6. Computational Couture#

This week I worked on learning how to:

  1. Use blender (open source Rhino equivalent) with the Sverchok (Grasshopper plug-in equivalent) to create parametric design, and
  2. 3D print aforementioned design on knitted textile (unsuccessfully....so far ;))

1. Designing in blender + Sverchok#

I found a good tutorial online that walked a beginner like me through blender and Sverchok. Following this tutorial was pretty straightforward, but you have to make sure you’re set up. To do so, follow the installation steps as listed here.

According to the tutorial, one can (1) design a 3D model (let’s call this “unit design”), (2) place the unit design into a user-defined grid, and (3) transform grid of unit designs into different shapes (twist, bent or tapered). Figure 1 shows the unit design I made in blender. My inspiration for this came from wanting to make feathers. I’m inspired by Behnaz Farahi’s work and also wondered how peacock feathers looked under the microscope, so I sought to replicate that.


Figure 1. My unit design


Figure 2. Node workflow in the Sverchok interface


Figure 3. Playing around parametrically - how does the z-coefficient change the shape of the array?

2. 3D printing design on textile#

Below are unsuccessful attempts at making my vision come alive:


Figure 4. (L-R) How I wanted the final product to look like (trial 0); First attempt (trial 1); Last attempt (trial 2)

Trial 0: Was printed without textile to see the feasibility of the design. The morphed unit designs were printed individually, and then A large layer height and fast printing speed was chosen in order for fast prototyping.

Trial 1: Printed on knitted cotton textile. The base printed OK, but as it neared the top, it turned into a hot mess and failed. I wasn’t present for the failure, so I had to replicate the failure. Based how it looked, some theories were (1) condensation in the lines, (2) printer moving too fast between high resolution patterns.

Trial 2: Second attempt of Trial 1. The travel speed for the nozzle was decreased to the slowest speed possible. A better print was achieved, but the same failure produced. Poor adhesion of the print to the bed was the cause of the problem and causing. See Figure 5.


Figure 5. Failure more for trial 1 and 2 prints. The 3D print adhered well to the textile (see Figure 6), but not enough to penetrate the textile and adhere to the printer bed


Figure 6. Underside of textiles being printed on for trials 1 and 2.

Some details#

Printer: Sindoh 3D printer
Filament: Polylactic acid (PLA)
Textile (for trials 1 and 2): Pink cotton knit

Things to try for success#

Files#