The Way of the Wobble

The meal is over. On the tablecloth there are corks, an orange, a few walnut shells, an empty glass and a coffee spoon. Those of us whose instinct is to see if we can somehow balance these objects, one on the other, are generally found to be annoying. Conversation falters. People wait for over ambitious configurations to collapse. Structural doodling is our way of playing at being engineers. We search for the point of balance as weight is transferred from one item to another. We load a folded card or a drinking straw until it buckles, replicating in miniature tests made with hefty presses and rigs in the basements of engineering schools.

The desire to play at building comes not so much from memories of the look of things as from memories of movement: memories as clear as the memories of smells or sounds, and sometimes more dramatic. One of my earliest is the shudder, pause, then the next shudder of an earthquake. Crockery rattled on the shelves. The whole house was in motion. We moved with it. There was not much damage – a cracked chimney stack had to be demolished (the house itself was timber frame), maybe there were broken plates – but the sense that a building is not a solid mass but something made only more or less stable by the stiffness of beams and walls, that the whole construction is flexible and unpredictable in its deflections, stayed with me. That was in New Zealand, in Wellington, where a fault cuts off one side of the harbour in a ruler-straight line. There, when you look down into excavations for new foundations you see concrete lap round a reassuringly dense weave of reinforcing rods. But you know that when the big one comes these buildings will shake and bend a little, even if they do not fall. And there are always forces which cannot be rationally planned for. Timber-frame houses exploded like paper bags in the great storm that hit Wellington in 1968.

I have another early kinetic memory, of swing bridges on tramping tracks. I remember their sway and the way the narrow-planked walkway bounced up to meet your advancing steps. It was worse if you didn’t cross one person at a time. Short in span, and made of wire for crossing streams rather than of lianas for crossing gorges, these bridges were otherwise pretty much the same as those liable to be cut away under Indiana Jones as he is chased through the South American jungle. New Zealand’s swing bridges came to mind when it turned out that the Millennium Bridge had a wobble. Reading about it I felt it in my feet.

Suspension bridge technology has been pushed forward by a need, on the one hand, and an ambition, on the other. The need is to prevent the rhythmic, self-destructive fluttering that can be generated when the cadence of a live load (marching men, wind-driven eddies) matches one of the bridge’s modes of vibration. The ambition is to cross wider spans with lighter, more economical structures. The two aims are to a degree in conflict, because light decks are likely to deflect in unpredictable and unpleasant ways. The result has been monuments that mark advances in civil engineering: Telford’s bridge over the Menai Straits, the Roeblings’ Brooklyn Bridge, and bridges by Freeman Fox over the Severn and the Bosphorus. A suspension bridge also provided what is probably – because recorded on film – the most widely publicised of all civil engineering failures, the collapse in 1940, owing to wind-induced oscillations, of the Tacoma Narrows Bridge in Washington State.

There is no reason to think that engineers make more mistakes than professionals in other fields, but they cannot bury them and are not allowed to forget them. A structure on which the safety of a trainload of passengers depends is a public performance. It’s easy to see that civil engineering could be, or should be, an anxious business and why engineers have, from time to time, been thought too conservative. Post-mortems on glitches and disasters make good reading for amateur engineers: they go back through the design process and tell you something about how engineers think.

An article by Tony Fitzpatrick of Ove Arup, the engineers on the Millennium Bridge project, describes the design of the bridge and the way its wobble was corrected.* One senses his satisfaction in being able to report that measurements made on the finished bridge matched calculated values for deflections of various sorts, and that the one effect that was not predicted fell outside the limits past experience suggested were relevant. It was in the literature but not yet incorporated in the bridge-building codes. Fitzpatrick’s report brings to life the diagrams in popular introductions to structural engineering, which assert that forces must be balanced and that tension and compression are not just about pushing and pulling but about making shorter and longer.

A few basic decisions about route and requirements set the agenda for the bridge: it was the central link in a pedestrian way running from St Paul’s to Tate Modern. There were good town-planning reasons for choosing this route. Closed to the north by the east transept of the cathedral and to the south by the brown brick wall of the museum, it is the closest thing we are ever likely to have to the streets radiating from the Cathedral and the Exchange that would have shaped London had proposals made for London after the Great Fire been realised. But the site brought constraints too. It is hemmed in by buildings on the north bank, and the profile of the bridge had to allow for the demands of river traffic, which set a minimum clearance for navigation, and by the height limits that apply to buildings round St Paul’s. Only a thin sliver of space was left.

