion. You're going to try to impart some geometry and structure. Initially, when we're just making steel, we're just trying to make plates or sheets or bars or long slender objects, and we have to worry about a number of things.
§3. Blast furnaces, cupolas, and slag [09:32]
We start off with iron ore, limestone, and coal. We put it into a blast furnace. A blast furnace is nothing more than a great big cupola. We're going to have these preheaters over here. And then eventually we're going to make steel in either an electric furnace, open hearth, or basic oxygen furnace.
Let's talk about a cupola first. The simplest version is a big tall shaft. We have a charging door at the top and we pour our solid materials in. There's a refractory bed here, but don't worry about that right now. The stuff just keeps falling down, you put your fuel in with the air, and you just blast it in there. Heat things up, and at the bottom your impurities will melt and form a glass — we call it a slag. You have your metal, and they just put a little clay plug in there. Every six hours, every twelve hours, they knock out that plug and the metal comes running out. After the metal comes out, the slag comes running out. Typically you'll take the slag off one side and the metal off the other side. This is a continuous feed furnace up here. That's how the Chinese built thousands of these all around China.
The steel blast furnace is basically the same thing on a much larger scale. This thing can be thirty stories tall. You put your charge in here and it starts out relatively low temperature and gets hotter and hotter as you get to the bottom. They call this the tap hole, the slag hole. The tap hole is a little lower than the slag hole because this is where the molten iron is. Remember, we started out with coke, which was coal that had burned off the volatiles, so we had carbon in here. What we end up with at the bottom is reduced iron oxide — we've reduced it with the carbon. So you get a lot of carbon monoxide coming off this, and that comes through here and heats up the brickwork. That gas goes off to run the rest of the steel mill. Carbon monoxide as a fuel gas — might poison everybody if they breathe it, but it's a fuel gas. You can burn it to heat up things in furnaces before you're going to forge or roll.
The other air coming in is preheated because the thing runs more efficiently that way. These are just two big heat exchangers. One preheats the brickwork; the other preheats the air and cools down the brickwork. And they just shift it over every twelve hours. These are huge things. Takes twelve hours to heat them up. A large part of the expense of the blast furnace is just these big brickwork lattice things, which are basically heat exchangers. So this is basically just a big cupola. You take fifty tons of slag off this side. You take three hundred tons of cast iron off the other side. This is carbon-saturated iron. Melts at about 1,450° or 1,600° Fahrenheit.
Student: Is there anything you can do with that slag waste product?
Not a lot. You can break it up into gravel and pave roads with asphalt. Blast furnace slag is not as bad as steel-making slag. You can spin it into slag wool like insulation — not Owens Corning pink fiberglass, sort of a black stuff. But it's a problem. One of the problems in the world is we produce a couple hundred million tons a year of slag which goes into landfills by and large.
Student: Reefs. You can make reefs.
Well, not with steel-making slag. There's a lot of impurities the environmentalists don't like. For blast furnace slag, it's fairly simple — kind of limestone and silica. So it's not too bad. You could make reefs, but how many reefs do you want to make? You can't fill in Boston Harbor anymore. So all this slag is still a problem. Everybody says, "Whether it's old rubber tires or blast furnace slag, let's mix it in with asphalt and pave the roads." Well, we don't have that many roads to pave for all the junk that we want to put in there.
People have been fairly creative about what they've tried to do with them. About fifty percent of it doesn't go into landfills — it goes into making insulation or products. But when you make this much of something, it takes a lot of whatever it is to make all that. Here's a little bigger version. You've got big blowing fans. There's a man sitting next to this for scale. It's just a cupola, but they call it a blast furnace just because it came up through kind of a different genealogy.
§4. Open hearth, BOP, and electric arc [16:21]
The open hearth is basically a reverberatory furnace. Here's your bath, and you have this big brick checkerboard to heat up the incoming air, with similar chambers on the other side. The hot gas comes in, goes across the top, goes out, heats up the checkerboard, and every half day or day they switch the flow. You tap it every now and then. To tap these things, you don't want to lose three hundred tons of steel, so they actually use an explosive charge to open the tap hole. Just stick some explosive in there and blow it apart. This is large-scale stuff.
It's a very inefficient process, because it's just radiation and conduction of hot gases across the top. What are we doing in these furnaces? All the carbon that got saturated in the cast iron in the blast furnace, we have to get out. We have to go from three or four percent carbon in the cast iron to about ten times less to make steel. You've got to oxidize the carbon in the air without oxidizing the iron, too. If you oxidize the iron, you're right back to the ore, and that doesn't do you any good. So it's a balance of how much oxygen you put in there — enough to burn off the carbon, which burns off preferentially to the iron.
The basic oxygen furnace is just a quick version of the same thing, except it's a smaller vessel. Here are your slag pots, and you have an oxygen lance. You have an oxygen plant next to this. The electric arc furnace, the third way to do it, is basically three carbon electrodes about thirty-six to forty-two inches in diameter. You're going to have three or four hundred volts at 50,000 amps, about 150 megawatts of power. In this, you can charge 100% scrap because the arcs can melt everything. In the basic oxygen furnace and the open hearth, you have to start with about seventy percent cast iron because you have to have liquid. You can throw some solid scrap in, but no more than about thirty percent, so you don't freeze your bath. If you don't start with a liquid bath, you won't get the chemical kinetics to burn off the carbon. The oxygen would just burn off the top surface layer. You need the convection in the liquid to bring fresh carbon to the surface so you can oxidize it away. So you have to start with virgin iron ore.
The BOP can make things in forty-five minutes. If you've got three hundred tons of steel in the pot — if the cast iron's worth a couple hundred dollars a ton, maybe $300 a ton today, and when you make it into steel it's now worth $400 or $450 a ton — you're talking about a value added of $150 a ton. Three hundred times 150 is $45,000. So $45,000 of value added per heat, $100,000 an hour worth of value added going through this BOP shop. And the actual value of the product, if you screw it up — you don't have to throw it away, it can be reprocessed — but if you had to throw it away, you've got over a million dollars an hour worth of product coming through this plant. So it's a substantial investment.
