CS_Su2012_01

On codes and standards, which comes out of the American Welding Society handbook, and it's got a reasonable definition. And I call them AWS [American Welding Society] definitions because they're not the definitions that the whole world uses, because there is no common set of definitions. They use standards in a generic sense to include code, specifications, recommended practices, classifications, methods, and guides. And I'm sure, since all of you have now worked in shipyards or on board ship or in some engineering duty, you've come across some of these things.

And as it says in the little handout I just passed out, um, these go all the way from merely being guides. Remember Pirates of the Caribbean, they're not really rules, they're sort of guidelines, okay, you don't have to do them. Other things will have the words "shall" and "will," and so you have to do them because it's in the contract document. Other things may have "shall" and "will" and you have to do them because some government body has adopted them as a matter of law, okay. If you build a pressure vessel in the United States, every one of the 50 states has adopted some form of the boiler and pressure vessel code. Anybody know what the boiler and pressure vessel code is? Was that a no, you're just— okay.

The boiler and pressure vessel code is sort of, I call it the granddaddy code. It's about 105 years old. It started because of a boiler in Brockton, Massachusetts, which blew up. And Mike's back there, he's from Brockton, but this was before Mike's time. Um, and I actually have a picture of it somewhere, but there's a nice picture of a mill building, that whole city block size, about five stories tall — that's the before picture. The steam boiler in the basement blew up, and the after picture is just level, like, you know, it was like a firestorm in Germany after World War II, just nothing left. And that sort of irritated a few people, and so the American Society of Mechanical Engineers formed the boiler and pressure vessel code.

And if you want to look at this, you're welcome to, but this is one out of about 20 volumes of the ASME. [Tom holds up a volume of the ASME boiler and pressure vessel code.] This is the 2004 edition, which is one of the ones I have on my shelf. But the code itself probably cost around $15,000. It's renewed every three years. If you buy it in one of the off years, um, you will get an addendum. So this was purchased in 2006, and there's a pink and then a blue addendum. The addenda are fairly thick. This is just on welding and brazing specifications. The whole code takes up about six— four to six feet of shelf space, okay.

If you're going to build a pressure vessel that could blow up and level the factory, you have to build it to some portion of the code. One part of the code, chapter three— um, these are Roman numerals— um, section three of the code is for how to build a nuclear pressure vessel. Section eight is in division one, division two, and now a division three. And division one is rules for designing a regular pressure vessel that's not a nuclear vessel. Section two is guidelines for designing it yourself. You can design it yourself without following the rules, but you've got to do lots of calculations. And section three, I think, is for vessels above 15,000 psi pressure. I mean it's extreme pressure, okay. And then there's a section on inspection, uh, non-destructive testing, welding, materials. I mean there's a whole shelf of these things. And it's a good profit-making thing for the American Society of Mechanical Engineers, but it also is the good practice that's been found as a result of many failures over the last 105 years, okay.

It turns out the first research contract ever given by the federal government was in the 1830s, because a steam boiler on the Ohio River, on one of the steamboats on the Ohio River, blew up and killed a hundred people or something. And so Congress gave a grant to the Franklin Institute in Philadelphia to try to figure out why the boilers were blowing up. They still blow up, okay, but not as often, okay. And, well, actually sometimes the consequences are more catastrophic than they used to be.

So anyway, um, I decided that I wanted to do something on codes and standards because frankly I didn't know anyone else who ever talked about codes and standards very much in universities. They talk to you about how to design things, and you guys are here because, uh, I think— I don't know for sure, but I think one of the reasons NAVSEA picks ten people and sends them to MIT is because an inordinately large number of graduates of the 2N program end up being the chief designer for the next ship for the US Navy.

So I had students who ended up being the cap— when they became captain, were head of the program for the DDG 51, and then what's the one that just got, uh, the Zumwalt class or whatever, that just got cancelled. Um, Millard Firebaugh [Firebaugh] was before my time, but he was, um, he was a 13A, which was before 2N, and he got a PhD here. He was in charge of designing the SSN 21, the Seawolf, okay. And I knew him when he was a captain designing the Seawolf. I knew him when Congress zeroed the budget for the Seawolf because the Soviets had developed a titanium submarine and Congress was, oh, they've leapfrogged us. Um, and uh, I knew him when he went back up to the Hill to get the money back. Uh, I knew him when he became chief engineer of the Navy. And I was hired by him when they found, when they completed 18% of the ship and had to tear it all apart in the early '90s, uh, along with a few other people to look over Electric Boat's shoulder and make sure they didn't make the same mistake the next time. Um, and then, uh, anyway he's still alive, he's retired obviously now, but anyway.

Um, but some of you, one of you is liable to be designing the next whatever it is. I don't, I guess it's probably going to be a littoral ship or something, it's probably going to be made out of aluminum. The materials of interest to you are going to change or have changed over the years, okay, so we'll talk about some of those things.

