CAS_Su2011_05

Okay anybody have any questions from before? We've been talking about casting and we're trying to impart geometry, structure, and homogeneity. The last time I spent a lot of time on continuous casting. I guess what I want to say about continuous casting right now is, the problem with continuous casting is it's a process that's too productive. Um, a big plate mill continuous caster, where you're casting something — it might be eight, six, eight feet wide and ten inches thick — the continuous caster will probably do three to five million tons a year. Well there's not that many places that need or have that much hot metal to be able to cast that amount of material.

So continuous casting is something that, uh — and I'm going to give you some examples in the copper industry — there's only so many continuous casters, big continuous casters, you can tolerate in the world. You can make small continuous casters, and I passed this around before. This is a piece of copper that was cast from a continuous caster, and the continuous caster is about half the size of this room. It's not very big okay. And you can make bar — I think this is inch and 3/8 diameter originally, and they were just going to turn it into electrical wire. Well, there's lots of electrical wire, and so you can cast this — you know it kind of comes out at a few inches a minute out of the caster — and then if you draw that into wire, that's a lot of miles of wire so far as that goes. But you got a lot of one product.

Now there are certain industries like the automotive industry, where you have lots of volume, and you can use something that's a continuous production mode. But most things, or many things, you really can't, you don't have, you can't absorb that much of something. The nice thing about, or one thing about continuous casting, we're producing simple geometries — rods or plates or something like that. I did show you the I beams, and that continuous caster is about two or three times the size of this room. A big plate caster in a steel mill is half the size of this main complex of buildings. Okay so there's all kinds of different sizes of things depending on what you're trying to do.

The other thing about casting is, if you're extracting the heat of fusion — we talked about — as you get to larger sizes, much more than a foot in diameter, you're talking about hours to extract the heat. Now, we also, and we'll talk about a little bit later as we get down to structure and homogeneity, sometimes we like to cast things very thin. I used to have some rapidly solidified amorphous metal, and if you cast it paper thin, and we can talk about that in a little bit — um, I couldn't find my amorphous metal sheets — but you cast a paper thin, you actually can make it extremely homogeneous, you can make it essentially as a glass. And so there's everything in between and it depends on what people are trying to do.

Now, so section size is one problem with extracting the heat of fusion. Another thing that people do is, uh, there's undercooling. And undercooling is basically the situation that if you have some temperature, um, temperature versus time, and you're extracting heat, if this is the melting temperature, the thing should solidify here. But in fact it won't nucleate, and it may go down in temperature, may remain a liquid down here below the melting point. And then when you finally do start to nucleate, as the solid starts to form it gives up the heat of fusion, so the thing heats up and it ends up on a cooling curve like that. So you get undercooling below the melting point because of this barrier to forming a nucleus of solid material. And that's because the solid material has a certain surface energy. And we don't necessarily want to turn this into a course on nucleation and growth, but you have to form some solid particles to start growing, and you need a little bit of energy differential, a delta T of undercooling, to do that.

Now, people have tried — NASA spent all kinds of money because it justified doing some experiments in space, and they just wanted to prove that they could spend money in space, um, and tell Congress it was science. But let's see if I can find my undercooling curve. Anyway, the, um, the more you undercool something — you'd like to undercool it such that the delta T, this would be delta T here, if the heat of fusion is less than the heat capacity delta T, then you basically could undercool, and when you nucleate you could solidify the whole thing all at once, instantaneously. And it turns out that they've done that, but only for one element in the periodic table, and that happens to be phosphorus.

Um, and now I can't find my phosphorus undercooling slide, but anyway, um, I'm sure I'll find it eventually here. Maybe we'll have to come back to it but let me just — [Tom flips through slides looking for the phosphorus undercooling plot.] oh I put it at the back as very end. Um, so this actually is a plot of melt supercooling where you undercool, and it's normalized to essentially this delta T equals to the heat of fusion. When you're below one on the normalized curve you're hypercooled, which means you've extracted more heat from the liquid than the heat of fusion. And the reason we can do this with phosphorus, doesn't have a very big heat of fusion, doesn't take a very big delta T. In fact the delta T here is on the order of 20° or so. But anyway the velocity that the solidification occurs ends up getting up towards, um, something not equal to the speed of sound but something fairly fast. Actually it's not close to the speed of sound, but anyway you actually cool, or you freeze very rapidly when you don't have to extract the heat of fusion.

Now one thing that was developed at MIT is a shape casting process, uh, where essentially they extract about half of the heat of fusion. This is a genealogy, you don't have to read it all. Shape casting processes where you use a metal mold and high pressure die casting — this whole thing is called semisolid processing. What do we mean by semisolid? You ever had soft serve ice cream? That's semisolid. It's about 30% liquid, it's about 70% solid little ice crystals. Uh, you ever had a Slurpee, okay, at 7-Eleven? Okay, that's about 50/50, 50% ice, 50% liquid okay. It's a semisolid, it's partly solid. So, you know, your Dairy Queen soft serve ice cream is actually liquid and solid ice crystals together.

Well it turns out Professor Flemings here, back in 1970, had a student trying to do a fundamental study of solidification. And the student ended up writing up his doctoral thesis, going off to spend his summer for a couple of weeks with the Army and reserves, and turn in his thesis thinking that — back then you didn't have word processors — that Professor Flemings wouldn't have the heart to change very many things, he'd be able to come back from his summer camp with the Army and graduate. Well Flemings looked at it and said, this isn't a doctoral thesis, and threw it back in his face when he got back, and said start all over again. And at that point they basically changed the experiment, and a year later he ended up with what's called rheo casting — for rheology is a study of fluids.

And essentially they would cool the liquid down and they'd stir it just like you make soft serve ice cream okay. And they were doing this on lead-tin alloys, but they showed they could get unique structures and they've already extracted half the heat of fusion. And so they thought this would be great for die casting, where you inject the metal into a solid mold, just push it in under pressure. And the problem in die casting is your cycle time. You got to push it in and hold the pressure on there until the thing solidifies. The bigger the part, the longer it takes to extract the heat of fusion. If you've already gotten half the heat of fusion out, you can cut your cycle time down by half.

So if you look at high pressure die casting, there's this whole technology of rheo casting — it's a slurry — thixo casting where you essentially freeze it hard but you've created this structure that it turns out when you heat it back up, you can end up with a rheocast type of structure, and thixo molding where you actually end up with something that's soft, and if you use dies at the same temperature you can actually mold and get a shape to this stuff, kind of like playing with Play-Doh. Um, that was all developed here at MIT. I was the student who helped the guy finish his first thesis. I left after he got back from summer camp, so my name is not on the patent okay. I was an undergraduate grunt for him. Uh, if I'd stayed working for him for another six months, I would have been the patentee of all this.

This is also an example of MIT's big mistake in patent technology, because they patented it to a company that sat on the patent for about 10 or 15 years and never did anything with the technology. And this is actually why MIT changed its whole patent position, um, in the mid 1980s, where essentially MIT, if they license a patent to somebody now, to some company, they have march-in rights. If you don't do anything with it, they'll take the license back and give it to somebody else. Because this technology lost about 10 or 15 years because they licensed it to a company that didn't want to see it come to market — they're, you know, because it was competition for them. So the company licensed it to keep it out of the marketplace. So MIT learned their lesson on that.

