SPEAKER 1: Our next speaker is Bob Silman, founder and president of one of the truly great structural engineering firms in the country. Silman Associates was founded in 1966, and his firm has worked on some of the most sensitive architectural restoration projects and structural projects imaginable, including famously Falling Waters and the Darwin Martin House in Buffalo.
Bob has earned universal respect among his peers and numerous awards, too many to recite, for his contributions to engineering and engineering preservation. Bob is also a firm and distinguished member of the AAP advisory council and has been a fantastic friend and advisor to me. Please help me welcome Robert Silman.
ROBERT SILMAN: What a joy to be here, and thank you. You know, we generally never get to see finished buildings. We're here only during the muddy boots phase, and then everything happens, and the building is commissioned, and we're on to the next job already. So it's very nice to be here and celebrate an actual finished product.
Thanks so much for everybody to allowing us to participate. And it was really a tremendous project, certainly the whole OMA team. Our own team I want to recognize. Also Alastair Elliott who is here today was our associate in charge and project manager. Laura Smith, engineer who did all the steel work. Ben Rosenberg, who did all this concrete work. I am only here as a spokesperson. They did the work.
Alastair today carried up with him this month's magazine in Civil Engineering, which is the publication of the American Society of Civil Engineers. What's the cover girl? Milstein Hall. It's terrific.
We're fresh up for that. Ted had asked me to sort of set the participation of our firm with Milstein Hall into our general practice. That's a hard thing to do, because we've done so many different types of buildings. It's a firm now that's 46 years old, 17,000 projects. How do you pick out one over another? It's like asking who is your favorite child. So I thought what I would do is just in the context of what we've done at Cornell, how does this building sort of fit in?
We've done on this campus, I guess, almost 10 projects. Our first one was biotechnology. This is a concrete building, and it was sort of extraordinary for Cornell to do a concrete building back about 20-something years ago. And it was great fun. It had a building where the structure imposed the discipline, because it was a waffle slab. So every two feet was a rib, and that was a module that was established. Also being a lab building, there were things like vibrations that were very important to control. And a really interesting job for us.
The Catherwood Library and the ILR School up on Tower Road. Again, a great departure for ILR, which had been housed for so many years in temporary buildings. When I was an undergraduate here, they were in these wooden shacks. And it was a great transition for them to be able to be in a real building for once. So that was great fun.
One of our notable projects was Sage Hall, now the Johnson School of Management. This was a huge undertaking in which there was a nostalgia to keep the look of the outside of the building, it being one of the earliest buildings on campus, I think, either the second or third oldest from the 1880s. But the inside of the building was totally satisfactory. It didn't meet building code requirements either for fire resistance rating or for handicapped access.
It was completely destroyed, the inside of the building. Totally gutted. We just shored up the outside walls and put a whole new school inside of it. That's where I first met John [? McCowan, ?] doing this project. And of course, it came out to really exceed everybody's expectations with this great central atrium that was once an open space, and it is now a terrific circulation space for the whole school. And that, in terms of context of what we had done really was a great achievement for this Cornell campus, particularly with the weather environments to have a glass roof.
The North Campus Residential Initiative, which it was called then. It's now called to Court and News Buildings. To take the vocabulary of how you build a college dorm and make something much more friendly out of it instead of having long, narrow corridors, the ins and outs in these corridors I think make this dorm a kind of unique one and a most wonderful one. The White Hall just to the right of this building as you're looking out from Sibley, when I was in architecture school here, that was the headquarters of the Architecture Department, White Hall.
It was, again, one of the early buildings. It had bearing walls that rigidly divided the building, and you could not pass from one sector of the building to the next, because it used to be half dormitory and half a classroom, and they wanted that division to exist. So you had to go outside and back in another door. We changed all of that, and you can see on the left side the picture of the corridors that now go all the way through the building, yet the outside looks exactly the same. This, too, didn't meet code. It was a wooden-floor building, and we replaced them all with concrete floors.
If you've not been there, get up to Sapsucker Woods in the Ornithology building. Again, very contextual. A lot of wood and laminated wood in this building. Looking out over the pond on the left side, you can see some telescopes there. You get yourself out-- almost from the inside, you can be outside the building.
