§1. Class logistics and the hydrogen-in-steel demonstration [00:02]
We did have a class yesterday but it was an introductory class, and I understand that some people thought that because yesterday was a recitation we wouldn't have class. If you want to sign up, you can get from my secretary Bing tomorrow the handouts from yesterday. It was just an introduction. We videotape the classes. We meet every day in general, 9 to 10. We will have about twenty-four live lectures this semester. We will finish up about the first week in March. After spring break you'll have to do about a ten-minute presentation. And then you have to watch some other videos, and you'll get all that from the handouts. There's not much to do in this class.
I have a bad attitude about the way we educate people in schools in the United States right now — we just teach them how to take quizzes. So we don't have any quizzes, we don't have any final exams. I would appreciate your asking questions so that I can diverge and think of a story to tell — just like you came in and said something about the snowstorm and how fast it's coming, so I tell the story of the Blizzard of '78, not that it's relevant to anything here.
[Tom pours Sprite Zero into a beaker.] All I wanted — I could have used club soda, but I wanted something that if I pour it in here you'll see the CO2 evolve. Foam. I want you to watch how it dissipates, and we're going to talk about it towards the end of class. This is sugar-free; I don't want to spill it all over the table and get everything sticky for the next class.
This relates to the Achilles heel of steel. Those of you who took the class last term, anybody know what it is? It's called hydrogen. And what you just saw — CO2 coming out of a soda — is timewise an almost perfect analogy for how fast hydrogen comes out of steel as it solidifies. Over the next 48 hours this will be flat and have no zing to its taste. In 48 hours you saw how fast hydrogen — or CO2 — came out and bubbled away. That's about how fast, once the steel starts to solidify, it's rejecting the hydrogen. You see some bubbles coming up here, and you'll see them at the end of class but there'll be fewer.
In processing of steel there are specifications that say if you introduce hydrogen from welding or electroplating or any other process, you have to put it in a furnace and bake the hydrogen out within one to 48 hours, because if you don't, hydrogen can form cracks in the steel under certain conditions. We're going to talk about that more than you ever wanted to know today and tomorrow. When you try to measure hydrogen content in steels — you weld the steel and then want to measure the hydrogen — if you don't quench it in liquid nitrogen within one or two minutes, you lose half your hydrogen, because it's coming out just like the foam in the soda. A visual will give you something to remember. I could have told you that, but it won't be the same as having poured three cans of Sprite into a beaker.
§2. Why we talk so much about steel [04:55]
Does anybody have any questions? I want to talk about welding metallurgy today, but I have to give a little bit of an overview. Last semester I talked about an introduction to material selection, and this is an article I wrote a couple of years ago on selection of materials and cost of materials. If I had to give you a three-page outline of the twelve hours of lecture I gave last fall, that's it.
Those of you who were here last semester know that I kept talking about steels and sort of apologizing. Why does he talk so much about steels? Steels are 95% of all metals made. If I look at tons produced per year, steel is 1.5 billion tons. The next most common metal is 45 million tons. Stone — I've estimated it's 6 billion tons in the world. Cement as a structural material is about 2.2 billion tons. These two are more than steel. All plastics combined are only 300 million tons. So the plastics industry on a weight basis is 20% of the steel industry. Yes, we will talk a fair amount about steel because it's the predominant structural material.
Now you can say, well, what about aluminum in automobiles? They've been making all-aluminum Audis for twenty years. They made an all-aluminum Duesenberg in the 1930s, or a Pierce Arrow — there's a picture I have of JP Morgan standing next to his all-aluminum Pierce Arrow. But those were not consumer cars like a Model T. The Ford Taurus is still made out of steel. You can buy a $90,000 Audi, all made out of aluminum. And you're going to soon be able to buy a Ford F-150 that's not all aluminum — not the frame rails, they're still steel — but the body is going to be aluminum. They haven't announced the price. This summer they'll announce it.