§5. Ingot casting and rim steels [21:42]
Those are the three basic types. In the old days, back in the 1960s and before, we basically would cast it into an ingot. There's lots of different types of ingots. You have a big cast iron mold, and the steel would be cast into different shapes. Depending on the chemistry of the steel — how much gas is in it — as the steel cools down, it will give off the gas, and it may actually start to boil. You'll get what they call a carbon boil. It's actually a carbon monoxide boil, but they call it a carbon boil. You can essentially fill up the whole vessel. If you've done something to get rid of the oxygen — like adding aluminum or silicon or manganese — you'll end up with no carbon monoxide boil and you'll get solidification shrinkage. You're going to have to cut off the top third of that ingot. To make really heavy plates above about four or five inches thick, we still have to use ingot technology. But today, ninety-seven percent of the steel in the world is made by continuous casting.
Student: What are the other variations?
These are what they call semi-killed steels, and it's not a very good structure. The other variations may have what they call hot tops on them to try to keep the heat on top. They make kind of a thermite reaction, put it on top, and try to get the hot top smaller, because you're going to throw away one third. If you want, I can dig up the Making, Shaping and Treating of Steel book — there's lots of variations. The old sheet steels they used to like to make this. It was called a rim steel. These are different amounts of action with the boiling carbon. You get porosity in your steel, but if you're going to roll it into sheet, it's all going to get welded together anyway. If you're going to make good plate and you're not going to roll it as much, you've got to go with a killed steel. If you get rid of the oxygen before you pour it, it's called a killed steel. This is a semi-killed steel. These are unkilled steels, or rimmed steels, with carbon boils. These are for sheet material. This is for plate material. Rim steel gave excellent surface quality, whereas plate steels you end up with inclusions and other things on the surface.
§6. Continuous casting and the productivity revolution [24:36]
Starting in the steel industry in the 1960s, we had this other thing that improved productivity. We haven't done much with the blast furnace in terms of improving productivity for the last four hundred years — it's all sort of incremental. But we went from the open hearth, after a hundred years, to the BOP, and we got this tremendous increase in productivity. About the same time, we went from ingot casting, which was essentially 100% of all steel, to now only three percent of all steel, between 1960 and 1995. We went to continuous casting.
In the first continuous casting, you come out of your melting furnace — your BOP or your open hearth — and you pour it into a ladle. The ladle goes into a stationary holding vessel called a tundish, and you pour it into a mold. These are huge water-cooled copper molds that vibrate back and forth. They may only weigh fifty tons or so, but they're vibrating them. As the steel comes in, the water-cooled copper extracts the heat of fusion. Remember I said you had to take out the heat of fusion to cast it. You come down here, the stuff is still hot, and initially it was just a big tall vertical furnace. This could be thirty stories tall again. Then you cut it off. They actually got to the point where they went from vertical to curved — they curved the steel while it's still hot. So you have great big rollers.
You're going to have five hundred tons of steel in the tundish. And this is probably about a billion-dollar facility right now to build one of these, maybe a couple of billion. Here's your tundish, here's your ladles, and you keep on bringing them over. You may run a continuous caster continuously for a month or two. You may have to shut it down every now and then to replace refractories and wear and whatnot, but you just run it twenty-four hours a day. You can't afford to shut this thing down. If you have a breakout — a breakout means all of a sudden you no longer have a solid shell and the molten steel just falls on the floor — that shuts down your continuous caster for about a week or two to clean up. If you're talking $100,000 an hour, we're talking some real money. People don't like to have breakouts.
Nowadays we don't have as many breakouts, but back in the '60s and '70s, you were lucky if you only had one breakout a year. It took some time to get to the point where they could control things well enough. When you come off, you basically can cut this stuff. Even though it's a big thick steel — we can melt something ten inches thick, some places go as much as twelve inches thick. They can't go much thicker because you've got to extract the heat of fusion. The thicker it is, the longer it takes to suck the heat out of the inside, and it goes as a square of time. The thermal heat flow equation always has the form: the distance heat travels is proportional to the square root of the thermal diffusivity times time. So the thickness X goes as the square root of time, or X squared goes as time. This is a material property — thermal diffusivity of the steel. If you go up from ten inches to fourteen inches, you've got fifty percent greater time. You've got to have fifty percent greater height and bigger molds and everything else. So about the thickest we can go is on the order of ten to twelve inches.
You're going to have some defects in there, so you're going to have to roll it to get rid of those. Generally the thickest plate we would make by this is two to four inches thick. Anything thicker than that, we still have to go back to old-fashioned ingot casting. Ingot cast molds can be thirty, forty inches thick. Typically, the old ingot casting lines, you have three hundred tons of steel — you might cast fifteen ingots, so it's about twenty tons of ingot. And if someone in the steel mill is not careful and they fall into the ingot, they bury one ingot from the heat. That's the way they did it — the graveyard, they just bury one ingot.
§7. Hazards of the steel mill [29:50]
When I was at Bethlehem Steel, there was one guy who had just retired, and I guess he couldn't really handle retirement. He came back to the plant one day and he dived into the bath of molten steel in the BOF. They were sitting there waiting to get chemistry back, and they saw a poof on the TV screen, and they looked in and they could see this body spread-eagled on the molten bath of steel, floating on it. They float really well. The density of steel's much heavier. He had committed suicide by diving in there. So what do they do? They'll bury one ingot when they tap the steel.
Steel mills are hazardous places. There were people working on the hot tops on the ingot casting floor, and every now and then someone might trip and fall in. It's actually a fairly quick way to go. But the old steel mills were just places to maim people. My second child was born in Bethlehem, Pennsylvania. Wonderful hospital for a town that size, because it was just a couple hundred yards down from the steel mill. In the old days, like the 1930s and 1940s, when you're rolling bar, it'd be coming through at forty miles an hour — a piece of hot steel a mile long, coming through at forty miles an hour, and a guy was supposed to grab it with tongs and put it into the next stand. If anything happened, they called it a cobble. The thing curved when it came out, and it might just cut his arm off — but it cauterizes the wound at the same time. It's hot. It just cut through human flesh like a hot knife through butter.
