SSW_S2013_07

So let me finish up a little bit on what we're doing. At the end of the day last time I had put up some of these — this thing that basically showed you the surface energy of different hydrocarbon chains, going from polyvinylidene chloride, which is two chlorines and two hydrogens on a carbon backbone with a surface energy of 40, all the way down, changing it through polyethylene, which is right here, which is four hydrogens, and getting down to fluorines — and you go one, two, three, four — to the Teflon, which is four fluorines.

Um, the fluorine bond is the most stable bond. The chlorine is a larger atom and gives you more surface energy. Chlorine tends to react with a number of things; fluorine is fairly stable. Fluorides don't like to react with things, so Teflon tends to have the lowest surface energy of any common material, unless you want to get into the noble gases. When they become liquids they have extremely low surface energies, less than a dyne, but that's only at very low temperatures. At room temperature Teflon is the lowest. And so, skating on Teflon with a little bit of water as a lubricant is considered to be better than skating on ice, except you have sharp skates and those sharp skates will cut into the Teflon, and ice tends to be self-healing okay, in terms of filling up the cracks.

Um, I'll talk to you a little bit in a few minutes about how to bond to Teflon, but if you're talking type one adhesive bonding, the lower the surface energy the harder it is to bond to. And for example, a few years ago — it's probably 15 or 20 years ago now — I had a student at Ford who was looking at — they were going to introduce a new type of plastic bumper okay. And the plastic bumper was poly— polyethylene ox— no, PEO, polyethylene oxide or something. Anyway I can't remember the full name of it. But this PEO had a very low surface energy and they had to paint it. So the problem was, whenever they tried to paint it the paint would just peel off. They eventually had to introduce a sandblasting operation to paint this, to roughen the surface. But you only wanted to roughen a little bit because you want a nice glossy paint job, so the roughness has to be probably a third or less of the thickness of the paint layer, and the paint layers on cars are pretty thin. So they had to develop a process to roughen up the surface.

And people do various things. They etch the surface chemically, they mechanically abrade it, they do a number of things to change it over to a type two mechanical interlocking okay. Another type of mechanical interlocking — several of you who've taken parts of the class before have seen this prop. [Tom produces a piece of Millennium cookware.] This is Millennium cookware by Farberware. It illustrates a number of things. It illustrates forming, if you take my deformation processing. It's actually an example of cold bonding. They have a layer of aluminum at the bottom here of the stainless steel. So they draw the stainless steel, they heat up the aluminum in a furnace, they put the pot over a die — a guy by hand takes some tongs, puts the aluminum on top of this, and there's a steel die holding it, and a 5,000-ton press comes down, goes wham, and just with pure normal pressure — is it pure normal pressure? Well, yeah it is. There's a little bit of shear because this thing didn't have some of these rounded edges and stuff after it comes out of the die — it does have rounded edges.

Um, but in fact when I saw this I thought, wait a second, this is going to be a lousy bond. But in fact I realized they wanted — they didn't know it, but they wanted a lousy bond. It will only give them maximum 30% bonded area, it probably only has about 10% bonded area. You don't need a whole lot of strength. I mean, look at the bonded area you've got, and what's this thing going to get, what force on it? It's just to conduct heat across the bottom. If you had 100% bonded area and you put this on the burner, the aluminum has a different coefficient of expansion than the steel and the whole thing would just distort dramatically. But this was the first time that I realized, if you're going to have dissimilar metal joints with different coefficients of thermal expansion, the best thing you can have is a porous interface, because you essentially want something to take up some of those thermal strains, and what better to take up thermal strains than voids — the empty space.

And so actually, based on that thought back in the 1990s, I had a student do her doctoral thesis, and she showed that you could actually get the strongest joint in ceramic-metal joints if you had about 10 volume percent porosity in the proper pattern on the interface. So you don't always want to get 100% joint okay. But that illustrates cold bonding.

