FW_Su2013_03

So in flames, we use flames for straightening things a lot of times. You put something together and you start looking at it and it just doesn't fit, okay. For example, here's a drawing. By the way, this comes out of a National Shipbuilding Research Program. This was in the — 20 years ago, it was August 1991. They had a program funded by the Navy and the Maritime Administration to take best practices in shipyards and try to publish them and let people know how to build ships. And so I used to get these regularly for free.

And so here's a situation. You got some valve body down here. This is probably a real situation. Real, real situation. And they basically do all the piping and they find the things don't fit together. Now these are big pipes and big flanges and you don't just kind of push them with your hand to get them in place. So what do you do? Well, you could cut this out and put a longer piece in, or you can do flame straightening or bending, okay. And basically what this handbook tells you is, if you want to bend pipe — this one was specifically on how to bend pipe — you basically make a flame heating with a little V shape at different locations.

And wherever you heat — one rule in residual stresses that you should always remember, and I've never seen it fail yet: if you heat the thing up, by heating it you will cause it to expand in one direction, by the time it cools down it always ends up taking a residual strain in the opposite direction. So if I take something and I heat a triangle like this, it's going to expand more on the outside than on the V point of the triangle, right, the heating triangle. That means this thing, while it's hot, is going to expand and cause this thing to bend over to be even worse out of alignment, right, while it's hot.

But in fact, what happens is, this colder part at the bottom of the V and the hot part up here — the hot part's trying to expand but it gets compressed and you get above the yield point of the hot material and it gets compressed while it's hot, on your right side. The left side is non-yielding, it's not hot enough to yield. And so this material, as it expands, actually is compressed compared to its room temperature state. And when you let everything cool back down, this shrinks, the wide part of the V shrinks, the bottom V never changed its shape, and so the thing pulls back into alignment, okay. In the relaxed state, in the cool state, you get residual stresses because you have temperature gradients in the material and you get to temperatures where the material is getting above its thermal yield strength.

So welding, we get residual stresses. Pipe bending, we put residual stresses in. I had a — there was a guy who used to be head of the Welding Department of Ohio State University, and he told me — he was an expert, Roy McCauley, has passed away now — but he told me that when they were building the Cincinnati Red[s] Stadium, okay, this is back when Pete Rose was still playing, so this is a long time ago, they had the world's largest I-beams for part of the roof section. And while they were trying to erect this, one of the cables on a crane broke and this beam came crashing down right across this big steel — not steel, concrete wall or whatever — and they ended up with a beam that looked like this rather than a straight beam. It was more than a year — like a year and a half was the lead time to get a new beam from the steel mill, because they don't roll that size every day, and you can't go to some hardware store and pick one up. They're not exactly in stock, okay.

So what were they going to do? They were either going to have to fabricate one out of plate, which would probably still take six months, or — Roy came in, showed him how to flame straighten it, okay. And they took something that had like a 30 degree angle on an I-beam and he straightened it right out by applying heat. You can apply the heat over and over. It points out, they generally like to limit the heating to three heats when practical. You can try to do this and keep jacking it over, you know, three times. Although you can do it more times, you cannot do this on heat treatable materials because you'll destroy the heat treat. So on carbon steels, most of your piping is just plain old carbon steel, you can do it. You can't do it on the Hy steels, you can't do it on your heat treatable aluminum alloys in most cases.

This is another book by a practical guy on flame straightening technology. He shows — he did a lot of work on castings. And he shows that — oops, let me show — okay, you see it, he had a casting that shows this kind of an I-beam, and after it came out of the mold it had a high spot right here. Well, you put the heat on top of the high spot and make the high spot higher while it's hot, but when it cools back down, it straightens out. And he's got rules — both of these — this has kind of rules of thumb, this guy has here up here. This book has the potential to save the metal working industry millions of dollars, which is in fact true. The difference between that book and the Maritime Administration books is, this one's quantitative. There's actually some science behind this, okay. There's a lot of practical art behind the other one. The guy was — he worked in the foundry and he had to straighten things all the time. Yes?

Student: [question inaudible — about residual stress relief requirements]

It depends. The ASME boiler and pressure vessel code does not require residual stress relief of welds less than about an inch and a quarter thick, okay. But when you get to very thick welds — actually you get above about an inch and a quarter, inch and a half — at an inch and a half you have to stress relief the welds, because otherwise you'll get brittle fracture from the residual stresses and you'll get fatigue cracking and stuff. So ASME pressure vessel must be stress relieved when you're above, say, inch and a half, okay. If you're below a half an inch there's no need to relieve residual stresses. Most structures have enough flexibility when the walls get that thin. And in fact, residual stresses and distortion are sort of a trade-off. The thinner the material, the more distortion, and the thicker the material, the stiffer the material, the more residual stress you build up. It's basically caused by the same thing — contraction, temperature gradients, and the material contracting differently in one area than in an adjacent area, and the material yielding in one area locally as compared to another area. Yes?

Student: [question about submarines]

If you're thinking about submarines — you don't stress relieve the walls of submarines. When you cut out a hatch, you put an insert plate. If it's more than an inch thick, you should probably think about some sort of stress relief. Now, there's stress relief by mechanical means, not always by thermal means, okay. And what I was going to point out — we can't talk in here on videotape about how thick the wall, the hull is of a submarine, but we all know it's more than an inch thick, and everybody up there knows too, okay, around the world does. In fact, they have to cover the modules down at Electric Boat, because the satellites up in the air can actually measure the thickness of the hull from a satellite, okay, within a fraction of an inch. Pretty amazing, but they can. And if they know the outside diameter and the wall thickness, they know what the dive capability is — okay, depth capability of that. I happen to know what some of those numbers are, but it's classified, right.

But — nuclear submarines, the hulls are not stress relieved, are they?

Student: [response inaudible]

Well, you create all kinds of residual stresses, but they are stress relieved. The first deep dive stress relieves the welds mechanically, okay. So if it was a surface ship of that shape or that thickness, you would have tremendous residual stresses. And so they actually — this actually brings me a reason I brought this in. This is not part of a battleship, okay, it is an arc weld on a fairly thick plate. Came off of a huge forging press. Oops, I better do something, that's getting rusty, but anyway, we can pass this around.

But this is about a 13 inch thick weld made with — lots, hundreds, I think I estimated five or six or eight hundred submerged arc welding — not somewhere — SMAW, manual metal arc welding passes. And we're gonna — oh I'm sorry, I'm not even — look at the weld, the weld is over here, okay. So the weld is — I'll pass it around but it's not the lightest thing — the weld is here, this lighter area, and it goes all the way down. Well actually goes all the way down from that tip all the way up here, okay. So this was a forging press that had been made in the 1950s. It developed a crack. If you had to rebuild this 7,000 ton press, you'd probably cost you 100 million dollars. So they wanted to repair the casting. So they brought in some good old boys from Tennessee — well, it's happening in Erie, Pennsylvania — they brought in some people, you can actually see the weld beads on the top here across the top.