Nowadays you walk through a steel mill and you have to have a reason to be down there on the floor. They basically have the people in protected pulpits with armored glass and everything else. So it's not anywhere near as bad as it used to be. But just like we now have all kinds of people in Cambodia with missing legs because of landmines, we used to do that in steel mills seventy, eighty years ago. We had people with missing limbs all over the steel mills, because it was just a hazardous place to be. It's not like that anymore, but there's a reason why unions have a lot of power if you look at what people had to go through in the old days. Anyway, a continuous caster could produce four or five million tons of steel a year out of one caster.
§8. Defects, segregation, and vacuum degassing [32:35]
After casting the steel, you really have to start worrying about getting rid of some of these other things. You've got to extract the heat of fusion. By the way, one of those forty-inch ingots may take two or three days to solidify. Once you cast those forty-inch ingots that you're going to forge some big thing from, you've got to let them sit there for two or three days. You can't jostle them. You can't pull them away in the railroad car, because if you vibrate them as they're solidifying, you're going to introduce defects. When you had ingot casting — and we still do have ingot casting, when you're making really large parts like a big shaft for a propeller — that costs a lot more money, because they have to cast it and that part of the plant, nothing happens for two or three days while it's solidifying. That's just basic heat flow. You can't speed it up. Nature gives it to you.
You can have shrinkage which gives you porosity at the top. You can have gases which create the boil. You can have liquids which end up as inclusions. You can have segregation — it's an alloy, it's got carbon, manganese, silicon, and other alloying elements, and they may not be homogeneously distributed. One of the issues is getting rid of the gas or the liquids that you don't want.
One of the things you can do is pour from your electric furnace into the tundish, add alloying elements at that point in the ladle, and pull a vacuum on this. If you pull a vacuum, you pull out all those gases dissolved in the steel — just like the rim steel will give you carbon monoxide boil if you don't want it, or if you have other impurities. This is not just steel. Hydrogen in aluminum — just the moisture in the air will introduce hydrogen into the aluminum. If you solidify that, you'll end up with Swiss cheese, full of holes. So you've got to do something to process the liquid. This is just a schematic of tap ladle, teeming ladle, big vacuum chamber, and exhaust lines of the vacuum tank. You're basically just going to suck all those gas impurities out. This is vacuum-degassed steel. Costs about $20 or $30 a ton more — vacuum degassing — but makes it very pure in terms of the gas impurities.
§9. AOD and the stainless steel revolution [35:41]
Another thing is vacuum oxygen decarburization. This initially started — actually this was invented in the basement of Building 8 here, and about ninety-eight percent of all stainless steel in the world is made this way now. Stainless steel is an alloy of iron and chrome. A lot of it is iron, chrome, and nickel. But you want low carbon — ultra-low carbon. Ultra-low carbon means 300 parts per million or less.
In the old days before the 1960s with the electric arc furnace, people had to heat it up, put a slag on, take that slag off, put a cleaner slag on, keep it in the furnace for hours to burn off the last amount of carbon. But because of some basic research done here in the basement of Building 8, they were looking at what happens when you bubble argon through the bottom of the steel bath. Instead of having a bubble in your molten bath of one atmosphere of carbon monoxide, if most of your gas is argon, the partial pressure can be a hundredth of an atmosphere of CO. If you just look at the thermodynamic equilibrium, you can drive your carbon down very low.
The simple reaction for those of you who are chemists is: carbon — underlined for dissolved in the steel — plus oxygen dissolved in the steel goes to carbon monoxide gas, and the equilibrium constant is the pressure of CO gas divided by the activity of the carbon times the activity of the oxygen. You want to drive this down as low towards zero as you can. You actually want to drive both of these down towards zero. In order to do that, you have to have a low CO pressure. If you're just getting a carbon oxide boil in the steel bath, it's 100% CO. That's one atmosphere. So you can't get the carbon down unless you get your oxygen up. But if you drop this by a factor of 100, by having 99% argon and 1% CO, you can get this down much lower.
So today, almost all of the super alloys, the nickel-based alloys for turbine engines, even carbon steels for high quality parts, aircraft landing gears and things like that, go through argon oxygen decarburization. You basically have a vacuum chamber of the molten steel, an oxygen lance you can blow oxygen in to get the carbon down, and you also bubble argon through the bottom. Here's the picture of argon oxygen decarburization — argon blown in here and just a big vessel.
Here's actually a better picture. You've got auxiliary oxygen to get the carbon down. Nowadays we actually first blow with nitrogen. You can't blow with all nitrogen because then your nitrogen in your steel will go too high and that can create problems. You start off with nitrogen, get most of the oxygen and carbon out, and then you finish off with argon. Argon is expensive, but it's stainless steel. This is a plot showing carbon content versus chromium content. Electric furnace practice was down like this — this would be 200 parts per million. You really want to get below 300 parts per million of carbon. The AOD practice goes this way, at much higher chromium content. In the old process, you had to get the carbon down low and then add your chromium as an expensive type of chromium. Nowadays we can use cheap chromium. Basically this process has reduced the cost of stainless steel in the world by about a factor of two per ton since the 1960s.
Here — this is chromium content versus temperature. You actually heat up your steel to higher temperatures because you're burning off carbon. You're adding heat when you blow the oxygen in there. The chromium goes down a little bit, you add a little bit more, but the carbon comes way down to these levels that you need to have good weldability and good corrosion resistance in your stainless steel.
§10. Who innovates in the steel industry [41:16]
So something initially developed at MIT — but it also brings up an interesting point. Someone did a study in the early 1990s of how many innovations in the steel industry were brought about by the steel industry, and essentially it was zero. Things like argon oxygen decarburization were done by a company that wanted to sell oxygen to the steel industry, and argon — the gas companies saw this as a way to sell more gas to the steel companies. Continuous casting was really brought in by companies that wanted to sell machinery to do the casting. The basic oxygen furnace was brought in by companies that wanted to sell the gas, pure oxygen, as opposed to air. No one can sell air. It's pretty free.
