SSW_S2013_03

This location, you haven't read the uh, you weren't here yesterday right see. But anyway, we talked about dislocations and how the theoretical strength of a chemical bond is on the order of millions of psi. If you actually took the bonding energy of one to three electron volts per atom and took that across one square centimeter, which is about what, 10 to the 16th atoms or something like that, you would end up with several million psi. Well, the reason metals are weaker is because of dislocations.

And you can take a telephone book — this is the MIT faculty staff directory from 2009 and '10, and I save these now — but you can introduce a dislocation, okay, and now all you have to do — and unfortunately they now put polymers in these papers, so it's a lot harder than it used to be. [Tom tears at the directory.] You basically just bend it and break one page at a time. And I've gotten about halfway through now. In the old days when I had nice brittle paper... But you could break these things very easily, but in any case, get enough of a tear in there, and once you've got the stress concentration you can just rip it. But that was the trick that people used in the old days when they ripped telephone books. And I actually did it once with the yellow pages, the old brittle yellow pages.

A dislocation will weaken a crystal, a ductile crystal, by about a factor of 10. Those are the two handouts from the other day that, last Thursday or whatever. So if I actually look — and can we see that the shades are not coming down in this room — so darken it up a little and blow it up. But this is a book on the plastic deformation of metals, and so it shows properties of whiskers of different materials. An iron whisker has a tensile strength of about 1.8 million psi. People have measured this experimentally. Aluminum oxide sapphire whiskers, 2.2 million psi. Silicon carbide, 3 million psi.

So it turns out silicon carbide is one of the stronger bonds we know. We know that because it's a very hard material, but the chemists know it because they measure bond energies in electron volts so far as that goes. So whiskers, or two atoms coming together, would like to bond, and they like to bond with tremendous strength. And as I pointed out yesterday, the first thing to inhibit bonding is surface contamination.

If you keep something clean and bring it together without any oils or absorbed gases or anything else on the surface, and you actually have two primary bonds, they come together, they will bond with tremendous strength. And they've done this in ultra high vacuum systems with gold back in the 1950s. They would take two very flat surfaces of gold and bring them together in an ultra high vacuum system after they made them very clean, and they would stick together with the strength of a braze joint, okay. They would still fail at the braze joint, and that's because, no matter — well actually does anyone know what why that is? What's the second inhibitor to bonding? It's not just chemical contamination on the surface. It's surface roughness.

So everything we're going to do in this module is going to talk about these two things, and what we do in the world to overcome these things. And that's going to relate to cold welding, soldering, brazing, adhesive bonding, and diffusion bonding, which are actually the topics. If you go to the outline that Dr. Belmar [Bellmore?] gave you the first day, this module is going to talk about cold welding, soldering, brazing, adhesive bonding, and diffusion bonding. So of those seven things, five of them I'm going to relate back to this type of surface contamination and surface roughness. So that's sort of the theme for the next 10 lectures. But in fact this is experimental proof that primary chemical bonds have tremendous strength on the order of millions of psi, even though most of the things we deal with in the world only have strengths about one tenth that.

This book is the plastic deformation of metals by R.W.K. Honeycombe. This is what I studied as a junior at MIT. And Professor Bever was an old German, used to make us put covers on our textbooks and bring them to class. Okay, so we don't treat you like elementary students anymore, but anyway, he did. He was the only professor I ever had who made us put covers on our texts. That's why mine has a cover on it.

Turns out if you go to — well before I put this up — if you go to things that try to classify all the different welding and joining processes, here's the American Welding Society welding handbook, which is, uh, not eighth edition, there's a ninth edition, but it's too thick to bring anyway. If you look in the back of this they will have two pages — is it two pages or three? I guess it's just two pages of fine print that you don't have to read all this, which is a designation of welding and allied processes by their letters. So the welding society likes to use acronyms. Adhesive bonding is ABD, okay, I don't know where you get D, but anyway. Arc welding is AW. Atomic hydrogen welding is AHW. Anyway, on these two pages they list 199 different welding processes. And most welding courses will take you through some subset of those, and they will describe each one. Can you think of anything more boring? Okay, which is why I don't do that. But that's how most people do it.