Um, I did— Kathleen gave out this MIT Faculty Newsletter article which I wrote back when they were looking for the previous president of MIT. Um, I actually wrote it so the Provost at the time, who had been a department head when I was the department head, and then he became Dean, and then he became Provost — I wanted to make sure that he did not become president of MIT. And he didn't, he's now president of Boston University. Um, but you'll see a quote in there from him, uh, and if you're careful and think about the quote, you'll realize why we didn't want him to be president of MIT. Um, but anyway.

Okay, so we're going to talk about codes and standards, but today I want to get into, um, not codes and standards but what— when we're dealing with codes and standards, it's sort of an agreement between the purchaser of something and the, uh, seller of what it is you're going to buy, okay. If you go to buy a house, you want so many bedrooms, you want, you know, such and such in the kitchen, you want such and such in the garage, and you have specifications, uh, for what you want.

Um, but if you do this, um, you also may specify things in a little more detail about the quality of what you want, so that you don't get, um, uh— if you don't want, if you want real plywood as opposed to oriented strand board in your house, you're going to have to specify it. If you want a certain quality of blueboard gypsum as opposed to cheaper gypsum, you got to specify it. Uh, otherwise people are free to come in and use whatever they want. So if you want to maintain quality you have to specify it, and you also have to be able to measure it.

And as Lord Kelvin said, I often say that— "When you can measure what you're speaking about and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind. It may be the beginning of knowledge but you have scarcely in your thoughts advanced to the stage of science, whatever the matter may be." So part of your MIT education, um, is to learn to be quantitative in your opinions. And believe me, if you can be quantitative in your opinions, you will win many arguments, okay. Because most people don't know how to be quantitative, okay. They just say, well I think— well, if it was, if building a ship or any other product is a faith-based exercise, then what you believe would be wonderful. But in fact, if you want to have some sort of agreement and not have lots of fights down the line, you need to specify things. And so you need to learn how to specify things, you need to learn, as Lord Kelvin says, um, how to measure things and quantify things.

So what I want to talk about today is what I call observables, okay. What can we observe, uh, what can we measure. And it turns out, whatever you observe is ultimately— and this is sort of, um, since I— only one of you is an MIT grad, right? And frankly, it's— oh, you're a MIT grad too, right? Okay, you're Air Force officer, right, okay. Um, and one of— is the Coast Guard person here yet? Not yet, okay. Um, uh, so we have one Air Force and one Coast Guard person, but a lot of what I'll talk about, I'll try to relate it back to Navy-type things.

But in any case, um, one of the things you'll learn is, um, one of the philosophies at MIT is to try to go back to the fundamentals. And you'll, in this Faculty Newsletter thing I'll talk about, you'll see I talk about what makes MIT unique. I've been here since 1968 as a freshman, all except three years. And, um, MIT is different than most other places. The only place that I can liken it to is Caltech, okay. And you can read the article and see why. Uh, except for the two of you that have been here before, uh, you're going to learn what it means to take a drink from a fire hose, okay. Um, the Navy too, you do that in the Navy too? Well, okay, you might find it's still a little different, okay.

Um, uh, but in any case, you know— well, you can read the article. And part of what I'm supposed to do, uh, with this class is help acclimate you to MIT, okay. Because you're going to, in your math course you're going to get the traditional MIT, okay, how to solve this differential equation and this, that, and the other thing, but there's more to life than just that. But part of it is going back and saying what are the fundamentals. And if you're going to measure anything, it's actually, at the most basic level, some form of energy interacting with matter. And since I'm a material scientist, we can talk about that.

Does anybody remember what the Heisenberg uncertainty principle is? Yeah, what it is? The more you know about the velocity of a particle, the less you know about its position, right. Or you can talk about momentum and time and things, uh, but, um, this is momentum or momentum and position. Um, whatever you know about one, or the change in one, or the uncertainty in one, times the uncertainty of the other, has to be greater than Planck's constant over two. Well, we're not usually dealing with things quite that, uh, precise, and it really usually only gets into, um, principles in quantum mechanics. But I want to talk a little bit about how precise we can be.

But I also want to talk— well, does anybody know the difference between precision and accuracy? Yeah. Precision is expressing something to a degree of certainty; accurate is saying something that is true but not necessarily precise. Right, okay. So it turns out that, um, that's a good definition. Reference is the actual value of what it should be, okay. I mean, whether it's true or not, we'll leave truth to the philosophers, okay. But the reference value is what you specified, okay, in your specifications or your classification. You specified something, and the precision is how well you can measure it and with what, uh, reproducibility. So they got a Gaussian curve here, so you might measure some value, and how accurate is it to the value that is the correct value, okay, or a reference value as it calls it here. So there is a difference between, uh, accuracy and precision. Precision is how precisely can you measure something. Accuracy is how closely does it conform to what it should be, whether that value is the truth or not.

It turns out— fundamental, uh, quantities that everything else can be reduced down to, you learned about three of them in your freshman physics. So whatever units you know, f is equal to ma, force is equal to— you can separate it into, force is kilogram times meters squared per second squared, or m/s squared, or whatever. Acceleration is m/s squared, right. So force is kilogram·m/s squared. Mass, length, and time, right. Remember that, and you did the dimensional analysis and the units had to cancel on both sides of the equation and stuff.

