core.form-ula

photo:Steph Goralnic

It had a good run. Now it’s finally coming down.
After a long 5 month existence, the Liquid Sky Installation at the PS1 MoMA courtyard is finally coming down. The project began when Ben Ball and Gaston Nogues of Ball-Nogues Studio in Los Angeles were invited to submit a design for the yearly competition. Immediately following their invitation Ben & Gaston submitted a posting on Archinect requesting the help of some experienced parametric modelers. At the time in my final semester at Stevens Product Architecture Lab, I responded to the post, hoping to get some of the younger students involved. After their enthusiastic response, I along with two other students Mark Pollock (currently with SHoP Architects) and Cory Brugger (currently in his final year at Stevens and doing research with SOM) decided to work with Ball-Nogues on the design of the submission, and its modeling and computation, with the hopes of building it upon its success. We dedicated equivilant to a course’s time during the semester on developing the project with the team in LA. Mark and Cory even went to LA during their spring break to help the duo finish the final model.

The team in LA began designing through physical models as they had done on a prior project Maximilian Schell (image01).

image01

Planning on using a similar technique as in Maximilian, we at Stevens began modeling the forms in parametic platforms and developing ways we could automate the tedious process of tiling the funnel forms with variably shaped petals. We were unsure of which platform was better suited to accomplishing our ultimate goal so we began by working tandemly in Generative Components (image02) and Digital Project(image03). Both programs required some programming to automate the petal instantiation.

image02

GC benefited from its speed of instantiation. However, initial setup of the base geometry, the difficulty in defining a complex petal form, and the limited fabrication planning ability, brought Digital Project to the forefront. The model was heavier and instantiation took more time, but the petal geometry became much more easy to define.

image03

Once we had the process for petal creation and propagation set up, we were able to prove to the jury that the size of the project would not be a concern.

Despite some tough competition, our design was chosen and we immediately began the process of fabrication planning and construction.

Below I have posted a sequence of images taken from the competition presentation and our final presentation at Stevens. They give a pretty decent narrative of the process of design, the project requirements, and the process of construction.

From an Idea to Reality.

The idea, in brief, was to create a colorful canopy that spans the extents of the courtyard. These funnel shapes would be supported by and reach their necessary height through the use of large wood tri-pod structures made of untreated utility poles. The canopies themselves are tensile structures constructed of individual petals of mylar sail cloth. Their shapes were determined using a sophisticated form-finding software which produced the skeletal frame which was then imported into the various modeling programs.

The primary programmatic requirements were to provide 1. Shade, 2. Seating, and 3. Water (images05,06,07) . These were achieved through 1. Canopy structures. 2. Net-Hammock Seating spanning the tripods (assembled largely in part by Sheila Pepe) and 3. Sprinklers atop the tripods and the now infamous “Drench Buckets” done by Jenna Didier and Oliver Hess of Fountainhead.
image05

image06

image07

Some additional images of the competition model.

Construction began with the delivery of the tripod utility poles. These had to be cut, assembled on the ground, and then hoisted into pre-made foundation holes. Throughout this time we were still finessing the petal shapes (image08,09). Once the tri-pods were up, their tops were surveyed and the coordinates were updated in the form-finding software (image10) . Once the final forms were determined, we then propagated the petals on each funnel. These petals were then laid out and labeled for fabrication (image11). A fabric cutter in Brooklyn was able to take the rolls of mylar and quickly cut and label the petals as well as their reinforcement patches. The reinforcement patches were necessary at the hole locations of each petal to prevent tear out since the stresses in the shapes due to wind exceeded the materials capability. Once the petals were delivered to the site, they were then punched, patched and grommeted prior to assembly. Assembly occurred in the courtyard, and could be achieved by around 4 people per funnel. Once the funnels were assembled, they were hoisted to their respective locations, attached to the perimeter cables and tensioned.

image08

image09

image10

image11

The “DroopScape”

Steven’s Product Architecture Lab’s only recognition…a 24×36 plastic sign.