I like to generalize a little bit, and we can take about 60 of those processes, or maybe even 100 of those processes — and we'll talk about a hundred in this course — and what are the underlying physical principles. There's another book by Peter Holcroft [Houldcroft]. I'm sure Peter is retired now. We're going to put Peter — so, but he was, uh, the director of research of the British Welding Institute. I think I mentioned, after World War II, there were three places that went to study brittle fracture of seals [ships] and one was the British Welding [Institute]. So this is a place where 500 scientists in the United Kingdom probably does 10 or 20 million dollars worth of research on welding technology each year. And Peter [Houldcroft] was the director of research.

And in this textbook he has his classification of welding processes. American Welding Society just lists a table, but [Houldcroft] basically — and you should have this handout in your, from your handouts. So here's welding process classification, and across the top he has shielding method. You can use a vacuum, so you don't have any contamination. Inert gas. You can use a gas which might be a reactive gas. A flux, which is a liquid typically — you can't have gaseous fluxes. No shielding is just a, sealing method, I guess. Mechanical exclusion, you just try to scrape the contamination away. But basically this is all trying to take care of surface contamination across this part of his diagram.

And here he has source of heat. And just like shielding has no shielding, source of heat is no heat, excuse me. Anyway that would be cold welding, cold pressure welding, thermal compression bonding — I'll talk what that is in a second. Mechanical — well that's not a source of heat, or actually it could be, there's explosive welding and it does get things warm, and we'll probably talk about it at some point. Friction, ultrasonic, thermochemical, electrical resistance, electric arc, radiant energy. And I made the point yesterday, as soon as someone develops a new source of heat, we tend to try to use it for welding. And what are we doing on this axis of this little chart? We're trying to somehow get rid of that surface roughness, okay. At least that's one way to think about it, and I don't think he necessarily thought about it that way when he put this together.

But when we talk about surface roughness — and I've been using these for nearly 30 years now, I have gotten some others but these are still the best, okay — [Tom holds up packing foam.] obviously it's just packing stuff right, okay, foam packing. If you look at a surface and it might look smooth — that table looks smooth to you, right? But if you had a microscope, that table would be rough, right. So I can talk about putting two surfaces together, and we can go back to this gold in the ultra high vacuum and see if we can get it to bond. Well, if the two surfaces are slightly curved, they don't really contact over the whole area. And that's why you can take two extremely flat gold surfaces, polished to what we call an optical flat, so they're extremely flat, and gold is somewhat soft, and you can squeeze them together in an ultra high vacuum, and then you try to pull them apart, and they've got the strength about 50 percent of the strength of a solid piece of gold.

And that's because gold is somewhat unique in that it doesn't form surface oxides in the air. It might absorb hydrocarbons, but in an ultra high vacuum there's no hydrocarbons, so you can get a perfectly clean, a nearly perfectly clean, surface. And if it was perfectly flat and you could bring them into contact, you would have a perfectly flat grain boundary. But on an atomic scale, you actually have some surface roughness.

And one of the things we can do to get rid of the surface roughness — [Tom presses two pieces of packing foam together.] if I put this together like this, I can see light through there, right? If I squeeze on it harder, I may not see the light through there, but there's still areas, that voids in there, okay. I can squeeze on this until I'm deforming the piece of metal or whatever, and turns out — I'm going to prove to you a little bit later — you still will only get about one third area of contact. And so in fact when they squeeze the gold together in an ultra high vacuum system, they get about one-third, one-half the strength of a solid piece of gold. Not because of surface contamination, but because of surface roughness. If you just squeeze two things together directly, you will not break down all those asperities, those little mountains and valleys, and get a 100 percent area of contact. You actually have to have some shear between the surfaces to break down, break off the mountains, and get down into the valleys.

