WM_S2014_14

Ready, I'm going to go ahead and start there. Okay, so this is just the review of where we were. We talked about 3D [3D — three Ds, i.e., the three methods/conditions] yesterday, and you want to be at least above 50 degrees Fahrenheit before you start — well not steel, just remove the layer of moisture that you can't see. But as you know, if you get down to frost temperatures you'll see the ice form on something in a turn point. But people, before, you know, at 33 degrees there's actually a layer of moisture in the [metal] that you might even be able to feel it with your finger but you can't see it because it's a water barrier. But once it freezes you can see the frost. Well there is a layer of moisture even at 40 degrees, and it's more moisture than you really should be blowing through.

Between 100 and 300 degrees F typically we're just trying to slow down cooling and keep the thing in this hundred to 300 degree range. But we know that will accelerate the [hydrogen] evolution, with bubbles of hydrogen coming in on the surface, and I show people unless you're an example stuff — that you can speed that up, it allows the hydrogen to escape. So in the first one, if you take this Venn diagram, and I hope you've noticed I've been doing them consistently — well, stress, hardness, and hydrogen content. Other people might put them different, but for this class I've been keeping stress at the top, microstructure and hydrogen content in that order.

So what are you doing on the first one? Removing hydrogen. On the second one we're moving — we're trying to shrink that circle. The third one, if you're above 300 degrees you get tempering and you reduce that. Tempering will reduce the hardness, and you're also trying to let the hydrogen escape even faster. But the real thing is you're trying to temper, so you're reducing the hardness by tempering, you're also getting [hydrogen] there. So you're actually reducing two circles. You haven't changed the residual stresses, okay, that support that goes.

And in fact this is a tempering curve I showed you before that basically — if you take a piece of steel and you quench it to its maximum hardness, for a 10.1 [10-series] steel the maximum hardness, which is a function of carbon content, maybe Rockwell C 37 to 38. You get above 150 C, 300 degrees F, and the thing will start to soften by tempering. And you temper between 300 degrees and 600 degrees Fahrenheit. You avoid 600 to 1100 [degrees] because you get something called temper embrittlement in steels. Which this is not a [Metallurgy] of Steels — this is supposed to be a welding course — but eventually we're going to see that we do post weld heat treatment up here in the 1100 to 1200 degree range.

And in fact, I might as well do it — here it is up here on the board. Post weld heat treatment is another process that we might do for difficult-to-weld steels, where we heat them in the 1100 to 1200 degree Fahrenheit range, typically for one hour per inch of thickness. And that one hour per inch of thickness — you can measure the temperature on the surface, you put it in some furnace, and when the surface gets up to 1,100 degrees you then now need to soak it so the temperature on the inside gets there. We might put a roast beef in the oven, I might get it done on the outside, the inside still rare, okay. So you need about one hour per inch of thickness in order, with the thermal conductivity of steel, to allow it to heat up in the center. I want medium roast beef, not rare or rusty.

So the problem here is for something between one and three inches in [thickness], one inch per inch of thickness is about right. You start going to 10 or 12 inch thick steel and it turns out you can sort of get away with a half hour per inch of thickness. Nowadays they have little computer programs that will tell you exactly how long it takes to get the inside up to a reasonable temperature, but in the old days, one rule of thumb you'll often hear is one hour for each inch of thickness. That only applies in the couple of inches thickness range; above that you can do a little quicker.

You tempered to lower the hardness. This is what this does in this temperature — you're tempering to lower the hardness, which means I'm really lowering the [hardness] under this subject. You're not losing the martensitic structure, but you're getting to be relatively soft compared to the maximum hardness, and that's going to be good in terms of lowering the hardness circle, okay. We're also relieving the residual stresses. The material at 1,100 to 1,200 degrees — this is called stress relief — and there are locked-in stresses, there's enough creep at the steel actually loses about — the steel actually loses about half of the team strength when you get to eleven, twelve hundred degrees. And that means the steel can actually start to deform and relieve those locked-in stresses. So a piece of steel that, if you quenched it and cut off the top layer, it might blow up because you took off the tensile or compressive surface layer — if you put it in 1,100 to 1,200 degrees and then machine it, it will stay flat. It's been stress-relieved.

