WM_S2014_28

Changez and we can make others. It will go through at least April seventh because I have two new people who have signed up for the course. I'd like you to attend three if you can; attend all of them, that'd be great. But it's nice to have someone to present to other than just me, right, particularly if I start snoring it's really doing okay.

So that's that for the time being. We were talking about — anybody have any questions before I start on titanium? We were talking about welding of titanium and I was telling you a little bit about Gurevich, who was this titanium expert. He just quit publishing in 1974 because basically the Soviets had decided about that time, most likely, that they were gonna build a titanium submarine and they were using a lot of his technology.

I didn't realize it, but there's more to this story. There was an exchange, and I actually put in — I stapled it in — the letter that I got from Medovar [Medvar?] that says "metal bar" down there in Russian. But here's the translation they sent me, since they knew that I didn't read Russian: "we're glad to send you the book." One of our colleagues, Bibi Medovar — I'll tell you maybe — will tell the story sometime, but he was one of the top three to five people at his welding institute in Kiev, which is the world's largest welding institute. And he was here on an exchange with some other Soviet scientists, and he was unusual because he was Jewish and he had risen to a very high position in the former Soviet Union. Okay, that's unusual.

But it was clear he didn't care about what other people thought, at least in the group he was the leader of. The group, and we all know who the KGB agent was — there was six or seven people, and if you asked a question of the Soviets, a group of MIT faculty, a group of Soviet scientists, and if one of us asked a question they would all look at Mariinsky. Mariinsky was the KGB agent. If you ever ask a question directly, a technical question directly to Mariinsky, he didn't know the answer. He was KGB agent, right?

So anyway, this is actually a humorous story, this has no real point to it. But Professor Flemings in this department was and is a very well known faculty member. But Professor Flemings — never, just like me, I can't remember Hannah's name — Professor Flemings could never remember anyone else's name. He used to call Professor Kent Bowen "Ken." Okay, this is one of his colleagues, right. He couldn't remember Mariinsky, so he would call him Marinchek, because it began with an M, it was sort of a Slavic name — Mariinsky, Marinchek, you know, it's all the same. Well, so he would call this KGB agent Marinchek, and the guy would kind of, you know — and he wasn't doing that. Flemings wasn't doing this to be mean or funny, he just couldn't remember the guy's name. Some of metaphors started calling him Marinchek, okay, in front of all the other Soviets, and of course a bunch of us who all knew was a joke. Flemings didn't know his joke — he was actually Flemings was the only guy who didn't know it's a joke. But here you have this senior scientist who happened to be Jewish just needling the KGB agent. Anyway, there's more to that story; maybe sometime when we don't have a class I'll tell you some more about Mariinsky and what it's like to be hosted by a KGB agent.

Some people are — were asking me about submerged arc welding. They said, "is that submerged underwater?" and things. Submerged arc welding is a process where you weld two plates in the flat position, you have a wire, and instead of a hundred or two hundred amperes you might have five hundred or a thousand amperes of current. So you put down a really big weld. This is one of the ways they make the Keystone [pipeline] steels — great big pipeline like the one, the Keystone pipeline. It will be a steel plate that's been turned into a U-shape and then forged into an O-shape, cold forged into an O-shape, and then they'll run a submerged arc weld in the center. And it might be three-quarters of an inch thick and you can make a great big weld. It's submerged underneath a bed of sand or flux. Okay, steel welding flux — the sand might cost you two or three dollars a pound, okay. And you've submerged the arc underneath the sand, so you have this arc that otherwise would be blowing sparks all over everywhere, be too bright to see, but it's submerged beneath this sand which melts and forms a glassy layer on top, which is the slag. Okay, so this is a submerged arc weld.

My first research contract on titanium at MIT was on submerged arc welding of titanium to get a good weld. And these were — they didn't have the best shape, but they had chemistry that was excellent, but I had to use optical-quality calcium fluoride crystals, which I was paying — buying scrap from this company in Western Massachusetts that was making these optical windows out of calcium fluoride, and I would buy their scrap cuttings from their single crystals, and I was paying $100 a pound for the scrap. And it turns out it would cost you about $1,000 a foot just to pay for the flux for submerged arc welding. Technical success, economic failure, okay. So that was one of the first things I helped the US Navy learn about.

