GL

I think you were involved in helping with the Barcelona Fish…

RS

Yes.

GL

You were also very involved with the Lewis House. I’m interested in the milieu you were working in and what brought you to that project.

RS

Back in 1979 I worked for Lockheed Martin. Before then, at school, I’d learned three-dimensional modelling through Applicon and Computervision. At Lockheed I worked on a program called CADAM, which was their home-grown system. Lockheed sold it to Dassault Systèmes in France, and then Dassault decided to write a three-dimensional program on top of that. Around 1982 I joined IBM because CATIA had come out publicly and I’d heard that IBM was going to start marketing the program to a number of different industries.

But well before all that I actually went to school, California Polytechnic State University in San Luis Obispo, to be an architect. This was in the early 1970s. And my dad, who worked for IBM, showed me how they were developing techniques to draw on computers. That was back in, oh—about 1968, I think, when my dad showed that to me. At Cal Poly everything was pencils, paper, and T-squares. I knew that the computer technology was developing, so that’s when I transferred to Brigham Young University where they were actually teaching it.

I thought: This is the future. And if they’re not teaching it in school, I’ll go another way; even though I’d always wanted to be an architect. So I went through ten years of IBM, worked in the aerospace industry, worked on the Space Shuttle and got involved in medical devices.

When the Berlin Wall came down, you know, Reagan and that whole political environment, the aerospace industry just died. IBM started laying everybody off. It was two days later that I got a call from Frank Gehry to my office. Well, it wasn’t Frank himself, but one of his staff, who called and said, “Do you know how to build a building in the shape of a fish?” And I told him, “A fish is kind of aerodynamic… Sure.” So that’s when I got involved with Frank. It was about July 1991.

Read more

GL

Do you know how they got to you? Was it through CATIA?

RS

They had been talking to Bill Mitchell, then dean at MIT. They were playing around with Alias, trying to do the Fish. Alias could create pretty pictures, but it couldn’t get any engineering or manufacturing information out of the model. Someone told them, “Well, you ought to try IBM and CATIA, because they build airplanes with it.” I was in the aerospace division in Los Angeles and somehow they got the number to my desk.

GL

That was for the Barcelona Fish.

RS

The Fish was first, but then we jumped right into the Walt Disney Concert Hall. We had done a lot of work on Disney, and we were doing it simultaneously with the Lewis House, so the designs fed off each other.

GL

What was it like working with Frank on those projects?

RS

When I got there I had no idea how they worked. Back then, in the aerospace or engineering industries, we would design right on the computer. But Frank did not like looking at the computer screen. He was completely invested in models and had a full model shop.

Frank would work out his paper designs and then make a model. I would follow, digitizing his models to build a computer model. Then he would look at the computer model and the physical model he’d built to check for differences and to see whether we actually captured his shapes.

GL

And so you also brought the digitizing arms and the scanning with your 3D modelling.

RS

Yes.

GL

And in terms of people in the office back then, was Craig Webb or anybody around? Is anyone left?

RS

Edwin was there and Craig was there at the time.

GL

Were things different between them and Frank? Would they jump on a digital modelling package or did they also work with paper and pencil?

RS

Well, it was interesting because when I first got there… I’ve got some funny stories. They had the Disney model on a table, you know, about four feet by four feet. It was my first day. I was walking around and saw they had a big piece of opaque glass, kind of a bathroom-style frosted glass. They would shine a light on the opposite side of the model to cast a shadow on the glass, and then they would tape paper along the outline and start sketching that shadow.

That was how they were going to get the elevation drawing! And I said, “Do you think you’re going to build a building this complicated with that method?” And the guy got really upset and said, “This is how Michelangelo built his buildings.”

GL

Ha!

RS

They had a plumb bob they would drop at different points to get the measurements off the model. When I came along, there was a lot of reluctance. In fact, most of them… I mean, they had graduated from highly prestigious schools, paid a lot of tuition, and were trained in those methods. And I came along with this computer saying, “You know, those methods are archaic.” It created some real frustration in the office.

And Frank also not wanting to look at the computer—he was not happy with it. But it was kind of forced onto him for Disney. Actually, the Music Center had basically hired me, then told Frank, “You need to use computer technology.” Frank was really reluctant. Because of that, all the people in the office who had graduated and wanted to become like Frank, were also reluctant to show interest. But then there were people like Craig—he and I worked really closely together. Craig, I felt, broke the mould in the office. He broke things open.

