WM_Su2014_23

Monday, because I have somewhere else I have to be on Tuesday now. So we'll finish by definition on Monday.

Okay, if you want to give it to me after class that's fine. Um, you know, if you've already ordered for Tuesday just have it delivered somewhere else and you can all have breakfast when you come in. How are — I mean after this class are you all going to come in at 7:30? No? Okay. Why not? I got in at 5:30 this morning. It reminded me of the good old days when I used to get in at 4:30. There's no traffic at all.

It's kind of misleading right now too I think because the screen is not really — yeah. So what — I can't remember, you come from New Hampshire or Maine or whatever? Uh, what time are you coming in? Yeah, but he hits traffic coming in from — you learned that the first day, right? Yeah. So, good old Boston and traffic, okay.

No other questions, let's go ahead and start. Today I want to talk a little bit about fatigue, and fatigue plot in welds applies not just to steels but to other alloys, such particularly aluminum, and I'll show you that in a little bit. But this comes out of a book on fatigue of welded structures by the Welding Institute. And the interesting thing — if I just have a straight bar of high strength steel, and I plot the strength of the steel — as a British book so it's newtons per square millimeter, but 700 newtons per square that's megapascals okay, 700 is about 100 ksi, uh yield strength steel, so this is HY 180 or HY 100, right about in here — and you see that as you go to your stress range for a life of 10 to 6, a million cycles, increases with increasing strength on about a 45 degree slope, okay.

If you put a hole in it so I have a stress concentration, the stress concentration lowers the fatigue strength, and it lowers it proportionately more for higher strength steels, and for lower strength steel it's not quite a 45 slope. If I put a weld in it and I load it transverse to a fillet weld, I get no benefit from strength, okay, in fatigue.

Now this was sort of a shock to Ford in the mid 80s. They were starting to build the Aerostar van in 1985 in the St. Louis plant, and they were starting to use high-strength steels to get lighter weight to improve fuel economy. And they were finding that people's axles were fatiguing — not a good thing, okay — on the rear axles. And so they had a huge program in the late 80s to look at fatigue strength of welds at Ford. But I mean other people in the construction business of, you know, bridges and things like that, already knew this and had known this for about 40 years. But the design engineers at Ford didn't know about it. And so they went ahead figuring that, oh, we'll get, we'll be on this curve, higher strength steels means higher fatigue life, when in fact they were on this curve — higher strength steels gave you nothing in terms of fatigue. And so their axles were — the welds on the axles were fatiguing. Well, they've since learned how to design for some of those things.

But there are a lot of design details and we're not going to go through all of them. Four different welds — this comes out of the same textbook by — originally it was Guthrie, but then it was, I came [Maddox], I guess was the guy who ended up doing the second edition of the British Welding Institute. These are attachment welds, okay. Um, obviously that little plate is not producing a lot of strength compared to the great big plate, the little vertical plate is too small. Or we've got something that adds a little deflection resistance for the flanges here. But they actually — if you look very carefully right at the end of this one, there's a little fatigue crack right here at the toe of that one, there's a fatigue crack, they're trying to show — here's a fatigue crack. Every one of these ends up becoming a location for a fatigue crack.

And in fact this type of detail right here, where they weld the end of the — they put a doubler plate on to stiffen up the beam or for whatever — that's the reason for the King Street Bridge failure in Australia, which is sort of a famous weld failure. I don't think anybody got killed but it was sort of a major bridge in Sydney or somewhere. And it just cracked. And what they had is they had a doubler plate and they had welded all around the tip. And people learned, oh, that's not a good design detail because you'll start a fatigue crack right there at the change in stress concentration.

Um, when we're doing bridges and buildings sometimes you have to splice a beam. For whatever reason if you splice a beam, here's a steel girder, so an I-beam, and so they welded here and welded here between the two I-beams. And this is actually a built-up beam, so it was welded here, and they have to put a stiffener in here because the welding is going to cause the distortion of that top flange. You distort the top flange, you no longer have the same moment of inertia, right, of the beam. You also have to have rat holes in here, okay, for — to avoid having intersections of welds. You intersect two welds of the T and that's a very highly restrained T intersection, okay. Lots of residual stress. The last weld to meet there has nowhere to go because it's surrounded on two sides by a big heavy piece of steel. And so you'll get cracking at that.