Well, it turns out there's actually seven things. Um, so this is an article that I just got, um, in March was when I first saw it. I was down at the National Institute of Standards and Technology, 'cause I was on the review committee for manufacturing. And that review committee, or that group that is doing manufacturing, now includes the group that takes care of the reference standards. We're going to talk about standards, and so for a second here I'm going to talk about reference standards. So if the whole world wants to agree on some standard, and we're going to— I'm going to buy something that's going to be one inch long, okay, well how precise is that going to be? It's going to be one inch plus or minus one-thousandth, plus or minus 50 millionths, okay. How precise does it have to be to one inch, and under what conditions, okay?

Can anyone think of, when you took high school chemistry you learned about, um, the volume of an ideal gas, right, and it was referenced to a standard state, which was STP, which was 298 Kelvin and one atmosphere pressure, okay, standard temperature and pressure. That's a standard, okay. If we're going to talk about standards, that's a certain type of standard. And in fact, um, uh, this group at NIST outside of Gaithersburg, Maryland, and out in Colorado, takes care of developing reference standards not just for the United States, but for the whole world. And that article I just gave you comes from a group, couple of guys at, uh, NIST— the NIST is what used to be called the National Bureau of Standards. It's National— now National Institute of Standards and Technology. And, um, this is an article written by a person, a similar person at a group in the United Kingdom, and the other person is from Sèvres, France.

Um, why Sèvres, France, anybody know? We got mass, length, and time. Do you know how we define time today? Yeah, I was thinking there was some standard. There is. Well, there once was a standard meter, there was a stick, um, in Sèvres, France. But they got rid of that. When they first started doing the standards, they had a standard meter and they had a standard kilogram. And, um, they had clocks that they were trying to keep accurate. What replaced the clocks, anybody know? It was just after World War II. It was right up here at the Harvard Smithsonian Observatory. If you go up Concord Ave towards Belmont, you'll see— you'll pass the Smithsonian— uh, Harvard Smithsonian Observatory. And there they developed the atomic clock, okay.

And the atomic clock— hey, I've got one in my office, I got one in my bedroom, I got one in my kitchen. You can buy them— [you] work on your rear watch. What else— most of you in this room have an atomic clock right now on your person. Quartz or— well, quartz and cesium. Well, cesium is correct, okay. That atomic clock is based on cesium. Before that we had quartz, and we were looking at the vibration of quartz, and that was— I don't know how many megahertz— and so we could separate time into millionths of a second with a quartz crystal oscillator. The cesium atom vibrates at 10 gigahertz, and so we can separate time into 10 to the 11, ten billionth of a second, okay. So the atomic clock, if I'm looking at mass, length, and time— the atomic clock for time, um, cesium ion, or Cs, not cerium, cesium— is 10 gigahertz, and that means it's one part in 10 to the 10th seconds is the accuracy of our measuring time, because of the research that was done up the road here in Cambridge.

Um, and now they beam a signal from, uh, Colorado Springs or Boulder, Colorado, from the National Institute of Standards and Technology. But how many of you have cell phones? You have an atomic— you have, you are using the GPS signal from atomic clocks. Every one of those satellites has its own GPS, has its own atomic clock. And an atomic clock — the one in Boulder, Colorado, which keeps the reference time for the United States and the world in general — is about the size of this room and costs millions of dollars. Well, they've condensed it down, so they— you now can get one that's not quite as precise, it's the size of a grain of rice, okay, with a little oscillating cesium atom. And if you can find something that, uh, reproducibly and reliably gives you a higher frequency than 10 gigahertz, we can get one better than one part in 10 to the 10th. But that's our limit on time.

Our limit on length used to be a standard meter stick in Sèvres, France. What is it? Light. And what they want to do in trying to measure things precisely, so that everyone can agree on everything, is they want to come up with something that— and in fact they defined it for us in March, they said we want to have something that, if you have enough money, anyone in the world could do the experiment to measure the property. So with the right equipment, people can measure the speed of light. And so now you can— the meter is now defined in terms of the speed of light, okay.

And so we have— what about mass? Mass is still defined by this platinum-iridium weight locked in a vault in Sèvres, France, and it is the international standard for mass. And so anyone who wants to measure it has to go to Sèvres, France. It's only taken out of the vault about once every ten years, and so far as they can tell statistically, it's losing 60 micrograms per century, okay. So if you can delay delivery on some product that you sell by the pound, you can— you know, you can make a little extra money, right.

Well, it does get to be a problem, um, in some of these things. And this little handout I gave you has some of the fundamental constants. I don't know if you're going to be able to read that, but you've got a, uh, handout— this is on page 230, okay. You don't have to really worry too much about this, but, um, some of the recommended values for Planck's constant, the, um, the electric charge, uh, Boltzmann's constant, the Avogadro's number, and these are the accuracy to which we know these things. One part in 10 to the 7th, uh, eight parts in 10 to the 8th, anyway. So, uh, um, that's our accuracy. These are the actual numbers, and you can see how many digits they've got, and the last two digits they've got in parentheses because that's the experimental error in making these measurements.