You're also removing hydrogen big time, because you got eleven, 1,200 degrees, the hydrogen is going to go out there actively. So what you've done with a post weld heat treatment is you reduce all three circles, so you get no overlaps. Over the — remember, that's the overlap moment, which is what causes — where you have a danger of cracking, in theory. Our little diagram. So those are the reasons for post weld heat treatment.

Now let me tell you, post weld heat treatment is not cheap. Post weld heat treatment will cost as much, usually, as welding the part to begin with. Post weld heat treatment isn't hard if you've got a part the size of this table, and you've got a furnace bigger than that. When you stick it in the furnace in air, you might get a little oxidation. You can clean that off with a sander, or you can peen the surface and knock the scale off, or whatever, you can paint over it if you want some places. It's not a thick oxide, it's not a terribly big oxide. There's a black oxide that's fairly adherent — it's what we call black iron piping, as they have.

In any case, so stress relieving is not too bad on something the size of the table. But what about a pressure vessel the size of this room? It's a bit of a problem unless you have the furnace. I think some of the pressure vessel shops do — I mean, there are heat treatment furnaces at a pressure vessel shop that could be 20 foot square by 30 feet long. In fact, I printed it out — perforce got to get up off the printer. If you go to Cooper Heat — Cooper Heat is now purchased by something else, but they for many years were a big company around the world to [do] this. So they had a picture of [an] 118-foot long pressure vessel that will be stood up as part of an oil refinery. Probably 3 or 4 inch thick steel, probably going to have one hundred atmospheres type of gas in it, 1,100 degrees or something, whatever. But it's thick enough wall, it's all welded construction, you're going to have all kinds of residual stresses there. If you don't do something about the residual stresses you have problems. You can have fatigue problems because the residual stresses — the fatigue's already got a head start. You might apply some other stress in service, but you've already kind of given it a head start with tensile residual stresses in some locations.

Cooper Heat will come to your site for a fee, and they will build a furnace around your part. And so it could be — the 118-foot long, when they actually — they have a little kind of railroad tracks, and they had a box furnace at 118 feet long full of insulation around it. Other cases, they made it a local stress relief. You have a nozzle in a pipe coming off a big pressure vessel and you just have to do the nozzle. Well, they actually have heating blankets, which are just little ceramic pieces with wires in them, and great big — bigger than a welding power supply — pump a couple of thousand amps to that resistance heater and heat the whole thing up locally. Put a bunch of refractory blankets around it and locally heat treat it.

There are all kinds of rules on local heat treatment, because local heat treatment somewhere — you're going to get into this intermediate range where you get temper brittleness, okay. You don't like to be in the 600 to 1000 degree tempering range, between here and here. You're going to be here if you're doing local heat treatment — obviously something's cold back here, something is in [between] — and you could damage your steel, depending on the type of steel, by doing local stress relief. So sometimes you have to stress — stress relieving the whole vessel. And in fact, there is really no pressure vessel so large that it can't be put in a furnace for stress relief. There are some vessels that are so large that we can't afford to do it.

And one pressure vessel that is so large that we can't do a thermal stress relief is called a nuclear submarine, okay. We're talking three, four hundred feet long, thirty feet in diameter. And even if you could have a vessel that big, it's already got a few other things inside that don't want to go up to eleven to twelve hundred degrees. You're going to destroy the insides of the ship, okay. So you can't really put it in a furnace. The submarine steel — you can locally — that's local stress relief, and you have to worry about what you're doing to everything else. Yes — the problem with the submarine steel is you've already heat treated it to get a maximum strength, and you want that strength. Pressure vessels — very rarely, not saying we don't use quench and tempered steels for pressure vessels, we do — but, you know, a nuclear submarine you're basically using all that strength for the strength of the vessel, to be able to dive.