Anybody want to bet what the increase in price is going to be? The base F-150 right now is about a $30,000 truck. You can put all the bells and whistles and get it up to 55,000, but it's a $30,000 vehicle. Anybody want to guess what the all-aluminum F-150 is going to cost? I'll bet you it's in the $40,000 base price range. It'll add a third to the price. That's my prediction. But it's going to blow away all the other pickup trucks on gas mileage. It's the most profitable vehicle for Ford — they're making $10,000 on every one. So they're going to take a big risk of coming out with a high-volume vehicle, the first people ever to do it. It was twenty-five years ago that I said it will be twenty-five years before we see a high-volume all-aluminum vehicle. Well, maybe it was only twenty-four years. But it takes about that long.
§3. Defining steel and the iron-carbon phase diagram [09:13]
So we're going to start out weldability talking about steels. One of you in last semester's class asked me, well, why are there so many steels? If you go look in handbooks you'll see hundreds of different steels. You first have to ask, what is steel? How is steel defined? It's an alloy of iron and carbon. Very good.
[Tom passes around an iron-carbon phase diagram.] I got to figure out the exposure on this thing. That's a little overexposed. You plot temperature versus the amount of carbon in the steel. If you look down the corner, you'll see the name of John Chipman. Those of you who've been to our departmental conference room — it's called the John Chipman Room. John Chipman came to MIT in the late 1930s. He was a physical chemist from Georgia Tech, and he taught people how to make steel with consistent composition, and he had a tremendous impact on this industry. This phase diagram he published in 1972 when he was 75 years old — he retired in '62. I remember going as a young assistant professor to meet with John Chipman and ask him some questions when he was in his 80s — he was still working on things.
This is the liquid region at high temperatures, above 1530 degrees C. The steels are at less than 1 or 2% carbon, and they have this region called austenite. The cast irons are also on this diagram — they're above about 2.5% carbon. So the cast irons are in here, the steels are over here. The interesting thing about steels, and one of the things that gives them a lot of their weldability, is they transform from body-centered cubic to face-centered cubic at a temperature somewhere between 723 and 910 degrees centigrade depending on the carbon content. Because of that solid-state transformation, the steels can give you very fine grain size, and very fine grain size in most metals gives you very good strength. I can weld two pieces of steel together and just let them solidify without doing anything special and I'll get 100% joint efficiency. That's not true in many of your other materials.
It is true in steel, and it's true in titanium. Titanium has a phase transformation like this, and a system that gives you a crystal structure phase transformation that leads to a refinement of grain size naturally will give you very good joint efficiency. You go to aluminum and you're going to see we have some problems getting 100% joint efficiency. What's 100% joint efficiency? You weld the bar together and you pull on it and you get the same strength as the base metal. We're going to talk about, well, do you want the same strength, or what happens if you use undermatching filler metal so it would always fail in the weld metal, or do you want overmatching, and how much overmatching? It turns out the answer is, you do want overmatching for good mechanical behavior reasons. We'll get to that as we go along.
§4. Carbon, low-alloy, and high-alloy steels — and what they cost [13:17]
This is the iron-carbon phase diagram. All steels are based on this. The steel industry, starting probably a hundred years ago or more, divided steels into carbon steels and low-alloy steels — actually they called them alloy steels, but we call them low-alloy now. And high-alloy. Carbon steels contain only carbon, manganese, sulfur, phosphorus, and usually silicon. If they have any of these other things — nickel or copper or chromium or molybdenum or high manganese, and high manganese I think is above 1.25% or so — they're called low-alloy steels. The alloy content might be up to 2 to 3%. High-alloy steels are greater than 4% typically. If you get high enough alloy content, you get steels that we don't even call steels anymore. When they're less than 50% iron we start calling them nickel-based alloys or cobalt-based alloys. There's an entire system, and when Mike Bonforth [Bibber] talks about stainless steels, he'll show you a genealogy of steels similar to what I showed the class last semester.