And so no one — MIT didn't make any money, other people make a lot of money. And in fact today there are a lot of automotive parts that are made that way with rheo casting, where they've already extracted half the heat of fusion. But if you have a fancy Canon single lens reflex, they have a magnesium body okay. It's a black camera body, but inside is this very thin walled magnesium housing to make it nice and lightweight. That's rheocast. And the reason it's rheocast, if you've extracted half the heat of fusion, you've also gotten rid of half the shrinkage, and so you can make finer parts, thinner wall parts okay. So far as that goes. So the technology is used a number of places.

Um, so in casting we're trying to impart some sort of — we talked about continuous casting, and I may talk about a couple of those, but in general we're trying to impart some sort of shape, and we've got this problem with shrinkage over here. And this shows a casting — just to cast a part like this, you may have to have a mold that's three times as high. If you just cast the thing into a sand mold — quite often for steel or cast iron we just, people used for thousands of years, sand, silica sand — that you form the shape of the part you want to make, but you have to pour in this extra metal. Initially you pour it in liquid all the way to the top here, but because of shrinkage you get this pipe. It looks a lot worse, but if this is a cylindrical part, the void volume in the center is small as you're shrinking the radius down to zero. But you can end up with a pipe that is deeper than the actual casting.

If you do some other things like put a chill block in here, so that you extract a lot more of your heat from the bottom, you actually can get rid of a lot of that pipe. And so it just — you're starting to shrink here, and so a lot of the shrink, the volume here of this is the same as the volume of that. Remember you got this radius thing you got to worry about. So the area, the volume out here at a large radius is much greater than the volume of that area there okay. It's r dA if you will. Um, in any case, if you were to cool entirely from the bottom, all you're going to do is push all your shrinkage up to the top and you'll have a very flat surface across here, if you think about it.

So by putting chill blocks into the sand mold you can change the amount of what we call rising in the casting. In the foundry industry this thing above the actual part is called the riser. You're going to cut the riser off and send it back to be remelted. Um, but if you put in chill blocks you don't have to have anywhere near as tall a riser okay. You also can improve the structure. So they also put in things like hot tops or radiation shields on the top, so that you get a little bit more one-dimensional heat flow. And you'll get — here you have a chill block with a shorter riser, but if you put a hot top on here you actually can get a very small riser. And if you put a radiation shield up here you can get things very flat. Anyway, you can change the cooling on these complex shapes by playing all kinds of games, and we do okay.

But a lot of — well, actually today computers are fast enough that you can calculate these things. Twenty-five years ago people couldn't calculate — it was kind of too complex a problem, the heat of fusion problem, where you have a step in one of the properties. For all you mathematicians out there, and you're all mathematicians this summer, if you look at the heat of fusion or the enthalpy versus temperature, as the metal heats up it gets to the melting point, and then you have this step which is the heat of fusion. And this is the enthalpy of the solid and this is the enthalpy of the liquid. Um, the slope of these things is what, anybody remember from your heat capacity, from your thermodynamics? It's heat capacity, I just said it okay. The slope of the enthalpy versus temperature is the heat capacity, if you ever took thermo, anyway. And so these have slightly different slopes because the heat capacity of the liquid is different than the heat capacity of the solid.

But this heat of fusion, this discontinuity — mathematicians don't like — and in fact it's called the Stefan problem, after a mathematician named Stefan. Stefan-Boltzmann constant, um, Planck radiation — the Stefan did work on that. But anyway, that creates a problem for all the computer jocks okay, this discontinuity. But it's a real problem. And so back, computers were not powerful enough 25, 30 years ago to really handle that problem very well. Today they don't have a real problem okay, they can handle it. Um, still not easy, still a lot of computation to handle that problem where you have an isothermal hold so far as that goes. Anyway, there's lots of things that people do to improve the heat flow, and to try to do something about the undercooling — or undercooling is one way, rheo casting is another.

Um, one of the most complex casting processes is investment casting. Um, and we make some of our most sophisticated parts this way, uh, even today. But in the old days it was known as lost wax casting. And it's been used for about 3,000 years. So if you can make a part out of wax — you can take wax and you can carve it. I made my wedding bands originally, um, for my wife and myself, by lost wax casting. I just turned some, uh, wax on a lathe, took it up to a dental shop where they make little bridges for people's teeth, and they had a lost wax caster. I paid the guy 20 bucks and gave him the platinum alloy I was making. He didn't have that in stock. Um, and he cast a bunch of wedding bands for me. In fact I have a couple of spares okay. I figured I was going to, we might get — videos rolling, I don't know — now my wife has a couple of the spares too. Um, I figured as we got older we would gain weight and our fingers would get fatter, and so I made them a different size.

So anyway, um, but lost wax casting, essentially you make a wax part. Today we actually have a permanent mold which you can machine some complex shape, and you inject wax in there, so you can do all this at, you know, less than 200° Fahrenheit. You pop the wax part out of the mold. You now basically take some wax rods and stuff, and you put these little parts and wax, and you weld them to the other wax. And so you go into one of these shops and you'll see a — this is hand work and women do better with hand work than men okay, and they take little hot knives and they basically just weld with wax, the wax to the wax, and they weld it up. And they'll make this little Christmas tree of parts with a riser, or sprue, they call them sprues also, which is where the metal's going to go. This is inverted upside down — you're eventually going to flip it over and you'll eventually pour the metal in this way.

But initially you just dip it into a very fine slurry of a ceramic powder. It literally looks like flour, but you mix it with some alcohol or something else. You dip it in. And I was looking for my newer brass rat this morning, that's where I saw some of my old wedding rings, wedding bands. Um, but I couldn't find my older brass rat. This is the big brass rat — I think I passed around once. This is lost wax cast, the — my class ring. Any of your class rings were also lost wax cast, or what we call today investment casting. Because you're going to take the plastic, or the — yeah, the wax part — and invest it in the layer of ceramic. And then you're going to put another layer of ceramic, coarser ceramic, on top of that, and make a porous shell.

So you complete your mold, it still has the wax inside. You stick it in an oven, you melt the wax out, you can recover the wax, reuse it if you want, make candles out of it, whatever you want to do. You then put this ceramic shell, which has cavities the shape of your part — you put it in a furnace and heat it up to bond the ceramic powders together. This might be a 2000° furnace, 2000F okay, because this is all ceramic now and you're going to bond this stuff together. But the first layer was this very fine flour. And the reason I wanted to bring my latest brass rat — you can actually take, and you can read on this one, I'll pass it around, you can read "Massachusetts Institute of Technology" across here. But even the one that fits on my hand, if you take a magnifier you can read the "Massachuset—" you know, all the letters of MIT going across there on my actual brass rat. If you use a fine enough flour okay, you can get very good detail in your part.