And this paper explains in the abstract in the beginning, there are only seven fundamental, uh, quantities that you really have to worry about, and those happen to be — if you're interested, which you, um — so it's a length, mass, time, electric current, thermodynamic temperature, uh, luminous intensity, and amount of substance, which is a mole, Avogadro's number. If you can define those things, most other physical— most other things, most other physical properties of materials and stuff can be defined, and they can be defined very precisely.

And you say, well, who cares, I mean do I really care about measuring such things so precisely? Well no, neither do I, um, but there are some people who do, okay. Um, here's an article by a guy who works for Fluke. And Fluke — anybody ever heard of Fluke before? Fluke makes, uh, meters for, you know, voltmeters and ammeters. They also make very precise reference standards. And if you get on the Fluke website and register and stuff, I had to register to get this. "The impact of new SI on industry" — he's talking about a change in the way they defined the volt in 1991, and it turns out it made a difference in the sixth decimal place of what a volt is. And all of a sudden Fluke's standard, uh, reference cell, which they sold for tens of thousands of dollars, uh, was— it was supposed to be accurate to so many parts in a year, and it was no longer accurate because they redefined what the standard was. And so they had to send a notice out to everybody.

And about things— this is from his paper, of some of the numbers that we just looked at. But here are the, uh, the seven— yeah, the seven quantities. Cesium atom, that's time. Speed of light in a vacuum is this many meters per second. It is now defined as that. So now if you can measure the speed of light, you can measure the length of a meter, because it's one over 299,792,458 of that speed, okay. It is defined. The constant— these fundamental constants of nature, and there's only seven of them, um, are defined in terms of certain quantities. This is the frequency of the cesium atom. Uh, and there you've got, um, 10 significant figures here, you've got nine significant figures. The Planck's constant is exactly— well, actually these last four, they're not exact because they got the X on the end. The luminous intensity of monochromatic radiation at 540 hertz— yeah, 540 times 10 to the 12th hertz— is 683 lumens. That's about a 60-watt light bulb, okay, whatever all that means.

Um, but you can define these things. And if you're dealing in parts per million, you have to worry about these things. But who's dealing in parts per million? Well, some people are doing better than that. There is— the most precise measurement ever made is being attempted in Livingston, Louisiana, and Hanford, Washington. Anyone from either the state of Washington or Louisiana have ever heard about this? It's LIGO, the Laser Interferometer Gravitational-Wave Observatory.

Okay, and the LIGO project originally was funded at the tune of $300 million by the National Science Foundation. It was a joint project between MIT and Caltech. The chief scientist of LIGO is Ray Weiss, who's a retired professor of physics here. But it's two Michelson interferometers. Anybody know what a Michelson interferometer is? You didn't learn this in freshman physics? You did— maybe you did, let's see if you can recall it. Anyway, well, they're going to try to measure gravitational waves with this. But a Michelson interferometer — you can have a bench-type interferometer, but basically you have light coming in, you split it with a mirror. One of them goes down this length, uh, typically you have them at 90° to one another, and you have some length one and length two, and you have a mirror at this end, so the same beam of light comes in, is split into two, reflects off the mirror, and they come back. And as the wavelengths beat together, you can measure the phase angle difference between them, and you can tell the difference between L1 and L2 very precisely, because now you're measuring the shift in phase angle between the two waves. It's the same beam of light, and that's a constant, so far as we know, okay. The speed of light, we think, is a constant in the universe, um, so far as we know anyway. And you can measure these two lengths very precisely because now it's a fraction of a wavelength of light, because it's the beating frequency, it's the phase angle, right. So Michelson interferometer is for light, measuring light is a very precise way to measure length, okay. Or the speed of light. If you know the length, you can measure the speed; if you know the speed, you can measure the length.

Okay, well LIGO, Laser Interferometer Gravitational-Wave Observatory, they have one of these in Livingston, Louisiana, and they have another one in Hanford, Washington, a couple thousand miles apart. And the length is 4 kilometers of each of the two things. And back here at the beginning they have a tungsten sphere which is your mass, and I can't remember what it weighs, like 12-inch diameter big spherical piece of tungsten hanging from a pendulum with a mirror on it. And they have high-power lasers which shoot— split the beam of light, they go down one end to the other to another mirror and come back, and so it's basically a micrometer to measure the length.

Now because the density of air can change the speed of light— [the speed of] light is defined in a vacuum. So they had to build stainless steel tubes 4 km long and pull a vacuum on it at about one, uh, one-trillionth or one-quadrillionth of an atmosphere. And the only thing you're pumping there, that's left, is hydrogen. And they had to develop special treatments for the surface of the stainless steel vessel so it wouldn't outgas hydrogen, okay. There are all kinds of interesting problems. I got involved in this about 15 years ago when they were designing it, because they wanted to know how to weld the stainless steel because of the precision they were looking for.

