DP_S2012_05

The material actually wrapped around on itself and you got blue on the outside okay, that was the little quiz the guy had. If you want to see another example of deformation where things are folding over, this is a little bit different. This was a tool that was probably in the MIT machine shop for fifty years. When they were working on the milling machine they just had a big steel bar — it was bigger than this before, I didn't want all that. But I was walking through the lab one day and I saw this thing with the mushroom head okay. So it's mushroomed, and you can see the deformation, you can see cracks forming. And I've had three or four cases where someone was using a hammer that had a mushroomed head and the piece broke off and flew in their eye, and they're blind in an eye now. So I kind of in my horror I saw that thing should have been thrown out about forty years before, when it started to mushroom. But it wasn't, so I confiscated it, and then I decided well why don't I just cut it off and I can pass it around class and show people deformation. Well there it is okay, there's deformation, and rolling over as something deforms inhomogeneously.

Now this was just because someone was taking this thing and they were pounding some work piece in the vise before they tighten it up to do the fine adjustment. You know, you got a 20 pound piece, your steel piece you're machining, and you need to move it a sixteenth of an inch, so you just sort of tap it with this thing. Well that had been tapped with that big heavy steel bar for about fifty years, and it started mushrooming, and it was basically an accident waiting to happen, when a piece chips off and hits someone in the eye or something.

So far as that goes, Carl pointed out to me that Kingdom Brunell, when I was talking about this press and the thirty-inch diameter cast iron shaft — Kingdom to Brunell is not a kingdom, it's a person okay. So it's the Isard Kingdom Brunell [Isambard Kingdom Brunel] is the guy's name I guess okay. And so I thought okay maybe this is someplace in the Middle East, but it's actually the engineer who designed this huge ship in the 1830s. So there's a correction to that. I'd never heard of him, but Carl had heard of it somewhere and finally Googled it and corrected me, so I appreciate that.

Beyond that, does anybody have any questions as we get started? If not, I want to ask you some questions. I've been talking about inhomogeneous flow, and I came up with about five different things that inhomogeneous flow affects. Can anyone think of one? When you have any inhomogeneous flow, what are the consequences of inhomogeneous flow in your deformation? Remember, if you got a Delta one or less you shouldn't have any inhomogeneous flow, but if you get to Delta of two to ten, you're going to get inhomogeneous flow. The material — you know, squares don't become nice square rectangles, they become parallelograms okay. And the parallelograms are not uniform in dimension from the surface to the center and everything. So I've shown you examples — what are some of the things that I've shown you that are the consequences of this inhomogeneous flow?

For example, what happens to the loads because of inhomogeneous flow? They're all the same: Delta is equal to one, it's p over 2 t y. Delta is equal to five, what is it? It's about two okay. So it increases the loads required for deformation. So instead of having a 10,000 ton press, you got to have a 30,000 ton press to make a forging okay. And that's because some little volume element in that forging first gets to form this way, and then as it flows around corners and stuff it gets to form back the other way, and finally it gets to form back this way. And you had to go through three shape changes to get one shape change right. So it takes three times the energy. Well that increases the loads. That's why the deformation zone geometry curve increases above p over 2 t y is equal to one okay. You get increased loads.

What happens to the microstructure? It becomes non-uniform right. I mean I showed you pictures of big grains and small grains all in the same thing. So it results in non-uniform [microstructure]. Well, usually we want uniform microstructure, and so inhomogeneous flow is not helping us in that way. It's not helping us because it increases the forces. I showed you how material can fold over on itself and create laps, so it can create defects. Should say can create defects, doesn't always.

Another thing — it's influenced by friction. So I've got some friction between my tool, my working tool, and my work piece. And if that friction changes, I'm going to get a difference in microstructure, I'm going to get a difference in loads, I'm going to get a difference in surface finish, I'm going to get all kinds of things. So I have to now have some way to control the frictional behavior while I'm doing the forming.

And the last thing I listed was can stresses — if you think about it, if I have this non-uniform flow and some of it's shearing once, twice, and three times, and other stuff is just shearing once to get that same shape change, in different parts of the part I got different amounts of work hardening. We haven't talked — so Dr. Mar [Mares] has talked about work hardening, but I have different amounts of work hardening. So the region that's harder than the other region, so you got differences. I guess that's another thing — different local differences in strength. So I got all these things going on because of inhomogeneous flow.