So then you take this out of the furnace, you actually put it in a can, pour some other sand around it, just loose sand to keep, give it some structural stability. Now you take your metal, pour it in, let it solidify. When you're all done you basically knock the ceramic off — it's just trash now — and there's your casting. This was your original wax pattern, and now this is your casting. It's lost wax, you lost the wax from the mold. And it's called investment casting.

And that's how we start to make these $6,000 turbine blades okay. [Tom holds up a single-crystal turbine blade, cut in section.] And they actually, when they do it, they actually have a little ceramic mold that has the shape of the inside of this thing, so that when they — how would you make something with an internal — this one's been cut in two okay, so you can see the inside. But when Pratt and Whitney would have sold this, it would have been a solid part like this. They make it by the investment casting technology, and they end up with a polycrystalline grain.

But then they take that polycrystalline structure that has the shape, while it still has its core inside — actually in the investment casting it's still got the casting around it — and they actually cool it very slowly in a temperature gradient furnace. So the furnace is about the size of this room, cost about a million or $3 million. You can make maybe 16 of these at a time. You see in this thing they're making multiple copies, they might make 16. Actually Rolls-Royce makes four at a time, Pratt and Whitney and GE make about 16 at a time. Um, Rolls-Royce likes to have better control in a smaller furnace than General Electric and the others okay. But it might take you six or seven hours to very slowly cool from the bottom of the chill. We don't want any shrinkage in here. We want, as I said in the chill block before, you actually going to pull it through a gradient furnace, so it's actually an induction heated furnace typically. And it may be 50° below the melting point here and 50° above. Over two or three inches you may have 100° Centigrade temperature gradient. And if you pull it through there very slowly, it will grow from the bottom and you'll push all your shrinkage ahead, and you'll end up with a perfect part if you're lucky.

Um, if you want — and this is actually — this one actually, the $6,000 blades are single crystals. They actually start out with a little wax pigtail on the original casting. And pigtail — it's actually a circle, it's a quarter screw spiral. And what happens is when you slowly start solidifying that, you actually will start with a seed crystal that has a particular crystallographic orientation. Because these things are going to be at high temperatures, and they have a greater high temperature strength with a particular crystallographic orientation. So they start with a single crystal they already made, they orient it by X-ray diffraction at the right orientation, and they start that and it goes around the pigtail. And if you get a nucleation of multiple grains, as they go curling around that pigtail three or four times, you end up selecting out a single grain. So when you get up here it grows as a single crystal okay. And you end up with a single crystal turbine blade which cost $6,000 a piece okay, after you've done all the metal finishing and everything else.

And you have to — the holes in there are either laser, electron beam drilled, under high intensity, or you can take a plunge EDM machine and you can plunge the holes in there and cut the holes. Because eventually, in order to get high efficiency in the engine, you want to put cooling air through there. The engine is operating typically today at something 3000° F and above. The alloy melts at 2400 Fahrenheit. If you didn't cool it, your engine would melt. But in fact it is air cooled. When we say cooling, you're actually using the compressor air from the turbine, which is about 1,000°. So you're cooling it with 1,000° air so you don't melt the blade, which melts at 2400 but can operate at about 2,000 in a nickel based superalloy.

Um, it's actually a fairly safe system in the sense that if you lose your compressed air, you've lost your turbine anyway. I mean inherently you can't run the turbine without compressed air. That's what the front end of the turbine is, if you know anything about turbine engines. Um, so there's always compressed air.

If we look at how that stuff fits together, we actually can make quite complex parts by investment casting. This is one where they didn't do that little seed crystal to make it single crystal, but they grew it unidirectionally and they ended up with columnar grains. So this one may not operate at quite as high a temperature, but we've oriented all the grain boundaries parallel to the stress direction, primary stress direction. This thing's a spinning rotor on a spinning rotor, and so the primary stress direction is radial. And the grain boundaries are the weak — the grain boundaries are the weak thing at high temperatures, weak part of the metal. And so we orient them in that direction, that columnar growth.

Well, if you went back to the 1940s and 1950s, and remember gas turbines really weren't developed until the 1940s, um, and you know the V2 rockets were some of the first, you know, commercial if you will, turbines. Um, and in the late 40s that's when jets — actually this, the Germans had some, um, experimental jets at the very end of World War II, jet aircraft. Um, but after World War II we started building jets, and those jet engines, it was just polycrystalline cast turbine blades, or machined out of wrought material. If you go up here to General Electric in Lynn where they build these turbines for the military, now uh you'll see machine tools just taking solid chunks of steel and making compressor blades by just machining them from solid bars of material. When you get to the hot section and you have to use nickel based alloys, they don't machine so well, they're very difficult to machine. Um, and so they actually cast them. In the '50s they'd be equiaxed castings, lots of little crystals.

If you look at something that's just cast normally as an ingot, and you look at the crystals, you would get something where you get some nuclei on the outside edge where you first poured into the cold mold and it starts to solidify. Then you get columnar grains as the heat is flowing out to the walls. So you get columnar dendritic grains heading towards, in the direction of the walls. And then the final stuff might be equiaxed. Well, in the 1950s the standard turbine blade was just a casting like a lost wax casting, like my brass rat, multi-grain. But then by the 1960s people were learning to cool it directionally, like using a chill block okay, but cooling in a gradient furnace where you have a temperature differential and pulling it through slowly. By the late 60s people were learning to make single crystals. By the 70s we had that kind of technology, uh, that blade's probably about a 1980 technology blade.

Um, today the walls are getting down as thin as 15,000 of an inch okay. Really thin walls. Um, the problem with going thinner is, if you have a little bit of core shift — you've got a ceramic core in the middle of that thing when you're growing this thing — and you get a little core shift of 5 or 10,000s, and you got no wall left okay. So the way they inspect those right now is, they basically put them in a CAT scan machine, and for a half million dollars you can buy a CAT scan machine for metals. You put one of those blades after you cast it, you put it in there, you get a full three-dimensional image inside and outside, just like doing a CAT scan on your brain okay, or on your arm, only it's designed for metals. And they do 100% CAT scan, they measure the wall thickness on that CAT scan, and they probably scrap, hopefully they scrap less than 30% of the production okay, at that point.

That casting, the single crystal, may be worth, um — the actual value added of the invested casting and molding might be $11,000. The alloy may be a couple of thousand, because nowadays we have up to 5 or 6% rhenium in those alloys, and if you look at the periodic table, rhenium is one of the Platinum Group Metals okay. I remember when I took my creep course from Professor Grant back in the 1970s, um, he came in and said, if you're going to make a jet engine — he'd been working on these alloys since the 1940s — he says, what's the best material to make it out of, has good high temperature strength, uh, can operate higher than any of the nickel base alloys or the cobalt base alloys? And we all sat there and we couldn't think about it. He says, platinum okay. Platinum goes to 1700° Centigrade as a melting temperature. You could operate platinum at probably 3100, 3200°. If you want to make a platinum engine, you could get another, you know, one-third improvement in your fuel efficiency of your aircraft by operating things hotter. But platinum tends to be a little dense, and so the stresses are up. It also happens to be a little pricey okay.

