Parametric dresses

Explorations in computer-aided couture

Following several years of working on flat surfaces, I felt a need to work more dimensionally. These dresses have enabled me to employ a wider range of materials and develop opportunities for inventive construction.

Initially designed using completely analog techniques of draping and patternmaking, the process has evolved to rely on parametric modeling via Rhino/Grasshopper and laser cutting.

In 2014 I acquired a 24” plotter, which allowed me to digitize my paper patterns so I could use traditional CAD tools for drafting and making more precise revisions. I soon realized that if I could draft parametrically, I could not only generate the geometry I was attempting to create intuitively, but I would also save time making revisions. In 2015 I started a laser cutting business and was able to use the laser to rapidly prototype my designs.

I am still revising the designs shown below but will ultimately reconstruct these dresses in final materials suitable for exhibition. These prototypes (or toiles) were made with inexpensive materials such as cotton muslin, acrylic felt, interfacing of various types, braided mylar tubing intended for wire sheathing, paper and chipboard.

Dress 1

Dress 1 pattern
Dress 1

Above:After weeks spent constructing my own narrow fabric tubes by manually cutting strips of bias tape from muslin, stitching them closed and then turning (a feat of sewing wizardry that involves a long pointed hook and a metal tube), I finally decided to outsource this task to a company based in Long Island City that specializes in making spaghetti straps. This spaghetti is white cotton/poly blend with a tricot fill for added body. It is placed over the pattern, glue-basted to the spines and then stitched.

Dress 2

Dress 2 pattern
Dress 2

Above:Most of the dresses are first designed on a half-scale dress form. The prototype is then deconstructed, flattened, photographed, drawn in CAD using parametric tools, enlarged and reconstructed on the full size dress form.

Dress 3

Dress 3 pattern
Dress 3

Above:The original half-scale model used a narrow nylon string to create the volumes that make up the design. I needed to find a material that would hold the same geometries as the string at a larger scale. I also wanted to find a flat, tape-like material that would be easier to stitch to the supporting harness. This photo shows a trial made with cotton twill tape. Ultimately I ended up with 1/8″ braided mylar tubing that is meant for housing wires in electronics projects. This particular product is only available with an iridescent finish—an unintentional but not altogether undesirable effect.

Dress 4

Dress 4 pattern
Dress 4

Above:Out of the five dresses, this dress has the most complex pattern pieces and takes the longest to assemble. It was the first dress that did not begin in half-scale and came to be digitized relatively late in its development. That said, it’s been through four complete, painstaking iterations, and counting! This photo shows the second-to-last iteration where the main improvement was to use newly digitized, plotted pattern templates and outsourced spaghetti material. The latest iteration (modeled above) was made with laser-cut acrylic felt. With each iteration I have moved towards a more controlled process, reducing opportunities for error during construction to better pinpoint any pattern-related issues. Although the felt does not have the same drape and appeal as pieced-together spaghetti-plus-spine approach, I was able to test pattern alterations in days as opposed to months.

Dress 5

Dress 5 pattern
Dress 5 Dress 5 Dress 5

Above:Similar to Dress 4, this dress was first designed at full-scale, directly on the dress form. Once the pattern was digitized, assembly was greatly simplified by generating a tear-away strip that allows the infill to be glue-basted onto the spines in sections rather than piece-by-piece.

Next project: Paintings