With an exponential increase in the possibilities of computation and computer-controlled fabrication, architecture is now facing a novel challenge. As architects, we can design for infinite resolution and density of information, controlling the deposition of millions of material particles. Yet there is a lack of design methodology to harness the full potential of computational power and computer-controlled fabrication. When compared to historic architecture, for example the Gothic, computational experimentation is still not able to successfully generate the same level of fine detail, hierarchy and information-density.
Bartlett School of Architecture’s RC4 in London researches computational design methodologies for large-scale 3D printing with industrial robots, taking logistical, structural and material constraints as design opportunities to generate non-representational architectural spaces with extreme information density. The research cluster investigates tectonic problems associated with large scale 3D printing and makes them inherent to the design. Computational models develop as strategies to organise material in space in response to specific structural and logistical input. All computing is targeted at directly generating and optimizing data for the robotic additive process, bypassing other steps such as meshes, contour generation or slicing. The computed forms are in reality direct information for the robot about where to deposit or bind material.
Computational methodologies are continuously tested on architectural chunks, to investigate and test the spatial effects of the proposed fabrication methodology. Students are asked to situate their research in relation to fundamental structural paradigms and strategies such as space frames, lattices or vaults to then subsequently challenge the existing model. The final proposals range from highly differentiated, heterogeneous space frames to finely articulated compression-optimized domes.
Alongside architectural speculation and computational methodologies, students also successfully constructed their own tools, tweaking the industrial robots to become large-scale 3d printers. The research investigated a variety of large scale robotic printing methods and design strategies, such as vectorial plastic extrusion and 3D-printed sandstone in combination with cast aluminum. Other research investigated ceramics, engaging material properties of clay such as hanging arches and strands to generate both ornament and structural efficiency, and robotic weaving of textile substrates for concrete spraying. The research cluster aims to redefine architectural construction methods, challenging not only the use of technology but also the physical and material logic embedded in architectural tectonics.
The project SpaceWires by Filamentrics ( Zeeshan Ahmed, Justin Yichao Chen, Nan Jiang,Yiwei Wang) intends to overcome the formal limitations of space frames, avoiding its homogeneity by introducing robotic plastic extrusion. Printed structures with a high degree of fine resolution and complex hierarchy can be achieved on a large scale within a relatively short period of time.
The use of custom design and fabrication algorithms makes it possible to amplify efficiency, where material capabilities such as porosity or connectivity are constantly linked to structural behaviour. The use of a recursive plastic extrusion is directly controlled from the custom made design application, which relates structural data together to a variety of design parameters, allowing the continuous generation of tool paths for to the robot. The project looks specifically at a combination of spaceframe-like triangulations and more continuous, sweeping curves. These two formal qualities correspond to two different fabrication methods with the robot, which allow material to harden in the air.
The Pixelstone project by Microstrata ( Maho Akita , Fame Ornruja Boonyasit, Syazwan Rusdi, Wonil Son) is also investigating the dichotomy between additive manufacturing fabrication and generative design, trying to establish a single bottom-up process which would redefine in particular how reinforced concrete is approached by the building industry. For logistical reasons sandstone is taken as a substitute of concrete as a compression based material, as both are powder based. A layer based process generates large pieces of porous sandstone which are later assembled in architectural scale prototypes. Initially the computational process starts out as an abstract network of relations between material in compression and tension.
Once this network is established a Cellular Automata voxelizes the network, every voxel represents a drop of binding material to be deposited by the robotic unit. To reinforce the sand stone in tension, a capillary network of veins and channels is left empty within the mass. Afterwards, Aluminium is cast in this hollow space; establishing a continuous tension reinforcement.
Find out more about RC4 and about other master level courses at the Bartlett School of Architecture by visiting the links below.
Bartlett RC4 Studio Masters: Manuel Jimenez Garcia & Gilles Retsin
Technical Support: Vicente Soler, Thibault Schwartz
Filamentrics: Nan Jiang, Yiwei Wang, Zeeshan Ahmed, Yichao Chen
Microstrata: Wonil Son, FaFame Boonyasit, Maho Akita & Syazwan Rusdi
WeaveShake: Kehan Zhang, Xin Wen & Yihuan Xie
Clayware: Wei Liu, Zuan Huang & Wenzhao Gao
The Bartlett GAD Director: Alisa Andrasek
B-Pro Directors: Professor Frédéric Migayrou (B-Pro Director) and Andrew Porter (B-Pro Deputy Director)
RC4 Exhibition – September 2014