So if you get rid of your surface contamination completely — gold in an ultra high vacuum — and you actually rubbed the gold together in a shear test, you could get 100 percent of the strength of gold, okay. So things want to bond, they really do want to bond, but because of these two problems they don't bond.

Now there are some processes that we've developed, because we use whatever we can and try to find some application. One is friction welding. In friction welding, you take a bar of metal, you take another bar and you spin it around. Depending on its diameter, but you might get to a couple hundred surface feet per minute, which is pretty fast. You can do this in a lathe, which is how this was actually — you can't always do it in the lathe, it has to be a special lathe with a very fast brake, okay. But the friction weld, you hold one end steady, and the other one you butt the thing up while it's spinning fast, and the friction generates heat. And it should be on my website somewhere, but there's a video of friction welding. There's a company down out in Indiana that makes friction welding machines.

Now what do we use friction welding for? We actually, in this case, we get rid of the surface asperities, the surface roughness, by just a lot of, lots of shear, right. How do we get rid of the surface contamination? This is basically just done in the air. Well, if you look at this, we do it by just excluding it, as [Houldcroft] says. He has mechanical exclusion up in that top corner, top right corner. And he calls mechanical hot pressure, cold pressure — and here he's got friction welding, he's also got ultrasonic welding. Instead of having a high rate of shear, with this thing's spinning around and you push it together, everything just starts to glow red on steel. You can actually see some of the oxide. This one was made 20 years ago, but in any case, it glows red and you just forge it out and it forms a nice little ring.

And I think it got rid of my black plastic pipe that I showed yesterday, but it's — that's actually not friction welded, but they actually heat up the two ends and they just forge them together and extrude out the plastic, because the plastic's got surface contamination. And you just get the same type of curled edge so far as that goes. So that's a friction weld.

Another approach is not to use lots of high shear, but just rub it together and cause fretting. Anybody know what fretting is? Something — when I have some mechanical device or component is vibrating, I just get a little bit of microscopic motion, and that's called fretting. And that microscopic motion — there's something we call fretting corrosion or fretting wear that can break up those surface oxides and get rid of my surface contamination locally and cause metal contact, metal-metal contact, true metal contact. So we can do that, and we call that ultrasonic welding.

We can take — and this is an ultrasonic weld that was made at Ford research a number of years ago on aluminum, has a very stable oxide, this surface contamination, and won't bond if you just stick them together. But if I put it in a mechanical press and hold one of them stationary — you'll see some little ribs on here — and the other one you'll see a different pattern, and then you just vibrate the thing ultrasonically. Anybody know what an ultrasonic horn looks like? A couple of your mechanical type engineers. So I'll tell you what an ultrasonic horn looks like.

[Tom draws on the board.] In an ultrasonic welder, you basically have a bar, and you shape it like this. You might put a wheel on the end, but I'm just gonna put a great big plug on the end. And you vibrate this thing back and forth this way, okay. Does anyone know why I taper it down like this? If I put sound waves through a solid, the sound energy — there's a certain amount of sound energy in this volume, and that displacement gets magnified because the same sound waves, but they get magnified by tapering the area down. You just think it's — the energy is force times the displacement, and you think of a displacement wave going through here. If you taper this down gently, you basically can magnify the displacement. And so I'm getting — I put my work piece down here, two pieces of aluminum, I have a big solid anvil down here.

This rests on, I apply some force from the top, and I just fretting wear these two pieces of material against each other. I break up the surface oxide by fretting corrosion or fretting wear phenomena. I expose aluminum metal to aluminum metal, and I get a bond. Now what's the problem with an ultrasonic bond on aluminum? I didn't really eliminate the contamination, all I did was redistribute it. It's still there. Half of the bond area is still aluminum oxide twice as thick as it had been before. So I never get 100 percent joint efficiency on a bond, unless I really do it. And if you actually — if I had a cross section of that little aluminum piece I sent around, you'll find they, with the heavy indentation on the sheet metal, they have an interface that I just increased the total area by about 400 percent. And therefore if I've got 200, even at 50 percent of it's contaminated, but it's got 400 percent of the original area, I actually can get a joint that's just as strong. So that will actually pull a nugget. If you put that in a tensile machine, you'll rip out a piece of aluminum from one piece to the other.

