The Beauty of Numbers

Inspired by his work with influential architects, Cecil Balmond searches for that exquisite blend of structural form and structural logic.

On a warm day this fall structural engineer Cecil Balmond stands in a Soho gallery surrounded by a huge metal fence. Chains rise miraculously from the ground—others float in midair. There is nothing supporting them. They’re neither hung from the ceiling nor secured to the floor, but instead are held in place by tension from steel plates that look like abstract gingerbread men. The installation, called H_edge, actually resembles a futuristic privet hedge from a postvegetation world out of Mad Max. Created by Balmond and his Advanced Geometry Unit (AGU) at Arup, the piece is nearly eight feet high in places and turns the gallery into a maze while casting bird- and flowerlike shadows on the floor and walls.

Balmond says the installation “was inspired by the Indian rope trick, from a model on Francis’s desk” and gestures to Francis Archer, a tall man with the tweedy aspect of an Oxford don. Archer had been a quantum physicist at Cambridge before he moved to math and then engineering, joining Arup. He shoves his hands in his pockets and tells me about the Chicago Tribune’s 1890 hoax: the story of an Indian holy man who threw a rope in the sky; a boy climbed up after it, and his body parts rained down on the transfixed crowd. Minus the body parts, it’s a good analogy for Balmond’s work—creating mythic-seeming structural possibilities.

Balmond—his hair balding into a tonsure, the overhead lights reflecting off the top of his head—smiles his gap-toothed grin. He explains how they’d used a fractal, a Menger sponge, to shape the H_edge’s maze so it repeats the pattern exactly throughout the space at different scales. Balmond has been the engineer for some of architecture’s most influential structures—both built and unbuilt—like Rem Koolhaas’s CCTV tower, in Beijing, and Daniel Libeskind’s Spiral for the Victoria and Albert Museum. Balmond was also the cocreator of four of the pavilions the Serpentine Gallery builds each summer in Hyde Park. Now he’s developing master plans in London and St. Petersburg, and designing buildings. His first, a bridge in Coimbra, Portugal, opened last November and hardly looks like the work of an engineer—no wires, no trusses, no obvious structure. Balmond has taught at Harvard and Yale, and now has a permanent chair at Penn. He is arguably the most famous structural engineer alive, with name recognition akin to that of the architects he works with—Koolhaas, Libeskind, Toyo Ito, Àlvaro Siza, Shigeru Ban. Koolhaas has worked with Balmond on nearly every project and competition OMA has undertaken—more than 30 collaborations—since the mid-1980s, when they first met. “That number alone is evidence of the interest and shared passion for a particular kind of research,” Koolhaas says.

“There’s no one else like him,” Libeskind says. “He’s exploring the whole issue of structure and pattern. It’s not just using virtual reality and computers—there is a whole ban-alization of architecture where people are just playing games and inventing ludicrous shapes. But with Cecil it’s not just a pretty facade with a different shape and traditional structure.”

In a real sense Balmond is at war with the box, with those pretty facades and blobby buildings. Form for its own sake isn’t good enough for him—nor is creating a new shape without rethinking the underlying structure. His solutions inevitably have an enormous impact on buildings, but it’s not as if he wants them to look like “feats of engineering.” Instead they appear so integral to each project that you can’t tell the engineering from the architecture. Which brings up all sorts of authorship questions. “The whole question of influence in my view is barely relevant,” Koolhaas says. “And I would even say so is the issue of authorship.” The issue nearly got Balmond sued by a young firm he worked with in the late 1990s when it accused him of taking too much credit for the work.

Still, with each project he wants to see how structure can be deployed. In his collaborations he insists on working with architects from the start. “If you do the outside of a building first, it cuts off the building’s possibilities and you’re limited simply to deciding where the columns go to support that facade—it’s all about just refining something,” Balmond says. Not surprisingly, he’s also opposed to the traditional ways engineers work, looking for maximum efficiencies and the positions of beams and trusses. It’s a view that seems ironic given that he’s the deputy chairman of Arup, the world’s largest engineering firm, with 7,000 employees worldwide.

