I'll huff and I'll puff and I'll blow your bridge up

New Civil Engineer reports that Swiss architects have created an inflatable bridge. Philipp Dohmen and Oscar Zieta have built a 6m long steel-skin balloon that can hold a load of up to 1800kg, according to load test at the Swiss ETH institute.

It's all very exciting, but not really news as such - the ETH's department of Computer Aided Architectural Design (CAAD) has been working on this system for some time, with their so-called FIDU-Brücke having been load-tested in December 2007 (FIDU stands for FreieInnenDruckUmformung, or internal pressure forming). Essentially, it's a prestressed steel shell structure, where the internal pressure prestresses the skin (maintaining it in tension even when subject to applied compressive stress) and hence prevents it from buckling, even in the absence of stiffeners.

Although Zieta's website shows lots of clever applications in product design, including furniture, it's hard to see how any inflatable structure can be robust enough to act as a real bridge, with the danger that any escape of air (e.g. due to failure of a seam or accidental puncture) would immediately eliminate all the structural resistance. The advantages, of course, include the extremely light weight of the structure, which also allows it to be readily deployable in difficult locations.

Nor is ETH-CAAD the first to develop an inflatable bridge.

The US Army has a deployable air-inflated system which can carry military vehicles weighing as much as 80 tons. This however, is really a causeway system rather than a bridge as such: unlike the FIDU-Brücke, it isn't subject to bending.

The University of Maine has also used inflatable plastic arches in bridge construction, although these are filled with concrete after inflation.

More pertinently, the "tensairity" system promoted by Airlight has been proposed for a number of bridges, and applied to some. I wrote about this previously because an Airlight bridge was one of the entries to the Leamouth Bridge Design Competition (pictured left).

The tensairity system uses inflatable plastic beams shaped a little like cigars. These are inflated under relatively low pressure, and the system relies on the use of struts and cables fixed to the inflatable membrane to carry most of the load: the balloon only carries shear and compression between these, and stabilises the whole system against buckling.

An 8m span test bridge was built and shown to be capable of carrying a 3500kg test load [PDF], easily putting the Dohmen / Zieta span in the shade, and with a structure lighter in weight as well. Unlike the metal-skin solution, a further advantage of the Tensairity beams is their translucency, allowing them to be lit internally for architectural effect.

The Airlight website illustrates a number of projects for which tensairity beam bridges have been proposed, but not built. As well as Leamouth, these include footbridges in France and Switzerland, such as the one at Giubiasco shown on the right. They have also been proposed for temporary use to support construction vehicles.

The most significant bridge to be completed is at Val Cenis ski resort in France, a 52m footbridge for skiers (also carrying substantial snow load) - pictured below. This is essentially a timber frame bridge with steel bars providing the suspension cable below - the two elements are separated by the Airlight balloons which hold the main structural pieces in place. It was designed by Charpente Concept, who provide further details of the design and construction on their website.

See also:

• Student tensairity bridge project at ETH
• Greenline blog discusses the tensairity Tubaloon

Tensegrity bridges: 5. Miscellaneous

So, tensegrity bridges. The Rome, Washington and Brisbane bridges are the major examples I've found, but there are one or two others to briefly mop up.

Bankside Bridge
This was a proposal for a crossing of the Thames as part of the Millennium Bridge design competition. The engineer was Mott MacDonald, I've been unable to confirm the architect. I have no information other than these two images:




St Petersburg
As part of a redevelopment proposal for the Apraksin Dvor area of St Petersburg, Wilkinson Eyre have proposed both a tensegrity roof, and a footbridge over the River Fotanka supported by a tensegrity "cloud". Again, no real details other than this picture:



Deployable bridge
This forms part of a thesis under preparation by Landolf-Rhode Barbarigos, who has a blog devoted to tensegrity structures. It incorporates "active" members (e.g. pulleys or telescopic bars) which help it find the right shape as it is deployed.




Bamboo Bridge
Designed by architect Michael McDonough, this is an unbuilt proposal for a 133m span bamboo truss bridge in Mendocino, California, described as using tensegrity principles. From the pictures, it's far from clear that it's a tensegrity bridge at all, looking more like a modularised cable-truss bridge incorporating some additional cable elements for stiffening.






