DP_S2012_06

Deformation Processing Spring 2012 Session · 8 sections 18 cases · Watch on YouTube ↗ all files
Layer 3 — readable edition

§1. Where we are in the book; shear and redundant work [00:34]

§1.p1

At the very beginning of the class I said I'm going to lecture out of Backofen, but even if you had a copy of Backofen you realize I'm not going chapter by chapter from the beginning. Hosford was Backofen's student, so his book is not all that different. It has a slightly different order. We've essentially done chapters one and two, which is stress and strain. Chapter three, we've done part of that, which is plasticity. I've been working mostly on chapters 10 and 11 in Hosford — chapter 10 is deformation zone geometry. It's chapter 7 in Backofen. If you want to follow along in the book, just ask me which chapter I'm working out of. I probably won't know, but I can probably find out quicker than you can.

§1.p2

Since I'm not giving you reading assignments, sometimes students have criticized me and said they couldn't follow where we were in the book. Well, why do you need to? You're not being quizzed. So I have this lousy attitude. I used to hate quizzes and things, okay.

§1.p3

One of the things I wanted to talk about, because I don't know if it's just my poor explanation or something else — when I talk about shear going this way and that way, there is something called positive shear, which, depending on whether you define a right-handed or left-handed coordinate system, would be positive shear, and then there would be negative shear. It's just a question of whether the top is going in the positive direction or the negative direction in your coordinate system. Shearing in that direction is a mirror image of that, and so one's positive and one's negative.

§1.p4

I've talked about redundant work, and I think you understand that if I go this way and that way and then that way, I've done three times the work to get one times the shape change. But I'm not sure you've thought about it in terms of a streamline. If you took a streamline of a little rectangle of material along this streamline, and it had to first curve this way, it would take a positive shear, and it comes along here, and where it's going straight, nothing is happening in terms of shear. Then as it has to turn this corner, it has the opposite curvature. So that's a double shear to do a dog-leg type of path. The whole thing started out as a rectangle and it ended up as a sheared rectangle, but in fact it had to go twice to get there in the case of a dog-leg. In other forming operations it may go through three times.

§1.p5

So think of the contours — whenever you turn a corner, whenever you have a radius to that flow line, you're introducing redundant work. If you really want to know about that, chapter seven of Hosford I think is slipline field analysis. I've shown you a couple of figures from Backofen and Hosford on slipline field analysis. It's where you get the 1 + π/2 — remember, on the deformation zone geometry, for Delta equal to one you have P over 2 t_y equal to 1, so it's just the compressive yield strength. You get up to Delta equal to 10 and it's 1 + π/2, and I said don't worry about it. Well, that π/2 is how many radians of curvature the material had to experience to take that shape change. π/2 is 90 degrees. So if it took a 90-degree turn to make the change — well, here it's not a full 90, maybe 45 and then another 45, so it averages somewhere in between. So that's a little more on that, because I was getting some glassy-eyed looks.


§2. Buy-to-fly ratio and near net shape [05:45]

§2.p1

I wanted to put this up and talk about, again — to a certain extent this was the forgeability of aluminum. If you started out with some little disc of material, you actually could in a single shot get quite a bit of deformation — you turn it into an I-beam basically. Whereas if it was a nickel-based superalloy you would not be able to do anywhere near as much. So it's very dependent on the type of material, the work hardening at the temperature you're doing everything at, and whatnot.

§2.p2

When we start looking at things like this, we get to something I haven't discussed in this class, but it's probably on some of the other videotapes if you take the welding and joining — something the Air Force calls the buy-to-fly ratio. Carl, what's the buy-to-fly ratio? In the manufacturing side of the Air Force they actually used to use it a lot in the '80s and '90s. It's how many pounds of metal you buy in ingot form or billet form, and in order to fly something, how much does the part weigh. So if you need to start with a billet of material that weighs 100 pounds, but the part that flies weighs 3 pounds, you have a buy-to-fly ratio of 33. That's not good when you're paying $100 to $200 a pound for the material. That means 97% of your cost — and let's face it, if you're talking about a 100-pound ingot or billet at $200 a pound, you're talking $20,000 for the part, and you're going to fly about $600 of that $20,000 of material. The rest of it goes into machining chips. You can recycle those at twenty cents on the dollar, depending — if it's a valuable metal you might get twenty cents on the dollar for it. So you've lost ten or fifteen thousand of value, just because these things have to be made in big heavy bulk and then machined away to much less.

§2.p3

When I talked about the Boeing wings, they start with plate and then machine away all the material they don't want, and the buy-to-fly ratio on that aluminum wing is probably something on the order of 20 to 1. There are some parts in aircraft engines and other parts — I've seen buy-to-fly ratios of 100 to 1. Well, not anymore. The Air Force spent billions of dollars in the '90s on what we call near net shape manufacturing, and we're going to talk about a little bit of near net shape today.


§3. Closed die forging and forging laps [08:57]

§3.p1

Now, I was just starting to get into casting defects — or forging defects, types of forging defects you might get. This is an example of a closed die forging. An open die forging I've shown you before — that's where you have the great big press and you just put a big ingot in there, and if I want to turn this cylindrical ingot into a rectangular bar, I just come down and hammer it with a flat hammer and the anvil is also flat, and eventually I make something rectangular or flat or disc-shaped. That's open die forging, and you get barreling because you have nothing constraining it on the side.

§3.p2

[Tom sets up a clay demonstration of closed die forging using a polyethylene sheet and high-temperature lubricant.] A closed die forging — we put lubricant on both. You take a billet and you forge it in a closed die. Now, even though clay is fairly soft, I have to put a lot of pressure on there because I'm getting to a fairly high Delta operation. I'm also going to squeeze out the flash around the edge. When I'm all done, if I've really squeezed this flat up against the thing — looks like it's not working very well even with lubricant. Well, there's my part. I've made the beginning of half an Easter egg. But I've got all this flash on the outside, and I've got to trim that, and that's part of my waste in my buy-to-fly ratio. That's a closed die forging, and obviously you can get much better dimensional tolerances. That's the wrong lubricant for this application. Fortunately the buy cost of this was not very high.

§3.p3

You get much better dimensional tolerances, although you still have to have draft angles. You can't make something with absolutely square sides because you can't get it out of the mold. In a closed die forging, when you're trying to fill some cavity, as you're trying to fill this mold, you will get situations where you actually get overlaps.

§3.p4

How common are they? They're not very common. I actually can only think in my career of one forging lap defect that caused a problem. Well, I take that back, I can think of two. One was — the Germans had forged a 20-foot-long shaft for a plastic injection molding machine, and it had been machined and everything, went into service, and about a year and a half later had a big fatigue crack. It had a forging lap that was about an inch long and about a quarter inch deep, and that's where the fatigue crack started. I mean, you get a flaw that big in something that's rotating. Of course the Germans said it couldn't have been ours — except you always know it's a forging lap because the whole lap surface is oxidized. It was formed at 2,000 degrees Fahrenheit in the air, and so it's full of oxide and decarburization if it's steel. It's unmistakable. So eventually they had to own up to that.

§3.p5

Another one was actually not strictly a forging lap. It was a piece of pipe that had been laid in the Gulf of Mexico underwater, and they were doing the final hydro test before they started using it to bring oil from the well onshore. This thing had been hydro tested twice at the plant in Ohio where it had been made, and it passed the hydro test, and then it got down and they had to coat it with a plastic compound to give it corrosion resistance in the ocean. Then they put it in the ocean and they hydro tested it again, and when they hydro tested it, it opened up. It had a flaw 80% of the way through.

§3.p6

What had happened is — in order to make that pipe, and it's actually sort of neat to watch this — they start with a hot billet at about 2,000 degrees Fahrenheit of steel, and it might be as long as this room, and it comes rolling down out of the furnace onto the bed. Some metal jaws that are room temperature or slightly above grab it, and then a mandrel comes through — a big hydraulic press, big long steel mandrel about 40 feet long comes through and just goes right through it. It's called piercing the pipe. It's the way you make seamless pipe. You don't have a weld in the pipe. It punches right through and turns it into a tube, and you're doing it with an outside die that's drawing all this over. You now made a seamless pipe. Of course it's all oxidized and got scale all over it. It goes down to the next station, still hot, and they hit it with a blast of steam on the inside. The steam now is going to expand because it's not at 2,000 degrees, and you just see it blow all that oxide scale out of the inside of the pipe. They would make you pay to go and witness something like this if it was in an amusement park. But I get people to pay me to go and look at these things.

§3.p7

What was I getting into all that? Oh, I was talking about defects. That particular one — a piece of the mandrel had broken off and had gotten embedded into the surface of the steel. The mandrel is relatively cold. They have hundreds of mandrels, and the mandrels can be used once before they get rotated back through a cooling process. The mandrels have to be cold, they've got to be pretty strong to pierce this thing, so they get used once, and then they come back about an hour later after they've cooled down. So they've got about a hundred of these mandrels just running through in circles. They fatigued, and this one had fatigued. No one caught it. They did a non-destructive testing inspection, but the defect was at the end of the pipe. The technique they were using for the main 40-foot length was an automatic eddy current inspection. Very reliable — this defect was 3/4 inch long and 3/8 of an inch deep. The problem is, on the ends they couldn't do that because they had to grab it for the eddy current inspection. So the ends weren't inspected by eddy current, but they go in there with magnetic particle. That's all done with a bunch of what looks like immigrant labor, probably illegal immigrant labor. They squirt little magnetic powder on the inside, and they've got a flashlight, and they look to see if they magnetize it and if they see a flaw. You can imagine that if you do that for 8 hours a day, you might miss something from time to time. They did. They missed it.

