§1. The Ballpark Franks hot dog cooker — 304 vs. 316 stainless [00:02]
If you come early I'll tell you stories. So this is the hot dog cooker story — a material selection and materials mix-up problem. Has anyone ever had Ballpark Franks made over here in Everett? I get a call one day. They need a metallurgist because their hot dog cooker is falling apart.
Their hot dog cooker is about three or four times the size of this room. It's a stainless steel tunnel filled with steam, and they make a lot of hot dogs over there. They hang the hot dogs up and send them into this steam cooker on a continuous conveyor — goes in raw, comes out cooked, and then they package them. They'd had this cooker for years, and all of a sudden the steel started falling apart. Here's one of the pieces, here's another piece.
The only reason I'm telling this story is because just last week we found some of the pieces as we were cleaning out some junk. I'd been wondering — I used to have some in my office, but the ones in my office just completely crumbled in your hands. [Tom shows a piece of the failed stainless steel.] This is the stainless steel, and it was just falling apart.
So I asked a few more questions. You say this is the same cooker? Well yeah, except we replaced some parts. Okay, so there was a change. Something doesn't work for twenty years and then all of a sudden fall apart — there has to be some event.
It turns out they had to replace some parts that were just worn out, because it does have moving parts in there. And the whole cooker was supposed to be 316 stainless steel. Does anybody know the difference between 316 and 304?
Student: Chloride.
Very good. 316 has two percent molybdenum. It just about doubles the price of the stainless steel. Molybdenum is going for $50 or $100 a pound right now, and two percent in a ton of material is forty pounds, and forty pounds times a hundred is four thousand dollars. So you can double the price of your stainless steel by having two percent molybdenum as opposed to 304. It's exactly the same composition except for that two percent molybdenum.
There's a grade 317 that has four percent molybdenum, and that's actually for implants in the human body, because the body is about the same salinity as salt water. When I was in elementary school they taught us that's proof we all came from fishes in the sea, because we had the same salinity in our blood as salt water. I don't know that's true, but that's what they taught me in Catholic school. I do remember the little video of the little fish growing legs and walking up on land. In any case, there's another grade the Navy's been looking at for submarines which is six percent molybdenum — so you're now talking about twelve thousand dollars a ton extra just for the molybdenum.
What happened is they got 304 stainless steel without the molybdenum in the hot dog cooker, and guess what — there's a lot of salt in hot dogs. 316 had been working for years. I could have brought in a book that basically says 316 will take 3000 parts per million chlorine, and 304 won't be corrosion resistant with a thousand parts per million chlorine. Whatever the level is, it depends on other environments, like how much oxygen is around. The new replacement parts were literally crumbling in the higher salt content. That's the most dramatic example I've ever seen of parts falling apart in 304 versus 316, just because of a little more salt in your hot dogs.
§2. The desalination pump bearing — same story, different industry [04:37]
I have a current problem where one company was making desalinization pumps to sell around the world. What are the major countries that use desalinated water for drinking?
Student: Middle East.
Arab countries, yes. I thought you said Paraguay for a second — they've got some of the wettest areas of the world. So they sold all these stainless steel pumps and they were all supposed to be 316. They put some little inserts in there that turned out to be 302, which is like 304. Within a year, the bearing was falling apart, because these little inserts held the screws that held the bearing together. It ended up destroying the company's reputation in the desalination business.
§3. Bonding determines properties: diamond and graphite [05:43]
Let's finish up on measurement science today, and then we're going to get into codes and standards. Today I wanted to talk about the fact that there are limits to properties.
It comes down to the difference in bonding. Essentially, properties of all materials relate back to the bonding of the materials. The example I like to use is diamond and graphite, because it's the exact same composition. The only thing that's different is the type of bonding. Most of you are material scientists, so you know about the differences in bonding. But diamond and graphite are pretty dramatic in their properties.
You've got sp3 bonding in diamond. [Tom shows a diamond crystal structure model.] It's a cubic crystal structure. The white lines are the face-centered cube, but it's actually called diamond cubic — a bunch of tetrahedra. If you look at one of these atoms in the center, it's got four carbon atoms around it arranged in a tetrahedron. That's sp3 hybrid bonding. You take four bonds, one s bond and three p bonds, and you end up with a little tetrahedron that is spherically symmetric.
Graphite is sp2 bonding, and every carbon atom has three neighbors all in a plane, and you get this layered structure. Same composition, but different structure. One three-dimensional structure, one two-dimensional structure. Very different properties. Diamond cubic versus hexagonal close-packed. The stronger bonds in diamond give you a higher density. Diamonds are transparent, graphite is opaque. In fact, diamond is the most transparent at low energy light. You can go out to 50 micron radiation in the infrared, and diamond is the only thing that's transparent out there. Even some of your ceramic crystals like potassium bromide only go out to like 30 micron wavelengths. Diamond will go out to 50 microns. It's the most transparent in the infrared of any material in the world, because it's one of the strongest bonds.