But on the inside of this, they came to me back 20 years ago and they said, uh, we sent this to Japan and we found cracks. When it got to Japan, it didn't have cracks when we shipped it from Brooklyn. What happened? And I thought, well in my mind delayed cracking is the same as hydrogen cracking. I said, you've got hydrogen cracking. So when — I got a magnet, and you take a magnet, as you get further up here you go from no magnetism to very high magnetism, because this type of 304 stainless steel transforms from austenite face-centered cubic to martensite body-centered tetragonal, non-magnetic to magnetic. And the greatest deformation is up here at the top. You roll this — you folded this up basically and drew it from a sheet. And so the greatest deformation is up here, the highest martensite, the greatest tendency for hydrogen cracking is up here. I said, I just don't know where the hydrogen came from.

And they said, well could that be because in order to put the Teflon on we wanted to have a porous surface? And so they had plasma-spray stainless steel powder on the inside of this to make a nice porous coating, which they could then infiltrate the Teflon — which has low surface energy so your eggs don't stick or whatever else right. We'll talk a little bit about why eggs stick and other things don't as much. But they said, we use 95% argon, 5% helium [hydrogen] — 5% hydrogen as the gas for the plasma spraying. I said, well that could be the source of hydrogen okay. And it turns out it was. Um, so far as that goes, but they were trying to get a rough surface. And in fact this was the top-end pot 20 years ago, or 15 years ago, however long it was. It was the top end, and at the time it cost $100. It would probably be a $200, $300 pot today okay. Because let's face it, you've got plasma spray and infiltrate with Teflon, you know, forge on this aluminum piece. But in this case they wanted to do mechanical interlocking to get the Teflon to stick, because otherwise Teflon just doesn't want to stick to things. So there is a trick, which I guess I'll show you the trick in a little bit.

But let's finish up — the how viscosity changes in the Stefan equation, which is the kinetics of just squeezing a liquid between two surfaces. You have solvent removal, polymerization, and we really didn't talk very much about solidification last time, but you can change your viscosity by solidifying the liquid. Hot melt glue guns okay — it's a good way to burn your skin right, if you're not careful. Um, asphalt on the roads is nothing more than adhesive holding little bits of rock and gravel and stuff together. Um, so there's solidification — we're using thermal energy to change the viscosity.

Um, so I think we've actually probably gone over most of these things, but I had a list of — we know Elmer's glue is solvent removal, postage stamp. What about Portland cement?

We call it polymerization, but we could call it chemical reaction okay, because most cases we're dealing with a polymer adhesive, but Portland cement is basically a chemical reaction. You take burnt lime, which is calcium oxide, and you add water and you get calcium hydroxide. It's not polymerization in the sense of a polymer, but it is a chemical reaction where you form a new compound. The water reacts with the calcium, the burnt lime. It's actually more complex than that — the chemistry of Portland cement is very complex — but basically in the simplified sense it's calcium oxide forming calcium hydroxide. Okay, we talked about epoxies. Um, airplane glue — you know the little stuff you bond these plastics together with — what's that?

Student: [inaudible]

It's actually vaporization. People used to sniff it to get a little high okay. There's volatile organics off of it. But in fact if you put too much on, it actually dissolves the plastic itself, and now you're actually creating a bond of plastic to plastic, and it's not really an adhesive bond, it's actually a weld with vaporization from it if it's thin right. So epoxies — you call epoxies polymerization right, but a lot of those are also extremely — there's a lot of volatile stuff in them right. There's a lot of — you use them, I mean it's — yeah, you can mix these methods okay. But the general idea of an epoxy is a two-part epoxy reaction okay. But yes, some of them are also solvent removal okay. So it's not as if — once you — I'm just trying to outline the types of physical processes. Sometimes we mix those, just like in the airplane glue, to a certain extent you actually do get polymer bonding across there because you're dissolving part of the little airplane model — plastic model that you're making — and you get polymer bonds all the way across the interface. You also, if you get that much of the solvent on there, you end up making a terrible looking joint okay, which you have to paint afterwards anyway.

