French civil engineer Michel Virlogeux went out on a limb to ensure his brainchild, the Millau Viaduct, got built as he wanted it, only to be overshadowed by architect Lord Norman Foster. Andrew Mylius met him in Paris for the engineer's angle.
Michel Virlogeux's business card says a lot about how he sees engineering - or at least his role as a designer. It declares him as an 'engineering consultant and 'concepteur' for works of art'.
Immodest the proclamation may seem, but with design of the dazzling Millau Viaduct (NCE 16/30 December 2004) and Pont de Normandie (NCE 12 November 1992) among the 100 or more bridges on his curriculum vitae, it would be hard to dispute. In the flesh, though, this dapper 58 year old is anything but assuming.
Ensconced in a shabby room loaned out to him by cable stay specialist Freyssinet in the industrial southern suburbs of Paris, Virlogeux explains in rapid-fire English and with a good deal of gesticulation that he sees himself in an 18 th century tradition of French engineering.
'In the traditional education of the Ecole des Ponts et Chaussees [where he studied], art was a part of an engineer's training. In the 18 th and 19 th centuries engineers studied drawing, and on bridge projects they had to develop decoration [as well as make their structures work].' Virlogeux maintains that aesthetics should still be central to the process of engineering design.
'This is not a one person view.
But, ' he laments, 'it's a tradition that's in decline.' Design fices are having aesthetic creativity squeezed out of them by increasingly slender fees, he believes. 'They don't have enough time to work, not enough money to develop ideas - and in the long term it means we're losing the best engineering students.' He also lays much blame for the loss of art in engineering on the use of computers.
They dehumanise the design process, he argues. And their ability to crunch the numbers involved in any problem, however bizarre or inelegant, actually stops engineers from asking fundamental questions about form and force.
'The engineer's role is to design a structure with a very clear flow of forces. If you have a very clear idea of how to organise your structure you are able, at least for the main things, to do very simple calculations to check the dimensions and forces.
Too many people now are using computers to see how forces pass through a structure. The relationship is upside down.
'Personally, I don't have a computer.' Virlogeux offers the maxim used by the western world's first known engineer, Vitruvius, as a guiding principle. 'Utilitas, fermitas and venustas - utility and functionality, stability and durability, and beauty. In that order. It all begins with utilitas.
'The greatest art comes from making things very simple, but very elegant and perfectly adapted.' In seeing the Millau Viaduct through from conception to completion, Virlogeux has set out to exemplify the synthesis of engineering with art.
He claims to have been unruffled by the widespread misapprehension, even in France, that UK architect Sir Norman Foster designed the leggy, multi-span cable stay wonder.
And lately the balance of attention has swung heavily towards Virlogeux, to the extent that the half to full day a week he is giving up to radio, TV and magazine interviews is becoming a cause for concern. 'I'm self employed, so when I'm talking to the press I'm not earning.' Even so, he admits he has been overshadowed by a man who played second fiddle throughout the whole design and construction process.
'It probably wouldn't have happened 50 or 60 years ago.
I am as much at fault as anyone - engineers have become very bad at communication.' Yet Virlogeux is emphatic:
He could not have made the Millau Viaduct the triumph it is without Foster.
'I am able to see what kind of structure is fitting to the landscape, and what is technically suited to the conditions and constraints of the location.
I'm able to develop the global proportions. But I'm not able to do the detailed shaping, and that's not a minor role.
'From the global idea you can form the detailed shape so that it expresses the flow of forces and can enhance the structural concept. This is something that, personally, I cannot do.' He backtracks to the inception of the Millau project in 1988, when as head of state transport agency SETRA's bridges department he started looking for an alignment for this section of the Paris-Barcelona auto-route.
Virlogeux decided to cross the valley by descending to the River Tarn from the north. He would leap it with a bridge with a fairly impressive 700m span - the length was important 'to limit the height of the piers'. The road would then continue up the steeper southern side of the valley on viaduct before entering a tunnel through the valley wall to reach the plateau above.
'Though it was not beautiful it seemed to me the only practical way. But as soon as the idea was fixed, a road engineer asked:
'Why do we have to go down [into the valley]- Why don't we go straight across on a viaduct-' It was absolutely clear, that was the solution.' Thrilled by his colleague's radical alternative, Virlogeux sketched various multi-span cable stay options and within a few months selected a seven pier design as his favourite. His department worked up 20 or so alternative designs, 'some of which were absolutely ugly, absolutely stupid', to make sure there were no better ideas 'and to prove that the multi-span cable stay viaduct was the best'. The list was winnowed down to seven.
But Virlogeux's plans were to be hijacked.
Design of the Millau Viaduct coincided with construction of his first really big bridge, the Pont de Normandie, and with a change at the head of SETRA: The director through the 1980s, an engineer 'who was interested in everything technical', was replaced by a politician.