§3.p8

So it wasn't that they hadn't inspected for it, but they missed it. Things will get through, even when you have very good quality control. If there are still people in the process, the error rate is still on the order of a percent or so. Just because something's been inspected, don't assume that it's good. You're still going to have some failures.

§3.p9

Another type of defect — I have never seen it personally, but I do have a sample of it here. [Tom locates and holds up an extrusion sample from Backofen's collection.] If you're extruding something, you get an end effect when you have the die trying to extrude out the hole. So you've got a die, you're extruding this, and at the end you basically get a dead zone. If you're extruding this way to make a cup, you're getting to the end, you end up with a dead zone that forms because the metal can't flow — it doesn't want to go too far in that direction and turn a 90-degree turn. It wants to have flow lines. So you can get a little defect — this is one of Backofen's samples, this is an extrusion, you can see it's the end of the extrusion press and it's got a little dimple right there. That defect is inconsequential, but in some cases that type of defect shows up in forgings, because people weren't careful about the end of the forging.

§3.p10

Here are some examples of forging laps out of the steel mill. Scabby surface — this case is a bloom, so it's actually been rolled. These are rolling laps. Semi-finished roll product — that one you can't see very well. Here are the types of things, semi-finished roll product. These little grooves are probably scale, it's been rolled in, and they could be an eighth of an inch deep. This whole thing is probably 6 to 8 inches wide. So when you see grooves like that, they're pretty sizable.


§4. Continuous casting and the LNG propeller shaft [19:55]

§4.p1

In fact at the steel mill you have a buy-to-fly ratio in a sense — they call it yield, whether it's a steel mill or aluminum mill, how much you cast to how much you ship. When I started at Bethlehem Steel in 1974, the overall product yield was about 65% for the whole corporation, all products. Bethlehem was the second largest steel company in the world at the time. They made everything from wire to I-beams to huge forgings, sheet metal — everything. They had an overall 65%. If they could have gotten to 67%, they could have doubled their profitability — maybe even with a 1% change. What happened is, somewhere between '65 and the early 1980s, over about a 20-year period, they went from ingot casting to continuous casting. The yield ratio got up from 65% to 97%.

§4.p2

If I cast an ingot, I have slightly beveled sides to the mold. The mold has to have fairly thick walls because if it's just a cast iron mold, it's going to have to carry out the heat from all this liquid. You fill it up to some level of liquid, but the liquid shrinks as it solidifies, and it is what we call a killed steel. The final product looks like this, and you have a big gap here.

§4.p3

In the welding course I actually talk about how one company ended up forging this casting into a 33-inch diameter shaft. I told you about my 33-inch diameter fracture in the mid '80s. This went as a propeller shaft on a liquid natural gas carrier, and it stopped dead full of LNG in the middle of the Philippine Sea. They were airlifting in Lear jets fittings so they could transfer the LNG to another empty LNG cargo tanker that was coming back from Japan. This is a whole series of LNG ships going from Indonesia up to Japan and back. This one was going up to Japan full, and when it lost power and couldn't go anywhere, full of LNG, you could have had a big explosion which would have upset a lot of people. May not have killed very many. But remember, those same LNG tankers come in once or twice a month into Everett, right through Boston Harbor. One of the biggest concerns that Homeland Security has for Boston is, all you need is a rocket-propelled grenade when that thing's coming in, and you will wipe out half of Boston. So they have a little security to try to make sure there are no terrorists with RPGs. But nonetheless, this thing stopped dead in the Philippine Sea.

§4.p4

The problem with the old ingot casting of steel was you threw away the top one-quarter of the steel. You were down to 75% to begin with. In spite of the scale losses in the forging mill, or the end crops in the sheet metal mill, you're lucky to get 65% when you lose the first 25% in your casting process. Since yield is such a big deal — at the time, the United States was making 150 million tons of steel a year here, and they were shipping about 100 million tons. There's your 60, 65% yield.

§4.p5

In the '70s — basically in the decade of the '70s, but starting in the '60s — they switched to continuous casting. In continuous casting you don't have to crop the top, because you just cut it off, and then the bottom of one is the top of the other. Your yield shoots up. Now in the worldwide steel industry, they still have to ingot cast some big forgings, where you may still have a 60% yield on your final product, but most of the product is continuously cast, and some steel companies have 97% yield now.

§4.p6

What happened to the steel industry — several things happened, but one of the things is, all of a sudden in a 10-year period you were producing one-third more steel than you needed in the casting shop, because you switched to continuous casting. All of a sudden your capacity in all your plants around the world is 50% larger than you need. There was a tremendous overcapacity because of continuous casting. Here is a wonderful development, but it nearly bankrupted the steel companies because they started competing with each other on price for the market because they all had excess capacity. The world decided the steel industry was a dying industry when in fact they had just had a huge gain in productivity, but that gain in productivity almost bankrupted them. Think of the irony. The moral of this story is, you have to be careful about what you wish for. Big disruptive changes in technology can be very painful.


§5. Rolling I-beams and the Bethlehem Steel story [25:32]

§5.p1

What I wanted to talk about — it may not be obvious until you think about it for about 30 seconds, but if I want to start with a square billet and make it into another smaller square billet — this is out of the US Steel book, and they show you the types of dies. First of all, you don't take a square billet like this and form it on its flat sides. You deform it on the edge, on the peaks. You go from being a square to being a parallelogram. Then you flip it 90 degrees and you take the long axis of the diamond and you squeeze it again. It makes sense if you think about it. It's easier to make dies, you're not trying to squeeze things from the side. But this gets sort of messy in terms of the shapes — if you wanted to make a straight I-beam, your final passes look sort of strange. The I-beam's coming through on a slightly diagonal.

§5.p2

The thing that built the steel industries in the 1880s was the railroads. Before we had the automotive business, the railroads were the big consumer of steel in the world. That's what made Andrew Carnegie the richest man in the world. He was the Bill Gates of his day. If I told you the story of Andrew Carnegie and my grandfather, I'll tell you that sometime. You start out with something that comes out of the blooming mill, which is called a bloom — sometimes called a billet or a rod — and these are the shape changes you go through to end up with an I-beam or a steel rail. So they've got 10 steps, 10 sets of rolls to make a railroad wheel.

§5.p3

Here's another example out of the same book, of the breakdown pass — you first just kind of squeeze it in and you get these things, you finally end up with the finishing pass. The thing that made Bethlehem Steel — Andrew Carnegie had started US Steel, and it was the king of all the steel mills in the world in the 1880s and '90s. His right-hand man in Pittsburgh was a guy named Charles Schwab. Not Charles Schwab the investor — well, Schwab, but a different Schwab. I can't remember the Schwab of Bethlehem Steel 1890 fame. They had a tiff, and Schwab went off and found this little iron mill in Bethlehem, Pennsylvania. Those guys had just learned how to put molybdenum in steel to make high-speed steel.

§5.p4

Without getting into all the metallurgy — molybdenum forms molybdenum carbides which don't dissolve until higher temperatures, so you can get these hard carbides, and the steel will retain its hardness to about 1,000 or 1,100 degrees, where ordinarily at 7 or 800 degrees the steel is starting to soften. You get what's called a secondary hardening peak because of these molybdenum carbides, and you can have a hot work tool steel that can hot work at 1,100 degrees Fahrenheit as opposed to 6 or 700 degrees Fahrenheit. This is going to improve the life of drills and all kinds of things that get hot when you work them. That was the first big invention that helped grow Bethlehem Steel.

§5.p5

The second one — Bethlehem Steel was the first firm to learn how to make I-beams. They learned how to roll passes like this for making I-beams, and they were close enough to New York City — the reason New York City now has buildings more than six stories tall is because Bethlehem Steel made I-beams. Manhattan was sort of flat. If you go to Manhattan, you won't find a lot of buildings over about 10 or 12 stories tall — actually those you won't find anymore because they turn them into 100-story buildings. But if you had gone there in 1900, you would have found very few buildings over about 10 stories tall, because brick and mortar just didn't give you enough backbone to build a building. Steel I-beams did. The other thing you would find in New York City in 1900 — 40,000 gallons of urine and horse manure. I don't know how many pounds of horse manure on the streets, because that was the biggest pollution problem in all the big cities around 1900. The automobile solved that. It just put out smog.

§5.p6

Student: Is there a heating step between?

§5.p7

Yep, you have to reheat it after each one. Well, you might be able to go from this one to this one, you've got a lot of mass, but then you have to have two mills, because usually these are bigger mills because they're bigger pieces. Once this thing's getting long like this, you have to — many times — I wouldn't say you have to do it every time, you might get two passes before you have to put it back in the furnace. It's a slow process.

§5.p8

Student: This is a rolling process?

§5.p9

It's a rolling process.

§5.p10

Here's another series of passes to make an I-beam. That doesn't look like it's going to be an I-beam at that stage.

§5.p11

Student: Was this all done by trial and error?

