§1. What materials manufacturing is [00:05]
This is a high-level introduction to what this new 36-unit course is about. The purpose of the course now is not just welding and joining, which is the NAVC requirement. It's now supposed to be expanded into materials manufacturing, which would include casting, forging, rolling — anything that imparts geometry. You can take a simple bar of material — this is a bar of copper, we'll talk about it at some point — but you might want to impart shape. [Tom holds up a forged impeller blank.] This is part of a forging that obviously is going to be some sort of impeller, going to be machined a little bit further. So you want to impart some sort of geometry.
Sometimes you're just trying to flatten something. [Tom shows a piece of split aluminum plate.] This is a mistake in rolling. They rolled this half-inch aluminum plate, and because of the way they rolled it, it split right down the center. Not very useful unless you're trying to split it down the center — and actually there are times when we are trying to split things down the center as part of the forming operation. That's one of the ways we make seamless pipe, we split it right down the center. So there are little tricks we use. You're trying to impart geometry, and that's going to give you some sort of function.
[Tom holds up an I-beam section.] Here's shape — here's an I-beam. When I worked at Bethlehem Steel, this is what we did, we made bookends. This was back in the old days when Bethlehem Steel still existed and it was the second largest steel company in the world. We had a beautiful research laboratory on top of South Mountain in Bethlehem, Pennsylvania. Bethlehem had gotten its start by learning how to roll I-beams. It became second largest after US Steel — Andrew Carnegie's company — and the number two person at US Steel went and formed a company called Bethlehem Steel.
Bethlehem had two claims to fame. One, it had learned to alloy steel around 1900 with molybdenum, and they learned to make tool steels that could keep their strength at very high temperatures — like drill bits. That was an important breakthrough. The other thing is they learned to roll I-beams. Some people say that the growth of Manhattan was because of Bethlehem Steel learning how to make I-beams. Until around 1900, buildings never went above 10 or 12 stories anywhere, because when you make them out of bricks and mortar you can't go but so high. You need a strong backbone to go very much higher, and that needed to be steel, and you needed steel with a particular shape.
Why is the I-beam such an interesting shape? What's the moment of inertia?
Student: It's the rotational equivalent of mass with respect to acceleration.
Right. It's basically a measure of stiffness in any direction. It's proportional to the stiffness — how easy it is to bend something. Trying to bend an I-beam in this direction, you've taken a lot of your mass out of the center where you don't need it. You only need the mass at the ends, and that's what the flanges are all about. If you put the mass on the ends, where you need it if you're going to bend the I-beam as a beam like the floor of a building. However in this direction it's not quite as useful. In twisting it's sort of lousy. So there are lots of different moments of inertia for a given structure. If you go through the civil engineer's handbook, they'll give you the moments of inertia in different directions for different shaped I-beams and channels and angle irons. That's what a traditional old-style civil engineer used to spend their life doing — not much more than a draftsman who knew how to read the steel construction manual, designing buildings and bridges, worrying about moments of inertia and stiffness, trying to get the greatest strength for the least weight. Because steel costs money.
§2. Shape, microstructure, and the cost of metals [06:54]
So shape is part of your function, and microstructure imparts function too. [Tom holds up a slice of continuously cast trolley wire.] This is a slice from a continuous casting. Anyone been on the Acela train between Washington and Boston? It's an electrified train. Electrified trains either get their power from a live third rail on the ground, or they have a trolley wire up above. This is a piece of the trolley wire that was supposed to hang between New Haven and Boston when they were upgrading for the Acela. It has the shape so you have little pinches that can hold the wire. It was supposed to be made originally from a casting of very high purity copper-silver alloy. This only has a tenth of a percent silver, so they left it for me — I can still use it, it's not worth much. But it has a particular structure; you can look at the grain structure on this longitudinal section.
So you have geometry, which is shape, you also have a microstructure of the material that imparts part of the function or the strength, and all of that has to be done at a reasonable cost. Anybody have an idea of the current cost of steel per ton?
Student: [inaudible]
Well, there's lots of different types of steel. First question you should ask me: what type of steel? Mild steel — that's a good choice. Mild steel in a simple bulk form like a simple bar or wire might be $500 a ton. Or rebar, concrete reinforcing bar — cheap steel. It's not always mild steel, it's actually medium carbon steel, but the cheapest there is today is probably going for $400, $500 a ton. That turns out to twenty to twenty-five cents a pound. Cheaper than bottled water per pound. Interesting thought, isn't it?
They call it mild steel, they really mean low carbon steel, but the typical term among the guys in the shop is mild steel. They mean a steel that's readily weldable, has no alloying elements.
