AM_F2019_04

It was a titanium nitride inclusion stuck, it was about half an inch in diameter, from the melting of these titanium alloys. And they triple melt them okay to get rid of the defects, okay, but so the defects still control the properties of that thing. But the defects are big macroscopic defects in that case because we're talking about a metal with reasonable ductility, okay.

If you're talking about a ceramic, I mean I had some guys in the office from a company last week and they're trying to make better ball bearings for jet engines. They want to make them out of [out of] silicon nitride because it's one of the toughest ceramics. With the size flaws they've got to be looking for, five microns in size, because five microns will be a critical flaw size in fracture mechanics for a ceramic. For a metal, maybe half an inch, two inches is a critical flaw size. You can have some pretty crappy welds in a metal and it still is not going to fail. Although if you have a crappy steel and crappy welds, it will. That's helped put my kids through college, okay.

Questions? Other questions? Okay, no other questions, okay. So we're supposed to be talking about additive manufacturing, and up here we have the Digital Alloys website, and I need to do a little full disclosure here. But before I do that, I don't want everybody to think that I think that there's never any great materials innovations. There have been some huge innovations over my career.

One's optical fibers. Corning worked on something that everyone said was impossible. They basically, you know how they make optical fibers, anybody know? They draw it. But what they do is, it's actually sort of a cleanroom environment of making glass, and they start with a very pure, might be quartz fiber that they've drawn down, and then they flame vapor deposit layers on this thing as it's rotating. And then we end up making something that's this big around and three feet long, maybe longer, I don't know, but it could be four or five inches in diameter and three feet long. And as they do it, they can functionally grade the composition of the glass.

Now when they heat that up, because glass has an interesting property, it's very close to being a Newtonian fluid. A Newtonian, being a Newtonian fluid — if I knew we were going to talk about this I would have brought my play-doh, a plate of something. Play-doh is not quite a Newtonian fluid, but if you've ever pulled play-doh slowly you can pull long stretchy things and it won't fracture on you. Well, the right temperature in the right glass compositions, you can pull this big billet of glass which you may have spent a million dollars producing, okay, because it's a very slow process. It's an additive manufacturing process, by the way, okay. You start out with something small and you keep on adding different compositions, layers, as you go along, and then you draw it down into a fiber it's ten thousand miles long or whatever, okay. And I think because it has different refractive indices as you go through the diameter, the light will go down rather than out. The light is trapped to go down a tube or a tunnel of easily refractive center glass, okay. That's an additive manufacturing process. We've been doing that for thirty-five or forty years.

And it's the first, I remember in the late 1970s copper was starting to go through the roof, okay. And New York City didn't know what to do. They were not, they could not put more wires underground to carry telephone conversations or anything else. Copper just took up too much space. And because Corning finally perfected the optical fiber, now you could have ten million telephone calls on one wire, okay. And the copper industry, okay — they were talking about the problem in the nineteen-eight, nineteen-seventies that copper pennies were going to be worth more than one cent, and people were going to start melting down all the pennies and selling it as a copper product. And that's one of the reasons what our pennies worth now, or what are they made out now? Zinc with a copper plating, right. All right okay, it's not a thin copper plate, it's a fairly thick copper plate. And now inflation's come along and now we're running into the problem that zinc is getting too expensive, okay. And now they're saying do we really need pennies, okay we'll just do nickel exchange, right, because of inflation and everything.

So times change, but the optical fibers changed the entire world of telecommunications on earth. And now rather than having telephone poles all over making an ugly neighborhood, which is where I grew up, we always had telephone poles, now they just bury these things under the ground, okay. And MIT's still got a problem, they got too many, if you walk down in the basement and look at the optical fibers they got bundles, and they, I don't know how many channels they're carrying, but they still don't have enough room okay, but it's for a different reason. But the optical fibers were a huge success and changed the entire industry.

