At the outset of this research we aspired to familiarize ourselves with the scripting plugin Grasshopper by attempting to design a chair capable of being digitally output, and whose ergonomic contours would be parametrically defined and adjustable. It took a number of weeks and hours of tutorials to begin wrapping our heads around the way to think in terms of grasshopper’s framework and learning the appropriate combination of components to use to achieve a desired outcome.
In the initial stages of the project we encountered the problem of rationalizing ergonomic characteristics in terms of functions that we could feed into grasshopper. We struggled with how were we going to devise a set of relationships in grasshopper that would adjust curvature of seat and seat back in response to a persons width and height. We decided to approach the chair in section, drawing seat and back at an angle of 104 relative to one another, an optimum number considering the ergonomics of a lounge chair. We then set a series of points along the sectional spine, and assigned curves to those points, perpendicular to the sectional spine, that would serve as the seat sections. At this point we had to decide how to determine the curvature of the seat sections and how their curvature would be parametrically driven.
Our initial instinct was to assign a cosine function, which is similar to the dish shape of a seat, to the seat section curves. By applying a domain and range function to the cosine component, we were able to isolate the concave portion of the cosine wave, which allowed us to adjust the depth and width of the wave while keeping it’s period constant.
After multiple iterations we came to the realization that using a quartic polynomial equation (x4+1=0) as a means of generating seat base in section, we could adjust the parameters of the equation to create a variable ergonomic seat base and seat back. Over the course of this project we both realized that the way in which we had set up our script indicated that were using grasshopper more as a means of generating ergonomic form, and not as a means of responding to ergonomic form as we had set out to do. From this realization we have come to understand the importance of an intent and methodology when using parametric modeling tools.
Although we were not left with a ergonomically responsive chair, nor a 1:1 product, we both
feel extremely confident and comfortable using grasshopper after attempting this sketch problem.
We are optimistic about future success using grasshopper as we continue to explore its capabilities
with our thesis projects.
Until Next time...
-nick + Max
Nick and Max,
ReplyDeleteI was really glad to see your description and the slick renderings of your chair. Creating compelling images showing the design and construction detailing(and pseudo ergonomic fit!) is a good way to bring the project to completion. Your photos of the wood mock-up are an excellent record. The main suggestion I have for the final post is that adding a simple 2D diagram showing the parameters that you varied would bring the text to life.
I know you were frustrated that you cut the chair of the wrong thickness wood, perhaps you will find another time to pick up the ideas and realize a one-to-one version. Since you are so adept with the lasercutter, it should be a direct transfer to do 2D profiles on the router.
I particularly liked seeing the variants. The playful slit pattern of Option 2 looks very attractive and more structurally stable.
I'm glad that you were able to make time to pull together these great images and write a description of what you did. Let me know if you are comfortable if I link to this page and pull some image for my digifab Flickr collection: http://www.flickr.com/photos/chengo/sets/
Hope you can keep going with this exploration of digital and physical explorations.
Nancy