§1. Steel productivity from 1600 to today [00:05]
We were talking about steel last time. Now I want to talk a little bit more about productivity. I was thinking about steel productivity this morning — I've actually thought about it over the years quite a bit, but this morning I decided to graph it. So if anyone's looking for a senior thesis, here is the outline.
Hours per ton on a log scale. When they were making nails in 1600 by hand forging, it was about 3,000 person-hours per ton. That's sixty hours a week — people didn't work forty-hour weeks back in 1600, they needed to eat too. And it was consolidating iron. We couldn't melt iron at 1600 degrees centigrade. We didn't have the technology until Bessemer came along. So people would start with spongy solid-state diffused iron, or they'd start with cast iron and just keep beating it in the forge until they burned out most of the carbon, and they ended up with a high-carbon steel. They made swords out of that, but they had to heat it over and over to burn off the carbon that's in the cast iron.
Bessemer came along in 1856 and taught us how to melt steel. Carnegie came along and became the richest man in the world ever, in today's dollars. He just applied economies of scale. I have no idea what the productivity improved from 1600 to 1800, but I'm sure there was some improvement. They built larger furnaces and they learned to do things better, but it was a gradual increase.
When we learned to melt the metal — that's why Bessemer's invention was so great. We could now burn the carbon out in the liquid state, get much faster kinetics, for all you materials-science kinetics folks. Carnegie made a lot of money, but he probably didn't improve the productivity all that much. It was still very much labor intensive. All of this was labor intensive.
Some of the numbers I do know. Back in World War II we were up in the tens of hours per ton, still labor intensive. In 1975, when I worked in the steel industry, it was about half labor and half raw materials and energy in the cost of a ton of steel. Fifty-fifty. But then we came up with the basic oxygen furnace, the continuous caster, mini mills. This was all process technology. And we dropped from let's call it ten hours a ton in 1975 to about a third of an hour a ton in 2013, okay.
So this is all process improvement, and it's changed the whole landscape. It used to be how cheap was your labor; now it's how much is your energy cost. Raw materials and energy costs are directly related too — you've got to use diesel fuel to run that backhoe, and truck the stuff to market. So if anyone's looking for a senior thesis, this could be talking about the productivity and what caused it. Wouldn't be much of a doctoral thesis, would be okay for a senior thing.
§2. The Bessemer converter and the open hearth [03:49]
Let me tell you a little bit about what a Bessemer converter is and why it helped. You've heard of Sir Henry Bessemer. In 1856 he learned — and I didn't know this for years — you make cast iron in a blast furnace, but it's got about four percent carbon in it and it's very brittle. You can make nice castings and pots, but that was all you could make. You could make wrought iron by hand forging. But you couldn't melt steel. Bessemer came up with a riveted steel vessel where you'd blow air in on the liquid. He had a little lip on the end so that when the gas came out, it had to pass the gas coming in. That preheated the air coming in.
The maximum temperature you can get for a flame burning in air with most hydrocarbons is about a thousand degrees C, maybe 1100 degrees C. The nitrogen in the air just keeps the temperature down — that's if you start with room-temperature air. If you started with thousand-degree air, you could get an extra thousand degrees. This is sort of obvious, but it wasn't obvious to everybody back then. I'm not even sure if Bessemer realized what he was doing, but basically he had a way to preheat the air by having the exhaust gases pass by it. Today a heat-exchanger expert would call this a counter-flow reactor, because the air is going one way and the exhaust is going the other. Counter flow, big deal.
So Bessemer did that, and now they could get those extra five, six hundred degrees Fahrenheit — two, three hundred degrees centigrade — that you needed to melt steel. You could melt iron. Now you have the kinetics not of diffusing carbon out of iron in the solid state, but in the liquid state. Materials scientists in here: how many orders of magnitude are there between liquid-state diffusion and solid-state diffusion near the melting point?
See, you never learned anything practical. Typical liquid-state diffusivity is 10 to the minus 5, 10 to the minus 6 centimeters squared per second. Solid-state diffusion, just under the melting point, is like 10 to the minus 12. So you get seven orders of magnitude increase in reaction time. That's pretty significant. That's a productivity increase.
