Stage 1 Part 1
I imported a One WTC mass with floors arrayed at 17-foot intervals. I developed a Dynamo script to parametrically control "Tower Top Rotation" and extract Gross Floor Area (GFA) metrics. By documenting six distinct scenarios, I captured the relationship between structural rotation and GFA. While the first 4 options fall within the GFA requirement, the other 2 options become too skinny, and thus fail to satisfy the GFA requirement. For next steps, I would adjust the building height, making it taller, so that it would have a higher GFA.
Stage 1 Part 2
I wanted to created an undulating tower whose âwaistâ would either undulate inwards or outwards, depending on the width of the middle floor âWidth_midâ. Within the Revit Family Editor, I created three floor plates: base, mid, and top. By lofting these shapes, I created an "hourglass" tower form controlled by the Width_mid parameter. I then loaded the family into a copy of the Dubai project environment and enabled Mass Floors across 32 levels. I tested six design iterations by varying the mid-section width from 100 to 150. As seen in the image above, is a direct correlation between the towerâs curvature and its GFA. While the first two options satisfy the GFA requirement, once the Width_mid gets to 120 feet, it exceeds the GFA requirement. For next steps, I would make the building shorter or reduce the amount of floors, to get under the desired GFA.
Stage 2
For this stage, I worked in Rhino and grasshopper. I created an hourglass tower with two âwaistsâ, one closer to the bottom and one closer to the top. I then generated 12 different towers with different waist parameters, which was increased by a scaling factor. I hit some initial hurdles with data tree alignment affecting the lofting sequence, but flattening the inputs resolved the rror. I used cross-reference logic with Anemone to capture every variation. Since the engine was processing faster than the visual updates, I implemented a brief Python "speed bump" using a sleep function to stabilize the display. This allowed me to consistently capture the results through manual screenshots. For the data output, I kept track of the area.