So if I'm trying to impart geometry, function, and all at a reasonable cost to my material, materials manufacturing is in fact a function of a number of things: time, geometry, cost, structure, composition.
§3. The Attleboro gold case: when time dominates cost [11:05]
When I first bought my house, just before I bought my house in the late '70s — still live in that same house — I got a job as an assistant professor consulting for a firm down in Attleboro, Massachusetts. These people make — you got that continuously cast bar of copper — they would do that in 14-karat gold. This firm makes probably thirty to forty percent of the gold stock used by jewelers in the United States. This firm is the reason Attleboro is known as the gold capital of the world. A lot of the jewelry in the world comes out of Attleboro. They used to make the gold alloys that other people down there would turn into MIT class rings — the brass rat. That's made by casting to impart a very complex geometry.
This company would start out with gold bullion and they would alloy it in their casting shop. They had ingot casting, two continuous casting machines — just like that bar of copper they were making karat gold plates, bars by continuous casting. They then would roll it down, clad it to make filled gold, they would make wire, they would extrude it. Virtually every materials manufacturing shaping, forming, was done in that one plant down in Attleboro, and still is. However, I learned in that business no one cared about what it cost to actually shape it, they only cared about how long it took. Because in that case all the cost was the interest cost on the gold. It was more valuable than all the processing cost. If I can make steel for twenty-five cents a pound, which is cheaper than bottled water, then the materials manufacturing cost is not very significant. But back in those days gold was only six, seven, eight hundred an ounce. You start figuring, casting a hundred thousand ounces at $800 an ounce — I would go down there and there would be millions of dollars in the pot. So they had to do it as quickly as possible. A lot of that bullion made it from start to finish in the plant in three days. It wasn't there very long before it came out as a finished product. So sometimes it's a function of time, sometimes geometry, sometimes the cost of the material.
Stainless steels can have a much higher cost. [Tom locates a book on bolting.] This is a book on bolting. This is approximate relative cost of fastener materials. Medium carbon steel, mild steel — relative cost of one. Stainless steels are two and a half times that. Austenitic iron base — these are super-duper steels, things like the Navy wants to build a stainless steel submarine — 12 times the cost of mild steel. Now you have to understand that HY-100, which they're building submarines out of right now, has not a unit cost, it's maybe 1.5, 1.8 times mild steel, maybe even two times. But the new austenitic alloys for the non-magnetic submarine will be 12 times the cost.
Nickel copper alloys — how many of you came out of the shipyard? They're using lots of Monel now in copper-nickel alloys, particularly in the piping. Even aboard ship, probably a newer ship, it's not mild steel piping anymore. Mild steel piping is just going to corrode in 20 years, and the ship's supposed to have a 30-year life. Now ships are supposed to have 50-year lives, and so they've decided it's cost-effective to make the piping out of something better that will last the life of the ship, rather than having to replace all the piping after 20 years. Any of you on one of the old ships that had carbon steel piping that you were constantly repairing and inspecting? It gives some of the enlisted men something to do, and the junior officers, gives them a lot to do.
Titanium, 75 times. In the old days I was involved in some of the early titanium submarine stuff. My first research contract was on welding of heavy-section titanium for the Navy. Austenitic nickel-based alloys, things like Inconels — your nuclear reactor materials — 100 times the cost. So we've got two orders of magnitude difference in cost of metals.
So when we're talking about imparting geometry and function at reasonable cost — time, geometry, cost, structure, composition — and quality control. Typically quality control can equal all your fabrication cost nowadays, because we get better and better techniques. It's just like medical diagnostics. Why is the cost of medicine going up? One of the reasons: we have all these quality control things, all kinds of diagnostic tests. In the old days they'd x-ray your arm, now they'll give you an MRI at about 40 times the cost of an x-ray. Because an MRI tells the doctor more and protects his liability.
Codes and standards many times are the limiting thing in materials manufacturing, particularly in the pressure vessel industry. The code puts severe limitations on any innovation. If you're in the nuclear business, it's very hard to come up with a better material for that reactor vessel, because you need to do about a billion dollars worth of testing to qualify a new material.
Environment, health, resources, politics — I put religion down here too. Professor Hosler here wrote what was basically her doctoral thesis on a book called The Sounds and Color of Power. Her thesis was that materials were developed through the ages because of religious symbolism. My thesis was most materials have been developed for military purposes — the military is kind of the leader in the technology. That was the conflict we put before the students. It really depends on what age you're talking about or what culture, but materials manufacturing relates to lots of different things.