§5.p12

Yeah, basically. Trial and error where they had to build huge rolls to do it. These are two steps — this is a step to make a channel, and this is to make an angle. Here's another one to make an angle. Notice this one to make an angle, you're basically rolling it and sort of peaking it. This one you keep pretty flat to make that same angle iron, you're just rolling a little peak on it, and finally you straighten it out. Here's a Z channel, or actually it's a piling — interlocking things, this one interlocks into that one. You drive them into the ground so you can dig a pit and work in it without the wall caving in on you. A lot of steel used for pilings. So this is what you have to do to form a piling. These last ones, I don't know if they're hot formed in that same type of mill or in a separate little mill. I think it's sort of interesting to look at this.

§5.p13

Nowadays they actually probably would do it on a computer, but back then, it was trial and error.

§5.p14

Student: Are processes still roughly the same now?

§5.p15

I'm going to show you how they do it now, but until 1985 or so, yes — for about a hundred years that was the way they did it. For some of the smaller beams it might be 20 passes, or over 20 passes to go through the mill. Gets a little pricey. What that meant is, some of the smaller beams were pricier than the big beams. It takes a lot more rolling and reheating and everything else.


§6. Attleboro gold tubing and die design [32:54]

§6.p1

So far as people doing things empirically — once, 25 years ago, I used to consult for a gold company down in Attleboro, Massachusetts. I think I mentioned, they went through seven tons of gold a year, and they had their own continuous casters for karat gold alloys. That plant is still there. It's been there for 113 years now, started in 1899. To make gold tubing — I actually got them set up with a gold welding machine so they could weld karat gold tubing longitudinally. I had to go to Steamboat Springs, Colorado one winter to see the guy who makes these — he just had a big machine shop in Steamboat Springs, and he wasn't building welding machines, he was out skiing. But before that they made gold tubing. What do you use gold tubing for? If you make heavy-wall gold tubing, you can machine it into little circular things called rings — a lot of wedding rings are made with gold tubing.

§6.p2

One of the big very profitable things in the Boston area — you go down to downtown Boston, there's a Jewelers Building, and there's a bunch of flute manufacturers. I guess now I'm supposed to pronounce it "flout-ist," not flutist — they're a flautist. Professor Harry Gatos, a Greek, graduated from this department and then went off to Lincoln Lab and headed up the solid state division, and came back here in the mid '60s. He was called Mr. Gallium Arsenide. He was the world's expert on gallium arsenide before it went into cell phones. He's the one who really did all the basic research to make gallium arsenide a viable semiconductor. He's also a very accomplished flautist — he's well into his 80s now. If you walked down Building 13 when I was a student, you'd hear him in there during the middle of the day playing his flute. He played for the Boston Symphony and a few other people as a guest soloist. He had gold flutes, silver flutes, and he had one of the six platinum flutes in the world at the time.

§6.p3

They made their tubes by a deep drawing process — just a great big long press coming in. Start out with a piece of karat gold and make it. They also would do gold-filled material, which is a clad material — clad on copper or brass, with gold on the outside. They asked me to redesign their dies, which is why I started thinking about this when you asked the question, was this all done empirically. If you looked at the dies, some of them — when you went from one die to the next, some of them had a less-than-9% reduction, and another one had a 35% reduction. There's no logic whatsoever. So I redesigned their dies for them. It's basically calculating logical steps of progression. Some of the dies at 40% might be cracking, giving high failure rates. The other ones at 9% could have been setting up stresses that actually could have been bad later on. Remember, I told you, you want to get a fair amount of deformation in. You don't want to do these little light passes. They can do more harm than good.


§7. Chaparral Steel and the mini mill revolution [37:39]

§7.p1

So now let me tell you the Chaparral Steel story. Chaparral Steel was founded by a graduate of this department, who's now retired. He's from Vancouver, but he worked for a Canadian steel company in the mid '70s. Gordon — his name is Gordon Forward — used to be on the department's visiting committee back when I was department head. Gordon realized in the mid '70s that it cost $200 a ton to make cast iron in the blast furnace, but you could buy scrap iron at the time, scrap steel, for $100 a ton. What do you do with the cast iron from the blast furnace? You put it into the steelmaking furnace and make it into steel. If it cost you $200 for the liquid cast iron, and you could buy steel scrap for $100 a ton, you have a $100-a-ton advantage on a $300 or $400-a-ton product. That's not a bad advantage.

§7.p2

Gordon and a few other people around the world realized this, and they started what they called mini mills. Instead of a steel mill as big as — well, the last steel mill built in this country was the Bethlehem Steel Burns Harbor plant, from 1965 to 1972. It took 7 years to construct, it cost Bethlehem $5 billion. Bethlehem was the second largest steel company in the world at the time. They started the Burns Harbor plant, which is still in operation. It almost bankrupted them. I think they were only about $5 billion a year in sales, and so they were spending a lot of money on capital. It'd be equivalent today of Intel having to spend $20 billion to build a fab plant, or a $60 billion fab plant. It can bankrupt a company if things don't go right. Or Boeing deciding to build a Dreamliner — it can bankrupt the company if it doesn't come out properly.

§7.p3

Bethlehem actually finished the plant, and by '74 they were having the most profitable year ever. That's when they hired me, and they've been going downhill ever since they hired me. But since then, no integrated steel plant anywhere in the world has ever been built by a company. There have been plenty of them built, but only by countries. They have been built by the backing of an entire nation. POSCO Steel, which became the largest steel company in the world in the late '90s, was built by the Korean government. That's where the financial backing came from. It's a political thing to make that big an investment.

§7.p4

The mini mill guys could build a mini mill to turn out product for one tenth the capital cost of an integrated steel mill, and they did. Today about half of the US steel market is made in mini mills. What Chaparral did — they were making rectangular blooms, and then they decided, why can't we make something a little closer to net shape? So they did. They started casting a dog bone.

§7.p5

They kept on improving that until they got to this — now it's almost an I-beam that they're casting, and they went from 21 passes for the smallest I-beams to about seven passes. Not only that — for rolled steel bars, before Bethlehem started doing this, 41% of all the I-beams were imported into the United States. This is early '90s. But by the time they did this, they wiped out all imports because they were so much more efficient by doing net-shape casting to make the I-beams and wiping out two-thirds of the rolling passes. They were exporting 20% of the product outside of North America, including Asia, which in the early '90s was just incredible.

§7.p6

They had 75% energy savings. They had inline reheat because they didn't have this big heavy thing you had to stick back in the furnace. They were starting with something after the first couple of passes that was so thin they could basically heat it in line with induction heaters, and not have to put it back into the furnace. So inline reheat — 28 minutes total to go through the mill. They ended up with a fine-grain chill-cast structure that had a higher tensile strength and toughness. They were able to roll these things to get a finer grain size, and if you know anything about grain size and properties, grain size is the only thing that gives you both increased strength and increased toughness.

§7.p7

At the time, the garden-variety steel is what's called A36, ASTM A36. It just so happens — coincidence — A36 has a yield strength of 36 ksi, 36,000 pounds per square inch. If you go to the American Institute of Steel Construction manual, they would have back in 1990 design criteria for 36 ksi steel, 50 ksi steel, and 100 ksi steel. The 100 ksi steel was quenched-and-tempered steel, fairly expensive, but they put that in bridges and stuff when they needed weight savings. The 36 and 50 was the standard. You get a premium on the 50 ksi steel, because you can save one-third the weight when you've got 50% more strength.

§7.p8

Chaparral, by doing fewer passes, finer grain size, better reduction, was able to make product that could pass for either A36 or grade 50. They could cut their inventory in half. They didn't have to distinguish what was A36 and what was grade 50 — order whichever one you want, we're going to ship you what we've got, which meets both specs. You're getting a better steel if you ordered A36, but we'll sell it to you for the same price as regular old A36, because we don't care. It's profitable enough that we're going to save money on the inventory. We cut our inventory down by half. Here's an innovation that made Chaparral a world-beater on the I-beam market by going to net-shape casting, getting rid of a lot of these rolling passes, and getting a better product. So it's sort of a neat little story of something that helped them become more profitable.


§8. Homer Research, executive dining rooms, and management failure [46:15]

§8.p1

There are lots of good stories about Gordon Forward and Chaparral. I'll tell you the story when I worked at Bethlehem Steel. In the mid '60s, Bethlehem had built a Taj Mahal. It was the fanciest research laboratory in the world, probably, for metals research. It's called Homer Research Lab, after one of their chairmen named Homer. It was just south of Lehigh University. Lehigh's on the side of a mountain. The Bethlehem plant is north of the river, south of the river is a couple of slums where the steel workers lived, and then going up the hill which is called South Mountain was Lehigh University on the side, and then at the top was Homer Research, and that's where I was hired. A $600 million facility — in the late '80s, Bethlehem, when they were going under, gave it to Lehigh University to get it off their tax rolls. It's like 20 buildings, and these really were a Taj Mahal.

§8.p2

At Homer Research when I was there, we had the corporate dining room, which was on a second floor with huge windows looking out on the vista of the steel mill. You could see the smog from the coke ovens over here, and the pollution from the blast furnace over here. They had an executive dining room, which I snuck into once at about 2:30 in the afternoon, just to say I'd been in there. This is where the big shots, like the Don Trautleins, would eat lunch. I remember seeing a strawberry the size of a peach. I thought, this is amazing, it was going into the executive dining room. If I told you all the stories of how the upper management took care of themselves — I lived near the airport, and they had three Lear jets which would fly the vice presidents down to Florida in the winter to play golf on the weekends. This is good corporate management in the United States circa 1975, which is one of the reasons that after 13 months I said, where do I want to be in 5 years? And the clear answer was, not at Bethlehem Steel. So I left and came back here.