In London, Daniel Bosia, one of AGU’s engineer-architects, guides me through the group’s office. A giant chartreuse fabric flower dominates the space. “It’s a structure with both tension and compression,” he explains. He adds that it looked like a potato chip “until you put the second load on, and now it’s like this.” Bosia gestures to the flower and says that when artist Anish Kapoor saw it he started sketching. Just beyond, a sewing machine stands in a corner. Sheets of paper taped to the far wall are emblazoned with words like complexity and pattern.

“The way we get to a design isn’t through a linear process like optimization or refinement of preconceived form,” Bosia says. “It’s more about discovering a system. H_edge is like that. It’s starting from testing systems—materials systems, numeric systems, and their opportunities in architecture.” To reach such ends they often use fractals and algorithms. The hallmark of an AGU member is perhaps being a math geek. The six-person group includes a computer whiz in virtual reality, mathematicians, and physicists, as well as a couple of architects, Bosia explains. He takes me past a model made of crisscrossing Popsicle sticks, the initial make-up of Shigeru Ban’s Pompidou outpost.

Balmond has two offices—a cramped one down the hall from the AGU and another in Bloomsbury, where he’s next door to the chairman and on the same floor with “the number crunchers,” as he calls them. The space has blue carpet and a gray desk, and just beyond him is the sort of facade he’d scoff at. He does in fact: at one point in our discussion he points at the windows and shakes his head dismissively. A caricature of Balmond, the kind of thing you’d get at a street fair, leans on a bookcase by his desk. In the drawing a devilish-looking Balmond twists a Rubik’s Cube; at the bottom it reads, “Cecil’s cube.” Clearly his fascination with math games isn’t a secret.

Balmond has the look of a slim Buddha, with gray-green eyes and deep lines carved around his mouth. He speaks softly, a hint of his native Sri Lanka in every word. He picks up the spreadsheet he’s studying for his board meeting tomorrow and talks about his role in the business as head of the buildings group, which includes hotels and hospitals, offices, and government and defense contractors among the 13 categories. He waves his hand like a shuttle loom explaining how the different sectors march across the page, and it’s clear how much he enjoys the business side. When asked about which of Balmond’s roles is predominant, business or design, Bosia says, “Cecil’s a businessman first of all. He knows everything has to work in a certain marketplace and that involvement with the business enables a different architecture to be built.”

That Balmond even became an engineer is a fluke. He wanted to be a musician—and didn’t really see engineering as a calling until his late 30s. Born in 1943 and raised in Sri Lanka, he had to choose at 13 between medicine and engineering. And although he won a competition to design a house at 19, he entered it only for the prize—a motorbike. Not long after, he left for Nigeria and then England, where he joined Arup in 1968. It was more than a decade later that he had an awakening to the possibilities of structure while working with James Stirling on the Staatsgal-erie, in Stuttgart.

Balmond convinced higher-ups at Arup, mentors like the legendary engineer Peter Rice, that the gallery shouldn’t have columns; but after Stirling won the competition, Balmond decided to use them. “Not because I’d lost my will, but I could see how they would help your experience of space.” Now when discussing structure, he describes it musically as “syncopation.” Talking about the Kunsthal, his first built project with Koolhaas, Balmond says that the red zigging and zagging ceiling beams go “cheung, choong, choong.” He looks up from his drawing. “Those rhythms determine how you read space, and only after the patterns are set do I look for a materiality or how to connect in the system. It’s a reversal from how I would have done it thirty years ago when I would have just looked for where to put the columns.”