[Tom passes the sample around with a rag.] And you can pass it around with this old rag just to keep from cutting yourself or scratching the table, and you can look at it. But that's the type of weld they used to put in armor plate on battleships. Now, you don't use battleships anymore, but it wasn't that many years ago, 20 or 25, they had the guy who blew up his — I guess you call them partner now — on the Iowa. People remember this? So this seaman had a boyfriend, and I guess they broke up, and so he wanted to kill his boyfriend, so he lit off one of the charges inside one of these 16 inch guns on the Iowa. Did a fair amount of damage when you light that off inside the gun turret.

And the — I remember this was back in the 80s, and I used to go down to David Taylor [Naval Ship Research and Development Center], Annapolis, during the summer, and they were trying to figure out how to repair it. Because no one knew in 1985 how to make the welds they were making in 1945, okay. What they used to do is they would weld — put in a layer, a certain number of layers — and then they come in and they mechanically peen the surface. And then they put in some more layers and they mechanically peen that surface, and that was their stress relief. They would mechanically stress relieve once they got probably every half inch. But no one really knew how many to put on and then peen, or how to do the peening, or how long to do the peening. And it turns out they decided not to repair that gun turret on the Iowa.

Now, within five years they decommissioned all the battleships. But most of you may not have even been born, but in the mid 80s they had the Missouri — Missouri was the last battleship built, if I remember. Anyway, they sent the Missouri over to Lebanon and they were shelling the Palestinians or the Arabs in Lebanon or whatever with 16 inch shells. And those shells had been made in 1940s, okay. So they kept that ordnance pretty well. But in any case, battleships became sort of an anachronism. And they had 16 inch armor, thicker than this, not a lot thicker than this, but something similar to this. And they would just have a bunch of little welders go all over that ship just keep putting on weld passes with stick electrodes. I mean tremendous expense if you try to do it by today's technology. I mean, if you use that technology today to build something up that big. But it doesn't really matter, because who cares if you've got 16 inch thick armor — all you need is the right shape charge weapon to go right through it, okay. The molten copper would just eat right through it.

I mean, I've seen down at Aberdeen Proving Ground where a shape charge went all the way through 36 inches of steel, okay. I meant to bring in my shape charge thing this morning, I forgot to do that. But in any case, everybody know how shape charges work?

Students: [murmured response] Vaguely.

Okay, well why don't I waste some time and talk about shape charges for a second. Shape charges were discovered in the 1880s by a guy working — I think at the U.S. Navy Brooklyn Navy Yard — and he was just working on ordnance and explosives, and he found that if you have things in the right configuration and set off some — an explosive — you can actually cut right through metal, or you can emboss metal in another metal. And in fact there's a picture — I used to have a book but I loaned it to one of your predecessors and he never returned it — but it actually from the 1880s, it has this "USN" that the guy embossed into a piece of steel using an explosive, okay, with a shape charge.

But a shape charge — the simple — the improvised explosive devices in Iraq and Afghanistan are basically just steel pipes. Regular old — they started out with four inch diameter pipe, and they put a piece of copper in here, very precisely machined copper — more precisely machined than any Iranian, Iraqi machine shop could be — a good machine, okay. They were importing these from somewhere — I could tell you which country, but I couldn't do it on videotape, and I couldn't really do it without violating some classified information. But anyway, you basically just take a steel pipe and you put this copper piece in here, and underneath you basically put your explosive in here.

And when you set it off, the copper inverts, if you will, it flexes. You remember the little clickers in high school or something, it kind of just caused the metal to flex and you make a clicking sound. Basically it inverts, and when it inverts, the copper starts to heat here from just the force of the explosion, is concentrated, and that copper starts to melt, and you actually get a jet of liquid copper shooting out of this thing. A rocket propelled grenade nowadays often has a shape — I'm going to get this right — but it has a shape like this, and then it's got the rocket back here, right, to transport it. But the head is basically just nothing more than copper, very precisely machined, that when the charge lets go here, it inverts the V and shoots out — maybe I got that backwards — anyway it shoots the stream of copper out in front.

And the kinetics of copper cutting — I told you yesterday about someone trying to use liquid copper jets for cutting. Well, the idea comes from 140 years ago when this guy at the Brooklyn Navy Yard found out that if you concentrate all the explosive energy and have it converge to the center of something circular, you can cause it all to melt and shoot out the jet of copper, okay. There have been debates on whether it is liquid copper, or whether it's just the kinetic energy of all that copper being mechanically deformed. And that's what they do at Aberdeen Proving Ground, is try to design these things and make better ones and more efficient ones. But I actually saw some work — and I'm not telling you anything — where they believe they proved that it actually is a liquid jet. I mean, I've talked to the guys who did the research. Yeah, you can call it spalling, but in any case, liquid copper will cut through steel in microseconds per inch, okay.

So what have they done to improve armor for this? They start using ceramic armor, because liquid copper doesn't cut through ceramic. It melts through steel, but it doesn't melt through ceramics at higher temperatures. And what do they do to defeat ceramic armor? Well, they just use kinetic energy penetrators, okay. That's where you get your depleted uranium. You just want a big heavy rod of material that you shoot at the material — you shoot at the thing at near sonic velocity, and it hits the wall of something, it'll break up the ceramic.

So what do they do to protect the ceramic? Well, they layer with other stuff. In this case, they use what they call active armor. An active armor is just nothing more than a layer of explosive on the surface. So you've got your steel, you've got your ceramic, and you've got a layer of explosive on the surface. And when the depleted uranium — or nowadays they're trying to use tungsten, because the Arabs got all upset in Kuwait of littering the whole desert full of depleted uranium, okay, it's slightly radioactive. People don't like radioactivity anymore. Anyway, so when it hits, it actually causes the active armor to explode, and that sends a shock wave down this thing. It's so highly stressed that the extra stress from this causes the whole thing to just shatter. The penetrator just shatters.

So what do they do for that? Well, now they end up developing a sabot that has — is two sabots in one, just like a train, a two-car train. And the first one hits the active armor, causes it to explode, and it's sort of sacrificial. And then the second one behind it goes right into something that no longer has any protective armor layer, and it penetrates that, okay. The technology to armor something and then defeat the armor is tremendous. I mean, I've rarely seen that kind of sophisticated technology, but people don't really know much about it because a lot of it's classified. None of what I told you was classified, okay. You can go to Scientific American type articles and you can read what I just told you, okay.

But the whole idea of shape charges — they use shape charges to bring down buildings, okay. A simple shape charge — if you go to this National Geographic special on the World Trade Center — they just take a piece of copper — a rod — and they machine it like this, okay. So it's just got a square quadrant cut out of it. And they put a stick of dynamite in here, a long stick of dynamite. And when this dynamite goes off, it inverts this V, and essentially you put this right up against the steel beam, and the inverted copper V just cuts right through the steel beam, okay. One stick of dynamite, one copper rod, properly machined, you can cut through a piece of steel, okay. Now it helps to know what you're doing in a little more detail than that, but that's all it takes, and they've been doing it for 140 years.

And that's what they did to shoot through tank armor. I mean, I remember one of my students worked for the Army in the mid 80s, and they had just come up with better sabots that could penetrate multiple layers of armor. And he told me they did a study on the field, and some major general or brigadier general was there in the Armored Division. And they took three old tanks and they lined them up, and they just shot right through all six layers of armor — you know, two layers on each side, right, of each tank. And apparently this major general just tossed his cookies right there on the battlefield, okay, because obviously anyone in one of those tanks was a goner. But it doesn't really matter, when was the last tank battle we had?

