§1. Ashby's plot revisited [00:43]
Good old Ashby. Michael Ashby is a real leader in materials selection and design, and he's a great material scientist. I'm not going to try to knock him for that — I'm going for something else.
I showed you on the first or second day of class this Ashby plot — materials used through the ages. It turns out I was wrong about the date. I went back to his books. I thought it was from one of his 1980 books; it was actually from Material Selection and Mechanical Design, and the first time I could find that plot was 1992. He has books going back to 1980, but they don't have that plot. So here's the plot, and that's where my slide came from.
[Tom struggles with the projector for several minutes.]
So Ashby came up with this plot in 1992. The first edition of Material Selection of Mechanical Design was 1992; the fourth edition is 2011 or 2012. When I presented this I said he was extrapolating around 1980 or 1990 and saying the amount of metals is decreasing and is going to continue to decrease, and the amount of other structural materials is going to increase over time. I happened to buy the fourth edition a week ago, and flipping through it, what he says in the fourth edition is he sort of corrects that.
In the caption he says, "The evolution of engineering materials with time, relative importance" — which he's plotting on the vertical axis — "is based on information contained in the references from 1960 onward, data for teaching hours allocated to each material family at the United Kingdom and US universities." So this is basically — and it's true — when I was a student in 1970, the curriculum was metallurgy. It was metallurgy and material science, and it didn't become material science and engineering in this department until 1973 or 1974. So it was primarily metallurgy, and that's correct. Except that's not what he said in 1992.
In 1992 he said: "The evolution of engineering materials, this evolution increasing pace, are illustrated in Figure 1-1, the materials of pre-history." He's talking about the fact that everybody taught metallurgy back in the 1960s. "There had of course been developments in other classes of material, Portland cement, refractories. The rate of development of new metallic alloys is now slow, demand for steel and cast iron has actually fallen." That's wrong. It has never fallen. Maybe in one year, but if you draw a trend line over a decade it's been increasing, and it continues to increase.
"The polymer and composite industries on the other hand are growing rapidly." Yeah, because they were so small. It's not hard to grow rapidly when you're small; it's very difficult to grow rapidly when you're big. It's what your baseline is. "And projections of growth, production of new high performance ceramics suggest rapid expansion here also." Well, this is right after the ceramics fever — they're going to make jet engines out of ceramics and all this other stuff. You're supposed to be the material property guy, you're supposed to know about strength and toughness — you've got the plots in your book. In fact, he was on the same bandwagon as everyone else in 1990, saying metallurgy is dead. The last department head in this department was a polymer guy, and he actually said metallurgy was dead, which I didn't particularly appreciate. Other materials are important and I'm going to talk about some of those.
§2. Selection criteria and the Spruce Goose [09:06]
A long time ago I gave a talk on material selection at a conference. If you look at different materials — strength to density ratio — aluminum is better than steel for typical strengths and densities. Composites are right in there, equal to steel. Very fancy composites can be significantly above, but they're materials that cost ten thousand dollars a pound. Plastics are relatively low even though they're the lightest weight of all these, as is wood.
But if I'm going to build something like an aircraft, you look at the strength to density — the specific weight of the material — we make aircraft out of all these things. One of the biggest aircraft ever built, Howard Hughes's Spruce Goose, was made out of wood. In the 1940s when you wanted to make the biggest aircraft in the world, you didn't make it out of aluminum, you made it out of spruce. They only built one of them, but it worked.
So there's competition for materials. From that same talk, on selection of structural materials: the number one thing I've been telling you is cost. Strength and fracture resistance — there's your strength and toughness — and that's what I've been saying is after cost what we're worried about. We really didn't worry about fracture resistance until the last 25 to 50 years, and some industries haven't even caught up yet. Availability. One of the strengths of steel is there's a lot of iron out there, and it's not hard to come by. That's the strength of aluminum too — aluminum takes a lot of energy, but they're both available, and that's why they're the two most common metals.
Wood — really good quality wood, we're running into the same kind of problem they had back in the 1500s in England. The really good big trees, the high quality trees, have sort of been deforested, and we don't have them like we used to. So people are using composites and making things look like fake wood. Fabricability — if you can't machine it, you can't cut it, you can't weld it very easily, that's one of the problems with ceramics. I showed you the ad for the watch, the little bezel. Ceramic is very difficult to machine. That's why you've got to buy a forty-thousand-dollar watch to have them make it out of ceramic.