There are a number of welding processes that were used for welding titanium, one of which I didn't realize until 1980, in the Alpha sub. And I was at a meeting down at David Taylor Ship Research Center in Maryland, in Annapolis Maryland, and all of a sudden I realized Gurevich had been publishing in the early 70s on electroslag welding. Electroslag welding is a process where you have a wire feed, you have two vertical plates, you put a water-cooled copper dam on either side, you put a little bit of flux in the center — you don't cover the whole length of the weld with flux, you put a handful of flux in here — you run the wire in, you melt the flux, and now it's the resistive heating of the flux that melts the wire and you just make a casting going up. And that's this piece I passed around yesterday, which I'll pass around again. If you were to — I got a rillette it — sometimes, it's been 30 years, so as it was etched you can see on the backside where it's not even etched you can actually see the weld. This is a two-inch thick plate, two-inch thick titanium plates with a gap here, you can see where the water-cooled copper mold was on here, and the weld is this thing that's about 2 inches by 2 inches. Very efficient process.

I came back from that meeting in 1980 down in Annapolis and had one of my students make what I think is probably the first electroslag titanium weld outside of the Soviet Union or France. I'd later found a paper in France where they tried it, but right on the other side of that wall we made the first electroslag weld. They classified it, so that work was done — I couldn't do any more of that, okay. And the reason they classified it is — I will just say, to the best of our knowledge, that's how the Soviets — one of the ways electroslag was one of the ways the Soviets built the Alpha subs. They didn't use submerged arc welding, they didn't use what the US Navy had spent millions of dollars trying to do, which is gas metal arc welding. The bulk of steel welding and most of the aluminum welding done in the world today is gas metal arc welding. Just feed a wire in, you have some argon shielding gas, works great. Okay, well the US Navy had been working on it, and they never really got it to work very well. I was doing research on it and I couldn't get it to work.

And in 1980 I went over to the Soviet Union and I got to meet with Gurevich in front of Mariinsky, the KGB agent. He was sitting right there in the room, and I talked to him for two hours. Gurevich was a wonderful man. He told me — he answered every question I asked and he answered them all honestly. And I said, "what about gas metal arc?" "Oh, we don't use that." And the US Navy had been working on a process that the Soviets had determined a long time ago didn't work. And I won't go into the reasons why it doesn't work, but nonetheless the Soviets used electroslag, and another process they used was gas tungsten arc welding, GTA, 'cause you just use a tungsten electrode, you use some argon shielding gas and you make a weld. I passed around the pacemaker the other day, but here's some titanium welds made by gas metal — gas tungsten arc. Usually gas tungsten arc is for little welds in eighth-inch or quarter-inch thick material.

It turns out Gurevich had done a lot of work in the late 60s and early 70s on a process which he called semi-submerged gas tungsten arc weld. It wasn't submerged under a bed of flux. If you made something that, let's say, was 2 inches thick, just like that electroslag welding, but you had it in the horizontal position, just like these two tables — if I wanted to weld these two tables together I could prepare the joint edges like this, machine them like this, and this could be 2 inches thick. And if I put just a little thin layer of flux under — remember, this flux could cost you a hundred bucks a pound, okay — put a little thin layer, just like painting on some slurry about a quarter or half millimeter thick, and then you come in with a great big tungsten electrode, a quarter-inch diameter, and use a thousand amps, okay, you can make a submerged arc weld bead that looks like that, half an inch deep. And if — then you can put another one on the top with some filler metal, and then you can flip the plate over, you can make another one over here, you can make another. In four passes you could weld a two-inch plate, okay. More efficient than welding steel.

And I can't tell you how I know they did this, but I know they did this, okay. They would have to shoot me if I told you. Okay, they're gonna shoot me, go ahead.

Student: [question about titanium filler metal]

Titanium filler metal — oh, the flux is actually a key. They were calcium fluoride, strontium fluoride, they might have some other fluoride sand — chlorides in there, okay. So they had a flux. It was just a little paste of these optical — what happens in here is, in fact, that piece that I'm handing around, its quarter-inch piece, that's actually an example of one that had a flux on one side and then there's another weld on the other side. One of them only goes about a millimeter deep, the other one goes all the way through, or nearly all the way through, about five millimeters. Same welding current. One of them had a flux and the other one didn't. What happens is the flux causes what we call Marangoni surface-tension-driven flow, and you get the hot metal from right in the center of the molten weld pool driving down and digging a deep penetration. If you didn't have the flux there, the Marangoni flow would be out from the hot center to the sides — give it a very wide, very shallow weld pool. Turns out in titanium with these fluxes you can change the depth-to-width ratio of your weld zone by a factor of five or even ten by painting this little flux on there.