So what Andrew Carnegie did, he started making open hearths. You'd have this container that would hold the iron ore, and you want to melt it. It would be the size of three of these classrooms in area, in a thin layer of steel, because you're going to blow the gases across the surface. I want lots of surface area — this wasn't a deep bath, a shallow bath, because I'm still getting reaction at the interface between the air and the liquid. The air would first go through some brickwork next to the open hearth furnace. About half the size of a football field and about two stories tall, they would make a latticework of bricks inside a building.
They'd blow the air in through these bricks and combust it with the fuel, which was probably coal gas. Then they'd exhaust it out the other way through another set of brickwork, heating those bricks up. About twice a day they would reverse the flow of the gas through the brickwork. So the brickwork that had been preheated would now preheat the incoming air. That was how they got preheated air, rather than the Bessemer converter.
This technology is still used today in glass making. In glass making, you can't have any carbon around — carbon gets in your glass and you get black specks, and people don't like that. So they use natural gas, slightly oxidizing flame, so you don't have carbon soot particles. But they want to get a high enough temperature, so they still — and I've walked through these in glass factories as recently as fifteen years ago. They have this big latticework of brickwork, and they run the gas through one way to heat up, with the exhaust gases going through another set of brickwork. Then when those bricks are hot, they reverse the flow and put the cold gas through the hot bricks. You just keep reversing the flow, and you can run these glass furnaces for four or five years at a time.
Unless — in this particular one, they had one of the reversing doors. They're big steel doors they have to lift, basically a big valve to shut off the gas. One of them got hung up, they couldn't close it, and they started overheating the brickwork. They had to close it within five or six hours, couldn't do it, had to shut the whole thing down. It takes about three or four weeks to start it back up. So they lost a lot of production because of just one worn-out door.
§3. Andrew Carnegie and the $25,000 check [10:08]
Student: [inaudible question about Carnegie]
There's actually a good PBS special on Carnegie that goes through this. Other people were doing it; he was just pretty good at beating down the labor cost. He was actually a wonderful man in many ways. He loved education — that's why we have Carnegie Tech, Carnegie Mellon University. Lots of libraries are named after Andrew Carnegie. He only had thirteen months of education, and he always felt deprived, so he wanted to make education available.
In fact, my grandfather got money out of it. I hadn't told you my story of my grandfather and Andrew Carnegie. My grandfather was born during the Civil War in South Carolina. It turns out my great-great-grandfather is probably the richest man in the South. He owned — I don't know if I should say this on tape — but he owned over 300 slaves. He gave my grandfather a slave when he was born, and that slave kept my grandfather alive for a year during the Civil War, when he got separated from his mother for a whole year. So my grandfather had sort of an interesting life. His father became governor of South Carolina after the war, then secretary of the treasury, and ended up in New York City on Wall Street. My grandfather ended up in Chattanooga, Tennessee, as the mayor of Chattanooga.
He decided Chattanooga needed a university. He got John D. Rockefeller — I don't know how this worked — to commit $250,000 if they could raise another $250,000, so they'd have half a million dollars to start the university. They went through everybody in East Tennessee and they were still $25,000 short. So my great-grandfather, who had been secretary of the treasury, knew Andrew Carnegie and got my grandfather and another guy an audience with him.
My grandfather took the train to New York and met with Andrew Carnegie. This all comes from my grandfather's autobiography. Carnegie said, "Well, I tend not to dabble in John D. Rockefeller's charities." So he's basically telling him no. As they were walking out, my grandfather said, "You know, the people in the South really need better education. We just don't have schools after the Civil War." This is a combat zone — bombed out, burned out. And Andrew Carnegie looks at him and says, "How much education do you have, sir?" My grandfather said, "Thirteen months." Carnegie says, "You come back tomorrow at two o'clock and I'll give you my answer."
So my grandfather comes back with his colleague at two o'clock, and the butler opens the door. The butler says, "Who are you?" My grandfather says, "Thomas Thompson." "Oh, Mr. Thompson — Mr. Carnegie left on the 5 a.m. train for Pittsburgh." The butler says, "Mr. Carnegie's fifth secretary will talk to you." He goes in, and the fifth secretary gives him a check for $25,000 and says, "Mr. Carnegie wants you to know he also only had thirteen months of education." My grandfather wrote that he figured if he had twelve or fourteen, he wouldn't have gotten the money, because he had the same number of months.
