Stage 1: Modeling a Parametric Structure
To begin my modeling process, I started out with a concept for my parametric design. I wanted to emulate shade structures that I loved from my memory, like the Gilmore Girls gazebo. I realize now that from the start this was a radial design (which was advised against), but in my plan I was going to make it square with more edges to add different dimensions for parametric change. Here are my notes from this stage:
Planning
- a modern take on the gazebo
- flexible space to be used for leisure, events, and respite while still be integrated into the outdoors
- encourage moments of stillness and movement
- parametric version of Gilmore Girl’s Gazebo
- controlling geometry = arc of overhang (parabolic) and number of archways surrounding the center
This led me to sketch out an umbrella design from a series of lines. I first created two lines from formulas to place parabolic curves at an intersection (forming an X). I used these as guidelines (labeled as such in my code) and then created 4 separate curves to form arches to connect the 4 vertices of the guidelines. I used a simple curve by points method for these since I had very specific point locations in mind.
However, I ran into trouble trying to create a surface from these lines. I learned from this process that Dynamo struggles to input the parameters of so many curves, especially with intersection points. I thought that my methodology would yield something more creative and exciting — as well as something aligned with my original design vision — but I ultimately pigeon holed myself with my design.
To attempt to create a surface I tried a variety of functions, including surface loft, joining curves, splitting curves, and ultimately had success with a curve patch based on only one of the original guide lines. This created a surface in 2 sections, half with 2 arches and the other half with the remaining 2 arches. This proved problematic when trying to outline my UV grid for panel placement. Since Dynamo read them as unrelated surfaces, points were placed overlapping and resulted in panels laying on top of each other (pictured). The teaching team and I were able to brainstorm a fix through the formation of a separate surface by loft calculation of the shade structure.
Stage 2: Transforming my Geometry
The problems persisted into the sine wave calculations. Because the arches were aligned along a specific line (connecting between each vertices), it was very difficult to create a sin wave in a close enough location to allow Dynamo to recognize it as a replacement for one of the arches. I attempted point distance formulas, sine by curve formulas, and even trying to forge a formula from scratch to try and encapsulate the start and end points of the arch specifically. I could never get a line close enough to successfully loft into the surface. Screenshots below capture this struggle.
Again, because Dynamo struggled to produce points within my unique surface area, it was difficult to create columns that would recognize the top as defined by the archways. The multiple attempts from my code are shown below.
Stage 3: Applying my design at different Scales
To make sure that my design was scalable from the start (since I am taking for 4 units), I created integer slider inputs to determine the height and width of the guide curves. The 4 arches were linked to the same inputs so that they could be easily rescaled, addressing stage 3 of this assignment.
From top to bottom: 200ft scale; 20ft scale; 5ft scale
All in all, I spent over 15 hours trying to make this design work successfully. In the end I failed, but I believe that I learned a lot along the way. I learned that the basis of successful parametric design means leaving things more open ended. Rather than basing my original shape based off of 3 simple liens, I wanted to realize a very stringent vision, which is not conducive to variable change. I also learned about the “sunk cost” fallacy. I was so committed to realizing my design, that I wasn’t open to pivoting strategies until I had spent 10+ hours on the same design. In the future, I hope to be more open to pivoting in my design process. Although I was not successful here, I learned A LOT about the various formulas and strategies to realize a singular objective. Attempting 5+ nodes for the same task allowed me to get to know Dynamo and its logic a lot better. I also learned that sometimes alternating software may be the solution! The teaching team showed me how Grasshopper and Rhino would have been much more effective in realizing this specific vision. I am frustrated at the outcome of this design, but I am proud of my hard work for persevering with my vision and am grateful for the help and support from the amazing TA team.