Uh, we talked about anaerobics, which are things like Loctite sealants. They're basically adhesives, and we put them on in a big bulk form. And what are they, most of those? Vap— it's vaporization. Most of them, many of them are water-based nowadays. Um, there's been a lot of work to get rid of all the organic-based solvents because volatile organic compounds are now monitored by the EPA and whatnot. So particularly in large volume uses like the automotive industry, they wanted to go to water-based solvents. The problem with water-based solvents — they tend to dissolve in water to make low viscosity once you get them in use, they also dissolve in water to give low viscosity after you've solidified. And so it turns out the automotive industry —

Well, it started in the oil embargo of 1973. Basically people would buy cars before 1973 and expect them to start rusting after about three years, and cars were not exactly disposable, but people didn't keep cars more than 10 years. And if you wanted a nice looking car you basically traded it in after the three-year warranty. That might seem a little foreign, but it was really after the oil embargo that all of a sudden — and I never quite figured out the psychology on this — but people started saying we want longer-lasting cars. I think there's probably a significant increase in other features in cars, whether it be electronics or other things, and the prices started climbing. Um, remember Henry Ford made his money by making cars that were cheap enough that even his factory workers could afford to buy one okay. And so cars were not exactly disposable, but they were limited-time use.

Well by the mid 1970s the automotive companies wanted to do something about the rusting problem. Um, you don't remember it because you weren't born, but I remember the Honda Civic in the mid '70s was a joke. Honda had been making motorcycles when I was a teenager. You know, I wished I had a Honda 50cc motorcycle. Um, but then I saw the pictures — an orthopedic surgeon in town showed the pictures of people who had accidents right in front of the hospital, and there was just nothing they could do for them when you go sliding along the asphalt for 100 yards. And so I kind of lost my desire to do in an auto — in a motorcycle accident. Um, but in any case, Honda went from motorcycles to — in the mid '70s they sold this little Honda Civic which sort of didn't look a lot different than a Mini Cooper does today. It was a small little vehicle, and it would rust in 12 months okay. They got this reputation, and in fact this is one of the reasons Honda is so big on quality today, because they were terribly embarrassed. The Japanese don't like to be embarrassed. But they just had a terrible reputation for quality and rusting.

Anyway, people decided they wanted to use galvanized steel. Well, the problem is they didn't really know how to weld galvanized steel. We know how to do it today, but all through the 1980s I had research projects on welding of galvanized steel in the automotive industry. They were doing it, but at tremendous expense and quality problems and stuff. So in the early '80s they started using adhesive bonding, and one of the easiest things to adhesively bond was just putting the two halves of the door together. You've got the outside door panel and you've got the inside panel, and you've got the latching mechanisms and stuff on the inside of that. But they started adhesively bonding these things together and they worked just fine for about three or four or five years, and after about five years the adhesives were attacked by the moisture — you know, road salts or whatever, rainwater splashing up in there — and these door panels were just completely — you know, the inner and outer were just sort of crimped together okay. They basically folded the outer panel over the inner panel by about a 3/8 of an inch, and they were rattling around in there, and you had all kinds of rattles in your cars.

I remember the Dodge minivan, which was the car that made Lee Iacocca and Chrysler — brought Chrysler back from bankruptcy the first time. Those minivans were — I mean, after, by 1985, they were just a piece of junk, the early minivans, because of the adhesive bonding. And that was partly because people were using water-soluble adhesives. And it turns out we tend to use some of the lowest-cost materials for adhesives. We also use some of the highest-cost materials adhesives — go from about less than 10 cents a pound to about $400 a pound.

So if I'm looking at postage stamps or the glue on envelopes, I use all kinds of things. And I talked about eggs sticking to the pan of a Teflon pot. It turns out proteins tend to want to bond — they have lots of nice radicals on the surface and they tend to want to bond to a lot of things. And so there was at one time in Massachusetts an industry that made a byproduct protein. Um, down here in New Bedford, what's the fishing industry — I just said it okay, the fishing industry made fish glues. And in fact does anyone know what gurry is? It's sort of a term from the old whaling days, but gurry — you have to go to an unabridged dictionary. Gurry is "fishing offal as the refuse from cutting up a whale or the oil, a slimy gummy substance" okay. Gurry is old codfish heads okay, or the inner guts of the codfish. And you can chop that up and make it into a liquid — you know, grind it, and it becomes a wonderful adhesive.