So they don't use thermal stress relief, they do stress relieve these welds, but they do it — with thermal energy you have to go to another type of energy. What type of energy would you use? Mechanical. The first deep dive, okay. The first time you dive in that submarine and go a little extra deep, okay, deeper than you want the regular officers of the Navy to take it — but you're going to do a proof load test, and you're going to take it down deep enough that you will mechanically relieve the residual stresses, okay. So you relieve the residual stresses. It's not them — don't stress relieve the whole [vessel] — but we don't do it at temperature, we do it mechanically, if we do it by taking it down very low in the ocean, okay. I can't tell you how low — comes again, that would give you some idea of what maximum depth capability is, which is classified, okay. It's classified because then they know how to set their depth charges and things like that, okay.

Third reasons for this — it's actually more operational. You can't go maximum speed at maximum depth, because if anything goes wrong and you all of a sudden start going down, you can get a little too deep too quick. So your maximum speed in a submarine is at some intermediate depth, above the maximum [depth]. [You're] walking on the edge. But anyway, so we stress relieve. But the best quality control technique — one of the best quality control techniques I know they use on nuclear subs, and that is the top management of the shipyard has to go on that first deep dive. They may not have made the welds, but they're responsible for [them]. And yes, some Navy people also have to go. So you have — and you invite some VIPs to your first deep dive — and this is a very good way of ensuring quality control in the shipyard, okay. This is where the rubber meets the road in terms of quality control, and the top management gets paid the big bucks to put their life on the line for the quality of the product they just saw. And obviously the Navy thinks it's a good way to — I don't know what the management's your pair of things [says], but they — why, they want to make sure that's a very good weld that they made on those balls, okay.

I told you — I'll tell you some stories of examples where someone may have done something, may not have done something. One example: when I was doing this to a certain standard, 'cause some people yesterday said they like the stories, okay. You have to remember the stories are good and they help you remember long term, but they don't necessarily help you — there's, it's not as dense an information transfer, but they actually sometimes have other information, does a lot of it.

So one type of question is, when do you post weld heat treat? When do you have high enough residual stresses — this part right here, that circle gets big enough? And I think I mentioned to you that somewhere around an inch or two in this thing you can have residual stresses. In fact in the boiler pressure vessel code, which is the ASME code for building pressure vessels, they do not require stress relief below 1-1/2 inches thick, except for certain types of weld geometries like a fillet weld, an inch and a quarter — maybe getting some of this backwards but anyway, somewhere — some welds an inch and a quarter you don't have to stress relieve, some welds an inch and a quarter you do, [at] 1-1/2 inches you pretty much always have to stress [relieve], okay. So it depends on the geometry as well, and they have these rules. It's UCS-56, I think it's the table — UCS means unfired carbon steel, 56 of the table in that section, unfired pressure vessels and carbon steel.

So what's the story on that? Well, the story on that is the Helms project in California. So the notes — I'm from California, so you've heard of the Sierra Nevada mountains, okay. So Pacific Gas and Electric, you've heard of them, okay. Big utility on the west coast. So Pacific Gas Electric around the late eighties, nineties decides — they have all these nuclear plants they built on earthquake fault zones and things like that just for safety, nonetheless — they decided they need something to store the energy. Because the nuclear reactor, it likes to produce energy 24 hours a day at a certain level. It doesn't like to be started up and slowed down. It likes to just slow and steady wins the race. This is the tortoise paper — that the story, nuclear reactors, tortoise, that just keeps on plugging along. But people, they like to come home at 5 or 6 o'clock, start fixing dinner and turn the air conditioner on, and so there's a peak in usage each evening around five o'clock. There's a peak in the morning before they go to work. In any case, usage of electricity is not uniform.

And so they like to level that out, so the nuclear reactors can work the way they like to work. So they decide to build a pumped water storage project. We have pumped water storage projects in western Massachusetts. Any place you got a hill and water coming down, you could build a pumped water storage project. And all it is — and they've been building these in California for a hundred years, okay. They call them penstocks.