Almost our favorite project, because we work with the students here who participate in something called the Solar Decathlon-- I don't know if Jerry Wells is here, but he was the faculty advisor. And this is, I guess, two or three years ago. It's on the Mall in Washington, D.C. Where student competitions erect these houses for a couple of weeks, and then the houses are demounted and moved and sold to somebody. Cornell came up with this marvelous idea of silo-like structures, and some of our younger engineers who were Cornell engineering graduates, as a matter of fact, volunteered their time and worked on this project. So it's not on campus, but it's of campus.
And then a project yet to be realized, but now back in action, the new Humanities building on East Avenue, the back of Goldwin Smith Hall, is going to have this giant glass enclosure, and inside of it two towers rising up that will have classrooms and offices. And in terms of technology, this will be one of the most advanced buildings on the campus. Except for this one.
And then we're at Milstein. And I would say of all of the projects we've worked at Cornell on, this one is the one that really is in a sense the most thrilling. I'm going to talk just about two aspects of the building, of making the building. One is the steel trusses, which are so prominent. And the other is what you're sitting on, the quote, dome, that really isn't a dome.
So the structural features, of course, from the outside, you're enticed by this notion of these trusses. You can see them through the glass. You know the old implication that an architect would design a building, make a set of drawings, give it to an engineer, and say build it, that's gone. Nobody really has that kind of mindset anymore.
And in this building, more than any other, of course, we had to integrate the structure and the architecture right from the beginning. It referred to the University Avenue problem. And of course, the original idea was to put a row of columns quite close to the Foundry building, span across University Avenue. 60-foot span, it's not so terribly unusual carrying a two-story building. That could have been done and done quite economically.
And the town was so insistent that those columns were dangerous, given, of course, the Ithaca climate and people skidding on ice and snow. They were worried about damage to be sustained by a column. My idea was move the road a little bit. We're not going to move the Foundry building, but you can realign the road. You know, big deal. And that could've fit under the building. We could have made it safe, we could've put bollards all around it.
They were intractable, of course, and what was the result was the cantilever. Of course now it becomes the vocabulary of the building, and everybody loves it, what, $2 or $3 million later? It sounds very romantic. And so that imposed upon us by the town was a requirement that we span with this great 50-foot cantilever.
In doing that, the first impulse was to say, well, just put a big giant beam underneath everything and hold it up. The problem with that is, of course, that we're trying to meet the four levels of Sibley, and it's an older building. It doesn't have a very generous floor-to-floor height. And by putting a very deep beam underneath, we would not be able to have a meaningful, usable space under there. So that didn't seem to work, and the notion of a truss was developed.
Just going back a little bit, before we get to the trusses, understanding how we got that far even. You have to dig a hole. So here's the site, and looking at the Foundry building, the hole was dug. And you can see just underneath the fence there is a series of pipes, and these are-- I don't know if I can get this going-- a series of pipes here that are indeed part of the foundation system. They're going to be used for what we call the drilled piers or the caissons.
In Ithaca, the typical foundation is drilled piers. And it's not something that's common around the world, but it is used a lot in upstate New York, and here you see a rig with a circular drill. There's a churn drill in there. It's hug-a-tree kind of a thing. And there are piers that go down and bear on rock. It's an economical system. Around here, you can see some of them sticking up out of the ground here, and that's what holds the building up.
But the most important thing is the element that we see, the trusses. When we decided that we would use trusses-- oh, and they did a study of the code requirements of the building, and found indeed that we would not need to fireproof the steel, fire protect the steel. This was a great revelation, because it became the structural aesthetic and vocabulary in the building to have exposed steel. In this room, you see a couple of columns, but of course, up on the design studio floors, you see all of the steel exposed.
And in doing that, we had to decide on an aesthetic that had a certain amount of rigor and uniformity to it and allowed integration of the mechanical systems into the structural system, because you could see everything. There was no suspended or hung ceiling that could conceal all of that. The lights, the pipes, the ducts, the structure all had to be integrated, and of course, done well before we fabricated the steel.
