In this assignment, I used Parametric Tower - Twisting Rounded Triangular Mass as my building form. As shown in Figure 3, I chose to flex top height and top rotation as my input parameters. The two new evaluation metrics are building cost including floor area and the wall surface, and solar insolation of wall surface per sqft.
Figure 4 shows the setup for solar analysis and figure 5 shows the custom node that incorporates all results based on the different testing parameters in each case.
The first new evaluation is the building cost including floor area and wall surface. Since the goal is to minimize the construction cost, building envelope could also cost a lot of money, I thought it is necessary to evaluate the building envelope cost. To evaluate the cost including the wall surface, I found the mass floor has a parameter named “Exterior Surface Area”, so I decided to customize the custom node from the class example to get the wall surface for each floor.
As shown in Figure 7, I added the wall surface area to the code block which computes the cost. For simplicity, here I assumed that the wall surface would have the same cost as the floor area, but it could be easily customized by adding a new computational logic.
Figures 8, 9, and 10 show how the wall surface solar insolation per sqft is evaluated. The meaning of this evaluation is to decide how much solar insolation potential that wall surface will have. I think this is more related to people in the building since ground and roof surface are not included, and more solar insolation could reduce the electric light in the building and save heat during winter thus saving energy. Here I divided the cumulative solar insolation by total wall surface area to reduce the effects that if the building has a larger wall surface area, it would also have a larger cumulative solar insolation potential. The way to get the total wall surface is shown in Figure 9, first I pull out the wall surface on each floor, remove all nulls and sum them up.
The custom node to select wall surfaces is shown in Figure 11, by setting the normalVector.Z > -1 && < 0.98, the ground and roof surfaces are filtered.
When doing Single-Objective Optimization, I think it is necessary to evaluate the building space efficiency since the taller building will have a larger cost but it doesn’t mean the space efficiency is lower, so as shown in Figure 12, space efficiency was defined as Gross floor area / gross surface are because we want to minimize the surface area, so I didn’t use gross volume as the denominator.
Figure 13 shows how the maximum and minimum values in each testing case (Construction cost, wall surface solar insolation, and building surface efficiency) are identified and merged in a list.
As shown in Figure 14, I modified the custom node from the class example for evaluating my case, the details of the custom node are shown in Figure 15. Here I set the weighting factors of the cost evaluation and space efficiency equal to 2 and 3, respectively; the weighting factor of wall solar insolation potential is set to equal to 1 because the taller building will have a higher cost but it doesn’t mean the space efficiency is low, they interact and building space efficiency should be the most important factor. While wall surface insolation is important but the cost will have a larger effect on the decision of adopting the design. The trade-off is the gross floor area, while a large gross floor area with relatively high space efficiency could lead to higher construction costs, which results in a lower score.
The higher cost should have a lower score so the equation used in this case is Math.Abs((columnValue[5]-columnMax[0])/(columnMax[0]-columnMin[0])).
The next step is to combine the computed score and test results, so I added the score to the end of the list and then export the results to an excel file.
The results are shown in Figure 17, 20 cases were tested and the top 3 cases were highlighted. The best case is case no.4 which has a score of 3.665, in this case, it has a relatively high space efficiency and gross floor area, and relatively low construction cost, but the total wall solar radiation potential is relatively lower than other cases, that’s also the best design in my opinion, although solar insolation potential is not the best among all cases, it balances the construction cost, space efficiency, and gross floor area.
The best case then was found and set as the input parameter for the building form as shown in Figure 18.
In figure 19, the wall surfaces were selected and panelized to compute the sun directness values, here I used Rect_Panel with Resizable Opening:Open as my input family type. As shown in Figure 20, the panels were colorized from green to red to represent the sun directness as well as the Frame Width. The recommend building form is shown in Figure 21.