And if you wanted to make a really fancy engine, you could use iridium. The problem with iridium, they only mine about 100 ounces a year in the world. So you would make a lot of engines out of iridium. But iridium doesn't melt until over 3000° Centigrade. You could build a 4,000° engine out of iridium. No one's ever done it. Well, actually we have built iridium engines, um, I take that back. For these deep space flights of NASA, when they're going out to Pluto or Saturn or Uranus, anybody know how they get their energy?

Student: Solar?

p35b Now solar, you don't have enough energy density. The sun, it goes off as r squared from the sun, and it's a long way out there. You — I told you about the thermoelectric things, right? Well, what's the heat source for a thermoelectric flyby? Some of these things take 15 years to go out there, right? Well, they actually take plutonium. And plutonium's got a, what, a 40,000 year half life or something okay. 30,000, long time, long enough to go to Pluto.

Student: 10 to the eighth?

p36b Well it depends on the isotope. Some of them are 10 to the eighth, but the worst part of it, usually it'll die down to a reasonable level of radiation after about 40,000 years as I remember. That may be a couple of half lives, I don't know, who knows. But I don't, it's longer than I need to wait okay. So they take a thermoelectric generator, just a semiconductor, and you have to heat one side and cool the other side. It's easy to cool things out there in far space okay. You just expose it to the, you know, have a radiator and let it radiate out to the cold. And you're pretty cold out there, you're down, you know, only a few tens of degrees above absolute zero. So you're cold on one side, the other side you have a plutonium heat source. Well that plutonium heat source is encased in an iridium shell. So they make an iridium sphere that'll go to 3,000° Centigrade, they put plutonium in it, they put it in a rocket and shoot it up, and they hope that the rocket doesn't fail and they spread plutonium all over the Atlantic Ocean. But they don't tell — that doesn't usually make the papers, that they're about to shoot plutonium, you know, nuclear warhead material up into space. Um, they don't do as many of those now, but it is actually a very big concern. About, you know, the environmentalists would not be pleased if they knew that you just dumped a couple of pounds of plutonium in the Atlantic.

Student: [How much plutonium?]

p37b Yeah, they would use a couple pounds of plutonium, I don't know exactly how much. But plutonium is pretty dense too, it's not a very big — you know, they, if it's nuclear warhead, they call them the pits okay. And I don't know, but I think pits are about, you know, I don't know exactly what critical size is, but they're not much bigger than a walnut I think okay. So they had built iridium engines, but it was plutonium was the fuel, iridium was just a shell around it. Uh, and iridium doesn't oxidize, it's Platinum [Group], right? Um, anyway.

So this is the ceramic core that goes inside one of these turbine blades that you're going to cast, and that gets etched out with acid afterwards. These are some compressor discs. Um, and here's actually the — this is actually the rheocast technology. This actually is the finished casting. These are actually some of Professor Flemings' drawings, they're in a book now. But I recognize this is a — the firing pin for a US Army, I think uh, M16 okay. The Army gave them millions of dollars back in the 1970s to see if they could injection mold a semisolid steel okay. And this is actually injection molding. You're pressing the semisolid material into the cavity, and the cavity actually is that firing pin. It didn't work, because you didn't have mold materials that could even take those temperatures. They could actually — they made some parts. Um, they could make 100 parts on one $100,000 mold okay, so wasn't exactly economical.

Um, but they do make all kinds of aluminum, zinc, and brass parts by injection molding. Um, and we make all kinds of complex parts, automotive parts, by investment casting. A lot of your, uh, fuel injectors and things like that, uh, complex parts. Oh here's another investment cast part. [Tom holds up an investment cast part.] Anybody recognize this? It's stainless steel, and they make another part that goes on the bottom, they weld that on, and then they take it up to a belt sander and they give it a nice finish. And you can read "Spalding" on it afterwards. Um, so that's another investment cast part.

Um, sand casting — which sand casting, you can make small parts. A lot of, I could have brought in a plumbing elbow, a cast iron plumbing elbow that's made by sand casting. Which basically — [Tom searches for the next slide.] brighten that up a little bit, again it's not brightening — anyway, there it's a little bit brighter. So you start out with two core boxes typically. These might be wooden molds, you can shape things, and you make them in two halves that are going to go together. And one's called the cope and one's called the drag. Drag is on the bottom, the cope's on the top okay. It's like book molds — in fact sometimes they call them book molds in permanent mold casting. When you're all done you're going to put these two things together, that gives you a cavity.

Um, or actually what you do is you get the core halves. Um, you might make a wooden part, you might make a metal part, that's basically from some of these molds in the core box. You basically then bury it in — you bury this thing in sand. But it's on two layers. You've got the thing, you got a piece of plywood, one half here and one half here, the cope and the drag. There it says cope and drag. And you'll make two half parts, um, and you'll make a cavity in the sand like this. I guess we're going this way — yeah, well actually I don't know what they're doing over there. You'll make a cavity in the sand, and then you'll put the two parts together, one on top and one on the bottom. You now have a cavity the shape of your core halves over here, and then you pour the metal in and you make your part okay. So that looks like it's some sort of plumbing part or something.

And we make things as cheap as a little cast iron elbow that you go to the hardware store and buy for 89 cents, right? If you buy it in volume you get it for 29 cents. Chinese are very — Chinese make a lot of these. And those little kilns, those little cupolas right? There's not a lot of people that make them in the United States anymore okay. Now we also make brass this way, we make aluminum this way, uh, we make all kinds of parts. Sand casting goes back thousands of years. They just poured sand into a hole in the ground.

And here, to show you a casting that was not much more — I just walked around the lab once about 20 years ago and found this. [Tom holds up a rough cast lead or lead-bismuth piece.] This is probably a lead or lead-bismuth alloy. You can see the crystals, but you can just see that someone just scooped out some cavity in some ceramic sand or whatever, and just poured it in. They're just trying to dump the leftover metal from a casting operation probably. And I just said, oh that's nice, it shows crystals and stuff.

Um, you also can make big cast iron parts that are the size of this room. If I've got a generator rotor forging — not generator rotor forging, we talked about generator rotor forgings yesterday — but if I have a, I have to have a housing to put that generator rotor in. Once I take that steel rotor and I wind it with copper, then I have to have a housing that holds the other windings to make my big generator. And that generator housing is about four or five inches thick cast iron, and it's the size of a small house. And there's a top and a bottom, and it's just a big sand casting, just bigger than the little nipples that you buy at the hardware store okay.

You also, if you go look at those nipples in the hardware store, you'll see a parting line on the actual final cast part. They don't show it here, but if this was the actual part you'd actually see a parting line. You can see where the two molds go together okay. So um, there's lots of different ways to cast things for shape.

Um, this actually is not a casting but it shows you large grains. This is actually part of a — originally it was part of a superconducting microwave cavity. Um, when they were trying to generate microwave frequencies for particle accelerators for the physicists, then we ended up using it as a susceptor in an induction furnace, and that's how it got these huge grains. The grains here are like a centimeter across, if you look on the sides of that. And then I think someone — looks like someone used it as an ashtray for a while okay. Uh anyway, I picked it up in the lab here once sometime.