So I don't particularly want inhomogeneous flow, but it's just something that comes out of nature and I can't always get rid of it. That's not to say it's always bad. Sometimes I can make a forging — this is a forging, it's a steel forging, and of course this was etched fifty years ago by one of Backofen's students probably. But basically you can see the grain structure. The original piece that was forged was probably a block that had this shape — just a big cut-off end of a billet, and then it got forged and stamped into this thing. Looks like it's going to be some part of a turbine after they machine it. But it turns out it has a grain structure from the original bar that came out of the rolling mill. It has a grain structure just like a piece of wood, and you can see the grains flowing this way. But then as they forge it, this bar, and form these wings on either side, you can see how the grain structure flows around here. And turns out that if you do the forging right, you can end up with a very beneficial microstructure that will help resist crack growth. And I'll turn that over to Dr. Belmar [Mares] for another lecture about how differences in grain structure and that flow can improve the mechanical properties, particularly of forgings. And you don't get that advantage from a casting necessarily.

This is another piece, and it's been etched again probably fifty years ago. And if you look carefully you can see the grain structure going around the corner and stuff. So the grain structure — it's like a tree trunk, that the grain grew from the trunk up into the branch. And so the grains go around the corner, and so you have more resistance to fatigue cracks and whatnot in this corner, which is a place of stress concentration, because the grain structures are not running perpendicular to that okay. So you can get some beneficial effects from inhomogeneous flow. Not everything is bad so far as that goes.

Actually I'll do this. I talked a little bit about coining a couple of lectures ago, and seemed like there were some eyes glazing over when I talked about Delta of an operation. If I take something I wanted to form, and I take a die — this just happens to be a die — and I squeeze it into the surface, I just coined a beaver into this piece of clay okay. But I didn't really change the shape of this pancake very much. I just did local surface deformation. Coining is a high Delta operation okay, inherently high stress. I'm not really taking that pancake and making a thinner pancake, I'm just embossing relief or surface detail into it. So coining, if you look back on his little plot, is out at Delta is equal to ten okay. So you just take a die — this being a die that has some sort of features — and you want to emboss it in the surface. So coining is a high Delta operation.

So I've just kind of picked up a few loose ends here. Student: Does the rate of coining make a difference? Well first of all, lubrication is not going to make a difference in coining at high Deltas, we kind of said that briefly. And I haven't talked a lot about lubrication, but lubrication won't make a big difference. In general when you're deforming things, metals don't start getting stronger with rate until the rate starts approaching the speed of sound okay. The speed of sound being about 5,000 meters per second. So yes it does increase with rate, but only at ballistic velocities okay, which is not where we're doing all our deformation, unless you're doing explosive forming, which people do okay. But in general — and I read this all the time, people say oh well it was formed at a very high rate and therefore the stresses went up — you know, those dislocations are going through there at about a third the speed of sound. And so unless the dislocation movement gets impeded, and you're not going to do that unless you exceed the dislocation velocity, you've got to get to some substantial fraction of the speed of sound before you start seeing increased loads.

Student: So what you're saying that's wrong is recrystallization, like when you do hot rolling? I think you're referring to hot forming. Yeah, I'm talking about cold forming. Or whatever — if you have something — what Dr. Belmar [Mares] is talking about: if in the forming operation you're introducing energy and you get nucleation and growth of new grains, now you can go from a very coarse grain structure to very fine grain structure, and it can be during the whole forging process. So recrystallization and a change in microstructure could change things. And when we talk about superplasticity, we're going to see a huge difference in the effect of rate on the deformation stress. So there are some things. But if you're just talking regular cold forming of metals, you're talking about dislocation velocities, and those are approaching the speed of sound, typically about one-third the speed of sound in the metal okay. Is that fair enough okay.

So now it takes a lot more energy in the machine to go ten times as fast, right? Because the power input's got to be ten times as great right. Total energy is not different, but you're putting the energy in one-tenth of the time, and so therefore the power of the machine gets huge.