§8.p3

At Chaparral Steel — I did a study once with a guy from Harvard Business School, and we spent a few days down at Chaparral and visiting Gordon Forward. Gordon used to always talk about the corporate dining room at Chaparral. The corporate dining room at Chaparral was a couple of folding tables, plastic chairs — not as nice as these, nowhere near as nice as these. Plastic forks and the little things, because they would go out and get some barbecue from the local takeout. That was the corporate dining room at Chaparral. So difference in philosophy between the people who were at one time printing money.

§8.p4

Talk about printing money — what made the American steel industry, after World War II, great? We controlled 75% of the world's production after World War II. Why? We had bombed out all the rest of the capacity. Hey, it's not every day that a businessman gets to go and wipe out his competition that way. But the managers — think about it. In 1975 when I was working for Bethlehem Steel, the managers had been hired about 30 years before, right after World War II, and they came in with this arrogance. All through their careers they controlled 75% of the world steelmaking capacity. At that time US Steel was 40% of the world steelmaking capacity, and Bethlehem was like 20%. The two of them alone were over half the world steelmaking capacity, because we didn't fight the war in North America. At least not a lot of bombs fell in North America.

§8.p5

These guys thought they were the brightest managers in the world. It's sort of like a financial guy today on Wall Street who walks away with a half-billion-dollar income. He thinks he's the brightest guy in the financial market in the world. He's not brighter than anybody else, he just happens to be in the right place at the right time. The Bethlehem Steel managers when I was there, the guys at the top thought they were the most clever people in the world. It's why they no longer exist. They thought they were great managers, and in fact they were flying to Florida on the weekends in corporate jets to play golf, and they were eating strawberries the size of peaches. They couldn't sustain that. Frankly, I thought it was stupid when I was 25 years old. Unfortunately I've seen it repeated again and again in a number of industries.

§8.p6

Kodak. They were printing money. They called it film, but it was money. They had like a 95% margin on film, and they would not give up their film. They invented a lot of the best digital photography technology in the world, but the film guys didn't want to give it up because they had grown up knowing that film was their bread and butter, and they kept their bread and butter until it was all stale and rancid. So it's not just the steel industry, it's Kodak. We could go through a bunch of other industries. People — it just amazes me. These people at business schools will write articles. I remember articles written by some of my friends, one of whom became Dean at Harvard Business School. He was writing articles on the steel industry that were just — I said, Kim, that's wrong. He says, no, it's not. He had an economics degree, he studied labor relations, and he was lecturing me on the steel industry. He became Dean, and so he wrote all these papers, and industry would pay him tens of thousands of dollars a lecture, because he was — but that doesn't mean he knew anything.

§8.p7

Anyway, this is how you make a railroad wheel. Again, the steel industry grew up from Andrew Carnegie through the 1920s living off the railroad. You start out with a block, you forge it into a disc, you then take that and forge it after — second forging — and you make a little bit of a net-shape indentation, and then you punch it to make a hole in the center, and then you roll it. These flanges right in here are actually rolled. See the wheel here vertical, see the punched hole horizontal, and they're rolling these flanges in a special mill, and then they cone it to give it the right contour in another operation. So that's how you make a railroad wheel — a mixture of rolling and forging. Just gives you an example of how people vary those things.

Cases referenced

  • Air Force Buy-to-Fly Ratio in Aircraft Manufacturing §2.p2

    The Air Force concept of buy-to-fly ratio (mass purchased ÷ mass that flies) introduced as a productivity metric. Tom cites 33:1 hypothetical, 20:1 Boeing wings, 100:1 some engine parts. Sets up the near-net-shape pivot to Chaparral later.

  • Boeing aluminum wing manufacturing waste §2.p3

    Reference to Boeing wings machined from plate, ~20:1 buy-to-fly ratio. Brief — back-reference to earlier lecture treatment.

  • U.S. Air Force net-shape manufacturing initiative §2.p3

    "The Air Force spent billions of dollars in the '90s on what we call near net shape manufacturing." Frames the lecture's subsequent treatment of near-net-shape continuous casting at Chaparral.

  • German plastic injection molding machine shaft forging lap §3.p4

    20-foot forged shaft, fatigue crack after ~18 months in service originating from a 1-inch × 1/4-inch forging lap. Used to illustrate that forging laps are diagnostic: oxidized surface from 2,000°F formation in air gives them away. German manufacturer initially denied.

  • Gulf of Mexico oil pipeline defect §3.p5

    Seamless pipe with a 3/4-inch × 3/8-inch surface flaw 80% through-wall, made in Ohio, passed two factory hydro tests, failed third hydro test after installation in Gulf. Cause: piece of fatigued piercing mandrel embedded in pipe surface. Defect was on pipe end, where eddy current couldn't reach; magnetic particle inspection (manual) missed it.

  • SS Libra LNG vessel tail shaft failure §4.p3

    33-inch propeller shaft forged from killed-steel ingot stopped dead full of LNG in the Philippine Sea on Indonesia-to-Japan run. Lear jets airlifted fittings for emergency LNG transfer to sister ship. Tom uses this to make the killed-steel ingot top-cropping point and the related Homeland Security RPG concern about Boston Harbor LNG traffic.

  • Bethlehem Steel Burns Harbor §7.p2

    Last integrated steel mill built in the United States by a private company. 1965–1972 construction, $5 billion, 7 years. Nearly bankrupted Bethlehem. Tom uses this to make the point that since then, integrated steel mills are built only by governments (POSCO via Korean government).

  • POSCO (Korea) government-built steel mill surpassing Nippon Steel §7.p3

    Cited as the largest steel company in the world by late '90s, built by the Korean government as the only way to finance an integrated steel mill at that scale.

  • Chaparral Steel §7.p1

    **The lecture's central case.** Founded by Gordon Forward, MIT DMSE grad. Mini mill at 1/10 the capital cost of an integrated mill. Near-net-shape dog-bone casting reduced I-beam rolling passes from 21 to 7. 75% energy savings, inline induction reheat, 28-minute total processing time, fine-grain chill-cast structure. Product met both A36 and grade 50 specs simultaneously, allowing inventory consolidation. Wiped out I-beam imports to US, exported 20% of product overseas including Asia. Folding-table corporate dining room contrasted with Bethlehem's executive dining room.

  • Gordon Forward mini mill competition strategy §7.p1

    Gordon Forward's economic insight: scrap steel at $100/ton vs. $200/ton blast-furnace cast iron gave mini mills a $100/ton advantage on a $300–400/ton product. Forward was on the MIT DMSE visiting committee during Tom's department headship.

  • Continuous casting and steel industry capacity collapse §4.p1

    Yield jumped from 65% (ingot casting) to 97% (continuous casting) over 1965–1985. Effective capacity grew 50% with no capital expansion. Overcapacity drove price competition that nearly bankrupted the integrated mills. Tom's "be careful what you wish for" — disruptive productivity gains were painful for incumbents.

  • U.S. Steel Industry Overcapacity Crisis §4.p6

    Direct consequence of the continuous casting yield jump. Frames the lecture's argument that productivity gain ≠ industry health.

  • Ford Motor Company executive dining room hamburger story §8.p2

    Anecdotal evidence of management-class extravagance at mid-1970s Bethlehem Steel: peach-sized strawberries in the executive dining room, three Lear jets ferrying VPs to Florida golf weekends. Used to set up the structural argument about why Bethlehem's management failed.

  • Attleboro gold alloy and casting operations §6.p1

    113-year-old (as of 2012) Attleboro gold plant, 7 tons/year throughput, in-house continuous casters for karat gold. Tom installed a gold welding machine sourced from a one-man Steamboat Springs machine shop. Used here to set up the die-design redesign story (next entry).

  • Gold tubing manufacturer dies redesign (Boston area) §6.p3

    Tom redesigned a sequence of drawing dies that had grown empirically over decades — some at 9% reduction, some at 35%+ — to a logical progression. Used as the on-page proof that even modern industrial die sequences were assembled by trial and error.

  • Kodak film digital photography failure §8.p6

    Classic incumbent's-dilemma example. Kodak had 95% margins on film and would not cannibalize the business with digital, despite having invented much of the underlying digital photography technology. Used as a non-steel parallel to Bethlehem's management failure.

  • Boeing 787 all-composite aircraft §7.p2

    Brief comparison — Boeing's Dreamliner program cited as an example of a single product launch that could bankrupt a $5–60 billion enterprise if it goes wrong. Parallel to Bethlehem's Burns Harbor bet.

  • MIT Commission on Industrial Productivity - Steel Industry Study §8.p3

    Likely referent for "a study once with a guy from Harvard Business School" at Chaparral. The "Kim" mentioned at §8.p6 (HBS friend who became Dean and wrote on the steel industry) likely Kim B. Clark. Tom uses this to make a point about economists writing on industries they don't understand.