'Nobody cared about the Normandy Bridge when we designed it - really, nobody. I made one design, straight. But when people saw it growing there was huge public excitement.
Suddenly, everybody had something to say about it.
'The new head of SETRA saw this and realised that, as soon as Millau became known, it would also attract intense interest. His opinion was that a project coming from within the administration would never be accepted.' Those fears were realised with Virlogeux's unwitting help.
'I made a big psychological mistake in 1992 by making a drawing, showing at the same scale the Millau Viaduct, [Gustave Eiffel's seminal 1884 truss arch] Garabit Viaduct and the Eiffel Tower. The Eiffel Tower was smaller than the tallest pier.
'This drawing was published in the newspapers, so everybody saw it, and everybody wanted to be involved in the project.
There were lobbies organised by contractors, design engineers and architects - mainly architects - and the director decided he wanted to make sure there were no other, better ideas.
'So he organised, not exactly a competition, more a definitive study. There was an absolutely huge response.
Seven architects, including Foster, and eight engineers were selected and asked: 'What do you think about the seven project proposals by SETRA. How would you improve on or treat them?
Have you other solutions-' 'Of the responses, only one was similar to mine. It's fair to say they were all more classical. Some of the solutions from architects were absolutely terrible - heavy structures with decoration, with balconies. . .' Virlogeux momentarily looks as if he has been sucking lemons.
Seven designs, including Virlogeux's multi-span cable stay brainwave and two other SETRA options, were shortlisted. But the upbeat public announcement was accompanied by a blow. SETRA's director declared that the internal design office, Virlogeux's team, would be taking the project no further. The project would be let to external engineer-architect teams.
With 20 years at SETRA, Virlogeux was torn. However, 'I had not developed such a beautiful solution to hand it over to somebody else. I believed it to be the best design, and I wanted to get it built.' He prepared to jump from SETRA and set up on his own to take his design forward.
After that, the partnership with Foster came naturally, Virlogeux says. 'Foster had been selected [as one of the seven shortlisted designers] because he liked this [Virlogeux's] SETRA solution the best - he was convinced of the solution. He made some drawings which we could not build - the columns and deck were too slender - but were expressing something very simple, straight, light, transparent. This was exactly what I wanted.' They joined forces.
Their working relationship followed a pattern tried and tested on most of the projects he has undertaken. 'First the architect has to give his opinion on the solution - how it's introduced into the landscape.
Clearly, Foster liked the design.' The tables then turn. Virlogeux asks his architects to sculpt his structures so they best express the forces at play. 'Personally, I can't do the sculpting, but I'm able to see when it's OK.
Sometimes I tell architects:
'No'. Virlogeux is adamant, he is always in charge.
'The bridge hasn't been designed by Foster, that's very clear. I'm the designer, he is the architect.' Illustrating the difference between design and architecture, Virlogeux carries 100% risk for performance not related to construction.
Responsibility for design was written out of Foster's contract.
At no time, however, did Foster's input infringe on structural good sense, though.
'Foster understands structures.
There's big fashion for architects to do crazy things, and engineers feel obliged to make them work, or to do big and striking things themselves to compete with architects. Some are taking risks.
One of the great things about Foster was he did not try to push us in a dangerous direction technically.
'The engineer must not be reduced to the man who does only computations. But nor must you reduce the architect to someone who just does the finishing touches. It's something that must be more integrated.' Foster's contribution to Millau is measured by its invisibility rather than by any obvious stamp of authorship, Virlogeux says.
The bridge's efficiency is laid bare through looking very simple. 'You have the impression that there is no work. This is absolutely the contrary.' The architect's touch can be seen in the way the columns rise apparently straight from the ground. Foster insisted that the viaduct's colossal pile caps be invisible. Finding the ideal form for the columns themselves involved an evolution through 40 scale models.
Millau's box girder deck flows seamlessly into the hillside at either end. 'We have designed very unconventional abutments - very narrow, with exactly the shape of the deck - where normally we'd make the abutment wider than the deck to make the transition easier. It was incredibly difficult, but it means that you don't see the end of the bridge.' Foster also eliminated intermediate piers from the first spans of the viaduct.
'An engineer is always a bit handicapped by what he's done before, ' Virlogeux explains.
On the Pont de Normandie Virlogeux applied a principle successfully tried on a small cable stay bridge he designed in the 1980s. Intermediate backspan piers were used to provide rigidity, acting as additional anchor points and so stiffening the main span. 'All the stays are working as backstays.
'I had that so much in my mind that I put them into the design of Millau. But in fact it's a bit stupid.
As the bridge is designed to be stable over its intermediate spans without backstays, of course it could also be stable over its back spans.
'Foster cancelled these intermediate supports in the backspans and I immediately saw why. It makes the structure the simplest possible concept.' If any doubt lingers about the engineer's feelings for the architect, Virlogeux dispels it. 'We're trying to find new competitions to go for together.'