I did want to say one thing about deforming or removing residual stresses. Before some of you got here, there was a little discussion between — what's your name, Anthony — and Dr. Belmar [Mares] over stretch forming of sheet and things. And Dr. Belmar [Mares] pointed out that many times we actually will, in forming the shell of an aircraft fuselage which is sheet metal over a bunch of beams or struts or whatever — which you're eventually going to form them into, well, if it's a straight cylinder of the fuselage it's just a single curvature; if you're up at the nose or the tail it might be a double compound curvature with curvature in two directions, in which case you are stretch forming that sheet metal. And as he pointed out, when you do that stretch forming you're relieving some of the residual stresses in the sheet.

Well, they don't do that just with sheet, they do that when they roll the four-inch thick plate to make the wings of the aircraft. Turns out aluminum, because of its high thermal conductivity and the fact that it has to be heat treated and then quenched in order to get the best strength in the quench-tempered alloys — you can introduce significant residual stresses in the aluminum plate or big aluminum bar. Remember that little washer I showed you that was supposed to be stress relieved? Well, if you go out to Davenport, Iowa, where Alcoa has the world's largest rolling mill okay, and the rolls are probably as wide as this room and they're about eight feet tall okay. It's bigger than any steel rolling mill, because they're rolling the sheets for the wings of the aircraft. So if you got a 747 or something, they're going to roll some great big sheet. It might be four inches thick, could be a little thicker even. You can use some really big ingots to roll aluminum. Aluminum's so light, if you try to do that in a steel mill you have a little bit of a problem with the crane capacity okay, but in an aluminum mill you can have — start with really big ingots. We won't get into casting technology for aluminum, I think we cover a little bit of it in my casting section.

But in any case, so they roll these plates down, and when they're all done they have significant residual stresses. So they have a machine — big hydraulic machine — that will grab that plate that might be twelve feet wide okay. And they grab it with these jaws and they stretch it anywhere from 1 to 3% to relieve the residual stresses. Is that 51 on like a T61 — on a T651, that's the 51 if I remember, is a 1% stretch okay. And the five is stretch straining, like mechanical stress relief. On steels we usually use temperature for stress relief; on aluminum alloys we usually use mechanical stress relief. They'll take bars that are eight inches diameter and they just put them in this great big machine, stretch them a little bit to relieve the residual stresses okay. Student: Which way? Stretching in the direction — rolling in tension in the rolling direction. Yeah, so in the long direction. So if you got a bar like this or a sheet like this, they're pulling it in the long direction and it relieves the residual stresses.

And that washer I showed you — they bought it in the T651 condition. It had been stress relieved, and then they cut a piece off, you know, washer shaped like this, and then they reheat-treated it. And when they reheat-treated it they reintroduced the residual stresses, and they never did anything to relieve the residual stresses. And they said oh well we bought it stress relieved. Yeah okay, but then you reintroduced residual stresses, and so when you actually shipped it it had 8 ksi residual stresses in it. And stress corrosion cracking for that alloy is 5 ksi, so they got cracking of the washers. They changed the heat treatment to go to an overage condition, which now you have like 20 or 25 ksi, and so you can tolerate 8 ksi residual stress.

But what do you do if you have something that's nice and flat like that, to relieve the residual stresses? You don't do a 651, you do a T63, which is a compressive stress relief. So you would compress it 3%, you know, you pancake it 3%, rather than stretch it. It's still mechanical stress relief. The parts that are high compression you will yield, and the places that are low compression, when you do that compression you will not yield, and so when you're all done you've smoothed out all the peaks and valleys of the residual stresses going across the part.

And I mentioned, I think I mentioned, maybe it was at one of the breaks or I mean after the class, that — did I tell you how the US Navy stress relieves the submarines? I don't think I told the class, I think this was the conversation right afterwards. I told you that steel is usually thermally stress relieved. You build a big pressure vessel out of steel and you weld it, and the boiler and pressure vessel code says that if the welds — if the steel is more than one and a half inch thick, it must be post-weld heat treated, because the welding process introduces residual stresses. One and a half inch is thick — the residual stresses will be equal to the yield strength of the steel, and that steel could now fracture in a brittle manner when ordinarily it would have been ductile. And I have a crane component that fractured and a seventy-ton water bag fell on top of a man. He hurt his back pretty badly, because they didn't stress relieve the welds, which were about two inches okay, on steel that was about two inches thick. No post-weld stress relief, you had residual stresses, and when this thing got bent, started getting a bending load that it wasn't supposed to get but it got it, instead of just bending over like a paper clip, it snapped. Not quite like a piece of glass, but it snapped, and the thing came down while he was underneath it. Now he shouldn't have been underneath it, but nonetheless he was, and he got hurt. So stress relief is important.