In order to run this thing, they had to align the mirror here with the mirror here within 1 cm over 4 kilometers. That's one part in— let's see, it's, uh, one part in 4 million, right? I think that's right. Four— no, in 400,000. One centimeter is one part in 400,000, okay. That's pretty precise. No one had ever surveyed— needed to survey 1 cm over 4 km before. They had to develop new surveying techniques using GPS. And in fact, anybody know what wave GPS is? I don't even know what it is, but I've heard of it. But wave GPS is, if you end up building a little ground transmitter right here to help synchronize all the satellites up there in space, you can get about a factor of 10 or 100 more accurate using wave GPS. You use the satellites up there in space, but you have a reference point here that helps calibrate all of them to a reference position. So they used wave GPS to do the surveying, the civil engineering surveying, because no civil engineers had ever done anything this precise before, okay.

Now, so far as precision— let's see, Kathleen, you were in Aero and Astro, right? You know John Hansman, Professor Hansman? There you go, you didn't know him. Okay, John was developing a GPS with wave GPS— the only reason I know about this, he was looking at race cars, and he wanted to see when a race car might start to fishtail. And so he was using GPS to monitor the movement of the tail end of the car, to see if it moved one or two inches off to the side, so that he could control the speed and brakes of the wheels to prevent a spin out, okay. Now that's pretty precise, okay, if you like to race I guess. Okay, um, to be able to use things like that.

So there are things you can do— who's doing this? Okay, um, in the LIGO project, they have one of these interferometers in, uh, in Livingston, Louisiana, and the other one Hanford, Washington. The Australians are now talking about building one, and then they'll have them 10,000 miles apart as opposed to 2,000 miles apart, because they can up their accuracy by doing that. Some people proposed building one on Mars, okay. And what they do is they build two of these things, and they're going to check the beating frequency of that movement of those tungsten spheres 2,000 miles apart, and make sure it's not just noise in the system, because you're trying to measure a shifting, and the amount of shift they're trying to measure from those spheres— this is all to confirm Einstein's theory of general relativity, or special, general relativity. They're trying to measure 10 to the -18 centimeters.

What's 10 to the -18 cm? What's the size of an atom? Size of an atom is 10 to the -10 m, so 10 to the -8 cm is the size of an atom. This is 10 orders of magnitude smaller than an atom. It's much smaller than a nucleus. They're trying to measure a shift that is orders of magnitude smaller than the size of a nucleus. The accuracy of the whole thing is supposed to be one part in 10 to the 22nd, and then they have to divide by frequency, because when you start getting these types of precision, you have to say, well, what's my sampling time? Okay, 'cause you're averaging these measurements over some time. The longer the time you average, the more your error— or the less your error, but the more noise, right. So there's a signal-to-noise ratio problem if you really want to go through it.

Here's— I asked Ray once, I was starting to think about this. Here is his notes when he explained it to me, okay. Not a lot of notes here, but basically, it's the most precise measurement anyone's ever had to make.

Now, um, there are other things. For temperature, for example, anybody know what— well I just put it up, uh, what's the triple point of water or triple point of any pure substance? You might know. Where the all three phases exist, right? The triple point is where solid, liquid, and vapor coexist. It is an invariant point thermodynamically — if you believe thermodynamics, and who even understands it, so how can you believe it? Um, but if I look at the triple point of water, there's pressure and there's temperature, and I have something that looks like this where this is vapor, liquid, and solid, and there's a triple point. And according to Gibbs, the phase rule says that's an invariant point. And if I can come up with some way to get solid, liquid, and vapor coexist, that temperature is uniquely defined. And in fact the temperature scale is defined as the triple point of water is 273.16 degrees Kelvin— or yeah, that's the triple point of water. What's the melting point of ice at, um, one atmosphere? It's 273.15. The triple point varies from the one-atmosphere pressure by a hundredth of a degree, okay.

But in fact, for $1,500 you can go and buy from Fluke these cells. [Tom holds up a triple-point-of-water cell.] And now these cells were probably developed by some national laboratory somewhere, but it's nothing more than either some borosilicate or quartz tubes — the quartz is more expensive — filled with water. But it's not just any water. This is multiple distilled seawater, such as super pure water — they've gotten rid of all the salts and stuff — because this water has to have the same isotopic concentration as the Vienna mean sea water, okay. People have measured lake water, rain water, ocean water, and they all have a different isotopic concentration of oxygen-16, oxygen-17, oxygen-18, hydrogen, deuterium, tritium, okay. And those will all change the triple point depending on your isotopic concentration. Because these things have accuracies of— oops, this is really getting bad, uh, let me close this down a little bit, the sun is coming out is one of my problems. Not that you have to remember any of this stuff, but if you just want to— oops, what happened here? Well let me just— I'll read it to you.

It turns out, um, the uncertainty is less than one millionth of a degree, or one ten-thousandth of a degree centigrade, um, for these things, okay. So if you want to calibrate your thermometers, you can buy these little special cells of water. You can also buy them in gallium, ultra pure copper, and you can do a whole set of temperature scales if you want to be accurate within a fraction of a Kelvin, okay. Now, we don't usually need that, um, we don't need it for, uh, in most cases. But some people do, um, when they're trying to make very precise measurements such as the fundamental constants.