So in the mid-1980s you will see coming out of Ohio State all these projects on trying to do Marangoni flow type of weld for gas tungsten arc. They wanted to make it work for steel. Doesn't work so well for steel; works great for titanium, okay. And the Soviets were doing in the 60s — I was the only person outside of the US Navy welding heavy section titanium. That's another story. In 1980 when the Alpha sub came out, and then I tried the electroslag, kind of a light went on at this meeting. I thought, that's why Gurevich was doing like this — like I had been trying to do submerged arc — and then I started thinking of his semi-submerged arc. So I came back and I did electroslag welding, semi-submerged arc with this little flux for gas tungsten arc, and then the Navy classified everything, and so I quit doing any of that work because I don't want to do classified — you can't do classified work on campus. I didn't want to even touch it. I basically sort of quit doing most of my titanium work at that point.

But Oregon Graduate Center, where they made the electroslag weld, from Oregon Graduate Center — classified contract, they allowed classified contracts. The guys at Ohio State in their welding department, they weren't working on titanium, they were working on steel or aluminum, and that way they kind of got around the classification. But the thing that really worked and what the Navy was working on was semi-submerged arc with the little thin layer of flux, or electroslag with a little thin layer of flux. And a whole chapter or two in this book is how to clean — how to make flux and purify it, okay.

But there are also other ways to weld titanium. One of the problems with titanium — actually I can show you one of my slides from the welding handbook. We talked about — titanium is allotropic. At low temperatures it's hexagonal close-packed, at high temperatures it's body-centered cubic. Here's a schematic phase diagram, and if you have an alpha-beta and alpha stabilizer like aluminum in your titanium, at higher temperatures with more aluminum you stabilize the alpha phase. In fact, titanium 6-aluminum 4-vanadium is an alpha-beta alloy that has both crystal structures. I'll show you more about that later.

But it turns out if you start looking at impurity elements, which is important in titanium — oxygen, nitrogen, and carbon — the strength goes up from about 40 ksi up to a hundred ksi with 2000 parts per million oxygen. The elongation goes down, but you can still have ten percent or fifteen percent, which is plenty of elongation in these alloys. So we actually add oxygen to strengthen titanium alloys and we lose a little bit in elongation, but it's still got enough ductility. Yep.

Student: [question about thermal expansion]

Now the thermal expansion is about the same in both the crystal structures. It's not — it's anisotropic here, it's not the same in both directions, but in fact this temperature — what is that temperature? That temperature is about 900 centigrade, okay. Everything's plastic enough at those temperatures that you don't run — they sort of have a little bit of micro strain to accommodate each other.

What you get is you get what we call Widmanstätten structure. This is slowly cooled titanium 6-aluminum 4-vanadium, and at the grain boundaries you'll get some alpha, and you'll get some beta and alpha here going different directions. If you cool faster — this is furnace cooled, air cooled — this is basically what you'll get in a weld, and it's ultrafine compared to this. And if you water-quench you'll get something even finer. It turns out this is super strong because of the fine structure compared to the furnace cooled, but you actually get what we call basket-weave morphology. And so you get grains crossing each other because you nucleate and you get a nice — to trap growth because of the anisotropic crystal. But that gives you — that actually gives you a very favorable, very fine-grain, very mixed-up crystal structure, okay, if you will, which is one of the things gives you good strength.

It turns out the welds in titanium have better fatigue strength than the base metal because they have a finer structure. I had a student in the early 80s working on helicopter rotors over here at Watertown Arsenal. The Army wanted to use castings. You cast titanium or anything else, you're gonna have defects in the casting. You can't throw out a fifty-thousand-dollar casting, you got a weld repair it. And the question was, where the weld repair is gonna be any good. Gordon proved the welds were better than the base metal, okay, because you've got a better cooling rate, gave us a finer structure. So anyway, I mean we don't have time to go through all the metallurgy of all these things, but the point is, I guess, titanium is relatively easy to get a good structure to weld.