Um, here's some zinc ingot. Just like pig iron, this is zinc. When it comes out, there's actually a zinc blast furnace when they want to make zinc, just like making iron ore, not quite as big in size. It may only be two or three stories tall, rather than 30 stories tall, because we don't make as much zinc. But they basically — it comes out just like the pig in the sand bar, remember, just cast it in the sand. Although in this case they probably cast it in a metal mold, so that you actually would say — I don't know if this says "New Jersey Zinc" — if it was the whole thing. I just got this because, you look at it, someone just broke it with a hammer and you can see the columnar grains from the metal mold okay. So the way you extract the heat will determine your structure of your steel, or your metal, whatever type of metal it is.

Now let me just finish with a couple of specialized casting processes if I can find them. [Tom flips through slides looking for specialized casting examples; can't locate them.] Well why don't we take our break earlier and I'll find them at our break and show them to you. Oh wait a second — yeah, I'll find them at the break. Okay why don't we come back in about 10 minutes.

[BREAK]

[Resuming after break.] Just lining it as long as all these tables up there, and they came in one morning and they were all gone. They had been borrowed on the no-return plan, right. So uh, copper is worth a lot of money. We actually had that when they were redoing this classroom okay, about five or six years ago. Uh, you know that there's copper pipes and things in the heating system and cooling system and stuff. And uh, I came in one morning and, you know, I get in early, and uh someone was talking about they had lost $8,000 worth of copper pipe, which actually nowadays is not that much okay, cuz copper prices have gone way up. And they said they didn't know who did it. I said, well you know I got a surveillance camera in the hall. And there is — this if you look at the fire door down there just down from my office, I put in a surveillance camera that would watch the door to my office and the door of the lab. And I said, we never look at it unless we had a reason. But we had a reason, they had lost all the copper for this. So I asked Don Galler, the tech, works for me, that took care of the camera — he's an electrical engineer, so he put in the surveillance camera to begin with. I said go look at it. And there we had a picture of one of the contractors who had brought the stuff in, when he was in off hours he came back to take it home okay. We had a picture, beautiful picture of him. So they filed suit against him, he lost his job.

Um so anyway, okay getting back to casting technology. So we talked about — as much as I want to talk about extracting the heat of fusion, I do want to say something about shrinkage. One of the advantages of continuous casting is you don't have to cut off a riser. All the shrinkage just gets pushed to one end. And if you make something that's in a big steel mill, if you're making that casting for three months — and you made how many miles — you only have one hot top to throw away, or one laser top to throw away, rather than thousands, right. And so you go from like a 65% yield — and yield in a steel mill is, how much, how many pounds of steel you sell out of how many pounds you cast. So if you cast 100 pounds and you sell 65, you got a 65% yield. And continuous casting took them from the 65% number up to over 90% within a few years. And as they learned to make alloy grades and other things, they actually are up around 97% continuous cast in the steel mill now.

Um there's two continuous casters right down here in Attleboro. I used to work with this company that makes — uses more gold than any country, company in the country. And they have two continuous casters. Well in the case of gold, you want to have as little scrap as possible, so it pays to have a continuous caster, even though you may only cast something that's 12 feet long. It might be a bar like this of karat gold, and it may only be six feet long. But if you only have to cut off one little 2-inch hot top as opposed to 20 or 30, it paid for itself. Even though a little caster like that might be 5 or 10 million bucks. Um, so far as that goes. And that's how this copper was cast, so far as that goes.

Now there are a couple of — and I talked about continuous casting is a little bit of a problem because it's almost too productive, unless you got a really high value added or high volume, high value added in gold carat gold. Now a lot of those casters are a lot busier than they used to be when I started working there 30 years ago, because now they have the contract from the US Mint to make the gold coin blanks. The US Mint is making some sort of gold coins or something now. I don't remember — Sacagawea — okay, the dollar. Is that a dollar or $20? I don't know, a dollar, but anyway. I guess it's plated with gold on one surface or something. Anyway whatever it is, they have the contract to make all the blanks that are sent to the US Mint for stamping to make the coins. And so they're nice and busy now. And of course when you're making that much of something, you know, you can use it. But they were using it even when they only made small amounts for jewelry.

Although we did make a 36-inch diameter coin, 3/4 inch thick gold coin, when I was working there in the 80s for the Vancouver World's Fair. They had a world's fair in Vancouver, British Columbia, and we cast 3/4 inch thick, 6-inch wide gold, 14 karat gold bars, and then I had to weld them together okay, without distortion, and make a 36-inch diameter coin. So they had this coin on display, which is like $4 or $5 million worth of gold. Of course they recovered it afterwards, they didn't, you know, but it was just for a display, for the world's fair for a few months.

Um, but there are some — there are a few applications of making copper, I talked about this one. But there is Southwire is a firm in, I think they're out of Georgia, uh, somewhere. Anyway, south of Atlanta somewhere. And the Southwire process is — and that's what I was trying to find, and I still couldn't find it, I must have left it at my office. Um, there was a picture of it. But basically the Southwire process is about a 6 foot diameter copper wheel that has a groove cut in it okay. It's just, it's like a railroad wheel, if you will. A great big railroad wheel. And the edge of the wheel has a mold cap in it okay. The diameter is down here and this might be about 2 or 3 inches wide okay.

So this wheel is just going to go around in circles, it's water cooled so you can extract the heat of fusion. And they have a steel band, so if I'm looking sideways on this 6 foot diameter wheel, they actually have a steel band that goes around a roller that forms the edge of the mold. So you have a steel band right here, and this is your cavity. And they just pour copper in here. The thing goes around slowly, it solidifies, extracts the heat of fusion, and at the other end you just chop off these big S bars of copper. And Southwire makes copper wire okay, for overhead transmission lines and all kinds of things.

There's not a lot of those in the world because we don't need but so much copper wire. And ever since fiber optics we need a lot less copper wire okay. In fact back in the 70s you could plot the price of a penny. And it was getting to the point — well, some of you are too young to remember — but um, back they used to make pennies out of copper okay, they were copper pennies. And what happened is they were plotting the price of copper, and they knew that sometime in the early 80s it was going to cost more than a penny. There's going to be more than a penny's worth of copper in a copper penny. And they decided that would not be a good thing, cuz people would just be melting down the copper pennies to get copper. So that's when they developed the current penny.

Anybody know what the current penny is? It's zinc with a copper plating okay. Which I always thought, this could have a big corrosion problems, but nonetheless uh, they did work that out. Um, anyway, uh, because they wanted something cheaper, and zinc cost about 1/10th the amount of copper. Nowadays they're getting to the point where the zinc penny is getting to the point where it's a little bit more expensive to produce, and it actually is costing the government more than a penny to produce a penny. So they'd like to get rid of pennies, but anyway.

But um, in any case, this is the Southwire process, 6 foot diameter wheel turning out kind of 2 and 1/2 inch diameter — it's actually not a circle, it's this you know half moon shape or whatever. Um, but you're going to go through and you're going to roll it into a rod and then you're going to draw it into wire. Um, it'll produce a lot of copper for the world.