Um, most of the time we use much simpler things. And so now I need you to break into— let's, one, two, three, okay, um— we need four groups. On these, three be a group, those three be a group, and then break into two other groups along here. I'm going to give you— this is sort of like high school physics, okay— I'm going to give you a bag of spheres. [Tom distributes bags of sample spheres to four groups.] Okay, here for each group, here are some things so you don't spill them all over the floor. Okay, uh, keep them— um, so you need to give one of those back to the back. I'm going to give you a sheet that tells you there are— not 19 things in there. This is one of these open-ended problems, um, in the sense that, uh, the different things here, there's a brass alloy, and I want you— you can take them all out and put them in the— what's going on here?

Uh, I can't get— I'll figure this, huh, the brightness— well, the sun came out as part of it. Let me darken the room again. Um, see if I can get a back. Anyway, there are potentially 19 different materials there, okay. And you're supposed to figure out how many. So if you can figure out— for example, I bet most of you can figure out which one is the brass quarter-inch sphere, okay. So put it aside and say you found it. And then there's some white ones, and there's some steel-colored ones, and there's a black one. See how many you can figure out. So we'll do a competition, and whoever wins can have another bagel, okay. I don't think you're going to be able to get them all, at least initially, or even in the— I don't think— could be— you're going to grade yourselves on this, I don't, I'm not going to try to grade you. And if you get a good grade that's good, and if you get a bad grade that's good too, so— you might want to put the ones you figured out in the, uh, in the— bounce. Actually that's not a bad technique if you have the data to go with it, okay. If you don't know what it means, then it doesn't help, but it's actually a good technique.

I'm only going to give you 60 seconds more, and then I'm going to give you something to help you find something. Okay, I'm going to give everybody a magnet to see if that— [Tom hands out magnets.] I'm not able to estimate the weight. Well, if you had a scale you could do a better job, right. Okay, you'd be more— this one is barely— so there's on these, if there's like— so is it the carbon and the steel?

Okay, time's up. You can play with them some more later. Let's go over it and see. Well, what was the first thing— what was the first property of material that you used? Color, exactly. And how did you pick out— what did you pick out by color? Brass, yeah, okay. Aluminum — some of you picked out aluminum by color, but you confirmed it was aluminum by what else, what other property? Wait, mass, okay, it's light. And some of you recognized the color of aluminum, okay, it's aluminum color, right.

Um, and then you ran into some problems, you had all these steels, right? We'll talk about that, you got five different steels. There was another thing you picked out, eight or nine, you picked out because it was rubber, right? But you didn't know which kind of rubber it was. Well, one of them doesn't bounce — you have a black one that doesn't bounce and is also like— okay, that's not very rubbery, is it? Hard? Uh, no, it's pretty— okay, then it probably is one of the rubbers. But hardness is a property. So we have color, we have density or weight, we have hardness.

Did— was anyone able to sort out, at least sort out, some of all these different steels by magnetism? Three of them are sort of the same level magnet, okay. Which magnet you have— the yellow— you have the yellow magnet, right, okay. Uh, you have a theory: the number indicates carbon content, and we think that more carbon is less magnetic. Uh, not— it's not carbon, it's crystal structure, okay. 1020 steel is body-centered cubic, it's ferromagnetic. 302 and 316 are not supposed to be magnetic, but 302 will be partially magnetic — you said one of them had a little, okay. And 440C should be just as magnetic as carbon steel, and tool steel should be just as magnetic. The numbers actually do have to do with carbon content to a certain extent, certainly in the 1020 series and 440 series. But um, it's not carbon content, it's crystal structure, and crystal structure changes.

Which is one of the reasons I brought the carbon fiber-ware pot. [Tom holds up a 304 stainless steel cookpot.] It's 304 stainless steel, just sort of like the 302. And down here where it's not deformed, it's magnetic but not very magnetic. Up here it's really magnetic 'cause it got deformed more. This particular type of steel— and the reason I have 302 balls in there is because the more they're deformed, the more ferromagnetic they become. They transform on deformation to become magnetic, okay.

But in fact, that little yellow magnet there, that's what a lot of quality control inspectors will walk around a steel yard with — a magnet — depending on what types of steels are going to be in their yard, to distinguish what steel, one steel from another, based on magnetism. In fact, it's the same thing that I do all the time, the reason I have these magnets in my desk. And that went on there. Is because someone brings me something, and it's a piece of steel, one of the first things I do is I put a magnet to it to see how magnetic it is. Now, I've got enough experience, and Kathleen picked up on it, that they're not all equal magnetism, equal ferromagnetism. The 302 is the one that's probably partially magnetic that you have. The ones that are strongly magnetic are 440, tool steel, and, uh, the 1020. But you could have actually distinguished the 316 from the 302 by the level of magnetism.