The Lewis House had a big impact on the office. Toward the end of the project Frank did not want to look at the computer, but Philip Johnson, who worked with Frank on this project, came and wanted to work with me directly. So we sat down and the two of us worked on this idea of colliding together a pear and a cube and then moving them around and looking at different juxtapositions of form. Philip got so darned excited about the computer. He was sitting there, just giddy, going, “Hey Frank, isn’t this really great?” Frank got upset and said, “No, I hate looking at computers,” and walked away. All these young architects were standing around watching this. And I think they got the idea that, hey, it’s okay to look at the computer now.

GL

I was a bit of a fly on the wall for some of this period. I was acquainted with Frank—I didn’t know him well. But in the Lewis House it looked like there was much more of a back and forth then: build a model, digitize it, and be sure the computer model is faithful to the physical model.

RS

The Lewis House was a big breakthrough in many, many ways. Before that, the Disney Concert Hall was geometrically defined surfaces and everything, whereas the Lewis House was just this blanket. We had to digitize that, and it was very hard to make a model. We were using 3D printing techniques and they were expensive back then. It wasn’t like you could chase a lot of iterations with 3D printing, so Frank had to look at the screen. He kind of had to accept the computer, and he did it mostly through Craig. This was where Craig got a lot of his experience.

GL

Yeah. I took a look at some of the tubes of drawings. I was working for Peter Eisenman at the time, and it was the same thing. You’d get centreline of geometry off of the digital file, then pinbar it and everybody would start drawing plans and sections over it—as if that was going to be a help.

But within a few years, for the Lewis House, you start to see CAD drawings. The CAD drawings are still drafted over centreline geometry, but it’s shifting. Not quite a BIM file, but at least the drafting was happening in CAD too. Were you involved in any of that? I could imagine you saying, “Look, we should build a high-fidelity 3D model of the whole thing and generate plans from that.”

RS

Well, we did.

GL

You did?

RS

I still have a set of drawings—through CATIA we could do section cuts at any angle or whatever, so we produced the drawings. We’d do all the backgrounds in CATIA and then hand it off to AutoCAD, and they would try to dimension it. I mean, how do you dimension a crazy three-dimensional shape? But the direction we were going, until it got shut down, was what I used to do in the aerospace industry. Boeing had come up with this concept of digital mock-ups, for example, that used a fully 3D model, from which you’d fabricate.

When you design a wing in the aerospace industry, you have your struts and your ribs. You take section cuts in the shape and then a cut becomes a cross section; in a sense, that’s like a boat hull. So how do you make a boat hull? You make the spars, and everything else.

Jim Glymph founded a company that made boat hulls and they agreed to build it for us. We sent them the two-dimensional cross sections, they would create the ribs and then we’d clad them—it was going to be copper-clad. If we had followed it through, that would probably have been one of the first buildings out of Frank’s office to go directly to a fabrication process instead of to drawings.

GL

That’s now Frank’s mantra. I think for Beekman Tower and other recent projects they’ve been trying to get rid of the 2D step as much as possible.

RS

That’s the right way to go, really, from a 3D model. A 3D model is the 2D documentation.

GL

Now, because you have a lot of other connections and experience, what was happening with the Autodesks of the world, and CATIA, and Alias and Wavefront back then? Were any of those companies talking to you and saying “Now, what are you doing here—taking an aerospace approach to architecture…?” Was there curiosity, or connection, or discussion, or anything?

RS

No! (Laughs.) Well, for one, Frank told me not to say anything to anybody about his approach because the office felt it was a proprietary process we were developing… They wanted to release things. Frank’s very business savvy, and wanted to be leading-edge.

They did hire Kristin Ragins, who came with AutoCAD experience to set up 2D drafting, and that was probably the only outside connection to Autodesk—just Kristin and her AutoCAD experience—and I don’t think we went anywhere near to talking Autodesk with that.

The idea was to keep what we were doing (and how) quiet, within the office, for a while—in the Lewis House days.

GL

The majority of the architectural field let the structural engineers adopt CAD first, especially in 3D, and so they got a lot of control. That’s the only explanation for why ARUP and other engineering firms are so big now. Because they had all the control—they had the 3D.