So what does the US Navy do? You can't take a submarine with all the components and stuff and stick it in a big furnace. That's what they do with a pressure vessel for something like a nuclear reactor — and that pressure vessel for a nuclear reactor is twice the diameter of this room and three times as tall, and they have furnaces that are that size. And they'll stick it in there at 1200° F and they'll stress relieve it okay. And if they don't have a furnace that size, they will build a furnace around it — a special furnace — and they can do this in the field, in thirty-foot diameter vessels. But you can't do that once you have all the equipment and the O-ring, the rubber and the plastic and everything else inside the submarine. So how do they stress relieve it? The first deep dive — you got hydrostatic compression and all those welds get stress relieved mechanically on the first deep dive. And I also told whoever it was — that's one of my favorite stories about quality control — because the requirement is that all the top management of the shipyard has to go on the first deep dive. And they make sure that's a quality submarine okay. It's one of the best things I know. The threat of extinction really sharpens, focuses the mind, right.

Anyway, so in any case, this week I'll be lecturing on Monday and Wednesday. Dr. Belmar [Mares], are you doing all the — oh no, Tuesday I'm doing Tuesday, which is a Monday, today, and Wednesday. So I'm doing today and tomorrow, and you're doing Thursday and Friday. Next week I think I'm doing Monday and Wednesday, and Dr. Belmar [Mares] will figure out if he's available all three days. I looked at the schedule this morning and it looks like we might finish the lectures before spring break, as I predicted we might. If not, we're not going to go very far beyond it. Someone was asking when are you going to do your presentations — well, I don't even know what your presentations are yet. That's next Wednesday you have to tell me, you have to fill out that little form and tell me what you want to present on. And then you may have to schedule sometimes to come see me one-on-one, and we can talk about that, or we can talk, spend time in class talking generally.

In any case, I will probably — we're probably only going to do two presentations a day, so it'd be seven or eight classes. Someone had to go somewhere for Easter or something — I'm sure we can work around your schedules. In fact, there's even people who are not in class, who can't make it to class because of some scheduling conflict, they're taking it by video — that's why we're videotaping. But they're going to have to find a day to come to class okay to make their presentation okay. But actually the presentations are sort of fun. We actually I learn from them, and hopefully you'll learn from some of these things too okay. Any questions on schedule and stuff?

If not, I'm going to do a little more on inhomogeneous deformation until you get tired of it. Oh, maybe you already are, but anyway. So here's the inhomogeneous deformation, just a circular grid patterns, and what happens to the shape of those. And then what I didn't talk a whole lot about was the formation of a dead zone. If your die angle is not right, then nature will say that the easiest shearing plane may not be the frictional plane of the workpiece. There may be a lower energy shear at some other angle, in which case you'll just shear the material — the full friction of shear of the material. And let's face it, full frictional shear, I mean the material shearing itself is sort of like having a friction factor of 0.5 okay, because the shear stress is about half the yield stress.

And so the material, if it decides it's going to find its own shear plane rather than the work piece surface that you gave it — if your friction is too high and nature decides that it's easier to shear along a plane at the shear stress half of the yield stress, then it will pick its own and you'll end up with dead zone of material. And I showed you some pictures of that type of stuff okay. There are some cases where you actually try to achieve that, and where you're shaving the surface of the material. For example, you want to get really good clean surfaces on wire sometime — for example they do this with welding wire — they actually will draw it through a die that has an angle that will just skim shavings off the surface, so you'll get rid of certain types of surface defects and things like that. So anyway, we have those types of defects which form.