This is a picture in the shop of a steel truss detail. The top cords of these steel trusses are all the same depth. Yet the stresses in there vary considerably. So at points where we have very high stress, you couldn't pick out of the steel catalog a beam that was strong enough or big enough. And we had to actually fabricate these out of plates. So they're little plate girders. Everything is, I think, is 24 inches deep uniformly throughout the ceiling, in addition to which we had to cut holes for the ducts in there as well.
When it came to making these, we didn't want them to appear very differently from rolled beams, and a rolled beam has a fillet in it, a curved shape where two plates meet at right angles. And the wells are in fact shaped like that. We're able to do this because of our new technology of automatic welding. It's done now more by machines in the shop than it is by human beings. You're used to seeing Rosie the Riveter or a welder with a stick and a mask, and they're running a bead of weld.
You still do that, of course, but in the shop, that bead of weld is done automatically on a machine. The feed is timed to go through, and they build up the amount of weld material. These plates are big enough, by the way, that the plates had to be preheated, because otherwise the heat dissipates too fast, and the weld will crack and not be sufficiently strong. So there was a lot of technology involved. The steel shop was in Canada.
And here's a picture of a joint, and you can see quite the elegance of the welds. I'm fascinated that Lord Palumbo is here today, because my dictum to the shop was, I want these welds to look like the welds of Farnsworth House. And those of you who have been there probably understand what I mean. You go to Farnsworth House, which is an all welded steel frame, and you say, where are the welds? They are so beautiful that they disappear. Here we couldn't make them disappear, because the forces are much greater. So you will see them, but they are beautifully crafted, I think, and I hope that they contribute to the structure.
Here's Laura Smith. Where is Laura? She's here somewhere. Standing on the truss. And this is in the steel shop outside Quebec, and you can see it was fabricated lying down, which is, of course, much easier to work on. They have all the machinery to flip it over on the other side when they need to do that. And it's really quite an elegant operation.
The erection of the steel-- the trusses came prefabricated, but only in pieces. They were too long to ship in one piece, because although the cantilever is 50 feet, the truss is several hundred feet long by the time you get end to end of it. So you couldn't put that on a truck. And it was designed, and you can see on the left side here where plates stick out that are not painted white, these are where the field welds are going to take place. We do not like to weld through paint, because it contaminates the welds. So we leave them unpainted and grind the welds smooth in the field, and then paint it in the field to match the rest of it.
You will notice here all the holes in the steel work. This is very carefully designed, coordinated with a mechanical engineer so that the ducts and the pipes are able to be located before the steel is fabricated, so they can make these holes in the shop, and we're not cutting them in the field, and scurrying around and having to calculate whether, in fact, the stresses will be exceeded or not.
This is the same slide you showed, Joe. This is the steel being erected on two of the columns. They'd land the truss on there. The crane holds it until it's stabilized with some crossbeams, and then the connections are made, and they are all field welded. Again, that takes a great deal of craftsmanship to do and do well.
Now the shape of the trusses is-- here's the field splice. And in fact, this was a case where they did weld through the paint, and you can do that by dissipating the paint at first and then welding. You also see a bunch of bolts in here so that the web connections for shear are bolted and the moment connections and flanges are welded.
Here's the one place where two trusses meet. It's at the northeast corner, this being the north/south truss, the other one being the east-west truss along University Avenue. And that, of course, is the very interesting joint where the two trusses meet at this point.
The drawings indicate that we have trusses of various shapes. And this is what the great fun in this building was. Here's a typical truss. When we came to design these, it wasn't just a question of arbitrarily saying, well, we're going to build this kind of a truss. Most of you who have studied anything about trusses understand that to be true truss, we want the forces to be axial, that is along the member, not perpendicular to the member, the stresses.
So it's much more efficient to carry stresses axially than it is normally in bending. Trusses where the diagonals meet at a panel point and considered to be pinned at that panel point, that is they have no rigidity, no strength against bending, are called normal trusses. They often have people's names attached to them, like Howe trusses, Pratt trusses, the town truss, there are various names.
Then there are the trusses that are like the center of this, which are only vertical and horizontal, are much more highly stressed at the joints. At this kind of a joint, we have a tremendous amount of bending moment which has to be resisted to make a rigid joint. This truss was invented by a man named Arthur Vierendeel, a great Belgian engineer more than 100 years ago in the very late 1800s. He proposed it, and they built the first one, I think, in 1902. So all of these stresses are named for people.