There are not a lot of titanium alloys out there. I mean, there are dozens, but it's not like steels where you have thousands, or aluminum where you have hundreds of alloys, because the market size is just too small. And in fact, there's not even a titanium mill. You know, they're steel mills, and if a steel mill produces two million tons of steel a year and there's a million — there's a billion tons of steel in the world made each year, that means there's got to be five hundred steel mills in the world. Well, that's a lot, okay. Aluminum — there's 45 million tons, and if an aluminum mill produces half a million tons, that means there might be a hundred aluminum mills in the world. They're smaller than steel mills. Well, titanium — twenty-six thousand tons. You can't justify building one mill. And so what do they do? TIMET — Titanium Metallurgical, right? TIMET is one of the largest producers of titanium plate and sheet and stuff in the United States. They rent time one day a week on a steel mill, okay. They do — I'm not kidding. And one of the reasons it takes so long to get something out of TIMET is because they only do processing one day a month on someone else's mill. It's a billion-dollar facility, and with — I don't care — twenty-six thousand tons of structural titanium in the world, you can't afford to build a billion-dollar mill. But you can afford to go to some steel companies, it's not doing too well, okay, and rent time on their mill. But they can't, you know, there's lots of changeover in terms of — you got to get rid of all the iron oxide on the surface, so you gotta go in there and clean that mill. You can't just do titanium on a steel mill with all the titanium — steel dirt and stuff. So they have to clean things off and stuff. But, and that's why it costs more. It's more than just the rolling time, it's the cleaning time of the mill. Titanium — a little titanium in the steel doesn't hurt it, so it's not a problem for the steel, but it is a problem for the titanium to get a bunch of iron in there that you didn't want, or iron oxide. So anyway, there's lots of processing things that go into all these things, and market size. Yeah.

Student: [question about melting temperatures]

Now, the titanium melts at like 1700 C or something like that, steel at about fifteen or sixteen hundred C — roughly 100 degrees C — has a transformation temperature at 900 C. Both of them, you know, so all your processing equipment, your controls, your furnaces, they're all the same, okay. Now what happens in the furnace is — they're just air furnaces in a steel mill, and you form a little oxide skin on the steel. It's a scale. You roll it, it just falls off. Or you take a sandblaster and you're blast it off, whatever. Or you put it in a hydrochloric acid bath and pickle it off — we call it pickling, when you dissolve it off. Titanium — you get a little oxide scale, diffusion layer might be a quarter of an inch — it might be a millimeter thick. You're gonna go in there and you're gonna machine that off the titanium because it's what they call the alpha case. Oxygen is an alpha stabilizer, and if you — your surface will — this high in oxygen, I mean it could be ten percent oxygen, it's almost going back to titanium dioxide. Just machine it off. And when they make titanium castings they put it in a hydrofluoric acid bath and they just dissolve the surface, because you can't machine a complex surface so readily, okay, so they just dissolve it away. Well, that just ups the cost again. You're talking about material that might sell for $100 a pound and you got — its your yields, the scrap is greater because you've got a machine off the surfaces and things. But, you know, so there's some things that steel can't do, like fly light.

Okay, when you need titanium — you need to build an SR-71 Blackbird, titanium skin, can't do it with aluminum, can't do it with steel, too heavy to fly. With steel surface basically over ages, doesn't melt but a good melt almost, because of the melting temperature of aluminum. Material choice — titanium. Hey, you could make it out of beryllium except you can't afford it, okay. I mean, the price of beryllium is another factor of 10 to 50 above the price of titanium, okay.

The reason we never built a titanium submarine — we did build titanium submarines. The Alvin — you know the nice little Alvin submarine, okay, the one's goes down, finds the Titanic and stuff. Originally built out of steel in the 1960s, 1970s. Who owned — anybody know who owned the Alvin originally? Actually who still owns the Alvin? Actually that's not true. Originally was owned by the Office of Naval Research. The US Navy built it for deep sea stuff, and they were interested in deep sea stuff because in the early sixties an American plane not too far from Spain dropped a nuclear warhead in the ocean by accident. And so — it wasn't open, soon as a big oops, and the Soviets — all the ships, you know, converged in that area of the Atlantic, and fortunately we found it before the Soviets found it, so we got our weapon back. There have been times the Soviets lost some things in the ocean and we actually got some of those things back.