Here is an article by Norm Augustine called "Social Engineering." Norm Augustine used to be on the MIT Corporation — one of the few non-MIT grads ever to make it to the MIT Corporation. He was an engineer at Martin Marietta, wrote a book called Augustine's Laws where he looked at trends; he's a strategic thinker. He gave this as a commencement talk at University of Colorado in 1993. He talks about the ages of engineering and how society has gone through lots of different things over time. He says we're now in the age of socio-engineering, where a lot of the decisions are not made for technical reasons, they're made for sociopolitical economic reasons.
Right now a number of communities are trying to outlaw bottled water, because it's that wasted plastic at the end. My lab is undergoing — just had its asbestos remediation. This room already had asbestos remediation, but most of MIT, the old buildings, the ceiling tiles were asbestos, the floor tiles are asbestos, so we were surrounded by asbestos. And now they're coming in at costs of $100 a square foot to get rid of the asbestos.
§4. The history of nails: trees, charcoal, and the English energy crisis [22:30]
I want to give you a case study — the history of nails. Nails have been around for thousands of years. The Romans forged nails by hand. [Tom passes around forged and cut nails from a Massachusetts firm.] The reason I want to tell you about nails is because it gives you a history of where materials manufacturing has been and where it's come. There's a book, a series of PBS articles on metallurgy — they wrote a book on it, called Out of the Fiery Furnace. 1986, Pennsylvania State University Press. The impact of metals on the history of mankind.
There were some very interesting things going on in England in the 1500s. Ironmaking required charcoal back in the 1500s. They hadn't started using coal. They would take trees and turn them into charcoal. Anybody know how you make charcoal from trees? You burn it in an oxygen-poor environment. You fell the trees, make a big pile, put dirt over that, and limit the amount of oxygen that can get in. You start the fire in this chamber like a brick kiln, limiting the amount of oxygen. As the thing heats up in this oven, you're burning off all the volatiles — the lignin and other things in the wood — and you're left with charcoal, which is basically carbon. The charcoal is relatively strong compared to wood, doesn't give off a bunch of smoke because you've smoked it out. Charcoal was the way to make iron.
The earliest known water-powered blast furnace in Europe began operating in Italy in 1463. Before that they would make cannons out of brass or bronze, but brass and bronze are about two or three times as expensive as steel. When they learned to melt cast iron in a blast furnace, they could make cheaper weapons. So I'm now proving my thesis over Professor Hosler's thesis — it was military technology. From that time onwards the new technology spread rapidly, especially when it was found that cannons could be cast from iron instead of bronze.
Over the next 50, 60 years the trees in England were being felled with greater and greater rapidity. The Weald of Kent was densely forested with mature oaks, beech, and chestnut. They stood on chalk, and on that chalk, beginning in Roman times, people had found rich pockets of iron. So they had iron ore, they had trees for charcoal, and the chalk itself was the flux they needed for melting. They had everything they needed, and so they started cutting trees.
But they had a problem because the great oaks were what they built the man-of-war with. Why is Old Ironsides called Old Ironsides? The Constitution is the oldest ship in your Navy. Reportedly cannonballs would bounce off. Why was it stronger than the British ships, which were also made of oak?
Student: Different oak.
Different oak — very good. It was Carolina live oak. There is now a society to preserve the Carolina live oak because it's stronger than any other oak in the world. That's why Old Ironsides survived and they said the cannon just bounced off — because it was the strongest wood.
For the HMS Victory, Lord Nelson's flagship, it took 2,100 tons of oak. It took a whole forest to build Victory. So the two uses of oak were in conflict. And now a third industry was glassmaking. To make glass you also need charcoal because you can't get all those impurities in the glass or you end up with a mess. In 1558 a law was passed forbidding the felling of trees to make coal for the burning of iron. But as with all politics, they excluded the forests of Kent and Sussex, where it really would do any good. It's like saying you're forbidden to grow corn anywhere but Iowa. Good old politicians even back then.
In 1559 a writer complained the price of wood had risen from a penny to two shillings. By 1581 the shortage of wood for shipbuilding was so serious that a further act was passed forbidding the felling of trees within 22 miles of the Thames, within four miles of the great forest of the Weald, and within three miles of the coastline anywhere.
By 1615 England was facing an energy crisis. So here was the first energy crisis. A royal proclamation in that year lamented the disappearance of the kind of wood "which is not only great and large in height and bulk, but hath also that toughness and heart as it is not subject to rot or cleave, and therefore of excellent use for shipping." Again, military technology. 1615 — what colony had the Brits founded back around 1615? Plymouth Plantation here in New England, or Jamestown in Virginia. If you go to Jamestown today, they have a glass furnace shop. Jamestown had a glassmaking shop because they didn't have any trees in England anymore. We had lots of trees over here and could make charcoal over here. What did they have in New England? They had Saugus Iron Works.
