Beetle drive

Installed in the V&A’s John Madejski Garden, the centrepiece of the museum’s Engineering Season, is a vision of how we might build in the future. The Elytra Filament Pavilion, built only from glass fibre and carbon fibre strands, is inspired by the wing shells of a specific type of flying beetle known as Elytra.

The structure as you see it is also a form in transition. Over the six months the installation remains on site, sensors in the canopy will relay real-time data on user movements beneath to the fabrication software, which will make decisions on where the next hexagonal canopy component will be installed, and, as a result, what its consequential loads and stresses will be.

Below the canopy, all this algorithmic information is being relayed to a pair of Kuka robots, whose task it is to make the new component in full view of visitors, ready for installation above their heads. This could be seen as merely whimsical but, according to its designers, it presages a well overdue re-evaluation of the construction industry, acting as a prototype to herald the fourth industrial revolution.

The Elytra Pavilion’s design and fabrication methodology is the creation of German architect Achim Menges with Moritz Dörstelmann, structural engineer Jan Knippers and climate engineer Thomas Auer through the University of Stuttgart’s Institute of Computational Design (ICD) and Institute of Building Structures and Structural Design (ITKE). It is the result of four years’ research by the two institutes into developing computational design and fabrication techniques for the industry – innovations celebrated annually with a new temporary pavilion on the university campus.

If the idea of academic computational design research summons up visions of bespectacled geekiness, a look at the ICD’s website might help dispel the assumption. The short videos showcasing each pavilion design boast high production values, snappy overlaid graphics and electro beats. This latest iteration in London is highly influenced by the 2013 pavilion, which, though ‘less highly calibrated’ in terms of fabrication techniques and with no sensory modification to the structure, as presented here, looked at utilising the properties of glass and carbon fibre to produce extremely lightweight and robust structures with the potential for large spans. Any structure weighing in at less than 50kg/m² can be considered ‘lightweight’; their 2014 pavilion came in at 4kg/m².

Here, each of the 40 unique 5m² hexagonal components, spread over seven supporting columns, weighs less than 45kg; so at 9kg/m², the whole 200m² pavilion will weigh less than 2.5 tonnes. As Achim Menges is keen to point out, that’s less than the weight of a 2m² section of the museum’s wall.

With the design of the Elytra Pavilion Menges thinks they’re at the cutting edge of investigating possibilities for new architectures and construction through biomimetics. ‘There’s two methodologies at play – the technological pull and the biological push,’ he explains. ‘We tend to concentrate on the latter, looking for a structural solution that has a biological comparator, such as Elytra shells, and then investigating if there are fabrication processes that might align with it to generate their form.’ But he admits it’s a rarefied world, demanding literacy beyond the usual skills of the architect.

‘When your design experiments with the exoskeleton of arthropods or shells of American lobsters, you need to work with not only engineers but biologists, palaeontologists and materials scientists – and even then there’s no guarantee it’ll work,’ he adds. But underlying it all is the ambition to create building systems using new design thinking and methods.

Specialist input

The design team worked with biologists from the University of Tübingen to ascertain the structural principles of the double-layer Elytra shell. This involved micro-tomography of the beetle shells and computational morphological analysis, including commissioning a particle accelerator and scanners to produce a 3D model of the Elytra structure to a resolution of 3-6 microns. Once the engineers and designers had formulated a methodology of mimicking the double-layer structure, they looked at fabrication techniques that would allow the rationalised hexagonal components to be individually spun using robot ‘effectors’.

Here, it uses a robot and an external positioner, linked to it via control software. The positioner has a steel framework attached to it, which acts as a template.

The robot feeds the resin-soaked glass fibre strands on to this framework, and which acts as a scaffold for the carbon fibre resin strands that follow. The computer controlled, double-skin winding method has been designed to run from one layer to the other as a continuous strand, to harness the material properties of the carbon fibre and give it strength as a ‘woven’ structural component once the resin has hardened. In accordance with its potential position and associated stress loadings, each component will be individually spun (which will take around three hours) and be unique to its place in the final matrix.

The funnel column forms that transfer the loads of the whole canopy structure to its seven steel posts are formed using an additional framing module connected to the robots.

As the installation is open to the public, components will be assembled manually despite a robot making them, but the intention is that this too will become an automated process. To help realise the paradigm shift in construction they’re seeking, Menges and his team still look nostalgically to Paxton’s Crystal Palace for precedents for scaling-up the pavilion construction methodology to the size of stadia. Fabrication logistics is everything.

‘We want to combine current construction technology, like cranes, with industrial robots to generate an integrated fabrication and construction process,’ says Menges. They are helped by the fact that 1km of carbon fibre weighs only 10kg, allowing them to employ novel techniques. Back in Stuttgart, they are working on a new pavilion using unmanned aerial vehicles as a material delivery system across larger spans. While the team is using commercially available materials such as glass and carbon fibre now, they’re liaising with other research bodies to source mineral fibres such as basalt and even graphene to ascertain their structural properties. The aim is to move away from hydro-carbon based to bio-resins – hardening agents that are more eco-friendly and, crucially, fire resistant.

Parametric or not?

One might assume this form generation is the product of parametric thinking, but Menges says it’s more complex than that. The iterative procedure, with designs born of natural forms interrogated in the real world for viable construction methods, has resulted in an almost random quality to this series of pavilions over the years. Some might see it as a thinking disconnect, but Menges is reassured by the fact that different problems are generating wholly different solutions unencumbered by ‘style’.

‘All the pavilions we’ve done look different because we’ve never started with an aesthetic endeavour – we can only evaluate the forms after we’ve realised the project,’ he explains. ‘While some of our ideas might align with para­metricism, where we diverge from it is that we don’t end up with an architectural style.

We’re governed by emerging technologies, structural or user behaviour, so are not looking at a predetermined architectural language.’
Despite the Elytra Pavilion’s emphasis on computerised design and fabrication, Menges realises that it’s one step at a time. There’s no replacement of the artisan with the machine, for instance, but ‘a synergy between the two’; and they've been working with Autodesk on 'Hive', an installation exploring human and robotic collaboration. Menges concedes the step change he wants may be a while off. ‘For now, the only place to carry out this kind of integrated research is academia, as professional practices are not yet geared up for such speculative enquiry,’ he muses. He observes that after steel reinforced concrete technology was first patented in 1860 it took 50 years for it to be adopted by the industry – ‘There’s a time lag for all innovations to make their way into the mainstream.’ Watching his robot spinning away purposefully in the courtyard of the V&A, the paradigm shift could be sooner than we think.