* DESIGN OF
THE YEAR 2018

This pavilion was born out of an investigation that began in 2013. Your team won an ideas competition for a temporary pavilion in Dhoby Ghaut Green. Tell us more about this genesis and the ideas your team was exploring.

Thomas Schroepfer (TS): In 2013, the Singapore Institute of Architects (SIA) ran an ideas competition that focused on parametric design. It is a technique that has been around for a while. For instance, the early projects of Frank Gehry, such as the Walt Disney Concert Hall, exploit the complex geometry that parametric computational modelling allows. What this means is rather than drawing a building, like in the old days, you define a project through computational code that specifies the relationships of one part to another. The whole project can grow, shrink and the geometry can change, but the code and relationships are fixed.

Our competition entry was defined by a computational model that is based on certain parameters we were interested in specific to Dhoby Ghaut Green. We wanted a structure that opens up and becomes denser in relation to sun path, wind direction and views from inside the structure. With a computational model, you can feed in such data about the environment and it would adjust the design of the structure accordingly. It is a highly customized approach where architecture is not a one-size-fits all. Instead, we can create a structure that responds to things that are always there, such as the sun or even shade provided by existing trees.

These ideas remained when we were asked to design a pavilion for The Future of Us exhibition; it is about creating a comfortable experience for an outdoor public space without relying on air-conditioning.

The pavilion is more than meets the eye. Each of the intricately perforated aluminium panels is unique and they come together to effectively shade visitors. Was this integration of form and function important for your team?

TS: The basic unit of our computational model is a triangle because it is a shape that allows for complex geometries and provides a good structural system. As we feed the algorithm information about the different parameters, we go from a triangle that is almost completely open, to one that gets incredibly dense inside. The design may look very irregular, but it is based on a very clear mathematical logic.

We were also very interested in the visual language that emerges from this model. We describe it as “foliage”—it is like being shaded by the beautiful trees along Orchard Road. The trees make it comfortable to walk on the street, and you also get a beautiful play of light and shadow. So, there is an aesthetic dimension to our model, but it is every much driven by a mathematical precision to 6  work out things like environmental data and structural information.

The computational model also considers the optimum amount of material needed to construct the design. Can you tell us more?

TS: Shells perform very differently depending on the geometry they have. We created many versions and simulated many geometries to get the optimal design. You can build a shell that is not optimised, but it is going to be much thicker. Our pavilion shell is only about 20-centimetres thick, which immediately translates to how much material we used, the weight of the design, etc. These address ideas of sustainability. By optimizing the design in this way, we made it much cheaper and more effective to build.

Despite its geometric complexity, this pavilion was designed and built in less than six months. How was this achieved?

TS: When the curators of The Future of Us asked if we could do a pavilion many times the size of what we had proposed for Dhoby Ghaut Green, we said, sure. The parametric model allows not just a change in form but also, to scale up and down. In the conventional model, there is no way you can do this in such a short time and in an entirely different location.

Each panel of our pavilion is constructed from two triangles folded on the edges and then manually put together. Our model is so precise that the top and bottom of every panel are different to further fine-tune their environmental performance. Everything can be different because in the digital environment, the machine does not care. Once we have defined the mathematical relationship, the computer subdivides the design, organises the pattern and so on. We don’t draw every panel by ourselves. Working with a digital model also meant architects and structural engineers were working on the same model and we could feed information directly into the machines cutting the panels. This allowed us to save time.

The pavilion we created was eventually reduced in size not because of the design but simply because production could not keep up. As we don’t yet have a 3D-printer which is large enough to print the pavilion, we had to find a way to make it out of sheet material. We chose to laser-cut triangles out of aluminium sheets; an algorithm was also developed to pack these on the rectangular sheets to generate the least material waste when cut out.

The pavilion was built by workers who had no special training. We provided the contractor with a map of which panel goes where—it is essentially a gigantic puzzle. The panels were all shipped to the site on time and there was no need for storage. There was no sophisticated technology needed for construction. Even the contractor was surprised at how quickly they could put this together.

The pavilion is but one outcome of the various digital tools you have developed. What other possibilities do they hold for the future of architecture, particularly in Singapore?

TS: When tools such as parametric computational modelling first became available, there was a lot of formal experimentation. There was often no performative aspect to it beyond being beautiful. We have since moved on. We can now generate beautiful forms, but we can also make them meaningful. You can put material exactly at the form and quantity that makes the most sense to perform best in the environment.

The true power of this technology is that everything can be customised easily, and it makes no difference. You can design a building quicker than a traditional approach where everything is standardised. This is also the main message: there is absolutely no reason to have buildings that are the same anymore. Our tools allow for very complex geometries so there is no need to have buildings that are all rectangular. There are a lot of reasons why people work with rectangular shapes, but for a project like the pavilion that sits in a landscape— that can be a driver for architecture. Look at the Supertrees or Flower Dome in Gardens by the Bay— their formal language is so different. As Singapore pushes to be a City in a Garden, it should not stop at the level of integrating greenery. Buildings can become landscapes too. This fusion of landscape and architecture is powerful for Singapore because of the climate. Just as airplanes are shaped in a certain way to direct the wind so they can fly, thecity can be shaped in a way to provide a comfortable environment too.