But Andrew Carnegie — how did he get to be successful? He just beat those laboring folks and ran them as hard as he could. So he wasn't the greatest employer. He was pretty good as a businessman at cutting deals with the railroads, because that was your big market, and buying out other foundries and other steel mills. He was sort of just like businessmen today making conglomerates. I think he was probably a pretty good businessman. He was a great educational benefactor, but I'm not sure you'd want to work for him.
§4. The basic oxygen furnace and the continuous caster [14:28]
In 1956 in Austria — remember, we had bombed out their steel making. They didn't have any open-hearth furnaces, and these things cost a fortune. You go to an open-hearth site — I've only seen one or two in my life, but they're huge, and they cost a fortune in capital equipment. Some people in Austria got the idea: let's build something that looks like the Bessemer converter, but let's blow pure oxygen on it. They made a small one of five or ten tons. Nowadays we make them at 300 tons, and they're about six stories tall.
They'll build an oxygen plant — someone like Praxair or Air Liquide or Air Products will build a facility right next to your steel mill, because you'll buy all the liquid oxygen they can produce. They sell liquid nitrogen so people can freeze their corn. So they blow pure oxygen, and this 300 tons of steel actually forms a froth of about three or four times its volume. It's a water-cooled copper lance, and you'd better hope that water-cooled copper lance doesn't start injecting water into the vessel, because you know what happens when water hits liquid metal. It explodes — you get a steam explosion. So they do a lot to design these copper lances. You've got water coming in at tens of feet per second to cool this off so it doesn't melt.
It forms a froth, and you can now refine that steel in thirty minutes — what used to take twenty-four hours — because you've got lots of surface-to-volume ratio. U.S. Steel, in all their great wisdom, in the early 1970s built the world's largest open hearth, 450 tons. Everyone else in the world was putting in BOFs. They ran their hundreds-of-millions-of-dollars open hearth, the last one in the world, largest one in the world, for a few years, and then they realized they had to put in a BOF. U.S. Steel did come up with the Q-BOP — "basic oxygen process" — where they inject the air from underneath, which has certain advantages. So it wasn't as if they didn't develop something. But you look at the productivity gains here, and you're seeing factors of ten or twenty or fifty by these process changes.
If I look between 1975 and where we are today, over thirty-five years, over my lifetime — I stole this out of a metals magazine. Here's the converter, the basic oxygen furnace, there's the oxygen coming in. You can see it says slag — you do have a slag in there, but the whole thing forms a froth. Then you have to degas it, because you've got lots of oxygen in this thing and it won't make very good steel. You can degas by argon bubbling, or by pulling a vacuum on it. They do both. Or you can just cast it as is and you get another type of steel, but usually that would be ingot casting. You're going to go to the continuous caster nowadays. You've got to get the oxygen out, because if you didn't, you'd end up with Swiss cheese for your steel.
We can make that in twenty or thirty minutes. It's actually about a forty-minute process, because you have to do the chemical analysis before you pour. You've got 300 tons of steel; you don't want to pour it and solidify it and find out it's the wrong composition. So there's a whole business in that: can you get a detailed analysis of steel in five minutes, from molten steel, to a hundredth of a percent of all your alloying elements? Because you're talking about a hundred thousand dollars an hour of profit in this vessel. Minutes count.
You pour the liquid into a ladle, and then the ladle goes to a tundish, which is just a bathtub to hold the stuff so you get a nice manifold of flow. You've got a mold which is water-cooled copper, and you have rolls, and the steel is still hot — it's still got liquid in the middle — and you roll it. The whole thing is going to be about six or seven stories, maybe ten stories tall, in order to have a large enough radius to roll this ten-inch slab of steel. Then you have to cut it on the fly. They will run continuous casters for a year without shutting down.
Student: [inaudible question]
No, this is actually a detail of the copper. This is the tundish, which has a little slag layer on it. It actually freezes against the copper — it's water-cooled copper. The startup can be a little dicey, but you only have to start off about once a year. Once you get running, because of the cooling, the stuff pulls away from the copper.
Student: Is there a problem with copper contamination?
Yes, you can sometimes get a little copper contamination before the thing starts oxidizing and forming a little slag layer on the initial startup. That can cause what they call copper checks on the surface. So the first fifty tons you cast maybe just go back to be re-melted. The answer is, on a continuous basis no, on a startup basis yes.