If you ever read about Howard Hughes and the Glomar Explorer, the Navy built — or the CIA built — a full-size ship that was sort of a plant. It just kind of went out to that part of the Pacific and it had all this high technology stuff, and it lifted a Soviet submarine off the floor. Yeah, it was a spy ship and it was supposed to look like a freighter. There are books written about the Glomar Explorer, okay. But it was all big — it was a multibillion-dollar CIA project to get Soviet warheads. We got our warhead back from off the coast of Spain, but it was a big problem.

And so they started working about how do you get to the bottom of the ocean, okay. And so they built the Alvin out of steel in the 60s. The US Navy turned it into a titanium hull in the 70s, and that titanium hull kept on going until about the last five or 10 years. I was on a study commissioned of what would it cost to build a new titanium hull and submarine. That was in the 90s, and we determined in the 90s you could probably still afford it, but you didn't have anybody who could build it anymore. In the 1970s we had this big military-industrial complex, they could make these things, and by the 1990s a lot of those companies were out of business. That's another story. But in any case, they finally did decide they could build it, and then they found the price to jump by a factor of five or six. Because if you remember about ten years ago the Chinese started buying up all kinds of metals and stuff — price of everything jumped. And we were supposed to try to build the new Alvin for — they had — NSF had twenty-five million dollars in their budget. Woods Hole, who was pushing this because they maintained the Alvin, they said we can build a new one for six million dollars. I said, bull, it'll cost you fifteen. Okay, well, I was closer the actual cost that they were building. The one they're building right now, I don't even know if they finished it — it's 50 million, okay, to build a new little six-foot or seven-foot diameter titanium sub.

In the meantime, the Navy tried — the US Navy tried — in the 1980s they built the Sea Cliff, which is a little bit bigger titanium — a little mini-sub, supposed to be like eight feet in diameter. And they tried their gas metal arc welding technology they'd spent millions of dollars on. They were building at Mare Island Naval Shipyard. They wanted to develop their own — the Navy wanted to develop their own technology. So they tried, and they finally gave up and went back to gas tungsten arc — not deep TIG gas tungsten arc, but just gas tungsten arc. Super slow gas tungsten arc, you can lay down two pounds an hour of weld metal, okay. And you got to really shield it.

When I was going to show you one of the problems with titanium — I told you it doesn't — it's not that it hates hydrogen, you can actually tolerate a fair amount of hydrogen. This is dew point of the argon shielding gas versus hydrogen content. Here's 60 parts per million and stuff. If your dew point gets very high, all of a sudden your hydrogen goes up, your hardness goes up, your minimum bend radius — you lose mechanical properties, okay. It doesn't crack, but you lose mechanical properties. You've got to do a lot of things to shield the titanium when you weld it.

Here in Gurevich's book, you'll see one of the things the Soviets used to do — they used to put the welders in spacesuits and put them in an argon-filled room so that the titanium was welded in argon atmosphere, okay. That's sort of an extreme, okay. If you look in the welding handbook today, they'll show you glove bags, okay. Here's a welder and the glove bag is filled with argon. In 1969 when I worked at the Naval Air Rework Facility in Norfolk, Virginia, we were repairing titanium compressor cases for aircraft. This is what we'd use. We'd put the whole diffuser case inside one of these bags, fill it with high-purity argon, and the guy would go in there and weld through the gloves. He'd be on the outside welding on the gloves, okay. This is standard technology. It was in 1969.

You may have welding — but you have all kinds of purging things. You might put a plate across here, you have these — you put this up against the weld and you have a diffuser, you bring in backing gas, they call backing gas. You bring in your high-purity argon, you diffuse it in here, it fills, it floods the area. I mean, you might end up using ten times as much shielding gas when welding titanium. And you can look at the color of the titanium when you're all done and you can say, oh, this is nice and shiny like the ones I passed around, okay, and they say that's a good weld, it's silver color titanium. Here's a straw color. There's different colors. It gets worse and worse — this is more and more air contamination.

And in fact, here's another thing I found off the web — or maybe this one's out of the welding journal. That's supposed to be a good weld and it gets progressively worse. You see, it's just oxidizing all the way over. The problem is it kind of goes through darker, darker, cleaner, but dirtier further out, okay. Right in here you've gotten so dirty that you no longer have the thick — the colors you're getting are because of the thickness of the oxide on the surface. Once your oxide gets very thick you no longer — you get these interference colors and it just goes back to its normal gray color. So people have tried for years — if they do — say, this is something I got off the web. You look at the color and you find out whether the quality of the weld is acceptable, and they go through all these colors and tell you what's unacceptable. White is when you've gone back to titanium dioxide, okay. Gray, unacceptable. What's doing gray at silver? Well, you know what it is. But when you actually look at a piece of titanium it's not so clear. So the US Navy had determined you can't use color to determine, because all it takes is someone like me that's colorblind, and they're all good, you know, or they're all bad, who knows what they are.