Crucial in that change has been his work with Koolhaas. The two met shortly after Balmond finished the Staatsgalerie. Koolhaas had completed only one building when he came to Balmond about a competition. “Rem was shooting from the hip, taking no prisoners—an intense guy, and I liked him immediately,” Balmond says. The two became close friends. Koolhaas was interested in interrogating the expectations of structure, Balmond in how structure impacted architecture. Now, Koolhaas says, “Cecil has a much more profound understanding of architecture, and I a much more profound understanding of structure. That creates a situation where we simultaneously talk about both and where ideas can be launched from either side in either domain. Collaborating with Cecil has allowed us not to differentiate our architecture but to produce a different architecture, yeah?” He waits for me to nod. “An architecture that is more fundamentally engaged with issues of structure than many of our colleagues’.”

Balmond pushes his chair out and calls for his assistant to bring a book he wants. He returns to sketching the Kunsthal, before drafting a box with crisscrossing lines on it for Ito’s Serpentine gallery. The temporary structure is perhaps the best expression of Balmond’s approach to numbers and engineering. “If you look, you can’t see what we did,” he says. He points to a photo of a building covered in a filigree of lines. “At the start Ito said, ‘Let’s have a box’—but what kind of box? So we go, ‘Okay, box, what’s different there?’ And at first we both drew some random lines on the roof.” He sketches out crosshatchings and a colonnade. “But when I was looking at it, I thought, A roof on stilts? That didn’t feel right, and if you’re drawing random lines, your own bias comes in. People can’t create the random.”

Instead Balmond started to think about an algorithm they could use. He laughs and says, “I know when I say algorithm it sounds like”—he makes a stagy gasp—“algorithms are the new science. But they’re nothing new. Algorithms are the oldest thing in the world. If you count, it’s an algorithm, 1 + 1, 1 + 1 + 1. Our numbers are algorithmic. If you do any ballroom dancing, if you repeat yourself in a sequence, that’s an algorithm.” He was looking for something to serve as the engine to drive their search. Balmond draws a square and takes a line from the middle of one side to a third of the way down the next side. “If you keep doing that over and over and over, half to a third, half to a third, you get this shape.” It looks like an asymmetric diamond. “See,” he says and sketches another square where he divides the line in half. “A steeple is just taking half to a half.” And indeed his rotating square looks remarkably like the inside of a steeple.

With the pavilion, “I wasn’t finding structure, I was looking for a pattern first. I knew this pattern could be a structure because of that steeple drawing, but it’s not architecture, it’s a pure algorithm run six times. To make it a piece of architecture, I want a network of lines.” His pen scratches on the paper as he extends the lines so they join outside the square. That diagram became his model. He cut out the four corners so it would fold over. “And voilà,” Balmond says, “I’ve got a box.”

He points to the corners at the base of the pavilion, but there is nothing there. Because the building supports the structure through the crisscrossing lines, the corners have no structural purpose, so he suspended them to make it obvious they weren’t carrying weight, while the load-bearing lines were made of thicker steel. The resulting building was a box, but one Balmond could approve of—without columns or anything that would make one think of a box; the pattern on it created the structure.

For CCTV, the mammoth telecommunications center currently under construction, the process was different: “totally opposite,” though the building still bears the hallmark of Balmond’s interest in pattern and how you can create something that defies the generic solution. From the outset he and Koolhaas worked closely on it. It started with a late-night phone call—and a mad flurry of faxes. Balmond sent one of a building that, he says, “was much more brutal than it looks now.” He sketches and says, “I knew with this doglegging form that the whole skin was being mobilized without knowing what it was. I wanted to get Rem away from any feelings of core. I knew we had to work hard on the skin, that we needed it because of the torsion from the shape. Plus it’s an earthquake zone, so I was worrying about all those things.”

In New York on the day the CCTV show opens at MoMA, Koolhaas sits hunched over a low table in the lobby of the Mercer Hotel. “We developed the building in a series of organic sketches and models,” he says, sketching one of Balmond’s early ideas, drawing the upside-down asymmetric U that has made the building so famous. He traces a quick spiral of tight circles around it. “Cecil was interested in finding a structure that wrapped the whole thing in an organic ringed system of structure. We pursued that for a while, then it got away from that idea to something more classical where it had to be divided in equal sections.” But neither of them would let go of the idea of a skin. “We felt that the sectional system wasn’t doing it much justice and felt this continuity,” Koolhaas taps his pen on the spiraling drawing, “was more desirable, but we asked, ‘Does every member have to work equally hard?’”