Why was Ford looking at it? They want to build all aluminum automobiles. I told you the story about all aluminum automobiles yesterday. People would love to build all aluminum automobiles because they're lightweight, fuel efficient. But there's several problems. One, aluminum costs two and a half to three times as much as steel. So steel might cost you 40 cents a pound as sheet metal, but aluminum will cost you typically, because of other reasons, about five times that, about two dollars a pound. And what's the value of a pound saved on an automobile? It's one — it's two dollars over the life of the vehicle. Well, if I paid that much for the material and I haven't even stamped it and formed it and everything else, then I can't afford to use aluminum, unless I'm building a hundred thousand dollar car. So I'm building a $20,000 car, I can't afford it.

You want to build a hundred thousand dollar car, any fool can build a hundred thousand dollar aluminum car if they want to invest the time and effort. It only took Alcoa and Audi about 10 years to develop all the processes to make it all-aluminum auto body. But people did it in the 1930s, only they didn't do it for a hundred thousand dollars. It was Andrew Mellon of Mellon Bank, and he was on the board of Alcoa, and he had more money than anybody, you know, most of the other people in Pittsburgh, maybe all the other people in Pittsburgh combined. And he wanted an aluminum car, so he didn't care what he paid for it. We know how to join aluminum, but...

What's the value of a pound saved on the aluminum of a commercial aircraft over the life of the vehicle? How many magazines they put on board, because a pound of weight on a commercial aircraft is about $200. Anybody have an idea of how much it is on a military aircraft? They don't make as many military aircraft. The higher performance, lighter weight, higher strength alloys — and it's about a thousand dollars a pound for the fuselage now. That number of $200 a pound was secret for many years. That was a competitive edge that Boeing and Airbus — except in 1989 we had a student go out and work at Boeing on the manufacturing program, and they came back and they found that they did a big analysis and they came up with $188 a pound. Close enough to $200 for me, okay.

So you actually have to look at what type of transportation you're doing. And I told you about space-based things, and it costs $20,000 a pound to get something into orbit. If you go back and look in the 1970s, for the space shuttle, the goal of the space shuttle at that time was to lower the cost of payload in orbit from $10,000 — sort of missed the target, it would be cheaper. Problem is they were going to have space shuttle shots twice a week, 100 a year, except they never got there because it was a little more complex than they had anticipated and a lot more expensive. So the space shuttle was an absolute disaster from a financial point of view. But it was great for NASA, bringing in billions, tens of billions of dollars every year.

Anyway, so [Houldcroft] lists all the different welding processes essentially in terms of surface contamination and surface roughness. Um, in a number of these — well, mechanically we have to add some shear or some fretting. It turns out that once at Sandia National Labs — anybody know what Sandia is? There's two of them. What's Sandia? It's a national lab — well thank you very much. Oh well, okay, that's sort of a tautology. Anyway, saying there's two Sandias.

Sandia in Albuquerque, New Mexico is — well let me back up. After World War II and the Manhattan Project, they went for these things that they wanted, for safety. And you see that in the navy nuclear program — there was Westinghouse Bettis and there was General Electric Knolls Atomic Power Labs for Admiral Rickover, submarines. And that was competition. The government was paying twice to make sure that things were done right, and someone didn't have a monopoly. And they had similar things, they had um, Lawrence Livermore Laboratory and Los Alamos Laboratory. Lawrence Livermore was out in California, run by the University of California. Los Alamos was also run by the University of California, so how may you be competing — but anyway, I'm sure they did, they were certainly competing for research funds.

But in terms of nuclear safety there was only one lab, and it was Sandia in Albuquerque, New Mexico. And Sandia had another office right across the street from Lawrence Livermore called Sandia Livermore. But it was all Sandia, and it was run by AT&T Bell Labs under contract for the government. And their job was nuclear safety. We wouldn't let anyone steal our weapons, and we also wouldn't have a nuclear weapon go off when it wasn't supposed to go off, okay. So that's still their mandate. And there was no competition, there was no competing lab, so far as that goes.