The solution was a suspension bridge of a novel sort. The main cables on most suspension bridges sag in the vertical plane; rods or cables hang from these in straight lines and carry the weight of the deck below. On the Millennium Bridge the main cables make very shallow curves both horizontally and vertically, and unobstructed views up and down the river are made possible by having the cables sweep below the level of the deck. In place of the vertical hangers of a regular suspension bridge, box girders with projecting arms link deck and cables. It is, all in all, about as close as a suspension bridge, where deck and cables are separate elements, can get to being a ribbon bridge (like those I remember from New Zealand), in which the cables themselves carry the deck.

The high-tension, shallowly curved cables are anchored to deep, piled foundations on both banks. The tension pulls these forward, and must not be transferred to the ground between the piles and the river in such a way as to affect the stability of the embankment walls. To protect them, a narrow space, filled with material that is softer and more compressible than the surrounding ground, was made on the river-facing side of both anchor points. When a structure’s foundations are given to movement, all of its parts are in a state of continuous mutual adjustment. But the important readjustment, the one that hadn’t been planned for, had to be made because of the bridge’s response to the shifting load generated by walking people. At this point Fitzpatrick goes into mathematics and I can’t follow him, but the conclusion is plain: the lateral forces were the ones that mattered, for while the vertical forces of regular footfall cancelled out, its lateral forces tended to amplify one another. As the sideways movement increased, people tended to fall in step with the bridge, which amplified its sway still further.

One solution, to make the deck stiffer, was discarded. It would have added considerably to the weight of the bridge, both physically and visually. Instead dampers (of more than one kind) have been attached. They stop the bridge swinging too fast or too far – a hydraulic doorstop controls velocity in much the same way. The dampers complicate the look of the bridge, but don’t seriously compromise it. You could argue that, like flying buttresses, or the visible suspension on some racing cars, they are a proper visible expression of forces at work on the structure. The final result, according to David Newland, technical adviser to the bridge’s funding body, is ‘probably the most complex passively-damped structure in the world’. (In an actively-damped system an input of energy is required to answer the force with a counterforce; in a passive system, the force is simply absorbed.)

Why aren’t there more famous civil engineers? Things are changing, but the credit for bridges, which are quintessentially the work of engineers, is still regularly given to the architects involved. Like accountants, civil engineers often work anonymously for big firms, and when, as in Arup’s case, there is a strong association with the work of many celebrated architects (from Utzon and Spence to Foster, Rogers, Gehry, Piano, Koolhaas and Libeskind), the engineer’s contribution is likely to remain obscured from view. Reading the English newspapers you could get the impression that Norman Foster had designed the Millau viaduct single-handed: the name of the French engineer, Michel Virlogeux, was hardly mentioned.

It was not always so. ‘We see the monuments of a brief heroic age of engineering,’ L.T.C. Rolt wrote about the tunnels, bridges, cuttings and embankments of Brunel’s Great Western Railway, ‘as remote from our world as that of the great medieval cathedrals.’ No individual now, he added, can have Brunel’s mastery of the whole of his science – nor, therefore, his status.

Ove Arup, the Dane who founded the firm that still bears his name, is one 20th-century civil engineer whose name does have resonance. Founded in London, the firm now employs more than nine thousand staff in nearly forty countries. One of the reasons for Arup’s success in Britain must be that he was a foreigner, and from a country with a livelier tradition in some fields – concrete construction in particular – than could be found in England. His manners were different; for example, when he wrote to competition winners, it was hard to tell whether he was congratulating them out of professional courtesy, or touting for employment. His education was unusual, too. He became an engineer only after he had decided he wasn’t quite good enough to make a career in philosophy. He didn’t stop reading philosophy, however; it made him a good critic of false connections between structure and aesthetics.

His connections with modern European architects and his sympathy with what they were doing put him ahead of the game when English building began to have a modern, to a degree Scandinavian, flavour. He did clever things with concrete working with Lubetkin on Highpoint (an ingenious use of shuttering) and the London Zoo penguin pool (the walkways were admired by everyone except the penguins). Earlier he had worked on bomb shelters (having been appalled by the lethal technical naivety of the first brick shelters) and on the Mulberry Harbour for the D-Day landings. His favourite piece of design work – the elegant 1963 Kingsgate footbridge in Durham – was not a collaboration at all. Arup have put the engineering bones into so many significant buildings over the last fifty years and had a hand in so many major civil engineering projects, from the Sydney Opera House to the Pompidou Centre to the high-speed Channel Tunnel rail link, it might seem that they exercise a monopoly right to participate in the best new buildings.