They could only do that because of this gallium arsenide stuff. And it turns out if you went back sixty years ago — actually, if you walk through the Delta terminal at Logan Airport, you'll hear John Kennedy, you know, here John Kennedy talking about we're going to send a man to the moon not because it's easy but because it's hard, okay, which is a nice way of phrasing it. But it points out that one of the things that came out of the space program — they don't say it's gallium arsenide, but they say LEDs. Well it turns out a NASA-funded, at Lincoln Lab initially and a few other places, studies of gallium arsenide. Anybody party the properties of semiconductors who said gallium arsenide would be a lot better than silicon, okay? Silicon came around in the 1950s. People who were studying all these semiconductors looked at the compound semiconductors and said boy, this 3-5 compound of gallia, gallium arsenide, knocks the socks off silicon, let's make it. Well no one knew how to make it.

There was a guy who'd been at the solid state division at Lincoln Lab, his name was Harry Gaydos [Gatos]. Anybody ever heard of Harry Gaydos [Gatos]? Well he's a professor in this department in mid-60s. He came, he had been a student in this department in like the late 50s, went in Lincoln Lab, headed up the gallium arsenide program where NASA was throwing money at it, and he's actually, he was known for many years as the father of gallium arsenide. I knew him when I was a student, I knew him well as a young faculty member, we battled over the Graduate Admissions Committee, he was chair of it, and anyway that's another story, nice guy. If you walk down the hall in building 13 you would hear him playing his flute. He was actually outstanding flute player. He had brass flutes, he had gold flutes, and he was one of only four or five people in the world who had a platinum flute, okay. And he used to come to me for advice on how to make flutes, because he knew I worked with the precious metals company down here in Attleboro that made tubing, precious metals tubing. So he would come to me and he would work with the artisans who were making these half-million-dollar flutes, and he would ask me how should we make this, how should we make that. I'd tell him how to braze them together or whatever.

But so he came to MIT and he worked on gallium arsenide. But at the time everybody thought it was an impossible task to solve the gallium arsenide production problem. They knew it had the right electrical properties, but they didn't know how to make it with the controlled defects, let's put it that way. There are a huge number of possibilities when you go from a pure silicon single crystal and intentionally dope it to gallium arsenide, where now you can have different layers of gallium and arsenic, you know, different clustering and everything else. So Harry became the father of gallium arsenide. And I remember as a student, in the Journal of Applied Physics half the journal was about gallium arsenide or gallium aluminum arsenide. Millions and millions of dollars going into it, no one could make it reliably enough. And then came, well thirty years ago the military, the gallium arsenide production after thirty years of research was getting better, and the military could make $500 chips out of gallium arsenide for phased array radar, and that was a big thing of how do you package this and the stuff, and I worked on some of that thirty years ago.

And then about a few years after that, Motorola came out with something called a cell phone. Anybody know how big the original cell phones were? They were about the size of a brick, okay. They didn't weigh quite as much as a brick but close to it. And they were flip phones okay, and they had terrible reception. My first flip phone cost me $3,000, okay, and it filled up my pocket. Okay, and if I walked across campus, half the way across campus it wouldn't work because the cell phone reception was lousy and the power was — anyway, gallium arsenide got better, cell phones got better, and somewhere along the way, and because of the frequency the gallium arsenide you can work at, you can use GPS, okay. And that's why GPS works on the phones, okay. You can't make a cell phone with silicon, okay, it doesn't work, you need the frequency for GPS and cell phones.

LEDs — what's the secret to an LED? A guy Sake [Nakamura] from, who is Japanese but he was, I think he was working on, I think he was working on for some company, but he made good quality gallium nitride. And gallium nitride, you could predict back in the 1950s it could produce blue light. And we already had red and green, we used to make stop lights out of GaAs LEDs back thirty years ago. But when he came up with gallium nitride, process of making gallium nitride to make blue light, you could get white light, and that has saved how many trillions of dollars in energy costs for lighting okay in the world. So that, you know, NASA will take credit for the gallium arsenide.

Neodymium-iron-boron magnets, okay. They have about three hundred times the strength of regular magnet. And I can pass this around, which is two neodymium-iron-boron magnets separated by a little piece of polymer or whatever, and they kind of like to stick together. [Tom passes magnets around the class.] This other one is a fairly large block of neodymium iron boron, cost me fifty bucks off Amazon. But I can put this in the drawer of my desk, and I've got like an inch and a quarter of solid teak desktop, and I can actually feel the magnetism. If I put a piece of metal, scissors, on the top of my desk, I can tell where the magnet is. And it's not that easy to pull these things apart.