Out of dozens of improvements in the steel industry in the world, none of them over the last seventy-five years were done by the steel industry itself. That's because we had this fantastic management in the 1950s and '60s in the US steel industry. They had bombed out all their competition. They were making money hand over fist and they thought they were the greatest managers in the world. Actually they were probably some of the worst managers in the world, but no one could prove them wrong because they had no competition. So it's, you know, "Well, it's too bad that Bethlehem Steel went away and all these other steel companies went away." Yeah, it's unfortunate those workers who were having their arms cut off lost their jobs. But to think that management lost their jobs — that's actually a positive thing.
I worked at the steel company for thirteen months before I said, "Where do I want to be five years from now?" and the clear answer was not here. By twenty months I was gone. Today I would use the word corrupt. They had three or four Learjets parked at the Allentown airport, and their primary function was to fly the executives to Florida on the weekends in the winter so they could play golf. The only other organization I've ever seen doing similar types of things was the US Air Force Reserves, where they had to get their flight hours in and they would fly cases of Coca-Cola over to Tokyo on a C-5A so the pilots could get their time in and they could go to the electronic shops in Tokyo and buy stereos cheap. That wasn't the same type of corruption. They weren't making any money. They could just buy cheap stereos for their friends. I could buy Coca-Cola in Tokyo cheaper than I could buy it in the United States per case, flown over air freight courtesy of the C-5A and the US Air Force Reserves. It's an interesting system of economics in the world sometimes.
§11. Induction, vacuum, and consumable-electrode melting [44:50]
Other types of melting. We have cupolas and blast furnaces to make cast iron. We have electric arcs — that's the electric arc furnace. We have induction melting. Induction melting is nothing more than a big pot. You haven't taken them to the foundry downstairs, have you? Not yet. There are two induction melters down there. They're about 200-pound furnaces. With a full charge they're a couple hundred pounds. But you can get induction melters that are 100 tons.
There's one guy, a graduate of MIT, who basically bought Inductotherm and Lepel — there were only two or three big induction melting furnace companies in the world. He ended up acquiring them all, getting sort of a monopoly. His name was Henry Rowan. He wanted to have a university named after him, but he didn't have enough money to convince MIT to change their name. He went to New Jersey Institute of Technology and they wouldn't do it for $100 million, or $200 million. So he went to another school in New Jersey — he was out of somewhere just north of Philly — and he convinced this other school to take his money. I can't remember what its name was before, but it's now Rowan Institute of Technology. Because he gave them a couple hundred million dollars.
So induction melting is basically just a big water-cooled copper coil. I used to have one in the lab that was not much bigger than your fist. It can go up to about 100 tons. Big heavy AC power leads. The little ones — because we have things called a magnetic skin effect — here's a cross-section. These are the copper coils, big crucible, basically an AC induction. If you're talking a 100-ton vessel, this might be in the 60 Hz frequency range. But if you're talking the little ones in the lab that can heat up foils, you might be talking a megahertz frequency. So there's lots of different induction units. You're using AC induction to generate eddy currents in the molten bath, or the steel can start out as solid steel but you essentially electrically resistive-heat by coupling to the steel with induction. Magnetic induction, AC effects, just like an electric motor — induction motor — you're expanding and collapsing the field and generating an induced current in the metal. You can melt all kinds of things. Nickel-based alloys, not just steels. In fact, relatively small amounts of steel are melted by induction melting.
Vacuum induction is just a more expensive furnace. If you want to make higher quality and get the impurity gases out that will dissolve in the molten metal, then you run it in a vacuum furnace. Ron Ballinger's got the old vacuum induction melter over across the way. We used to have a big vacuum induction melter. Vacuum induction is typically the way that you melt things like turbine blade alloys — nickel-based alloys, things that are going to go into products worth thousands of dollars a pound like a jet engine. Very high quality melting, very pure and nice and clean.
Other processes. There's what they call consumable electrode melting, which is just a big single arc furnace where you actually cast a steel electrode and then you have a water-cooled copper ingot mold. Water out, you just run electric arc and melt this just like a great big welding electrode, depositing metal into a crucible. Expensive, but if you do it in a vacuum you can get rid of a lot of inclusions. This material would be VAR — vacuum arc remelted. Typically for a turbine blade, they'd be vacuum induction melted, and typically for the best materials it has to be all virgin material. Virgin nickel, virgin chrome, virgin molybdenum. No scrap, because scrap will have impurities and oxides on the surface that'll get into your material.
One of these turbine blades will go for about $6,000 when it's all done, each. So you're not going to skimp on the incoming materials. You're going to do vacuum induction melting and vacuum arc remelting. And it might be double vacuum arc remelted. If I have a critical shaft or a landing gear on an airplane — VIM-VAR. It'll probably be double vacuum induction melted, maybe triple. Each time you melt it in the vacuum, you're burning off some more inclusions and getting rid of some gases and getting down to a very pure high tonnage material.
§12. Electroslag and the Paton Institute [50:41]
Another process which is not used so much in this country, but was used a lot in the former Soviet Union for political reasons — well, not just political. They would cast the steel and remelt it in electroslag, where you have a molten glass which is a slag. You run a bunch of current through here, and the resistive heating in the slag causes the electrode to melt off. You cast an ingot underneath in a water-cooled copper mold, and you just drop the bottom of the furnace. It starts to solidify in the bottom, and then you just lower it and you can make big long ingots.
The only time I was ever in Russia, in the Ukraine, was back in 1980. I went with Julian Szekely, who was a professor here who had been a student in Hungary. He left Hungary in '56. I said, "Did you have a gun in your hand?" — he was a college student at the time. He said, "Well, yeah." I said, "How'd you get out of Hungary?" He says, "I just took the train to the Austrian border and walked across." But he did have a gun in his hand in the '56 revolt in Hungary. Anyway, Julian was a steel expert. When we were over there, the Soviets were very proud that half of the steel they produced was electroslag remelted. At the time we would only do this because it's an extra process — it costs you an extra $100 a ton to improve the quality. And Julian says, "Yeah, but that was because the stuff they made the first time around was such bad quality. They had to electroslag remelt it to get it up to the kind of quality we would get out of our BOP," which was probably true. Their stuff was just junk. They had to do electroslag remelting to get rid of all the big inclusions the size of your fist. Can you imagine a piece of steel for a ship plate that has an inclusion as thick as the plate? It just pops right through and you've got a hole in your plate. That's what the Soviets produced. So they had to do electroslag remelting to melt these inclusions and make a better ingot.