RS

Yes, I agree.

GL

What role did engineers have in the Barcelona Fish, Disney and the Lewis House? Were they very involved in the design, or were they more or less just providing the structural services?

RS

Well for all of Frank’s stuff we actually did the structural schemes because the engineers didn’t have a 3D model, or a context to model them in a 3D environment. And you know, you can’t say: “Here’s a drawing, here are plans; you assemble the structure.” We had to drive it. Really, we were the drivers and also, because CATIA had a great three-dimensional structural package, we could actually do the modelling of the steel itself, then hand it over in the end. For the Fish, the very first one, SOM was the structural engineer. They did the 3D modelling and we did the wireframe.

GL

That’s interesting.

RS

On Disney, I forget the name of the company in Los Angeles, but I went over and camped in their office for a number of weeks. We just worked with them directly. So I think Frank probably had a unique situation because of his crazy architecture. With me working through the architecture office, we drove the design, and then the engineers would analyze it, and then come back to us with their change and recommendations.

GL

And at the time, the whole idea of centreline geometry was the only way people knew how to do it, although everybody knew it was crazy. Centreline geometry doesn’t really do anything but give you dimensions. And if you’re not dimensioning all the members, it doesn’t make any sense!

RS

Yeah—it gives you lengths, and you can get loads, maybe, through the diagram, but you don’t know how it fits within that. That’s partly why you have to leave a lot of room, because you don’t know precisely. You can’t be as efficient at design.

GL

You already mentioned going to a boat builder, but was Permasteelisa in the mix at this point? Were there fabricators using this kind of technology who were ahead of the game?

RS

No. Permasteelisa worked with us on the Fish. I went over to Italy with the CATIA system for three weeks and introduced it to them. They’d just got started with the concept of 3D modelling for the Fish, and then they were brought in to Disney later on. I don’t think they were at an earlier stage if I remember correctly, before the Lewis House. Bilbao started up somewhere in that same time frame, too, I think. There were a lot of players that were looking, and getting interested. But there was still a lot of fear. I can remember contractors would come in and a lot of them would say, “This can’t be done—can’t be built.” And it took the proof of actually following the process through and, in a sense, that’s what Bilbao did. It proved that it worked. That broke open a big logjam for the contractors.

GL

Were you involved in any of the mock-ups? I remember for the Venice Biennale in 1991 Frank built a big mock-up of the undulating limestone wall for Disney; rather than putting in the Lewis House.

RS

Yeah, I did all that. I did all the modelling, the design, and worked with the stone fabricators on that one.

GL

And at that point the CATIA package was set up to generate things like toolpaths, that you’d then send to a fabricator?

RS

Yeah, CATIA had CNC back then because it was there for the aerospace industry. And—I forget the name of the stone company, but they actually went off after to design and build solenoid-controlled machines that could cut their stone following a toolpath. So they got into it big time from that experience.

All that was pioneering. In fact, talking about the software, I was thinking that for the Lewis House we were going from V3 to V4. In V3, there was no real time dynamic shading. You could only freeze an image and run it off batch mode to produce a nice and pretty shaded picture.

Before that, we had to have techniques for everything we did, so we could analyze what the shape looked like, by doing it in a wireframe. When the Lewis House came along, V4 came out with dynamic shading, so you could rotate the shaded image on the screen in real time. If we hadn’t had that then I don’t know how we would have done those shapes because they were so complicated. So that was another thing I feel like I really helped spring forward in function.

GL

Yeah, it’s funny. The guy who works for me is an aerospace and naval architect, and the shapes they do and the whole structural relationship between inside and outside is pretty simple when you get right down to it. There is a lot of refinement, but you don’t have surfaces that are folding around and through each other and…

RS

Yeah.

GL

So was there any exchange, did any of this ever go back into aerospace?

RS

I don’t know whether it went back. When I first got to the office they were asking, “What’s your experience of aerospace compared to this?” I’d say, “Well a wing section, for example, is one-directional with airflow compensated because of laminar flow, but then that section is pretty much the same along the entire wing.” Where the wing spars connect to the fuselage there’s a little bit of complexity, but nothing like what Frank was pushing. And on the Lewis House—I still think those are probably the most complicated shapes he ever did.