Fall 2021
Gabriel Lipkowitz
Instructor Glenn Katz
Modular architecture has a relatively long and well established history, its attractiveness lying in the fact that prefabricated components can be readily transported to site and assembled thereafter. The challenges and limitations such an architecture to date, though, is the limited geometrical freedom associated with these components.
Frank Lloyd Wright’s Ennis House in Los Angeles (1924)
At its most extreme, in brick laying, every component is as simplistic geometrically as possible, and while one may argue that such simple components can be assembled into more interesting forms, e.g. Thomas Jefferson's serpentine walls at the University of Virginia, the structural mechanics of ensuring the simple units stack correctly still severely limit the design flexibility.
Other experimental forays into modular architecture are plagued by a similar problem. In the 1920s, Frank Lloyd Wright invented a precast concrete modular block system, called textile blocks, that he realized in six constructed homes in and around Los Angeles, eg the impressive Ennis House. However, Wright was limited to only two types of blocks: plain and decorated. Moreover, the geometry of each of the blocks was identical.
And finally, we have Buckminster Fuller's tensegrity geodesic dome, whereby an astounding number of hexagonal steel frame panels would be airlifted by the US military to sites of operation, or to cover all of New York City. Again, though, like Wright, Fuller was limited to a series of repeating units, rather than differentiated ones. Fuller's Dymaxion House, also unrealized, had to rely purely on similarly standardized mass manufactured components, for the mass production of residential units. One wonders if the hesitancy with which the public greeted the idea, and the it's ultimate lack of realization, was not related to this lack of geometrical diversity.
These historical precedents of modular component based architecture offer inspiration for how 3D printing, limited in scale as it is, though impressive in its geometric versatility, may be able to make a nonetheless significant contribution to architectural design.
The ultimate goal is a rapidly prototyped architecture, whereby not necessarily entire buildings are reshaped at the designer's will, which is likely unrealistic, but that interior spaces within them can be adapted and shaped dynamically to fit the evolving needs of its inhabitants. Once fabricated in a custom 3D printing factory, these components can, as is typical for prefabricated elements, be transported on site and, e.g. be assembled by 6 axis degree of freedom Kuka robotic arms, as innovative experimental architects such as Fabio Gramazio and Matthias Kohler have at ETH Zurich. So called swarm robotics have been pursued as a particularly efficient and versatile method for assembling components on site.
By this construction logic, the overall structural frame of the building could be comprised of, again, traditionally fabricated elements of necessarily larger scale than 3D printing can offer. Within this typically constructed skeletal frame, however, a variety of heterogeneous interior spaces can be rapidly and continuously 3D printed, with the printers themselves even housed within the structure itself. (See Fig. 1)
Fig. 1: Combining interior geometrical versatility via 3D printing with traditional methods for structural strength and scale
In this way, 3D printing may be able to advance the dreams of architectural parametricists. With 3D printing, instead of the mutations in the parametric design purely occurring and being instantiated digitally, as is the case in e.g. Greg Lynn's explorations of animate form, with just the final realized design offering a (limited) glimpse of the design iterations that were cycled through, here the parametric design process would be interlinked and interwoven with construction and inhabitation.
In addition to imparting greater geometric freedom, 3D printing also promises to embed more interesting functionality into the modular components that are assembled on site. While the interconnection region between 3D components need necessarily be stiff, for instance, the interior of these components could be of an entirely different material, e.g. elastic if optically translucent or electrically conductive. (See Fig. 2)
Fig. 2: White is stiff material, red is elastic material, printed seamlessly by 3D printing.
3D printed components may be combined with traditionally fabricated ones by a jointing approach, too, such that the geometry of the printed structure is effectively extended by e.g. a steel beam. (See Fig. 3.)
Fig. 3: The parametrically designed object is 3D printed, whereas the bars are traditionally fabricated.
Finally, if sensor elements can be incorporated into multimaterial 3D printed parts, then construction can more effectively tracked and clashes detected between model and reality, which is a major contributor to waste and inefficiency in the construction industry.