Student: [question about shapes]
Yes, I'm going to show you that in a second. They didn't originally — they would just do rectangular bars or round bars, mostly rectangular bars because it's easier to bend them, or slabs, which would be up to a ten-inch-thick slab by eight feet wide. But they now roll I-beams. These are some of the details that go along inside this bath. They show the steel kind of wavy here, and it has a skin. If you have a breakout — breakouts are really exciting in a continuous caster, that means the skin broke and you dump about fifty tons of steel on the ground. So you don't want to be walking under one of these things, just in case. It will kill you. And it typically shuts the caster down for several weeks. You're talking a four- or five-hundred-million-dollar machine here, down for weeks. Millions of dollars of loss.
§5. Mini mills, Chaparral, and four-strand rebar [21:56]
Running continuously for a year at a time, lots of productivity gains there. There were several gains. There was the BOF, which went from twenty-four hours to burn the carbon out of the steel to about forty minutes — twenty minutes to actually burn it out, another twenty minutes to wait to get your chemistry back and pour the steel. You melt 300 tons of steel every forty minutes. Start figuring the profit on that.
The continuous caster and the BOF all came in around the 1960s. The American steel mills, they were still using 1910, 1912 technology. Did I tell you the story about my being a young engineer at Bethlehem Steel? They had a meeting for one or two weeks in the summer — they took all 500 of their new college hires, whether you were an accountant or whatever, and very much segregated those who had a college education from those who didn't, basically the hourly workers from the salaried staff. It was a real rich-man-poor-man type of society.
They'd put us in an auditorium for half a day, and the other half they'd take us down to the plant. We actually got to see how steel was made on large scale, which is pretty dramatic. You also learned why you went to college — you didn't want to have one of those jobs. Wonderful hospital, where my second child was born, because the steel company had built the hospital for all the iron workers who had lost their arm or their leg in accidents. It was just a factory to maim people, back in the old days. Nowadays they all sit in the air-conditioned offices with computers. But just forty years ago, it was a real hellhole to work in the steel mill. And in Andrew Carnegie's day it was a lot worse.
The next big rise in improvement, the thing that dropped it down to thirty minutes a ton or twenty minutes a ton for production, was the rise of the mini mills, about 1975. Some guys looked at the price of scrap, and they realized you could buy steel scrap for a hundred dollars a ton, and it cost $180 to make cast iron. Someone said, eighty dollars a ton extra profit on a three-hundred-dollar-a-ton product, I could make money on that. But they couldn't build these huge things. They started building small steel mills that might only cost $250 million rather than $5 billion of capital equipment. They used electric furnaces. The big steel companies all laughed and said, you can never make good quality steel out of an electric furnace, you have to have virgin iron ore. They laughed at them. Well, the mini mills make about sixty percent of all the steel today. They just kept eating away.
Anybody heard of the book The Innovator's Dilemma by Clayton Christensen? None of you? About ten or twelve years ago this was the greatest thing in business management. Clayton Christensen is now a professor at Harvard Business School. He was writing this book in the mid-1990s, and he's got three examples of the innovator's dilemma. The dilemma is: whether you're Digital Equipment trying to build mini computers while IBM's making mainframes — you're coming in at the bottom of the market, and the innovator was IBM making computers, and they started making bigger and fancier and more expensive computers. Digital Equipment came in and took over the bottom end of the market and ate away until mini computers got to be powerful enough for desktops. That's one example. Another is the mini mills taking over the integrated steel mills. The book became a top-ten New York Times bestseller.
I'm actually referenced in there, because my daughter, one summer when she was in college, got hired by Clayton to do some research, and she was supposed to find out what it cost to build a steel mill. She came home one night and said, "Dad, I can't find anything. I've been looking for a week, can't find anything about the cost of building a steel mill." I said, "Well, I've got a slide I can give you." So the next day I brought in my slide, and that's why I'm referenced. I estimated that it cost Bethlehem Steel from 1965 through 1975 a billion dollars to build the Burns Harbor, Indiana plant. I think I told you, since then no private company has ever built a steel mill, because it would cost ten or fifteen billion dollars today, maybe twenty billion dollars to build an integrated steel mill. Only countries have bankrolled that. Same thing with Boeing and Airbus designing a new airplane. Intel putting in a silicon foundry — these investments get so large that it has to be a country that subsidizes it, because no company can take that type of risk.