But Sandia is now the largest national laboratory. It's about three or three and a half billion dollars a year. The federal government spends about 700 billion dollars a year, right — no, not 700, but there's 700 national labs at about a hundred billion dollars a year, but Sandia gets like two or three percent or four percent of that, uh, for nuclear weapons safety. So Sandia is actually one of the premier national laboratories. And they spend money where they used to spend it like it's going out of style. Anyway um, they sort of had a blank check. If you went and told Congress, well it's for safety of nuclear weapons, who's going to deny you, right?

So they actually decided one time to do ultrasonic welding of aluminum. I have no idea what the application was. But they basically took two sheets of aluminum and they put their, the sonotrodes — these are called sonotrodes, which are your ultrasonic welding tips. And they made a huge radius on, and I can't even draw it as flat as it should be, but this was your big copper tip. There's another one down here. And they would vibrate these things side to side, but they would vary the pressure coming in. As they did that they would first clean off just the center, because that's where the contact was. But as they continued to press it in — this was on thin material, this was like a four inch radius on this hemisphere. If you, I saw one of the electrodes, to you it looked flat, okay. If it's a half inch wide it's got a, no, a four foot radius — I think it's a four foot, right. And that was flat, but it did have a radius. And they would increase the pressure as they're vibrating this, and they would take that aluminum oxide and they would just sort of push it out in a circular wave. So they got 100 percent area of contact, okay, as opposed to what Ford does and stuff.

And I'm sure they're either doing this as basic research to prove they could do it, which is one of the things the national labs does, or they actually have some application on a nuclear warhead, which is why they would have done it, where they wanted to cold weld aluminum. I just tell you about that not because it's a viable commercial process, but I said yesterday almost anything can be joined if you're willing to pay the price. And so if you understand the physics of surface contamination and surface roughness, and you have to eliminate those — ultrasonic welding is a cold welding process that tends to leave behind little gobs of surface contamination all over the surface, but if you're clever you might be able to sweep it away, which is what they did, okay. So as long as you understand these are the two things that you have to overcome if you want to make a successful cold weld.

So what else do I want to say. Um, so anybody have any questions on that? Let's talk a little bit about — if you don't have any questions, you don't ask questions, I won't be able to address into things. Yes?

Student: [inaudible question about ultrasonic welding configurations]

And what they do is they replace the sonotrode with a wheel. And so they vibrate, and it just, you just pass the thing under it like you're running a sewing machine, squeezing it between the wheels.

Student: [inaudible question, apparently about ultrasonic welding of plastics]

There's no real difference, other than plastics are easier because they're mechanically weaker, and so you don't need as much pressure as you do for metals. Plastics also, it's advantageous for plastics because you're trying to generate a little frictional heat at the interface, okay. And as you do that, um, it's easier to do it in plastics because they have lousy thermal conductivity compared to metals, so you don't need as much energy from your sonotrode, okay, this sonotrode thing, because they melt at lower — they get hot at lower temperatures, they hold their heat better, and they're weaker mechanically. But you're just trying to rub down that surface roughness, okay. And you still have to get rid of contamination — maybe you don't have oxides on plastics, but you got plenty of oils.

What you're talking about — if this is my plastic part, I may put a tip on it like that, okay. And yes, they do, but it's more common in plastics because it's cheap to put that in a plastic. It's a lot more expensive to put it in a piece of metal, okay. So I have seen projections, sometimes called projection ultrasonic welding, okay, where you put these little projections here. These are macroscopic — that might be a sixteenth of an inch deep, okay. And you're just melting that surface and pressing it in, okay. Usually in metals, um, we have big machines with a lot more energy, okay, that are doing the ultrasonic welding.