This is one I've seen a couple of times. I may have told you about the aluminum tanks on cement trucks, but this was one of the details — of, they had these water tanks on cement trucks to clean off the trucks. And they had a saddle arrangement. And right there you have a change in stiffness. And so as the liquid is sloshing around in the tank, that change in stiffness, you're going to get a fatigue crack right there at that location. So there's lots of design details.

And in fact, for years and years, the American Institute of Steel Construction had the Manual of Steel Construction. I think this is the — it's the eighth edition, okay, for whatever that's worth. And in the eighth edition it has a whole chapter, actually — since none of you — has anyone ever seen this? None of you are civil engineers. This is — this used to be sort of the bible for civil engineers, um, if they had to design things with beams and stuff. It's a whole bunch of tables of I-beams and their properties. So this is a W8 by 67, so it's an 8-inch web and it's 67 pounds per foot is the weight of this thing. And you can be down as light as 30, 31 pound — these are standard shapes. These are called W shapes. They got — they'll have, uh, angles and, uh, C-sections, channels. Just a whole bunch of tables of this.

But if you go further up, there's a section appendix B on fatigue, and in this section on fatigue there's this big table which I'm not gonna — well going the wrong way — um, big table, and then a bunch of — it's a big table of four pages, a bunch of design details for weldments. So rather than trying to do it out there, I've got a copy of it in my notes.

So if you go to the American Institute of Steel Construction Manual, it will tell you if you have plain material what your example of, uh, what fatigue class it can be — an A, B, C, D, E, or F class for fatigue. And the A is bare metal, no holes no welds. F is a very highly restrained T intersections and things like that, lousy fatigue strength. And so there are design rules in this manual. And it's the bible for someone designing — used to be the bible for someone designing a bridge or a building. If they're going to design the I-beams for the thing, and they have bending moments they calculate, they can go in here and get the moments of inertia of standard shapes. But if it's going to be a bridge, which is fatigue loaded as opposed to a building which is not usually fatigue loaded, you have to do a fatigue design.

And so this thing goes on for table after page. I guess I have two of them. Here I simplified it, and then it finally gives you actually two or three pages of these details. And here's category A, which is just a simple bar. Two I think is B, which has no welds on it. Three is C. And you get more and more complex in your details. Notice they don't like to put welds on the end of that plate there — that's the King Street Bridge failure from like the early 70s in Australia. So they don't permit certain types of wrap around welds in most cases. Sometimes you might need to do a wrap around weld because you've got a problem with moisture getting in between or something. Sometimes you might want to consider whether you couldn't grout it the end as opposed to weld it. If you're going to weld it, you're going to have to make sure you're not going to run into problems with fatigue strength.

Well it turns out the aluminum folks have copied this. Originally — well, we have, just like we have an AWS D1.1 for steel of the structural welding code — actually this is the wrong one, uh, I should have aluminum here. Oh, I didn't bring the aluminum. Okay, well there are structural welding codes. This is 1.6, this is for stainless steel. There is one for aluminum which I didn't bring with me. I had it out this morning. But what happens if you look in the structural welding code for aluminum, it will tell you to go to the Aluminum Design Manual, okay, which is put together not by the American Welding Society but by the Aluminum Association. And if I go in here, it tells me how to design aluminum structures. And what do we have? In fatigue we have the same table. And if you go over to the next page you have some of the same type of — they plagiarized from the steel guys, okay.

So it's the same principles. Um, the problem with aluminum is you get bigger residual stresses because you have more shrinkage. Steel shrinks three and a half percent on solidification, aluminum shrinks six percent. So the stresses are greater. And aluminum also has one-third the modulus and so it flexes more, uh, and that creates problems. Do you have a question someone? No. Okay. Um, so you have to pay attention to a number of factors. But the surprise for a lot of people, particularly Ford 30 years ago, was that when you weld it you lose all your advantage of high strength, okay. And that goes — same type of thing for aluminum.

The reason for that is that weld toe profile is never perfect. I mean if you could make a perfect weld — and in fact that's what Guthrie back in the 60s kind of made part of his career at the British Welding Institute. Was showing that you could get that fatigue strength with higher strength steels up if you went in and you — he did several things. One, you can machine it. Well that's a little expensive to go along a fillet weld. But if you have a fillet weld between two plates and you just put a weld in like this, you'll get a stress concentration of that toe. You may have an undercut anywhere along the length of that, you know, going back in three dimensions. So anywhere along the length is going to be where the fatigue crack starts. And there's a very high probability you're going to have a higher stress concentration than at one cross section, okay. The worst cross section along that length is where the fatigue crack is going to start.

