WM_Su2014_32

Ready, we'll go. So um, I'm gonna, I'll do a little introduction since the people on the tape didn't see this before. So in 2004, 2003 thereabouts, the National Science Foundation wanted to know if they could afford to build a new Alvin submarine to do these deep submersion science over the next fifty years or so. And so they commissioned a study from the National Academy of Sciences to look at future needs, and a lot of that, mostly it was the future needs of the people who were going to be using the vessel. But they also wanted to know if the twenty-five million dollars they had in their budget was sufficient to build the ship.

So it turns out, um, I'll hand it out, and you have here an appendix which I didn't really, well I sort of generated this version of it, but most of this was I stole from a report Woods Hole Oceanographic, which wanted to be the operator of this, had done a study in 2003. And they estimated it would cost 6.5 million, okay, this is the total down here, for a new Alvin submarine. Now it didn't, if you read the footnote, it didn't include design, assembly, certification, and testing. Okay, so they kind of left a few things out. But they had just finished building the Jason-2, which is a remotely operated vehicle that went to similar types of depths.

And the two big costs, or of course Jason didn't need a titanium hull because it didn't have any people, it was remotely operated vehicle, but if you come down here somewhere, you're going to find syntactic foam, a quarter million dollars, out of a Jason cost of 2.3 million. Okay, the total cost of the Jason. But you can see how they broke all this stuff down. And then Woods Hole had estimated the weight of a new Alvin and the cost at 6.5 million. Now, you know, I looked at this, I said you got to be kidding me. You can build an unmanned that weighs 7,200 pounds for 2.3 million, and if I calculate that, I think that's about three hundred a pound okay. Or you could build a manned vessel which, you know, has got to be more expensive because there's certain things you have to have on a manned vessel that you don't have to have on a remotely operated, that weighs thirty-four thousand pounds, and that's going to cost you two hundred dollars a pound. So I didn't believe this from the get-go.

But you can see that on this one, ten percent of the cost, or eleven percent of the cost, was syntactic foam. What is syntactic foam? Anybody know? You don't know.

Student: [inaudible response]

Yep, exactly. And fortunately for the Navy, the um, uh, oil industry is using lots of syntactic foam on their jack-up rigs, you know, in things. Uh, but you need it for buoyancy, and basically you start out with these little quartz spheres that can take the submergent pressure without collapsing. Their sphere — they're hollow spheres, not, I say made out of quartz, they're made out of a glass, and I think it's certainly, it's certainly a very high silica, so it has very high strength, relatively high strength. And then you infiltrate this whole thing with polymer resin. Okay, the polymer resin has to be something that won't biodegrade or get eaten up by marine organisms, because this stuff is going to be wet for fifty years right.

So it turns out they actually end up using a two size sphere, because you can pack more densely. You can prove to yourself you can pack spheres more densely if you, if you packed everything almost completely tight, you would still have interstices. And turns out if that one is one sixth the diameter of these, you can fill up some of the holes. Uh, it turns out you can pack with about fifty percent density with one size of the monosized spheres, and then you infiltrate with plastic. But if you use a distribution of sphere sizes, you can get higher densities, and I think they're up to like sixty or seventy percent, which is pretty incredible.

Now, and the relative density, I can't, I used to know, it was like two pound, two pounds per cubic foot okay. And if water is, uh, or they're approaching, I mean you can look it up on Google, it'll probably tell you. If water is eight pounds — no, sixty-four pounds cubic foot — so it was like eight pounds, weighs eight pounds, I don't remember. No, two, maybe it was twenty pounds a cubic foot, it was, yeah, that's probably right, it's about one third that, about one third the density of water. So you get lots of buoyancy. I mean, it's not an air tank, but if you start trying to build air tanks and pumps and everything just blow out the air tanks, you wouldn't be able to do it. So they — and the other thing about syntactic foam, you can basically machine it okay, it's a little hard on the tools, but you can use carbide tools and stuff, so you can machine it to whatever shape you want.

So if you actually looked at the Alvin itself, inside this outside hull, of course you have the inside hull which is a sphere, and that sphere is right here, and here's your propulsion, and you've got all kinds of other stuff back here, and you can go to the web and get that stuff. But anyway, every space in here is filled with syntactic foam machined to just the right shape to fill up all the space inside that external skin. So syntactic foam was pretty expensive, but just in general I didn't believe it.