There's actually a political reason why they were so interested in electroslag. In Kiev, there was a place called the Paton Institute. It's still there. During World War II there was a Dr. Paton who was looking at electroslag welding, and he developed this for welding of armor plate. He was very proficient at getting the Soviet tanks repaired and back to the front to fight Hitler. After the war he was a hero of the Soviet Union. Stalin decorated him. He was in charge of all the scientific money that went into the Ukraine — and that was about twenty-five percent of everything the Soviets did. He had his own institute, which should have been five hundred people, but because he controlled all the money it was about five thousand people. He had grown up doing electroslag welding, so they did all kinds of electroslag processing, and he convinced all the Soviet officials that this was the savior for their steel industry. And it probably was, because they couldn't make decent steel to begin with. You read things about how wonderful the Soviet steel technology was because they did all the electroslag remelting. It was only because the first part of the process was junk.
§13. The productivity revolution of the 1980s [54:50]
So why don't we take a break and come back at about 8:35. To show that I'm not the only person who believes iron is important, Rudyard Kipling put this together. He said, "Gold is for the mistress, silver for the maid, copper for the craftsman cutting at his trade. 'Good,' said the baron sitting in his hall, 'but iron, cold iron, is master of them all.'" Steel has properties that are vastly superior to many other things.
To show you something about the steel industry and what happened because of switching over from the basic open hearth to the basic oxygen process and because of continuous casting — you had a tremendous increase in productivity. If you looked at the decade of the 1980s, the United States used about 100 million tons of steel. There was about 76 million they produced that they sold, and another 25 million they imported. Except for the recession in '82, this is basically fairly constant across the decade. Believe it or not, it started out at 100 million tons and ended up at 100 million tons. The employment, on the other hand, dropped by a factor of two, from a little over half a million to a little over a quarter million. So what happens if half the people produce the same amount of material? That means in ten years the productivity, or tons per person, went up by a factor of two. This was the same decade where everybody thought the steel industry was dying. In fact, it wasn't dying. It was reinventing itself by going to BOP and continuous casting.
It also switched to a lot of production coming out of what we call minimills. This is a picture of a continuous casting line. It was curved up here behind here. When the steel is hot like this, if you want to cut it as it's coming down continuously, you just take an oxygen jet. You don't have to do anything else. You just blow pure oxygen on it and it does oxy-acetylene cutting right through it — any thickness you want, ten inches, twelve inches. I've seen people cut three-foot-thick steel with a solid oxygen jet, as long as it's hot. In another couple of lectures you'll get to flame cutting and I'll explain how that works on your videos.
This happens to be one where they're not just doing a big flat ten-inch-thick plate. This was a steel company in Midlothian, Texas, run by Gordon Forward, a graduate of this department. They actually were getting to net-shape casting. They were making I-beams and they were casting something that was almost a final I-beam product. It took them a while to get there. Here's a slide of one of their earlier products where they had, instead of a simple rectangle, a partial dog-bone shape. Eventually they got to the one I showed you before. They cut their rolling costs down by about two-thirds if you got near-net shape. You can roll much faster and more productively, and so you got other follow-ons in productivity. You started out with continuous casting and then went to other casting technologies.
But for some things like big shafts, you can't do continuous casting. You have to do ingot casting. This illustration just shows them stripping the cast iron mold away from the ingots after they've been solidifying for a few days. They're typically corrugated on the surface because you don't want them sticking to the mold. As the steel shrinks, you want it to pull away so that you can strip it off. It's sort of a pain when you can't strip it off. There's a big hydraulic ram here trying to push the thing down if it does tend to stick.
§14. Surveying and the SS LNG Libra [59:38]
I wanted to do a little case study, since you guys are in the Navy. This one I do on one of the other videos, so you can fast forward, but in the revised way I'm putting this course together, this is a good place to talk about it. Anybody know what surveying is in shipbuilding?
Student: You don't — you're checking the status of production of a ship, right?
You've got the American Bureau of Shipping, you've got Det Norske Veritas, you've got Bureau Veritas. If you're British Petroleum — actually, in the Navy, you have SUPSHIP. SUPSHIP is a surveying function. In the commercial business, you would be a surveyor. If I'm BP and I'm building a $2 billion oil rig to go drill in the Gulf of Mexico or the North Sea, I want to make sure that steel company does it right, or that shipyard does it right. So I will hire the American Bureau of Shipping and they will send their surveyors in. They will live full-time at the shipyard, just like a guy at SUPSHIP does, and walk around checking processes, checking quality control records, doing all those types of things that many of you are familiar with. Also, if you have a failure, Lloyd's of London or somebody will send in some surveyors to say, "Why did we have this failure?"
This is a case study of a failure that occurred in October 1980. This is the ABS report — "to certify the undersigned surveyor did at the request of the owner's representative." They use sort of an old English type of language format because it's going to Lloyd's of London. "Attend the SS LNG Libra of Wilmington, Delaware" — it's about 125,000 gross tons, I thought — "as she lay off Davao, Davao City in the Philippines." Basically she had lost power, fully loaded with LNG, going from Indonesia to somewhere in Japan, and she just stopped dead in the water.
The Libra had been built right down here in Quincy Shipyard, which used to be a Bethlehem Steel shipyard back in the days when I worked for Bethlehem Steel. It's now an auto parking lot, the big crane and stuff. These were LNG vessels built in the 1970s. I worked on a project to develop a new steel skirt. They had these big aluminum spheres. If you've ever seen these LNG vessels, they're like five spheres the size of the vessel. They were like 125,000-ton vessels — huge. They made these huge aluminum spheres that were like twelve inches thick aluminum at the equator. They carried the liquid natural gas, and those spheres had to sit on a cylinder, and I was working on the steel for the cylinder.