The ultrasonic welding of plastics is probably the largest application for ultrasonic welding. Lots of plastics, because of all these things we said about their properties. Now where do we use friction welding for metals? The knuckle of the drive shaft on most automobiles. The drive shaft is a tube that goes from the engine to the differential of the car. And you have a universal joint. A universal joint is basically just a U-shaped piece of metal and another U-shaped piece of metal with a cross piece, a cruciform in between, shaped like a cross, and bearings. And this thing has two degrees — how many degrees of freedom? Two, I guess two, right? You basically can wobble along that axis, along the axis that you're spinning your shaft. So when you hit a bump you don't put tremendous stresses and fatigue your drive shaft, because the universal joint can take the angular misalignment when your car flexes and stuff. Usually you have at least one if not two universal joints on every car.

To make it lightweight, because to carry the torque, you only need the outside part of the tube, the wall of the tube. The inside wall, the inside core of a tube in torsion carries no load to speak of. The stiffness goes as the fourth power of the radius. And so if you take the core out of things you can make it lighter weight. The drive shafts on a nuclear submarine have big holes in the middle, they're tubular drive shafts. But you have to make a — use a universal joint, you have to take this forging and weld it to the tube. And they do that by friction welding, because if you look at that friction weld I passed around, that will be made in one and a half seconds in the machine. Just put another one in, just — and just spinning up on, like a lathe, and it's just squeezing them together and forging them together, and your thing will glow red hot in one second and you squeeze it in another half second. So universal joints on cars. All kinds of plastics, little trinkets of plastic. Where did you see all this ultrasonic devices? Oh yeah, so medical, plastics which they're going to throw away.

An ultrasonic weld or a friction weld makes very rapid, very high quality joints, very quickly. It also, if it's not set to the right parameters, it makes scrap very quickly, okay. And the thing is, there's not much you can do about it to repair it, and it usually just turns into scrap. So if you set it up properly, it's running well, very efficient. No filler metal, you know. Once you get the machine set, it just runs itself automatically, usually they have self-feeding things to put in it.

There are some very critical friction welds. If you go up here to General Electric Lynn — anybody know what General Electric Lynn is? We're in Massachusetts. They build jet engines, okay. Now they keep on wanting to close it, because most of the General Electric jet engines are made in Evendale, Ohio, which is outside of Cincinnati. But they don't close General Electric Lynn, although they've been talking about it for 25 years, being pretty close to it. And the reason they don't close it is mostly because they have some very good engineers who don't want to live in Cincinnati, would rather live in New England. I'm serious, okay. I've talked to some of the vice presidents at General Electric and I say, you know, we don't know — you know we could save a lot of money if we consolidate it all to one location. And Evendale's probably eight or ten times the size of Lynn.

But why is General Electric Lynn an important factory for General Electric? It's the first. Who founded General Electric? Two people. One was a guy named Thomas Edison, okay. And the other was a guy named Elihu Thompson. Who was Elihu Thompson? Well, Thomas Edison had over 400 patents. Elihu Thompson had 380. They were number one and two in the country. Elihu Thompson was a professor of electrical engineering in a place called MIT, okay. And if you read the leadership article, there's a quote in there from Thomas Edison — actually I had to find it, it's a great quote, it should be in the beginning, okay.

December 17, 1911, Thomas Alva Edison was quoted: "There is no question but that the Massachusetts Institute of Technology is the best technical school in the country. I have found the graduates of Tech to have a better, more practical, more usable knowledge as a class than the graduates of any other school in the country. The salvation of America lies in the Massachusetts Institute of Technology." Hey, where are your Brass Rats? I don't wear my Brass Rat anymore, but anyway.

MIT was a very practical school when it was founded, and it still is practical in some ways. But in any case, General Electric Lynn was the beginning of the General Electric corporation. And it's now, uh, they still make jet engines there in the May, for another couple years, um, but they want to close it eventually. Anyway, so where was I — I was beside the faculty newsletter. Oh, I was talking about, I was going to tell you about friction welds that are made up here.