If you go up here to Saugus, Massachusetts, it was the first iron works in the United States. They have a blast furnace up there.
§5. From Saugus to the basic oxygen furnace: a 600,000% productivity gain [32:10]
At that time it took approximately one person-year per ton of iron. If you figure a person-year back then was 3,000 hours — they probably worked at least 60 hours a week, six 10-hour days, they didn't work on the Sabbath but they probably worked more than 8 hours a day during the week. They had to chop down the trees, make charcoal, put it in a blast furnace, cast the pig iron. Why is it called pig iron?
If you go up to Saugus — this would be a good place to take a little field trip, take your kids, your spouses, it's a national park now, doesn't cost anything — they will show you the bottom of the blast furnace. A blast furnace is just a big column, but at the bottom you tap a hole and the molten iron comes running out onto the sand floor. As it runs, the guys would dig a trough in the sand. So when the molten cast iron comes out, it would run down this trough into a manifold, and off the manifold would be other little troughs. The liquid comes down, fills up this manifold, and fills up each one of these little things. These were like the pigs on the sow, and this was called a sow bar. It weighed four or five hundred pounds, the pigs only weighed 100 or 200 pounds, but this was like the suckling pigs on the sow.
The sow bar might be what you turn into cannon, and you had to have lots of men just to even move that. The pigs were what you would make things like nails out of. To make nails you basically hand-forged them, or you could roll them into sheet. Most of those nails are either hand-forged or cut from sheet. They didn't really have a lot of wire-drawing technology back then. As a result, nails were very expensive. They would burn down an old building and sift through the ashes to recover the nails to reuse them. Can you imagine doing that today? Because nails are among the cheapest of our wire products right now.
Today it takes a half hour or less per man to make a ton of steel. We've seen a productivity increase — materials manufacturing as a function of time, materials, composition — over the last 400 years we've had a six-thousand-fold increase. That's a 600,000 percent improvement in productivity to make steel. Not bad. And we still see how those things change in my lifetime.
When I worked for Bethlehem Steel 35 years ago, half the cost of making steel was labor. Bethlehem Steel, second largest steel company in the world: forty-five percent of their cost was materials — limestone, iron ore, coal — forty-five percent was labor, and ten percent was profit. Today it's ninety-five percent materials, ten percent labor, and negative five percent profit. Actually it's changed a little bit in recent years because the Chinese have started using so much metal.
One of the things that happened, again in my lifetime, is they went from ingot casting to continuous casting. In the old ingot casting technology, one-third of your steel you had to recycle. You had to cut off the tops of the ingots. In steel mills of the 1960s you'd have a big mold and you'd cast the steel in there and you actually would have to cut off the top third because of shrinkage. One of the numbers we always had to worry about at Bethlehem Steel was trying to improve the yield. We were at about sixty-five percent overall for the whole corporation, which meant for every 100 pounds of steel you poured you only could ship 65 pounds.
When they went to continuous casting, which was being developed in the 1960s and really came into its own where almost all steel by the 1980s was continuous cast, this number went up to 95 to 97 percent. Today it's at about 98 percent. With continuous casting there's no top shrinkage. You just cut off the bar — the shrinkage is only at the beginning and the end. A good continuous cast nowadays starts and then doesn't stop for a week or maybe two. They cast continuously for one or two weeks. So there's not all this wasted material. That was a dramatic increase in productivity.
Another increase in productivity in the steel industry: they went from open-hearth furnaces, where to make 3 or 400 tons of steel it took 24 hours to refine the steel, to burn all that carbon out of the cast iron. In the late '50s, in Linz, Austria, someone found they could blow pure oxygen into the steel bath, and pure oxygen will burn carbon faster than air. To do it in a 300-ton bath is a little more exciting. Now they can burn the carbon out of iron not in 24 hours in an open hearth, but in a basic oxygen furnace they can do it in 20 minutes with a supersonic jet of oxygen. The vast majority of the pure oxygen in the world is used to make steel. They just blow a super jet, about this big around, of pure oxygen into the steel bath of 300 tons of steel. Really exciting. Anyone ever been to a steel plant and seen a BOF in operation? You want to see fireworks, it's really neat.