Student: [Gulf War, 19]91.

'91. Oh yeah, you're right. We had them in Iraq, sort of. But the real tank battle where we were fighting really fighting other tanks — you're right, we had one in [19]91. Was [it] like '72, the Suez War? And the average lifetime of a tank when the tanks were sort of equivalent capability was about two and a half minutes on the battlefield. I mean, helicopter gunships with rocket propelled grenades and stuff just shooting, you know, with, you know, whatever in a barrel, okay, they just wipe these things out in no time.

There's actually a book called Armored Cav[alry] by Tom Clancy. It's the only non-fiction book that Tom Clancy's written that I know of. And it's actually one chapter in there talks about — you write in the '91 Gulf War — have you ever read this one? Okay. So they start the tank assault, and the M1 Abrams tank — one of them gets stuck in some mud near a river, and the other tanks try to pull them out and they can't. So they say, okay, well you stay here, we'll send in a crane to get you out of the mud, and we got to continue our drive towards Baghdad or whatever, which they never got to, right. They were called back.

Anyway, while this guy's stuck in the mud, apparently three Soviet — what were they, T-1s or whatever they called, 72s — T-72 tanks they come over and they see this American tank stuck in the mud, and the first one fires at the Abrams and his round just bounces off the American armor. And the Abrams turns his turret, aims at him, wipes him out. The American ordnance was more effective against his armor than his ordnance against the Americans. So another one fires, and it just bounces off the American tank, and he turns the turret, okay, wipes them out. The third one decides discretion is the better part of valor, so he goes over the sand berm to the other side of the sand berm, and the Abrams scanning the sand berm — and he can see the heat from his exhaust on his infrared scanner — and he shoots through the sand berm with his penetrator and wipes him out, okay. Couldn't even see the tank. He went through the sand.

So these things are pretty amazing capability. And so you're right, the last tank battle was '91, okay. But it wasn't as exciting as the Suez War, where you had, you know, hundreds of tanks coming at each other like cavalry charges.

Student: [follow-up]

Yeah, and I was actually thinking in terms — but that was the last real tank battle, and the average lifetime of a tank was two and a half minutes, okay, on the field. So tanks are sort of like battleships, they're sort of an anachronism in today's warfare. But that's another story. In fact, today's warfare is sort of an anachronism. I remember as a kid, during Vietnam, you'd get a body count every week, from — or every day from Vietnam, right. And the concern was, how many Americans got killed in Vietnam? You didn't care about how many Viet Cong got killed or others. Over the years, it's become — actually not as many Americans get killed. I mean, the count — there's still a count, but it's nowhere near as bad as the 50,000 that died in Vietnam or whatever it was. But now you get the count of the enemy, and a lot of the people, civilians, believe that you're not supposed to hurt the enemy in the war either, okay. You're supposed to have be able to incapacitate them without doing anything harmful to them, right.

So war is not what it used to be. Of course, war — you know, Clausewitz, his definition of war, right — you guys must have taken some strategy — it's an extension of policy by other means, right. Anyway, so let's go back to arc welding.

Student: [question about flame cutting versus flame straightening]

Flame cutting to what, too? Oh, well in flame straightening, you're not going to have an oxygen jet. In flame cutting, you're actually burning with this oxygen jet. The flame that you start flame cutting with, you just heat up the surface just like you're doing flame straightening, and that's about a thousand watts per square centimeter. And a flame is good for straightening a pipe or plate or whatever else at about a thousand watts per square centimeter. You don't really want to melt the metal, you're moving it back and forth. In fact, that's one of the things — one of these things had a picture —

[Tom shows the handbook.] Flame straight[ening] uses oxygen to burn the fuel, but not to burn the iron, okay, in oxidizing and reducing sense. You should be using a neutral flame, so that you don't — okay, if you use a carburizing flame you could be screwing things up, okay. And in fact, these books in the beginning will tell you about the different types of flames, and you want to use a neutral flame. And you basically can use line heating. Actually there should be something in here, okay. This one I wanted to show at one point. There's three different heating patterns that you can use. And since this is the maritime book — the whole method where you weave back and forth to make your triangle, the line method, and then you can actually do spots across the surface. Just to the 12 o'clock position, you're not heating down at three o'clock or four o'clock.

But in none of that, all you're using oxygen for in that case is to burn the fuel. And — I didn't bring those overheads with me — but in flame cutting, you actually have the fuel coming down, but then you have another thing in the middle, that when you've gotten the thing hot, you put pure oxygen in there. And the pure oxygen burns right through the steel. So in flame cutting, you lit — we call it burning of the steel, it literally is burning the steel. You form iron oxide. The iron oxide is a liquid slag, it condenses. There's no boundary layer, and so there's no limit to heat transfer across the boundary layer anymore, like there is in surface heating with flames.

When you're doing flame straightening, you got a boundary layer there, you can only put in a thousand watts per square centimeter. You get rid of that boundary layer, and you can now have a hundred thousand watts per square centimeter, which will cut just like the shape charges, liquid oxygen and liquid copper can cut right through steel. A jet of oxygen essentially burns as effectively, and gives you the same type of 100 times the heat transfer coefficient to the surface, and just burns its way right through.

There's actually a video — I got to find — it was on the website. When I looked on the website the other day, it says it doesn't find it on the server. But someone did laser flame cutting. You can use other heat sources rather than just flames. You can use lasers if you want. In fact, laser cutting of carbon steel is the typical way of stamping out steel sheet nowadays. They have automatic machines the size of this room — big — and run in big sheets of steel instead of having mechanical stamping to shear out little eighth inch widgets or something. You basically have lasers that come along with an oxygen assist, and the laser provides the heat to get the surface up to ignition temperature, and then an oxygen jet follows along. You can cut at 200 inches a minute. And so you can take a sheet of steel, and within five minutes you can take a piece of steel that's eight feet wide and as long as this room, and you can turn it into all kinds of parts and scrap, okay.

I actually had a situation out here in Western Massachusetts. One of these machines might cost two or three million dollars, and you can run them lights out in the factory. You just program them, and the sheet is fed in, and you turn off the lights and you tell it what to do all night. You come back the next morning, you've got thousands of parts sitting there, okay. One of these machines caught fire in the middle of the night in the lights out factory, and destroyed the million dollar machine. And people looked at the fire — the remains — for three days, and they couldn't find the ignition source. So the people who wanted to sue the company that made it, to get their money back on their messed up machine, said okay, we think it was the metal powder that ignited. Very finely divided metal powder has a lot of surface area, and if it gets hot enough it can ignite.

I mean, it's a problem in metal powders. This powder can go off just like in a sparkler — it's metal powders that are burning, okay, in a controlled way. But there have been many fires in barges that are carrying machine shop turnings or grindings and parts and stuff. And if they ignite, it's like a flare going off, and it might be a 10 ton flare, okay. It'll burn for a while, and it burns really hot, because it's a solid fuel. So anyway, I was asked to — after someone else had done the fire investigation — I was told oh, they want to go to mediation, they want you to help mediate it. I said, oh okay, well I don't even know anything about it. So I called up the company in Finland and I said, well, is this oxygen-assisted cutting? They said yeah. I said, oh, well there's the answer.