Layer 2 — cleanup edit
p1 00:34

Does that work? Okay, maybe I switch that one, I don't know. Um, in any case, I think you may not remember, but the very beginning of the class I said I'm going to lecture out of Backofen, but even if you had a copy of Backofen you realize I'm not going chapter by chapter from the beginning of Backofen. Uh, Hosford was Backofen's student, so his book is not all that different. Um, but it does have a slightly different order and stuff. We've essentially done chapters one and two, which is stress and strain. Uh, chapter three, we've done part of that, which is plasticity. I've been working mostly on chapters 10 and 11 in Hosford, okay, which actually starts as chapter 10 in Hosford is deformation zone geometry. It's chapter 7 in Backofen. You know, if you want to follow along in the book — I realize, just ask me which chapter I'm working out of. I probably won't know, but I can probably find out, uh, maybe quicker than you can so far as that goes.

p2 01:33

So, um, I apologize. I'm usually not very — since I'm not giving you reading assignments, sometimes students have criticized me and said I couldn't follow where we were in the book. Well, why do you need to? You're not being quizzed, you know. So I have this lousy attitude. I used to hate quizzes and things, okay.

p3 01:53

Um, one of the things that I wanted to talk about, because I don't know if it's just my poor explanation or whether it's something else, but when I actually put up — it's not very dark, let's open this up and brighten it a little bit. Okay, when I put up, um — when I use these and talk about shear going this way and that way, there is something called positive shear, which, depending on how you define whether you're right-handed or left-handed coordinate system, there would be positive shear and then would be negative shear. And it's just a question of whether the top is going in the positive direction or the negative direction in your coordinate system. But shearing in that direction is a mirror image of that, and so one's positive and one's negative.

p4 02:54

And I've talked about redundant work, and I think you understand that if I go this way and that way and then that way, that I've done three times the work to get one times the shape change. But I'm not sure that you've thought about it in terms of a streamline. If you took a streamline of a little rectangle of material along this streamline, and it had to first curve this way, it would take a positive shear, and it comes along here, and where it's going straight, nothing is happening in terms of shear. And then as it has to turn this corner, it has the opposite curvature, and so it's another. Okay, so that should be a double shear, okay, to do a dog leg type of path. And that is how the whole thing started out as a rectangle and it ended up as a sheared rectangle, but in fact it had to go twice to get there in that particular case of a dog leg. In other forming operations it may go through three times, okay.

p5 04:00

So think of the, uh, the contours — whenever you turn a corner, whenever you have a radius to that flow line, you're introducing redundant work. And in fact, if you really want to know about that — I'm not going to go through it, at least not in any detail, if I do go through it — but chapter seven of Hosford I think is slipline field analysis. And this is a — I've actually shown you a couple of figures from Backofen and/or Hosford on slipline field analysis. But it actually — it's where you get the 1 + π/2, okay, as the — remember, on the deformation zone geometry, for Delta is equal to one you have a P over 2 t y is equal to 1, so it's just the compressive yield strength. You get up here to Delta is equal to 10 and it's 1 + π/2, and I said don't worry about it. Well, that π/2 is how many radians of curvature the material had to experience to take that shape change, okay. So π/2 is 90 degrees, isn't that right? Yeah, okay, uh, if I remember. Okay, so if it took a 90 degree turn to make the change — well, here it's not a full, maybe 45 and then another 45, so maybe it is, you know, a π/2 or something. It averages somewhere in between, okay. So that's a little more on that, 'cause I was getting some sort of glassy-eyed looks.

p6 05:45

And we talked — oh, okay, I wanted to put this up and talk about, again, not about forgeability of different alloys — to a certain extent this was the forgeability of aluminum. If you started out with some little disc of material, you actually could in a single shot get quite a bit of — you turn it into an I-beam basically. Whereas if it was a nickel-based superalloy you would not be able to do anywhere near as much, okay. So it's very dependent on the type of material, the work hardening at the temperature you're doing everything at, and whatnot.

p7 06:24

But, um, when we start looking at things like this, we get to something that I haven't discussed in this class, but it's probably on some of the other videotapes that you might see if you take the welding and joining — something the Air Force calls the buy-to-fly ratio, okay. Carl, what's the buy-to-fly ratio, do you know what it is? Okay, uh, in the manufacturing side of the Air Force they actually used to use it a lot in the '80s and '90s. It's how many pounds of metal you buy in ingot form or billet form, and in order to fly something, how much does the part weigh. So if you need to start with a billet of material that weighs 100 pounds, um, but the part that flies weighs 3 pounds, you have a buy-to-fly ratio of 33, right? Uh, that's not good when you're paying $100 to $200 a pound for the material. That means that 97% of your cost — and let's face it, if you're talking about a 100 pound ingot or billet at $200 a pound, you're talking $20,000 for the part, and you're going to fly about $600 of that $20,000 of material, and the rest of it goes into machining chips. Which — okay, you can recycle those at twenty cents on the dollar or something, you know, depending if it's a valuable metal you might get twenty cents on the dollar for it. So you've lost, you know, ten, fifteen thousand of value, um, just because these things have to be made in big heavy bulk and then machined away to much less.

p8 08:08

When I talked about the Boeing wings, they start with plate and then machine away all the material they don't want, and the buy-to-fly ratio on that aluminum wing is probably something on the order of 20 to 1, okay. There are some parts in aircraft engines and other parts that they had — I've seen buy-to-fly ratios of 100 to 1, okay. Well, not anymore. The Air Force spent billions of dollars in the '90s on what we call near net shape manufacturing, and we're going to talk about a little bit of near net shape a little bit today, so, uh, that's one of the reasons I'm going over that.

p9 08:57

Now, I was just starting to get into casting defects — or forging defects, types of forging defects that you might get. This is an example of a what we call closed die forging. Uh, an open die forging I've shown you before — that's where you have the great big press and you just put a big, uh, ingot in there, and if I want to turn this cylindrical ingot into a rectangular bar, I just come down and hammer it with a flat hammer and the anvil is also flat, and eventually I make something rectangular or flat or disc shape or whatever. That's open die forging, and you get barreling because you have nothing constraining it on the side.

p10 09:38

Closed die forging on the other hand — I guess I can do an example for you. [Tom produces a polyethylene sheet and lubricant for a clay demonstration.] This is a polyethylene sheet, but the reason I bought this is when I was demonstrating or trying this out for myself this morning, I found I needed a little lubricant, so we put our high temperature lubricant on there. Um, here's — sorry, this is the best I could find in my office, I don't have that many open dies in my office or closed dies. A closed die forging on the other hand — we put lubricant on both. I actually did measure the amount of material here. You take a billet or something and you forge it in a closed die. Now it turns out, even though clay is fairly soft, I have to put a lot of pressure on there because I'm getting to a fairly high Delta operation. But I'm also going to squeeze out the flash around the edge, right? And when I'm all done, if I've really squeezed this flat up against the thing — and if my lubricant worked, I didn't try it with lubricant earlier, I brought some lubricant because it didn't work very well without lubricant. Looks like it's not working very well with lubricant either. Um, well, maybe I'll take this one off, and there's my part. I've made a little beginning of half an Easter egg, okay. But I've got all this flash on the outside, um, and I've got to trim that, and that's part of my waste in my buy-to-fly ratio, right? But that's a closed die forging, and obviously you can get much better dimensional tolerances. That's the wrong lubricant for this application, is it. Anyway, um, fortunately the buy rate, the buy cost of this was not very high, um, in any case.

p11 11:23

Uh, you get much better dimensional tolerances, although you still have to have draft angles. You can't make something with absolutely square sides because you can't get it out of the mold, I mean, so far as that goes. But in any case, in a closed die forging, when you're trying to fill some cavity, if you just think about it — it's not all here's the flash and — but as you're trying to fill this mold, you will get situations where you actually get overlaps.

p12 11:52

Now how common are they? Well, they're not very common. I actually can only think in my career of one forging lap defect that caused a problem. Well, I take that back, I can think of two. One was a — um, Germans had forged a 20-foot long shaft for a plastic injection molding machine, and it had been machined and everything, went into service, and about a year and a half later had a big fatigue crack. And it had a forging lap that was about an inch long and about a quarter inch deep, and that's where the fatigue crack started. Duh, I mean, you get a flaw that's that big in something that's rotating. Um, and of course the Germans said oh, couldn't have been ours — except you always know it's a forging lap because the whole lap surface is oxidized. It was formed at 2,000 degrees Fahrenheit, okay, in the air, and so it's full of oxide and decarbonization if it's steel, and everything else. It's unmistakable. So eventually they had to own up to that.

p13 12:56

Another one that I had was actually not strictly a forging lap, it was a piece of, uh, pipe that actually finally had been laid in the Gulf of Mexico underwater, and they were doing the final hydro test before they started using it to bring oil from the oil well out in the Gulf onshore. And this thing had been hydro tested twice at the plant in Ohio where it had been made, and it passed the hydro test, and then it got down and they had to coat it with a plastic compound to give it corrosion resistance in the ocean and stuff, and then they put it in the ocean and they hydro tested it again, and when they hydro tested it, it opened up. It had a flaw 80% of the way through.