If we want to look — and this comes out of Backofen — if we want to look at the types of strains, let's look at the strains down here. This is an extrusion that Backofen shows. And so I've got about a — I don't know if this is actually axisymmetric or plane strain — probably axisymmetric. It looks like it's about a third across here compared to all the way across here, which means it's a 9:1 reduction in area — you square it right, square one-third and you get one-ninth. So it's got nine times the length as it did originally. And these are equivalent strains. And right here the material back here has not started to flow significantly, it's got zero strain. As you get closer and closer to the die, we actually have a dead zone here — if you start seeing this stuff kind of curl back on itself, so there is a dead zone over here. The stuff coming right down the center, this is the amount of stretch it's getting as it goes through. It gets up to a strain of about 200%, yeah 200%. This must be plain strain anyway. It's an equivalent — unless it's a true strain, could be a true strain, we'd have to go back to the paper.

But in any case, we actually have regions over here where we have the boundary layer in the friction, where we're getting so much redundant deformation, the stuff is work hardened, and we basically are getting strains over here of four. We have twice as much strain there as we have here, a measure of inhomogeneity. If you actually look at this plot, the local — where's the straining greatest? We actually have dead zones over here. The straining is greatest right as it starts entering the die, and you see these patterns of strain here. Actually, I'm sorry, the straining is greatest right over here. It's 0.9 here, it's 1.5 as it turns this corner — the stuff is really getting stretched. This is an incremental strain okay, this is a Delta Epsilon plot — incremental strain.

Student: How does the dead zone — is that visible on the outside of this part? You would see it. In fact I actually showed you something — let me find it. [Tom looks for a sample to show.] Is that etching, or is that — that's etched. This is brass. And you're seeing all the high strain right in here, right on the corners. Here's the dead zone — that's the original casting okay. That's all dead, everything's being sheared right along here okay. This one is actually axisymmetric if I remember, it's a billet. Student: This one needs to be cut to be seen? This one was cut to be seen, yes. This is a cross-section. Student: Visible on the outside? You might actually see large grains if you etched it on the outside. You'll see a different grain structure here than here. This is fine grain okay, this has been deformed — locks the stretch. This is starting to — this is getting deformed so much that the grains are pancaked, and if you went to do a heat treatment on this you would find this heavily worked material — I'm trying to figure out why these green laser pointers — this heavily worked material will recrystallize, whereas this will not because this has never been deformed, it has no stored energy of cold work. This is heavily cold worked. This stuff may or may not recrystallize, but if it does you're going to end up with a bunch of different grain sizes.

Now hopefully when you've extruded the whole thing and it comes out, everything's so deformed that you get — if you deform it enough, everything becomes homogeneous again okay. But as it's going through the process it's not homogeneous okay. And I actually showed you once before that if you do certain types of wire drawing, and you try to do just like a 10% draw rather than a 40 or 50% draw, you actually can make things very inhomogeneous, whereas a 40, 50, 60% draw, you start to get homogeneous again because everything's deformed so much that it's got a lot of work in it, and when it recrystallizes it gives you a nice uniform structure okay.

So you know, the worst possible situation is some sort of intermediate level or low level of work. You actually have to put a lot of work into the material to get a uniform product out. But that's partly because — you want to, in order to get a uniform product, in some ways you want lots of extra deformation okay.

And in fact, remember the trolley wire I sent around? Phelps Dodge made that by something called the Conform process. It was invented over at a National Laboratory in Great Britain, but Conform was a great big hydraulic machine, and you could throw powder into the machine, and it had a circular die, a circular cavity, and it sort of was like kneading. It would take all those powders, and they get 1,000% equivalent plastic strain. And the only thing — you can put lead through there, you can put aluminum, you're starting to reach the limit with copper, but the loads are so high that if you try to put steel through, all you're going to do is crack your steel dies. The steel is not strong enough to deform steel that much. You can deform aluminum because it's soft, you can sort of do copper, you can't do brass very easily, you'll crack your dies, you'll have almost no die life. But the stuff comes out, and you may have an effective reduction in area of 80% from some rod you put in. But the actual — I'm sorry, you may have an area reduction of maybe 80%, but the cold worked redundant deformation could be 99% reduction in area. You have just — it's like taking it and kneading that dough over and over in redundant deformation over and over. Very energy intensive process, but you end up with tremendous amount of stored work, very fine grain size, until that stuff was almost super plastic, which is why it had such good elongation. In fact, so much that the British guys who wanted to make an extra two million dollars profit said oh, you don't make our spec because you had too much elongation, you were too ductile — as if having too much ductility is a bad thing okay. But anyway, I don't know that any of them went to jail, but I think they were certainly scared when they lost their passports and things.