So if you actually look in here, my estimate was, basically said — Woods Hole had estimated a titanium sphere cost of two million dollars, okay, out of 6.5 million. So I said, but they didn't have design, assembly, certification, so I said well let's double the 6.5. I mean, they wouldn't, you thought I was going to go through and do the design analysis and the cost on all this? I mean, that would be a half million dollar job and I'm a volunteer on this. So I had to think of a nice simple way to do this. I said well let's take the 6.5 and make it 13 million if you include certification and design and assembly and all this other stuff.

And that's about right. I gave you this thing from 21st Century Defense Needs that talked about cost of welding is ten percent, inspection is ten, you know, management is ten to twenty, and all these other things. So if you do that you kind of say okay, well let's take the basic stuff they're talking about here, the components, and then double that for all these other things. And I said, I basically said, what did I, what'd I write — "the greatest risk in meeting the cost estimate of a new HOV is the cost of fabricating the pressure hull." Okay, of all these things on here, the two million dollars for the pressure hull was the biggest problem.

And Woods Hole had estimated two million dollars in the 2001 estimate of thirteen million. I'd taken the six and a half and I just doubled it, okay, to thirteen million. "The sources of the original titanium plate and the forging facilities used to fabricate the Alvin and Sea Cliff titanium pressure hulls in the 1970s no longer exist. The United States has extremely limited industrial experience in welding heavy section titanium. Soviets and Japanese have some." We could go over there and have them build it for us. Soviets could have done a great job, believe me, they built these — other, anyway.

And so I said basically, um, the final report, um, I basically came up with sixteen million dollars. But the sixteen million is basically almost ten million above the six and a half million that Woods Hole was trying to sell the NSF. Because, oh well, we're not gonna, you know, we don't have to pay — who's gonna design it? They expected me to design? I mean, you know, I'm not gonna design it. Um, so, uh, we determined that for sixteen million you might even be able to buy a man submersible and a, maybe another ROV, okay.

Uh, the other, one of the other constraints was if this thing was going to weigh 34,000 pounds, you really couldn't use the Sea Cliff because the Sea Cliff was going to weigh 40,000 pounds, because it was a bigger hull, it'd have to be a bigger Alvin. And I think I mentioned, you have to come up with a new surface tender. That's a hundred million dollar ship. And they certainly weren't going to come up with that. They're going to have to do something about that surface tender because that ship's not going to last for a hundred years okay.

Then, and they're about to start it, and uh, we had the, I don't know if it's the financial collapse, actually I guess the big question became, who was really going to fabricate it. And we did do a little bit of looking. Ladish Corporation [Ladish Co.] in Cudahy, Wisconsin does big forgings for the oil business. And then there's Cameron Iron Works down in Houston. And so there's some people who could make these things, and they make things for the oil industry, but none of them had any experience with titanium since the 1970s when the Navy was funding things. And all that experience had gone away, those people were retired long ago.

And we didn't get a, well it wasn't our job to go out and get quotes. But when they went out to get quotes, these companies didn't really want to do it, and the first quotes that came in after the study — and we had estimated sixteen million — they were coming in at like twenty-four million. Not for the whole thing, if you kind of — so my sixteen million quote, even though I thought it was two and a half times what Woods Hole, you know, I'd done two and a half times what Woods Hole had done, uh, my quote was looking pretty lame at that point. Uh, the final cost as I understand it was about forty million.

Okay, so Woods Hole at six and a half million, cost forty million. Tom Eagar thought he was being very generous by going up by 250 percent over Woods Hole. Uh, but Woods Hole didn't make any sense. Like I said, two hundred a pound for a heavier vessel that's man-rated, as opposed — you know, common sense said that was a stupid quote to begin with. Okay.

So let's talk a little bit about titanium welding in general, and first talk, like we did in most cases, about titanium alloys. There's not as much titanium used. I don't know if — I should have my little thing I showed you in the very beginning that talks about pounds of material used in the world, or tons of material. So remember steel is 1.5 billion. Titanium's 165,000 okay, tons per year, of titanium metal. Now titanium oxide goes into paint on the walls okay, so far as that goes. So they might mine half a million or more, million tons of titanium, but most are paint and pigments and things like that. Titanium metal is about 165,000 tons a year.