Tube Bridge
Andreas Kirchsteiger has come up with a concept for a tubular bridge comprising rings connected by woven fabric (reminiscent of Cecil Balmond's Weave Bridge, perhaps). The support struts connect to the rings, but otherwise it's closer to the pure tensegrity idea than some designs. I can't really see how it would work - how are the rings stressed together, and how is overall bending actually carried?

Footbridge using "Simplex" modules
This was proposed in a thesis by Valentín Gómez Jáuregui, and is intended for lightweight short-span modular footbridges.




So, that's it for this series of posts on tensegrity bridges. How can I summarise? Firstly, it seems likely that a structure complying strictly with tensegrity principles (no two touching struts) is unlikely ever to be suitable for a real footbridge. Secondly, that tensegrity can be taken either to inspire a fairly orderly and visually conventional cable truss structure, or used as the jumping off point for chaotic visual complexity. This shows how central is the designer's choice in adapting any structural system.

Over half-a-century after it was first conceived, tensegrity still throws down a huge challenge to the structural designer - conventional habits of design, both in terms of structural analysis and visual order, simply aren't sufficient.

Tensegrity bridges: 4. Kurilpa Bridge, Brisbane

Kurilpa Bridge, currently under construction in Brisbane, Australia, is one of the most audacious and technically ambitious bridge designs of recent years. This AUS$63m footbridge over the Brisbane River (pictured right) is due to open this September, nearly two years after construction began. When it does open, it should be acknowledged as a unique urban landmark up there with the London Millennium Bridge, Gateshead Millennium Bridge, or Alamillo Bridge.

Designed by Cox Rayner Architects with the engineers Arup, the Kurilpa Bridge is being touted as the world's first tensegrity bridge to be built. Other tensegrity bridge proposals have yet to get beyond the planning stage. While acknowledging the Kurilpa Bridge's uniqueness, I'll come back shortly to whether it is, in fact, a tensegrity bridge.

The bridge was the winner in a design-and-build competition, and it's always pleasing to see something radical to result from D&B, which usually gets a bad press for its inability to generate high-quality design. The structure is being built by Baulderstone Hornibrook contractors and Beenleigh Steel Fabricators (construction photo shown, right), amongst others, and it's impossible not to be impressed at the bravery of anyone taking on a project like this.

The span connects Tank Street in the city centre to Kurilpa Point in South Brisbane, and establishes a cycle route loop through the city, which also passes over the Goodwill Bridge, another Arup and Cox Rayner design.

The huge contrast between the two bridges comes both in style and cost. Goodwill Bridge's slender arch is rational, efficient, harmonious - and relatively inexpensive. Built for AUS$20m, I estimate it cost about £3,500 per square metre of deck, which is extremely good value for a large-span landmark footbridge (compare for example, Calatrava's Sundial Bridge which was about £10,000 per square metre; the London Millennium Bridge, about £12,000 per square metre, even before modification; or Gateshead which came in at about £22,000; I'm ignoring inflation in all these figures [and using today's exchange rates]).

In comparison, Kurilpa Bridge is irrational, visually chaotic, disruptive, and possibly one of the most expensive fixed footbridges ever built. Guessing it to be twice as long overall as its 128m main span, and of similar width to its cousin, I reckon it cost £19,000 per square metre. This may be less than Gateshead, but Kurilpa doesn't sit up and do tricks.

Some readers of this blog will be shaking their heads by now. Yes, I know, it's the value of a bridge that matters, not its cost. Judgements will differ on whether Kurilpa Bridge is a beauty or a blot, but it's undeniably a landmark, an innovation and a substantial technical accomplishment. I guess only the locals can judge whether the value is commensurate with the impact on their tax dollars.

So, is it so expensive because it's a tensegrity structure, a form inherently difficult to build and not especially efficient for this sort of span? Well, its marina-full of spars and cables has the visual complexity of a tensegrity structure and appears to be based on tensegrity geometries, but it's not strictly a tensegrity bridge at all. The bridge deck is a continuous member carrying both bending and axial compression, and is stiffened laterally with what appears to be conventional bracing. There are several locations where struts interconnect (that tensegritarian sin), most obviously over the piers, which support the conjunction of two mast struts as well as the compression strut of the deck. So, it's tensegrity-ish, but not tensegrity-proper.

