(Note: despite me turning on the show mass form and floors, it doesn’t save with that setting for some reason).

Stage 1: Revit Conceptual Masses

Part 1

I was having a rough time on the computer cluster computers getting the screenshots on time, so I’ve opted for a video taken with my phone to show the other variations.

I didn’t realize it was only one parameter that we had to vary at this point, and I got ahead of myself by doing two parameters. in this set up, I chose to vary the top outer radius, and the top rotation. Since I’ve been working in Rhino for the most part, I definitely had a tough time even setting everything up and going further into the next parts and stages. In terms of my evaluation metrics, I also modified the custom node to be able to support reporting the gross volume as well. In this first section, I did not even get close with my metrics, and the gross floor area was around 500k SF for the most part. Additionally, the number of floors made it probably more expensive than it could’ve been and I should’ve considered that further.

This part took me a really, really long time to figure out everything. I think I tried at least 5 different types of profiles, and in the end most of them had some form of a bug that made it fail or just not work. I really liked one where it looked like a curving tear drop, but unfortunately, I found out that I could not loft three surfaces together (I had set make surface from closed loop to true), and when I turned it back to a profile of lines, it would fail the constraints. I was stumped for a long time, and decided to move on, but I wish it could’ve worked.

In the end, however, I chose to create a profile of two circles with a filleted connection in between. In the base family file, you could flex the top left and the bottom right radii as you please. In practice, for the final result, I did end up setting up some formulas to relate the radii and fillet size in each level (ratios, circle A > circle B in bottom, circle B > circle A in mid, etc.). I ended up flexing two parameters which were the top rotation and the middle rotation. For the most part my floor area metrics all fell within the range with a couple going above and below the requested range. The surface area is a bit much, which makes sense because a lot of it is going towards the sharp angle since I set the radii in the center so small (which also makes me note that perhaps this is not the most structurally sound of a building…). In terms of cost, I should’ve made the top smaller than the middle to make it less expensive, but on another note, the height of the structure was constant at 600 ft.

Points to Ponder #1

By exporting the values to Excel, we’re given a good overview that provides the metrics and details for each variation really quickly. This is helpful to help determine which forms are appropriate for our client and also be able to understand what we’re looking at more in depth. I can see this also being used for other metrics as well. It’s faster than going between each variation slowly, and comparing their metrics in your head.

My flex parameters for this are as follows:

• Base Radius
• Base Number of Sides
• Mid XY Offset
• Mid Height
• Mid Rotation
• Mid Radius
• Mid Num Sides
• Top XY Offset
• Top Height
• Top Rotation
• Top Radius
• Top Num Sides

For my series in my report, I ended up setting up base radii from 140-160 with a step size of 10, mid height values from 150-270 with a step size of 30, and story height values with step size of 2. I created the custom node, very much like the examples given, but added some of the parameters mentioned above as well allowed a third variable to be flexed.

Points to Ponder #2

There may be better variables out there, but in my experience I feel that the base radii as well as the other radii play a huge role in the amount of floor area—it makes sense because the radii determine how large the cross section will be. I also found that smaller story heights will definitely add to the amount of floor area (with the caveat that we have lower ceilings) because less height dedicated per floor means more floors. Additionally, found the mid height also plays a small role, as long as the radius of the mid value is large enough. If it tapers between the top and the mid, then we want the mid height to be high as possible so the tapering doesn’t reduce our floor area as much.