In his biography Peter Jones shows that Arup’s achievement was as much to do with human relations as with technical innovation. He was a catalyst for cultural change in his profession; his monument was the firm and its ethos. Although in the early years Arup was better off than his colleagues, the financial arrangements of the partnership were essentially egalitarian. Some of his partners have said that he was more of a father figure than a boss. The most illuminating asides on what it was like to work with him come from the letters, notes, comments and memoranda of Ruth Winawer, his secretary for twenty years, now quoted in Peter Jones’s book. His response to outside complaints, she said, was to investigate them rather than immediately defend the firm’s patch, which wasn’t always popular with his colleagues. There was something of the spoilt child about him – the long chopsticks he kept in his top pocket to sample delicacies from other people’s plates; the refusal to take notice of the fact that, though he had a penchant for smart cars, he was a lousy driver – and he could, I imagine, be irritating if you were immune to his energy and charm.

Arup was always good with people – with his family and with the talented people he surrounded himself with. One of them, Peter Rice, an exception to the rule that engineers don’t have reputations outside the profession, became resident engineer on the Sydney Opera House at the age of 28. He went on to make significant contributions to the Pompidou Centre, the Lloyds building, I.M. Pei’s Louvre pyramids and Kansai international airport. That the firm was able, in the end, to get on very well without Arup himself is greatly to his credit.

Arup was already 50 when the partnership was founded in 1946. The Opera House job, which followed a few years later, brought it great prominence. Alas, as it progressed, a deep antagonism developed between Arup and the designer, Jørn Utzon. Utzon was fighting off pressure for changes from the client that would have compromised his vision of the building, but he had little or no experience of managing a project on this scale, and had ideas about structure that were more romantic than logical. Even the much praised roof was in a sense wrong-headed – a brilliant design executed in what may have been the wrong material. In reminiscences of the early 1930s, Arup describes meeting ‘a number of young people who really were interested in new ideas’, who ‘were in love with an architectural style, with the aesthetic feel of the kind of building they admired’, but determined to design in reinforced concrete even though the job would have been done better and more economically with other materials. One wonders what would have been the outcome if Utzon had given Arup the shape of the roofs and been open-minded about the materials to use. That sort of dialogue between engineer and architect is now, to judge from accounts of later jobs, the rule.

In 2002, Cecil Balmond, Arup’s current deputy chairman, published a book, informal, whose language and the structures it describes are as far from Fitzpatrick’s account of suspension bridges as the Millennium Bridge is from one of Telford’s. The analytic tools – computer analyses and simulations – are much the same, but for Balmond engineering involves as much exploration and pattern-making as it does problem-solving. Here is his description of his exchange of ideas with Daniel Libeskind about the now abandoned V&A Spiral:

We exchanged metaphors. If the form were closed it could be a mineral deposit, or if an open transparent steel-framed building, it could be a lantern or beacon. If it were heavy it could be hacked out of granite, or was it buildable out of special masonry? The images helped loosen the thinking and helped us look for the radical … As Libeskind developed the semantics of spiral I looked at the syntax of connectivity and implied movement. It was not just architecture and engineering but a wider endeavour being moulded.

Such fluidity of thought is possible because calculations that were once laborious and expensive are now easy and virtually free. This brings its own dangers. Balmond goes on to say that architectural form must not become an ultimately meaningless ‘gesture of software’: ‘Our brains, random as they are, do not act without interior connections – there are rules – albeit ones hidden to us.’

The simplest such rule might stem from intuitions about structural stability which tell us to keep out of a building that looks as though it could fall and crush us. The idea of a hidden rationality takes us back to ancient sources – the mathematics of the golden section and what we used to think of as Pythagorean harmonies – that went rather well with those ‘regular framings of closed squares and rectangles’ that Balmond nonetheless says he wants to question. The mathematics he looks to now – topology, set theory and so forth – offer architects inspiration when they want to break the rhythms of spaced post and beams, the regularity of circles, parabolas and catenaries and all those other shapes calculated by the same simple formulae that generate the conch shell and the soap bubble. There is an understandable taste for refreshing dissonance, even for the illusion of danger, in a world where the undecorated orthogonal geometry of new buildings presses down on the earth all around us. How far Arup would have approved of the mould-breakers’ endless flow of free calculation is hard to say. It would have pleased him that the webs, shards, bubbles and carapaces of new buildings come with no promise of automatic human improvement. That engineers can now contribute so early in the process to the design of buildings would surely have delighted him.