Except I remember back in the early '80s when I would pay my estimated taxes in Massachusetts, they used to give us envelopes to mail in our quarterly estimated taxes, and I will swear that they were using gurry for that adhesive, because it was the worst tasting — I mean I finally got to the point I would use a rubber sponge rather than lick that stuff. So it kind of left a double bad taste in your mouth — had to pay taxes and I also had to taste the gurry. I'm positive it was gurry okay. It was terrible, never tasted anything like it. Um, fish glue was the first liquid glue that reached commercial importance and was the forerunner of all household glues okay.

You've also heard about horses' hooves. Lots of gelatin in a horse's hoof okay. Gelatin is a protein that makes a nice adhesive. Most proteins — milk okay, the leftover part of the milk, the whey from making the cheese. You know, there's not a lot you can do with your curds and whey. You can sell the cottage cheese but the whey is not much good for anything. So um, starches from potatoes, from corn — any kind of starch or protein that's a waste product becomes one of these less-than-10-cents-a-pound glues okay.

And they were probably paying a buck a pound for the adhesive that was going on those Chrysler doors, and other people — I mean Ford and GM had problems too. But eventually they got to the point where now they use about $12, $15 a pound glues that have about a 10-year life, rather than a 3-to-5-year life. And when you go to a more expensive glue formulation, you get something that — in the old days it would be organic-based solvents and therefore it would have good water resistance — um, but now, as of the 1990s, they've developed things that are water-based. When the water goes off, they now are sort of impervious to water, so they have some hydrophobic things in there mixed in. And so the chemists do some magic tricks to take something that's hydrophobic and put it into water solution okay. If you're careful you can play some of these tricks. Um, but you're also now doing some fairly sophisticated chemistry.

Um, it turns out one of the major players in this is the Lord Corporation, which — their headquarters is now in North Carolina, but their original home was up in Erie, Pennsylvania. And it turns out Thomas Lord's grandfather started the company based on a patent from MIT of how to bond rubber to steel. Now this patent was back around 1900 or so, and at the time they needed to bond rubber to steel for all kinds of railroad applications, shock absorbers and things like that. And it was out of Erie, Pennsylvania — one of the big things that's made in Erie even today is, that's where General Electric makes all their locomotives for railroads. Um, so the Lord Corporation got started as bonding rubber to steel, but they're into adhesives, and one of their major businesses now is these high-tech adhesives for automotive.

Um, turns out Thomas Lord did not have any children, and he ended up leaving his part of Lord Corporation — which was half the corporation, and it's privately held but it's probably several billion dollars a year — he left 50% of it to four universities and the Cleveland Clinic, and one of those universities was MIT. And so if you go look in the laboratories down here you'll see — the infinite corridor hallway, there's a little plaque there, it says the Lord Corporation, or Lord Foundation of Massachusetts. And the Lord Foundation of Massachusetts essentially is the 10% of the Lord Corporation — essentially it's not owned by MIT, but the MIT investment portfolio manages it for the Lord Foundation, and their only goal in life is to give money to MIT. So whatever the Lord Corporation declares as a profit each year, 10% of it comes to MIT, or a dividend. Um, so anyway, paying for some of your education.

Um, we talked about surface roughness, and when we get to surface roughness we're talking about something that's going to compete here between type one and type two. This would be a relatively smooth adherent, and if you have good wetting you wouldn't get any gas bubbles in here. If you have a very rough surface you're always going to get a few gas bubbles. Um, but if you have type two you'd like to have a rough surface, to a certain extent. If you have type one you don't want any bubbles in theory, but a little bit of increase in surface area just means you've got a bigger bonding free energy, if you will, to try to pull the adhesive in there. So you actually have competing — do you want a thin joint or a thick joint, and that sort of gets back to the duct tape question okay.