Recyclability. The most used metal is steel — a billion tons a year — and steel is about 80 to 90 percent recycled. Everybody thinks of recycling of aluminum, people picking up beer cans and soda cans, but aluminum is only 50, 60 percent recycled. It's not recycled as much as steel. There are a number of reasons. The aluminum alloys are so different in their classes that if you try to mix them all together you end up with a mishmash that you can't use for much of anything. People are working on new ways to separate the alloy content out of the aluminum alloys.
But there are two materials that are higher than steel in usage. They're not metals. Stone, and cement — and wood is there too. You go to the hobby shop — Ken Stone recycles wood. He made some things for me, all from recycled wood. He used to use the old MIT desks. The legs were solid cast iron, and the tops were bird's eye maple, beautiful maple, two inches thick solid. We got rid of them rather than refinish them. Beautiful wood. So Ken recycles that. But mostly wood is consumed. It's a renewable resource.
The one we make one and a half billion tons a year of that we don't recycle is cement. You can't keep putting one and a half billion tons a year of big old rocks — once you break it up, you have an earthquake, you destroy the building and send it to a landfill — you can't keep landfilling one and a half billion tons a year. You want a problem to work on, figure out how to recycle cement. At some point in the next 10 or 15 years people are going to realize you can't keep landfilling all the cement. We've had tremendous growth in the use of cement — it used to be half a billion tons, now it's one and a half billion tons and continuing to grow, particularly in the third world.
Repairability. This is one of the Achilles heels of all your fancy composites. How do you repair them? You went to all this trouble and processing and extra expense to make this fancy material, and then you have a hard time. There are other criteria too — I actually have a list of 10 or 12 different things.
However, even Rudyard Kipling knew that — "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." So what I've been telling you is not new. It's been known for a long time.
§3. Productivity and the MIT Commission [17:11]
I want to switch into a topic that's not necessarily material selection, but I lecture as I feel like — productivity. I'll tell you a little bit about Made in America: Regaining the Productive Edge. Michael Dertouzos was a professor running the computer science lab at MIT. Richard Lester has been an administrator around here for 20 years. And Bob Solow is an economics Nobel Prize winner. This is the MIT Commission on Industrial Productivity. As a younger faculty member I got involved in this. It started about 1985 at a meeting in the School of Engineering conference room. It was published in 1989.
In the mid-80s the Japanese were knocking our socks off in manufacturing, and everybody thought the problem was regaining the productive edge. In 1984 and 85 I went to Japan for a year on sabbatical working for the U.S. Office of Naval Research. I was visiting research labs and manufacturing facilities all over Japan for a whole year, and I said, this is not better than what we have in the United States. It might have been making better quality, but at the time there was a significant competitiveness issue. It was 250 yen to the dollar after World War 2 — it had been like 350 yen to the dollar. Today the exchange rate is about 100 yen or 90 yen to the dollar.
Back in 1985, 250 yen to the dollar, a 10,000 yen note was a $40 bill. My housing was pretty expensive — the U.S. government paid $36,000 for me to rent a house in Tokyo for my family. Food was expensive, particularly if you wanted to eat like an American. Back then a box of Cheerios could cost you $15. That'd be like $35 or $40 today. Because they had to import it from the United States. If you wanted to eat like a Japanese you could eat pretty cheap, and if you wanted to eat like Chinese you could eat really cheap in Japan.
I had a certain advantage — I was a U.S. government employee for that year, and I could buy things at the military exchange. C-5As would come over from Seattle on the weekend. You'd go out to Yokota Air Force Base and there'd be 10 of them, and this is the National Guard basically taking a quarter-million-dollar flight from Seattle to Tokyo to get their training time in. What were they offloading? Big pallets of Coca-Cola. So I could buy a case of Coke in Tokyo cheaper than I could in Boston, and they'd been flown across by the military.
I remember the first week I was there, I showed up at church and they said — I was the only person who was a government employee at the time — can you go down to the embassy and buy us hot dogs, because they were going to have a social and they wanted 300 hot dogs. A hot dog in Japan would cost you like two bucks apiece, just for the hot dog without the bun. I could get them at the commissary for 50 cents a pound. So I had to go down to the commissary and buy all these frozen hot dogs. One thing I learned: if you're ever going to live abroad, live under the U.S. State Department rules. Those people know how to live abroad.
