How much detail should you include about wall, floor and roof composition at different stages?
◦ Conceptual design
◦ Preliminary design
◦ Design development
◦ Construction documentation
Answer: In conceptual design I would barely include any. One generic wall and one generic slab is enough, because I am still testing massing and the numbers only need to compare options fairly. In preliminary design I would pick the main system, like wood stud over a concrete podium, and a rough R-value target, since that affects energy, carbon and wall thickness. In design development I would model the real layers with real thicknesses and thermal values, like I did this week, because now the wall thickness changes the floor area and the analysis needs it. In construction documentation every layer, material and detail has to be exact and match the specs, since someone is going to build it.
You moved a design from Forma into Revit this week.
◦ What survived intact, and what did you rebuild?
Answer: It kept the geometry really well. Levels, footprint, walls, floors, windows, rooms and stairs all came in where they were, so I did not have to redraw the building. What it lost was meaning. Every wall was the same generic type with no layers, the floors had no real structure, and the levels had computer names like B01-L03. Forma knew where things were but not what they were made of. So the handoff saves a lot of time on shape, but all the decisions about materials and assemblies still have to be made by a person in Revit.
◦ What does that tell you about what each tool is for?
◦ Knowing where the handoff lands, what would you do differently in Forma next time?
Many manufacturers provide Revit families for their products.
◦ What is the advantage to the manufacturer of giving them away?
◦ What is the benefit to you of using them over the generic library families?
◦ At what point does it become appropriate to switch from generic to manufacturer specific?
The core is located early and moved rarely.
◦ Why does it get fixed before most of the building?
◦ What are the consequences of a core that does not stack vertically?
◦ What did locating the core force you to change about your Week 3 layout?
Your Insight result did not match your Forma result.
◦ What accounts for the difference?
◦ Which number would you show a client, and how would you explain the gap?
◦ Does a more detailed model automatically produce a more trustworthy prediction?
Answer: Forma said 19.80M kgCO2e over 60 years and Insight said 36.84M, with the same envelope and interiors scope. Embodied actually dropped from 2.82M to 0.49M because Revit knows my real walls are wood stud, while Forma assumed 8 inch concrete everywhere. Operational jumped from 16.98M to 36.36M because Insight defaulted to an Office building with default systems, so it thinks the building needs way more cooling than apartments would. Right now I would show a client a range, not one number, and explain that the two tools assumed different building uses, so the operational part is not comparable yet. Once I fix the building type in Revit I would show the Revit number, because it is based on the real assemblies. And no, more detail does not automatically mean more trustworthy. My Revit model is way more detailed, but one wrong default setting made the total almost double. A detailed model is only as good as its assumptions.