So they make titanium turbine disks, okay, up here for jet engines. And it turns out one of these disks, after it's been forged and heat treated and machined, might be worth $25,000, okay. And it's got to be joined to another one in a geometry that's very complex. Sometimes they electron beam weld these things, which we'll, you'll get in another module. Up there, and they have big electron beam welders, and I have stories on that. But they also use friction welding, because sometimes the geometry is such that the two surfaces you want to friction weld together — it's basically sort of two tubular surfaces, or a flat surface of the tube that comes up against it. You can't even get the electron beam down in there. So sometimes they friction weld two $25,000 parts to make a $70,000 part. And they really want it to be a good friction weld the first time, because otherwise you end up with $50,000 worth of scrap, okay.

So they have spent lots of money instrumenting the force and the pressure and the displacement to make sure they get a good weld. Because a lot of times, once you've made that weld, it's very difficult to inspect it to make sure it's a good weld. And all you really know is, did it collapse the right amount in the right amount of time on other welds that we made, that we tested, that were good, right, and destructively tested to make sure they were good. So you try to make it reproducible.

But many times we make welds and they're not very inspectable. Okay that spot weld in that ultrasonic joint in aluminum, it's not very inspectable. Just think about it — surrounded by a crack, right? I mean, you have two sheets together and you weld it one spot, well there's a non-weld shrouding it 360 degrees. How do you inspect that and know that you've got the weld you want, okay. So I remember in the automotive industry they were trying to — I remember spot welding of aluminum, spot welding of steel, but this one guy came in, he said, well all we need to do is we'll make our welds and then we'll come in and we'll hit the thing really hard with a hammer, and if it breaks, it's not good enough, and if it holds up to this big hammer blow, that's good enough. And one person said that's the most ridiculous thing they had ever heard. I mean, you know, you're going to try to destroy it to see if it's any good, right? I mean, you could be introducing damage in that big hammer blow that means it's now going to fail at one tenth. Anyway, it's just a, it's not the way to do it, but that was one person's solution.

Um, so we talked to — let me start talking about primary bonds now, and let's look at other questions. I appreciate questions, see how I was able to digress for 10 minutes. Anyway, so let's talk about the strength of bonds. So we have primary bonds, and what are the three types of primary bonds? All those you took 3.091. There's — covalent, very good. Where are we? Metallic. We forgot — we're, this used to be the metallurgy department, actually it started out as the Mining and Geology Department. Metallurgy didn't enter the name until 1988 — or 1888, 1988, anyway.

The typical bond energy in electron volts is one to three electron volts for all three of these, okay. Lead melts at low temperatures, is probably down one electron volt. Tungsten is the highest melting metal, probably around three electron volts, whatever. Covalent — what's a good example of covalent bonding? Diamond, silicon. Diamond, probably around three electron volts. Ionic bonding — salt, sodium chloride. Probably sodium chloride melts around 900 C if I remember, so it's probably two electron volts. But I got aluminum oxide sapphire, second to diamond, and that's sort of covalent and ionic, but let's say it was — let's take, there must be something up there at very high — calcium fluoride or something, that's pure ionic, nearly pure ionic bonding, melts at, really relative high — yeah okay, zirconium oxide, probably three electron volts, melts at 2000 C, right.

And the strength of the bonds is related to the melting point. And in fact the strength of the bonds, if I draw the Lennard-Jones potential, which is the energy versus distance — the strength of the bond is the depth of that well. Okay, this is the energy of bonding right here, and the deeper that well, that's — this is the depth of the well of the bond. And it turns out I can have deep wells. I also can have deep wells that have very narrow curvature down here. And what does that mean? Actually, the deeper it gets, the sharper the curvature at the bottom, because the distances, the separation, aren't all that different.

Turns out, the first derivative of this is the force, and at zero force it's at its equilibrium separation. It doesn't want to get bigger or smaller. The second derivative is the modulus. And so something that has a very deep bond or a high bond energy tends to have a high modulus. You go back to the chemical properties or the chemical bonding, and most of your physical properties relate directly to the bonding. Which is why in 3.091 we teach students material science based on chemical bonding, okay.