§15. LNG dangers and the Qatar explosion [63:28]
By 1980, one of them had a tail shaft failure. The next part of the report — this is a little two-page report — describes when he went to the Philippines. This was an out-of-the-way area. The LNG vessel is all cold liquid, and it will slowly heat up. It may take a month to heat up — you're just basically boiling off gas, natural gas. If you don't have power and a storm came along, you could have a big fireball.
Such big fireballs have happened. Has anybody ever heard of any with liquefied natural gas? In the late '40s, if you look up LNG explosions, you'll see they wiped out a fair-sized amount of Cleveland, Ohio, when a tank of LNG blew up in the late '40s. There's another one that you will not find if you Google it. They blew up the entire harbor in Qatar in the Persian Gulf when a tank of LNG exploded, back about thirty years ago. That one has been sort of quieted down. Shell designed the steel tank for holding the LNG, and they did it to 1940s specs — only they built this in the 1970s or '80s, and it was a brittle fracture. We knew a lot about brittle fracture, but they were using a 1940s API spec to build it, and it had tremendous welding defects, lousy steel, everything built to a British standard. They blew up the harbor and killed a lot of people.
You can't find out a lot about this. Qatar took Shell to the World Court over this, and Shell won, because their argument was the Emir of Qatar had a lot of enemies — it was really an RPG hit that was fired at the tank by one of his enemies that blew up the tank. Had nothing to do with the lousy steel and the welding defects. I didn't work on it. Professor Pellini [Pelloux], who's retired from MIT, actually worked on that and told me some of the story. It turns out for several decades Shell was no longer welcome to do business in Qatar. Although I have heard that Shell was back in Qatar.
§16. Davao City harbor transfer and the broken tail shaft [66:08]
When they actually got and did the dry docking survey, which was in November — the other one was in October — in Davao City, which is just a little nowhere harbor in the Philippines. I don't know exactly where it is, but they towed it there. Another one of the empty sister ships, coming back from Japan, came alongside, and they actually transferred over a three-day period all the LNG from one vessel to the other so they could unload this vessel and take it to Singapore to get it fixed. They were flying pipes and tubing around the world in Learjets just to get all the parts there. It was sort of an exciting thing to make this transfer in a harbor in the Philippines. If they had blown it up, I guess they would have killed a few monkeys, but there wasn't a lot of other stuff there. It was successful and they transferred the product.
She lay afloat in dry dock at this shipyard on the 28th day of October, and he looked at the propeller shaft and found — on the second page it says the tail shaft was found sheared in two pieces and was renewed at this time and found satisfactory. By that they meant they put a spare in. They had only two spares out of this fleet of six ships, and they had made eight ship-sets of propellers. The propellers look something like this. They had a bearing right here, a six-foot-long bearing. They had a thirty-three-inch-diameter solid steel shaft which had broken right here, and a couple of other shafts. That's the propulsion system, and the reason it stopped dead in the middle of the Philippine Sea was because the tail shaft broke. So the question is: why did the tail shaft break?
§17. The shrinkage pipe and the Seattle steel mill [68:26]
It broke in part because there's a big hole in it. Here's a Singapore shipyard worker. This is a big void. This is a fatigue fracture on the outside. This is a brittle fracture in the center. That's the scale. What happens is, when you make the ingot — because you can't make this by continuous casting — when they made the ingot from a killed steel, the killed steel will have what they call shrinkage porosity, and it will create what they call a pipe. You're supposed to cut off the top third of the ingot.
The problem was, the steel mill in the United States that bid on this project was in Seattle. You don't usually think of Seattle as a big steel mill mecca in the United States. This was a fair-sized steel mill, but it wasn't a really big one. I can't remember how many tons the shaft was — maybe forty or fifty tons. This was at the limit of the largest ingot they could cast. The largest ingot you can cast depends on the crane capacity. You've got to be able to lift the mold off the ingot. You've got to be able to move the ingots around. You've got to be able to forge the whole thing. You should cut off the top third, but because they were at the maximum size of their ingots — maybe the ingots were 200,000 pounds, I don't remember; it's been twenty-five years since I worked on this — they couldn't make a bigger ingot. They had actually had to throw out the first two or three that they tried to make, because they found defects after they forged them. They were losing their shirts on this job.
This is not forging of this particular ingot, but it comes out of a steel-making book. It shows what happens when you're trying to forge a big piece of steel. This is an ingot. This is another ingot that's been partially forged and they're going to put it back in the furnace. Some of these black spots — you may end up with oxide that regrew in the furnace. This furnace will be at around 2,200° F. In order to forge the steel, you're going to forge it at around 2,000 or above. In order to heat this up, it may take a week in the furnace. Remember, X squared is proportional to the time. I told you ten inches is about as fast as you can do that — might be forty-five minutes — that's with water cooling on the surface. When you're trying to heat it back up and you've got something that's fifty inches in diameter, it takes a week to get the heat in to get it up to temperature and uniform. During that time, you may form one-inch-thick oxide on the surface. When you forge it, all that oxide is brittle and it will fall off.
The first time I ever went through the forge at Bethlehem Steel — they had just finished forging and putting some stuff in, and two guys, it was July, it was hot, and two guys bare-chested went in there with shovels and started shoveling the slag off the floor where they'd just done the forging and throwing it in a dumpster. They were throwing this hot slag — probably at that point only 1,000° F, glowing slightly red — just throwing it over their shoulder. As one guy would go down, the other would come up. They were like two cylinders in a two-cylinder engine, going up and down in perfect tandem. One of them, of course, was throwing it over the other guy's shoulder, because the dumpster was over there. I looked at that as I was a twenty-four-, twenty-five-year-old kid, and said, "How can they do that?" The guy says, "They've been doing it for thirty years, so they learn to get in sync. And if you didn't learn to get in sync, you got a face full of thousand-degree slag." So there's incentive to learn how to do it. And never make a mistake.