p14 13:47

What had happened is, in order to make that pipe — and it's actually sort of neat to watch this — they start with a hot billet at about 2,000 degrees Fahrenheit of steel, and it might be as long as this room basically, and it comes rolling down out of the furnace onto the bed, and then some metal jaws that are room temperature or slightly above room temperature, since they've been doing this all day, um, grab it, and then a mandrel comes through — and a big hydraulic press, big long steel mandrel about 40 feet long comes through and just goes right through it. It's called piercing the pipe. It's the way you make seamless pipe. You don't have a weld in the pipe, and they just punch — just punches right through it and turns it into a tube, and you're doing it with an outside die that's drawing all this over. You now made a seamless pipe, okay. And of course it's all oxidized and got scale all over it. And it's sort of neat — the, uh, it goes down to the next station, and it's still hot, and they hit it with a blast of steam on the inside, and the steam of course now is going to expand because it's not at 2,000 degrees, and you just see it blow all that oxide scale out of the inside of the pipe. Just hit it with a little bit of steam on one side and the other end comes shooting out all the scale all over. It's sort of, you know, they would make you pay to go and witness something like this if it was in an amusement park, right? But I get people pay me to go and look at these things. Anyway.

p15 15:25

Um, so what was I getting into all that? Oh, I was talking about defects. That particular one — a piece of the mandrel had broken off and had gotten embedded into the surface of the steel. So the mandrel is relatively cold, by the way. They have hundreds of mandrels, and the mandrels can be used once before they get rotated back through to go through a cooling process. So the mandrels have to be cold, they got to be pretty strong to pierce this thing, and so they get used once. Well, they get used once and then they come back about an hour later after they've cooled down. So they got about a 100 of these mandrels just running through in circles. And so they fatigued, and this one had fatigued. No one caught it. They did a non-destructive testing inspection, but it was at the end of the pipe. The technique they were using, except the ends where they had to grab it, for the main 40-foot length was um, an automatic eddy current inspection. Very reliable, 'cause this defect was like 3/4 inch long and 3/8 of an inch deep. Um, the problem is on the ends they couldn't do that because they had to grab it for the eddy current inspection, and so the ends weren't inspected, but allows that you can go in there with magnetic particle. And that's all done with a bunch of looks like immigrant labor, uh, probably illegal immigrant labor, but they go in there and they squirt little magnetic powder on the inside, and they got a flashlight, and they look to see if they magnetize it and see if they see a flaw. And you can imagine that if you get to do that for 8 hours a day, you might miss something from time to time, and they did. They missed it.

p16 17:09

Um, so it wasn't if they hadn't inspected for it, but they missed it. Things will get through, um, even when you have very good quality control. If there's still people in the process, the error rate is still on the order of a percent or so. Just because something's been inspected, don't assume that it's good, okay. Uh, you're still going to have some failures.

p17 17:31

In any case, you can have lapping defects. Another type of defect which I'm showing you, 'cause it's in the books and they talk about it — I have never seen it personally, but I do have a sample of it here. I did have a sample of it here, um — huh, what happened to my sample? Uh, oh here it is. Okay, this is — if you're extruding something, you get an end effect when you have the die trying to extrude out the two — the hole. Oops, you can't see that, can you? What's a problem with this thing? Something. Okay, so if I have a die — oh, not too far, this thing does have a lot of overshoot on the — okay. So you got a die, you're extruding this and at the end you basically get a dead zone. I mean, if you're extruding this way to make a cup, in this case you're just extruding and you're getting to the end, you end up with a dead zone that forms because the metal can't flow, it doesn't want to go too far in that direction and turn a 90 degree turn. It wants to have flow lines, right? And so you can get a little defect — this is one of Backofen's samples, this is an extrusion, you can see it's the end of the extrusion press and it's got a little dimple right there. So that defect is inconsequential, but in some cases that type of defect shows up in forgings, uh, because people weren't careful about the end of the forging, okay.

p18 19:07

Um, we can go through this as much detail — well, if I'm talking about defects in forging laps, this is — let's see if I can — these are some examples of forging laps out of the steel mill. Scabby surface — this case is a bloom, so it's actually been rolled. These are rolling laps. Uh, semi-finished roll product, that one you can't see very well. And here's the types of things, semi-finished roll product. These little grooves in here are probably scaled, it's been rolled in, and they could be an eighth of an inch deep, okay. This whole thing is probably 6 to 8 inches wide. So when you see grooves like that, they're pretty sizable things.

p19 19:55

So in fact at the steel mill you have a buy-to-fly ratio in a sense — in fact, they call it yield, at a, whether it's steel mill or aluminum mill, how much you cast to how much you ship. And when I started at Bethlehem Steel in 1974, the overall product yield was about 65% for the whole corporation, all products. And Bethlehem was second largest steel company in the world at the time. They made everything from wire to I-beams to huge forgings, you know, everything, sheet metal. And they had an overall 65%. If they could have gotten to 67%, they could have doubled their profitability — maybe even 66%, maybe a 1% change. Now, what happened is somewhere between '65 and, uh, in the early 1980s, over about a 20-year period, they went from ingot casting to continuous casting. And what happened then to the yield ratio was it got up from 65% to 97%, okay.

p20 21:03

So if I cast an ingot, I actually have slightly beveled sides to the mold. And by the way, the mold has to have fairly thick walls because it's going to — if it's just a cast iron mold, it's going to have to carry out the heat from all this liquid that's in here. You fill it up to some level of liquid, but the liquid shrinks as it solidifies, and it is what we call a killed steel. The final product looks like this, and you have a big gap here. You take — the welding course I actually talk about how one company ended up forging this casting into a 33-inch diameter shaft. I told you about my 33-inch diameter fracture in the mid '80s. This went as a propeller shaft on a liquid natural gas carrier, and it stopped dead full of LNG in the middle of the Philippine Sea. And they were airlifting in Lear jets fittings so that they could transfer the LNG to another empty LNG cargo tanker that was coming back from Japan. This is a whole series of LNG ships going from Indonesia up to Japan and back, and so this one was going up to Japan full, and when it lost power and couldn't go anywhere, full of LNG, you could have had a big explosion which would have upset a lot of people. May not have killed very many. But remember, those same LNG tankers come in once or twice a month into Everett over here, okay, right through Boston Harbor. And one of the biggest concerns that homeland security has for Boston is, all you need is a rocket-propelled grenade when that thing's coming in, you will wipe out half of Boston if you do that. So they have a little security to try to make sure there's no terrorists with the RPGs. But anyway, nonetheless, this thing stopped dead in the Philippine Sea.

p21 22:54

Well, the problem with the old ingot casting of steel was you threw away the top one-quarter of the steel. You were down to 75% to begin with. So in spite of the scale losses and the forging mill, or the end crops in the sheet metal mill or whatever, you're lucky to get 65% when you lose the first 25% in your casting process. Well, it turns out, since yield is such a big deal — well, and actually at the time, the United States was making 150 million tons, okay, of steel a year here. Supp — I guess M-M for million — and they were shipping about 100 million tons. There's your 60, 65% yield.

p22 23:38

Well in the '70s — basically in the decade of the '70s, but starting in the '60s — they switched to continuous casting. In continuous casting you don't have to crop the top, okay, because you just cut it off, and then the bottom of one is the top of the other. And so now your yield shoots up, and now in the worldwide steel industry, they still have to ingot cast some big forgings, where you may still have a 60% yield on your final product, but most of the product is continuously cast, and some steel companies have 97% yield now.

p23 24:14

And what happened to the steel industry — several things happened, but one of the things that happened is, all of a sudden if in a 10-year period you were producing one-third more steel than you needed in the casting shop, because now you switched to continuous casting, all of a sudden your capacity in all your plants around the world is 50% larger than you need. And so there was a tremendous overcapacity because of continuous casting. Here is a wonderful development, but it nearly bankrupt the steel companies because they started competing with each other on price for the market because they all had excess capacity, okay. And so the world decided the steel industry was a dying industry when in fact they had just had a huge gain in productivity, but that gain in productivity almost bankrupt them, right? Think of the irony of this, right. So the moral of this story is, you have to be careful about what you wish for, right, okay. Big disruptive changes in technology can be very harmful or can be very painful — not harmful, but painful.

p24 25:32

In any case, what I wanted to talk about — it may not be obvious until you think about it for about 30 seconds, but if I want to start with a square billet and make it into another smaller square billet — this is actually out of the US steel book, and they show you the types of dies. First of all, you don't take a square billet like this and to form it on its flat sides. You deform it on the edge, on the peaks, right. And so you go from being a square to being a parallelogram, okay. And then you flip it 90 degrees and you take the long axis of the diamond and you squeeze it again. It sort of makes sense if you think about it. Um, it's easier to make dies, you don't have — you're not trying to squeeze things from the side. Um, but this gets sort of messy in terms of the types of shapes you might — if you wanted to make a straight I-beam, your final passes look sort of strange, okay. The I-beam's coming through on a slightly diagonal.

p25 26:45

This is an example of — the thing that built the steel industries in the 1880s was the railroads, okay. Before we had the automotive business, it was the railroads were the big consumer of steel in the world. That's what made Andrew Carnegie the richest man of the world. He was the Bill Gates of his day. Um, if I told you the story of Andrew Carnegie and my grandfather, I'll tell you that sometime. Anyway, um, so you start out with a, uh, something that comes out of the blooming mill, which is called um, a bloom, okay, which is just — sometimes called a billet or you call it a rod — and these are the shape changes you go through to end up with an I-beam or a steel rail, okay. So they've got 10 steps, 10 sets of rolls to make a railroad way wheel, okay.