So one of the things that he did is, he said well, you can try to machine it to have a nice smooth profile like that, and that's better. You can give a full penetration weld so you don't have the notch in between, and that strengthens it for fatigue. He went in and he actually, to get rid of the metallurgical imperfections here, he did what he called TIG dressing. So you make a fillet weld and then you come in and you weld with a little TIG torch, just remelt that toe of that weld on both sides. So now you're making three welds rather than one, but you can double your fatigue strength, okay, if you want to do that. However, it's still a problem.

He worked on this 50 years ago. But five or ten years ago, Caterpillar — one of the larger welding companies that does more welding than most others in this country, they're the ones in Peoria that have more welders, and Caterpillar employs five thousand welders in Peoria, Illinois, which is their home — Caterpillar was having a big study, because their earth moving machinery is fatigue loaded, right, big heavy welds, fatigue loaded. And they were trying to get welding procedures that would give you a nice profile that would improve the fatigue strength, because they're fatigue-strength limited. They'd like to use higher strength steels but as they try to go to higher strength steels they couldn't get the benefit in fatigue. So this was a major program for Caterpillar. A lot of proprietary work. They came to see me a couple of times and I said good luck. But they have had some luck. You know, if you spend tens or hundreds of millions of dollars on a problem you can make some progress. So they've made some progress and most of it's proprietary.

You go look at Caterpillar welds 25 years ago, you go through a Caterpillar plant and half the welding was still manual welding, because they got some of the same problems that you have on a ship — that you have lots of details and they're not just simple straight lines, okay, or they're intersections and things like that. But they've gotten to more robotic welding, they've gotten rid of, you know, some of the people. And they've got — for consistency shape sake — one of the problems if you're holding that electrode for eight hours a day, and that's what you do in a Caterpillar plant, uh, you're welding all day. At least in a shipyard you have to walk back and get some more electrodes, you know, it's a break, right. In a Caterpillar plant you just keep working.

Uh, so they've put in more robots, they get more consistency that way, and they have been able to develop welding procedures, changing the torch angle and whether they weave the bead or whatever, pulse current welding, lots of different things to try to control that shape, that profile, and get better fatigue strength. So that they can use, or gain the advantage of going, in their case from a 50 ksi material like an EH-32 or a DH-32 to a 60 or 70 ksi high strength low alloy steel, okay. They don't usually have quite the hardened ability, hardenability, or the strength requirements, because they have lots of deflection requirements. You may have deflection requirements but they're a little different, okay, than theirs.

Okay, any questions on fatigue? Yeah, it does. If you try to under match this, you're localizing the strain. I mean you actually have here a weak material and a strong material. The toe is the geometric stress concentration, okay. It can also be a metallurgical stress concentration. But if you have a mechanical plasticity concentrate or difference, you're going to be concentrating everything right there, that one spot. So under matching filler metal will be a problem. Over matching filler metal doesn't buy you anything in fatigue, but under matching could localize the strain in the weld metal, and you could get more plasticity, and so it probably makes things worse. Okay, that's your question.

Yeah, so there's, I know — well, we just talked about a redeeming social value for under matching before class, okay, these brackets on the Coast Guard cutters. Uh, because they had plenty of weld area. They were welding a high strength steel because they needed a deflection of the beam. But they had plenty of weld area holding it together, so you can go to lower strength. And they wanted, um, better corrosion resistance in the weld than any other material. And they also wanted toughness, which we'll talk about — we haven't talked a lot about lower temperature brittle materials. That's one of the problems with the Liberty ships. And it is generally a problem.

If not, let's start stainless steels. Okay. So the genealogy of stainless steels, and this figure comes out of a book by, by John Sedriks — unless John's retired in the last couple of years, he was head of corrosion research at the Office of Naval Research, so he gave out all the basic research and corrosion for the U.S. Navy. But he also wrote a book on corrosion of stainless steels which is one of the better books on stainless steel metallurgy in general.

