Knit Partitions for Privacy in Capsule Enviroments
@ Massachusetts Institute of Technology in Cambridge, Massachusetts, USA
In collaboration with:
Mariana Popescu
Tailored Materiality Research, MIT Architecture / Schwartzman College of Computing
GitHub repository
This is the design portion of a broader research project aiming to understand privacy behaviors and mechanisms for spaceflight.
Lack of privacy is one of the biggest stressors in long-duration spaceflight. We especially want light, foldable solutions that can easily accompany astronauts in their capsule. In this project, we explore using machine knitting as a rapid prototyping tool for a flexible lightweight partition, whose opacity may be adjusted through integrated inflatables. Variable opacity allows us to balance affording privacy and social opportunities, as well as fine-tune privacy needs for different types of activities that may need different visual access (e.g., sleeping vs. working vs. “showering”).
Visual Privacy
We use knitting to quickly iterate through different structures and porosities, yielding in different levels of openings. This allows us more precise control over the iteration process to find suitable levels of visual privacy, as opposed to relying on commercial ready-made fabrics.
high porosity
medium porosity
low porosity
CNC Knitting
CNC (Computer Numerical Control), or machine, knitting, is a computerized and semi-automatic additive manufacturing process, much like 3D printing. Instead of melting and depositing material, knitting uses strands of material (yarn) and creates loops using a bed of knitting needles, which are hooks that are preprogrammed to perform three actions: holding, drawing, and releasing yarn. Through a combination of these three actions, over a bed of 1024 needles, a language starts to emerge that becomes functional and/or aesthetic.
“Printing” the first prototype on a Steiger machine
We work on a Steiger double-bed machine, which is capable of independently and dependently knitting on its two beds. The machine is able to work with up to eight yarns from the left and the right side, to combine material properties or, more interestingly, dissagregate material properties based on where they are located in the knit file. The machine uses bitmaps (.bmp), which can be created in an image editor (e.g., Adobe Photoshop) or more robustly, in code via Python PIL.
Example bitmap, where each color/pixel represents an operation
The vocabulary of each needle, combined, creates endless possibilities. We could knit one yarn on the front needles and a different yarn on the back needles, which creates pockets of fabric, like two walls. We can alternate the yarns that go on the front and the back, which creates tight interlocked stitches. We can hold half the needle still and knit on another half, which creates extra loops and starts to create 3D shaping for the fabric. We can knit all the way to the top, then miss the loop, which allows a whole column of loops to drop, creating extra space. All of these local operations create global representations, and we are able to form entire geometries through the combination of three simple actions.