- For 2 or More Units: Create Two New Evaluator Nodes
General logic :
As seen in the picture below, my main code structure is to add the two custom nodes to the function from Module 5, which added inputs to the function and added two columns to the output tables.
Custom Node 1, Evaluating floor costs :
My first evaluator was the cost of the floors, depending on the surface area and the cost per level. I chose this evaluator because I think that the cost of the building is one of the main factors for the investor, and we can present the results in terms of dollars, which is very clear for everyone.
The first step was to get the area of each floor at every level through âBuildingFormMassFloorV2â, where the element was input. From there, the mass floors were selected and the respective areas were extracted, along with other parameters that were not used in this assignment.
The next step was to estimate the cost through âEstimateCostByFloorâ, where the output from the previous step, the floor areas, is input along with building properties such as the height between floors, the cost per square meter at the lowest and highest levels, respectively $5k and $10k per square meter, as well as the highest level. The cost of each level, depending on its position, is then interpolated using a code block. After that, the areas are multiplied by their respective costs, and everything is summed after âcleaning the listâ of null elements. This gives the output, which is the total cost of all the floors.
BuildingFormMassFloorV2
EstimateCostByFloor
Custom Node 2, Evaluating the view :
My second evaluator was the view of the building, which I computed using a relative (to a landmark) View Score. This score does not have a direct physical meaning, but it helps compare the views of the different alternatives. This is important for the investor, as having the best view possible can attract more business.
To implement this evaluator, the main inputs were the number of U and V panels, the building we are designing, and the building we are using as the reference for the view. In this case, I added a Mass object with a height of 828 meters, corresponding to the Burj Khalifaâs height, northwest of our building, though not at its exact coordinates, as the reference for the view. First, I panelized the outside surface of the tower, which I extracted using the GetSurface node. After obtaining the panels, I had to flatten the list in order to produce a compatible output for the following steps.
Finally, to compute the actual âview score,â I used âComputeDirectnessToObjectâ, which takes as inputs the panel surfaces and two elements: the building we are working on and the one we are using as the reference for the view score. It gets the coordinates of both, computes a vector using these points, and normalizes it. The dot product of this vector with the normals to the panels, evaluated at their centers, defines a score, which is then remapped between 0 and 1. This list is summed, giving the view score for that alternative, which is the output of this part.
Panelize
GetWallSurface
ComputeDirectnessToObject
Results Table:
Note on angle : The higher the angle, the less the twist in this case.
We can see that the area and volume parameters decrease as the twist increases, and naturally, we can expect the floor cost to decrease as well. As for the view score, we can see that it is highest when the angle is 60 degrees. The âbetterâ option depends on the ownerâs preferences, but, for example, an angle of 60 degrees could be a good choice, as it maximizes the View Score and is not very far from the cheapest option (4.58%), which could also be harder to construct due to its higher twist.
Point to Ponder :
Do the new evaluation metrics that youâve designed capture the meaningful differences between the building form alternatives?
The two evaluation metrics capture meaningful differences between the building form alternatives because they reflect two important but different design priorities. The floor cost evaluator shows how the geometry affects the approximate economic performance of the tower, while the view score evaluator captures how the changing twist influences the buildingâs orientation toward a desirable landmark. Together, they allow us to compare alternatives not only in terms of size and cost, but also in terms of a qualitative design benefit that may increase the buildingâs value.
What other metrics would be useful to compute to help understand and make the case for which alternatives are truly better than others?
The previous metrics do not fully determine which alternative is best. The cost metric is simplified, and the view score is relative rather than a physical measure of visibility. Other useful metrics could include structural efficiency, such as estimated lateral stiffness or the amount of material required; usable floor area efficiency; solar exposure; and a more advanced view analysis that accounts for obstructions and the number of occupied spaces with direct views. These additional evaluators would provide a more complete basis for deciding which form is truly preferable.