Facade automations generate a pattern across a whole building in seconds.
◦ What does the automation know, and what does it not know?
◦ Where did the generated pattern need overriding, and why?
◦ If the automation produces something that analyses well but looks wrong, which wins?
Answer: The automation knows geometry and rules. It knows the wall sizes, the module width I gave it (13 ft), what goes in each module (window, balcony door, balcony) and the WWR that comes out. It does not know what is behind the wall or who uses it. It put residential balconies on the ground floor shops because it did not know Floor 1 was retail, and it would not let me add entry doors to a wall it controlled because it did not know people have to get in. It also did not care that the default module was 57.3% glass, which put the building over the 50% limit. So I overrode it in two places: I took the module off Floor 1, gave the shops a 60% storefront and placed three doors by hand, and I swapped the tall side window for a shorter one to get the module down to 44.8%. If the automation makes something that analyses well but looks wrong, I think the design judgment wins, but only after I figure out why it looks wrong. Usually “looks wrong” means it is missing something the analysis does not measure, like entrances, privacy or how the street sees it. The numbers are a check, not the designer.
What drives a project team's decision about floor to floor height?
◦ From a developer's perspective?
◦ From a designer's perspective?
◦ From an engineer's perspective?
◦ From a builder's perspective?
Answer: For a developer it is mostly money and count. A lower floor to floor fits more floors under a height limit and uses less facade and structure per floor, which is why I kept 11 ft and got 7 floors. For a designer it is the quality of the space: ceiling height, daylight reaching deeper into the unit, and a ground floor that feels like a real storefront. That is why I think my 11 ft ground floor is too low and should be about 15 ft. For an engineer it is the depth needed for structure and ducts. An 11 ft floor leaves maybe 9 ft of ceiling after the structure, which is tight once you run mechanical, so they would push for more. For a builder it is repetition and cost. The same height on every residential floor means the same stud lengths, stairs and facade panels, which is faster and cheaper to build. So the final number is a compromise, and in my case the ground floor is where it should break from the rest.
Last week we chose a massing option based on site evidence. This week you had to fit a program inside it.
◦ Where did the program and the massing disagree, and which one gave way?
◦ What would have had to be true for you to change the massing instead of the program?
◦ Which of this week's program decisions will the structural engineer and the mechanical engineer assume you got right?
Answer: I’d say they mostly agreed, which is why I kept the Courtyard Block. Where they disagreed was the unit sizes and the ground floor. The U shape made the corner units come out big, so my average unit is about 803 sf instead of 750 and my mix is 14/43/29/14 instead of 15/45/30/10. The program gave way, because the massing already hit the GFA target (102,395 sf) and did well on sun and daylight last week. I would have changed the massing instead if the program could not fit at all, like if I could not get 2 stairs within travel distance or if the unit count fell way short of the target. The structural and mechanical engineers will assume I got the core right: two stairs at the ends of the arms and the elevator and shaft at the base of the U, stacked the same on every floor. They will also assume the 11 ft floor to floor and the stacked typical floor plan, since structure, risers and ducts all line up with those.
Total carbon can now be estimated before materials are specified.
◦ How much confidence should a team place in that number at this stage?
◦ Which decision made this week has the largest effect on it?
◦ How would you present it to a client without overstating its precision?
At some point a conceptual model has to stop being conceptual.
◦ How do you know when a proposal is resolved enough to move into Revit?
◦ What is the cost of moving too early? Of moving too late?
◦ Which parts of your Forma model do you expect to survive the handoff?