Why is that important? Why are strong magnets important? Motors. When I was your age and I was rebuilding starter motors on cars, they were this — they were size of a football and they weighed about forty or fifty pounds, out of copper. Now what's the size of starter motor in a car? It's about the size of your fist, right? It's not the size of football, because you have, you know, you now have magnetic fields that are ten to twenty times stronger than we looked, you had before. And the power you get out of the motor goes as the square of the magnetic field, and blah. You never could have had power seats in cars, other than maybe you go forward and back, okay. Now you got twenty to twenty-one degrees of freedom in your car with how many, eight or ten motors in each seat? Why? Because the motors are these little things the size of a big golf ball, or a golf ball, a little big grape, okay, and they're powerful enough to move the seat, okay.

Who developed neodymium-iron-boron? Nope, General Motors. General — I think General Motors had the basic patent. And General Motors knew what they wanted to use it for in the 1980s. Sumitomo came along and there was a big fight with General Motors and stuff, but it is really General Motors. And General Motors was going to bring it to market but there wasn't enough neodymium in the world. General Motors thought about getting into the primary production in neodymia, but then when other people heard that General Motors was interested in neodymium, they kind of came to the marketplace and said we'll do it for you. General Motors didn't want to be a primary metal supplier, but they wanted neodymium-iron-boron magnets, okay.

Who was working on it before? And General Electric came up with neodymium iron. It turns out I went as a student in the early 70s — yeah really seven— it's probably about nineteen seventy, we went to a field trip, Professor Flemings took us, one of the classes, on a field trip to General Electric Research in Schenectady, New York. They weren't working on neodymium iron because General Electric hadn't — their general, General Motors hadn't invented it yet. They were working on samarium-cobalt. The problem with samarium-cobalt was samarium's expensive. The great thing about neodymium iron, there's plenty of neodymium in the world, just didn't have a lot of it in metallic form to make magnets, but iron was cheap and neodymium could be cheap. Samarium-cobalt was always going to be expensive and still is. It has certain advantages.

They don't, you — I have two all-electric vehicles, Chevy Bolts. They don't use neodymium iron in the car in the electric motors in my car. Anybody know what they use? Dysprosium iron. Why? Because neodymium iron starts to lose its great magnetic properties at 140 degrees Fahrenheit. Dysprosium iron will go to 180. And when someone's gunning that car they're liable to get it hot, particularly if you're in Arizona, and then all of a sudden you lose your power if you go above the transformation. Who else is interested in dysprosium iron? The US military. They have a number of applications such as nuclear reactors powering ships that would like, the attacked, like to have that extra forty degrees, the safety margin, for some of the applications. So there's all kinds of things that follow on from these different things.

What else? Met glass. Dave Hill talked about met glass. Here's a sheet of met glass made in his plant that he commissioned down in Conway, South Carolina. [Tom holds up a sheet of met glass.] It was basically on a big copper roll, and the physics of squirting a very thin layer of fluid against these copper rolls and solidifying this at about half a million degrees per second to make a metallic glass was able to produce this. And this was to go in those transit, electrical transformers. It wasn't what they did initially. The first application, because most of the stuff was off-spec, the first application — anybody ever remember going to a video store where you got videotapes, and they had — you ever remember going through, actually you still, certain high-value items like cell phones or something, put this little plastic thing on there, and if you walk through the metal detectors in the store it'll send off a beep? That glass, the only thing, the only material in the world with that low of magnetic susceptibility, you can't counterfeit a little piece of that strip. In one of those little plastic things, you take one of those plastic things apart after you get home, it's got a little piece of met glass at it. That's what kept Allied Signal's met glass division going for the first two or three years, because they couldn't make good production quality transformer sheet, and they sold little strips to deter shoplifters. And they still do the off-aged product.

Same type of story with a new material in the 1950s. It was called plastic, and it was called polyethylene. And they built a huge plant, Karemera [Carbide?] who built it, but one of the big chemical companies built a big plant to make polyethylene, had too much variability in properties, and they were about to close it. And someone came up with a product where they could sell all this off-grade polyethylene. It's called a hula hoop. And the hula hoop craze of the 1950s came about because this chemical company needed to get rid of all these polyethylene that was off-spec. They had to make little hoops for a year and a half until they could get their production processes under control to make uniform quality product for other applications. That was the first high-value-added plastic ever made, was polyethylene, and if it weren't for hula hoops it never would have made it to market okay. Yeah.

