A massive but intricate timber structure that took shape outside London’s Victoria & Albert Museum in time for the London Design Festival in September provided a visible symbol of the Museum’s commitment to the Festival as well as being a bravura example of imaginative design. But it was also much more – a demonstration of the power of collaboration, of what can be achieved when some of the best imaginative, analytical, and practical minds come together to create something that is at the limits of the possible.

Called the Timber Wave, the 12.5m high structure was designed by architect AL_A with engineer Arup and specialist fabricator Cowley Timberwork. Made using American red oak, it is the third collaboration between the American Hardwood Export Council and the London Design Festival – and the most ambitious. Despite its huge size, it uses furniture-making techniques and bent timber in a manner more common on a much smaller level.

Amanda Levete, the founder of AL_A, said, ‘Normally I feel uncomfortable if a design doesn’t have a function at its core, but the purpose here is very clear: to create something that is self-supporting, that is made out of timber with minimal steel connectors, that goes 12.5m high and doesn’t need to be tied back to something.’ She described it as ‘The most structurally complex project that we have ever undertaken, which was deeply stressful but ultimately successful.’

The idea, which evolved after the team rejected several earlier concepts, was to create a self-supporting arch, with one end standing on the pavement and the other on the shallow steps of the museum. It has a direct visual relationship with the intricate carved stone arch of the museum entrance.

Structurally, the form is a series of trusses, but this is not visually obvious because the elements of both the cords and the braces are all curved, so that it is not easy to read structurally. And, of course, it made the structural engineering far more complex. Ed Clark, a director at Arup, said, ‘We worked very hard to embed structural stability. The distortion complicates the load paths.’

What made it particularly complex was that, because the structure is synonymous with the appearance of the piece, any engineering changes immediately affected the architectural concept. So there had to be an iterative approach, with Arup sending changes back to AL_A, which would make adjustments, leading to Arup needing to recalculate and make further amendments. Andrew Lawrence, the timber expert at Arup, said ‘There was just a tiny intersection of the circles of budget, time and engineering. Somehow we have got there’.

The design uses glued laminated timber (glulam) and one issue was that the tighter the curves, the thinner the laminations had to be. The other challenge was with the jointing. The chords were broken down intoshorter lengths,connected at complex joints which typically also brought together four braces and, on the most highly loaded part of the structure, two ties as well.

All the joints are in stainless steel, as the tannins in the oak would attack mild steel. The braces are flitched with steel plates, with bolts passing through the lugs at the end of them. Clevis joints at the end of steel rods join the chord members. Although there was a certain amount of standardisation introduced, the geometry of every joint was different.

The chord members are built up from 7mm-thick lamination of timber which Cowley Timberwork put together and glued, then placed in moulds to achieve the desired curvature. In total there are 24 curves, but Cowley managed to create them with only 12 moulds, re-profiling each mould to suit a second curve. Originally the curved braces were to be created in the same way, but Cowley suggested that it would be less time consuming to build them up from just three flat 20mm laminates, and to cut out the curved shapes. The line of the grain runs with the main body of braces, and the curved ends, where the grain alignment is not ideal, are reinforced by the steel flitches. A similar method was used to make the straight ties, which were then cut to be circular in section, with square ends.

All the timber, supplied by a number of donors, was of a high structural grade, but Cowley carried out further selection, ensuring that the outer laminations were of the very best quality, since these had to bear the greatest loads. It carried out extensive testing of joints and of the timber itself, to confirm that all the calculations were right, and in fact it performed even better than anticipated, leading Levete to say, ‘We could even have been a little bit more daring’.

But this design was probably daring enough. Erection was originally planned to be from cherry pickers, starting at one end and then connecting each of the nine sub-assemblies as it went up. But in practice this proved not to be feasible, because it was not possible to achieve the very accurate alignments that were needed in order to join the pieces. Scaffolding had to be brought in, and the museum entrance remained closed for longer than had been anticipated. Speaking at the official opening (the structure was completed just in time) Levete said ‘To be inconvenienced by design is really important sometimes’. For the public, as for the design team, the stress and inconvenience were justified by the result.