INSTRUCTOR: K. Steinfeld
UA Pratt Institute School of Architecture

Unraveling the Componentized Carpet the six techniques of non-generic surface componentization

There’s an elephant in the computer lab. It’s clear to see that some digital techniques, while offering the promise of a highly tuned and specific surface articulation, have brought us instead to another version of the generic. While the range of techniques of form production accessible to designers continues to expand, the scale of the population of resulting forms has revealed the deeper topologies shared between them. Despite all their difference, all these componentized carpets are starting to look the same. This formal similarity has its roots not only in the cultures of design practice which produce this work, but in specific resistances and tendencies latent in the design technologies they employ. As is true of all forms of representation, these technologies present us with a bounded and filtered perspective on our work – they aren’t just doing things for us, they are doing things to us. While a grasp on the former is straightforward, an understanding of the latter requires both technical mastery and critical distance. This seminar seeks to develop this understanding through a comprehensive unpacking of one of the most dominant tropes in contemporary digital design – the componentized surface.

We look to move beyond the inbuilt resistances and tendencies of surface componentization methods, and to enable design possibilities of greater richness and architectural potential. To achieve this, we will study six advanced techniques, each of which is calibrated to address a specific resistance common to the practice. The course will alternate between seminar and workshop formats. Every two weeks, a new technique critical to a more sophisticated practice of componentization will be introduced. This will be followed by a workshop session where this technique will be put into practice on a project of each participant’s choosing, culminating in a final project.

1. the Tatter

Participants will learn to account for trimmed and other non-rectangular surfaces, in order to incorporate a richer set of guide surface geometry.

2. the Pinch

Participants will learn to account for complex manifolds, such that surfaces which display periodicity or self-intersection may be componentized.

3. the Broken Grid

Participants will learn to generate non-quadrilateral subdivision topology, so that all manner of space filling regular tessellations of a surface may be employed.

4. the Stitch

Participants will learn to gracefully handle edge and seam conditions, in order to differentiate components at joints and edges.

5. the Handle

Participants will learn to produce robust and useful models of control, such that their parametric models may be calibrated precisely and easily.

6. the Field

Participants will learn to create context-responsive differentiation, so that components may react to their surroundings in a meaningful way.

Prerequisites

This course is intended for students who already posses a minimum understanding of algorithmic design techniques. Minimum proficiency in any programming language (Visual Basic, Java, C#, Python, LISP, C++, etc.) or scripting on any platform (Rhino, Maya, AutoCAD, Blender, Processing) is a requirement.