When they went to forge it, the problem was they hadn't cut off enough of the end. Just to give you another idea of this — this is just a big chain link. Here's a man standing here, and this is what holds it when it's about to go into the forging press to break down that ingot. This is the same thing going into the furnace. This is the forging press. You guys saw forging downstairs just like this, right?
This is the other end of the shaft in Singapore. Some hard hats there. And here is the pipe. This is the shrinkage pipe. They didn't cut off enough of the top of the ingot because they really didn't have the ingot size capacity in this mill to make this product, but they wanted to get the business. They bid on it. They thought they could do it. The first two or three they had to scrap because they found defects when they forged them, and they were losing their shirts because they had already scrapped three.
§18. The ignored inspection report [74:41]
What happened is the technician, the lab guy, went out with his ultrasonics. He does an inspection on the end and he sees this football-sized hole in the middle of the shaft on his ultrasonics. He does exactly what he's supposed to do. He goes to his supervisor and says, "I found a defect in this shaft." And his supervisor throws him out of his office. He says, "I don't want to hear about it" — because the supervisor was getting pressure from his boss because they had already had to scrap the first couple. So what does the guy do? He goes back, he writes it up. He has a little drawing. He shows the little circle on his ultrasonic indicator. He shows it's twelve inches in diameter — the indication he found with his ultrasonics was twelve inches in diameter. He puts it in the file. They ship it to Quincy, Massachusetts.
Back then, people were learning not to rely on your supplier for quality control. So they didn't do any incoming inspection at Quincy. They put the shaft in, it worked for three or four years until the fatigue crack grew and it broke in two. Quality control works — but you have to have people who are willing to accept the consequences of the results. They weren't, in that case. So it ended up costing a little bit more money when the failure occurred in the field.
§19. The clink defect [76:21]
Now, the rest of the story — I was going to tell you about the clink. This defect was a two-part defect. They sent me to Singapore. I go out there in a dirty old lay-down yard and I take this picture as it's sitting there. They had excised a piece. You had the fatigue fracture up here, the brittle fracture around here. They had excised a piece for mechanical property tests on the steel, and they did it right at the interface between the fatigue crack and the brittle fracture. Here's your cavity, which was because the ingot had this shrinkage porosity. Why did the brittle fracture occur?
This is what's called a clink defect. I will tell you that in the last twenty-five years, when I've met other metallurgists — not that I talk about this every day, but it could be a couple dozen metallurgists I've talked to and asked them if they know what a clink is — I've only found one metallurgist who knew what a clink was. I didn't know what it was. Professor Pelloux, who did the Qatar thing, didn't know what it was. I was only thirty-five years old when I saw this for the first time.
A clink occurs — they hadn't necessarily machined it like this, but they had this big ingot with a hole right here. If you let the whole thing cool down to room temperature, you get brittle fractures in steel around room temperature. If you put this in a hot furnace at let's say 2,000° or 2,400° Fahrenheit, what will happen — it might have to go into the furnace for a whole week before you forged it. The inside doesn't heat up right away, but the outside starts to heat up. The outside, as it expands, goes into a compressive stress state. If I'm plotting the temperature gradient across, the temperature is high on the surface, still cold in the inside.
What happens as you heat it up? The outside goes into compression. What happens to the inside? Goes into tension. You have to balance your forces. So I have a tensile stress state here with a defect right here. I've got tension on this defect caused by compression thermal stresses. The inside is still at room temperature. This steel has not been worked. It doesn't have fine grain size. It's still fairly brittle even at room temperature. They call it a clink because you can hear it — it's like a rifle shot — clink — when the brittle crack forms. So in the old forge shops in the 1920s and '30s they called them clinks. It's an internal flaw.
So when it got to the shipyard, it wasn't just this big void. It actually had a crack. If you want, I'll make it a V-shaped crack — running all the way up to where the fatigue was. When they put it in service, the only solid metal was that fatigue area right there. That crack was already there, and it formed in the steel mill. The guy on his ultrasonics probably saw something here, but this was so thin when he's shooting from the far end that he didn't really see the crack. He only saw the big void. He could have turned his sensitivity up and he might have seen that crack, but his boss didn't even want to hear about the football-sized void. So why bother to do any finer searching?
Student: Does the fatigue crack only happen when there's a void there, or does it happen at other times?
It can happen other times. It's easier because the void creates the stress concentration. You get clinks when you have a stress concentration. What they do to avoid clinking is, when you first put it in a furnace, you put it in a first furnace that may be at 1,000° F. You put it in that furnace for two or three days until the inside starts to warm up above the ductile-brittle transition temperature. So when you put it in the hot furnace, the steel is not brittle. But the steel when they first put it in a big furnace like this was brittle. They may have actually done this — we never got to this type of detail from the steel mill. But because of the void, there was a problem.
The reason I'm telling you this — a number of times in the last thirty years, when you're dealing with really heavy section stuff, and not even such heavy section stuff, you'll often see in the specification it says you want to heat up something to 2,000° and it will say first heat it up to 800, 900, 1,100, hold it for so many hours, and then heat it up all the rest of the way. This is the reason. They learned this back in the '30s and '40s. By the time we saw this in '85, it was an old metallurgist in Britain, Pat Smedley, who knew about clinks. Reggie Pelloux, who had studied in the '50s, didn't learn about clinks. That was old ancient technology. People had solved that problem years before. The problem is, as George Santayana, the philosopher at Harvard, said, "Those who cannot remember the past are condemned to repeat it." They ran into a clink problem, and it all had to do with trying to heat the thing too fast, setting up bigger stresses. They had a flaw, and whether they did a two-step heat treatment or not, they ended up getting a crack — a clink.
Frankly, if they hadn't had this fracture, that football shape never would have been a problem — because, those of you who are submariners, you know you use tubular shafts anyway because you want to get the weight out, and it's only the outside surface of that shaft that's got the big moment of inertia. You actually drill the center out. So that football hole was not the reason for the fatigue crack here. It was because we actually had a crack here of about two feet in diameter on a thirty-three-inch-diameter shaft.