Yeah — well, I've heard of these pump water storage things already. A penstock is just a pipe coming down the side of the mountain. So if I've got a hydroelectric plant, I got water up here, I have turbines down here, and I have to connect them with a pipe — then that pipe is called a penstock, okay. Some of these are pretty big. In this case, we have a lake there already existed in the Sierra Mountains at about eight or nine thousand feet, and another lake at about 6,000 feet, and they needed to build a penstock between the two of them. For the energy requirements and the size of the lakes, they decided this should be a 22 foot diameter pipe, okay. This is a pretty good size. And anyway, 22 foot diameter pipe, 188 psi water, actually so hard that you need feet underneath it [for] about 7,000 feet. Because most of this penstock, they just blasted a tunnel right out of the rock and then lined it with steel and concrete and built this thing, and now it's supported — you don't have to have lots of structural support for it. But there was this one canyon that they had to cross, as I remember it's about 300 feet, plus 160 feet, something like that.

So they had to cross this canyon — nature didn't give them rock there to support it. So they came out of one tunnel, across the canyon, into another tunnel, and they had to build a 22 foot diameter pressure vessel that would not be supported anywhere else. Well, it was convenient somewhere in the middle here they could put a manhole at the bottom of this 22 foot diameter pipe, so whenever it was drained they could take the manhole cover off, put a ladder up there, and somebody could crawl up in there and walk around. Then to start — what fun, right? Well, so they were welding this 22 foot diameter pipe up in the mountains. You don't just kind of put that on a truck and carry it up to the mountains because the roads aren't there, okay. You gotta build it on site, field-fabricated 22 foot diameter pipe. This particular 22 foot diameter pipe — a pipe section with the manway, reason I remember it — was about 160 feet or 140 feet or something across the canyon. Some of them were 40 foot long sections, as I remember there were three of those, with a 20 foot long section in between.

So here we have granite mountain. There's on the bottom of this there's a little manway, okay, little manhole where someone can crawl up in there, down about the six o'clock position, mm-hmm. So they built this thing and they had to build it on site. They roll the plates down in Fresno, some big shop, they take the sections of rolled curved plate up here, they set up something, they start welding them together, okay. Now these — as well, something like that — together, easiest to do vertical welds or won't [break] your lunch to do them well. So you stand up the pieces, several of them, weld them together, so far that goes. So they'd weld them together and they were starting — they cut a hole in here for the manway nozzle. And they were starting — well, it was getting to be winter, the snows had started in October, it was now mid-November. And they were going to shut the whole thing down just after Thanksgiving. They wanted to shut it down before Thanksgiving but they were a little behind schedule. And they couldn't keep it going all winter because there would be 20 feet of snow there by December.

So they got the guys welding inch-and-a-quarter thick — nozzle here, doing fillet welds on this thing. And so the fillet weld looked something like this. This is the flange and this is the pipe coming down, okay. So they're making that go well, and this is big enough — it's stick weld, something — let's stick weld it in the field, stick weld. So you have some hydrogen in the air. Well, fellas, it's a pressure vessel steel, but it's sort of a mild steel, might be points requirement and we're talking a medium carbon, need some 3D, okay. So they preheated it — but in fact, and this is actually the wrong weld, that was the plan. This is another weld up here where it goes to the top of the pipe, and they were welding the overhead position. And, you know, they preheated it, because we had pictures showing three feet of snow on top of this thing, on the inside of this pipe. This line horizontal, okay. Well, it must have been preheated to 100 degrees, because she had three feet of snow to get you to quarter it away, right? Okay, well, the supervisors had told their families they were gonna be gone for Thanksgiving. So they left in mid-November and left the welders to do the work themselves.

Okay, so they didn't bother to shovel the snow off the inside. There was a lot of work, sir — like, you preheated it on Friday and we're welding on Monday, right? So they just welded this thing with snow on top of it, an inch and a quarter away. I don't think they got very good preheat, okay. And they're welding with stick electrodes, and who knows, maybe they're not even drying the stick electrodes — there's no supervisor around to check, right? They're just finishing up, and the sooner they can finish up the better chance they get to be home for Thanksgiving, right? So it's not as if they're trying to do the best job in there, or anyone's looking over their shoulders. So they weld this thing and they let — you know, they get out of there before Thanksgiving. The whole thing gets frozen at about minus 20 for the rest of the season, okay, the winter season. They come up in the spring, snows have started to melt around the April timeframe, mount this thing across the canyon, and they load it up. 188 psi.