Obviously a lot of that goes into aerospace. The next largest application is corrosion resistance. Marine applications are very small okay. But there are marine applications, and the Navy has been leading the charge on heavy section titanium for the last sixty years okay. Titanium comes in a number of grades, one through twelve, according to the ASTM which grades these things. And it turns out, I'll show you later — well actually be able to show you here — the difference in grades one through four is basically just increasing oxygen content. Just like carbon hardens steel, iron, oxygen hardens titanium without great reduction in properties. Now if you, the higher hardness the lower the toughness and whatnot, so there are some trade-offs.

Now it turns out for some chemical applications you actually put palladium in there. Palladium is the same price as gold okay. But there are some very severe corrosion applications where you can get pitting. And if you put a little bit of titanium-palladium in there, you know, you don't usually find something like palladium in a metal alloy that's not a precious metal. But for some chemical plant applications even titanium is not good enough resistance for pitting. In other cases they use nickel — nickel is right above palladium in the periodic table — and it's obviously much less expensive.

Uh, there's titanium six-four. I'll talk about the six-four type of alloy in a little bit, but it was developed over here at Watertown Arsenal in about 1945 or '47 or something, when titanium first started to become available in larger quantities. There is this alloy titanium-6211 — six aluminum, two niobium, one tantalum, and 0.8 moly, and they call it Ti-6211. That is the U.S. Navy 100 ksi strength material. Has no other application that I know of other than the Navy buys it, because they did a lot of work at David Taylor to prove it out as the Navy's choice of strength, toughness, weldability, etc., back in the 1960s or so.

Classification of alloying elements. Turns out titanium, like steel, has a transformation. I didn't, well I can show you this okay. With most alloys, or with many alloys, titanium goes from, uh, at low temperatures on a phase diagram with alloy content and temperature — uh, titanium at low temperatures is hexagonal close packed, we call that alpha phase. At high temperatures it's face centered cubic — not facing, body centered cubic — we call that the beta phase. And as you read about titanium you'll see alpha and beta and alpha-beta phases. Many of them are two-phase. Titanium six aluminum four vanadium is two-phase alloy, alpha-beta.

Um, and you, one thing about titanium is, one of the most reactive elements with gases in the air okay. I guess I can tell a story about that, but — in any case, the alloying elements that stabilize the alpha phase are aluminum, oxygen, nitrogen and carbon. The beta phase is most other things in the periodic table. Um, and the neutral ones are tin and zirconium okay. So we use tin and some zirconium in some of the fancier aircraft alloys and whatnot. You do get precipitation hardening varieties of titanium, but they're mostly used sheet metal in the aerospace industry so far as that goes.

You can dissolve lots of aluminum into titanium in the beta phase. So here's the titanium-aluminum phase diagram, and here's the — oops, let me make sure I got the right side, yeah — this is the titanium side. And putting aluminum in stabilizes beta, okay, more aluminum you put in — actually I'm sorry, stabilizes alpha. But you can put a lot of them into the beta phase. If you put too much in, you'll get intermetallics, which you can use as precipitation hardeners but in your beta alloys. But in general you want to balance these so that you have your transformation where you're starting to work it, so you can use the advantage. Just like in steel where you can change the grain size with temperature.

Okay, in steel you go to FCC and then you come down and you can get a change of grain size to FCC [BCC]. In titanium you start with BC, you start with BCC, and you come down, you can go to HCP and you get a reduction in grain size. And if you cycle between those, you get finer and finer grains, to a certain extent, and get better toughness and better strength. So titanium does have these different crystal structures, which aluminum does not, and that gives titanium a certain advantage in terms of being able to produce different types of microstructures.

Um, typical types of properties in titanium — this is out of a materials processing book, a magazine. Room temperature fatigue properties. The wrought annealed titanium can give you strength, this is 100 ksi, this would be the strength of titanium 100, over here, and out at lots of cycles, over a million cycles. So titanium can give you higher strength than steel at lower density, which is obviously one of the reasons why you'd like it for a submersible. Because if you're going to spend all this money on syntactic foam, you'd like not to have as much syntactic foam, you'd like to have something that automatically comes to the surface.