In London Balmond answers that question easily. “We quickly realized that if you take the generic skin, it changes its patterns as it moves around the body, so without studying it we just drew zones to begin.” With a traditional building—the Hancock Center, say—all external bracing has a set pattern, and Balmond didn’t think they needed that here. “With typical engineering, you take a peak condition, take a section and spread it across the whole project, and to me that’s wasteful. I’d rather make this work here and that work there if it has a place within the architecture.” So he put his ideas into the computer to see what happened. “Sure enough it was giving different patterns all over the place with different stresses,” he says and smiles.

He and Koolhaas decided to put material where they needed it and leave it off where they didn’t. For the skin they created a diamond pattern that could be deployed at different scales. The pattern itself never changes, but because of its serially repeating nature they could use a larger version of it with less bracing where they needed less support, and where they needed more a smaller one, with tighter diamonds and additional bracing. “Rem understood that this would give a power to the building. He knew this form needed something—and this was it, purely structurally driven.”

“You know,” Koolhaas says, “neither of us is interested in radicality per se. It’s more of a kind of resistance to the obvious—or a further investigation of the obvious.” That questioning of expectation is what has informed Balmond’s bridge in Coimbra, where he was fighting against the foregone conclusion that when you step onto a bridge you get to the other side, the teleological imperative of a pedestrian bridge.

Now his bridge crosses the water in a zigzag, like a lightning bolt. His original plan was inspired by skipping stones and the low arched bridges in Oxford. He presented the idea and told the clients, including the town’s mayor, a story about skipping stones, and everyone left the meeting happy—except for Balmond. “I was on the plane and kept thinking of being on the bank and looking at the water. I knew I hadn’t done justice to the site, and I didn’t want to go across straight. I do not want to be told that the moment I step on this bridge, as soon as I do that, I’m going to be getting off on the other side. So I just drew two lines that never met.” He joined them with a platform in the middle, and it worked: the bridge became even sturdier.

Balmond jumps up from his chair. “See, thin bridges like the Millennium Bridge often wobble,” but here that horizontal between the zig and the zag gave it more strength. He stands with his feet hip-width apart, then one in front of the other spaced about a foot or so apart like his bridge, to demonstrate. “It’s more stable.” So now one-half is braced underneath along the far side instead of in the middle, as bridges are normally. The other half is braced on the opposite side so that each half is a mirror image, and the span an elegant curve. It barely seems to be about the structure at all, just this delicate move across the river.

That night we have dinner with his family at his house in Hampstead. Afterward we go up to his music room. He’s fighting a cold but rallies there beneath the silver-foil ceiling. The room has a decidedly Moroccan vibe with large cushions and low tables, kilim rugs, and Sri Lankan batiks. The stereo is perfectly balanced. He props my chair with pillows so I’m at the perfect height for the acoustic experience, and he starts playing Miriam Makeba and then Rattlesnake Annie, “a huge influence on Willie Nelson,” he says.

Balmond has thousands of records, everything from jazz and R&B to country, including rare Sun recordings of Johnny Cash. As he talks, I can see in him the young man who’d just moved in with his wife, Shirley, but was too poor to furnish the place. Since then the space has become the music room, and is, perhaps more than anywhere else, the summation of who he is. His eyes close as he listens and sways to the beat. “Here, listen to this.” He puts on a record of Zulu chants.

“Isn’t that powerful—like prayer?” Balmond says, his words precise and lilting. “It’s my escape, this.” He waves his hand around the room. “It’s all I need at the end of the day, an hour here and I’m refreshed.” Watching his rapt expression as he listens to the song, I understand what he means when he’s talking of structure and pattern and rhythm, why it’s all music and numbers. He says the patterns in algorithms allow him to go further than his own mind would let him; so too does the music, and the two together let him do amazing things.

Recent Programs