So we thought that the Japanese were these master manufacturers, and it never quite made sense to me in the mid-80s from what I'd seen over there. They had lots of inefficiencies built into their system. Finally in the early 90s I read something in The Economist that pointed out the United States was the most productive nation in the world in terms of person-hours per whatever metric — tons of steel, pounds of aluminum, computer chips. We actually had the competitive advantage in almost everything, including agriculture at that time. Now Thailand is the most productive rice grower because they've got the right climate and soil. But we are the most productive nation in the world, and have been for the last century. We passed the British.
The important thing is to read the very beginning of this book: "To live well, a nation must produce well." It was written by a committee led by these three people, and I served on one of the committees that looked at the steel industry. There's a section in there on the steel industry. Paul Krugman — does anybody know who Paul Krugman is? He's an economist. He did all his really good work here at MIT. He wrote some very good books that were understandable and used correct principles. He then went to Stanford, which is when he won the Nobel Prize, and now I think he's back at Columbia. He's an MIT product.
So here's Made in America: "To live well, a nation must produce well." Paul Krugman used to say, "productivity isn't everything but in the long run it's nearly everything." That's where I talked about nails. I have a plot here on productivity of steel — hours per ton on a log scale versus time on a linear scale.
§4. From Bessemer to the basic oxygen furnace [24:26]
Steel was very labor-intensive back in the old days, and then the Bessemer process came along — we learned to melt steel. Before, we basically either started with cast iron, or if we made wrought iron we basically diffusion-reduced it from the ore and ended up with a sponge which we then had to forge by hand into wrought iron — which was steel basically, a lower-carbon cast iron. Then Carnegie came along with economies of scale and built huge steel plants, based initially on the Bessemer process but later others — the open hearth. Then in the 60s came the basic oxygen furnace, continuous casting, mini mills which were smaller, a way to use a lot of scrap.
We went from labor-intensive in 1975, when I worked for Bethlehem Steel — labor was 45 percent, materials 45, and they actually made a 10 percent profit. In the 1980s it was 55 percent and they made a 10 percent loss. That's another story. Raw materials and energy intensive today. That's what it takes to make steel. Doesn't take much labor — we're down to 20 minutes per person-ton.
Why? Let's talk about what happened in the steel industry. People asked me about that before, and I went through a little bit of it but didn't give the whole story. I told you the story of Saugus Ironworks and the problem they were running out of wood in England, so in the 1620s they started an ironworks here, the first one in the United States because we had a lot of forest. They needed the energy, which came from wood.
This is something out of the Metals Handbook on casting — development of foundry technology in the United States. Here's a plot of the Chesapeake Bay in the 1700s. Northeast Forge in 1735, principal forge and furnace in 1724. In the 1700s there were all kinds of forges around the Chesapeake Bay. This became the center of the iron industry in the 1700s, which was the beginning of the metals industry in the United States. Saugus failed in the first 10 or 20 years and went out of business, eventually because of the wood. In Virginia they had all the water power down on the Rappahannock and the other rivers — and you could eat oysters for lunch. They used to feed oysters to the slaves and to the lower-paid laborers because they were so plentiful. Rich people didn't eat oysters — that was poor people's food. Nowadays I had an oyster yesterday in San Francisco that cost me three bucks apiece.
Let's talk about some of these processes. The Bessemer converter came along. You couldn't melt steel with a regular fire. You can stick a poker of steel or even cast iron in a hydrocarbon fire burning in air — we make furnace tools, the hearth and stuff, out of cast iron, and it doesn't melt. But Bessemer built a converter where you had the liquid iron, and you'd blow the air in, and the exhausting air would come out and sort of preheat the air coming in. It was a very crude way to preheat the incoming oxygen. If you preheat your oxygen or your fuel, you get a higher temperature flame. With this particular furnace design he was able to blow air in, get this kind of forced heat exchange, and melt steel.