And here's a hybrid, and maybe my name will be on it. Maybe they'll name it a Silman truss. Actually, I hope not. This wasn't as arbitrarily decided upon as it may seem. We worked very closely with OMA after the function of the building had been decided upon, because there were clearly areas where they needed to have horizontal pedestrian circulation. People needed to be able to walk through virtually unimpeded. And those are the areas that you can see where the Vierendeel truss predominates where there are verticals.
It's also the area where the stresses are mostly the lowest, except let's say right here and right here. When you have a big cantilever like this, the maximum stress is right at the column, on either side of the column. So here we have a diagonal with a pretty steep rake. Here on this end, not quite as steep as maybe we would have liked. And that's quite efficient. On the other side, we couldn't do that, because they needed to have circulation. This may not be as efficient, but we know how to do it. We can analyze it.
These things are all analyzed now on a computer. And because of this hybrid notion, the computer doesn't care. It will work through the results. And that was beautiful, because we would develop a scheme, OMA would say, eh, that doesn't really work with our circulation. You could move the angle of a diagonal slightly, push the button, and get an answer very quickly. I hope we were quick enough, We could play around with this.
In the old days doing it manually, it wasn't so easy. You had to go through all the calculations every single time. And it was very sensitive. As you move these things very slightly, it is amazing how the results change. We were working against a budget, and that makes things always much tighter. So thanks to computers and thanks to the collaboration effort, we were able, I think, to satisfy the needs and leave this as a beautiful exposed steel structure, in certain cases with steel diagonals where there isn't any need to interact. And of course, you have the little fence at the base so you don't bang your head, hopefully.
And then other areas where we had to have pure Vierendeel kind of truss. This is really a hybrid. And it's very exciting, and I think from the outside, it makes you want to come in and say, what's going on in there? Why is it like that? So certainly from the quad side, and you see that, it looks more like a conventional truss, and then all of a sudden, as you get back into this area, it's not anymore. And that's, to me, a very exciting thing.
We've talked about mostly stresses in the trusses, but that isn't the critical thing in a cantilever necessarily. People want to know, how much is it going to move? We call that deflection. And who wants to know that? Well, the people who put the curtain wall up, particularly, but also the users. We want to promise you that you're going to wind up with a flat floor, so that when the kids are tired from designing, they can play marbles.
It's not so easy to achieve that. Because don't forget, when you're pouring concrete on the end of this truss, the truss is going to go down. It's going to deflect. Then you have to pour more concrete to make it level, but pouring more concrete makes it go down more. So we try to calculate these things. We know it's going to go down under the weight of itself and the concrete, so we build it with a slight camber up hoping that it's going to come back level. Doesn't always.
We actually did a calculation for five different loading conditions. It's very hard to see these, but A, B, C, D, E are the deflections under various combinations of loads. The dead load alone, that is the weight of the building. The dead plus the live, that is the people in the building. When you only have the concrete base slab, when you have the base slab plus the topping slab, all these combinations. And for every point, we have the five conditions, A, B, C, D, E.
The tip of the middle truss at University Avenue goes down about three inches. That's fine. We know that. I mean, we prepare for that. The guys who did the curtain wall design front and the contractors for the curtain wall are absolutely insistent that they know how much movement is going to be out there, because the width of their joints is dependent on how much the truss is going to move. And you know damn well they want the skinniest joints possible.
Anybody can make a fat rubber joint in there and take up a lot of movement. That's not the point. How can we make it as delicate and narrow as possible? So they're very insistent on that, and I have to say without the use of computers, we never could have done this kind of accurate calculation. So in a sense, God bless the computer.
This picture is the floor before the concrete just poured. You can see steel studs here and that the beams underneath are made composite with the concrete slab, making them much more efficient. Here's the slab with the topping already poured, and it's being protected and cured with this polyethylene on it.
You may see these big floor boxes out there and wonder what they are. But inside of them are the tubing and the valves for the radiant heat. A lot of this building is, of course, out over the exposed weather, and it's Ithaca, and it's very nice to have a warm floor . So this is a radiant heated slab. There's a lot of boxes out there, and look at the number of tubes in each. Amazing.