So we went to the mediation. The other side comes in and they say, well, all the metal powders that are formed from the cutting operation got down on the bottom of the machine and they ignited. And then they asked me to make a presentation. I said, well, that's all very interesting, but they used an oxygen cut, so all those metal powders are already iron oxide. You can't burn wood twice, okay. It's already ash. It's already oxidized. And then I went — and it turns out, eventually we got them to withdraw their lawsuit, because they didn't have a clue what started the fire. It wasn't metal powders burning because they had already been burned, okay.

But kind of a long-winded answer to this whole thing. But the difference between flame heating and flame cutting is, you have a big oxygen — a source of oxygen. In a steel mill, you don't even need the flame to bring the hot steel up to temperature when it comes out of the continuous caster. It's glowing sort of yellow-white hot. All you need is an oxygen jet. Don't need to get it to ignition temperature, it's already there. You put a pure oxygen jet on there and you can cut right through it. And they do — the steel's coming out, coming down the line continuously, and they just have an oxygen jet that goes with the speed of the steel running down the line, and just — you know, the jet's moving across, but it's moving along the length at the same speed that the thing is, and so in a sense the path is diagonal, but the cut in the steel is straight across.

And one of these things I'll show you — not this one, but another one of the lectures, if you get to — it talks about continuous casting of I-beams, and you kind of see the flame cutting. They just cut the I-beams with just a straight oxygen jet, and guys will go in there — in the old days, when I — you know, old days, 40 years ago when I worked for a steel company, they'd have guys go in there with their hot suits on and just an oxygen jet, and they kind of stand on some platform next to a hot piece of steel and just spraying jets of pure oxygen on it and cutting right through it. Lots of sparks, okay. People — they usually OSHA makes them keep people out of those environments nowadays. But anyway, other questions?

See, I told you, I'd rather tell stories in the lecture. So the thing in arc is an electrically augmented flame. Even with a jet burner I can get high enough velocities coming out of here to blow that boundary layer — oops, did I — what happened? [Projector/overhead malfunction.] Oh, does this guy fall asleep? Okay. Well, I'll have to fix that another time, okay. So we have plasma — we have — I have to go to the old-fashioned way, chalk.

[Tom moves to chalkboard.] A jet of gas coming down against the surface, and you have a boundary layer. The higher this velocity across here, the thinner that boundary layer. And you can go to jet burners where you have near sonic velocity of the gas coming out, and you still got a boundary layer that's a fraction of a millimeter or a millimeter thick. And that limits your heat transfer. The maximum you can get is about 2,000 watts per square centimeter, okay.

Now, it turns out that you can defeat this if you actually add electrons. And we talked about this the other day. If you actually have an ionized gas, even though the flame might be at three thousand degrees Centigrade for an oxy-fuel flame like oxyacetylene, and the arc may be at ten thousand — it turns out the gas is not going any faster. It's going about 500 miles an hour, it's not quite sonic velocity, but it's pretty fast. And you don't really get that much thinner a boundary layer. And you're going to get with either one of these with no electrons about 2,000 watts — well I already wrote it down over there, 2,000 watts per square centimeter with no electrons.

My abbreviation for electron is an E with a minus sign, okay. Now, if I actually have electrons — if this is a plasma arc — there be electrons, and if I have enough voltage across this cold boundary layer — and the cold boundary layer may be something on the order between 1500 Centigrade, the melting temperature of let's say steel, and this free stream velocity — of either three thousand or ten thousand — but out here let's just call it three thousand Centigrade. This is a non-ionized gas, it's too low a temperature to give off ions. And so the electrons have to punch their way through just with pure voltage and kinetic energy, okay.

So I'd put up this overhead before. It showed a welding arc has looked something like this, and it has gas coming off the surface because there's some electromagnetic forces — we'll talk about later today or tomorrow — that cause a 500 mile an hour wind coming off this thing. And if you plot the voltage versus distance across here, you'll get something that looks like this. And this is if this is the anode or the cathode. And that's basically the voltage necessary to strip the electrons out of the metal and get them into the plasma phase. This is the plasma column. And this is the anode voltage drop, and this is — and we mentioned before, the plasma column is about 10 volts per centimeter. And this is about 5 volts, let's say, and this one's about 5 volts also — 3 to 5 volts, something like that, depending on the material that's in there, okay.

It turns out, because of the higher mobility of the electrons compared to the ions, in the plasma, 99% of the current's carried by the electrons. The big heavy ions are 100 times slower. I mean, you got a voltage gradient, you got an electric field here, and the positive ions are trying to go up and the electrons are trying to go down. But in terms of winning the race, you get 99 electrons down here for every one ion that ends up going into the electrode up top, okay. So most of the flow can be considered electron flow, okay.

Now it turns out — you guys probably didn't bring your notes, did you? And my overhead's not working now. Well, I have to tell you — I'll skip that, and maybe tomorrow I'll tell you how a fluorescent light works. I usually call this my fluorescent light lecture. But I kind of need the graph to do that.

Now there's a paper in your notes by Metcalf and Quigley, if you want to read it. But this what I'm going to do right now comes right out of this paper by Metcalf, Quigley, Swift-Hook and Gick, okay. And it's the heat flow to the surface due to this electric current flow. And the total heat flow is — the heat, Q is often used for heat — is Q of the electricity, or the electrons, I'm sorry, the convection and radiation.

Now this is actually pretty basic heat flow. Mechanical engineering heat flow course will tell you have heat transfer by conduction, convection and radiation. Conduction is — if I just had a piece of copper here and I wanted to see how fast the heat went through the copper, that's conduction through this — copper solid copper. If the air is completely still, there's conduction through the air, but the conduction through air is a lot slower. The conduction through a gas is a lot slower than conduction through a liquid or solid. Why, anybody know?

Student: [response about atom spacing]

Yeah, well, yes, proximity of the atoms, okay, if you will. Does anybody know what the relative density is of water versus steam at 100 degrees Centigrade, and what the ratio of densities is? About 800 or a thousand to one if you calculate it. That means that water molecules in liquid water at 100 degrees Centigrade are basically in contact with each other, okay. They're touching each other, you can't get them any closer together except under extreme pressures, and even then it's not very much closer together. The vapor — the gas — the steam — is about — let's just use a thousand, because I can take the cube root of a thousand in my head, okay.

So if the gas phase, the vapor phase, is a thousand times less dense, that means the distance, the separation between gas molecules in steam at 100 degrees Centigrade, is 10 atomic distances, 10 molecular distances, okay. Now that's a worthwhile thing to know, because if I start talking about how high of pressure can I really get by compressing a gas — well, you can get about a thousand times compression at room temperature, because then all of a sudden you can't tell it from a liquid, okay. You actually get to a point thermodynamically where they call it the critical point, and you can't tell the difference between the gas and the liquid, okay. From a thermodynamic point of view, it's completely compressed. So really about 15,000 PSI is the maximum type of pressures you can generate in steam systems. There are some that go to twenty thousand, so you know, let's not worry about a full significant figure.