I could put lithium batteries up there, but in fact we had, we've had batteries for a hundred years. Lithium is just a smart small improvement, okay. To me it's not a groundbreaking new technology. They're just using lithium instead of magnesium or manganese, huh. It's better, but it's not a hundred times better. Well, okay, it's a question what you want to say is groundbreaking and stuff. These were sort of really disruptive technologies. Lithium — I mean first of all, the thing that made Sony Walkmans viable was neodymium-iron-boron magnets. You could make a videotape player or a tape recorder to play music, but you couldn't make it light enough and small enough and have enough battery life unless you had neodymium-iron-boron magnets. And that's what Sony came out with, which they got from Sumitomo the magnets and stuff. But that allows you to miniaturize things. Before that, a tape player was the size of a football. You want to carry a football around in your pocket while you're running, jogging? No, okay.

So anyway, yeah, you could put lithium batteries up there, I wouldn't necessarily be opposed. Other questions? So this is not all additive manufacturing, but actually optical fibers is additive manufacturing. So let's talk a little bit about additive manufacturing unless you have another question. This is the Digital Alloys website. This is where they are actually making these little titanium pieces for Boeing, that's what they're making here, okay.

Now, I haven't — I thought I'd told you the story about how I am a member of the Digital Alloys scientific advisory board, and when they called me originally I said no I'm not interested. And the guy said, he kept calling me and he said won't you at least come for a visit. I said no. And then I found out that Sal Barriga, who had been the IT guy who took care of — he wasn't one of my graduate students but he used to take care of the IT for my group and I used to pay him to do that — but I found out Sal was working for Digital Alloys. And so the guy called me up, I said okay out of a courtesy to Sal, one of my former students in a sense, I will listen to your spiel. So you're going to come to Burlington? I said no, I live in Belmont which is all of twenty minutes away. I said no, you've got to come to my house, okay. So I made him come to my house. We sat around the dining room table and he showed me a little cube, one centimeter cube, not all that different than that titanium piece. And he started explaining how he did it. I said well okay, I will go out there and I will see what you're doing. And I went out and I saw this manufacturing process which they called Joule printing. And it's a little different, and I could tell they had at least a factor of ten advantage in cost over most other people, and maybe they had a factor of a hundred advantage in cost. And so I thought okay, this one might not be a bunch of sales okay schlock. So I agreed to be on scientific advisory board.

So if you think about Digital Alloys, they've got Don Sadoway, Alec Slocum, and myself, and Dave Hardt on their scientific advisory board. And then the founders of Desktop Metals, which is a powder metallurgy process and there's a machine right down the hall — our Chris Schuh, Yet-Ming Chiang, Ely Sachs who is one of the founders with Mike Cima of 3D printing in the old days, and Craig Carter, okay. And we really don't talk to each other about it because it's a conflict of interest. I'm just telling you this story about, for full disclosure. I'm not saying that Desktop Metals is the be-all and end-all of additive manufacturing. The point I want to get across in these lectures is additive manufacturing is a very specific process and it has very specific limitations, okay. I didn't want to have anything to do with it until I thought, well they've got a process that I could probably, from my welding background, I could probably help them with. And so okay, I'll help them. But it's not because I think that they've got the great some great thing that's going to take over the world okay.

So if I now look at things, and I had suggested that when I saw that, Alex Huxstep [Huckstepp] who's their vice-president of marketing or something, some such thing — I had put together a series of blogs, it's now eleven blogs that he started. So that I decided okay, I was going to teach something on estimation this term, I decided okay, students are always asking me about additive manufacturing and other people, so why don't we just do that. So Alex Huckstepp put together these eleven blogs on the website, and you should go to them and look at them. But we're going to spend a few days this week going through some of them, and I'm going to tell you my story of them.

If I go through this, take some quotes right out of Alex's first part one, "The Business Value of Metal Additive Manufacturing" is his first blog, part one. This doesn't look like it's the updated version that I had done this morning, let me check. Well it's got some of the updates, this doesn't have others, okay, good matter. So he says — and you got to remember he's vice president of development, he wants to sell his product, but I think these are fair, he is fair in talking about both strengths and weaknesses, and most people don't talk about the weaknesses.