The roof, of course, with many skylights on it. A green roof. This is before the topping was poured. Here's with the soil on it and the north-facing skylights. The Sibley dome behind. Again, as [? Sheryl ?] said, too bad we couldn't have that as an occupied roof. It's designed that it could be someday, and maybe a great donor will happen.
Talking about the dome that you're sitting on. This is interesting, because it was a post-poured dome. In other words, we built the building first. And then with the luxury of an enclosed space, and that's an important luxury in Ithaca, we were able to construct the dome. So we left reinforcing bars sticking out getting ready for the dome. Kent, you're sitting right here now, OK? No, maybe not, actually. That's the end of the flat floor. That's in the back of this.
And all these rebar with the orange ends to them so you don't poke your eye out are there. The great diagonal columns was poured, but nothing else here. This is the diagonal column. Up at the top, the same thing happened. You have the rebar sticking out waiting for the dome to happen.
The formwork for the dome is as complicated as the concrete pour itself. And of course, this is the negative part of the dome, which gets taken away, but it's every bit as important as the positive part. And formwork carpentry is a real art, because it has to be perfect and yet removable.
The top surface with plywood, and you can see the joints out there. But they are not exaggerated, and it really, I think, was extremely successful. The reinforcing of the dome was prodigious, as you can see. There's a lot of reinforcing bars there. This is calculated by the computer. We use the word dome, it isn't really a dome. In the old days before computers, we were only able to analyze shell structure that were mathematically describable. Now a dome is one. It's a surface of revolution. You have an axis and you have an arc, and you rotate it around.
So mathematically, we could write an equation for that. You could form matrices and you could solve these things mathematically. In these freeform structures, even if it was a dome, with so much of it cut away at various points, it's almost impossible to analyze by hand. But we have this finite element method of analyzing things now on a computer. All we need is the geometry. Now that's not so easy, either, accept that God bless OMA, they came up with some beautiful drawings.
And now in CAD, this is Computer-Aided Drafting, we are able to export their file out of CAD, import it into our structural analysis program, and all the geometry is then done. The computer picks it out from the picture. So the hardest part is done. Now all we have to do is put a load on it and tell it what the material is, what the allowable strength is, and it analyzes this thing. It's magnificent. It's fabulous. So we're able to do things that we were never able to do before.
This is what you're sitting on. The quote, dome surface underneath is built up with these egg crates. They were filled with Styrofoam, and then the step slab was poured on top of that. Here we are looking at the underside of the finished slab. The slots, of course, are eventually the light fixtures. So there's a lot of stuff going on here. The surface, I think, came out absolutely beautifully. The eyebrow from University Avenue, a piece cut away, really mucks up an analysis, because it has this very, very, very heavy edge, which changes what we call the boundary conditions and introduces bending into this shell structure that otherwise shouldn't be there, that we would hope not be there.
And then the columns go through it, and the load is picked up there. It's very complex. Here's a computerized screenshot. We divided it up into plates that were about-- I think they were two feet square, roughly. Making them translucent, you can just see what's underneath it, because the bridge and the stair was modeled with it, because it's all part of the structure, and it's all together there. And the green of the support elements all around it.
So it's an extremely complex structure, and Ben did a great job in our office, and I think it came out just beautifully. Similarly, you can model all of this stuff. The bridge, stairs, and all of that. In order to know what it would look like, we built a mock up of this. Here's the formwork. Here's a slab being lifted, looked at from the underneath side. Is this satisfactory or not? And you mess around with it until it is satisfactory.
Again, this is an unusual concrete truss that holds up the stair and the bridge above. And the flying bridge that goes through. Celebrate Milstein. We had a great time with it. I hope you enjoy it as much as we did. And thank you, all.
We've received your request
You will be notified by email when the transcript and captions are available. The process may take up to 5 business days. Please contact firstname.lastname@example.org if you have any questions about this request.
From lectures by Rem Koolhaas, John Reps (M.R.P. '47), and William Forsythe to an exhibition of work by Simon Ungers (B.Arch. '80) to a party unlike any the college has thrown before, Celebrate Milstein Hall energized the AAP community as 500 alumni and guests reconnected with 300 faculty, students, and staff for an exhilarating weekend.