Aug '08 - Aug '09
Tomorrow is our final review... this quarter, our studio instructors were Dan & Amale of WORKac, and the project brief was taken from an international competition for the House of Arts & Culture, Beirut, Lebanon.
The first few weeks of the quarter were dedicated to research on the context - Beirut and Lebanon: geography, history, culture, arts, etc - and some time was spent on architectural precedents. A few weeks in we started tackling the program... after a few attempts at diagramming program areas and adjacencies we jumped in to massing... the full story (more or less) is documented here on my flickr page, but what follows are my final boards... with some work-in-progress commentary. Consider this a draft of what I'll present tomorrow.
With WORKac as our instructors, it was clear that this studio was to focus on program. The competition brief included a fairly complex table of spaces, which was divided into the above categories. Dan & Amale encouraged us to rethink the program by combining spaces, thinking about sectional adjacencies, etc. In their lecture at OSU they suggested that their own practice dealt with an expanded definition of program - one that encompassed sustainability, urbanism, and material practices as well as the typical layout and functionality of spaces.
For my project, I chose to pick up the line of inquiry initiated by Paul Virilio's "The Oblique Function" (in Architecture Principe, 1966) and advanced by Rem Koolhaas/OMA in the Jussieu Library project. Virilio's text contends that in "the future" architecture will do best to concern itself less with verticals and horizontals than with inclined planes... when walking on a flat surface, no direction of travel requires more energy than any other, wheras inclined surfaces introduce a phenomenological and energetic directionality... the implications for programming seem obvious: there are certain inclinations that are habitable, and certain that are not. In the OMA project, folded floor plates are used to define programattic gradients: for example from grand public spaces to intimate private zones. This is acheived primarily through the modulation of ceiling heights and floor plane inclination, with walls eliminated. I took this Virilio-OMA approach to programming: spaces are defined through the single gesture of a manipulated floor plate.
Secondarily, I wanted to continue the architectural project of HdM's Prada Tokyo, wherein structure, space, and facade patterning become merged. In Prada, the facade patterning extends through the volume of the building as enclosed tubes, the structural diagrid becomes the facade.
The ambition of my project is to combine these two approaches: to define spaces through the relationship between adjacent folded floors, and to define the facade patterning, structure, and interior volumes with a single system. In our cultural research I had hit upon the underlying geometry of arabic screens: many patterns can be generated from a tiling of octagons and squares inscribed with various figures. To avoid a blatantly referential facade pattern I decided to use this underlying geometry as a base.
One desire of the House Of Arts And Culture competition was to create an open, accessible, public center. It was to be prominent without being ostentacious. Highly visible yet unobtrusive. I decided that the basic massing could be a simple extrusion of the site boundary, and the major program elements would be pushed to the exterior in the spirit of openness and connection. By pushing the program to the exterior, I could begin working with the unwrapped facade as a generator. Program areas were placed in elevation according to their specific functional requirements (entry on ground level, for example) and for contextual reasons (the main exhibition space was placed above the adjacent highway, for better views to the south.)
I overlaid this programmed unwrapped-elevation with the base pattern defined above, and began to manipulate it. The pattern could grow or shrink to accommodate the relative sectional size of program areas. The floor configuration required for each program element could be built into the facade pattern. For example, the seating rake in the auditorium could easily be achieved as the diagonal side of an octagon...
The macro-scale pattern could be used as a rough guideline, but to define the smaller scale facade divisions (and the floor plates that would be extruded from the facade) I turned to Grasshopper (Rhino plugin), and set up a parametric pattern based on tiled octagons.
It starts with an array of evenly-spaced squares, each connected by diagonal lines to its neighbors. This array is affected by 3 attractor points which control the size of each square by a function of its distance to the three points. The diagonal lines are drawn between diagonally-adjacent squares, which means the angles vary, but the squares will always consist of verticals and horizontals, and their midpoints are always evenly spaces. As such, if this were to be constructed it could be done with custom-cut panels, each unique, but all equally-sized.
The attractor points were placed where they were for several reasons: if the octagons were to be open and the squares to be closed, it would be desirable to allow the facade to open up towards the southwest into a kind of Briese-Soleil. And following on Virilio's theory, it would be useful for the pattern to take on the appropriate angles where neccessary: for theater seating, for staircases, or flat floors for exhibition spaces and workshops. The grasshopper definition was designed with this in mind. The base grid was set with regard to floor level heights, so the facade generated in this way would be of appropriate scale.
After the facade pattern was designed, the geometry was generated and wrapped back onto the building envelope. The floor plates were then developed based on this facade pattern.
Palm trees were a requirement.
Finally, I think the project is fairly successful. As always, it feels like I didn't quite have enough time, but perhaps it's for the best. I thought i would need to go in and get the plans to a better level of detail, but I think now that that would derail the argument about redefining program based on floor level geometry, so I'm willing to leave them blank. In preparing for tomorrow, I'm expecting some criticism on that (and on my physical model - not pictured) but I think even if it's solely a geometric exercise I can live with how it turned out.