"Metal additive manufacturing also known as 3D printing promises to deliver a fourth industrial revolution." Whoo, how exciting, right? So this is sort of the sales pitch okay. Everybody's scrambling to figure out how to do this, lots of money going into it. "3D printing can deliver part geometries, hollow structures, lattices, that are impossible to produce with conventional manufacturing." True, okay. Absolutely true. "Most processes either cut material away or shape" above, "all those approaches have restrictive design rules which constrain the performance of the part." He's at least being honest, there are design rules and you've got to follow those rules. Other people say we're going to take over the world, all those subtractive manufacturing people, they're dead, okay. Give me a break. We're going to talk about why those types of statements are garbage. Anybody who comes in with that type of hyperbole, just ignore him.

This is one of his things, and he talks about — he's got a number of ways that he's gone through looking at, but they're not all from him, I think he references pretty well where he's taken things from a lot of places. Prototyping — that was the first application of 3D printing. I told you, prototyping to make artificial hips, okay, as a mold. You could do things very quickly, you couldn't make customized hips. Spare parts production cost — I've told you that the US Navy has this huge program with what they call seventeen admirals with forty stars, okay. You know a star is, you know, one star, off by, a rear admiral's got one star, a vice admiral's got two and a, right, a full admiral's got three stars. So if you've got a committee of seventeen admirals with forty stars, that's one step below the Chief of Naval Operations who actually probably has four stars anyway. And there's not a lot of four stars, okay. In fact there were five stars but those people were in World War Two and they're dead now, okay.

So, and so he goes through and talks about conformally cool tooling and, okay, where you can put cooling passages and things. General Electric has a facility that I've been invited to in Cincinnati, where they are, a huge facility, and supposedly they are now making four hundred different parts for jet engines. Well from what I told you the other day, why jet engines? Because we're talking two hundred to two hundred twenty thousand dollars a pound as the value of a pound saved. And if I can go put certain cooling passages or fuel passages in a nozzle that goes into a jet engine, and can save millions of dollars worth of fuel costs, I can probably afford to make something with additive manufacturing. If that was General Motors and it's going to go in a car, no way, because the value of a pound saved in the car is two dollars over the life of the vehicle. And even if it's in a better place, $200 a pound, you're going to see that there's some problems.

If you go to part two, he has some pictures of some of these conformally cool parts and things and you're supposed to go and read some of the — at least look at these things, they're not terribly long, you can go through most of them in ten or fifteen minutes, and there was only eleven. Production speed is part two okay. "A manufacturing technology is viable only if its throughput can keep up with the demand." Well duh, okay. I mean that's sort of an obvious statement, right. But we're going to look and see how well can additive manufacturing keep up with the demand. Well he's got another slide that I had never seen before about six months ago when I saw, when I found this blog site. This is Joule printing, he put of course what's first is Digital Alloys, and then electron beam — I've showed you parts that we made thirty years ago by electron beam, arc plasma, laser. A lot of these are the different techniques. And he's kind of, directed energy deposition, powder bed fusion, powder and binder. And here all the different companies okay. And GE actually has a business venture called GE Additive okay. Trumpf lasers — I mean Coherent lasers, they make a laser-based, Coherent lasers they make a laser deposition process.

And I think I told you that back in 1975 the Office of Naval Research gave a, I think it's a seventy-five thousand dollar contract for Pratt and Whitney to take their twenty-five kilowatt laser, which was like a half million dollar research laser, okay, to make a turbine disk by laser glazing they called it. They just took the powder and they poured the powder on, and then the laser came in and melted it to make a part about six inches in diameter, which they then spun and did a test to see if that super strength — was it carbon nanotubes types of strength, okay. Took them a month to make a six-inch diameter part on a machine that you could buy for about two million dollars back then. So how economical do you think? It wasn't economical at all and no one even considered that, it was an attempt, a demonstration to show feasibility.