§6. Glass: bullseye, plate glass, and the Pilkington float process [39:46]
You've got this glass blower named Blaschka. Blaschka was making little colored replicas of sea urchins — people would bring back these little sea animals, he'd look at them under the microscope and do a model in blown glass. Someone said, well, can you try to do a flower? So he did, and he shipped it across the Atlantic, and it kind of came across and was broken, but they could still tell there was a lot of promise. So some rich woman over in the Back Bay gives a bunch of money to Harvard and they brought Mr. Blaschka over from Prussia. For the next 60 years he and his son would take plants like a cashew from Brazil or India, and they would copy it in glass in full color. They would also put it in the microscope and show you whether it's a monocotyledon or a dicotyledon. So it's a scientific exhibit and an art exhibit all at once.
One of the things they didn't do was all the materials science to know whether the glasses were going to be stable over the next 100 years or whether they would decompose. Apparently only about 30% of them are still on display. When I started out they were free at the Peabody Museum, then they started charging 25 cents to get in, now it's up to like 10 bucks a head. You still only get to see about 30%. In the 1930s, his son, after the original Blaschka had passed away, apparently started getting interested in diseased fruit. So you can go in and see a pear on a branch and it's all shriveled with fungus. So that's the glass flowers.
What is bullseye glass? [Tom shows an image of bullseye glass from Sugar Hollow Glass.] It's a spun-out plate. They have a departmental glass-blowing facility — they put hot glass on a stick and then shape and form it with tools. You'll see the same thing in Corning, New York.
Bullseye glass was the way they made glass plate or glass windows back in the days of the ironmaster at Saugus Iron Works in the 1620s and 1630s. You would start out with your blob of molten glass on a metal pipe, because you're going to want to be able to blow through the pipe to blow a bubble in this hot glass. You then flatten the end with a mandrel tool while you're spinning it. They would then cut a hole in it, put it back in the furnace to soften it again, and then spin it into a disc. They would cut out the little window panes as squares from the flat disc, but the bullseye was the thing that was left over that had been connected to the pipe.
In the old days, bullseye glass was not taxed because it was considered junk. It was scrap, it was not pretty. The people who could afford good glass for the window panes were the wealthy people, and the poor people would buy the bullseye glass because it was cheaper, it wasn't taxed, and they would put it above their doors just to let light in. If you go to the Old Manse, the building where Mrs. Hawthorne scratched a poem in the windows with her diamond ring, they have bullseye glass above the window. If you go to Williamsburg, Virginia, or Jamestown they will tell you about bullseye glass. So in the old days bullseye glass was the cheap stuff. Today, if you go to Sugar Hollow, you buy a little piece of bullseye for 100 bucks a pane. It's the premium stuff because you want to pretend like you're poor when you're really rich. But it is handmade.
Glass got more sophisticated in the 1800s, rather than just a little hand job making a little disc you could cut window panes from. That's why old windows were never big panes of glass — they were made from just spinning this little thing and cutting it by hand. They actually got to the point where they could take big glass melting furnaces and pull out whole sheets of glass. Hot glass is sort of interesting — it can flow like taffy and you can stretch it out. They got to the point where they could pull out 6-foot wide sheets.
At the turn of the 20th century, Pittsburgh Plate Glass had a building that supposedly was a mile long beside one of the rivers in Pittsburgh. They would have this glass melting furnace and pull the glass out, let it cool, put it through an annealing furnace because the glass will develop residual stresses and can fracture very easily until you anneal it. Then the greatest length was because they had to polish it because it wasn't very smooth. If you go look at old buildings from the 1880s and 1890s, you can tell which were the original panes of glass — the originals have got all kinds of waves on them. I can go out here at MIT in my office and you can look and see which panes of glass might have been original in 1916 or 1917 when these buildings were built, because they're not as flat.
Around the early part of the 20th century, the Pilkington company in England developed the Pilkington process for float glass. The float glass process still pulls out the sheet of glass, but instead of putting it on a table to cool and then go into the annealing furnace, they pull it out onto a bath of molten tin. Tin melts at around 400 °C, but the other good thing about tin is it doesn't vaporize until about 2,000 °C — it's got a very low vapor pressure. They would pull it out and the glass would anneal on this molten tin, and the top surface would just smooth out by gravity, and the bottom surface would be smooth because it's on top of this bath of molten tin. That's the float glass process. Everything that's in your house today is made by the float glass process.
Instead of a PPG building that goes for a mile in Pittsburgh, they now have buildings that may go for 150, 200 yards long. The actual melting furnace is maybe twice the size of this room for the hot furnace. The annealing part is probably similar. It's all continuous — continuously pulling out glass, feeding glass in the other end, melting it and pulling it out, floating it on the tin. No more polishing of the plate glass, and it's very flat, very clear. The top surface is better than the bottom surface, because the top surface is just floating in air and gravity makes it smooth.