Now there's one other application for copper in continuous casting. If you go up here to Burlington Vermont, you'll find the Hazelett company. And the Hazelett company basically takes a big wide steel band on rollers, and they water cool it with a water spray on the backside. And they have another steel band coming in like this, and they pour molten copper on here. And they cast — about sheets of copper come out of this thing, um, this water cooled on both sides, about 2 inches thick and about 3 feet wide. The steel band's about an eighth of an inch thick, water spray on one side, to extract the heat of fusion. And you're basically making copper 2 inches thick.

Now, why do we do these for copper? Copper has a very high thermal conductivity. It's easier to extract the heat of fusion from something with high thermal conductivity. You can do it in thicker sections, and in this case they can use thin steel bands. Now it's a real mess if you break one of these bands, and you now have molten copper in contact with water and you generate steam and explosions and other things. Uh, so you have to know how to do it.

What are you using all this copper sheets for? Well it turns out every copper mine essentially, in making the copper, if they want to make really good copper, they make a copper that's got a lot of impurities in it, and then they want to go through an electroplating process. And so these are copper anodes. To make really high quality copper, they essentially take big sheets of copper made by the Hazelett process, which has got all kinds of oxide and impurity and other stuff in it, and they electroplate it. And they make electrolytic, what they call electrolytic copper, sometimes called ETP copper, electrolytic tough pitch. We don't want to get into the tough pitch, but it has to do with how you deoxidize.

If you go through a copper plant — actually there's one in Reading Pennsylvania that I went through once, I had a student do a thesis down there. Um, they actually use green tree trunks, the lots of sap in them, and the way you get the oxygen out of the copper that you've melted is you throw a big tree trunk into the bath of copper. And it creates a boil as all the hydrocarbons from the tree trunk are boiling off, or burning off. And it's called poling the bath, because you're using a telephone pole to stick it in the bath okay. And you get this froth, and it's the hydrocarbons burning off there are deoxidizing the copper bath okay. And I think the pitch, the electrolytic tough pitch comes from the pitch rosins that are in the tree trunk. I'm not sure if that's true or not, but I seem to remember that from years ago. But some of this is sort of ancient technology okay. Um, so far as that goes.

Uh, any questions on that? Yeah.

Student: How long do the steel [bands] last? Do you replace it every so often?

p64b Yeah, every couple days, I mean you know they know approximately how long they last. They've done a lot of work, and it's actually a titanium bearing steel. And I mean they've done a lot of work on these things. I went up to the Hazelett plant once. Can't remember why they invited me up, and so they gave me a nice little tour. People have been trying — actually that's probably why they brought me up. The Hazelett folks have been trying to make thin steel strip using their process. But there is no — you can't put steel on steel. And the problem is, steel doesn't have the thermal conductivity of copper, and so it stays hot longer, and so you end up wearing out your bands. Um, and you can't run a little laboratory scale one of these things, you actually have to run one that's about 3 feet wide okay. So it's sort of expensive to do.

Uh, people have been trying to make continuously cast steel strip for 50 years, and no one's really been able to do it. They still make it as a 10-inch continuously cast plate, and they roll that down. But if you — there's a lot of steel sheet used in the world, and if you could start it out as quarter inch thick strip to begin with, it'd be worth a fortune. Billions of dollars if you could do it. But lots of people have tried, but no one's been successful. Okay because it's hard. I mean you're only working with something that melts at 2,000° Fahrenheit. You get to steel, you're working with something that melts at 2600° Fahrenheit, and there's a big difference. That last 600° makes things a lot harder, particularly when you're just using a steel band and there is nothing else. I mean the nickel base alloys, that turbine blade melts at a lower temperature than steel. So use an iridium band — yeah, and there's 100 ounces of iridium in the world each year, and it goes for about $1,000 an ounce okay. So yeah, you could use an iridium band. Or a platinum band — a platinum band okay, and it goes for about $1,000 an ounce, and there's a lot more platinum, right. You use a rhodium band okay, goes for about $3,000 an ounce okay. Yes there are some things, yes okay, you're right, you probably can, but no one's done it yet okay.

Uh, so people now — so there's a limit to continuous casting and part of that problem is, actually it's too productive okay. Now let's talk a little bit about structure in casting.

Um, the heat transfer across the interface — this is sort of an interesting plot in terms of distance. This would be steel, solid steel, this is liquid iron carbon alloy. We have the composition of the liquid, and we actually have a big drop as we go across here. The liquid has this carbon concentration. When you get to the interface, you go from the composition of carbon in the liquid to composition of carbon in the solid. It's sort of like the Stefan problem — you got a mechanic, you got a, in this case not a temperature discontinuity but a, uh, a concentration discontinuity. And then it drops down into here. But what you start out solidifying is not the composition of what you end up solidifying. Your liquid and your solid don't have the same concentration. And that's a function of the phase diagram.

How many of you have never seen a phase diagram before? You've all seen the phase diagram before, somewhere in some of your Navy training or something okay. So a phase diagram is nothing more than a plot of temperature versus, in this case carbon. This is the steel diagram. At high temperatures you got liquid up here. This is your cast iron alloys over here, these are your cast steels up here. This is called a peritectic reaction, we don't have to get into the details of that. But what it shows you is the composition of the liquid is richer in carbon than the composition of solid. This is a two-phase region, this is the gamma phase. And that's generally true for any alloy. Pure copper, you don't have to worry about it. Pure aluminum, you don't have to worry about it. But any alloy that's a binary alloy, you generally have a liquid that is richer in one phase, or one component, than the solid okay. And that means that you get what we call segregation.

So if I go back and look at my challenges and the issues, we've talked about shrinkage and how we can do continuous casting or use risers or chill blocks. We've talked about extracting the heat of fusion, and if you got really big things you have to wait hours or days. We get down to very thin things and we'll talk about, in rapid solidification, where we melt things, we freeze things at a million degrees per second, we get really thin, like paper thin literally. But in segregation, what we have when we have a two-phase alloy, we actually get the liquid that will separate to a higher concentration than the solid, because of this type of phase diagram.

And if I look, uh — oh, I can probably do it with this type of phase diagram here. [Tom points to a binary phase diagram on the slide.] Uh, this is the liquid, this is a solid, this is another solid. I will always have a liquid that's higher in concentration than the solid. This is the liquidus line, this is the solidus line. And what happens is that, cools down from some liquid temperature down through lower and lower temperatures. I will have a solid that forms like this, but I will get coring. And it doesn't, it very well, but you can see some kind of different colors in here. The liquid that forms will actually get a core around a rich alpha phase.

And I guess what I can show you is, if you look at a real structure, this is one from one of my students' doctoral thesis on stainless steel where he used a special etchant. And you actually can see the concentration gradient of iron, nickel, and chrome okay. The core that's fairly white is the first FCC non-magnetic stainless steel to form. And it actually was sort of round in cells. But as it transformed, as cooling down, it actually rejected some of the chrome and nickel, and with this particular etch it would etch a different color. Actually originally — this is in black and white now, but originally it's a color etchant. You see all kinds of rainbows of colors. But you can see in this type of picture the coring on a microscopic scale. It's not uniform in concentration. So one of our goals was to get something that was homogeneous, and to control the structure. But depending on how we extract the heat of fusion we can get very different structures. I'll show you some of those. And we can get segregation within the casting.