But in fact, one of the problems you have if you're going to do a metal additive process by directed energy, and you're going to have some part, and you're going to put a thin layer on top of that by melting it with some beam, under some beam, you can say well we'll just speed up the process and go faster, just like the met glass process, right. Well the met glass process spins that stuff out at sixty miles an hour, okay. You can start figuring out how many cubic centimeters that is when you got something that's six or eight inches wide, it is a reasonable number, pounds per hour. But this process, if you try to go faster, you actually are defeating yourself. Does anyone know why? Anyone knowing your heat flow? Yeah. Now you can go to higher power, but if you go, it turns out, you have to take my welding course. But this process has to be at 10 to the 6 to 10 to the 8 watts per square centimeter as the incident heat flux. If you don't have a million watts per square centimeter, you're not going to get the high speed of melting. If you have more than a hundred million, you're going to drill a hole through here. It's called laser hole drilling, okay. You've vaporized the material and blow it away. At a million watts per square centimeter you melt thin layer, but at a hundred, above a hundred million you're just going to vaporize the material, you're putting it in too fast.

There are natural physical limitations, and some of those are controlled by the properties of material in Fourier's, this Fourier's, Fick's law that says the heat is minus K dT/dx, right. We know what K for metals is, it's fairly high okay, and you're going to have a temperature gradient. And this is the watts per square centimeter — Q is watts per square centimeter, I know that's going to be somewhere in this range of ten, a million to a hundred million. I know what K is, tell me, it'll tell me what dT/dx is. But in fact if you go to Fourier's second law, you find that Fourier's second law always has something that is alpha over the square root, or no it's x is the thickness of this layer, x over the square root of alpha t. And Einstein actually worked out a diffusion equation mathematically and showed this will always be the argument. If you have an exponential it's going to be e to the x over square root of alpha t. If you have a Bessel function, cylindrical or spherical geometry, it might be x-squared over alpha t, which is this quantity squared. This is a dimensionless number, and it comes out of the form of Fourier's second law. The differential equation, well you can go through it, parse it around, and Einstein showed you always get something of this form. It has to be a pure number inside an error function or an exponential or Bessel function. There's all kinds of ways to solve the specific geometry, but you're always going to have this.

This says that the faster this — alpha is a function of K over rho Cp, where rho is density, Cp is heat capacity. So these are all known quantities. This is time. If I want to go faster, x gets thinner, and it gets thinner faster the time gets longer. And eventually the faster you go the thinner this gets until you have no melting, you're just heating the surface, and you're going so fast you're not even melting it. And if you try to go higher power densities you start vaporizing the surface. So you have physical limitations, vaporization. So all of these processes that are directed energy deposition, arc plasma and laser, okay, laser, electron beam, all of these have inherent limitations. Joule printing doesn't actually melt, it just heats it until it's nice and soft, and it just, you know, by resistive heating. And I happen to know, because I've been studying heat transfer in terms of welding for thirty years, this has about a ten to a hundred fold advantage okay in terms of the heat in the process. I don't have to melt things, I don't have to put in the heat of fusion and then take it out, I just have to get it hot so it'll lay down. And you saw the picture of it, right. That's not actually melting, that's actually just getting so hot it sort of fuses together. Yep.

If you have joints — and when they first made these you did have little tubular porosity and it was weak. They spent about six months or a year before they finally found ways to get rid of that. So now they are a hundred percent dense okay, which is why Boeing's interested okay. I think they can make a hundred percent dense titanium bars. I'm interested in tensile bars. But I pointed out to you actually, I should have, fine, pass that's rough. [Tom passes a sample around.] I pointed out this MMPDS, you cannot apply something on a commercial aircraft unless you follow the material properties in this book put together by the Defense Department and the Federal Aviation Administration. And titanium 6 aluminum 4 vanadium, which that alloy is, goes from page 556 to the next alloy at 5-120. So I've got over sixty pages of fatigue curves, tensile curves, and your material must meet these specs or you can't put it in an aircraft.

So why are they making a hundred parts for Boeing? Tensile specimens. They're going to do fatigue — Boeing's going to do fatigue tests and tensile tests and impact tests, because Boeing has to certify to the Federal Aviation Administration before they put it on an airplane that it matches the properties that everybody else has measured for titanium 6 aluminum for the last eighty years. Where was titanium 6 aluminum developed? Anyone live in Watertown? Anyone been to the Watertown arsenal mall? No, the Arsenal Mall used to be the US Army's Watertown Arsenal when I was a student, and in 1945 that's where titanium 6 aluminum was, to the invented, developed. Titanium was a new material available after World War Two, and 6 aluminum 4 vanadium was the first commercial alloy. It's still the workhorse material alloy for titanium alloys.

