I’ve been interested in textile structures, tensile or inflatable, ever since 1973, the year I graduated as an architect. That year, a friend from my graduating class, Francesc Albardané, translated a book about Frei Otto into Spanish. We were fascinated by the work of this German architect, and became immediately passionate about lightweight structures. In 1978 Albardané and I had the chance to work together on the design for an inflatable roof for a swimming pool in Sabadell, Spain. It was then that I understood the gaps in my technical knowledge, and I decided to pursue a doctorate in this kind of construction. While working on my thesis, which I submitted in 1981, I developed the first version of the Tess computation software, which, with the advent of Windows a few years later, became WinTess. Ever since, I have continued my research into the development of software for calculating tensile and inflatable architecture.

Bit by bit I have specialised in the design and calculation of these structures, which to start with were unusual but which have become more and more commonplace. At first I worked as an architect, designing and building my own projects. Over the years, other architects have sought me out as a consultant. WinTess, initially designed for private use, has become a piece of commercial software, so my work as a consultant has gradually taken over.

Concurrently, I teach at the architecture school at the Universitat Politècnica de Catalunya. It is important for students to learn about these structures because they are the source of so many pieces of playful, contemporary architecture for sports or industry. We also dedicate one semester each year to lightweight structures. Few architecture schools offer this programme and I am regularly invited to teach or lead workshops at other architecture schools in Europe and the Americas.

The Bouncing Bridge

TP Arquitectura i Construcció Tèxtil, a Spanish construction company specialising in textile and inflatable structures that I have known for many years, came to see me one day to show me the Bouncing Bridge project. They had come across it on the Internet and had contacted AZC to propose making it. The architect in me was immediately seduced by the bridge’s design and I was sure it would spark a lot of interest. On the other hand, as an inflatables specialist I saw the incredible challenge such a project would pose if realised for everyday use. I had never built anything on water – it’s not an architect’s usual environment! Even less so when it’s moving water. So first of all I concentrated on static studies, saving the dynamic studies for a later stage.

The project has two key elements: firstly the ring shape, which constitutes both structure and foundations, and secondly the stretched membrane, a secondary structural element but essential to the function of the bridge. It quickly became clear that a certain number of questions could only be resolved empirically. For financial reasons and to improve analysis of the results, we made a first prototype at a scale of 1:10, approximately three metres in diameter, which we wanted to test for structural behaviour, stability on the water and feasibility of construction.

The shape is not created by simply twisting a ring, but, similar to a sewing pattern, by joining lots of different pieces that together will generate a curved, three-dimensional form. Working on a small-scale prototype created a considerable amount of extra work for us when devising the pattern. The number of pieces had to be the same as for the full-scale bridge, so that we would be able to judge the smoothness of the curves. Geometric design was very slow because we had not yet developed the necessary software. We did, however, finally manage to build the first prototype and to test it. First of all in dry conditions, testing it with different loads and under different pressures. We had to take into account the proportional relationship between the 1:10 prototype and the full-scale object, so that we could analyse manufacturing details for the object (behaviour of the materials, dimensions, joints) and the reactions to the tests (how much it would move when in use). The model’s behaviour lived up to our hopes, as WinTess had predicted. Next we tested its stability on water in a domestic swimming pool. We were satisfied with the results, but the size of the prototype did not allow us to test the impact that bouncing on the trampoline membranes would have on the bridge.

This was why we built a second prototype, this time at 1:3, measuring 10m in diameter, which would allow us to test the project’s behaviour on water but also to observe how the trampoline would work under real conditions. A swimming pool would no longer suffice, so we took the model to Banyoles Lake. The tests on the water, with several people jumping on the trampolines at the same time, were a success.

After this we were very keen to build the project full scale, 30 metres in diameter. But we also needed to evaluate the tests we’d carried out in order to improve our method. Principally in pattern cutting – I wanted to be able to put together this kind of tubular structure more efficiently. And in fact, a new project came along that allowed us to do just that!

The Peace Pavilion

The Peace Pavilion is a smaller scale project, with limited civil liability but of a more complicated design. Its beauty stems from the different perspectives afforded by the twists of the ring, a fairly complicated 3D structure.

Thanks to our experience with the Bouncing Bridge, we did not need to build a prototype, because we already knew that the principal difficulty would be in making the tube, more specifically designing the pattern. The curves for this project were more complex than those of the rings for the bridge.

Designing the pattern manually with the help of WinTess, as we had done for the bridge, would have been very laborious considering the number of pieces required by the structure. Each change to the design would have meant redesigning all the pieces. So I created a new module for the software that was able to define the tubes in 3D and to automatically set out the pattern. This was a lengthy task but the result was spectacular: we were able to create various patterns for this type of tube in seconds! This allowed us to adapt the shape of the pieces in order to optimise the amount of material used and to precisely define the curves of the final object.

Unlike the bridge, the pavilion’s structure would not have to withstand significant dynamic loads, but just carry a simple, waterproof, transparent canopy. However, it is far from straightforward designing a canopy for such a complex three-dimensional structure, requiring parametric structural modelling software rather than geometric. We were therefore able to use WinTess.

The result entirely lived up to our expectations: the images speak for themselves.

And next?

For the Peace Pavilion, the next challenge would be to design a more permanent structure. For a temporary event we were able to ignore the effects of snow or wind, the characteristics of the ground or the durability of the project. But for a permanent structure, all these considerations would be primordial. We look forward to an opportunity to prove that this is possible.

In the case of the Bouncing Bridge, we have not yet realised AZC’s project for three 30m-diameter inflatable rings. Moreover, if the project needed to be permanent or semi-permanent, its structural complexity would be greater and we would have to study the implications of different locations, a calm or fast-flowing river, a lake or a canal. However, I am certain that we could successfully develop this trampoline bridge.