Carbon steels historically were dirt cheap. They were priced differently than low-alloy steels because it was much more difficult and expensive to make low-alloy steels. How much cost was there? If I add nickel to steel at a 1% level I'm adding $130 a ton. What's the typical price right now if I went out to buy some carbon steel plate or bar? Anybody know what it costs per ton? That's if you're buying in low quantities. If I'm General Motors I'm going to be paying $400 a ton — about 20 cents a pound. So the cheapest carbon steel bar stock is probably $400 a ton. Reasonably good quality automotive sheet might be $1,000 a ton, so it's 50 cents a pound. This is, I'm going to go out and buy 100,000 tons — I'm General Electric Appliance Park where they make washing machines and refrigerators, or I'm General Motors. They're going to negotiate prices of $400 to $1,000 a ton.
If you're talking low-alloy steels, the prices can be anywhere from $800 a ton all the way up to $10,000 a ton or more. High alloys, you could be in your $4-a-pound range — $16,000, or $8,000 is $4 a pound. If you get to the high-nickel alloys, you could be paying 30 or $40,000 a ton. So we build nuclear reactors out of these things, a great big thick plate, but it's not cheap.
Nuclear reactors cost right now — what's the cost of building a nuclear reactor? You can't do it because the environmentalists will take you to court for ten years. But if you could build a nuclear reactor what would one cost? Billion? Ten billion is a good number. Four billion, I'm not sure you're going to have a very friendly country to build in. You could build one for 4 billion. And it takes about ten to fifteen years to get certification, which is one of the reasons the cost keeps jumping. They use the best alloys. But again this gets into a matter of cost. If you're going to make an automobile, you're going to make it out of carbon steel if you can, or a very low-alloy steel.
§5. Why steel is attractive — formability, hardenability, and martensite [18:00]
What is it that makes steel attractive? There's a number of things. But one of the things we want from all kinds of material — since the dawn of creation people have liked to have something that's easily formed. I can take clay and I can form it when it's soft and make a paper clip holder. Then I can put it in a furnace, heat it up, and we call it pottery, and it comes out nice and brittle but hard. So it has wear resistance, and I can put water in it, and I can grind my cornmeal in it. So I'd like to have something that's hardenable — formable and soft at low temperatures, and somehow transform it. Pottery clay was one of the first things they learned to harden, but other things can be hardened, and steel can be hardened.
Steel can have strengths from 30 ksi — or if you're a megapascals person, more and more students want to use megapascals — 200 megapascals up to 700 megapascals, even up to 1.4 gigapascals. That's 200 ksi. And it can go higher than that. Aircraft landing gear are 250 ksi, or 1.7 gigapascals. So you can get tremendous hardness out of steel, and that's because the steel can transform at high temperatures to the face-centered cubic austenite phase. If you quench it you can get a metastable phase that is not body-centered cubic ferrite — you can get something called martensite. And martensite is as hard as a steel file.
There are other alloy systems that can be hardened. We're going to talk about aluminum, but aluminum is precipitation-hardened. You heat it up, solutionize everything, quench it, and then you bring it up to an intermediate temperature — we call it tempering. You get little precipitates, and some people will tell you that's nanotechnology. Well, the Wright brothers used nanotechnology in their engine at Kitty Hawk. It was an aluminum-copper alloy that had been precipitation-hardened — aluminum for lightweight because it's an aircraft, precipitation-hardening so it'd have decent wear resistance. And it's still sitting in the Smithsonian today.
§6. Cast iron, the Saugus Ironworks, and the Bessemer revolution [21:06]
Cast iron is more than 2 or 2.5% carbon, more than carbon steel, because cast irons are eutectic. If I look at the melting point, cast iron — you'd like to have something low-melting to make it easy to cast. They call it cast iron because if you took iron ore — where you weren't looking for carbon — you would mix it with coal. Or originally, if you go up here to Saugus Ironworks, they mixed it with charcoal, which is carbon, a wood source of carbon. When you do that and try to melt it, the carbon will alloy with the iron and form cast iron.