20 hours later — there's some workmen there in the little canyon, and they have a little porta-potty. Guys coming out of the porta-potty just finishes his business, he's looking about 50 yards ahead of them at this crossing, and he hears a big crack. He sees water streaming out of this thing, and he starts heading for high ground. And he makes it to high ground along with a couple friends, but in the meantime the whole thing blows apart, throws fractured — oh yeah, they were supposed to post weld heat treat it. They didn't post weld heat treat the living body, okay. So they had plenty of residual stresses. They hadn't done the post weld heat treatment. It probably had plenty of hydrogen because they didn't preheat properly, and it had a high hard microstructure because they didn't [preheat] it properly. So they did about everything wrong you can imagine.

And if they knew that — it had gotten to 188 psi, didn't crack for 20 hours. So that's where Tom [Eagar] says hydrogen cracking, no, delayed cracking — hydrogen has to diffuse and get to this place. Even though it's nine months later, this is the oldest hydrogen crack from [the] welding. But you got to remember, they froze it for most of the winter at minus 20. The hydrogen can't get out. It doesn't do a lot of damage because it's stuck, but it can't get out either, okay. So what does — it warms up, and then you stress it at the applied stresses, and wait 20 hours, I think it's 20 hours, okay, more than eight hours, and how recovers. But the thing was apart.

Well, they had a valve up at the top lake in order to shut it off, except it must be a 22 foot diameter valve, and they didn't have anybody at the top lake to do that. So these guys hustled up there in their pickup truck — they have these dirt roads — and they get there, and it takes — it's not a small valve, it takes a half an hour to close. In that half hour the upper lake dropped 50% of its height, and the water's coming rushing down the canyon, into the canyon, coming out of this end at 188 psi, 22 foot diameter, 4 million horsepower, okay. Hits this granite wall, erodes it away a hundred feet back. They just made a 140 foot canyon into our area, even crossing, okay. In the meantime, the water has to go down the canyon, down through the woods, and end up where it's supposed to end up, this — in the 6,000 foot lake, okay.

Now, a few bears and raccoons and other people were very upset by this, okay. It disturbed their habitat quite a bit. It also built this beautiful sandbar beach we're about a hundred yards down at the lower lake, okay. So, but it's all about failure to preheat. So you're post weld heat treat and hydrogen — Pasiphae guess electric — okay, Pacific Gas Electric — actually, I'm sorry, Pacific Gas Electric was the owner. The company that did it was American Bridge, division of the U.S. Steel, okay. They were the contractors back in those days.

Now I'll tell you the rest of the story a little bit. But it went to trial in Fresno, California in the early 90s. Why did we go to trial? It was a hundred million dollar loss on an eight hundred million dollar project. Pacific Gas Electric really didn't want to have to take it to trial, but they wanted to recover their money as a loss. And they went to the California Public Utilities Commission. They said, we want our hundred million dollars that it cost to repair this thing as part of our hundred [—] eight hundred million dollar project, and we want to put it in the rate base, and this will raise everybody's electricity rates by some tens of a cent or something. And the Public Utilities Commission said, if you say it was a defective weld, and if you don't take American Bridge to court to sue them and get your money back, we will not allow this hundred million dollars in the rate base, okay. Wow.

So here's a hundred million dollar extortion that the Public Utilities Commission is giving to PG&E — listen, Pacific Gas Electric'd be delighted to just forget about the lawsuit and just put it in the rate base and just spread the peanut butter all over all the citizens of California. But the Public Utilities Commission is there protecting citizens. So they go to trial. Six-week trial, that was the last witness, but in any case we lose. PG&E definitely cared, they had tried. And now they go back to Public Utilities Commission and say, well, the jury decided it was a foundational problem. Our supports, foundations for the support pipe were inadequate, they shifted, and had nothing to do with high poverty level. So long story about how American Bridge argued that — in my opinion they were wrong, but some of our metallurgists, which are some of the highest-paid that were just in the world at the time —

We're also long unpacked. At the time I was charging like 150 bucks an hour, this is like — and the top-notch network is from this big [outfit] located [in Palo] Alto, their experts return for 275, almost double. Why? And I used to go around saying, well, if you get $275 an hour for the wrong answer, what's the right answer worth, hey, you know? That's when I learned a lot, from that, 25 years ago, about how to bill, okay. You give someone the right answer it's worth a lot of money. You give the wrong answer, it's not worth two cents. Who says — what's John Wulff's thesis [advice]?