That computer right there has a glass screen. That screen is made — and I can only tell you a little bit about this because I had to sign a confidentiality agreement to go through this plant — but it's made by the float glass process, and the furnace is actually no bigger than this room. It's not anywhere near as wide. They pull out strips of glass from the melting furnace. So it's got the air surface and the tin surface, and then they split it in two and they drop it and reweld it basically by putting the two hot pieces so the two tin surfaces are welded together on the inside. Two pieces of bread coming together with no peanut butter in the middle. Now you have a float glass surface on each surface — each free surface is now a float glass surface. That's how they make something very thin. As it falls by gravity for several stories, it's falling through furnaces that control the temperature so it just slowly thins down and stretches out like Silly Putty. They end up with something that's only fifteen or twenty thousandths thick but has two air surfaces of float glass.
Very profitable business. Start with sand, end up with the outside surface of a computer screen. But they really control that technology. I've been through a number of military facilities — you guys don't know anything about security until you've been through a computer screen plant. So that's how glass has gone from a hand art, where bullseye was the cheap stuff, and now it's still a hand art making bullseye and selling it for 100 bucks a pane. But float glass has shrunk plants from a mile long to a couple hundred yards. When you really want to be precise for TV screens or computer screens, you get even fancier.
§7. Fasteners and the variety of nails and bolts [59:31]
Let's talk about fasteners. You've seen the nails and we talked about the productivity increase with nails. [Tom holds up an assortment of fasteners from Home Depot.] What's this? A cotter pin — a different type of cotter pin. Here's the regular type of cotter pin you're familiar with. This one's just like a safety pin, operates the same way, it just locks on rather than bending it. Here's another one that slips on and locks. There are washers in all kinds of sizes and shapes — fender washer, lock washer. Here's a bolt for hanging things in the wall where things spring out. There are plastic bolts — this is the latest and greatest for going into gypsum board, big threads. You drill a hole, screw this in, and if you want a metal screw you screw the metal screw into that. There are tens of millions of different types of metal fasteners.
We don't use nails anywhere as much as we used to. [Tom passes around coated and uncoated nails.] These are similar nails but one of them's got a coating on it. Anyone ever used resin-coated nails? It melts the resin from the friction. Have you ever tried to pull one out with a claw hammer? You'll break the wood before you get that nail out.
When I used to do Cub Scouts when my sons were that age, one of the things in the Wolf book is to have them drive a screw. The first time I had the scouts do that, unless I made the hole really bigger than it should have been, they couldn't really drive that screw without stripping the heads. You can use a Phillips head, which is not as easy to strip as a flat-head screw, and now we've got Torx and things that make assembly easier. I drilled all these holes the proper size to screw them in, but these seven, eight-year-old boys were still stripping the heads. What's the solution? I went and got a bar of soap.
So you can use resin-coated nails to lock them in place with a coating, or if you're trying to get a screw in and it's twisting too hard, you lubricate it. Soap's not a bad lubricant, and then the kids could do it just fine. Here are nails — zinc coated, but they've got a twist to them, so when they go in they twist and help lock them in. I think they're deck nails. Here's a nail that's very lightweight. You know what material that's made out of? Aluminum — that's even lighter than galvanized. This is a galvanized nail. The twisted nail was a galvanized nail. Here's a coated screw — a deck screw, coated for similar reasons.
We've gotten more and more sophisticated — it's not just some blacksmith hammering them into shape. Even the blacksmith hammering them into shape — some of them have little bulges on the surface so that when it goes into the wood it won't pull out as easily, because the wood has to expand beneath the surface, and because the wood will creep back with the moisture, it essentially locks it in. People do all kinds of things to try to improve things. Nylon screws — very corrosion resistant in most things, but not in everything. There are expensive bolts.
§8. The Morgrip propeller bolt and the precision of bolt stretch [65:20]
[Tom holds up an Allen-head cap screw.] This is a high-strength Allen-head bolt, sometimes called a cap screw. You often find it in good expensive machinery, equipment. It's just a very high strength steel bolt. [Tom holds up a much heavier bolt.] This is the heaviest bolt I have. This is called a Morgrip bolt. They're used to hold the flutes of the propellers on ships. It's got a hex head, and it's got an Allen head in the center here, because it actually has a hole right down the center.
Anyone ever used a bolt that had a hole down the center? What's it for?
Student: Well, they use an ultrasonic tool, and as you tighten it, the bolt actually stretches. The bolt stretches as you pull on it, and this has to be a specific amount of space between the end of the hole and your actual...