Um, duct tape — we think of it as very good, and first of all I've never seen duct tape adhesive dry out, not in my lifetime. And I've got some duct tape in my office that's 20 years old, and maybe there's a little difference, but it holds — whatever the rubbery constituent and other stuff in there, it doesn't have a very volatile component to it. It mostly relies on the rubberiness and the stickiness. But a lot of what duct tape does when we fix things, we use lots of area.

And so if you start thinking about it, adhesive bonding is inherently a weak bond. We didn't do anything to get the surface clean, get clean metal contact, we've left contamination there. And so the worst thing you could think of would be a simple butt joint like this, where I have that much area. But in fact I should have a nice picture here somewhere — yeah, I do. Um, we tend to do a number of things to increase the bonded area of joints when we're using adhesives. One thing we can do is we can just get 40% more bonded area by using a scarf joint, we can use step joints, we can put straps in between, or double straps. And so now you can make adhesive joints that are stronger than the base material in simple tension, and also pretty good in shear as well. But you still never quite trust the adhesive joint for really long-term use.

I remember 30 years ago, the guy who was my NSF monitor told me the story about going through the Boeing plant. And at Boeing they were very proud of the fact they were now using adhesive bonding to join the skins and wings of aircraft. And so they told him this story and how they had improved that adhesive bonding technology to the point they could now use it to make commercial airliners that would last 30 years. And so they took him out on the floor, and yes, they were using adhesive bonding, but then they were putting rivets in. And he says, well, why are you doing that? And they said, oh, just to be sure, because in 30 years adhesive bonds are going to get weaker. Over time you get some corrosion at the interface as things migrate in.

Um, but in fact those adhesives they're using in the aerospace industry typically might cost $400 a pound okay. They're specially formulated, they're polyimides, which are expensive anyway because of the way they're made. Um, but you go from making envelopes or postage stamps where you're using garbage starches and proteins from fish and horses' hooves and milk residues, at less than 10 cents a pound, up through $15 a pound now today to bond automobiles together for something that's got moisture resistance, all the way up to aviation stuff where you're talking about life safety and you might be paying $400 a pound for it.

You also do things — I didn't bring my piece of armor, but my armor — my little glass armor — you guys have seen that before right. It's four layers of glass and one layer of polycarbonate that's adhesively bonded. And they have to go through all kinds of processes. I mean, it's all hand layup, and they have people with little rollers squeegeeing on the adhesive, because you don't want any bubbles. The bubbles are going to make a lousy window and the President won't be able to get a clear view of the guy who's trying to shoot him okay. Um, anyway, they squeegee them on, and then they put them in a vacuum unit. That'll help. Actually, you can't really pull those bubbles out — they have to put it together pretty well. But they actually put it into a press inside a vacuum system and squeeze everything together and try to make a defect-free adhesive joint. It's not easy to make something that doesn't have some bubbles, but you end up trying to come up with lots of surface area.

Well, if you're talking planes or sheets, this is the way you have to get more surface area. It's easier if you have fibers. If you have a bunch of spaghetti noodles and you want to adhesively bond those, you've got lots of surface area on those fibers. So what's the common material that is adhesively bonded fibers? Fiberglass boats okay, or other pieces of fiberglass. It's nothing more than a polymer adhesive with fiber structure. We call it a composite. We don't think of it as an adhesive, but it is — it's an adhesively bonded fiber okay, whether it's glass fibers or plastic fibers or something else. But even easier is if you have particles — well that's asphalt on the road okay. We don't think of it very often, but there are some composite materials that are particles that are adhesively bonded with something. And essentially you don't have to worry about the strength of the adhesive joint when you get to fiberglass or you get to particle composites, because you've got so much surface area per unit volume that the strength of the joint is not critical. We do have to worry about it when we get to joints that are simple little tensile joints, and we have to increase the surface area by scarfing or putting doubler plates on or things like that okay.