That wasn't very efficient. Then came the basic open hearth — what Andrew Carnegie used — where you just had a big furnace with a layer of liquid iron about three feet thick. You start figuring out the density of that iron. We used open hearths from the 1880s to the 1970s. When everyone else was going to the basic oxygen furnace, U.S. Steel — being the conservative people they were — in the 1970s put in the world's last open hearth furnace at 450 tons, probably the world's largest steel furnace ever made. They were going bigger and bigger in scale.
To do this and preheat the air in a big furnace, first it has to be a fairly shallow bath because you're getting the reaction of burning the carbon out of the cast iron which comes from your blast furnace. You blow the air in and the oxygen burns away the carbon, leaves behind an iron liquid with less carbon. You control the time of burning to get the carbon down to what you want, and then you can lift up this whole — it's about a thousand-ton furnace if it's 450 tons of iron with refractory and walls. Some of these would actually tip up — pretty massive things. Some you would just tap out of a little hole. You build ceramic, put a little piece of dynamite in there, blow out the hole, and all this liquid iron comes rushing out — pretty hard to stop. You have to have something to catch it, and if you don't, you have a mess.
§5. The Copperweld disaster [30:37]
I remember one of my classmates one summer worked for Copperweld Steel. Harvey came back and told the story that the foreman in the melt shop was an hourly employee — the kind of guy who maybe graduated from high school but he was making a lot of money. When I worked at Bethlehem Steel, the foreman in the melt shop was an hourly worker probably making forty, fifty thousand dollars a year. I was making about twenty thousand dollars a year as an engineer. The foreman had to turn out about 300 tons of steel every hour, he was responsible for it, he had about 30 years working in the steel industry, and he made more than almost all the college professionals — not as much as the plant manager.
So the guy's sitting there in the control room, and the engineer comes in, looks up at the gauges and says, you haven't tapped the heat, I want you to go ahead and tap the heat. Now this whole thing is about a 500-ton vessel with a couple hundred tons of steel, on a trunnion with great big bearings — you basically tip it over and pour it into a big ingot on a big railroad track, then carry it away to cast it in the molds. The engineer says, you haven't cast the heat, I want you to go ahead and cast it. The guy says "but —" and the engineer says I don't care, I told you to cast it, do it, and walks out. The foreman's an hourly employee, the engineer makes more than him. So he does what he's told, he taps it. Problem was, there was no railroad car there yet — that's why he hadn't tapped it. So he taps it right on the floor of the casting shop. The floor of the casting shop is usually just dirt, but it takes about a week to clean that up, and five or ten million dollars, because you've now welded all your steel parts in your cast shop with this big puddle of 200 tons of molten steel which solidifies. You have to wait for it to solidify and then go in there with oxygen lances and cut it up. Both of them lost their jobs. The moral of the story is, have enough respect for that hourly employee and listen to what he means when he says "but." He may not be completely stupid.
The problem with the old open hearths is you had to preheat the air. You had a basic open hearth which could be a quarter the size of a football field, and then two big buildings of two or three stories of brickwork — a big lattice. Initially you'd fire it up once every two or three years and run it for two or three years before shutting down to replace the refractory. The exhaust gases would go through this brickwork, preheating the bricks. The brickwork might be half the size of a football field, all inside a chamber, and then the gases would go up the stack. From about 12 hours before, you had preheated these bricks, and the air comes in — that's how you preheat the air going in — and then you switch over every 12 hours. You're just using the thermal mass of these bricks to contain the heat. You're taking your waste heat going out to heat up bricks, then you switch over and your air coming in is heated by those hot bricks. Fairly inefficient process — took about a day to produce 300 tons of steel.
In the early 70s — though in the late 1950s some Austrians had developed a process where they took pure oxygen and blew it through a lance onto the steel, and in 30 minutes they could burn all the carbon out, where the old process took a full day. There's about a factor of 50 in productivity. The BOF shop has a footprint one quarter that of the open hearth shop. Everybody in the world — the Japanese and others — except U.S. Steel and Bethlehem Steel, went to the BOF. The American companies were 10 or 15 years behind because they were used to doing things the traditional way. The BOF is only filled about ten feet deep and it's about three stories tall.
When you actually blow the oxygen in there at sonic velocity — it's liquid oxygen, and an oxygen company will build a refrigeration plant right next to the steel mill to supply it, whether it's Linde or Air Products. The whole thing turns into a froth two or three stories tall of molten steel and oxygen reacting and burning rapidly. Pretty dramatic to see from a distance.