Now that casting can be on a micro scale like we see here, or it can be on a macro scale, on the size of the whole casting. To give you some idea of the different types of structures that you can get, here are a number of different stainless steel microstructures that John — he was looking at — number six different stainless steel compositions. But you can get all kinds of different types of structures. This is one that forms in the solid state. This is a structure from what we call dendrites. These are also some of the remains of dendrites. But you get all kinds of cast structures with different length scales. Some of them very fine, we're going to talk about that, some of them very coarse.

Instead of just dendritic structures — this is a similar dendritic structure, dendrite comes from the Greek word that means trees, and this is sort of, if you turn it sideways, a tree with branches. And you can also get cellular structure where you have cells, and the black outlined, these are individual fingers of material that are growing, from as the heat's extracted.

If you look at it in a schematic, you actually will form two different phases in some alloys, and you can have cells grow where you have the alpha phase and the beta phase. Or you can, if you want to talk stainless steel, that could be cells of face centered cubic and body centered cubic crystal structure iron. One would be magnetic, one would be non-magnetic. They can have very different properties. In fact that's the definition that J. Willard Gibbs gave to a phase, is uniform throughout in properties — not necessarily in composition, but in properties. So a phase hopefully is homogeneous in properties. But when you have two phases, it's by definition not homogeneous in properties, because each phase has its own set of properties. One of them might melt at a higher temperature.

If I get John's thesis out again, I'll show you where the ferrite has a higher melting point than the austenite, and it actually gets fingers of ferrite sticking out into the melt as you're solidifying. This is where you actually get a breakdown and instability called dendrites, and you can see the little branches on the trees. This is cellular solidification. This is a type of cellular solidification that we sometimes called the thicket. It's a very fine cellular structure. So you can get — there are more cast structures, in all these aluminum, copper, iron, nickel, titanium alloys, than there really are wrought structures so far as that goes.

Actually here's a good picture in aluminum — or, no this is lead-tin. In fact this is probably a Flemings picture. I think this was on the cover of his book originally okay. Um, but these are, uh, 20% — 10 — 20% directionally solidified. These are probably 10 dendrites growing into a lead matrix, or vice versa. I guess it's got to be one or the other. But you can see the dendrites melting at a higher temperature. You got liquid, the darker liquid here, and this is the all solidified stuff. And you get these dendrites form in different ways.

When you have an ingot casting, you actually will get the same initial nuclei forming on the outside when you first pour it in the mold. That's fine grain. Then you'll get columnar grains, and then you may have dendrites form in the center. And through all this you get fluid flow in the liquid as it's cooling down. Remember in a big casting this could take hours or days to cool down. And the liquid is getting enriched in one of those other elements. So the first stuff that solidifies is very rich — in a ingot cast steel, a rim steel, this would be a very pure iron alloy on the surface. So we could get the best sheet properties from ingot cast steel okay, because the surface was very pure iron. The internal composition might be slightly different. You're going to roll it into sheet, you don't really care. But in plate you didn't want that. So we have to worry about how we solidify these things.

But as the fluid flow goes on, here it's due to differences in buoyancy. And part of that buoyancy is the composition differences. I got molybdenum which is very heavy compared to steel. And in the iron-moly alloy I can get tremendous segregation because the differences in densities of the different elements. And that's what we call macro segregation.

If I look — should have some pictures of some macro — here we go, macro segregation. This is an ingot showing macrosegregation. In this case they show it as pluses and minuses. Minus, my alloying element is deficient down here, it's enriched up in here. This is called A segregation, the ones on the outside. The ones in the center are called V segregation. And it's basically just due to fluid flow. But fluid flow among the dendrites — this is fluid flow between the solid limbs of the casting alloy. So it gets sort of messy okay, to analyze. But people have measured it, um, and they probably still are analyzing it.

But um, so actually this one actually shows you some of the terminology. So you have your riser up here, positive segregation in the riser, branched columnar dendritic zone, V segregation in the center, and in the equiaxed grain zone, negative zone that we saw before. We have a chill zone down here, and a columnar zone. So you really have some sort of composite from the original casting. And that's a problem, um, in a lot of cases. You want to break that down.

And so you may later forge the product to get rid of it. If you're lucky, if it's iron or titanium, those alloys will actually go through a crystallographic phase change at intermediate temperatures. Iron goes from face centered cubic to body centered cubic at around 900° Centigrade. Titanium goes from BCC to HCP at a similar type of temperature. And you can cycle the material in the solid state and break down some of this stuff. But it turns out, if you want to diffuse away the compositional gradients from a big ingot, you can wait a couple of thousand years in a hot furnace. The rate of diffusion of the elements over long distances is so great that macro segregation cannot be annealed out. Micro segregation can be annealed out, but macro cannot be annealed out.

So what do we — we take something that's got large grains and lots of segregation, and we work it, we forge it, we hammer it down to a smaller shape. You don't use, you know, cast plate. If we could use cast plate we could save a lot of money. I mean people make pressure vessels out of 4-inch thick plate. If you could cast it as 4-inch thick plate and get uniform properties, we would, but we can't. If you want 4-inch thick plate you better start with 16-inch thick castings, and get 75% reduction to break up this macro segregation and homogenize it. It's not all that different than kneading dough. It's just, we don't kind of fold it over on itself. But actually the old blacksmiths did fold it over on itself okay, when they had the sponge iron. They basically would literally fold it over on itself, just sort of like kneading dough, over and over. Now you oxidize away a lot of your stuff. But nowadays we put it in a big hammer press and we forge a big ingot into a small round.

And so if you start out with a 40-inch diameter ingot you can forge it into a 12-inch bar okay. And you can machine things out of that bar. The bar won't have very much macro segregation. But in fact a lot of the specs will require you to slice, take a baloney slice off the end of that ingot, that 12-inch ingot, and etch it with a special — they'll tell you what type of etchant to look for segregation that still remains from the original casting process. Now this is not so common if you're making railroad wheels, but if you're making aircraft engine parts it's very common okay. The specifications going to tell you to do a macro etch test, slice a piece off, make sure that you don't have a lot of segregation okay. So you have to get it homogeneous.

Student: So this is a major issue in those generator rotors right?

p84b Yeah, those things could be 60, 70 inches diameter as the ingot, and they'll forge them down and they may end up at 30 or 40 inches diameter. But the area reduction is still about — if you go down twice in the diameter you get 75% in the area. And in general you need about 75%. You'd love to be able to get 90% okay.

This particular ingot, which actually has a little bit, very small amount of silver, like a tenth of a percent — this I was brought in because, um — did I tell you this was going to be the trolley wire for the Amtrak New Haven to Boston extension okay. The overhead line. And the company that was making this wire, the largest wire diameter wire they had ever made was 5/16 of an inch diameter, and they had made it from this inch and 3/8. Now this starts out as very large grain. This is not segregation here but it's the grain size.