So if you're going to get redundant deformation, it's best to get a lot of it okay. Which is why to a certain extent you can't start with things and just do 70 or 80% deformation. You really like to get above 90% deformation on something, if you're going to get uniform products.

One of the things that we do — it's not very good focus is it — let's see if that helps — is we actually do a ring test, to try to get the friction coefficient and the amount of barreling and other things. You basically just forge a little ring — a tube, thick wall tube — and you'll get a change in the inside diameter, which could get larger or smaller depending on the friction boundary conditions and the outside diameter. And it's called the ring test, and it is a test for forgeability of a material without going through — you can get, if you have low friction you'll get something that turns into a straight-sided ring; if you have high friction you can decrease the internal diameter, you get bulging on the sides; and you get some idea of how forgeable your material might be with a relatively simple specimen.

Now people have studied this test in some detail, so that they can actually tell you — if you look at the decrease in the internal diameter of the ring, which can be negative which means it gets larger, or positive which means it gets smaller, versus the amount of deformation up to 60, 70% squeeze okay. So if you start something that's one inch high and you end up with a third of an inch, you're down to 67% deformation right, so it's one minus — anyway. These are the various friction coefficients. Mu is the friction coefficient, from 0.57 all the way down to zero friction. You'll basically get an increase in the internal diameter because there's no friction. With lots of friction you're going to essentially take that internal circle hole and forge it together okay. So you can actually do this test and get an idea of what your friction coefficients are, which is actually a fairly simple test and simple way to get an idea of your friction coefficient.

Because if you're doing this on some nickel-based super alloy from some jet engine, you're doing it at 2200° F. It's kind of hard to get in there and do a lot of friction tests at 2200° F. It's much easier just to kind of smash it, look at the internal diameter, external diameter, thickness reduction, go to that plot, and say okay, which — where does it fall on the friction coefficient? Because what are the lubricants we use? Glass okay. You spray on a glass frit, and there are people like Acheson chemicals [Acheson Industries] who make different types of glass for different forging temperatures okay. Because you want a glass that's still viscous, more viscous than honey, but not so viscous that it's still brittle like glass. And you don't want it so fluid that it just flows away, and you don't have anything that's going to stay there at the interface. So you got to have — so glass is one. The other is they use a lot of graphite compounds okay. Because graphite, it might burn in the air eventually, but you're coming down with a big forging press and wham okay. So the big forging press doesn't take long — you basically got the whole thing occurring in a fraction of a second.

Anybody know what the biggest forging presses are? Okay, the biggest one I've ever seen, and there's only two of them — the Air Force built them in the early 1950s, because they were starting to build bigger and bigger bombers and they needed bigger and bigger parts, forged parts, because they didn't trust welding. And they still don't. And so they basically built two 50,000 ton presses, and one of them is right here in Grafton, North Grafton, Massachusetts, at what used to be the Wyman Gordon plant. It's now owned by Krupp Steel, a German steel company — they bought Wyman Gordon. The other one is located in Cleveland. Anybody from Cleveland? Okay, Alcoa owns that one. Alcoa has a forging plant in Cleveland and they do great big aluminum forgings for aircraft. The Wyman Gordon plant tends to do titanium and nickel-based super alloys. It's a 50,000 ton press.

And I've been to it once back in the early 80s when it had a crack. But the base — if this is the forging floor, which is basically a dirt floor, and you have these columns which go up about five or six stories okay. And I don't remember — they're about four feet in diameter, it's probably five or six, or four, six of them or whatever. And you have a big hammer up here, that's a hydraulic hammer, that's going to come down. The base underneath goes down for six stories. I had to go down in an elevator to get down there to see one of the beams at the bottom okay. They had a series at the top, and at the bottom they had a series of I-beams which had a cross-section that looked something like this, and each one of them was forged. The top surface was machined within ten-thousandths of an inch because it had to be the machine base. So here's your I-beam giving some depth into the board. This was about the thickness, here was about — if I remember, nine inches the web thickness — and the whole thing was like six feet tall okay. It weighed if I remember 300 tons for each one of these beams. And there were six of them across the base okay, and there were six of them across the top. And one of them had developed a crack at a weld over the at that time thirty-year history of this thing being a forging. And the estimated cost to replace the whole unit was two billion dollars in like 1983 okay.