p26 27:40

Um, here's another example, I mean I can show you examples all day, but out of the same book, of the breakdown pass — you first just kind of squeeze it in and you get these things, you finally end up with the finishing pass. And in fact, the thing that made Bethlehem Steel — Andrew Carnegie had started US Steel, and it was the king of all the steel mills in the world in the 1880s and '90s, and there's a guy, uh, his right-hand man in Pittsburgh was a guy named Charles Schwab. Not Charles Schwab — well, Schwab, but I — Charles Schwab's the investor, I think it might have still been Charles Schwab, but anyway, um, I can't remember the Schwab of Bethlehem Steel 1890 fame. But, um, they had a little bit of a tiff, and Schwab went off and found this little iron mill in Bethlehem, Pennsylvania. And those guys had just learned how to put molybdenum in steel to make high-speed steel.

p27 28:38

And without getting into all the metallurgy of that, it turns out, um, molybdenum forms molybdenum carbides which don't dissolve until higher temperatures, and so you can get these hard carbides, and the steel will retain its hardness to about 1,000 degrees or 1,100 degrees, where ordinarily at 7 or 800 degrees the steel is starting to soften. But you get what's called a secondary hardening peak because of these molybdenum carbides, and you can have a hot work tool steel that can hot work at 1,100 degrees Fahrenheit as opposed to 6 or 700 degrees Fahrenheit. Well, let's face it, this is going to improve the life of drills and all kinds of things that get hot when you work them. And so that was the first big invention that helped grow Bethlehem Steel.

p28 29:24

And the second one — Bethlehem Steel was the first firm to learn how to make I-beams, okay. They learned how to roll passes like this for making I-beams, and they were close enough to New York City — the reason New York City now has buildings more than six stories tall is because Bethlehem Steel made I-beams. And New York City — Manhattan was sort of flat, okay. If you go to Manhattan, you won't find a lot of buildings over about 10 or 12 stories tall — actually those you won't find anymore 'cause they turn them into 100-story buildings, but if you had gone there in 1900, you would have found very few buildings that were over about 10 stories tall, 'cause brick and mortar just didn't give you enough backbone to build a building. But steel I-beams did, okay. Um, the other thing you would find in New York City in 1900 — what would you find? 40,000 gallons of urine and horse manure. I don't know how many pounds of horse manure on the streets, 'cause that was the biggest pollution problem in all the big cities around 1900. So the automobile solved that. It just put out smog. So, anyway.

p29 30:47

Um, that's 10 or — no, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11. Student: Is there a heating step between? Yep, you have to reheat it after each one. Well, you might be able to go from this one to this one, got a lot of mass, but then you have to have two mills, because usually these are bigger mills because they're bigger pieces. Once this thing's getting long like this, yeah, you have to usually — well, many time — I wouldn't say you have to do it every time, you might get two passes before you have to put it back in the furnace. It's a slow process. Student: This is a rolling process? It's a rolling process, okay.

p30 31:18

This is — here's all the pass — this is another series of passes to make an I-beam. You end up with — so that doesn't look like it's going to be an I-beam, does it, okay, at that stage. Um, Student: [inaudible] this was all done by trial and error? Um, yeah, basically. And but trial and error where they had to build huge rolls to do it. These are two steps — well, this is a step to make a channel, and this is to make an angle, okay. If you're — I mean, I can actually — this is — here's another one to make an angle. Notice this one to make an angle, you're basically rolling it just and sort of peaking it. This one you keep it pretty flat to make that same angle iron, keep it pretty flat, and you're just forging sort of a — you're rolling, not forging though, a little peak on it, and finally you just straighten it out, okay. And here's a Z channel, or actually it's a piling probably. Actually this is a piling, okay, um, a piling, or interlocking things, that this one interlocks into that one. You drive them into the ground so that you can dig a pit in the ground and work in the pit without the wall caving in on you. Uh, a lot of steel used for pilings. Anyway, so this is what you have to do to form a piling. These last ones, I don't know if they're hot formed in that same type of mill or they're probably a separate little mill that does that. Anyway, it's — I think it's sort of interesting to look at this.

p31 32:54

Nowadays they actually probably would do it on a computer, but back then, it was sort of trial and error. Student: Are processes still roughly the same now? Um, I'm going to show you how they do it now, but until 1985 or so, yes — for about a hundred years that was the way they did it. And actually, for some of the smaller beams it might be 20 — I can't find my overhead from 10 years ago, but, uh, the example I gave you — they, for some of the smaller I-beams it might be 20 passes, or over 20 passes to go through the mill, okay. Gets a little pricey. And in fact, what that meant is some of the smaller beams were pricier than the big beams. If you think about it, right, it takes a lot more rolling and reheating and everything else.

p32 33:47

So far as people doing things empirically — so once, 20 years ago — 20, actually probably 25 years ago — I used to consult for a gold company down here in Attleboro, Massachusetts. And I think I mentioned, they went through seven tons of gold a year, and they had their own continuous casters for karat gold alloys. And that plant — it's still there, um, it's been there for 113 years now, started in 1899, so right over the door. Um, but that plant, um, to make gold tubing — now, I actually got them set up with a gold welding machine, okay, so they could weld karat gold tubing longitudinally. Um, I had to go to Steamboat Springs, Colorado one winter to, uh, see the guy who makes these — he just had a big machine shop in Steamboat Springs, and he wasn't building welding machines, he was out skiing, right. Um, but anyway. But before that they made gold tubing. Now what do you use gold tubing for, anybody have an idea? If you make heavy wall gold tubing, you can machine it into little circular things called rings, okay, uh, so a lot of wedding rings are made with gold tubing.

p33 34:58

But one of the big very profitable things in the Boston area — you go down here to, uh, downtown Boston, there's a Jewelers Building in place, and there's a bunch of flute manufacturers. I guess now I'm supposed to pronounce it "flout," right, a "flout"-ist, not a flutist, they're a flautist. Have you heard that? Anyway. And, so, Professor, uh, oh — name — um, Professor in this department, Harry Gatos, is a Greek, and he graduated from the department and then he went off to Lincoln Lab and headed up the solid state division, and came back here in the mid '60s, and he was called Mr. Gallium Arsenide. He was the world's expert on gallium arsenide before it went into cell phones. He's the one who really did all the basic research to make gallium arsenide a viable semiconductor. He was also a very accomplished — or he is also, he's well into his 80s now — uh, very accomplished flutist. And if you walked down Building 13 when I was a student, you'd hear him in there during the middle of the day playing his flute. Uh, he played for the Boston Symphony and a few other people as a guest soloist and stuff. Uh, he had gold flutes, silver flutes, gold flutes, and he had one of the six platinum flutes in the world at the time, okay.

p34 36:22

Um, but anyway, they made their tubes by a deep drawing process like this, and, um, just great big long press just coming in. You know, just start out with a piece of karat gold and make it. And then they also would do gold-filled material, which is a clad material, clad on the copper, brass, with gold on the outside. And they asked me to redesign their dies, which is why I started thinking about this — you asked the question, was this all done empirically. If you looked at the dies, some of them — when you went from one die to the next, some of them had like a, less than, I think one of them had a 9% reduction, and another one had a 35% reduction. There's no logic whatsoever. And so I redesigned their dies for them. It's basically going and calculating, you know, logical steps of progression, okay. Uh, some of the dies that had 40% or something might be cracking, you know, giving failure, high failure rates. The other ones at 9% could have been setting up stresses that actually could have been bad later on, okay. Remember, I told you, you want to get a fair amount of deformation in. You don't want to do these little light passes. They can do more harm than good. What you have — oh, okay.

p35 37:39

So anyway, um, there — oh, so now let me tell you the Chaparral Steel story, okay. So Chaparral Steel, founded by a graduate of this department, who's now retired. He was — he's from Vancouver, um, but he worked for a Canadian steel company in the mid '70s, and what Gordon — his name is Gordon Forward, he used to be on the department's visiting committee back when I was department head. Um, but Gordon realized in the mid '70s that it cost $200 a ton to make cast iron in the steel mill or in the blast furnace, but you could buy scrap iron at the time, scrap steel, for $100 a ton. Well, what do you do with the cast iron from the blast furnace? You put it into the steelmaking furnace and make it into steel. If it cost you $200 for the liquid cast iron, and you could buy steel scrap, steel for $100 a ton, you have a $100 a ton advantage on a $300 or $400 a ton product. That's not a bad advantage.

p36 38:53

And so Gordon and a few other people around the world realized this, and they started what they called mini mills. Instead of a steel mill that was as big as — well, the last steel mill built in this country was the Bethlehem Steel Burns Harbor plant, from 1965 to 1972. Took 7 years to construct, it cost Bethlehem $5 billion, okay. Bethlehem was the second largest steel company in the world at the time. They started the Burns Harbor plant which is still in operation. Um, but it almost bankrupted them, okay. I think they were only about $5 billion a year in sales, and so they were spending a lot of money on capital. Um, and it'd be equivalent today of um, Intel having to spend $20 billion to build a fab plant, okay, or a $60 billion fab plant. It can bankrupt a company if things don't go right. Or Boeing deciding to build a Dreamliner or something like that, it can bankrupt the company if it doesn't come out properly.