In the 1600s they didn't have the phase diagram, but it was cast iron — it melted at a low temperature. They could then take that cast iron and beat it and forge it, heat it up — that's the blacksmith folding it over and making swords and stuff. As he did that, the carbon would burn off the surface, and if he did it enough — burned off enough carbon — he would get something even more malleable, more ductile. That was wrought iron rather than cast iron because he had wrought it on the forge. And if he did it long enough and didn't oxidize it all the way to nothing, he would end up with steel.
So historically they would take iron ore, reduce it with carbon, and end up with cast iron. Then they would work their way back across the diagram and if they're lucky they could end up with little bits of steel which were tremendously expensive — all that work that went into things. I've estimated the person-hours per ton in the 1600s was about 2,000 to 3,000 person-hours per ton of cast iron or wrought iron. Today it's about twenty minutes a ton. Look at that productivity gain. What was the big productivity gain? In 1856, Sir Henry Bessemer taught us how to build a vessel so we could burn the carbon out of the cast iron more efficiently without oxidizing all the iron back to the ore. That was the beginning of the Iron Age in 1856. And who was the Bill Gates of his day, the richest man in the world? Carnegie. Andrew Carnegie was, in his day, richer than Bill Gates or Warren Buffett on a proportional scale. And it was all because of the railroads and steel.
So the reason it's called cast iron is because that's how it was made originally. You take charcoal and iron ore, reduce it, and you end up with cast iron, which is right here on the diagram. All the cast irons are somewhere around — this is about 4.3 or something — I can't read it here. But you can have cast irons down here, you can have cast irons up here. You don't go above 5 or 6% carbon in cast iron because you run into other problems.
§7. Sulfur, phosphorus, manganese, and the chemistry of slags [24:30]
Student: [question about historical European steel quality]
In the United States we had some iron ores very low in phosphorus, and so we could make good steel. The Czechoslovakians were some of the best steelmakers in Europe, and their iron ore had a lot of phosphorus and sulfur. They could not get consistent results. It wasn't until John Chipman came along in the 1940s that people were really taught how to get consistent results. You don't just take iron and carbon and put them together. You put them together with a flux, and that flux might be limestone, it could be calcium silicate — which the geologist calls wollastonite, just a rock. They might throw sand in there. They throw these different things into the steel when they're making the cast iron, which you're going to then turn and burn the carbon off to make steel out of.
But if you have very much silicon or sulfur, you've got a real problem. The iron-sulfur diagram has a really deep eutectic with sulfur. Sulfur is right underneath carbon on the periodic table. The eutectic with sulfur is an even deeper eutectic than carbon's, and it's below the hot-working temperature where you're going to forge that steel. So you'd form liquid iron sulfides on the grain boundaries — they call it hot shortness or hot tearing. When you try to roll or forge the steel it just crumbled to pieces because of the liquid iron sulfide. Someone learned, if you throw some manganese in, the manganese will form stable manganese sulfides that melt above the melting point of steel. So manganese is merely a way to tie up the sulfur impurity that comes from your coal.
Then they had problems with phosphorus, and they had to learn to use what they call basic refractories — calcium-containing refractories. Acidic versus basic — I hate these terms, but the people were chemists back then, and they knew acids and bases. The silica-based slags were considered acid, and the basic slags were based on calcium and magnesium — calcium hydroxide, magnesium hydroxide. Silica would tend to form acidic mixtures. They make no sense to me today, but they made sense to people who were chemists working with water-based systems back then.
I went back to a fifty-year-old book on welding metallurgy, written in the mid-60s, and he didn't even list the basic oxygen furnace process, which is how all steel is made today, because it didn't come into being until the 1970s. Even today, you read specifications and it says it can be made by the acid process or the basic process — that's just a question of, you've got iron and carbon, and what is your other constituent? Your slag is going to be made out of wollastonite or sand — actually a bunch of these things — but what's the composition of it?
John Chipman came along and taught people — well, other people learned in the 1880s, you throw some manganese in. I read one book and they said the definition of steel is an alloy of carbon and manganese, because there is no steel that's made with zero manganese. In the old days, before World War II, there would be a rule when you're welding that your manganese-sulfur ratio had to be greater than about three, and preferably greater than six. You had to have enough manganese so you didn't get this iron sulfide intermetallic in your weld metal and have the weld break from cooling shrinkage stresses. We don't worry about that anymore. No one has sulfur that high anymore.