And in fact, we have gauges in welding of stainless steel, we have gauges which is nothing more than a little magnet on a little beam-weight balance, and you measure the lift-off force. And based on the lift-off force, you can tell what percentage is ferritic stainless steel and what percent is austenitic stainless steel — two different crystal structures. We actually use that property in specifications for making stainless steel welds, so far as that goes. Just different— all I'm trying to illustrate for you is different properties.

What else was anybody else able to pick out? Tungsten — you actually have tungsten carbide there, right? But it's very heavy, right, you did it on density, right? Well, non— and it's non-magnetic. Now, not everybody had tungsten carbide. Some of you had silicon nitride, which looks very similar, it's dark gray, but it's light. But it's non-magnetic, okay. And people actually use silicon nitride balls as ball bearings at high temperatures in jet engines, okay, so far as that goes. So some of you may have picked out silicon nitride, but you probably didn't know enough.

What about aluminum oxide? I actually saw you were dropping things and weights and stuff. One of them had a nice hard clang to it, was very white. Remember that one? Yeah, it's hard, that's probably aluminum oxide compared to all your other things. Polyethylene — anybody know how you would check your polyethylene? I didn't do it for you, but if you had a big lighter and you light polyethylene, it will smell like candle wax, because it has the exact same chemical composition as paraffin wax. Polyethylene is just a super long extended chain paraffin wax, and it smells like candle wax when it burns. So one of the ways to check, if you got a little piece of polymer and you don't mind destroying it, is you light it and smell it, okay. Now you got to be careful, there are some polymers that give off cyanide when you— so anyway, uh, but they're not very common.

And stuff. Plus the color of polyethylene is sort of a translucent, okay, although it can be any color 'cause they put colorants in it, but the natural stuff. Nylon, uh, is somewhat harder. Anyway, it's hard to pick up some of these things. Polypropylene — you can't tell the difference between polyethylene and polypropylene even if you smell them, because they both sort of smell like candle wax. If you look at their chemical composition, you realize they're basically hydrocarbons and they're fairly similar — slightly different ratio of carbon to hydrogen, um.

In any case, there are multiple properties. These two rubbers I can't tell you right now, except back in my office I have the bags I got these things from originally. And um, if I went and measured the hardness with my thumbnail, I might be able to sort those out, if I had a standard to compare it to, right. But you didn't have a standard, so you had to use your kind of own common knowledge. I couldn't have done much better than most of you did, uh, even though I've been studying materials for 40 years, okay. But you did— you know how many did you get, you think? Okay, but a positive ID on how many? Four, five, okay? How'd you do? Yeah, you'll say four or five 'cause everyone else says four or five, right? Hey, you all won, you can all have an extra bagel, okay.

But the point here is, there are physical properties of materials, and they're not very sophisticated. We're not measuring things to six parts per million, okay, in order to tell the difference. And in fact, what I brought out over here— oh, there it is. Um, let me tell you that positive materials identification — anyone ever heard of PMI, positive materials identification? You haven't run into that in a shipyard or something? It's that the alloy is what it's supposed to be, okay. And it's part of a quality control program. None of you have actually worked in a shipyard yet, okay, and you haven't— you didn't get into PMI, okay. You don't have to, it depends on what you're doing in the shipyard.

But let me tell you that in some chemical plants, a chemical refinery, I've seen data of different steel pipes, and they've gone in and done a very careful positive materials identification, and 6% of all the pipes are mislabeled as to what type of steel they are. And you say, well, it's no big deal, it's all steel. Well, not necessarily. I was involved 25 years ago in a case up in the tar sands in Canada, where someone was supposed to repair— they had cut out a little piece, and they were putting in an 18-inch piece of 6-inch steel pipe, welding it back in. They had to open it up to get in to clean out the pipe, a bunch of coke deposits formed inside the pipe and they had to clean it out 'cause it was getting Hardening of the Arteries and couldn't get the oil to flow through anymore. So they cut out a piece, went in there with their reaming tools and stuff, cleaned it out, and then they welded a piece back in.

Well, the piece they welded back in was plain old carbon steel. It wasn't 1% chrome-moly steel, which has better sulfur, uh, corrosion resistance. And about three years later the thing had thinned down from like a quarter-inch wall to a wall that was 0.007 inches, about four sheets of paper. And the pressure it blew open, started a fire that cost a billion dollars, okay. People were not happy, okay, particularly people who had to spend the billion dollars, okay.

So positive materials identification is a big deal. And in that case they were blaming the welder. "Well, he should have known when he was grinding it that it wasn't carbon steel." And I said, how would he know when he's grinding it? Now, it turns out, if he was working in a steel mill or in a shipyard in the 1930s or '40s or even '50s and '60s, he might have known. Or his father or grandfather might have known. Because there's something that we used to use for PMI in a steel yard, uh, fifty, sixty, seventy years ago, called a spark test, okay.