p37 39:54

Um, so Bethlehem actually finished the plant, and by, uh, '74, they were the most profitable year ever. That's when they hired me, and they've been going downhill ever since they hired me. Um, but since then, no integrated steel plant anywhere in the world has ever been built by a company. There have been plenty of them built, but only by countries. They have been built by the backing of an entire nation. POSCO Steel, which became the largest steel company in the world in the late '90s, was built by the Korean government, okay. That's where the financial backing came from. So it's a political thing to make that big an investment.

p38 40:38

Well, the mini mill guys, they could build a mini mill to turn out product for, um, one tenth the capital cost of an integrated steel mill, and they did. And it turns out today there's actually interesting stories about this, but today about half of the US steel market is made in mini mills. Well, what Bethlehem did — not Bethlehem, but Chaparral did — they were making rectangular blooms, and then they decided, why can't we make something a little closer to near that shape? And so they did. They started casting a dog bone. And I'm — don't know if I'm going to take all these out, but here's a closer view of the — now, this — the problem is back there. No, it's — well, there might be — back there is a backlight, okay, now I got to figure out the backlight. Um, there you go. Does that work? Doesn't help much, it's these stupid — I got to find a way to store these films, but anyway, that's basically the same thing.

p39 42:11

But the one I do want to take the time — they kept on improving that until they got to this, which is still got problem. That's the backlight now. Uh, got to get to the upper, okay. Well, fy — um, well, I'll pass it around, you can look at it if you want. But now it's almost an I-beam that they're casting, and they went from 21 passes for the smallest I-beams to about seven passes. Wow, okay. Um, but not only that, um, it used to be — I read some of my notes here from years ago — oh, um, for roll steel bars, um, before they — Beth started doing this — 41% of all the, um, I-beams were imported into the United States. This is mid early '90s. But by the time they did this, they wiped out all imports because they were so much more efficient by doing net shaped casting to make the I-beams and wiping out two-thirds of the rolling passes. And they were exporting 20% of the product outside of North America, including Asia, which in the early '90s was just incredible.

p40 43:16

They had 75%, um, energy savings. Uh, they had inline reheat because they didn't have this big heavy thing that you had to stick back in the furnace. They were starting with something after the first couple of passes — it was so thin they could basically heat it in line with induction heaters, okay, and not have to put it back into the furnace. So they did inline reheat, 28 minutes total to go through the mill. And they ended up with a fine grain chillcast structure that, um, had a higher tensile strength and toughness. Um, they were able to roll these things to get a finer grain size, and if you know anything about grain size and properties, grain size is the only thing that gives you both increased strength and increased toughness.

p41 44:06

And at the time, the, uh, garden variety steel is what's called A36, ASTM A36. It just so happens — this just coincidence — A36 has a yield strength of 36 ksi, okay, 36,000 pounds per square inch. If you go to the manual, the American Institute of Steel Construction manual, they would have back in 1990, um, design criteria for, um, 36 ksi steel, 50 ksi steel, and 100 ksi steel. The 100 ksi steel was quenched and tempered steel, fairly expensive, but they put that in bridges and stuff when they needed weight savings. Um, but the, uh, the 36 and 50 was the standard. And of course you get a premium on the 50 ksi steel, 'cause you can save one-third the weight when you got 50% more strength, right.

p42 44:57

Well, it turns out that Chaparral, by doing fewer passes, finer grain size, better reduction, was able to make product that could pass for either A36 or grade 50, either 36 ksi or 50 ksi. And they had — they could cut their inventory in half. They didn't have to — what was A36 and what was grade 50, order whichever one you want, we're going to ship you what we got, which meets both specs, and you're getting a better steel if you ordered A36, but we'll sell it to you for the same price as regular old A36, 'cause we don't care. It's profitable enough that we — we're going to save money on the inventory. We cut our inventory down by half, right. So I mean, here's an innovation that made Chaparral a world beater, okay, on the I-beam market by going to net shaped casting and getting rid of a lot of these rolling passes and getting a better product, okay. So it's sort of a neat little story of something that uh, helped them be — become profitable, um, or more profitable.

p43 46:15

Um, there's lots of good stories about, uh, Gordon Forward and Chaparral. I remember — well, I'll tell you the story when I worked at Bethlehem Steel. Um, in the mid '60s, Bethlehem had built a Taj Mahal. It was the fanciest research laboratory in the world, probably, for metals research. It's called Homer Research Lab, after one of their chairman named Homer, and it was on top — it was just north of Lehigh, well just south of Lehigh University. Lehigh's on the side of a mountain. Actually the Bethlehem plant is north of the river, south of the river is a couple of sort of slums where the steel workers lived, and then going up the hill which is called South Mountain, was Lehigh University on the side, and then you get to the top was Homer Research, and that's where I was hired, and I worked. And $600 million facility, uh, in the late '90s — or the late '80s — Bethlehem, when they were going under, gave it to Lehigh University to get it off their tax rolls, okay. But it's like 20 buildings, and these are — it really was a Taj Mahal, um, and whatnot.

p44 47:32

Oh, so, at Homer Research when I was there, we had the corporate dining room, which was a second floor with huge windows looking out on the vista of the steel mill, and you could see the smog from the smoke of the coke ovens over here, and you could see the pollution from the blast furnace over here and stuff. Anyway, um, they had an executive dining room, which I snuck in once at about 2:30 in the afternoon, just to say I'd been in there. But this is where the big shots, like the Don Trauttleins, would eat lunch. And I remember seeing a strawberry the size of a peach, okay. I thought, this is amazing, okay, it was going into the executive dining room. Um, and if I told you all the stories of how the upper management took care of themselves — I lived near the airport, and they had three Lear jets which would fly the vice presidents down to Florida in the winter to play golf on the weekends, okay. This is good corporate management in the United States circa 1975, which one of the reasons that after 13 months I said, where do I want to be in 5 years? And the clear answer was, not at Bethlehem Steel. And so I left and came back here. I don't know if that's a good idea or not. Anyway.

p45 48:39

Um, in any case, at Chaparral Steel, so I did a study once with a guy from Harvard Business School, um, and uh, we spent a few days down at Chaparral and visiting Gordon Forward. And uh, Gordon used to always talk about the corporate dining room at Chaparral. And the corporate dining room at Chaparral was a couple of folding tables, plastic chairs, not as nice as these, nowhere near as nice as these. Um, and then you'd have plastic forks and the little things, 'cause they would go out and get some barbecue from the local takeout, right. Uh, that was the corporate dining room at Chaparral. So difference in philosophy, um, between the people who were at one time printing money.

p46 49:22

Um, and talk about printing money, let me — I think I'm probably just going to show this one other thing, and then you have to — what made the American steel industry, after World War II, great? We controlled 75% of the world's production after World War II. Why? We had bombed out all the rest of the capacity. Hey, it's not every day that a businessman gets to go and wipe out his competition that way, okay. But the managers — in think about it, 1975 when I was working for Bethlehem Steel, the managers had been hired about 30 years before, right after World War II, and they came in with this arrogance, all through their careers they controlled 75% of the world steelmaking capacity. At that time US Steel was 40% of the world steelmaking capacity, and Bethlehem was like 20%, okay. The two of them alone were over half the world steelmaking capacity, because we didn't fight the war in North America. At least not a lot of bombs fell in North America.

p47 50:30

And these guys thought they were the brightest managers in the world. It's sort of like a financial guy today on Wall Street who walks away with a half a billion dollar income. It's because he thinks he's the brightest guy in the financial market in the world. And he's not brighter than anybody else, he just happens to be in the right place at the right time. But the Bethlehem Steel managers when I was there, the guys at the top thought they were the most clever people in the world. And you know, it's why they no longer exist, okay. They thought that they were great managers, and in fact they were flying to Florida on the weekends in corporate jets to play golf, and they were eating strawberries the size of peaches, okay. Um, and they couldn't sustain that. And frankly, I thought it was stupid when I was 25 years old. But unfortunately I've seen it repeated again and again in a number of industries.

p48 51:24

Kodak. They were printing money. They called it film, but it was money, okay. They had like a 95% margin on film, and they would not give up their film. They invented a lot of the best digital photograph technology in the world, but the film guys didn't want to give it up because they had grown up knowing that film was their bread and butter, and they kept their bread and butter until it was all stale and rancid, okay. So it's not just the steel industry, it's Kodak. And we could go through a bunch of other industries. Um, but, um, people — it just amazes me, these people at business schools — and they'll write articles, I mean, I remember articles written by some of my friends, one of whom became Dean at Harvard Business School. Um, he was writing articles on the steel industry that were just — I said, Kim, that's wrong. He says, no, it's not. Right. He had an — he had an economics degree, okay, he studied labor relations, and he was lecturing me on the steel industry. Now, he became Dean, and so, you know, he wrote all these papers, and industry would pay him tens of thousands of dollars a lecture, because he was, you know — but that doesn't mean he knew anything, okay.

p49 52:44

Anyway, um, so this is how you make a railroad wheel. Again, the steel industry grew up from Andrew Carnegie through the 1920s living off the railroad. You start out with a block, you forge it into a disc, you then take that and forge it after — second forging — and you make a little bit of a net shaped indentation, and then you punch it to make a hole in the center, and then you roll it. And so these flanges right in here are actually rolled. See the wheel here vertical, see there's the punched hole horizontal, and they're rolling in a special mill these flanges, and then they cone it to give it the right contour in another, uh, operation. So that's how you make a railroad wheel — mixture of rolling and forging. Uh, just gives you an example of how people vary those things, okay. So I'll see you — well.