They have built all-aluminum deep diving submersibles. The Aluminum Association built the Aluminaut back in the 1960s or '70s. You have to use high strength aluminum, and that has stress corrosion cracking problems, so that's the only all-aluminum pressure hull that I've ever heard of. So people tend not to do it. Titanium has fantastic corrosion resistance, so it has all those advantages, just so costly. Cast titanium doesn't have the same types of properties, but if you do hot isostatic pressing, you can bring the properties up so that in many cases it's approaching the base metal strength, actually sometimes it can exceed it, so far as that goes. So I don't think we need to go through that.

If we go to the room temperature properties of these — this is the pure titanium, commercially pure titanium grade 1, 2, 3, 4, that had higher and higher oxygen. You can see you can go from a yield strength that's not anywhere, not as good as carbon steel, to something that's as good as your high strength aluminum alloys, by just changing the oxygen content. And you don't lose too much in elongation, from thirty percent to twenty percent. Most people use grade 2, because grade 2 gives you 45 ksi, which is fine for making heat exchanger tubes. People love — I mean when I was your age, people would make heat exchanger tubes out of brass or admiralty brass or something like that, and they'd have a twenty or thirty year life. Well it got to be more and more expensive to replace things, so more and more of the utilities and uh, oil companies and stuff started using titanium tubing for these things.

Not the titanium tubing won't corrode — I've had under-deposit corrosion attack on titanium tubes if you just let dirt get in the system. One time in New York — I think I tell you about the titanium fire, these guys were spraying water on a titanium fire, and you just, when you do that you get titanium oxide plus hydrogen, and so you actually end up creating a bigger flame because you're generating hydrogen when you spray water on titanium that's burning. In any case, um, where was I going with — I don't remember where I was going with that.

So people, oh, you can, they got, they basically let some river water in and they got silt in the bottom of their titanium tubes, and they assumed oh well titanium won't corrode. It will corrode underneath a layer of silt okay. You got to keep everything clean to prevent corrosion, what they call under-deposit attack. So anyway, they had to replace the whole titanium heat exchanger.

Here's your Ti-6211, has a 100 ksi yield strength. The elongation doesn't look great but it's got actually — they don't show toughness here but the toughness is, it's designed for excellent toughness. Titanium six-four, 130 ksi, but it's only got 135 tensile, um, so, but this is the workhorse alloy for, and actually 150 in the aged condition. Uh, this is the workhorse alloy for a lot of your forgings that go into aircraft and things like that. But the Navy likes their own alloy. It's very expensive to have your own alloy folks, but that's what they do.

So, um, the Navy still uses titanium, of course, in thin sections for tubing, for piping. We heard about that, although you need to not mix it with some other things, we learned, so far as that goes. But there still has been an interest, although I think it's a decreasing interest, um, ever since the Alfa sub came out. The Soviets built the Alfa — became public knowledge that they had the Alfa sub in about 1980 — and they leapfrogged us in submarine technology for about two years, until these things started cracking due to the creep-fatigue interaction and stuff.

But in the meantime Congress was very upset. And at one time when Millard Firebaugh was the — Millard Firebaugh was in your program, which used to be the 13A program, and he got a, he's allowed to stay an extra year or two and got a PhD, and then he became the designer of the Seawolf submarine. When I met him he was a captain in the late '80s, and then later he became chief engineer of the Navy, and then when he retired he became a vice president at General Dynamics — it was an Electric Boat. And in fact if you look back at this, these National Academy reports have to be reviewed, and the person — one of the, list, all they do is list the reviewers, you don't know who the reviewer is going to be, it's anonymous peer review until they actually publish it. And here's Millard Firebaugh who was at Electric Boat in 2004. And he was the only person among all these who could have said anything intelligent about whether my estimates were okay on cost. He didn't dispute them. But we didn't know the Chinese were going to start buying lots of metals and stuff.

In any case, let me talk about the Alfa subs and how they might have been built. Um, the first thing I'll put up or pass around was submerged arc welding. And what is submerged arc welding? Well, this is a submerged arc weld in titanium that I made about 1977 or '78 okay. And it was made with a calcium fluoride flux. We took single crystal optical grade calcium fluoride and we crushed it up. And submerged arc welding is a process where you can take heavy plate, and you pour the sand on the surface — it's a granular flux — and then you take a bare wire and you strike an arc, and the arc is submerged beneath the flux. It's not submerged in water or anything, it's submerged beneath the flux.