Now let me tell you the trick of how to bond to Teflon. And I used to have — I can't find it okay, I've been looking for it for two years now. Uh, we actually did this test for years and years. I had heard about this — but of how to bond water to Teflon. And ordinarily the problem with water on Teflon — water, I showed you a thing once, said it had a surface energy of 72 ergs per square cm at room temperature. And if this is one of your idealized asperities, water on Teflon is non-wetting because the Teflon has a surface energy of 18 okay and the water is 72. So the water doesn't want to get in there. It has a lousy contact angle, and you can do whatever you want — squeeze on this all you want — and you'll never get the water to really coat down inside that, get mechanical interlocking on that Teflon. So that's if you're trying to do it by surface energy flowing in there.

But you can flip it around. If you could flip it around and deposit the water on the inside like this — and you can, if you vapor-deposit in a temperature gradient okay. So if I had steam up here and I cooled my surface and I vapor-deposited — I ended up depositing water inside these things — the surface energy is now going to hold that water in there. And you say, well why is that important? Well it never was important for about 15 years as I was teaching this course. And then I got involved in some aviation failures, and people were saying it's carburetor failures because of icing of the carburetor okay.

If you are at — even if you're at room temperature, with the lower pressures and the fuel vaporization and everything, even at room temperature or ambient temperature, 70 degrees, depending on the humidity in the air and stuff, you can develop ice in the carburetor. And if that blocks the carburetor then you're in trouble — you get no fuel to your engine and it just doesn't run as well okay. And if you're in an airplane that's really not a good thing. And one of the problems is, you're up there very high and it could be low temperature — even if it's a nice day here, it gets colder as you go up. And so icing in the carburetor is something you have to worry about, and so they put heaters in the carburetors, and there's a little switch in the cockpit and the pilot can turn on the deicing switch okay, to heat up the carburetor so you don't get water deposited.

Well, someone told me that the Canadian aviation authority — and I have since gotten a copy of this report and looked at it — but they ran tests to see if putting Teflon on the surface of the carburetor would prevent ice from building up on the surface. They were also interested in deicing of airplane wings. I mean, if you've ever taken off in the winter at Logan and they have to go and spray your aircraft with deicing solution okay — do you know what that solution is? Usually it's polyethylene glycol. Um, however I think they may have switched recently to polypropylene glycol. Um, one of them is toxic, and they have to take the planes over to a little part of the area because they now have a little dam around there so they can collect all the polyethylene glycol, because the EPA gets very upset if you just let it get into the ocean. But now I've noticed they've been deicing just back from the gate, which means they've switched to polypropylene glycol — which is a food additive.

And for years I knew that there was this difference — one was toxic and the other wasn't — and I just read about a year ago the difference. Polyethylene glycol — it's an alcohol, and the body burns it and makes methanol. And what does methanol do to you? Turns you blind and does all kinds of other things okay. It's why you're not supposed to drink white lightning and stuff, because if you really are going to drink alcohol you should drink fairly pure ethanol without a lot of methanol in it okay. And the problem with polyethylene glycol — the body metabolizes it to methanol, which will turn you blind. That's why polyethylene glycol is not a food additive. Polypropylene glycol, with an extra carbon — what does it oxidize to? Meth— uh, ethanol. And so you get a little high. If you want to put that in your baking, it's actually often in baked goods. But in any case, because polypropylene glycol is quite a bit more expensive — and you can kind of tell which one they're using — more and more the airports are going to polypropylene glycol, even though it's more expensive, but they can just do it right there at the gate, because if it gets in the water supply, so what, the fish get high okay. Anyway, they probably get high the other way, but they'd be blind anyway.

Um, in any case — if you vapor-deposited — so I actually had a reason to run a test. So we just went to the store and got a Teflon-coated frying pan, cut out a flat sheet out of the bottom, soldered on a cooling pad underneath. And we had two sides — one side we — well actually, the whole thing, we put some drops of water on, and then we took a — when it was cold, we misted, or actually steam, you know, basically steamed it. And when you were all done you had these drops of water frozen on there, and you had a film of frost frozen on there. And you could take a steel blade and you just pop those drops of water off because they were just sitting on there as liquid on top of the Teflon and they never bonded. You could not get rid of that frost — you'd have to scrape away the Teflon before you could get rid of the frost.