But they could make forgings half the size of this room, if you wanted, that weighed 50 or 100 tons, out of some pretty fancy materials at temperatures of well above 2,000 degrees okay. The dies would be the size of this room okay. The anvil — the base anvil, which is the bottom die — would be approximately the size of this room, probably the thickness, the height of the room and stuff. A lot of these anvils were cast.

I once had a fracture in a 33-inch diameter propeller shaft on a liquid natural gas tanker. And I was kind of — I was a young — in my 30s — I was kind of teasing Professor Pelloux that — Professor Pelloux is actually still alive, he's retired, but his specialty was fatigue and fracture — and I said I bet you I got a bigger fracture than you've seen, because 33-inch diameter shaft fracturing is pretty good size fracture. He says I don't think so. I said what do you have? He said he had an anvil base to a big forging press, and the fracture was like 10 feet by 20 feet okay. That's a big fracture. You start figuring that if you had to pull it in two at, you know, 50,000 pounds per square inch, there's a lot of square inches in 10 by 20 feet okay. So they exist.

Anyway, so the biggest forging presses are 50,000 tons. The Air Force had plans in the '50s to build a 100,000 ton forging press. But in fact they couldn't just build it anywhere — they found a granite mountain in Colorado for the base, and they were going to build this thing in the base of the mountain because they couldn't afford the foundation. I mean this thing has a huge foundation, it's going down like ten stories into the ground right. And anyway, um, they never built it, because basically what happened was electron beam welding came along, and so it became more economical to design things for electron beam welded. These are things like the center box for the first swept wing fighters. Remember they had swept wing fighters where at sonic speeds they'd be like this, and at supersonic speeds the wings would go back. So that whole box that was kind of — if you think of the wings as your arms, it was your chest part, you know, and this was your shoulder, you know, of the aircraft, the shoulder joints — they learned to make that by electron beam welding okay, in the late 50s. And so they never built it, but they had plans to build it okay. But it would have cost a fortune, I mean in the 1980s these were going to be two billion dollars okay.

To give you some idea of the types of friction coefficients you get with these graphite and glass lubricants for forging — this actually is out of a friction in metal forming book. What a great title right? Everybody wants to have a copy of that at home. And here are some of the alloys: so 6061 aluminum, workhorse aluminum alloy; titanium alloys; stainless steel; Waspaloy, which is a high temperature nickel-based alloy for jet engines; precipitation hardened stainless steel; workhorse titanium; Inco 718 [Inconel 718], the workhorse nickel — aluminum alloy, a lot of it used in jet engines; Udimet, a nickel-based super alloy; 7075 aluminum. So the specimen temperatures are 800 — well you can get here 2100 is the highest temperature. I told you 2200. Well, the highest temperature here. But the frictional shear factor as measured in a ring test was anywhere from 0.18 to 0.53 okay. And the contact time — you was asking how long, someone was asking — the contact time is a fraction of a second, is measured in tens of milliseconds, if you will. The ring ratio — and this is outer diameter, inner diameter, and reduction or whatever. And the lubrication system B is basically a graphite with a caustic pre-coat, and C is a glass okay. These are all Acheson types of things.

But they can get these things, and then the friction is going to have a big effect on the loads — how many tons you have to have. I mean, you have a lubricant with a lower friction factor, and you might be able to forge something with a 10,000 ton press as opposed to a 15,000 ton press okay. These presses are — I don't know that I've seen any really big ones made recently. I mean I was working on a weld repair — actually if you take the welding course I bring in a 20-inch thick weld repair, and that was a weld repair on a big head of a forging press, that was like a 7,000 ton forging press. And that stamping, that at the time was stamping out truck wheels, aluminum truck wheels. So if you want to go look at a truck wheel, eighteen-wheeler on the highway, look at the size of the forging — 7,000 ton press to make that aluminum forging okay. And the dome of that was about ten feet high, and it was 80% of it was square and it was 80% of the way across the room in the short direction. So it would have filled up a third of the volume of this room and weighed about 350 tons if I remember.