It was this — was called Union Melt in 1936, there's a patent on it. And the process was used, is still used widely for steel. Uh, there's a letter from uh Franklin Roosevelt to Winston Churchill in 1943 talking about this wonderful new welding process that was helping build ships, okay, and this is the process. Invented by Union Carbide, and that's why it was called Union Melt. Well, it turns out you can weld titanium by this process. This was my first research contract as a young professor.

Turns out Union Carbide, before it changed its name to Praxair and other things because of the Bhopal disaster, had taken a contract from the Navy to do submerged arc welding of titanium. That was in the mid '70s. The Navy was still interested in trying to weld heavy section titanium, but you had to keep the oxygen and nitrogen low in order to get good toughness in your welds. And so um we tried submerged arc welding. Um, the flux was costing me a hundred dollars a pound okay. Now you can get it cheaper okay. Um, but all we could get that had the purity we needed, uh — and it would be very high, it was very hygroscopic, if it picked up moisture and you welded with it you'd get terrible oxygen levels and it wouldn't be any good.

So you were going to have some problems with that. The other problem is you would melt about a pound and a half of flux for every pound of titanium. Now titanium is lighter than steel, but in steel you might melt a third of a pound — you might melt a third of a pound of flux for every pound of steel that you put as weld metal deposit. Here we had like a tenfold difference. So I did a little economic analysis for the Navy and showed, boys this is going to be really expensive.

That's kind of where I was about 1980, when uh the Alfa sub was announced. And David Taylor started holding some conferences which were — well, the conferences were not classified because they couldn't get everybody's security clearance, but Congress was all upset, they wanted to do something about, um, could we, you know, could we build a titanium submarine and how fast and what cost and whatnot.

So, it turns out about that same time I had a chance to go to the Soviet Union with Professor Szekely. And I may have mentioned, Professor Szekely was this Hungarian who had gotten out of Hungary in 1956 during the revolt, went to Imperial College, and he had started at MIT the same week I did. He was a full professor, I was an assistant professor, but we were pretty good friends. And President Carter had started this um, an exchange, a scientific exchange with the Soviet Union. But, and then that first exchange they had a metallurgical project funded by the National Science Foundation, but really I think some of the money came from the State Department.

And the head of that metallurgical exchange was Professor Nick Grant of this department, for the whole nation. And Professor Grant had grown up in the Soviet Union until he was five years old, and he spoke Russian, and he was a well-known metallurgist, so they put him in charge of this. And the first time they had thirty-seven scientists from the United States, they all wanted to go over to the Soviet Union, see what was going on, hadn't been able to get in for years, this was like 1978. The next time they had like eight people go in that exchange. And the next time it was just Professor Szekely, and he asked me if I would tag along, because he didn't want to go there by himself. So I had — I knew I wouldn't get a chance to see the Paton Institute in Kiev for many years, because Reagan, President Reagan was shutting down this program. The National Science Foundation had a little bit of money, the Soviets wanted to keep it going, but um, we were kind of the last two.

So I got to spend a week in the Soviet Union, a few days in Moscow, a few days in Kiev. And I got to meet with Gurevich — see there's his name. Well, that is his name, Cyrillic I guess. But S.M. had been publishing — this is an unclassified translation of Metallurgy and Technology of Welding Titanium and Its Alloys. It was translated by the U.S. Air Force in April 1980. There's the translation, here's the actual book. And here's Gurevich's gift to me of his book, okay, been a little worse for wear.

It's got some interesting pictures in it. You have to keep things nice and clean when you're welding titanium, keep the oxygen out. And here's some people in the Soviet Union wearing basically spacesuits. They're in a chamber where the chamber is filled with argon, and they have oxygen pumped into their spacesuit okay. So they're welding titanium components. Now a lot of this book is how to make high purity fluxes okay. I don't have some specific things, but they have a number of things in here on how to make high purity fluxes. So I got to sit down with Gurevich for about two hours, and he would answer every question that I had. And I could tell he was answering them honestly, because he was just a scientist, he didn't care about all these politics.