I had mentioned there's relatively few titanium alloys out there. They have very good strengths in general. The pure titanium may be fairly low strength — this is grades 1, 2, 3, 4, and that's just increasing the oxygen from about a thousand ppm up to four thousand ppm or so. And you can see the strength doubles as you add the oxygen, your ductility should go down — thirty percent down to twenty percent — it's not bad ductility, okay, so far as that goes. What else do we have here? We have different alloys. You get two alloys that have a hundred and thirty, 140 ksi. Titanium 6-aluminum 4-vanadium, 135 ksi, not bad, huh? Stronger than most steels. When you get down here, there may be some here at 180, 220 range. These things are sheet material, Air Force developed, and you can't get them in anything other than sheet material — can't get them in bulk, okay. So any questions on that? Yeah, skins on aircraft, yep. They want the maximum strength they can get and they're the Air Force — they spend all kinds of money, okay.

To tell you about — I think I told you yesterday — so I'm riding back on a commercial aircraft across the ocean in 1980, I read the International Herald Tribune, there on the front page, "Soviets develop Alpha submarine," okay. They probably said titanium, they didn't call it Alpha then, but this was their titanium submarine. Well, I mentioned Congress was very upset, the Soviets had leapfrogged us. There's all kinds of stories that go on for the next 10 years about titanium submarines. I told you how the Soviets ended up mothballing these things. Two years later they were noisy, they were full of fatigue cracks. And the US Navy has never built a big titanium submarine. They built little models, okay, little research submarines.

About 20 years ago I was involved in a project for Draper Lab. They came over, they said, "we have a project to build two four-foot diameter, 20-foot long titanium submarines." Is this classified? No, the fact we're building the submarines is not classified, what they're gonna do is classified, okay. So I helped them with designing a four-foot diameter titanium submarine. Turns out four-foot diameter titanium submarines — only about a quarter-inch thick wall. You're not trying to go to the bottom of the ocean. And I still don't know — I have no idea what the mission of these little subs were, of these two subs. But we basically went down to places like Pratt & Whitney. They're building titanium four-foot diameter pieces all the time. They called jet engines, okay. And so we used aerospace technology to build the components and then weld them together and stuff, and they were successfully welded them.

I guess I could tell you — the rest of the story is when they came to me, they were — this was a rapid prototyping thing, this early 90s, and rapid prototyping was a big deal in the US Navy then. And they said, "we have to do this in like three or four months." Well, fine. And then I get the contract about two months later, through all the contracting things at Draper, and they were gonna pay me $50 an hour. I thought, excuse me, I haven't charged $50 an hour for 15 years. And they said, "well, just adjust things accordingly." So I gave them a daily rate, okay. And some days I might work two hours and they billed for a day, you know — there was different ways you can do these things. And so, but because I'd already spent the time, we'd not negotiated a price, okay. They know — I think I probably told them what my consulting rate was, and all of a sudden their contracting people say, "oh, we can't give you that fee." That is bull. If you were ever working for the federal government and they tell you that they can't pay your full consulting rate, it means they don't want to pay your full consulting rate. If President Obama wanted to hire a consultant, you don't think he'll pay the full rate? Of course he will. It's just these are little minions over there, and they were trying to get me. They said, "wow, the rate for MIT professors is $600 an hour." I said, "you telling me my colleagues are over there working for $600 a day?" Or five or $400 a day or something — it would be $400 a day. Oh, yes. Anyway, so hey, you know, I was working on the project, it's sort of fun. I've never built a titanium submarine before myself, so I worked on it.

And then I get the contract for the next year, and it's the same rate. And I said, guys, forget it, okay. At this point I had done all the fun stuff, now it's just dog work. I said, we got to come up with a — I wrote a letter back, and I said, it's time for Draper Lab to come into the 20th century, okay. This is in the 1990s, right. Apparently you got copied — I'm told it was posted on every blackboard, because the engineers in Draper were completely frustrated with their accounting department, okay. They knew the rates were ridiculous. The same thing when I worked with MIT Lincoln Lab. Oh, you know, we have —