So for years I had read about this and talked about it. Um, it turns out that I finally found an application for it, which is in defrosting of aircraft. And then the question is, should you put Teflon on your carburetor or not? Well, the answer is, it depends on what's causing icing. If it's water droplets because you're flying in rain — let's say you were flying last night here in Boston and you had drops of water that are near the freezing point — when they go through that carburetor, they're like the drops that can't do wetting down in the surface asperities, no mechanical interlocking. Teflon would be great. But what if it's just the humidity in the air, and that's cooling and depositing on the walls of the carburetor, in which case now you're getting frost buildup that you can't get off with a sledgehammer? Um, and so should you line your carburetors with Teflon or not? It depends on what type of icing you're getting. And the real answer is, no, you should turn your heater on, you idiot. I mean, if you're a pilot, you're supposed to know that. And if you forget to do it, it's Darwin's principle. You know, Darwin's revenge, and you deserve to go. Okay, if you can't remember your training okay.

Um, so other things about adhesives. Um, I showed you aluminum oxide and how we roughen up the surface by anodizing the surface of aluminum oxide. Just after World War II, someone discovered that you could phosphate steel. And I apologize, some of you, if you took one of the other modules would have seen this toy before, but you know I've only got so many toys after 35 years. Um, but what I have here is a cold-formed part. [Tom holds up a small fastener.] This is a zinc-plated mechanical fastening piece that allows you, if you're making particle board furniture, cheap stuff, you can basically embed this in the surface, drill a hole, embed this — it's got little knurled edges — and then put a screw through it. So if this comes in the bottom, it's got a shoulder, and you can put a regular old machine screw through this. It's threaded on the inside, and you can make a strong joint to a particle-board piece of junk furniture okay. Um, so there's lots of little parts like this.

But in the old days before World War II, you couldn't afford to make these things. A part like that, instead of being five cents, might be a $1.50, because you had to do it in an automatic screw machine. Well, what happened is someone found out that if they phosphated the steel — iron phosphate on the surface of steel forms a very porous surface, just like the aluminum oxide anodizing on the aluminum. And it used to be, when I had this originally 25 years ago, it would feel slippery — in fact I still feel a little slipperiness to it. This is just a sheared blank of 3/8 inch steel okay, and you can actually see the shear marks. They just cut it off in the cold head — it's called a cold heading machine. And the next blank, all they did is take one of these. I actually went to the plant, and you have to wear your ear muffs on a plant like this, because basically in a cold heading machine you're doing cold forging of these parts, and these machines are slamming together at about three to five times a second, and it's just going bang. It's about — it's above 125 dB, and 130 is the threshold of pain. If you weren't wearing ear muffs — and I'm sure the people who work there probably do lose their hearing after a while.

But they would phosphate the steel and then they would put calcium stearate — you know it as soap okay — which is a lubricant. But now the lubricant is in this thicker porous phosphated surface, so you're not just putting on a film. So if you go to certain parts of the country that have hard water and take a shower, you can feel the film of slimy — you can rinse in hot water all day long but you still — your skin still feels a little slippery. And that's because you've got a little layer of calcium stearate on you. The calcium stearate has a polar — they typically draw it like this, and then they just kind of do something like this. But this is a hydrocarbon backbone, and this is non-polar, and this is polar.

And so in terms of surfaces, if I've got a surface like this, I get the stearate molecules line up on my surface, and all of a sudden I end up having — it's like grass, a grass of polyethylene if you will, because this is nothing more than a polyethylene chain okay. It's a grass of polyethylene rooted to the surface of whatever it is by the polar bonds here. And the calcium stearate is a great lubricant. Uh, there's oleates and citrates and all kinds of things. One time I wanted to go through and figure out the genealogy of all these. There's lactates, oleates — if you go back to the Latin root, I think oleates may mean sheep, and it turns out sheep fat has a certain number of carbons in the chain, and stearates are like 20 — it's one of the longer chains. But you lubricate it with calcium stearate and now you can have lots of sliding. Remember, lubrication is opposite of welding.