An ASM spec originally written in the 1920s or 1930s will say the sulfur must be less than 0.05%. By the 1940s they were saying less than 0.04%. By 1975, when Tom Eagar was working for a steel company, the sulfur was typically less than 0.01%. And the Japanese were making it at 0.005%. They had modern equipment and they were beating our socks off in quality of steel. 0.005%, 50 parts per million, was about the best they could do in the 1920s with any regularity. They didn't know how to control it until John Chipman told them what types of slags to use. I can go to an ASM spec today and they'll still say the sulfur has to be less than 0.05%. Are you kidding me? No one makes 0.05 steel anymore. They haven't made it since before 1950. Sulfur does a lot of bad things for toughness in steel, but we've learned how to make it better and better. We can regularly get down to 0.002 or 0.001 sulfur today if we want, but we don't usually need to.
§8. The Saugus blast furnace and the origin of "pig iron" [30:53]
Student: [follow-up question — would like to hear more about slag]
If you go up to Saugus Ironworks, they've got this big shaft furnace, and it's on a hill. This is the grade of the hill, and this is about 15 ft above the bottom down here. They would pour in charcoal and iron ore — mostly that's probably Fe2O3, they basically got it out of the bogs here on the seacoast. They'd add sand or limestone. Limestone is calcium carbonate. It could be magnesium carbonate. Sand is basically silica. It could have some magnesium, manganese oxide, titanium dioxide — lots of things in it depending on what you got.
You throw these things in the top of the shaft furnace — they used to call them, or they still call them, the blast furnaces. They blow the air in here. There's a great big bellows down here, and they have a water wheel pumping this bellows and blowing air in. You're burning up the charcoal, generating heat. It's really hot down here, it's cool up here because you're dropping the stuff in. You have a particle bed reactor. You're sending CO2 — or actually carbon monoxide — up the stack. And down here you'll generate, on a hearth of sand, some cast iron.
This cast iron will also have a slag that forms on top, and it's basically a molten glass — a calcium silicate molten glass. If you pour it out on the ground you can hit it with a hammer after it cools off and it shatters like glass. You can find obsidian, which is a natural glass formed from volcanoes, and it's basically just melted sand. This becomes the slag. The slag is where you want your impurities to go. Some of the sulfur will go up the stack, some of the phosphorus will go up the stack, but a lot of it gets trapped in this molten glass. So now it's a calcium-silicon-phosphoric acid glass, or it's got sulfates in there. It's just a complex liquid ceramic mess.
People in this part of the country could make really good steel because they had really good raw materials. They didn't even know what made them good. People in Czechoslovakia had to deal with all this high-phosphorus ore, and they were beating their head against the wall until someone figured out — oh, let's throw some manganese ore in here, we get better results, and it doesn't crack when we start to forge it.
I want to tell the story of pig iron. Does anyone know how the term pig iron came to be? About once a day, you would tap the bottom of the furnace. You'd have a little clay plug that you put in around this sand hearth, and you would go in with some steel rods and poke away that clay plug, and allow the liquid metal to flow out of the furnace.
If you go up to Saugus you see the guys — the floor around the caster was basically just about 2 feet of sand. They would dig a little trough in the sandy floor for the liquid metal to flow out through. So it's coming out of the furnace through a trough, and they would flow it into what you would call a manifold. Off the manifold they would have little castings. The metal would come and solidify on the sand floor — they've dug this stuff in the sand, it's like people at the beach playing in the sand. These were the pigs. It's called pig iron. They would make little castings — it might be a 40, 50 lb casting like this. It's a piece of pig iron.
This bar — anybody know what this bar was called? It's called the sow bar. What's a sow? Some other pig. Those are the piglets nursing off the sow. The sow bar might weigh 400 or 500 lb. If I'm going to make a cannon, I use the sow bar to forge the cannon from.