Right. Most of the time the scouts just torqued it until it got in, and the screw was done when it got into the wood. But if you're building a nuclear reactor or something, you've got to make sure people torque the bolts. Actually, if you're just working on a car they ought to use a torque wrench on the bolts. Bolts have a problem if you under-torque them or if you over-torque them. What happens if you under-torque them? They don't have as much clamp and they can come loose. What happens if you over-torque them? Too much stress and they can fatigue and become really loose, they break.
When you have a really critical application — not some lousy little thing like a nuclear reactor, but something like an aircraft engine — you actually measure bolt stretch rather than bolt torque. Because if you measure torque, it depends on the friction on the bolt, the friction coefficient, whether it's lubricated or not on the threads. It also depends on how well the threads are cut. We talk about the class of threads — Class 1, 2, and 3 — which is how closely the male threads meet the female threads. Class 2 is a good bolt, like one of these. Class 3 I think is the highest, Class 1 is garbage. A very sophisticated bolt will actually have a shank that is smaller diameter than the threads, so that you actually have solid metal all the way down. You don't thin it down where the threads are, because if you do a stress analysis on this, you'll get a more uniform stress if you actually have the shank a smaller diameter.
They often put a little ultrasonic thing in there and they can measure the stretch. They also sometimes just put a rod down there and measure the length of the hole.
Student: The application I was experienced in was on the dive side — some Navy divers were replacing some blades on propeller hubs.
Right. If you're diving, it'd be a lot easier to use an ultrasonic tool than to get in there with a little micrometer trying to measure something in ten-thousandths. When you've got your gloves on and stuff, ten thousandths is sort of a joke. This bolt, as I remember, came off an Israeli Corvette propeller. They were cracking — there are cracks right under the head. When we were cleaning up the lab a couple of weeks ago I retrieved one.
The propeller bolts are about as sophisticated a bolting operation as I've ever seen, using Morgrip bolts and measuring stretch. They also measure stretch on the connecting rods on aircraft piston engines. If you've got a Lycoming engine or a Teledyne Continental engine, when they tighten up the connecting rod, the connecting rod goes with a bearing against the crankshaft, and they have two bolts holding it on. Those bolts have ground top, head, and bottom end, and they're only about two and a half inches long. You take a micrometer and measure the bolt before — it's ground to within 2.623 inches plus or minus one thousandth of an inch. After you stretch it, torque it, you don't measure the torque — well, you can approximately measure the torque, but you don't rely on torque because of all the variations in friction. You actually measure the stretch, because Young's modulus doesn't change by very much, it's a lot more precise than relying on friction.
It might be 8 thousandths of stretch on a two-and-a-half-inch-long bolt. You can take 30 million times 8 thousandths over two and a half inches and calculate what the stress is in the bolt. By using that technique you can get down to plus or minus 5% variation on the clamp-up of the bolt. Whereas if you're just torquing something, your variation will be plus or minus thirty percent. Torque wrenches are not very precise. Many times in nuclear reactors they do use stretch — measuring bolt stretch is a lot more precise than measuring bolt torque. Many people don't do either, and fortunately when they don't do either I often get to put another one of my kids through college.
§9. Bolt stress, tolerances, and high-strength alloys [72:02]
If you look at the stress on a bolt at different locations — A, C, D — you'll get peaks in your stress. They put a nice fillet on here, but they didn't thin down the shank compared to the threads. You have a stress concentration underneath the head, you have a stress concentration right where the nut hits the first thread, and then it decreases. You have some pretty significant variations.
If you look at bolt tolerances — if you take Class 2 threads, this is a magnified version of the tightest clearance where the two fit really close. This one's at the maximum and this one's at the minimum and you have very tight clearance. This is the amount of slop you've got in Class 2 threads, which is the typical good thread. There's good, better, and best. This is the better Class 2, and that's how much slop you've got. Bolts are actually treated sort of like springs.
If you want to look at how strong bolts can be, we've got lots of different materials, just like nails — we make them out of aluminum, out of bronze, out of lots of things. ASTM A325 are the standard bolts you put in I-beams if you're building a building. Why would you use bolts to build a building? It's a lot easier to slip a bolt in than to get up there and weld the thing. I can slip the bolt in with the cranes putting the beam down or the column down, slip the bolt in, put the nut on, and things aren't going anywhere. I can then come back and tighten them up. That's a lot faster and can actually be better tolerance in an erection than if you weld at the wrong angle and the thing's out of place and then you have to grind the weld off.
One of the advantages of bolting is it can be faster in the field than in the shop. Sure you'd rather weld everything together, but bolting has certain advantages. A325 is a typical high-strength bolt. A490 is an ultra-high-strength bolt. The problem with the A490s is that they're very susceptible to hydrogen embrittlement cracking. They have a very high ultimate strength — 165 ksi, about 100,000 in shear, which is the way they're usually loaded. A286 — that's the 12 times as expensive stainless steel as carbon steel. The strongest material on here is MP35N.