In your dies, in your metal dies, you can form it, size it, and then start to extrude a hole, extrude a deeper hole, form a flange. This little piece right in here — oops — is actually the one piece of scrap. You end up punching out a little hole, and that is the one piece of scrap. You end up curling it, threading it, plating it, and those are the steps in making that cheap little part, all in one little thing. Um, so cold heading has made it possible for us to make all kinds of inexpensive metal parts. If you go through a big automotive part, most of the steering knuckles and things on the car are basically made not necessarily by cold heading but by warm heading. They might be doing it 400, 500°, and you might have parts that are four or five pounds okay, basically forgings that are warm-forged because of the ability to do adhesive bonding of the lubricant to the surface okay.

Um, any questions on any of that? Nope, okay, well if not — you don't have a lot of questions. Is it all just crystal clear? Let's go on to diffusion bonding. We've talked about the two limiting things — I told you you'll get sick of this — are surface contamination and surface roughness. We spent a lot of time on cold pressure or warm pressure welding, friction welding, um, ultrasonic welding and whatnot. If you go to — those are things where pressure was the primary thing breaking down, mostly in shear, those surface asperities, and you had to clean the surfaces or extrude the contamination away. But in diffusion bonding, we're mostly going to use temperature to diffuse away the surface impurities, and we're going to use temperature to make it such that we can just flatten these surface asperities under pressure. So we're typically going to use 5,000 to 10,000 PSI pressure.

And essentially — you have this in your handouts — this page somewhere in your handouts. Um, this comes out of an old welding handbook for diffusion bonding. So you have your grains, and you have initial asperity contact right here and right here, and these are voids — these long elongated things are voids. That's the initial contact. You press it down, and I'm going to talk about contacts now — you never can get more than about 1/3 contact area, and we'll prove that at some point. Uh, the first stage, deformation and initial boundary formation — you're going to be at high enough temperatures, typically above 8/10 of the absolute melting temperature, and you will get diffusion across the interface and you'll form a grain boundary. These voids will shrink as little asperities, and you'll have a linear grain boundary across there if you look at this on the microscope. That's stage one.

Stage two, you actually will have some of those interfacial grain boundaries remaining, but some of the voids will get trapped on the inside of the grains. And the third stage is when you trap the voids and you no longer have any of that original interface grain boundary. You actually have the grains have recrystallized and grown across the interface, and except for those porosities, you now have a perfect joint. Diffusion bonding is one of the ways to get a near-perfect joint. This is — I think I've shown it to you before — part of a development of titanium diffusion bonding for the F20 — uh, is that the Raptor? Okay — for one of the engines they did down at Pratt and Whitney, so far as that goes okay.

Um, what's happened is, there are certain metals that are very easy to diffusion bond, and these happen to be things that have oxides that are easy to dissolve into the material. And those are, fortunately, things like iron — although we tend not to use it for a lot of iron because we don't have the really expensive parts; this is an expensive process — titanium, nickel, and a number of other things. But among the common metals, iron, titanium, and nickel, and in particular titanium. But the things that can't be bonded very easily are aluminum and magnesium. Why? Because they form very stable oxides that won't dissolve into the base material.

What happens in titanium — our poster child for diffusion bonding — above 900° centigrade titanium will consume its own oxide. In fact that's been a problem in the titanium jet engine business. They've actually had titanium fires in the engine if they run the engine improperly and they get above 900° centigrade. The protective titanium oxide skin diffuses into the titanium. Now you expose fresh titanium to hot compressed air, and you start a fire. And it's like having a flare go off in your engine, only the fuel is the engine. And you end up with a hollow shell with a bunch of white powder all over everywhere — it's called titanium dioxide. Um, and so this was a big concern. It happened several times. The Air Force was not happy. Um, commercial guys were worried it was going to happen to them, but it really was because people were pushing their engines too hard so far as that goes.

Um, I guess I'll save till tomorrow — or actually not tomorrow, I'm up next on Thursday. Dr Belmar will be here tomorrow and Wednesday, and then I'll be here on Thursday and Friday of this week. So we have class every day this week okay. And on Thursday I'll talk to you about super-plastic forming and diffusion bonding, which is one of the ways we make jet engine parts.