But in any case, you can only compress gas so much. But in terms of heat transfer, the heat is transferred by the molecules. Each molecule carries a certain amount of heat. So liquid heat transfer is about a thousand times faster than gaseous heat transfer, okay. So in terms of conduction — we're conducting through a gas, gases are great insulators — this is going to be a low value. Convection — well that's because the gas is stirred up and moving. It's not — this would be gas stationary, this would be gas with movement, and that gets to be a very messy problem. But even convection is on the same order as conduction in these types of systems.

Radiation — how much radiant heat is there between me and the wall? Not much. So it turns out radiation is even less. We're going to take a break in a second. But it turns out of the total heat, conduction is about 4%, convection is about 2% — this is all out of that paper by Metcalf and Quigley — radiation is 1%, and it turns out the electron flow is going to be about 50% of your total heat that's in the arc. And that's because not 100% of the heat gets into the workpiece surface — some of it goes off in hot gases that just heat up the welder, the person doing the welding. Just like with flames, it was 90 to 98% of the heat was just wasted. In an electric arc, something on the order of 40% of the heat is just wasted. Only about half the heat — 60% of the heat — actually gets to the workpiece. But the bulk of it, 80% of the heat that does get to the workpiece, is due to the electron flow. And we're going to talk about that when we come back from break, okay. I'll go get my shape charge. I don't know if I can set up a video thing, I'll have to get it figured out later.

[After break.] Okay, I'm gonna have to think of some stories to tell rather than lecture, but anyway. So we just talked about the total heat in the arc — only about 60% actually gets to the workpiece, but 80% of that 60% comes from the electron flow. And the electron flow Q_electron — and this is all in the Metcalf and Quigley stuff — is equal to the current times the work function voltage, which I'll describe, plus the anode voltage drop, plus the Thompson voltage.

And the Thompson voltage is the Thompson of Lord Kelvin — okay, [no — Thompson was Lord Kelvin's name; Tom slips, then corrects: Benjamin Thompson [actually William Thomson, Lord Kelvin]] — but in any case. So the electrons are — the electron flow heat is clearly proportional to the number of electrons, that's the amperage, I, current, right. So the more current, the more heat. Now this stuff — and this is just current times voltages. This is the work function voltage, let me write that down. This is the anode voltage drop, which we talked about right up here, okay. That's this 5 volts, and that's going to be about 5 volts. And this is the Thompson voltage.

Now the Thompson voltage — if we start thinking the electrons are up here in this plasma that's hot, like 10,000 degrees, 12,000 degrees — and if I look at my notes I can do the calculation for that. Might as well use the board, since I can't use this board right now. I want to know what the Thompson voltage is. And I know that Boltzmann's constant times temperature is equal to some energy in electron volts. And in fact — this is basically saying the thermal energy is equal to some voltage. Or if you want voltage, is equal to kT over the electric charge e, okay. Now k is Boltzmann's constant, T is temperature, and e is the electric charge. Temperature is on the order of 12,000 degrees.

I'm using 12,000 for a reason, you'll see here in a second. It turns out kT — most Boltzmann's constant times temperature — any people who [are] double E's in here? No double E's? You're a double E. Do you know what the kT is in terms of the equivalent noise voltage at room temperature? Okay. The electron — there's thermal noise in electrical circuits, and it turns out when you do those calculations for thermal noise, room temperature at 300 Kelvin roughly is equal to 25 millivolts, okay. So without going through and writing out Boltzmann's constant and the e — so 0.025 volts equals 300 Kelvin, okay. If you just put the electric charge in coulombs, and Boltzmann's constant together with temperature in Kelvin and voltage in volts, and you'll show that 25 millivolts equals 300 Kelvin.

So if I have 12,000 Kelvin and I divide that by 300, I'm going to end up with — 40 — oh yeah, a factor of 40. So if I just take a ratio and proportion, I'm going to find that the Thompson voltage is 40 times 25 millivolts, which is one volt. And that's kind of the electrons up in this plasma above the ground state energy. The ground above the ground state energy of [an] electron is defined as just an electron sitting there in space doing nothing. But at 12,000 Kelvin, it's not doing nothing, it's whipping around with a lot of thermal energy. How much thermal energy? At 12,000 Kelvin, it has the equivalent of one volt of energy — one electron volt, if you will. But the voltage that's going to go into this equation, the Thompson voltage, is one volt, okay.

So if I think of some sort of energy state for the electrons, the zero of energy is an electron at rest. I actually have some electron here at 12,000 degrees — this is an energy scale, 12,000 degrees is one volt of kinetic energy. And that electron is going to go from here up in the plasma column down to the anode surface. But it's got to punch through — it's got to have enough electric field to punch it through this neutral region right here. And that's going to be on the order of 5 — 3 to 5 volts — and we'll just say the anode voltage drop — I'll just make it 5 volts. So I'm just going to make it 4 volts because it's somewhere in between 3 and 5. So on the order of 4 volts to punch through, from the hot plasma in the plasma column through the boundary layer.

And then it has to drop down to being an electron in the metal. All we're doing is taking an electron up here in a hot gas, punching it through the boundary layer, and then letting it relax and become one of the electrons in the metal. It turns out there's something called the Fermi level, which is the highest energy state of electrons in the metal, okay. The electrons are bound to the metal, they're at a lower energy state in the metal than if they were in the gas. And the highest energy state is known as the Fermi energy. And the difference between the Fermi energy and the zero electron state is called the work function. And the work function is on the order — let's call it 5 volts. It can be 3 volts, [for] sigma anyway. But I did one, four and five. The voltage in the arc has a total of 10 volts to get the electron from the hot plasma in the middle of the arc down condensed into the metal.

Will cause that electron to lose 10 volts of energy potential. And times the current gives me the total amount of heat transferring the electron from the plasma into the metal, okay. Now it turns out, you can sort of reverse this and say — okay, if I've got the anode — I got my electric arc here, and this is my cathode and this is my anode — it turns out that my cathode is losing electrons, they're coming out of the cathode and they're going into the anode. And I know — I just calculated the heat into the anode is about 10 volts times the current. The heat — the electrons coming out of this electrode up here are actually evaporating out of the electrode. And if you evaporate off particles that have some energy, you're losing the heat of vaporization of those electrons. They were bound into this metal by the work function, and to pull them out, I actually had to pull them out with the cathode voltage drop. And basically, as they evaporate out of here, they carry away the heat of vaporization. So the electrons coming out of the cathode — we have electron cooling of the cathode.

This is — it turns out it's going to be important when we start talking about when to use reverse polarity, you want to use straight polarity. This is what we call straight polarity in welding — the electrode is negative. Most of our welding when we have a consumable electrode is done with electrode positive, which we call reverse polarity, okay. If this is a tungsten electrode — gas tungsten arc welding is usually done with straight polarity. And in fact, we like to use tungsten — and I have a big tungsten — here's a big tungsten, and I have some other tungsten here. [Tom holds up tungsten electrodes.] We'll pass around if you want, it's not all that interesting. There's just tungsten rods that would be used for gas tungsten arc welding.