I used to say it's the strongest bond, the carbon bond. In fact, the silicon-oxygen bond is stronger than the carbon bond. But carbon is second only to silicon-oxygen. Anybody know where we have silicon-oxygen bonds?
Student: Silicones.
Siloxanes, yes — all the silicone family of polymers are silicon-oxygen backbones. Our hydrocarbons have backbones of carbon chains, like polyethylene. So graphite is opaque. What's the difference between something that's transparent and something that's opaque in terms of electronic structure? Come on, some of you material scientists, you've learned all this — you've got to start integrating what you learned.
Student: Band gap, free electrons.
Diamond has no free electrons to absorb light. Graphite has free electrons. The sp2 hybrids leave a band in the band gap that's a smaller distance in terms of electron volts, so it can absorb visible radiation. Diamond sp3 has a huge band gap. Unless it's got impurity centers — or color centers, as we sometimes call them — to absorb the light, there is no color to diamond. It has the largest band gap. That's why it's transparent all the way out to 50 microns. Graphite is opaque because it has free electrons, because of the sp2 bonds. There's essentially a free electron for every atom between these planes. These long bonds here are the free electrons. They will conduct electricity, so graphite's a conductor and diamond is a great insulator.
Thermal properties: diamond has the highest thermal conductivity of any material, because it has the strongest bonds and the stiffest crystal structure. Graphite is anisotropic — great thermal conductivity in the basal plane, not so good across the hexagonal plane. Mechanical wear: diamond is an abrasive, graphite is a lubricant. How much different can you get? Mechanical hardness: very hard and soft. Abundance: rare and abundant. Machinability: difficult to machine diamond, very easy to machine graphite. Tremendous differences in physical properties of materials depending not just on composition, which is what we usually think about, but on bonding methods. Even if the composition is identical, there are tremendous differences.
§4. The Lennard-Jones potential and the limits of properties [12:51]
There are limits to these properties, and everything's related to the bonding — the Lennard-Jones potential, the energy potential for bonding. So if this is the zero of energy, this is positive energy and negative energy of bonding. Take an atom right here, another atom out here at some distance — the potential energy of bonding looks like this. The equilibrium distance of separation is here. Out at infinity there's very little attraction. There's only attraction within three or four atomic distances. Ten atomic distances away and those atoms can hardly see each other.
What's the distance between molecules in the air in this room? The difference in density between a vapor and its condensed liquid or solid is about a factor of a thousand at room temperature. You can prove all this from the ideal gas law and looking at the density of the solids. If I take the cube root of a thousand, I get a distance of about ten. So there's ten molecular distances between each molecule in this room. There's no significant attraction between the molecules. It acts almost like an ideal gas because they're way out here.
The depth and sharpness of the bond, and the closeness of it, makes a stronger bond. Silicon-oxygen and carbon are much stronger than hydrogen-carbon. Your mechanical properties are actually determined by the curvature at the bottom of that potential well. Your melting temperatures are determined by that and by the number of electrons involved in the bond. What's the highest melting element on earth?
Student: Tungsten.
Very good. On the periodic table I've highlighted in yellow the four elements that have a melting point above 3000 Kelvin. As you go down the periodic table, you go from 2200 to 2900 from aluminum, but to 3700 for tungsten. Seaborgium, if we actually had enough to melt, would probably have a melting temperature higher than tungsten — that's what the periodic table tells us. Even if we can't measure it, we can predict it. Why are these the highest melting metals? Because they have the most bonds — the most f-shell electrons participating in the bond. It's not just the depth of the potential well, it's also how many electron-electron interactions are coming together to hold the thing together. Melting temperature is a function of bond strength, which is a function of the number of electrons participating.
§5. Measurement and matter: how energy can change what it measures [16:48]
When we're talking about measurement, the very first day I started out saying: if you're going to measure something, you're talking about energy interacting with matter. There are various types of energy — electromagnetic, mechanical, chemical, thermal, nuclear. You can also get combinations of these. Piezoelectric is basically mechanical and electric. Thermochemical. There are various ways to combine these, but nonetheless, there are only certain types of energy.
The whole science of thermodynamics that you've learned is basically trying to quantify these energies. The highest strength chemical bonds are just a little over three electron volts for carbon or silicon-oxygen. You've got other bonds that are down around one electron volt. These are the energy for the primary bonds. You have secondary bonds that are ten times weaker. So when you're trying to measure something, you have to worry about whether you're going to do something to change the material. Remember Heisenberg said: you try to measure the position of something and you hit it with a photon, and if you're talking about quantum mechanical things, that photon is going to move the thing so the position changes — and that leads to the Heisenberg uncertainty principle.