In fact, it's a complex variant on the good old cable-stayed bridge, with a substantial dash of the inverted fink truss (as at Forthside or Royal Victoria Dock) thrown in. This seems most obvious from the various construction photographs (several at Wikipedia, for example, as well as the one shown here), showing cantilevered construction using the conventional cable-stayed principle. So the astronomical cost, I assume, can only be the result of taking a very economic form of bridge and doing as much as is possible to eliminate its advantageously simple regularity and buildability.

I admire rather than enjoy its aesthetics, although I can see that the bridge may feel different in real life than to the photos and visualisations included here. The seemingly random cable and strut angles provide little in the way of reference for viewers, there seems to be a conflict between the large scale of the structure and the feeling of visual instability that might be present. Like many bridge engineers, I like a bridge where the structural principles are clear and comprehensible, which is unlikely to be the case for anything tensegrity-ish. But I do admire its audacity, the willingness to install something that works against the orthogonality of its surroundings, a provocation which offers restlessness in place of reasurance.

Further information:

• on Wikipedia
• on Structurae
• at Queensland Department of Public Works

Tensegrity bridges: 3. National Building Museum, Washington

While the TorVergata footbridge was an example of a very well-ordered application of tensegrity to bridge design, the proposed footbridge at the National Building Museum, Washington DC, is something else entirely.

Designed by Wilkinson Eyre and Arup, the proposal is to span 35m across the Museum's huge Great Hall, connecting galleries at a high level. I think it was first exhibited at the Venice Architectural Biennale in 2004, and I'm far from clear whether the design is still being developed, or stands any chance of ever being built.

Like Wilkinson Eyre's previous Challenge of Materials footbridge, a glass-floored span in London's Science Museum, the Washington bridge is intended to be an exhibit as much as a practical crossing. The struts are proposed to be glass tubes, with internal lights and sounds that respond to changes in load as people cross the bridge.

Quite how practical any of that might be is hard to judge (the detail of the joint between the tubes and cables would seem to be the toughest challenge). So instead, I will comment only on the general layout and the appearance of what, by any standards, is an extremely unusual structure.

Chris Wilkinson, one of the two founders of Wilkinson Eyre, has written [PDF] of how the firm strives for "lightness" in design, both in terms of weight and visual impact:

"As architects, we are attracted to tensegrity structures for their visual lightness and their efficiency. They offer the maximum strength for a given amount of material, which keeps the member sizes slender and light. This is particularly relevant to bridges, where long spans can be achieved with slender suspension structures, such as our Metsovitikos Bridge in northern Greece. Cable structures have movement and life, which adds to their appeal. When a bridge structure moves in response to your weight as you cross it, you know that it has been designed for efficiency. A certain amount of movement in structures is generally a good thing, so long as it is controlled within defined limits."

Structural efficiency is a difficult concept. It's far from clear that a tensegrity footbridge really does use less material than a more conventional design, but efficiency of material is certainly not matched by efficiency in construction. The Washington design would be exceptionally difficult to assemble, requiring substantial temporary support. To make it sufficiently rigid, it may impose significant loads on the existing building structure.

To me, the structure's visual lightness is somewhat doubtful: although it is developed from a series of simple tensegrity cells, the end result looks like a jumble of scaffolding freeze-framed in the act of collapse. While that offers an intriguing challenge to conventional ideas of structural elegance, it seems to be me to be visually very "busy" and hence at odds with the stated aspiration - it may be 'light', but it isn't 'quiet'. Indeed, something of the challenge involved visually is given in another Wilkinson Eyre quote:

"The underlying geometry is based on a series of tetrahedral cells, replicated numerous times to accumulate a visual mass capable of asserting itself within the extraordinary scale of the museum's courts while remaining essentially light."

This fundamental tension between the desire to assemble cloud-like mass and simultaneously to dematerialise the structure has the potential to result in an exceptionally interesting and unusual experience. I can't bring myself to like the design, visually it's just not an aesthetic I admire, but I do admire its ambition and radicalism.

Further information:

• "Tensegrity bridge" at Wilkinson Eyre website [PDF]

Tensegrity bridges: 2. Passerella TorVergata, Rome

So, with an introduction to tensegrity out of the way, let's see who would actually be crazy enough to try and design such a bridge.

First up is a proposal for a 32m span footbridge at the Tor Vergata University in Rome (pictured, left). I first read about the proposed design in a paper by Andrea Micheletti (and others) presented at the Footbridge 2005 conference (abstract available online at "The tensegrity footbridge at TorVergata University in Rome" [PDF]).