Anyway, I don't care if you've got the best type of manufacturing process in the world. Unless you can make it uniformly enough that you can meet the specs, the statistical process control specs, in that manual, you can't sell it into the aerospace industry. Well guess what, these things are costing like twenty thousand dollars a pound, you're not going to sell it to the automotive guys, I gave you that lecture the other day, right. This has to go to the people — and you're going to have to meet those specs, you've got to make a hundred of these, at least a hundred of these, before you can even — you have to, you got to get enough statistics, and it says you have to do ten heats and ten parts, ten heats and ten samples in each lot to get this statistical process control data so the FAA believes that you can reproducibly produce a product okay to go into the aircraft.

Now are they going to make parts like this? No, of course not okay. They're going to do the same technology, and they're liable to make something that's sort of flat and has some complex shape if you look there on it. I don't know what Boeing's application is, but they might make some little bracket. I've actually seen parts that kind of look like this okay. That would be very expensive to make. You would throw away eighty percent of your material if you just machined it from a solid block of titanium and that costs money to do all that machining. And if you can start with a near net shape piece — do you want a surface finish like that? Of course not, you can't use that surface finish, they're going to have to machine that part. But that's what they do with all their parts now anyway in building a commercial aircraft, whether it's aluminum or titanium or something else. I'm sure they have an application where you can do this layer by layer technology, make a complex near net shape part, and save money okay.

So let me jump ahead, I was going to give, will do it the other, tomorrow. Let's look at this is a slide I put together this morning, cost ratios for additive manufacturing versus wrought or conventional manufacturing. Water we know has one gram per cc as a density, and that works out to one metric ton per cubic meter. These are supposed to have cubes in there, I didn't have time to. Titanium is 4.5 grams, that's just the density of titanium, 7.9 for steel, 4.5 metric tons per cubic meter, 7.9 metric tons per cubic meter. The cost per metric ton of wrought material in titanium is $200,000 a ton, it's a pricey material okay. Steel to make a metric ton, it's about four hundred dollars a regular ton but metric and a per cubic meter, to make a cubic meter — this is not per metric ton, this is per cubic meter, I guess I better change that, dollars per cubic meter wrought, two hundred thousand titanium, 3200 in steel okay. Steel has a big cost advantage over titanium, might be heavier but it's got a big cost advantage.

If I do additive manufacturing, if I go in now, Huxstep [Huckstepp] in I think it's part four he does the cost of additive manufacturing, we're going to get to that, but he kind of shows kind of an average number is five dollars per cubic centimeter. Five dollars per cubic centimeter works out to five million dollars per metric ton. The ratio between this price and this price is twenty-five to one for titanium. For steel it's fifteen hundred to one. So why are they looking at titanium? Well it turns out it's easier to make by their process okay, that's the real answer. They do stainless steel, but you're going to have a hard time, folks, beating the price of anything in steel by additive manufacturing. You're going to have to be looking at titanium or aluminum, which is even harder from an additive manufacturing and bonding point of view. So titanium is not a bad place to part start.

Concept Lasers had their US production facility — they're an Austrian, German company. They had two machines up here in Woburn, Massachusetts, and they were making titanium parts for racing cars, and they weighed about two or three pounds, and they were selling them for thirty-five thousand dollars, and it took about three weeks to produce them, each one. It's not a fast process. They use a million dollar machine. We'll go through some of that economics. But you, I mean, I just, I dug through these numbers this morning, took me five or ten minutes, but you just start doing the metrics. I can tell you why they're working on titanium, I can tell you why they're not working on steel. Anybody who's working on steel is an idiot, okay. I'm sorry, I can prove it.

So that's what I'm trying to tell you. There are niches. Boeing's found a niche, and there are many other niches, but a lot of the stuff you're going to see in the additive manufacturing is not a niche, it's just a sales pitch okay. And if you like that sales pitch, I got a bridge in Brooklyn, okay. I'll see you tomorrow, we'll talk about some more, and I would encourage you to go through the blogs, they don't take too long to go through.