And if you look at properties of metals in general — this is steels, this is aluminum, this is copper, this is virtually everything. You can add alloying elements and everything as much as you like. I don't know if what I'm going to grab — draw here, but you have what we call, um, strength — well, strength and toughness are the two properties that we're interested in. And strength is the force of fracture okay. I'm just going to measure how much force it takes to break something. And so we talk about the yield strength of the material, and it has to meet some minimum value. Toughness is the energy of fracture. Glass is actually fairly strong. It's actually stronger than steel. That's why fiberglass composites can be reasonable strength. The fibers are stronger than steel. But they're not very tough. They shatter, they're brittle, there's not a lot of energy.

And in fact actually what I can plot here is stress versus strain. And you go up, you yield, this is sigma-y for yield, and it may keep on going up to some sigma-U for ultimate tensile strength, that's the maximum stress. And then it'll break. Well the area under this curve is the energy that it took. Force through a distance if you will, stress times strain is the work necessary to deform this bar of metal or this piece of rubber or this piece of glass. Well glass has a stress-strain curve that looks like this. There's no area under the curve, no energy of fracture in glass. It's a brittle material, it doesn't stretch. Rubber on the other hand doesn't have much strength, but boy it stretches. And there's a lot of energy under that stress strain curve for a rubber band. You can shoot it across the room. You try shooting a piece of glass across the room okay, can't store a lot of energy in it.

So the one thing in metals that will give you both strength and toughness is fine grain size okay. You want both. The Liberty ships taught us you want toughness. We knew about strength in 1880. That's when they first started doing tensile tests, and that was the specification until the 1950s. After the Liberty ship problem, they basically realized — research at Naval Research Lab, British Welding Institute, and MIT were actually the three places in the world that really did the study of the brittle fracture of steels which was the Liberty ship problem — and found you needed to include toughness, the area under the curve. You wanted ductility.

Last week I was down at the Army where they're making armor and stuff, and they say their three metrics are strength, toughness, and ductility. Well, this strain is ductility, strength — well, what's toughness? It's just the area under the curve. So what I wrote up in my report, it's nice that they have these three metrics but only two of them are independent okay. If I got an area here, the area is just equal to the length times the width, right. So you only need two of these to define your strength and your toughness, which is the two things you want. But they like to measure all three okay. But I'm not sure they thought it through enough to realize that only two of the three are independent. But nonetheless.

Um, so what happened here is, they had the large grains. To break down this grain structure in a very pure material like this copper, they were supposed to go from this cross-section down to about 5/16, so that would be about an 80% reduction in order to get the strength and the fine grain size and this ductility. They really needed about 90% reduction, but their continuous caster couldn't make anything any larger than this. They had never made wire — this was twice the diameter, four times the cross-sectional area of any wire they had ever made. And I had two days to decide whether to accept $6 million worth of this product, to reject it. And I rejected it. And then they sued the government.

And when they sued the government, you don't sue the government unless you're squeaky clean okay. It turns out the contract — cuz Amtrak's owned by the government right? So the contract for this stuff basically said the government put certain specifications on the strength and the ductility of this trolley wire. This is the trolley wire right here okay. It was going to start at this diameter, it's going to end at that diameter. It didn't have a 90% reduction in area, it only had about an 80%. And their wire didn't have enough ductility. The real problem is, because it had still had large grains, it hadn't been worked enough to get to small grains. It actually wasn't very smooth on the surface, it had little waves. But when you're going 100 mph, little waves become big bumps okay. And so when they tried to run the train down the track, it was like the 4th of July with the sparks coming off as it bounced along the conductor okay. And of course if you had those arcs, this thing would have worn out within months okay.

So anyway, I rejected it. They counter sued, because they had a fixed price contract and they could have made an extra few million dollars if they had been able to buy this stuff from this one small firm. And they had justified going to that firm cuz they said, we tried to source everywhere in the world and no one else would bid on this specification. Which wasn't true. They were lying. Phelps Dodge, the second largest copper company in the world, headquartered in New Jersey, had been begging this company, this British company who was building the extension, had been begging them to let them put in a bid. Because they had a new process that had what we call redundant deformation. And maybe when we get to deformation — I don't know if we'll get there, but basically it really needed the metal, it really worked it over okay. In fact this is one of the ways the Army's looking at some new magnesium armors now, the same Conform process that was developed in one of the British labs, where you really get a lot of work into the metal, um, and get very fine grains.

They actually had made something that was too ductile. That's what the company said — well, your material has too much ductility, too much toughness, it's too good for us so we won't let you bid because you don't come all within our spec. This is what they said. And then they turned around and told the government that this other company was a sole source and they had to use their material. And then Tom Eagar came around and rejected it. And so it turns out once they rejected it, it's going to slow up a $300 million project. But all of a sudden now the government said, well why don't you go to Phelps Dodge because we understand they were interested.

Well it turns out when this thing went to the lawsuit, the company came in with their technical argument why Tom Eagar was wrong okay, on the properties of the material. The government came in and said, you just defrauded the federal government. They yanked the passports of the British managers, took the computers, and basically turned it into a criminal trial okay. So you don't fool around with people who can bring in — actually it wasn't the F — it might have been the FBI that came in, but anyway, it's a long story but the story is a little bit longer than that, but that's basically what happened.

And when they saw the legal response from Amtrak they all of a sudden, you know, they realized they had committed a sin and they were in trouble. So, but anyway, it turns out the wire that went up was the Phelps Dodge material, and it's great material. Great ductility, great strength, great toughness okay. It's got all the properties we wanted and more okay. But that was why it was defective according to the British. Of course the real thing the British was looking at, they — it was $2 million cheaper okay on a $6 million order. And so they decided that would be great material. This junk would be great material.

Except that was another thing that did — Amtrak sent them in order, said do not put any of your junk material up. And they did. Right down here in Providence they put up about 6 miles of track with this wire. And that's the way we knew that it would spark, cuz they ran a train down it, was just a 4th of July all over the place. Um, but uh, that was their big mistake, because it wasn't clear that I could back up with data my assessment that this material wouldn't be, is good enough okay. But once they ran it down the track and you saw it sparking, we all knew it wasn't good enough. It became obvious it wasn't good enough. But until they ran the test that they were told not to run okay, we didn't have the data to really back us up. But anyway. So sometimes I look out — okay any questions on any of that?

Okay I think that — I in fact I know right now that I'll be here Thursday. Um, I won't be here tomorrow, I got to go to Washington. But uh, I should be here Thursday. And actually Thursday we're going to go through some forging. I don't know that we're going to get to rolling or powder metallurgy. You'll just have to take this course in the fall or some other time, or just ignore it um for the future. But that will be — the 12 extra hours we've been — this is the fifth class and Thursday will be the sixth class, and we've been two hours of time. So um, I'll still be around. Maybe I'll come to a couple of the class, uh, and watch the videos or something, see if I still believe what I said back then okay. So I'll see you Thursday.