When I was at Bethlehem Steel, one guy retired and decided to commit suicide, so he came back and jumped into the bath. You float very well on steel, by the way. They could see the body spreading gold on top of the bath. In a case like that, what they do is bury one ingot in the cemetery. The only problem with the steel is the phosphorus increases, because we have a lot of phosphorus in us. Phosphorus is an impurity in steel — so that had to degrade the steel a little bit.
After the BOF converter, you may do argon bubbling to get even lower carbon. You can do degassing — actually pulling a vacuum using this venturi technique that was developed at MIT mechanical engineering in the 1940s. You can degas the steel, pour it into a ladle, transport it by the cranes into a tundish, which is stationary, a holding vessel, and they control the rate of casting into this little vibrating copper mold which might be about 10 feet high, water-cooled copper. The steel comes out red hot and you actually can bend it with rolls, turn it horizontal and cut it off into big slabs about 10 inches thick.
§6. BOF chemistry and the Chipman legacy [37:23]
Student: [Question about the oxygen reaction.]
You're shooting the oxygen in there and there's so much surface area between the gas and the liquid that you get a very rapid reaction. First, there's no nitrogen around to mess things up. The oxygen combines with carbon to form carbon monoxide. One of the reasons it's nice and dramatic is you get a carbon monoxide gas jet coming out that's about 10 or 20 feet tall. You don't stand right next to it, you stand about 30 or 40 feet away. You can feel the radiant heat at 3000 degrees Fahrenheit. You're getting huge amounts of surface area between liquid and solid, burning that carbon out — a little controlled explosion, a huge deflagration.
Student: [Question about whether you also burn the iron.]
No, it doesn't, until you get down to very low carbon. The thermodynamics are such that you can get down to about 0.05 percent — about 500 parts per million carbon — before you start to oxidize the iron. The people who worked all that out: John Chipman here at MIT, who was head of the materials department, and after him John Elliott who was one of his students, and Tom King who was department head when I was an undergraduate. All the world's steel-making technology, the chemistry of it, was developed here at MIT by John Chipman.
John Elliott told me that when Nixon opened up China, Elliott in the late 70s was one of the first scientists to go from the United States. He visited one of the universities there and they had a book called Elliott and Gleiser. John Elliott and Molly Gleiser had written it on thermochemistry for steel making. They had it enshrined in a glass case at the library, because Mao Tse-tung didn't allow them to purchase European things. Harold Larson's got a case of those volumes down in the basement. He was a student with John Elliott at the time.
We have a Chipman room, and over half the chaired professorships in this department come out of the steel industry, in part because of John Chipman. Back in the 1920s, 1930s, 1940s, it was sort of a crapshoot to get the right chemistry for the steel. You'd melt it in the basic open hearth furnace and hope through trial-and-error experiments you got the right chemistry. You could end up with 300 tons of the wrong chemistry steel, which you'd have to downgrade or re-melt. John Chipman came along — he was a physical chemist from Georgia Tech — and applied the same principles of physical chemistry to high-temperature physical chemistry, not aqueous chemistry which is what the chemists were doing in water baths. He applied those principles to high-temperature slag baths and steel, figuring out how the elements partition between the slag phase and the metal phase. Now we can hit those chemistries right on. You get one bad chemistry out of a hundred heats today, whereas in the old days you'd get 10, 20, 30 bad heats.
That's one of the things about productivity. If you go in the Chipman room you'll see the Benjamin Fairless Works award. The world's largest steel mill at the time was the Fairless Works of U.S. Steel. There's lots of metallurgy that goes into the continuous caster — the tundish where the stuff is flowing around, the water-cooled copper molds, solid flux and liquid, a vibrating copper wall that's contracting. You actually have the molten zone supported by rolls, and sometimes you'll have a breakout. This little continuous caster you see here is about 10 or 12 stories tall — a pretty good size building. You want to build a continuous caster like this today, probably about 2 billion dollars.
If you have a breakout you're going to drop your whole tundish right onto the floor, just like they did at Copperweld — but that guy did it sort of intentionally because he was irritated with the jerk engineer. There's all kinds of microchemistry going on — solidification, dendrites, pushing of inclusions. Tremendous technology has been developed in the last 50, 60 years, but most of it was developed between the 1940s and 1980s. Once the steel industry lost its productivity, they closed the research labs. Steel research went down, there weren't any jobs. Now they're coming back and they want to hire students, but there are no students — everybody's working on semiconductors.