And it turns out different steels, when you grind them, will give different sparks. You didn't know that, did you. And if you go look up "spark test" in Wikipedia, okay: high carbon steel, manganese steel, tungsten steel, molybdenum steel — they all give different-looking sparks. You didn't know that, did you. Well, this is a chrome-moly steel, the kind of thing that ball bearings races are made out of. [Tom takes the chrome-moly sample to the grinder.] Watch the sparks. Wait a second. See the secondary sparks on— this is carbon steel. Doesn't look a lot different to you, does it. But there are guys who've been doing this for 30 years and they can pick out the difference between those with dead accuracy, okay.

Uh, let me get another one. Here's 440C. Can you see the difference in that? This is a stainless steel; these are both carbon steels. These burn up in the air. Stainless has got a chrome oxide, doesn't burn very well, okay. Um, 304 stainless. Okay. Well, I can tell stainless steel from non-stainless steel, but someone who's really discriminating on a spark test, they could pick out every one of these because they do it every day. You know, I look at someone else's children — twins, okay, or triplets — I can't tell the difference, right, but their mother can, okay. Even their father can, okay.

Um, so there are some very simple tests, um, that work, that we use all the time, whether it's magnets, whether it's spark tests. Um, there are spot tests you can buy. [Tom produces a chemical spot test kit.] Little kits — this is not one of them, but a little kit like this that has a bunch of drops of different chemicals, and you sand the surface, put a drop of the chemical on, and you watch how— see how it foams.

Anyone ever gone to a pawnbroker and tried to sell some gold? Neither have I. But a reputable pawnbroker, assuming there is such a person, um, a reputable pawnbroker will take out some nitric acid and he'll put a little drop of nitric acid, and he will look at how the nitric acid reacts on the gold, the carat gold. And he can tell usually within about plus or minus one carat number what the carat is. He can tell 14 from 18 and from 10 karat, okay, by how vigorously the nitric acid reacts with the other elements in the gold, okay. Simple little tests. And there are whole chemical spot tests, if you go look and Google it, you'll find all kinds of little simple little tests.

But that's not what we use today. For $30,000 you can have this little gadget that's no bigger than this. [Tom holds up an XRF analyzer.] And it's basically got a radioactive isotope, and you just stick it up against the surface of something and you pull the trigger, and the radio isotope generates X-rays, and it's got an X-ray detector in there, and it will give you a computer printout of all the elements in the metal. So we don't have a lot of people who know how to do the spark test with accuracy anymore, 'cause now we have people who go out with a $30,000 instrument and pull the trigger, okay. But still, you go down to laydown yard for pipes and 6% of them are liable to be mislabeled, okay.

So who cares? Why should you care if you're in the US Navy? Well, a few years ago when we were doing our corrosion test— our corrosion course in January, actually was first or second year we ever did it, I had, uh, one student get up and give a presentation. It was during the second Gulf War, and they were on a destroyer. And now, the war didn't last but a couple of days, right, but it was the first day, it was the first few hours, and this one destroyer was there in the Gulf and they were going out on— what they were— their assignment, whatever they were supposed to be doing, and they lost their 3,000 psi air system.

What is it, 3,000— [the] 3,000 psi air system runs on the ship probably most things, okay. It's sort of your pneumatic system, and it runs lots of things just like electricity runs lots of things. The compressed air system runs lots of things. And when they lost their system, they had to pull back. The admirals weren't happy, they lost a capital ship in the middle of the thing — not that they lost the ship, they just couldn't use the ship, okay. So you can imagine how pleased your commander would have been, okay. You can imagine how pleased the captain of the, uh, of the destroyer was to know that his career was now over, okay.

Um, and so afterwards they looked at this brass fitting, and it was stress corrosion cracking. And so the student got up and gave this little presentation of what they had found. It turns out another student in the class was working in the Philadelphia Supply Depot about the same time. And when they found that it was the wrong brass alloy — it was not the correct alloy that was specified by NAVSEA for this application, and that's why it had failed by stress corrosion cracking in this application — and so of course they go and say, well, check our Supply Depot and make sure that all the stuff in stock is correct. 80% of the stock in the warehouse was the wrong alloy, okay. And so that student wasn't— that's not what she had planned to present, but, you know, as part of our discussion she says, oh, I was in the supply at that time and we found 80% of the stuff was bad, okay.

So PMI is important, okay. You— going back to Lord Kelvin, you have to know how to measure what you've got to get some real data. It doesn't have to be sophisticated by how you do it. But it creates big problems. It created a billion-dollar problem up in Canada. It created a ship that had to come offline in the middle of an assault, uh, not good things, okay. Uh, so mistakes happen, but they happen because people make mistakes, and because of those mistakes they can get multiplied, okay.

That's enough for today. Um, we'll do a little bit more on observables and measurements, then we're going to get into codes and standards tomorrow. Um, this week we're probably going to have live lectures, I think, every day. I won't be doing it on Wednesday morning, I've got to go somewhere, but Dr. Belmar is going to talk about factors of safety, okay, uh, on Wednesday. And you can start watching your videos as soon as you get them from Jerry or whatever, okay. We're going to stay in this room. You can turn it off, Mike. Um, we're going to stay in this room.