Student: So you talked about friction as being a good thing to reduce load, but is it also fair to say that it would reduce your inhomogeneous flow? Not all the time. I mean it really depends on the Delta factor. I mean if I have higher Delta and lower friction, it might be the same as lower Delta and high friction. So there's a trade-off between friction and Delta okay. The height working ratio. If you end up with low friction and low Delta, it's going to be more homogeneous, no question okay. But I can't just say in general, because the two of them are competing. High Delta, high friction make it more inhomogeneous; low Delta, low friction make it more homogeneous okay. So low friction usually means more homogeneous. Sometimes you want more inhomogeneous because you need that extra kneading to get the grain structure refinement and things okay.

So there's a lot of experience and art to these things, there's no question about that okay. People have big fancy models, and the finite element models for predicting the deformation in forging are not bad — a heck of a lot better than sheet metal okay. Sheet metal is still a very difficult computationally problem, and people have tried, believe me. That's the bread and butter of the automotive industry okay, is the stamping process to stamp out the sheet metal. And if they could — you know, well, a set of dies for a new high volume automobile — and high volume is 100,000 units, 30,000 units a year, 100,000 units over three years — set of dies is somewhere between 100 million and a couple hundred million dollars okay, for the sheet metal forming dies okay. That's just for the tooling. And then you got to pay operating costs and steel going through it and everything else. So it's pretty impressive.

Any other questions on that? One thing I did want to show you is that when we work things, they get hot okay. You're putting a lot of energy in there when you're deforming these things. And this is from — I don't know, it's probably measured somehow — but wire drawing in some die, and the material is coming in at 86° F and it's coming out at 136 — 130° centigrade. You got a 212-degree temperature rise due to the friction. And you can look at the friction on the dies. Now what does that mean? Well if this is in wire drawing, they often use an oil-based lubricant, but the lubricant that works at 30° C has also got to work at 300° F okay. So 86° F — and not a lot of your oils can go above that. You are limited in your drawing speed. I mean when they're doing wire drawing, it's sort of impressive to see these things spinning at sixty miles an hour. They're drawing at sixty miles an hour, or they would go higher except their lubricant will break down and turn to carbon okay. So you're limited by your lubricant and how hot it can go. But those are types of things, and that means your frictional boundary conditions are changing a lot so far as that goes.

Let's talk a little bit, so I can try to finish up forging here in the last couple of minutes. Different materials forge differently. Classification of metals in order of increasing forging difficulty: tungsten alloys are most difficult, nickel-based super alloys are right behind them, molybdenum alloys — molybdenum melts like 2600°C, tantalum and niobium are both up there — well, tantalum is 3,000 centigrade, niobium's like 2400 centigrade melting temperatures. So you got your iron-based, nickel-based super alloys; titanium's in the middle; simple nickel alloys are not so bad. The easiest of all is aluminum okay. And low carbon steel — carbon and low alloy steels — is actually fairly easy.

What you can deform — this is an aluminum kind of I-beam type of shape. This is what you could do if it were a nickel-based super alloy. You obviously can get lots of metal flow in aluminum for one forging. I mean this started out as a disc and ends up as an I-beam; this starts out as a disc and ends up as a dog bone. So you got to have more dies if you want to end up with this kind of shape. You got to have like four or five dies in the nickel-based alloy to get to the same shape change that you would in aluminum alloy.

So if you look — another way to look at this is the increasing flow strength of the forging pressure, decreasing forgeability. The nickel-based super alloys very difficult down here. The easiest stuff is aluminum, carbon steel, alloy steel. Stainless steels are kind of in the middle, and you got all kinds of other things around here — here's Waspaloy and aluminum and stuff. So it's all — I would say that forging twenty years ago was all sort of empirical. Because of the advances in computers and finite element analysis, it's not all empirical anymore. You can do some pretty good calculations of the forgeability of things for deformation processing. But it still takes big presses, very expensive equipment. But it ends up with something with lots of deformation, lots of homogeneity, good grain structure in terms of fracture resistance and things like that. So forgings are usually considered high quality parts okay. I'll see you tomorrow. Have a nice rest of your Monday.