If I use electromagnetic energy to try to measure the properties of something, I could be altering the properties. One example: if you walk out in the sun and stay there for a while, you'll get a suntan. What are you doing, and why is it only in the sun and not indoors? If the sun was shining through the windows, you wouldn't get a suntan through the windows. It's the UV reaction with your skin. The energy of ultraviolet radiation is on the order of a couple of electron volts. It's sufficient to break primary bonds in your skin, and that's what causes the suntan. You can't get a suntan through the window because the window blocks the UV. Visible light is only about one and a half electron volts or less, so it won't break down the bonds.
Ultraviolet radiation breaks bonds. If you're trying to measure with ultraviolet, you're liable to change the material by breaking primary bonds. With visible light, you could be degrading polymers with their secondary bonds. If you get high enough energies, you actually can start nuclear reactions. At about five million volts, with electron beams, you'll start transmuting elements to radioactive isotopes — you start knocking protons or neutrons out of the nucleus. It takes a lot of energy, but all of these things can interact depending on the limits.
I'm not going to go through all the types of the electromagnetic spectrum, but here's a little summary of what happens in different energy ranges. The main interaction with matter for radio waves is essentially that you can cause electrons to move around — that's like an antenna. You can move electrons in a metal if you have free electrons. There's not a lot of energy in radio waves. Remember what's the energy for electromagnetic radiation? Einstein won the Nobel Prize for this. E equals h nu. He didn't win it for relativity — he won it for the photoelectric effect. Energy equals Planck's constant times the frequency. You learned that in freshman physics. You haven't forgotten all your freshman physics yet, have you?
§6. Ashby plots and the limits of properties [25:09]
So these are the properties. The best way to describe the limits is with Ashby. Who is Mike Ashby? He was from Cambridge University in Great Britain, and he was Lorna Gibson's professor. He's recipient of the Acta Materialia gold medal — one of the great material scientists of the 20th century, at Cambridge now. Ashby plots came out in the early eighties — plotting two properties on log scales. He plotted strength versus modulus, and engineering materials fell into surprising patterns. Reinforced laminates, woods like ash and oak, porous ceramics. All your other woods — ash and oak — make handles. About five orders of magnitude across, but within each family, only one or two orders of magnitude, because each one has a particular type of chemical bond that only varies depending on orientation, like diamond and graphite — electron volts per bond.
Look at some other properties — very good correlation. Thermal conductivity lies on a 45 degree line. You start out with diamond, which has the highest thermal conductivity of anything. It's got the strongest bond, the highest hardness — that leads to mechanical vibration, which leads to very good thermal conductivity. What's the difference between thermal conductivity and thermal diffusivity? Anybody remember? Thermal conductivity divided by heat capacity equals thermal diffusivity. They're correlated. Polymer foams have a thermal conductivity an order of magnitude lower than anything else — the lowest in the world, because they're not solid.
[Tom produces a piece of charred polyurethane foam.] You can buy this in a little spray can. They use this as spray foam insulation. It's the best material — light, nice and white. A little white cube of this, I could stick it in the furnace and it gets darker, toasted like a marshmallow at 500 degrees, and it actually starts to burn. They used this as spray foam in the 1970s. It was all over the place, because it's the best insulator. When you just lay the batting in there it expands to fill the eaves, and it's such a good insulator.
I know a guy at Woods Hole — he was rebuilding his house and his house burned down last year because of the way it was installed. This is the fire — you can see it's charred. They were squirting it along the eaves, and they didn't pay enough attention. You can see the surface is a little darker on this side, and you can see the internal layer heating up. With a thermocouple down in here, 100 degrees. You can put something very hot here and you won't feel it through — it's a good insulator. Faster — there's data on this — more oxygen on the surface, the polymer actually forms carbon, and it starts to overheat. If the heat reaches the oxygen, it'll burn down your house. So it's safe when installed with instructions, except when the omission gives the right instructions, sorry.
The point is that there are limits. Modulus versus thermal expansion has a little curvature to it. This is from material selection in mechanics — a whole Ashby plot. Then around 2000, Ashby became a very wealthy man. He had lots of materials properties data, and he started a company, Granta. Instead of looking at things one plot at a time, you can look at multiple properties at once if you pay for his computer program. Professor Gibson and the department use it — a thousand dollars a year. You can specify: won't melt until a certain temperature, has this modulus, this property — and it will show you possible combinations. They don't always tell you about toxicity — you put in beryllium and it's toxic. They don't always tell you the cost either. So you still have to have someone with a job interpreting what comes out of Granta. There are ways to look at the limits of what products can do.