Student: [Question about the vibrating copper mold.]
Because you don't want it to stick to the walls — you're trying to get it to move through there, so you vibrate it. You don't want to weld to the copper wall, and you want to reduce the wear on the copper. That copper's got to last for about three or four years before shutdown. The Japanese have run continuous casters for over a year without shutting them down — running 24/7 for over a year. These things have to be fairly robust, and vibration is one of the things —
Student: [Follow-up about space inside the caster.]
Well, you've got these water-cooled copper molds, 10 or 20 feet high, and you've built up a skin half an inch or two inches thick depending on your casting speed. The actual contact is just up in the top. It's cooling away, and you've got radiational cooling from the cold water-cooled copper that's helping grow that skin. One reason you can bend it so easily is that the first part has a liquid core. If you're not careful about the radius of curvature and the casting speed, you can have a breakout — all that liquid steel falls down through everything else — probably a hundred-million-dollar loss. They've got control algorithms and everything else. A lot of incentive to run it as fast as possible but not too fast. Breakouts are not very common — probably not more than one breakout in the world every year or two today.
§7. Mini mills, dumping, and government subsidy [44:51]
When they first started continuous casting in the 1960s, a lot of people said that process will never work. Some people said going from 65 percent yield — pounds of steel produced per hundred pounds cast — to 97 percent is a big enough increase in profitability that they were willing to bite the bullet and do it. American steel companies were too conservative. Boards of directors were not willing to take a risk. It was people who were willing to take a risk, but some of those people were in foreign countries where the government was backing up their risk. It's sort of like the Airbus-Boeing controversy today — how much is the government supporting both of them.
In the United States, well, we don't support Boeing. How much money does Boeing get in defense contracts every year? You don't think that's a way to support things? What about the new Air Force tanker program, which went to a consortium that included Airbus on the first bid? All the Boeing people went to their congressmen and forced the Air Force to reopen the bid. And guess who won that time? Boeing. Now Airbus is pretty much directly subsidized, and they're always fighting about how you're illegally subsidizing your aircraft industry. It's a tremendous export industry. What's Boeing today — 40, 50 billion dollars a year worth of exports, and fairly profitable in general.
Student: [Question about whether Boeing is always subsidized.]
All the time. Half of Boeing is government contracts. We don't subsidize in the same way — we subsidize in a different way. But our subsidy is okay, we say. The Europeans say no it's not. These are some of the externalities. It was a real problem for the American steel industry in the 1980s, because everyone had convinced Congress that the Japanese were not dumping steel — and then the Romanians, and then the Koreans, and now the Chinese. They've been fighting this for 40 years.
Yes, the American steel industry was losing a tremendous amount of money in the 1980s, but the reason — I showed you these plots before — productivity had doubled, consumption was constant, employment went down. That was because they went to BOFs from open hearths, they went to continuous casting, they built mini mills, and mini mills could use 100 percent scrap.
Before the mini mills, they had to use 70 percent virgin iron ore to make the steel from cast iron, and only 30 percent scrap in the BOF. An electric furnace can make steel out of 100 percent scrap. You can use John Chipman's principles to purify the scrap to get the chemistry you want. It's a lot harder in aluminum, because — this goes to your question about burning out the carbon before you burn the iron — if you start burning away iron, you're going all the way back to iron ore, which is a lot of wasted energy. In aluminum, you'll burn up the aluminum before you burn out the copper and other things. In steel, you can burn out a lot of these. Certain things you can't burn out — bismuth in steel goes to the grain boundaries and produces a brittle product.
Fifteen years ago people were trying to get lead out of solder, and they're still trying to get lead out of solder. I'd go to these conferences and they'd say, we've developed this bismuth alloy. I said that's great — so you're going to start putting pounds of bismuth in every automobile, which means we will no longer be able to recycle steel automobiles. That was dead on arrival, but these metallurgists working on solder didn't know anything about steel making. They didn't realize nobody was going to put bismuth alloys into the soldering on an automobile, because you can't — you'd have to separate the scrap. You can't just shred up the automobile and dump it into the furnace, you'd have to separate it out. We do separate some things — we take out the airbag explosive, because that has killed people in the steel mills when they're shredding and all of a sudden a piece of shrapnel goes through someone's arm. They do some things, but you can't start ripping all the wires out — you couldn't afford to recycle the steel.
