From what I learned from the AEC Global Teamwork, I feel like the most design trade-off occured between disciplines. Therefore, I will not lock myself inside my own discipline, but explore more interdisciplinary options.
Step 1 - Generative Design Framework
In my design task, I will focus on the highrise building design from different discipline’s perspectives.
1) Design decision : Architecture vs. Real Estate
- Description : Build a high-rise building with a certain height on a specified land sizes. Usually, the government will have requirements for the urban planning. We have to design a building in a specified volume.
- Design Variables
- Location on the land
- Geometry (Form of the building)
- Evaluators
- View Quality
- Profit (Renting revenue - cost)
- Land usage
- Most Important Tradeoffs to Consider
- Cost vs. Appearance Quality
2) Design decision : Structure vs. Construction
- Description: we want to design a structural system for the building form, however, the consturction perspectives also needs to be addressed. So, both the structural performance and constructability need to be consider.
- Design Variables
- Structural system selection
- Member sizes
- Evaluators
- structural performance (resilency)
- prefabrication rate
- construction time
- cost
- Most Important Tradeoffs to Consider
- structural performance vs cost
- structural performance vs construction time
3) Design decision : Sustainablity vs. Financial
- Description: Most of the time, we want to our building to be net-zero or achieve LEED certification. It needs a lot of involvement with all disciplines and also the budget control, to be practical.
- Design Variables
- Material Selection for structural system
- Scale of Solar PV and other electrical generation
- Scale of Natural Ventilation
- Evaluators
- Life cycle assessment
- Cost
- User experience
- Embodied Carbon
- Most Important Tradeoffs to Consider
- LCA vs Cost
Step 2 - Generative Design Study
Objectives :
I want to keep exploring the first design decision, architecture vs. real estate decision. I form the problem as building site and form selection in a given constraints (adjacent buildings, land sizes, a range of building heights, minimum functional space). With the given constraints, I want to shape my building to reach the maximum view, maximum land usage and minimum cost.
Model:
In order to not make it to complex, my idea can be achieved by adjacent box building (from last module 6.A.3) and a loft building form with three cross-section (base is circle, the upper two is ellipse) The inputs can be seperated into two parts:
- Loaction:
- x & y of building base center
- rectangle land sizes
- Geometry:
- Base: radius
- Mid : mid heights and a,b radius for the mid ellipse.
- Top : top heights and a,b radius for the top ellipse.
- floor heights
Variable & Evaluators
Design Variables :
- Location: I set the relative x,y location, X normal & Y normal, based on the rectangle land position
- Top Heights is vary from 20 to 25 m considered the relative heights with adjcent building.
Constant :
Another Paramter, Mid heights, mid b radius ratio (0.5-1.5), top b radius radtio (0.5-2) [compared to base radius], can be regarded as the design variables, but in order to make this study simple, I set them to be Constant.
Floor heights, I set them to be 1 to align with the building height.
Base radius is determined by the maxiumum the circle can have to tangent the boundary.
With pre-trials results, the generative design tend to design no volume form, which is not realistic. Therefore, I set the constraints to the design volume. (1500-2500)
Evaluators :
- Ground Floor Area → The area of design base circle. The owners want to make the maximum use the provided land since it is much cheaper Strategies.
- Cost Estimation :
- Weight view values :
I set the floor construction fees to be linear with the heights. The total cost estimation can be calculated :
Previously, the values in Module 6 is decided by the center, but it is biased since the floor has large perimeter that means it can have much better view. Therfore, I take this part into the consideration, mainly the view counts per floor x the floor perimeter:
Interpretation
1) Optimal Location exploration : from my understanding, the generative design can help me find the optimal location in order to have better views. Intuitively, it is a place with less obstruction by other adjacent buildings.
2) The trade of between the view quality and cost. In order to have better view, the building has to be higher to avoid the obstruction, which turn out to be more cost. I expect parento curve to represent this phenomenon.
3) As I said in pre-experiment, the optimal building tends to be very slim, which is not what I want. Therefore, I set some constraints related to volume.
Step 3 - Generative Design Study Results
1) View Points vs Cost Estimation
The result bascially aligns with what I expected. We can observe an upper bound as between weighted views vs. cost estimation. So, there is a clear trade-off, better view quality requires greater cost. Also, we can basically estimate the average cost for the conceptual design, pretty intuitive.
2) Optimized Location in the Land
Very interesting results, except one point at very far left. Another values of x normal are located at around 0.8. Y normal different bring too much different for the results. Therefore, we can decide the center location of x normal, which greatly minize the solution space.