So, in any case, this is a figure that I've seen reproduced other places, but it first came out in John Sedriks' book. And this is the genealogy of stainless steel. So around just after 1900, somewhere between 1900 and 1910, might have been someone in Sweden, was throwing chromium in steel. And they found that if you get more than about 12 chromium all of a sudden the steel would no longer rust. They called it stainless. It's not, actually. You call it corrosion — CRES, corrosion resistant something, anyway.

I have to be careful with CRES because the people in the nuclear business use — that's that same acronym for something else, okay. A controlled residual element steel or something like that. So it's not stainless. So I mean when I see CRES, I mean, whoever threw that up the other day I kind of did a double take. And then everything — whoever was you? No. You threw it up. And I kind of did a double take because he used the same acronym for two different things.

Anyway. So what soon became the norm was an austenitic stainless steel 304. And Sedriks has got the data in his book for about 1995. But 40% of all stainless steel made is 304 stainless. It's what we call 18-8. You know, you go by dinnerware and it says 18-8 stainless. 18 chrome 8 nickel. Well, so he's using 19. There's a range. And the range is 17 to 19 and it's 8 to 10 percent nickel. So 18-8 is what we used to call it, but now he's just using 19 chrome 10 nickel to use the high side of the alloy content. It's weldable, um, and that's one of the reasons it's the most widely used. 40% of all stainless steel is 304.

Now you can add other things to it. 303, add selenium or sulfur, okay. So it's 303 add sulfur, 303 Se is a designation for selenium. You don't see very often. Sulfur will embrittle the steel and makes it non-weldable. Okay. Anybody know what a Sub-Zero freezer is, or refrigerator? Okay. My son bought a house and had one. It's a twelve thousand dollar refrigerator, okay. Uh, it's top-of-the-line refrigerator. Uh, so I had a problem with Sub-Zero. They had a little bracket on the hinges for the door, and they were supposed to have — it was supposed to be carbon steel to carbon steel, and then they did a little TIG weld for this pin, and the pin was part of the hinge for the door. And someone made the pins for them and they welded them but they didn't notice that they were getting little cracks. Because someone had machined the pins out of a free machining steel which had extra sulfur in it. So they had these little cracks. They sent the refrigerators out there and they had to go out and repair like 3000 of them in the field.

Has anyone ever seen the statistics on — if you catch something at the design stage it'll cost you a dollar to fix it, if you catch it at the fabrication stage it might cost you ten dollars to fix it, if you catch it out in the field it might cost you a hundred dollars or a thousand dollars. Well, this was sort of the thousand dollar problem. In fact I think it was a seven million dollar thing to go out and repair these things. And the people who buy twelve thousand dollar refrigerators are people who can afford to sue you, okay. And so, I mean — and you know this is a high-end, uh, they make Wolf stoves, so if you want a six thousand dollar stove you can buy a Wolf stove from them.

Anyway, so it makes it non-weldable because the sulfur forms an iron sulfide. Remember I told you it had to have the right manganese to sulfur ratio to tie up the iron sulfides or otherwise you end up with a slurpee, a liquid phase iron sulfide at the interface, and that cracks when the material is shrinking. So it's there for machinability, it also destroys weldability.

There's something where you increase the chromium and lower the nickel and create a duplex stainless steel. Duplex means it's — this should be 100% austenitic, this is basically a mixture of austenitic and ferritic. And this in the regular state has got 40 ksi yield strength. This can have 80 or 100 ksi yield strength because you got both FCC and BCC crystals, and it gives you higher strength. Problem is — this is basically — has no problem with hydrogen cracking in general. This is just like any other ferritic steel and will crack with hydrogen, and you have to have very careful welding procedures.

There are precipitation hardened stainlesses where you add copper, titanium and aluminum, and you can get precipitates with nickel titanium, nickel aluminum, or just straight copper type precipitates in the iron. And you can get 180 ksi. This is what the Navy used to use, the PH stainlesses, on their hydrofoils. They needed — I think it was the rudders. It was either the rudders or the propellers. But they wanted high strength, they wanted lightweight, and they were going to PH stainlesses. The most of the hull was aluminum, but back in the drive end they were using some PH stainless. And we created a few problems back there, but nonetheless.

