WM_S2014_07

Well anyway, so the phase diagram is such that for manganese and sulfur, right here there's a eutectic. There's two liquids that form at very high temperatures in the manganese sulfur system, and the sulfur-rich liquid, which is basically manganese sulfide, forms a eutectic and melts at 1570 C. That means it's solid in the liquid steel, because the steel melts at 1538. Turns out if you look at the periodic table, the only two sulfides that melt above 1500 centigrade — 1538 centigrade — are molybdenum and manganese. And molybdenum, when you alloy it with iron, forms the lower-temperature sulfur, this. Okay, this is the iron-sulfur system, and you can see there's a very deep eutectic, just like the iron-carbon system, except it goes down to 988 C. So when steel melts at 1538, you get a little sulfur liquid in your iron, or you get some sulfur in the liquid iron, and you will have liquid iron sulfide, unless you throw some manganese in.

Manganese is unique on the periodic table. It's the only element that will tie up the sulfur, okay. And until we had John Chapman and Nick Grant and other people who were faculty in this department, who taught people how to get the right slag composition to draw the sulfur impurity out of the liquid iron, people didn't know how to make iron-carbon steels with lower than point four sulfur, or much lower than point four sulfur. So they always put manganese in, and they still do. Manganese is not harmful — in fact, it's beneficial to strength and toughness and other things. It's relatively inexpensive, and it's really one of the only — it's the main use for manganese in the world, because on average steel has half percent to one percent manganese. If you've got half a billion tons, that's a market of several million tons of manganese.

So steels aren't iron, manganese, carbon. They often have iron, manganese, silicon, and carbon, because with sulfur and phosphorus impurities — I'm just putting impurities in parentheses — we usually talk about a carbon steel by just talking about these elements: iron, manganese, silicon, and carbon. The silicon is sometimes put in to get the oxygen out of the steel. Once you get the carb— burn the carbon out, you're burning it out with oxygen, which means you're throwing a lot of oxygen into the steel. If you don't get a lot of that oxygen out, you'll get bad properties. So you throw some silicon in to tie up the— actually make a little quartz particles, SiO2 inclusions in the steel, which are not terribly harmful. So that's some general steelmaking technology. Anybody have any questions on that?

Okay. If I go back to some of these phase diagrams again, transformation diagrams, we talked about hardenability. And hardenability is the relative ability of the ferrous alloy to form martensite when quenched above the upper critical temperature. And in fact, we want to know the hardenability in terms of the thickness of the steel that can be heat-treated to a given hardness. Hardening is sort of like Jonathan was asking about — can I harden just one surface or so of the steel? It turns out I can.

If you go to a book on heat treatment, and I'll pass this around since you can't read it, it's just the table of contents of a book on heat treatment. And most of the contents are on heat treating of steel. Why? Cause steel is 95% of all metals. So you have heat treatment of steel, you have surface hardening of steel. What do I mean by surface hardening? Well, Jonathan asked earlier about — if you get a case, you can just heat the surface of the steel. If you want a tough core, which is low-carbon, but you want a high-carbon surface because it's good high hardness and wear resistance such as on a ball bearing race, you may carburize the steel. You start with a low-carbon steel and then you heat it up with a laser or an electron beam, you carburize it in a gas or in a liquid, or you nitride it. You can do a lot of things to just heat the surface of that steel and make it very hard, and that's what we do to make ball bearings.

Then they talk about heat-treating equipment, heat treating the cast irons, heat treating tool steels, stainless steels, and then at the end of the book they have a chapter on every other type of metal, okay. No need to copy this — this was to show you there's lots of different types of hardening that we do because it's such a common material.

And the last time, I was telling you that we basically had — the hardness of the steel is a function of the carbon content, and I think I probably haven't handed this out. So here's a copy of this curve, and this shows the hardness of the steel. There's hardness and hardenability, and hardness is how hard — indentation hardness — the steel gets, and it's merely a function of carbon content. And about point six carbon it tends to level off. But the low-carbon steels are not very hard, the medium-carbon steels are intermediate, and the high-carbon steels — this is why we kind of talk about high-carbon steels are the high hardness, medium are intermediate, and low-carbon are low hardness. Merely a function of the carbon content. But the depth of the hardening, how deep I can do that when I quench it, is a function of the alloy content.