The bridge is planned to cross a highway near to the University's Faculty of Engineering, and the design has been developed by the tensegrity systems group there. They chose a tensegrity design for its "sense of transparency and lightness", and their design takes a modular approach, connecting together a row of five "expanded octahedra", each of which is an internally self-stable tensegrity structure (pictured, right).

In combining the modules (see image left, showing two modules added together), much of the simplicity of the basic element has to be eliminated. If you compare the two pictures closely, you'll see that several additional cables have been added to provide adequate rigidity in all directions, and also that where the modules join, two struts come together at a common node. A strict tensegritarian, if there were such a thing, would regard this as a no-no, but it's a common and perhaps unavoidable fudge in tensegrity bridge designs, acknowledging that the basic principle really isn't well suited to this application. More information on the design development is available online in "Una passerella pedonale tensintegra per il campus di TorVergata" [Powerpoint].

Form finding techniques have been used to optimise the geometry of the final design (see image on the right). Preliminary engineering studies have shown the design not to be susceptible to vibration, and to weigh about the same as a conventional footbridge design for the same span.

The bridge is proposed to be built by assembling the modules off-site, folding them down for transportation, assembling and prestressing on site, and finally erection by mobile crane. This is not in itself significantly more onerous than a conventional bridge, with the exception of the need for prestressing, which has the potential to be complex.

Visually, the adoption of a modular approach eliminates much of the tangled chaos that is common in tensegrity structures - the bridge seems well-ordered, visually comprehensible, and not entirely unfamiliar. That makes me wonder why you would opt for tensegrity at all, if a similar truss bridge structure could be provided more economically by conventional means, while offering greater freedom to be shaped by the designer. The question in this instance is whether the real transparency attained justifies the effort.

While this bridge seems very unlikely to be built, it's not entirely impractical, and particularly nice to see academia seeking to bring the abstruse and exotic into the real world.

Further information:

• "Modular Tensegrity Systems: The TorVergata Footbridge", Andrea Micheletti [PDF]
• "Una Passerella Tensintegra nel Campus di TorVergata", conference presentation [PPT]
• "Una Passerella Tensintegra nel Campus di TorVergata", Livio Ponzi [thesis in Italian, PDF]

Tensegrity bridges: 1. Introduction

This September, the world's first bridge of any significance which claims to be designed on the tensegrity principle will open in Brisbane, Australia (Kurilpa Bridge, pictured right). While drafting a posting on the bridge, it became obvious that some other proposed tensegrity bridges merit more than passing attention.

So over the next few posts, I'll be covering as many tensegrity bridges, built, unbuilt or never-really-intended-to-be-built, as I can. With that in mind, I guess some sort of introduction to the concept of tensegrity itself might be helpful!

It dates to the work in 1948/9 of American futurist Buckminster Fuller, and artist Kenneth Snelson. The term "tensegrity" (tensional + integrity) was coined by Fuller, who in his 1975 book Synergetics defined it as follows:

"Tensegrity describes a structural-relationship principle in which structural shape is guaranteed by the finitely closed, comprehensively continuous, tensional behaviors of the system and not by the discontinuous and exclusively local compressional member behaviors ... The integrity of the whole structure is invested in the finitely closed, tensional-embracement network, and the compressions are local islands."

Got that? As applied to structural design, it's normally taken to refer to a space-truss system comprising elements only in axial tension and compression, where no two compression struts are in contact. By maximising the use of slender tension members to provide the system's rigidity, tensegrity structures are seen by promoters of the concept as minimising material use overall, as well as offering visual transparency.

Perhaps the best known tensegrity structure to be built was the Skylon, erected in London in 1951 for the Festival of Britain. It's not a "classical" tensegrity form, as there are compression "elements" in contact with each other, where the support elements meet the ground. An example of a structure which fully satisfies the tensegrity principle (i.e. is self-contained) is the Skwish (pictured left), a child's toy constructed from wooden struts and elastic ties.

Most tensegrity structures resemble, at their best, an elegant cat's cradle of wires and struts. At their worst, they are more like a tangle of epileptic scaffolding. Their inherent flexibility, lack of robustness and visual non-linearity don't seem to lend itself to bridge structures, which are by their nature normally very linear in form.

So it will be interesting to see how various designers have risen to this challenge in designing tensegrity bridges ...