In the old days, we sometimes used pure tungsten, but usually we used to use thoriated tungsten with 2% thorium oxide powder, okay. That was very common in the United States. Historically, I don't know exactly why they started using thoriated tungsten, but it turns out thoriated — thorium oxide is a very easy electron emitter. It will give up its electrons very easily, and it turns out it will cool down the surface of the tip of the electrode, the cathode, by lowering the phi voltage there — the work function voltage. And this actually has been noted in various experiments. People put pyrometers on the electrode in gas tungsten arc welding, and if they measure the hot spot — this is temperature versus distance — they find it's very hot, and then it actually cools off at the surface as you get down here where the electrons are coming out. The electrons cool the tip of that electrode, and they heat the surface of the anode as a result.

About two-thirds of the heat, 60% or whatever, goes to the anode, and some of it's lost to the air. But only let's say 20% of the heat gets into the cathode from the conduction and convection and radiation, okay. And the rest of it is lost to radiation and hot gases coming off.

If you stand next to a gas tungsten arc, you can get a sunburn. In fact, I went to a conference back in the 1980s in New Orleans, and I had stopped in Pittsburgh the day before at Alcoa, and they were welding these great big liquid natural gas container spheres for ships, LNG ships. And this was — they had to weld like four to eight inch thick aluminum for these huge spheres that were going to contain the liquid natural gas. And they developed a welding process they could use up to 600 amps and really put the metal in there and make these big thick welds in aluminum. And they asked me to look at it. So they gave me a handheld torch, they didn't give me any gloves, and I held the torch and I watched this plasma arc, okay, I watched it for 15-20 seconds. And the next day at the conference, I had this big red burn on the lower half of my outside of my hand, and I thought, how'd I get a burn? It was all red like a sunburn. It was a sunburn. Aluminum, which is very reflective, and a 600 amp radiation in 15 seconds, I had gotten a sunburn from holding it. The reason it wasn't up here is the torch actually shielded the upper part of my hand. Finally my shirt — you know, they should have given me a glove, okay, to keep from getting sunburned. In 15 seconds I developed sunburn, which didn't show up until the next day.

But in any case, there is a lot of radiation and heat loss and hot gases coming off. But two-thirds of the heat is here, 20% is up here, and some fraction is off over here is radiation and stuff, and hot gases. That's the portion of heating between cathode, anode and the air. Yep?

Student: [question about the work function being a drop below from the anode voltage]

Right. The anode voltage essentially — if you thought of the electron coming from whipping around at very high velocity here, to being stationary right on the surface — well, stationary is the zero energy state. If I then go into the metal, it's actually whipping around and has some velocity in the metal. But the appropriate way to think of the energy in the metal is that the lowest energy — or the highest energy state the metal will allow — which is called the Fermi energy in solid state physics. So it drops from being an at-rest electron to being part of the metal, bound to the metal atoms, okay. And it gives up that energy.

Student: A lot lower than I would have guessed for the arc going from, you know, the tip of the lecture without the metal. So it's only about 10 volts to have an arc, 30 volts —

Oh, is that what you — well go ahead, sorry.

Student: Machines a lot higher than that.

Well, that's because the — no, it's a — it turns — there's two things going on. If you run a gas tungsten arc, the power supply will be putting out 10 or 12 volts, okay. You might lose a couple of volts in the electrical leads going up to the arc, but if you have big heavy copper cables, you're not going to lose more than a couple of volts. You don't want to just be heating up copper cables, right, that doesn't do much good. So I use big heavy copper cables. And when you run a gas tungsten arc, your actual operating voltage — probably a TIG arc is about 12 volts. When I run reverse polarity — we'll talk about it in a second — you actually are putting all your heat in here and you might be running 25 to 32 volts or something like that.

The power supply has a load line operating. If you look at the voltage-current characteristics of the power supply, it might have 80 open circuit volts, okay. That's with no current flowing. The power supply puts out 80 volts, and it's a drooping characteristic for the power supply, and the operating point will be where the plasma load line crosses. And that might be — for straight polarity TIG arc, it could be 12 volts for straight polarity, for a reverse polarity it might be 30 volts, okay. It's a lot less than the open circuit voltage, because when you draw down on the power supply, it drops the voltage because of its own internal resistance and everything else. And so that's the — what we call the power supply load line. If you go look at the manual that comes with the power supply, they will show you the power supply load line.

And for different types of welding, we use some things that have what we call a constant current power supply, or constant voltage power supply. Constant voltage will have a drooping power supply, constant current will be more steady. And it may have actually — the constant current — this would be constant current, but we might actually run a constant current that has less slope to it. And we can talk about why. When you run constant current, when you run constant voltage in your power supply — when you do TIG welding you're going to run constant voltage power supply, when you run a lot of your MIG welding you're going to be using constant current. And nowadays with fancier power supplies, inverter power supplies, it's just a flick of a switch to change the electrical characteristics of the power supply. In the old days you had to change 500 pounds of copper transformers and inductors and things, and so you had to have — it was a different power supply for one or the other. Nowadays it's the flick of the switch in a solid state rectifier — or actually it's an SCR, silicon controlled rectifier, okay. Any other questions?

But what all I wanted to do is say, the heating is not uniform. And in fact, the electrons will cool the cathode and heat the anode. And [in] gas tungsten arc welding, I'm not trying to consume my tungsten electrode, I'm trying to melt the workpiece. So of course I use direct current — or direct, a straight polarity, or electrode negative — because I'm not trying to melt the tungsten, I'm trying to keep it cool, keep it from melting. The reason I use tungsten is because it has the highest melting point and it won't melt off on me, and I want to put all my heat into the workpiece. When I'm using a wire electrode that I'm trying to consume, I switch the polarity because I want to put most of my heat into the anode, because I'm trying to melt this off and fill up a groove.

And [in] gas tungsten arc, I'm trying to melt the edges of the sides of the plate, so I put the anode as the plate. In gas tungsten or gas metal arc, or MIG welding, I'm trying to fill up a groove, a little V-shaped groove. I need enough heat down at the cathode to melt the edges, but I don't need to melt very deep. I'm trying to take drops up here and spray them into that little groove to create the weld, and so I want most of my heat at the anode, and I flip the polarity when I use gas metal arc welding, okay. It sort of makes sense in physics and even in electrical engineering — not that the two are that different, okay.

So as far as gas tungsten arc welding goes, it turns out that you can do some things to constrict the current path. Obviously, if you have a current path that's very broad as opposed to one that's very narrow, here you're going to get a higher heat intensity on the surface if you have a narrower current path than if you have a large current path, a big broad one.

And it turns out — I mentioned two days ago, just in passing, I wouldn't have expected you necessarily to pick it up, but our electric arcs are very strange beasts in the sense that if you cool the outside, it gets hotter on the inside. And what you're doing is, you're taking that plasma — if you cool the outside, you're condensing some of those electrons turning them back into neutrals — and so in order to get the current that you're imposing on the power supply, that current has to concentrate in the core. I get a higher density in the core of the arc and the temperature goes up, okay. So an electric arc plasma is a very strange beast in [that] if you cool the outside, you actually get higher temperatures in the center and a higher heat intensity on the surface, okay. It's sort of counterintuitive, but it makes sense if you realize that all you're doing is condensing the electrons here, they're going back and combining with ions where you're cooling it. That forces you to ionize more of the center, which lowers the resistance of the center, which allows more current density to go through there, more heat intensity, because the heat intensity is going to be equal to the current density on that surface, okay. Directly proportional to the current.

