Personally, computational design is most rewarding when you have to rely on it to build something you would have otherwise never been able to accomplish both technically and aesthetically. This euphoria is exactly what I experienced during A-week at Tulane this past month.
A-week, more formally known as Architecture Week, is a student-organized tradition at the Tulane School of Architecture where a student committee selects visiting architects based off of their professional work to lead an intense design build effort in three days. This year, the architects consisted of Jason Kelly Johnson and Ripon Deleon from Future Cities Lab, Ronald Rael from Rael San Fratello, Hart Marlow from su11, and myself from CASE. Hart, being my best friend from undergraduate, teamed up with me to lead one of the three groups.
Tulane SOA summarizes the week as follows.
The objective is for the visiting architects to offer students a new perspective on the changing landscape of architecture, technology and culture. Over the span of three days, students participated in workshops led by the visiting architects that engage the students in learning new and innovative digital modeling processes, design charrettes and the full scale fabrication of a project. A-week is an important part of a student’s education as its goal is designed to empower students with knowledge and methods that are redefining the future of the industry.
Our workshop was a huge success. We had a group of 12-14 students, mostly 3rd year with a few upper year sprinkled in, all with no experience in Maya nor Grasshopper. What the students did have was an impressive work ethic, tolerance for each other, and a basic understanding of digital fabrication techniques.
The focus of our workshop was on the design and fabrication of a full scale canopy that was informed by invoking dynamic site constraints. Using the Dynamics Engine within Maya, students produced dynamic curve formations informed by their own site investigations of the Tulane Quad that self assemble into a dynamic system. When played out over time, this system progressively attracted to itself, or cling, resulting in dense cluster areas.
These clusters were analyzed based off of control point density and magic using Grasshopper. These clusters then formed the basis from which a structural system could be derived from. Using a series of custom Grasshopper tools built for the workshop and the dynamic curve system as a rig, the students formulated a layout. After the primary structural elements were designed, additional custom Grasshopper tools allowed the students to skin the structure by selecting which primary elements should terminate the span of a pedal. Since the primary structural elements could be at any angle, and since we need to bend the conduit with low tech equipment, a planar solution had to be developed. When the two terminating structural elements were selected, the tools generated a modifiable intermediate segment, derived by intersecting two planes. The result is the ability to create very complex shapes all while using only planar bent conduit. The skin, its profile, and its graphic was then mapped onto the structure using a third set of Grasshopper tools.
At this stage, the students broke into two teams. A STRUCTURE team generated the output to fabricate and bend conduit tubes as structure. Using a mid-range conduit bender to form the conduit pipes into shape, a flat table and plywood as a plane to assemble, and 10 hands, the students were able to bend and assemble a unit in about 15 minutes. I am always amazed at how well things come together when you follow the direct output of the model. Of course during assembly, it matters that you are as precise as you can possibly be during fabrication.
The SKIN team then designed how the skin both looked and performed during the day. After unrolling the skin/pedals using Grasshopper, Acrylic sheets were CNC milled. By replacing an end mill with a sharpie, the team used the CNC to draw a pattern on the pedals that were later scored with an knife. After painting the graphic on the pedals, they were riveted together on the ground before lifted into place to get riveted to neighboring pedals.