An ongoing project that I started a little over a year ago was the design and production of hand-sanded wooden boxes. The project started out as a tool to escape boredom, but quickly turned into a study of holism, the part to the whole, and families. A rule set up initially was that no two boxes could be exactly the same. However, each box must share characteristics, or share similar “genetic code.” To achieve this, the development of the boxes took on characteristics of previous boxes such as a certain behavior or manipulation acted on the box, or a distinct aesthetic detail. The difficulty was not only translating the characteristics to other boxes, but the development of new characteristics to inject into the family of ideas.The way the boxes are constructed, which is by subtractive means through sanding, limits the types of behaviors and manipulations that can be enacted on the original form. I eventually reached a point where I was having trouble designing new forms that where original enough from the others, but still maintaining a sense of holism.
After experimenting with grasshopper for a couple of months, I decided to recruit it into the box experiments. The first step was to recreate the primitive form which was pretty simple, a box. However, the box had to be constructed in a way so that it could transform with precision and control. To accomplish this, the box was formed using 15 points that would be weaved together using a Spline curve to form a rectangle. Each point could be manipulated in certain controlled directions mimicking sanding on the surface. This rectangle was then duplicated 5 times in the z direction and finally lofted to form the final form. Next, I began studying the boxes already developed to determine the behavior and manipulations performed on them, so that I could recreate them in the grasshopper digital model. What I found was that there were very few different behaviors performed on the boxes (three to be exact: move, rotate, and scale). However, it was the combination of these behaviors that created the many different forms. From this understanding, I began setting up parametric controls replicating these behaviors. After some tuning, the grasshopper definition was able to reproduce each hand-sanded box by simply adjusting certain parameters.
From this point, I was able to begin exploring new possibilities and new combination of moves that I would have not discovered via my own design process. Initially, I adjusted the parameters manually, but quickly realized that I was restricting the possibilities of the script by trying to find patterns and similarities with the already fabricated boxes. To really utilize the power of Grasshopper, I applied random number generators to the parametric controls to get forms completely unexpected, but still using the same language of behaviors setup with the original boxes. The result was a seemingly infinite amount of configurations that stemmed from the very complex, to basic formal gestures. An exponential increase in new ideas were spawned that would not have been realized from manual manipulation. The idea of this Grasshopper definition was not to generate final box designs that could be milled precisely by a CNC router, but instead, to generate new ideas that fit within an already established family of handmade boxes. Not only did this definition succeed in generating thought provoking designs, but exceeded my expectations of what forms could physically be produced from a subtractive method of box building.
Above- A new box based on ideas generated by the definition