Here, you can add lots of manganese, or some nickel or nitrogen to lower the nickel, and get higher strength. These are not used very often, okay, this is kind of older steels. You can take out the nickel to save money and lower the chromium, and you end up with martensitic stainless steel. I passed around these scissors before, surgical stainless, okay. It's basically a 410 or 420 stainless. A 409, which I don't know if he has 409 on here, it's basically muffler material, which is basically this material, you know, down here. Martensitic — if you try to weld it, means it gets hard and brittle, okay. The reason they use it for surgical tools is they actually harden them. I mean you don't want them to wear and things like that. So they actually harden it, and it's basically surgical stainless.

You can add some molybdenum. Two percent molybdenum makes 316, four percent makes 317. Is often used for — in the body where you have high chloride concentrations. It's not very common. 316 is fairly common — that's what they were supposed to use in my hot dog cooker, okay. They didn't, they used 304, and it just corroded, okay. So this is for pitting resistance mostly in chlorides.

If you go to a low carbon variety, we're going to talk about that in a little bit. This improves the weldability and the corrosion resistance, okay. This one we'll talk about in a little bit. When you go even higher than in molybdenum and nickel and add some nitrogen for corrosion resistance, you get to the super austenitic stainless steels. 10 or 15 years ago the U.S. Navy was thinking about getting rid of HY100 and building a super austenitic stainless steel submarine, because it's non-magnetic, okay. It doesn't have magnetic signature that someone can find in the ocean. Um, I think they've gone away from that because they realize those terrorists probably are not going to be using superconductors in the sky, okay. And it's also sort of pricey, okay.

Uh, 321 you can add some titanium, does sort of the same thing to tie up the carbon as these low carbon varieties here. 347 you add niobium or tantalum, does the same thing as titanium. 430s are the ferritic stainless steels, have very good corrosion resistance, and you can get to very highly alloyed with like 30% chrome, uh, and you get to the super ferritic stainless steels.

If you go up this way, here's your — more corrosion resistant, more temperature resistant, higher chrome and nickel. Not up to 30 percent, but well actually 330 is about 30. And you keep on, you go way up, and replace the iron with the nickel. So this is 70 percent iron. This one is, uh, and 20 percent uh, or 10 nickel. This one is 60 percent nickel and twenty percent iron and twenty percent chrome. These are your Inconels, okay. So far as that goes. And you use those for better corrosion resistance and higher temperature corrosion resistance. And if you're in a pressurized water reactor like Westinghouse makes — they made their whole pressurized nuclear reactors out of Inconels, okay. Very pricey, very good corrosion resistance. But a number — and they all have welding problems so far, which is why I have a job. Um, okay.

If you look at the iron-chrome-nickel phase diagram, this is part of a ternary phase diagram. It's plotted not in a triangular form, which is what a pure metallurgist would do, but in a rectangular form. If you look at the nickel and chromium content, so 18-10 is right here in the austenite, okay. You can, you know, it costs money to put chrome and nickel in. And you can get right here on the edge, or right up here 18-8, it's right in here. And that's your 304 stainless steel, it's austenitic, but you can see it's got the minimum alloy content, uh, to make it austenitic. Down here when you have — you're down here in ferrite at room temperature, and this is your regular carbon steels. You get up here with low nickel and you get your martensitic stainless steels.

Um, sorbite, huh? Anyway, I know what troostite is and I know what sorbite is, I don't know if I know what the future is. Anyway. Um, you can get our austenite-martensite — this is what happened when the — we made the Egyptian tanks, and they used an austenitic filler metal with the armor steel, and they ended up in this region and they got cracking because of the martensite. Some of these surgical stainlesses are over here, anyway. You can get in these regions, you can get the duplex stainless steels, but you have to be careful about your welding procedure, so far as that goes.

Now it turns out back 30 years ago when General Electric was making their nuclear reactors out of stainless steel, 304 stainless basically, they found that sometimes they would get cracking. And it wasn't hydrogen cracking, it was — eventually that they tied to the little bit of sulfur in the steel. And it would also cause pitting corrosion. You'd have a little sulfur inclusion and then you go in there and you study where your pits are, and they're right there at the sulfur inclusion in the steel. But they found that on solidification a lot of times the steels would crack, and actually probably have a — may have a picture here of the type of test that they developed in the 1970s at Rensselaer Polytechnic. At the time RPI was sort of the school in the country for producing PhDs in welding.