MP35N is one of the most fantastic metals I know. It's got a composition that's basically 35% nickel, about 20% chrome, 10% molybdenum, and the balance is cobalt — about 40% cobalt. Cobalt's not cheap, it's very expensive, three, four times the price of nickel. Molybdenum is even more expensive than cobalt. Chromium is just typical stainless steel. But it's a nickel-cobalt alloy — these are very expensive bolts, probably more expensive than titanium, on the high end of the nickel alloys. So rather than the 100 to 1 figure I gave you for fasteners before, this is on the 200 range. But it's non-magnetic, and it gets its strength from cold work.
A guy in International Nickel named Gaylord Smith invented this alloy back in the '60s. It's great for a lot of things. I've recommended it for aneurysm clips that go in the brain for medical applications. Won't show up on magnetic resonance imaging, won't mess up things for NMR. Very corrosion resistant, as good as titanium alloys in most cases. It's got very good toughness.
This stuff can have a tensile strength of 300 ksi, which is as good as the best steels for aircraft landing gears. It can do that with 10% elongation and 50% reduction in area. There's no other material I've ever heard of that can give you that kind of ductility with that kind of strength. So among all metals this is way out there. The problem is, it gets its strength from cold work. You can make it into 2- or 3-inch bolts if you start with a 6-inch diameter rod, but you can't make it into plate. You can make it into wire. So it's fine for medical implants and little sheet things — little clips for the brain. And it's good for some high-strength bolts. I've heard of people using it on propellers, I haven't seen that myself. I've also heard that there are four MP35N bolts that hold the fuselage of the 747 to the wing of the 747.
You don't talk about holding the wing to the fuselage, because when you're up there flying around, the fuselage is just being lifted up by the wing. All that's keeping it on the ground is keeping the wing from falling down, but up in the air all it's doing is keeping the fuselage from sliding off the wing. There aren't a lot of bolts, and they're MP35N bolts.
§10. HY-80, HY-100, and the limits of high-strength steels [79:41]
A286 is twice as strong at high temperature. It's a very complex stainless steel, heat treated in special ways, very highly alloyed. It's not all that different than HY-200 in composition. Are you a submariner? After the Navy developed HY-80, they wanted to go to HY-100.
The Nautilus was the first ship that had HY-80 in it, sometime in the '50s. They had all kinds of hydrogen cracking problems, such that some of the first subs they actually welded with austenitic stainless steel, because they couldn't solve the hydrogen cracking problem. It took them about 10 years to really get their arms around that, and then they had the Thresher problem and they had Subsafe. They didn't actually start putting HY-100 into ships until the late '80s. There were two ships before the Seawolf that they put modules — cylinders in the center of the ship — with HY-100. I'm familiar with that because my first student ever still works at Electric Boat as a welding engineer. He was in charge of welding up those HY-100 cylinders. When the Seawolf had all its cracking problems, the two captains who were in charge of the ships came to me and said, well, what about our two ships with the HY-100 that are out there? I said, don't worry about it, just inspect them when they come back in. They're not going to fall apart tomorrow — but they could fall apart if you didn't inspect them properly.
Originally the Navy spent $50 million in the '60s developing HY-130, which was really supposed to be an HY-150 but they downgraded it. I remember having to go meet with the chief engineer of the Navy around 1990 and explaining to them that they were trying to sell the Seawolf as an HY-130 ship, and I said, no, it's going to be a big enough challenge to go to HY-100. It turns out it was. I said, you're never going to build a ship out of HY-130. But there were people in the Navy who said, well, we spent $50 million on it in the '50s, it must be worth something. I said, just because you spent $50 million doesn't mean it was worth something — let's be realistic. They also tried to develop an HY-200, and they did build some research vessels that were sort of classified, and there have been some very special applications.
A286 is almost like the Air Force 1410 alloy, which is a 10% cobalt alloy. Very sophisticated strengthening mechanism, but very intolerant of hydrogen. That's part of the problem with HY-130, HY-200 — no one's ever really solved the hydrogen problems and been able to keep the toughness up. So I've run over a little bit today, but we'll finish up bolting tomorrow, and go into casting. We'll go into materials manufacturing — I'll tell you why nanotechnology does not work for structural materials, and then we'll get to casting. You'd never see a Buckminsterfullerene skyscraper — on the moon maybe. When we get to the material selection part of things, I'll explain why cost-wise there is about five or six orders of magnitude difference between the cost of a commercial ship and a spacecraft.