But we discussed gas tungsten arc welding, gas metal arc welding, submerged arc welding — uh, what else did we discuss? We discussed — well, those are most of the things that we discussed. And I think I mentioned to you, he basically said oh no, we don't use gas metal arc. This is what the U.S. Navy had been hanging their hat on, this is what they tried to weld the Sea Cliff with. This one will give you let's say ten pounds an hour. This one, gas tungsten arc in titanium, may give you two pounds an hour. And so at Mare Island they tried this, they tried for six months, they couldn't do it, kept on having to cut out the welds, and finally they finished the Sea Cliff with — you know, Sea Cliff was about two-inch thick plate — they finished it with gas tungsten arc, way over budget, very slow, very expensive process.

I had worked on submerged arc welding, and it turns out Gurevich had been publishing about submerged arc welding in the 1960s. About 1972 he just sort of stopped publishing. Well, this was probably about the time they decided they were going to go full scale and build a titanium submarine, and he was probably told you don't get to publish in the open literature anymore. So, we were curious about how they had built the submarine, because we looked at the cost of trying to do these things and they were just prohibitive. Of course the Soviet Union had a different scale of economics okay. If some top politician or admiral or someone said do it, you might do it even though it's, you know, takes ten percent of the gross domestic product, you would do it, because you're told to. Uh, here we actually had to follow a little bit more of an economic rule on things.

Anyway, so we talked a little bit about submerged arc. Um, like I said, he answered all my questions. But it was when we got back, um, and I think it's been long enough now that I can say — I, the one time I ever really used a security clearance, the Navy actually had some evidence of how they may have built the submarine, and I can't, I'm not going to tell you how they had this evidence, but we looked at things.

And actually, it was even before that, before they had me, you know, got my security clearance and I went to see it. I was at one of these unclassified civilian conferences. It was a two-day conference at Annapolis at David Taylor. And the first day was industry day, and the top aerospace companies, Boeing, McDonnell Douglas, Northrop, and all these others were coming in, they were talking about welding heavy section titanium. The heaviest section we heard of in that first day was three-eighths of an inch. One person had done one weld I think, and three-eighths of an inch. Most of the rest of stuff, heavy to them was a quarter inch, in the aerospace industry.

So the second day, it was myself, because I'd been working on one inch thick submerged arc welding, and the um, and the Navy had been working on one, two, and four inch thick titanium. And they actually had welding tables down at Carderock that were made of uh four inch thick titanium. Well why did they have this? Because the Air Force had tried to build a B-1 bomber, and then it got cancelled, and they had tremendous amounts of titanium they had to order ahead of time. This was all 6-4, and so it was all supposedly in some unused runway at some Air Force base. They had thousands of tons of titanium in all kinds of shapes and forms of titanium 6-4 for the B-1 bomber program. And the Navy convinced the Air Force, when their B-1 bomber was canceled, to give them that titanium. And so down at David Taylor, they wanted to buy a new welding table and they didn't have the budget for it, so they just went out and got some four inch thick titanium plate and made their own okay. Anyway, seeing that table was probably worth quarter million dollars if you had to go out and buy, maybe more.

Um so anyway, we're at this conference and we're talking about things, and all of a sudden that afternoon, um, one of those afternoons, I all of a sudden realized how Gurevich had welded a lot of the Alfa sub. Because I was thinking back to having read his papers prior to 1972, and he had talked a lot, he'd written a lot of papers on electroslag welding. Now I didn't pay too much attention to it, because electroslag welding was the process that Boris Paton, the father of, or — I don't remember the first Paton, I don't remember the first Paton, who had repaired Soviet armored tanks in World War II and became a hero of the Soviet Union for repairing tanks and started the Paton Institute in Kiev. His specialty was electroslag welding.

Where he used two vertical plates, heavy section plates, these could be one to four inches thick, I mean I've seen people do eight inches thick. And you put two copper dams, water-cooled copper dams, on the side, to make a little vertical cavity. And you bring an electrode in, you put some flux in. Here now you don't have to use, you know, three pounds of flux for every pound of weld metal, you can use one pound of flux and put down fifty pounds of weld metal okay. So the flux is not a big cost.

So I had to actually asked one of the guys at Carderock come out in the hall with me, and I said I know how they did it, they did electroslag. At least that's one of the processes. You can only make vertical welds like this, but if you go down to Electric Boat today or to Quonset Point, that's how they make the basic hull longitudinal welds. They turn the cylinder — if the cylinder is this way in the water, they turn it this way to weld it. And they make four welds approximately. Actually maybe you make more than four welds.