Now, it turns out there's some other bonds. What are the non-primary bonds? Secondary bonds. You remember the name — van der Waals. A van der Waals bond has a strength of about a tenth of an electron volt. And there's something in between here, metallic and van der Waals. Hydrogen, exactly. Hydrogen bonding is just a form of van der Waals, but it's so much stronger — it's about three tenths of an electron volt — that we sort of give it its own classification. It's also very ubiquitous, okay. And so a lot of materials have properties, many of your polymers have properties based on their hydrogen bonding, not really their van der Waals bonding. They have all of these.

But the point is, if this has a strength — I didn't take down my thing, but if the metallic bonding of tungsten is 3 million psi, or 3000 ksi, and for aluminum oxide it's 3 million psi, and ionic has something similar — what's the strength of a hydrogen bond? Three hundred thousand, or three hundred ksi. What's the strength of the van der Waals bond? About 100 ksi. Well this is as strong as steel — except we have defects.

And there's a quote from Bob Sprague, was head of materials at General Electric Evendale, Ohio in the jet engine business. And he has a number of great quotes. He was just a very frank guy. I never met Bob, but I know the guy who replaced him, and the guy who replaced him. And Bob Sprague had a bunch of great quotes. He said, whenever you first hear about the properties of new material, write it down, because those are the best properties the material will ever have. And there's Jim Williams, who replaced him. And Jim Williams, by the way, was a titanium expert, and he went on to become a dean of Ohio State and stuff. But anyway, Jim used his corollary: whenever you first hear about the cost of a new material, write it down, because that's the lowest cost that material will ever have, okay. We tend to glorify these new materials.

Well, Sprague's — another one of Sprague's quotes is, physicists think that structure determines properties of a material, structure being atomic structure, of how the atoms are arranged. But material scientists understand defects control properties. And that's where we get back to the dislocations. A metal has one tenth the strength of its theoretical strength because it contains dislocations and vacancies and all these other things.

I can take two very similar materials, okay. [Tom holds up samples.] What's this crystal structure? Hexagonal close-pack — it's actually graphite. It actually says it right here on this little tag. You can tell also you can tell because the atoms are black, okay. The — this is diamond cubic. This is also carbon. This is coal — this is graphite, it's black. These are little diamonds, embraced onto a little substrate, okay. So I brought some diamonds and some graphite — two different materials, two materials have identical composition but vastly different properties, okay.

And um, somewhere here I have a little sheet — there it is — that tells you the different properties. Why does this one, what type of bonding in terms of sp orbitals does this one have? Anybody remember? sp2 — triple bonds, that's why you get the hexagonal structure. sp3 — uh, tetrahedra, or yeah, tetrahedral bonds. Tetrahedral, yeah, tetrahedral bonds, okay.

So the bonding type — and we'll finish up here, I'll just run a minute or two over, if I don't get yelled at, my instructor comes in. The property — I think this is in your handouts — diamond sp3, graphite sp2. This is a planar three to 120 degree bond. This is a three-dimensional tetrahedral bond. Diamond cubic, hexagonal close-pack. You can see the two of them. Densities are quite different. One's transparent, one's opaque, okay. One's an insulator, electrical insulator, the other one's a conductor. One's isotropic and has high thermal conductivity, the other is anisotropic, has moderate thermal conductivity — that's just an average of high and low. Mechanical wear is abrasive, and a lubricant. Mechanical hardness, very hard and soft. Abundance, rare, abundant. Cost, very high and low, okay, low to high. So I mean tremendous difference in properties.

And when we get together — I'll be here on Friday, Dr. Belmar [Bellmore?] will be here the next two days, I'll be back on Friday from the red eye, and we can talk about some things again. But I will tell you about how graphite is a lubricant on Earth and an abrasive in space. And that was a big surprise at the beginning of the Mercury — before the Apollo program, the Mercury program. Some of the first space shots, things quit turning. They had put them together with graphite as the lubricant. And when they got into space, they found that graphite was no longer a lubricant, it was an abrasive, okay. And it all has to do with these fundamentals of bonding and contamination and things like that we're talking about.