Anyway, so anyway, we lost, knowledge — and we lose. This big-shot metallurgist from this thousand-person consulting firm had said three days after — in the second day of the inspection, three days after the failure he had been there — and he said something about it was American Bridge's fault. Good idea to start slandering them in the paper after only 72 hours. And so they not only didn't have to pay a hundred million dollars, they received seventeen million dollars in punitive damages for slander, okay. Because the [metallurgist] just hired by PG&E to investigate it was on site, basically made a press announcement, okay.

In the meantime — I firmly believe if anybody wants to — I'll go seconds when I have an hour, not today, but let me tell you the John Wulff story, because it's really a good story I learned. He told it to me as a freshman. [Tom looks for a slide or note.] I'm trying to find the lake. Yeah, it's called the Helms project. AGLMS, that's all I know, I never knew the names of [them]. I find it amazing. It's a pumped water storage project, okay. If you look at pumped water storage Helms project California, it'll show up somewhere.

Anyway, so the John Wulff story about charging fees — my favorite is actually [Steinmetz]. Mrs. — uh, the story of [Steinmetz]. Do I know who Steinmetz was? So Steinmetz was a great mathematician who basically learned how to use imaginary numbers to explain what AC electricity [does]. And he worked for General Electric, but he also consulted, brilliant man. And at one point around 1900, Westinghouse or somebody put in this big generator at General Electric or something. It's time to hire Steinmetz as a consultant. And for two days, he asked for seven drawings and he asked for a chair, and he sat in front of this huge generator — built around like [a] electrical generator — but the drawings, tried to understand why this thing wasn't [working]. They just installed it, it wasn't working. And finally on the second day he asked for a piece of chalk. And on the second day with the chalk he draws a little box and puts an X in it. He says, cut the steel case open here, cut this connection in the copper windings of the generator, and you know, sealed back up, and that will fix it. And they did, and it worked. And General Electric was very happy, and the people of New York now had [lights].

And Steinmetz sends a bill to General Electric for $1,000. Now in 1902, or 1910, whatever this was, a thousand dollars was a lot of money. And the chairman of General Motors [Electric] writes back to Steinmetz and says, well, we thank you for fixing the problem, but how can you justify a thousand dollar bill? It's time, that's right — writes back: chalk, one dollar. Knowing where to put the X, 999 dollars. Okay, so it's what you know that you're hiring when you get a consultant.

So John Wulff, late 40s, early 1950s, gets a call one morning. Boston Edison, who generates electricity, had just had overnight all the stainless steel in their an electrical generating plant just fall apart by corrosion. And John Wulff gets a call. And he says, I know what your problem is, but I'm not gonna tell you the answer unless you agree to a $5,000 fee. Well, $5,000 in 1950 or whatever it was, was like two or three years salary for an engineer. It's a lot of money. And they're kind of, well, if you're right you'll pay.

So it turns out John Wulff was an interesting old [guy]. I'd had energy — when I was a freshman he had the office that I met everyone for the first time, and I was taking [his class] anyway. He was a crusty old [guy], okay. And in fact, there in World War II he knew there was something going on at MIT that involved uranium, okay. But he didn't know what it was, because he was German, and they didn't let anybody of German or Japanese [extraction] know anything about certain things. But John Wulff had been studying trace metals in different ores in the world, and he knew there's something interesting. For the departmental cocktail party reception or something one night, he says, you know, there's a lot of uranium in the oil in the Balkans, okay. And the next morning, he said there were two FBI agents in his office asking him what he knew about uranium. So he was right — there was something about uranium going on, and part of the Manhattan Project was here. Morris Cohen, John Chipman were working on things, anyway.

So anyway, why — get back to Boston Edison. So John — they agreed to the $5,000 thing if he's right. And John Wulff then tells them, he says, well, you just started bringing in oil from Venezuela. They said, yes, this was the first shipment of oil they just opened up — the oil [arrived]...