You won't know where the data came from. I sat there in Saginaw, Michigan, at an airport with one of my graduate students, after visiting GM back in the early 1990s. I was thinking about this problem of integrated mills and mini mills, and I just scratched out the numbers. I pulled them out of my head, but they're right there in Clayton's book. So if you want to know where's the reference, well, I'm probably as good a reference as anyone else.
So the mini mills — originally they wanted to make the garden variety. The cheapest stuff is concrete reinforcing bar. It just has to be the right diameter, and it's got the lowest price per ton of virtually anything. So this company called Chaparral Steel, which was started by a guy Gordon Forward, who graduated this department back in 1975, saw the hundred-dollar scrap price versus $180 cast iron price, and he was one of the guys who said you can make some money there. He started Chaparral Steel down in Texas.
Student: [comment about spelling]
Yeah. As Gordon says, you only pee once on Chaparral. A lot of people try to put two P's in. This is one of Gordon's slides. They started out, instead of square slabs out of the continuous caster, with the first slabs that just had kind of notches to reduce the amount of rolling in the rolling mill. Chaparral was going beyond rebar — they were going to try to get into the business of structural shapes. Here's what's coming out of the continuous caster as of about year 2000. You can see — here's the I-beam shape. It's almost an I-beam, but it's actually cast.
They were very successful, such that in the 1990s, when everyone else was importing steel into the United States, they were exporting steel to Japan. They were down in 1990 to one person-hour per ton. Net-shape casting of structural drove costs down by one-third, to the point that concrete could no longer compete against steel for building construction. I'll talk about four-strand rebar rolling. The new facility they were going to put in in 1990 was going to be 0.89 hours per ton, and they had sales of $450,000 per employee.
Why is that important? Anybody know what the typical sales per employee of a manufacturing firm in the United States is? It's about $150,000 per person. I had a student do a thesis at the Boeing door product center once. This is where they make the doors for the aircraft. There are five or six doors on a 737 or a 747. There's the door you walk in. There's often the door on the other side where they bring the catering in. There might be a couple of those. There might be three or four emergency exits. There's usually a rear door, ramp and stuff. All kinds of different sizes, from emergency exits to the full-scale door. And in the cargo planes they have doors the size of a garage door.
So the door manufacturing center at Boeing was a very interesting business. It was completely deterministic manufacturing. You knew what your order book looked like three years ahead, and it only took you one year to make the doors. The aircraft got ordered and they started laying the keel three years ahead of time. You had three years to get your doors made. You may not start until two years later, but you knew exactly how many aircraft were going to be built and how many doors were going to be needed and what their shape was going to be.
They operated it in accounting at Boeing as a cost center. I asked, "What's a typical door cost?" Because every rivet hole they drilled three times: first a rough drill, then a reaming drill, then a third time — I can't remember what it was — to get everything just right. It was all hand work, the guy with the drill. I watched all this and said, "What's it cost to make a door?" "Well, we don't know, because we're a cost center, and we only sell to Boeing, so there's no price change." I said, "Well, how many employees do you have?" "250." "How many doors do you make a year?" "750." I said, "Oh, okay, your door is costing on average $50,000." One of the managers says, "Oh yeah, that's probably about right." They didn't have a clue what their doors cost.
I could figure it out, because I knew that typically at Boeing you had about $150,000 sales per employee. You'd better have $150,000 sales per employee in most manufacturing firms. Because if you're going to pay the guy sixty or seventy thousand, then pay the overhead on top of that — it's going to be ninety thousand dollars — you'd better have something for materials and everything else, and making a profit. If you only have $70,000 sales per employee, you're going to be out of business real quick. So $450,000 sales per employee is a very good business. Used to be one of the metrics I would use when I worked in manufacturing. Without people knowing it, I would ask, "How many employees? What are your annual sales?" I'd do the division in my head and figure out right away whether this was a good company or not. Well, maybe that's the problem.
The sun has come out. Let me tell you about four-strand rebar rolling at Chaparral. It was a very interesting company, because — it was the first company I'd ever been to where, if you asked if they made a mistake — I used to ask, "Who got the blame?" The first day I was there, people looked at me funny, like, "What do you mean, blame?" They didn't understand what blame was. I said, give me a break, the managers are always blaming the employees. "Oh, no." They just didn't understand what I meant. I was there for about two days doing a study with a guy from Harvard Business School, and the second day, or that night, I figured it out: at Chaparral, you didn't get blamed for trying something new and failing. You got blamed for not trying something new. They expected you to innovate.