One of the things that comes out of this is, you don't control the voltage in welding. I mean, you might impose some voltage on the power supply, but the heating comes from the work function, and you don't have much control over that — that's a material property. The anode voltage drop, which is a plasma property, you don't have much control over that. And the Thompson energy, which is the temperature of the plasma, you don't have much control of that. What you have control over is you can dial up the current, okay. If you want more heat, you get more current. And the heating value of an electric arc is directly proportional to the current. And you'll have all kinds of explanations out of welding handbooks and other places about, this is such and such happen because you increase the voltage — you don't have control over the voltage. Nature picks the voltage, in terms of the temperature of the plasma, Thompson voltage, the anode voltage drop, and the work function voltage.

Now you do have some control. You have these different types of electrodes. I mentioned thoriated electrodes. It turns out the cathode voltage drop from a pure tungsten electrode may be 5 volts. From a thoriated tungsten electrode it might be two and a half or 3 volts. And so you cool down — or, well, you actually could do less cooling of the surface, but in fact, it's easier to strip the electrons out of that cathode with the thorium in there. You can actually run twice the current density with a thoriated 2% thorium oxide powder mixed in with the tungsten. You can run twice the current through that electrode for a given size electrode.

But people don't like thorium, because it's slightly radioactive, just like they don't like depleted uranium all over the ground in Iraq, okay. And I said, people have learned to not like radiation as much as they used to. Back in the 1920s, really wealthy people used to drink radium, because they thought it gave them an upper, okay. They felt better. And there was one very wealthy guy — I mean, some people started finding health effects after they did this for a little while, okay, of drinking radioactive radium. It might make them feel good for a little while, it's sort of a short-term upper, but some of them started getting aches and pains and the bones started breaking and things like that. This one guy, he was just convinced, and he drank it until the day he died, and he hardly had any bones left in his body when he did die, okay. And he died from radiation poisoning. But it was like — I don't know what the cost was, but it was hundreds of dollars per drink, okay, but he would drink some radium every day, okay. So people used to like radiation. When they first discovered it, they thought it was a good thing, okay. But nowadays they think it's a bad thing.

Well, what people have done is, they've gotten rid of the thorium. The Japanese about 30 years ago did a lot of work on this, and they have cerium oxide, lanthanum oxide, and yttrium oxide — very high melting oxides. And the oxides will emit the electrons — they're easy electron emitters, okay. And so instead of 2% thorium, nowadays — it's hard to find thoriated electrodes — but they actually have a different color on the tip. This big tungsten electrode I'll probably pass around tomorrow, but it has a red tip, if I remember — the red tip — this is thoriated tungsten. But this one is probably good for 500 to 1000 amps, okay. You don't see these this size very often, okay. They did use these for certain types of welding that's sort of interesting to the Navy.

In the sense that — when the Alpha sub was first kind of came on the scene, 1980, right — what's the Alpha sub, everybody know what the Alpha sub is? You don't — you haven't seen The Hunt for Red October? Okay. Tom Clancy. The Alpha sub was the attack submarine, all-titanium attack submarines, the Soviets dropped on the world in 1980, and it was a real eye-opener for the U.S. Navy. [The Navy] had been looking at titanium for submarine hulls for 30 years. You know, this was 1980 — 1950, in the early 1950s. Well, I give you the full story — probably during World War II, but — and I think because the Manhattan Project, MIT mechanical engineering developed some very effective vacuum systems for what they call vacuum metallurgy, okay. And you now could melt reactive metals under a full vacuum, whereas they really didn't have the technology to do this on an industrial scale before that.

So all kinds of new materials on the periodic table — new metals became available. Titanium, which if you tried to melt it — you can't melt it under a flux really, because it'll react with the flux. You can't melt it in [air], because it will burn like a flare, okay. But titanium, tantalum, niobium, molybdenum, tungsten — well, tungsten had been used before because of light bulbs, but they did it all in the solid state, they didn't melt any of these things. But all of a sudden these things became readily available, and they actually could do vacuum processing. My thesis advisor in the 1950s was growing tungsten single crystals and looking at the mechanical deformation of tungsten. He was growing these crystals in a vacuum system at ultra high vacuum, and when he would send his samples out for chemical analysis, the chemistry lab would keep the sample because it was purer than any tungsten that they had, because it had been melted under vacuum. And they started using it as their standards, okay.

But so right after World War II, titanium became available. And over here at Watertown Arsenal, which is now Watertown Mall, okay — but the Army had an arsenal there since the 1860s. They used to try to make cannonballs at Watertown Arsenal for the Civil War. And they tried to — people knew that if you had small little drops of metal and you melted them and let them fall to the ground, they solidified in the air, they would be nearly pure spherical shape. And so during the Civil War, they actually tried to build towers over there to drop cannonballs — molten cast iron for cannonballs — and see if they would hit the ground and solidify. They didn't know a lot about heat transfer in the Civil War, and they realized that the tower they had to have was about 50 miles tall, okay, to get the heat transfer. And they also didn't know that there's the ratio of surface tension forces to gravity forces, and they would get ellipses and not spheres. The surface tension forces have to dominate, and that's only below about three millimeters in size. But they didn't know about dimensionless numbers, and chemical engineering had not yet been invented.

Does anybody know where chemical engineering was invented? MIT, okay. That's another story, but anyway. It wasn't invented until about the 1880s. So in the Civil War, they had a big government research project to try to make spherical cannonballs by dropping them from towers, and all they did was make a bunch of splats on the ground. And they never could build a tower tall enough. But nonetheless, Watertown Arsenal after World War II came up with Titanium 6-aluminum 4-vanadium, which is the workhorse titanium alloy even for today in almost the aerospace industry. And it was all because of this vacuum technology that came out of MIT's mechanical engineering. The building which — one of the buildings is the Sloan School building now, originally was the National Research Corporation, then became Norton Research. And now there's a firm over here in Newton called HC Starck. It makes tantalum metal for tantalum capacitors and other things. So there's a whole technology of reactive metals.

Where was I going with that? I'm killing a little time here too, until I get my thing fixed. Yeah, okay, the Alpha sub is made out of titanium. So the Navy had been working since the 1950s, early 1950s, on trying to develop titanium for a submarine. And I mean, you go down to David Taylor in Annapolis, and they would be welding four inch thick titanium, okay, in 1975 or whatever. And there was a problem — titanium is very reactive in the air, and it gets contaminated with too much oxygen, it becomes brittle and things. But the Naval Research Laboratory had identified certain problems with — when you go down a deep dive, you've got a lot of compressive stress, and something they called the creep-fatigue interaction. And they had never really solved this problem. Titanium under compression — continuous compression — if you put some cyclic load on it, just, it doesn't like it, and it will start forming fatigue cracks.

Nonetheless, in 1980 I was flying back from a conference in Europe, and I read on the front page of the International Herald Tribune that the Soviets have the Alpha submarine. This is their new attack submarine. It could dive deeper than the collapse depth of our deepest depth charge. It could go faster underwater than the fastest destroyer on the surface, okay. So some people were very concerned that we'd lost this huge strategic advantage. And the Navy was embarrassed before Congress because they had leap-frogged us in technologies, just like they did with Sputnik in '57.