And they developed something called the transverse strain test [Varestraint test]. So you would basically have a TIG torch, you'd have the material you want to study the weldability of, you'd have a radius block, and you start making your TIG weld across here. And as you did it you would load this thing and bend the thing over a certain radius block, while it's hot. And then when you're all done you'd look at it and count the cracks. Here there's pictures of the cracks, okay. Here's the instantaneous solid-liquid interface when they applied the pressure. They would — this was the time they started deforming it as it was being welded — and they would just open up little cracks. And if they went in there and studied that they'd find that the sulfur would go to the austenite, which likes hydrogen but doesn't like sulfur. Uh, would end up — it would crack at the austenite boundaries. This metal could be all austenite, or it could be a little bit of austenite-ferrite, was good they found. And they'd like to have a little bit of ferrite. Uh, they found about three percent ferrite was good because it essentially absorbed most of the bad sulfur. These steels weren't high in sulfur, they just would crack, particularly in the highly restrained joints and whatnot.

So originally we had in 1949 — or actually 48 — we had this thing called the Schaeffler diagram. And he basically did the same type of plot, which is sort of the phase diagram I showed you before, where you have ferrite down here, you have all ferrite, you have ferrite and martensite, and then up here you have austenite. This is that little corner of 18-8 or 18-12 stainless steel. And he could plot on here how much was austenite or how much is ferrite. So this is 100% ferrite, go all the way up here, to zero to five percent — we found it's nice to have your weld metal right in here. If you're welding something down here, so we use 308 filler metal, which is slightly adjusted in its composition so it will give you like three percent ferrite in the weld metal, and you wouldn't get cracking because of the sulfur in these highly restrained weldments, okay.

So then the guy DeLong came along. He was working for — I can't remember — it was a Bill DeLong. In 1973 he came along and he came up with another plot, and he introduced the term ferrite number, okay, to the welding community. Here's the Schaeffler line, here's the 18-10 kind of corner, austenite. And you could look in here and you could see if you wanted to be two or three percent or six percent ferrite in your weld metal to avoid cracking. This could allow you to calculate what you should have, how much dilution, what type of welding electrode. So we tend to use a 308 or 309 welding electrodes for different thicknesses, if you've got heavy section stainless and you don't want it to crack, okay.

The problem is, with the super austenitics, you can have zero ferrite, because that was what the Navy wanted. They wanted super austenitic, it had zero ferrite so it wouldn't be magnetic. I showed you what would happen with that, my little Fiberware pot, you know. It was like five or ten percent ferrite down at the bottom. It had been deformed some and transformed some, and you get up to the top and it was probably close to, uh, eighty or ninety percent — actually in that case martensite, but magnetic steel, so far as that goes.

Okay, so that's general things on stainless steel. Anybody have any questions on that? Hopefully we'll finish up some of the stainless steel today. There are the ferritic stainless steels, and the ferritic stainless steels basically — this is the CCT curve for your, for 410 stainless, which is like 12 percent chrome. And remember I was telling you on carbon steels you had to get past this nose within like one second to avoid — to be able to get martensite, and avoid getting the austenite transformed to ferrite. If you wanted to get martensite you had to quench it very quickly. Well, you go to a 410 stainless and now instead of the nose being all over here on the log scale at one second, the nose here is now at one hour. So it doesn't matter what you do with one of these types of steels — when you weld it you're going to get martensite. Unless you weld it and then put it in a furnace and slow cool it over about a day or a week, okay. This stuff is really hardenable. Very high hardenability with 12 chromium. It has extreme hardenability, which means it's very difficult to weld. Okay, let's not say we don't weld it, we do weld it, but we have to be careful about how we weld it. And it's susceptible to hydrogen cracking, okay.

So I guess I can now tell you, um, a little bit of the story. I guess I've told you enough. Um, so last March — it's middle of March — someone calls me up and he says, uh — a contractor here, a construction firm here in Boston, was building the Miami Art Museum. And it was supposed to open last December, which it did. But it was about 50 yards from Biscayne Bay. And it was a, uh, it's like a 70 million dollar building. And they were gonna have a grand opening around, you know, the middle of the winter when all the people from New York, all the donors, are going to be down there. They want to have a grand opening. And they had a Swiss architect and he built — he designed this thing —