You go through every shop, and on the wall they'd have that week's profitability and sales. You got a bonus based on those sales. So when they went out and bought equipment, those people felt like they were spending their own money, because in a sense they were. They took ownership. Gordon Forward set up a great management system. He had a corporate dining room which was not quite as fancy as the chairs and tables here. They'd bring in a barbecue from some takeout place in Texas, and that was the corporate dining room — you'd eat with plastic forks and spoons. The barbecue is okay, and nothing wrong with plastic, it works for forks and spoons.
That contrasted to Bethlehem Steel, where they had a special executive dining room — I never got to eat there. At the research labs we had a very nice dining room, but it wasn't anything like the executive dining room, and it didn't have the quality of food. There's a great story, the only one I remember in Iacocca's book about Ford. Henry Ford — or whichever Ford was chairman of Ford Motor Company while Iacocca was there in the 1980s — used to love hamburgers. Wherever he'd go around the world, he'd always want a hamburger for lunch. He said, "You know, I never get hamburgers anywhere near as good as I do here in the Ford corporate executive dining room." Iacocca wanted to figure out what the chef was doing. He goes back afterwards and says, "What makes the hamburger so great?" The guy pulls out a prime rib eye steak, puts it in the meat grinder, and says, "Here." Just quality of starting materials.
That's our philosophy at MIT too. Admit good students, you probably won't screw it up too badly. So the guys at Chaparral — another thing Chaparral did, they forced their production employees to spend one week a year with the sales people going to visit customers. They'd ask the customers, what is it you need? So the guys in production would know what the customer wanted, rather than hearing it through the grapevine. That was the philosophy.
When they went out there, they found that rebar comes in sizes. Number three rebar is three-eighths of an inch. Number five rebar is five-eighths of an inch diameter. The smallest rebar is three-eighths of an inch. They found out there was like a twenty, thirty percent premium per ton for three-eighths-inch rebar. They decided, we want that extra profit, and we want to run our rolling mill faster to get the extra profit from making three-eighths-inch rebar. The problem with three-eighths-inch rebar is you can't make very many tons when it's only three-eighths of an inch diameter. You can turn out a lot more tons of five-eighths, right? Five-eighths is 25 over 9, almost three times the tons per inch if you're running at the same rate.
Typically a steel hot mill runs at sixty miles an hour. So they tried to speed up — seventy miles an hour. It turns out there are instabilities. The stuff starts getting waves in it when you go too fast. They couldn't run it faster. One of them got the bright idea, why run it faster, let's run it slower. Let's break the steel into two strands. So here's one strand coming in, and they split it into two strands. Another guy got the bright idea, well, if we can do that, we ought to be able to split it into four strands. Here's four strands at the top. They slowed it down by a factor of two but got four times the output, and they got double the productivity, and they took over the three-eighths-inch rebar market. Good for them.
Can't remember your first name again. John? Johnny? Johnny said his dad always told him to go horizontally if everyone's going up. Well, I'm sort of, if you're going up, go backwards. Whatever it is, don't go in the same direction as all the other lemmings. They probably don't know what they're doing. Actually, they probably do know what they're doing, and they're going to beat you.
§6. Turbine blades — metallurgy and single crystals [39:52]
Let's go to another material. I'm going to talk about turbine blades, and the productivity of turbine blades. The productivity is not necessarily the manufacturing of them, it's the operating temperature. For every fifty-degree-Fahrenheit increase in operating temperature of that jet engine, there's about a five-billion-a-year savings to the commercial airline industry in fuel. It's just thermodynamic efficiency. Delta T over T is your thermodynamic efficiency of your heat engine. Sophomore thermo. Get a higher delta T, get a higher maximum T, you're going to get more efficiency.
Back after they invented the jet engine in the 1930s — didn't do too much. Hitler had some jets, and the V-2 rockets, and some other aircraft jets at the end of World War II.
Student: [comment, possibly about the Me 262 being the first German jet]
And that was the Germans, right? Then we killed a lot of test pilots until Chuck Yeager broke the speed. We had jets after World War II, based on German technology. I didn't know it was the Me 262, but there were the buzz bombs and all that. They thought they were going to turn the tide after D-Day by being able to bomb London again. But they didn't.