And over the next few years, there was a lot of hullabaloo of what do we care about Hy steels, you know, if the Soviets have titanium subs. One of my friends who used to work with — Rocky Flats used to be a nuclear weapons facility out in Denver — one of my friends at Colorado School of Mines, when I told him about the problem of the depth charges, he says, not a big deal. Just hit him with the right size depth charge — right type of depth charge — and you can hit them anywhere within three mile depth. I said, well, if you start using that type of depth charge — well, but if you start shooting at their subs, you know, you're probably already that type of war anyway, so it was sort of a moot point.

Nonetheless, there was a real crisis for the Navy, and there were workshops held down at David Taylor Annapolis. And I went to one, and one day was Navy day, and had a bunch of guys in the Navy talking about — it was this workshop, was a welding heavy section titanium. And the Navy guys came in and talked about welding one-inch thick titanium to four-inch thick titanium on the first day. I was the only guy outside of the U.S. Navy on Navy day, because I started out — I didn't bring it back with me, but I showed you one of my one inch thick titanium welds the other day from the early — that steel — from the early, from the mid 70s, okay. My first research project, I was working on welding heavy section titanium, one inch thick. It's all I could afford.

And the second day, they had American industry come in talk about heavy section titanium. And for them, heavy section titanium, the heaviest thing they talked about was 3/8 of an inch thick, okay. These are Boeing and McDonnell Douglas and all these aerospace guys. For them, a heavy section piece of material was 3/8 of an inch, okay. So you just — you go to this workshop, it was just completely disjoint between what the Navy thought of heavy section and what the industry thought of heavy section. But anyway, I got to see some — we call foreign technology at that time, and now it's more than 20 — it's yeah, it's 30 some years now since I saw it, so it's probably — I don't even know if it's classified or declassified.

But if I was trying to weld some heavy section titanium, and I looked in the scientific literature up to about 1972 from the Soviet Union — in the open literature I would find there was a guy named Gurevich at the Paton [Baton] Institute in Kiev, and he was Mr. Titanium in the Soviet Union. And he would talk about electroslag welding of two inch thick titanium. I've got a piece of two inch thick titanium that was electroslag welded, and maybe I'll bring it in tomorrow. And he would also talk about gas tungsten arc — at what he called semi-submerged arc was what was in the open literature. And they would use huge tungsten electrodes like this, and they'd use a thousand amps of gas tungsten arc current. And they could weld — without giving you the design of the weld joint — they could weld two inch thick titanium at four passes, okay. Better than submerged arc welding of steel in terms of productivity. And they did it with some of the physics of this gas tungsten arc process.

Or let's say they could have done it. But by 1972 all of a sudden Gurevich quit publishing. Just he was gone. And that was because they had — that was probably the time they decided to start designing and building the Alpha sub, and he was not allowed to publish anymore. So in 1980 or so, President Carter had gone out because of the embassy thing in Iraq [Iran] — Iran — and Reagan was in. And there had been an exchange program in electrical metallurgy. It had been headed up by Professor Grant at MIT, Nick Grant, had actually lived the first five years of his life in Russia, and he had emigrated when he was five years old, but he still spoke some Russian because his parents did. And he was head of this big national program in the United States in exchange with the Soviets under the Carter Administration.

And there were the huge — the first exchange was like 40 senior metallurgists from U.S. universities and stuff going over to the Soviet Union, there's an interesting story about that. And then the second one there were like six or seven people, and the National Science Foundation was running this exchange because it was headed up by the State Department and stuff under Carter. But the Reagan Administration was trying to slam the door on it. So the third exchange consisted of Professor Szekely [cichelli] from this department — he's passed away — and me, okay. We were there. And he wanted to go because he actually grew up in Hungary before the revolt in '56. He left Hungary in the revolt. I said, you didn't have a — you know — and he went to Imperial College, got his PhD, and then became a faculty member [in] the United States. But I said, well, you weren't carrying a weapon or anything? Oh yeah. I said, how did you get to the border? He said, I just took the train, got off the train, walked across. But he had a weapon with him, and he was one of the revolutionaries in the 1956 Uprising.

But anyway, he had learned some Russian and stuff, and he had a National Science Foundation contract, and he didn't want to go over there by himself. So he said, do you want to go? And I was just this young 30 year old assistant professor, associate professor. I said, yeah, I'd love to see the Paton Institute. So I got to see the Paton Institute, and I spent two hours with Gurevich asking him about how do you weld titanium. I didn't say, how do you weld the Alpha submarine, but I was asking him scientist to scientist about how do you weld heavy section titanium, because the U.S. Navy had been focusing on gas metal arc welding. And I at that point was starting to do some gas metal arc welding in addition to the submerged arc welding. And when I showed him the gas metal arc — I said — oh, no, we don't use that. So whatever I did — what the U.S. Navy had been hanging their hat on, the Soviets didn't use, because it was an unstable process. And we later learned over the next four or five years of why it was unstable.

Gurevich was completely honest with me. He was just a scientist, he didn't really care about the politics or anything. There was a KGB agent sitting right there in the room, right in the same table with us, and they were answering every question I asked, because they wanted to keep the door open. That it started under the Carter Administration, Reagan was trying to slam it, and this was politics at a higher level. I was just interested — and the Navy was interested in my learning whatever I could about how they built the Alpha sub, because the U.S. couldn't figure out how they could afford to do it. They couldn't figure out how the Soviets had solved the creep-fatigue interaction the Naval Research Lab identified as a Achilles heel of titanium submarines.

And it turns out the answer to the Achilles heel and the creep-fatigue interaction came within two or three years — you're smiling, you know what happened. Yeah, they cracked. Soviets never studied it. Apparently they didn't know about it. They built several submarines, and within three years they were in dry dock — or not dry dock, they were — and you know they were — what do you call — they were taken out of service. Insofar as the ability to track them underwater, these cracks made them so noisy that you didn't need any sophisticated sonar, you could hear them from miles away apparently. So it turns out the Alpha sub — which shows up in the Tom Clancy Hunt for Red October, they sent the Alpha subs after Captain whatever-his-name [Ramius] that he had stolen from the Soviets — and it turns out the titanium sub became sort of a — Soviet titanium sub became sort of a non-issue.

And it turns out, in the end, the U.S. Navy was right to be as cautious as they were about just going off and building. You could build one titanium sub for 10 steel subs, so you have to ask yourself, would you rather have 10 steel subs or one titanium sub? They're little trade-offs like that. But anyway, the Navy decided not to build titanium subs. And then Congress was just livid that the Soviets had leapfrogged us. And I can tell you other stories over the next five years about how Congress took away all the money for the SSN-21 when Millard Firebaugh [Fireball] was Captain — he was a graduate of your program — but he was designing the SSN-21 in the mid-1980s, and Congress zeroed his budget for the next year. Said, we will not build another steel submarine while the Soviets have titanium submarines, and we don't even want to build titanium submarines, we want to leapfrog, we want to build composite submarines, okay. So I went to a four day workshop on that, and that was a laughing mess, okay. I mean laughing though. Anyway, so we'll see you tomorrow.