§8. Chaparral Steel and the 3/8 inch rebar [50:10]
What happened in the 1970s is people like Gordon Forward and Ken Iverson realized you could buy scrap steel at $100 a ton on the open market. There are millions of tons out there. Cast iron cost $200 a ton. You had a $100-a-ton advantage on something that sold for $400 a ton — a 25 percent cost advantage to use 100 percent scrap. Except the quality of the scrap was not very good because it had all these alloying elements, and about the only thing they could make was reinforcing bar for concrete. It just has to have 60,000 strength and be the right diameter. You don't have to have toughness requirements — let's face it, the concrete has no toughness, so you put the steel in, it's tougher than it was.
You could make rebar, but rebar was garden-variety low-quality stuff. And the Bethlehem Steel manager said, "we don't care about mini mills." This is what Clayton Christensen made his name for at Harvard Business School — the innovator's dilemma. Here you have these little mini mills, and one of his three examples is steel technology. My daughter worked researching things for him one summer, and I'm a reference in there on what it cost to build a steel mill. Rebecca came home and said, "Dad, Clayton wants me to find out what it cost to build a steel mill and I can't find it anywhere." This was 1995, and no one had built a steel mill since Bethlehem Steel in 1965. It had only been countries that built them, all mixed in with the government finance.
About a year before, I'd been sitting in the Saginaw, Michigan airport with one of my graduate students, and I'd sketched out on a piece of paper the approximate cost of an integrated steel mill, a mini mill, and a micro mill. A mini mill you could build for like 100 million dollars. An integrated steel mill was 15 billion, by my estimate. It cost Bethlehem Steel in 1965 5 billion dollars. That was their whole year's gross product, and they almost went bankrupt. By the time they hired me, it was the most profitable year ever — that's why they could afford to hire me. But they almost went bankrupt in the previous six or seven years.
No one could take on a 5 billion dollar investment back in the 1960s. That's like Intel taking on a 15 billion dollar investment today, or Boeing a 15 or 20 billion dollar investment. These companies are betting the farm on building a new 787 Dreamliner or a new computer fab. Other countries — it's the whole country. There's never been an integrated steel mill built since Bethlehem Steel Burns Harbor by a company. Plenty of integrated steel plants built, but by countries doing the financial backing. No private investment taking on a 20 or 30 billion dollar investment for an integrated steel plant.
What they did at Chaparral Steel — Gordon Forward is a graduate of this department, a Canadian, retired in British Columbia now. Every now and then, if you were an employee at Chaparral, you had to go spend one week a year with the salesman in the field to find out what the customer wanted. What a unique thing. They found out that the smallest diameter rebar, three-eighths of an inch, got like a 20 or 30 percent premium, because you could only go so fast through the rolling mill. So first they came back and said we're going to roll faster — we're going to take over the 3/8 inch rebar market because there's more profit there. So they tried to go faster.
This stuff is going through there 60 miles an hour, and if you go faster you get instabilities, and all of a sudden you've got hot band steel that breaks and goes whipping around. If someone's there, cuts your arm right off. My oldest daughter Rebecca was born in the hospital in Bethlehem, Pennsylvania — beautiful hospital, because they sent maimed steel workers down there every day. Cut your arms right off.
So they tried to speed it up for a while, and finally they said no, we're doing it the wrong way. We shouldn't try to speed it up, we should try to slow it down. They slowed it down — here it's coming into the rolling stand as one bar and coming out as two bars. They split it in two. They had to go slower to do it, but they got twice the productivity out at a slower rate. They did that once, and decided they could do that again. Here you see two bars coming in, and four bars coming out. They did it twice. So they're running this mill slower, but they're getting four times as much product through. They took over the 3/8 inch rebar market.
I'll tell you about that tomorrow. The quality actually went up — that was another benefit. They proved that with I-beams in less than a year, and that's the rest of that story. You can read about it — not in this book, but there's another book I was involved in reviewing on Chaparral Steel's product development with Steve Wheelwright at Harvard Business School. You can read about it if you're really interested.