Jet engines were operating at about 800 degrees C, 1500 degrees Fahrenheit. They were using materials like stainless steel, then they got into cobalt alloys, nickel-based alloys. The operating temperature here is the firing temperature, which is hotter than the metal itself. You can see that until about the 1970s, the firing temperature was following the material's maximum temperature, and this was all metallurgy improvements. Hey, Course Three.
Then there started to be a big divergence. What happened in here? This turns out to have been metallurgy too. They went to directional solidification. It was a guy who had been a faculty member here, Mert Shank, who went down to Pratt & Whitney. There's a guy here, Tony Giamei, retired from Pratt & Whitney, who wrote a little article. Here's the turbofan engine. Over here on this page — Professor Flemings used to have these same three on a little plaque back in the mid-1970s when this stuff was all kind of new.
The conventional casting gave you a grain structure that wasn't fine grain. But in creep failure, which is how these things fail in service, they fail along the grain boundaries. He has a picture of metal failure at the grain boundaries. The grain boundaries are the weakest part in creep. So first thing they did is they did directional solidification, and they got the crystals all growing this way. The stress is that way, so the grain boundaries are parallel with the stress, and so they're not harmful. Then they got to single crystals.
They had to grow these very slowly. Professor Flemings and others worked out that it's the growth rate divided by the thermal gradient, G over R, that's important. You're now growing things — it might take you a whole day to grow some of these single-crystal things.
Here's a little plot that's a more refined version of what I just showed you. Metal temperature capability. You had conventionally cast, you had columnar growth, you got single crystals going off like this. From 1980 to 1999, if you look at this firing temperature, this is all directional solidification that allowed these things to happen. Then they got to air cooling. I've shown you the single crystal that has the air cooling on the inside, and that's why you got this big jump here. Right now the firing temperature is about 3000 degrees, the melting temperature of the alloy in the engine is about 2,400 degrees. If you lose your air cooling, your engine melts.
Student: [question about cost]
These materials are worth five to ten thousand dollars a pound in the engine. The average cost of the airframe might be 200 — on a military engine, who cares, it could be a thousand or two thousand dollars a pound for the airframe. But the engine is worth more, because it spins faster. So the energy cost here is insignificant. This is all processing equipment and time cost. And steel — it's all energy and raw materials today in that business.
It's partly raw materials now in the engine business. To get these best single crystals, CMSX-4 was like a six percent rhenium alloy. Rhenium is a platinum group metal, pretty pricey. I remember taking Nick Grant's course. Nick Grant was a professor in this department. He made his career after World War II on high-temperature creep, and he was developing some of the alloys people were using. First day of class he came in and said, "What's the most creep-resistant material to make a jet engine out of?" We all started guessing about different alloys. He says, "Platinum. Melts at 1700 C. It's got a lousy density — density is important in a jet engine — but it doesn't oxidize at high temperatures, and platinum is the best metal." Of course it was all laughable, because everyone knew that platinum was too expensive to use in an engine.
Until — around 1995, 2000, we started putting in CMSX-4, which is like three percent rhenium, and CMSX-5, which is like six percent rhenium. Remember those little pigtails that I passed around? Dr. Belmer talked about — I said I wanted those back. They're about six percent rhenium. There's about a pound, and about six percent of that is platinum-group metal. Those are about two thousand dollars apiece. We found about twenty of them down in the lab when we were cleaning it out. I collected two. I told everybody else not to throw them out. They've got about a thousand dollars apiece worth of platinum-group metal in each of them. They typically recycle them.
§7. The Air Force Mach 17 study — limits of materials [47:29]
So the jet engine business — if we go back to early days, it was metallurgical composition. Then it was metallurgical processing. Then it became air cooling, which is a mechanical thing. Since I only have two minutes left, I'll tell you a story. I'm sitting on a committee about seven or eight years ago. The Air Force had asked the National Academy of Sciences to run a study of how they should spend their $300 million a year research budget in propulsion. It was one of the biggest NRC committees I've ever been on. There were about thirty-five of us. Two of us were in materials. One was Neil Paton, who used to work for TIMET, and then myself.
I was sitting next to this one guy who was the designer of one of these first air-cooled engines, a guy from Pratt & Whitney. He was in charge of designing that engine. Other people on the committee were all chief designers of most of the engines of the last twenty years. Pratt & Whitney, Rolls Royce, General Electric. First meeting, going around the room, they asked us what would be the most valuable contribution to get higher-temperature engines for the future. I'm sitting right next to this guy, eighty-five years old, famous person in designing engines.
