The stunning Millau Viaduct in the south of France was opened late last year with great ceremony by head of state Jacques Chirac. But without doubt, the brighter media star was a figure standing alongside the president, UK architect Norman Foster.
Foster, whose winning design has been transformed into eight soaring spans of concrete and steel, is world famous, with a string of high profile, international projects to his name.
His structural engineering counterpart on the project, Michel Virlogeux, is also well known - although his fame is pretty much limited to France. He was acclaimed as lead designer of the Pont de Normandie and Millau has reinforced his excellent credentials.
Virlogeux's high level of public recognition is in part cultural.
Engineers have historically enjoyed far higher prestige in France than in the UK. Had Millau been built in the UK, though, Foster would almost certainly have generated more column inches than the scheme's civil engineers.
Engineers have to put up with such overshadowing, and by and large they shut up. But it does not sit well with them. Foster felt the profession's simmering resentment during the infamous London Millennium Bridge debacle. The Thames footbridge suffered from unexpected and alarming dynamic excitation. Tempers briefly boiled up after Foster distanced himself from the problem, landing blame squarely on the engineer.
So in talking to NCEI about what the architect can bring to big civil engineering projects Foster treads carefully, emphasising that bridges are primarily engineering projects.
'That is the dominant aspect, ' he says. 'We do not bend the engineering to suit our ideas but work with it, and perhaps even draw out the possibilities.'
But as soon as he begins to explain how the Millau design developed you can understand how well architecture interacted with engineering, and how important it is. Certainly the structural engineers on his team thought so; among half a dozen entries competing for France's most prestigious job in 1993, his was the only one which had the architect as the spokesman.
'The others were persuasive and insisted that I presented - although I only did it on the basis that any engineering aspects should be queried as a direct conversation between the jury and the engineers.'
What he did was to jump straight in at the highest point, confronting the selectors with a philosophical challenge, he says.
'I began by asking them a question - was this to be a bridge over the Tarn which celebrated the river crossing in a heroic manner?
'If so, ' I said, 'then we are not your team, because that is not what we have done. We believe this should instead be a link from one plateau to another - the river is just an incident, albeit important, along the way.' That decision meant knowing the terrain, Foster says, the feel of the landscape, the shape of the huge valley, and careful study of the many upward or distant horizontal sight-lines. It it also meant a creative dialogue with the engineers from the outset about what was possible.
That could be done, he says, because the right individuals were able to interact - especially he and Virlogeux, who was with consultant Setra during formulation of the scheme, and then moved to become an adviser to the client once the project began detail design and construction.
Collaboration is always 'a question of individuals and the chemistry between them', Foster says, 'even though they are, of course, the focus for a whole team effort'. With the right individuals the whole team can be enabled to work together, dissolving the boundaries between engineer and architect, he claims.
At Millau the fusion produced a bridge consisting of seven slender piers 'marching across this 3km wide landscape'.
Getting the piers right aesthetically involved striking a delicate balance. They would be equidistantly spaced. They had to be economical - structurally efficient and buildable. And their number had to be optimised. Each pier represented massive cost, so there could not be too many of them. Yet, excessively long spans would create their own deck construction difficulties and cost increases. The selected compromise was a span of 342m (NCEI May 2003).
As they grow up from the valley floor the sculpted, tapering piers are designed almost to disappear, says Foster. The large 'needle eyes' at their tops begin the process of lightening the structure.
Trompe d'oeil contributes further.
'A sculptor will tell you that if you want a piece to disappear in the landscape it should be dark. I remember a conversation with Henry Moore that a bronze would do this - to make it appear again it should be light.'
Most viewers will see the Millau crossing against a blue or white cloud sky, so the cables are white to minimise their impact.
'You can only see them looking from above - but that is for the birds, literally.' He quickly sketches in with dark pen on a recent photograph to show the difference - with cables fully visible the structure is much heavier, 'another animal altogether. 'And that is completely cost neutral for the engineering, ' he adds.
The effect is helped by a lightweight deck in steel, proposed by winning concessionaire and contractor Eiffage, which will operate and maintain the link for the next 75 years.
'We suggested both concrete and steel options, ' says Alistair Lenczer. A civil engineer by training, including time at Arup, he now plays an important role in speaking engineer's language within the Foster partnership. He is project architect for Millau and works on other schemes such as Wembley Stadium.
'The steel allowed for a slimmer diamond cross section deck. And it also reduced the number of cables needed in each fan by two.'
Foster was involved in the selection of the design and build contractor and says he was pleased with a design that 'went in the right direction'. His overall concept for the bridge, especially for the hollow curved sides to the piers and the concrete finishes, was firmly pinned down as soon as detail design got under way.
As Eiffage developed detail drawings they were examined and commented on by Foster's six strong project team. Final decisions were made by the client after round-table discussions with engineers and contractors.
Eiffage's contract specified that the architect's form should be followed closely. But that did not mean conflicting with engineering needs, Foster emphasises.
For example, a curve in the line of the road, on a large 20km radius, is a key part of appreciation of the bridge by drivers, says Foster. These are after all the 'inhabitants' of the structure, he says. 'It makes all the difference.'
However, it fitted well with the curving approaches of the motorway at either end of the viaduct. These were used as deck fabrication and launch yards; construction of curved deck sections meant no additional land take was required.
Foster maintains that it is critical to follow a project through from start to completion. 'You have an interest in every stage - not simply 'we are the designers' but looking at the detail all the way through.'
Close monitoring and particularly the use of full scale mock-ups of elements such as the deck wind shield are important, he says.
For Foster, work on the viaduct is a part of a philosophy which concentrates on infrastructure and urban spaces as much as buildings. Cities in the modern world make their impact not by individual buildings but as a totality of links and connections, he says.
To work on those connections means an integration of the work of the architect and the engineer.
Just who is leading depends on the kind of project and those creating the work, he claims. But he is adamant: If a civil engineering project is to be architect led, the designer must be able to convince with engineering arguments.