Student: Oh, okay. So if you drive your car real hard and shut it off, that ticking sound — is that the same thing going on?
No. This is all thermal stresses. Unless something in your car is getting up to a couple thousand degrees. I suspect you have a bearing problem. If you're hearing clink, clink — you may hear a clink, but it's a different clink.
I was actually down in a metallurgical shop in Dallas, Texas, once. They had a shaft about this big. It wasn't thirty-three inches, may have been twenty-four inches. They didn't have a big football in the middle, but it was a brittle fracture followed by a fatigue fracture on the outside. I asked the metallurgist who ran that shop, "That's an interesting fracture you've got out there in the back." It was just laying there. We were there for some aircraft part to look at. He said, "Yeah, we haven't been able to figure out what did that." I said, "What did it come off of?" He says, "That was one of the shafts for a turbine for a big windmill." All the big windmills generate electricity and have big shafts. I said, "Well, I know what that is. That's a clink." This was probably ten years ago that I saw my second clink. I've only seen two clinks in my life.
§20. Big-shaft forging and the limits of capacity [85:15]
First of all, you're not going to see them on little shafts. You're going to see them on really big shafts. And if you think about it, we don't do as many really big shafts as we used to. The forging shops that do these things don't exist like they used to. That was actually one of the projects I had when I was at the steel mill. They asked me if we could weld together two forty-inch-diameter cylinders — basically two ingots or two forgings. The reason was: the world's largest single part made out of metal is a 750,000-pound — so 375-ton — generator rotor shaft, and this is for like a 1,000-megawatt electric generator. So this is a General Electric or Siemens-Westinghouse. These are the largest electrical generators made in the world. Outside of maybe the Soviet Union — I'm sure they could do this if they could make decent quality steel — there's one facility, Creusot-Loire in France. There may be one facility in Japan, maybe Tokyo Steel. I can't remember the name of the big steel company that does the big forgings in Japan.
There were two in the United States. One was the Bethlehem Steel plant in Bethlehem, Pennsylvania, and the other was US Steel. Those facilities back in 1975 would have cost $2 billion to build new. They didn't want to spend that kind of money. You could buy half of a steel plant for that back then. Now to build a facility like that might be ten or fifteen billion. That's the whole facility to forge the shaft. This is not the steel-making part. This is just forging it, machining it, heat treating it. And the heat treatment — you can't just stick these things in a furnace and lay them horizontally. You actually have to do it vertically. This shaft might be forty or fifty feet long. So you've got to have a forty-foot furnace, typically buried in the ground, going about four stories down.
The reason we had the one at US Steel and Bethlehem Steel is because they'd been put in at the beginning of the twentieth century to make battleship gun barrels. Remember, they used to always want to have two facilities. Actually, you guys don't remember that, but back even twenty-five years ago, that was one of the big things. They always wanted to have two shipyards. Actually, you still have two shipyards for subs. But that's one of the only legacies that still has that. You don't have two shipyards for carriers. The subs are more conservative than the carriers.
Bethlehem Steel and US Steel — the facilities had been put in like 1910, 1915, to make battleship gun barrels. But now they were the only two places outside of Japan and France that could make these huge generator rotor forgings. At the time, the growth in electricity demand was such that the world was not going to have enough capacity, because you could only make something like a half a dozen or a dozen of these things a year. It takes two weeks in the furnace to warm it up. You may have multiple furnaces for forging and heat treatment, but just to put one through the whole process might take more than a year. You might have a dozen of them in production, but at different stages of production. There was no way based on heat transfer to get more production. So they wanted to know if they could make smaller ones and weld them together and make big forgings. I had to figure out how — could we weld forty-inch-thick steel shafts together?
We never did it. No one wanted to spend the money to actually do the trial. You could have used electroslag, you could have used narrow-gap welding. There were a couple of welding techniques that could have done it. Every now and then people run into these things. About twenty years ago, the guys over here at the MIT Plasma Fusion Center came to me and they wanted to weld twenty-inch-thick by about four-foot-long stainless steel together. This was going to be for the top dome of the International Thermonuclear Reactor that they're building in France now. But there was no place they could forge something this big in stainless steel. So they wanted to know — actually, if there was, it was going to be too expensive — if they could weld a bunch of smaller pieces of steel together. So every now and then someone wants to weld a piece of something that's twenty inches thick. Can it be done? Yes. But you can't make a lot of test welds and have them screw up.
§21. Closing the Libra case [90:46]
Just to show you the microstructure in the Libra steel — it had not only a brittle steel, but the microstructure was fine from the heat treatment and everything. But it had small cracks that had porosity, and cracks that had formed during the brittle fracture, and cracks growing from those defects. So that was the cause of the Libra failure. It's an interesting story in casting technology — the pipe, a shop that bid on a project that was a little too big for them, and an inspection procedure which caught the defect but didn't get past the managers who were afraid of losing money.
Student: When it was found that was faulty, that was deliberately sent out?
Yes, that was why I was involved. The legal action was such that it was clear the steel mill was guilty of making something bad. But — I don't remember all the political reasons — the ship owners, the company based in New York, basically wasn't getting along with General Dynamics, the ship builder. This is the commercial ship builder down here in Quincy, but General Dynamics owned that yard. Originally that yard was owned by Bethlehem Steel. They had built it for World War II. They built a lot of Liberty ships down there. Bethlehem Steel had sold it to General Dynamics, and they had a number of problems. This was a really high-technology ship to build, these LNG tankers. They were some of the first really big LNG tankers in the world, and they had a number of problems with them.
There was a lot of bad blood between the shipyard and the owner. So the owner decided to get together with the steel mill, and the two of them legally went against the shipyard. It was like a $15 million loss, and I think the shipyard ended up paying six or seven or eight million — even though their problem was that they didn't do incoming quality inspection. But it was really bad blood between the shipyard and the owner. I always felt that the steel mill in Seattle had really gotten away with murder. Not murder — well, no one got killed. I'm going to turn the tape off for just a second.