I was all ready to say, look, we've really exhausted the limit of our materials technology to go to higher temperatures, and we're going to have to improve things by design, as opposed to pushing the limits. I'm thinking of this graph. But he's speaking before me, and he says, "We've really pushed the limits of design and we now need better materials. And I got United Technologies' Board of Directors to spend $18 million of their own internal money developing new oxidation-resistant niobium alloys." So I had to change what I was going to say so I didn't completely contradict him, but also say, well, we really have a challenge in materials to go much further, because we do.
People have tried niobium alloys, but to make an oxidation-resistant niobium — if you know anything about the reactivity of niobium with oxygen, I could have told the board not to spend $18 million on that stupid project. They spent the $18 million and they got nothing. People have spent hundreds of millions of dollars trying to come up with higher-temperature alloys. But these refractory-metal alloys are very reactive. Sometimes they're called refractory alloys, sometimes they're called reactive alloys. Titanium, you get above 900 degrees and it just will go on fire like a flare. They've burned out engines because the compressor blades are made out of titanium. If you don't run the engine properly, instead of the compressors being at seven or eight hundred degrees, they can get a temperature spike up to 900 degrees, ignite the titanium, and now you just melt through your engine and end up with an empty core. It doesn't fly very well at that point.
We're really pushing the limits. But the really interesting thing — I'll take two more minutes. The Air Force explained their goals, and their goal was to have a Mach 17 non-man-rated engine. If it's not man-rated, basically this is for some cruise missile. I said, "So you're going to go Mach 17. We already go Mach 17 in space, what's the big deal?" "No, this is in the air." They wanted to go Mach 17 in air. I said, "Well, have you ever heard of frictional heating?" Okay, I didn't say that. But the problem is, when a spacecraft at Mach 17 is coming into reentry, what happens? They have to have these carbon tiles that absorb the heat. The frictional heating will give temperatures of 3000 degrees centigrade.
There are limits to materials. There are only so many materials that melt above 3000 degrees centigrade. I can count them on one hand. Tungsten, about 3300, and I think iridium at 3100. There's less than five that melt above 3100. The ceramics — but even that, they were talking about skin temperatures of 4000 degrees centigrade. I said, "Well, there's a limited number of materials." "Well, we're going to water-cool the skin." Let me explain water-cooling the skin.
They wanted to do this for the National Aerospace Plane in the mid-1980s. They were going to have about a three-millimeter-thick piece of copper — they'd use copper for thermal properties. They were going to have liquid hydrogen as the fuel tank, and they were going to have four-thousand-degree — or maybe it was three-thousand-degree — frictional heating in air on the outside. You want to rely on that? If you have a leak, what happens when hydrogen comes in contact with air at 3000 degrees centigrade? Kaboom. These people in the Air Force who dreamed these things up, they spent billions of dollars trying to design a National Aerospace Plane. They were going to do scramjets to get Mach 17 in air.
Another one of their goals was they wanted to have a plane take off, light enough, fly 25,000 miles around the globe, and land. I said, "Wait a second, you should only have to go 12,000 miles to get anywhere in the world." They said, "No, we cannot expect to have any bases in the future anywhere other than the continental United States. No one's going to let us have an air base anywhere in the world in the future. So we have to be able to take off in the United States and land back in the United States and go anywhere in the world and bomb them within so many hours." That was their goal. Nice goal, but I don't know if I want to pay for it as a taxpayer.
The scramjets — I said, "Why do you need to go Mach 17 in the air?" This is just a few years ago, eight or ten years ago. We once found Osama bin Laden. If we could have gotten the ordnance to him in twenty minutes, we could have taken him out, but we didn't have anything that was fast enough to get the ordnance from one of the carriers in to where he was. They located him, but they had to get it there in twenty minutes, and they would have needed something that could go Mach 17 in order to get into Afghanistan or Pakistan or wherever they had found him within twenty minutes.
So they have these goals. Some of them are just beyond the limits of a materials imagination, or even a mechanical engineer's imagination. But the people who are dreaming them up are not engineers necessarily. They're people who watched Star Trek in their youth, and they believe that everything on Star Trek is real. Okay, I'll see you Monday.