§7. Concorde and the limits of practical engineering [31:07]
One of the videos I'd like to talk about — the Concorde in the mid-eighties, and the SST built around 1980, the supersonic transport. Instead of a 747 at about 40,000 to 45,000 feet, going out into the edge of space where there's no air, if you can get to Mach 16 you could go up and out and back down. That was the SST plan. They said for a commercial — the British supersonic transport from the 1980s showed the capability of how they could build something. The one in the United States — we actually looked at the Concorde, with the French and the British. Anybody know what happened? A plane hit some junk on the tarmac, on the Concorde's wheels on takeoff.
They only had a few of them, and they were expensive. I flew the Concorde. I had to come back on a Friday evening. I would pay the extra $2500 — it was first class, everybody wanted to be on it. They only had full time service, just flying across with maintenance on the ends. The seat I booked was 1B, six weeks ahead of time. I was in business class, the company paid for business class — that was for $2400. Early 1990s, twenty years ago, $400 extra out of my pocket, but I'd never flown an SST. We left London Court at five o'clock and arrived the same day on the local clock. Most of the time you're in the boost and slow-down phase, and because you're going so fast across the Atlantic — you could only go supersonic when you were over the Atlantic, so there were only certain routes. They made it uneconomical between New York and basically London. You could take the route via Bermuda. People complained about breaking the sound barrier. They had a little display that told us we were doing Mach 2.3 or so.
People who fly first class make their reservations ahead of time. There was a king — a tycoon — who'd fly British Airways, and British Airways would hold a seat for him on the Friday flight at noon, London to New York, and he'd come back on the two o'clock flight to London over the weekend. Probably money going into illegal bank accounts, a different world. He told me his schedule — New York, then Cairo, somewhere in Delaware, then Bangkok, and he'd be back. He was going around the globe that week. If you have a lot of money, you fly all the time. The other thing was — the tray table for dinner was broken, so he had to hold it with one hand and eat. As a passenger on this airplane, you don't have time for maintenance. Same point with the helicopter tour outfits — you've got to have time to do maintenance on the aircraft. Once the helicopter comes, you shouldn't be flying anymore.
§8. Codes and standards — ASTM and the Alberta refinery dispute [45:29]
Codes and standards. I put up one society's definition. I'm going to hand out the Ben Franklin chapter on lightning protection systems. Here's an article, a chapter out of a welding handbook. People who weld worry about codes — for you and your companies. The American Welding Society is one organization that produces standards. They include specifications, classifications — there are lots of things that fall under standards. Some of these have force of law, some don't.
ASTM — American Society for Testing and Materials. Headquarters in Philadelphia. For almost the last hundred years they've published the annual book of standards. The irons and steels volume cost me a thousand dollars years ago — they've gone up dramatically. The annual book has grown five to tenfold over the last hundred years. I want you to take an ASTM standard and go through it: who's doing it, what the scope is. Look at the one you've got — A106 is for seamless carbon steel pipe.
The pipe in your basement, if it's seamless, not welded pipe — it does look like it's seamless, I don't see a weld going around it — that's ASTM A106. Welded pipe is A53. Plate steel — A36 is what civil engineers use, and there's a higher strength steel A441. Has a set of codes — in hard copy they would take up about twenty feet. For a material specification you might spend a dozen pages, but for polyurethane foam, on how to make it, how to evaluate it — it's much longer. It's a put-together by industry and government, basically. At the bottom here — specifications under jurisdiction of the steel and stainless steel committee, current July 10th. They tell you what day. We'll get into arguments — if I say July 11th, what about the prior code? They tell you what was previously published. The seamless piece edition was 1990. It tells you the scope: from this diameter to 48 inches, from this wall thickness up. If there are some requirements or limits, it tells you what other code to look at.
Some of these are American National Standards. The Department of Energy, the nuclear engineering code — the Nuclear Regulatory Commission will adopt ASTM standards. If you want to build a nuclear reactor, you go to their standards, which incorporate ASTM standards. ASTM, fifteen years or so ago, basically said: people will know how to measure properties, so we can call out a standard and know exactly what we're getting. These things get revised by committee. It greatly simplifies — if I want to build a high rise in downtown Boston and I need steel that has certain properties, I can get cheap foreign steel and check it against ASTM.
The reason I'm giving this background of what we're going to go over tomorrow is because I was involved years ago in a case — a new refinery up in Alberta. There was a big recession at the time, the economy was in session, and they were doing all kinds of things to get steel pipe. The company building this new refinery — I don't know, $500 million — said: we'll find some pipe, we can't find it, we'll go out, they specialize in stuff. They got ten, twelve inch, twenty inch pipe, and were fabricating it and making a refinery. When they did the hydrostatic test — putting water in, pressurizing it — a crack appeared. They brought us in. They condemned the pipe based on this crack. They said: we want our six million dollars back, or two million dollars back, because we're going to yank it all out, and that part of our refinery is built, and we're going to charge you for putting the new stuff in. There's a big fight, which I'll get into tomorrow. The fight arose because people don't know how to read a standard in plain English.