So if I need to understand whether I can weld a steel, I don't want to get something that's too hard. I want to get something that I can heat up and let it cool down and not have everything fall apart on me, okay. So what I do is I try not to lose my— [Tom searches for a handout.] I may need one of those back. Say, can I have all this back? I got one up here somewhere, but I don't know where it is.

So here's my transformation diagram in this Metals Handbook article here. My function of carbon, four — ninety-nine percent martensite, fifty — ninety percent and fifty percent martensite if I cool it down as a function of carbon. These are the cooling curves for the austenite-ferrite, austenite-bainite, austenite-martensite. And what we see here is it takes time. And right here, the A curve — I can cool and get one hundred percent martensite in an 8630 steel, which is a little bit of alloy, three-tenths of a percent carbon, but I have to cool within 10 seconds to harden something to a sixteenth of an inch deep. Well, that's still pretty fast, and it's got a little alloy content. If that was carbon steel, I would have to probably cool within one second to get a sixteenth of an inch deep. That's not very much depth of hardening, so far as that goes.

If I go to more highly alloyed steels, which is probably on the next page — well, the next page shows you some different types of heating or cooling. Quenching and tempering: you quench, get down to room temperature, you have to bring it up to temperature because the martensite's brittle. But that, when you — after you form the martensite, if you heat it up and temper it at somewhere between 300 and 600 degrees Fahrenheit, you can soften the steel and get something that's very hard and still ductile. You can do all kinds of other things. This is called austempering: you cool it down into an oil bath or a lead bath, hold it at this higher temperature, let it transform and then cool it down, you get bainite, which is the stuff that has very good ductility, but is expensive because you have to have this fancy quenching.

Here are some curves. This is a 1080 steel, a 5140, 1034, and a 9160. And you can start looking at these things and seeing the amounts of time that you have for transformation. These low-alloy steels, ten, twenty seconds, and the stuff will have transformed. But the problem with welding is welding can cool down in two or three seconds. And so if I weld a steel, if I do arc welding on a steel, I will end up with something — I can end up with something that's going to be very hard and brittle unless I do something to control the properties of steel.

So until I did some other things, for about thirty years I worked with a company that was a spinoff of mechanical engineering here at MIT, and they would call me up from time to time, and they would ask me, "Well, we're building, we're redoing something" — like Dr. Belmar [?] said, they're redoing something like the Longfellow Bridge. They've got this old steel in this old building, and they want to know how do I weld it. What's the welding procedure that they want me to develop, the type of heating and cooling that they have to have for some particular steel? And I say, "Well, what's the composition?" They say, "Well, we don't know." And I said, "Then I can't tell you, because I need to know how much carbon, how much alloy content is in the steel. I need to also know the thickness of the steel." And if they give me that information, I then go to this little book called Weldability of Steels, which is a fourth edition about 1985, and in the back it has about 20 pages of a table.

And the table looks something like this. You've got a copy of it now. There are 208 different steels in this table. I'm giving you page one. Steel composition — well, I need to know the composition of the steel. I need to know the carbon range, the thickness range. These are the first two things I need to know. Then they will tell me — and I will go over this tomorrow, I'm finally getting to weldability of steels — it'll tell me the preheat and interpass temperature for making that arc weld. It will tell me what happens if I use low-hydrogen electrodes, it'll tell me if I don't use low-hydrogen electrodes, tell me if I need a post-weld heat treatment, it'll tell me if I need to do painting. It will also go across, tell me what that grade of steel — what the nominal, what the spec says the composition should be and things. So that one little table, and about 10 years' experience as a welding engineer, and I can draw up a procedure as required by some book such as the structural welding code.

So here's a structural welding code, and we'll talk about procedures and things like that. But if I want to weld a steel, there are hundreds of them out there, and they all behave a little bit differently, and some of them are easy to weld, some are very difficult. So that's it until tomorrow.