Please enter the following info in the fields above:
- Your Name as the Card title
- The link to your Module 6 folder in our Autodesk Construction Cloud project
Please also type the first few letters of your first name into the Link to Design Journal field, then hover over your name from the list of matching records and click the blue plus sign to link this entry to your Design Journal.
Then, share your Design Journal entry here (replacing these instructions) ... Click the text area below the headers and just start typing your response. There's no need to add new properties.
Please include:
- Images/screenshots showing:
- your original building form
- the recommended building form based on your evaluation and analysis
- Images showing the node logic in your new evaluator nodes
- Images/screenshots or links to your the summary tables (created in Word, Excel, Google Sheets, or any data table tool) showing the input values tested and the values computed for each of the reported parameters
- Brief descriptions outlining:
- Your Single-Objective Optimization scheme (combination/comparison/ranking approach)
- An explanation of why you consider the recommended building form to be the “best” choice
Here is the optimal design option with a twist of 40 degrees and a bottom width of 220’. The building was panelized and using the sun’s directness value I adjusted the opening and color of the panel to provide visual feedback.
The inputs for my script are shown here as a the parameters that will be flexed, the range of values for those parameters, and the object being flexed.
Using the list.map and cartesian.product node I run all these values through a modified version of the evaluation custom node from module 5 with a few extra evaluation metrics added to it.
Once all the combinations have been tested the data for each case is compiled into a list and exported to excel.
Once the optimal parameters have been selected unfreeze the two groups above and freeze or disconnect the testing node and set the element parameters to the desired value in code block on the left-hand side.
For the custom node “Sun directness per Area” the inputs are the element and the u and v values to panelize the surface.
The element is used as a form for a surface which is panelized based on the u and v values, a normal vector facing outward is placed in the middle of each panel, and using a vector dot product between the normal vector and the vector of the sun the directness is computed. Remapping the values from 0 to 1 and summing them up for the entire building gives us the total sun directness for the form. I remapped from 0 to one because a lack of directness from one panel does not reduce the directness of another so a range of 0 (absolutely no directness to 1 (directly facing the sun). Although the name suggests this score is normalized by the gross surface area I decided to remove that feature and felt the total sun directness was a more appropriate metric. This metric is important because it gives a good approximation of how well the building is exposed to natural light. Many people enjoy natural light inside whatever building they are in so a higher sun directness value means the building will be more appealing to potential renters. I was primarily interested in investigating the effect that the twist of the building would have on this parameter.
The next node is the floor area efficiency which tells me the amount of gross floor area per volume. This is a fairly straightforward metric that tells how efficiently is the space enclosed in the building being used. The more floor area the more space there is to be rented out, so it is important that we maximize the amount of floor area enclosed in the building. I was primarily interested in investigating the effect that the area of the bottom floor relative to the top of the building would have on this parameter.
Then I evaluated the estimated cost and revenue from the building to determine the potential financial gain from each design. Similar to the example where cost was evaluated I evaluated the revenue as well with some key differences. The price per floor can be varied exponentially instead of linearly, I chose to vary the revenue per floor parabolically. I included the rent at the first as an input to ensure the rent at the bottom floors does not get too small. Also, I used a 20-year building life for my analysis and a 2% rent increase per year so the total rent is multiplied by 20 and then 1.24 (which I calculated in Matlab). I was primarily interested in investigating the effect that the area of the bottom floor relative to the top of the building would have on this parameter.
The cost per floor followed the same logic as the example however I assumed that this cost would be covered by a loan so I assumed a 10-year loan with an interest rate of 6% so I multiplied by 1.45 to account for the interest on the loan. I was primarily interested in investigating the effect that the area of the bottom floor relative to the top of the building would have on this parameter.
Inputs for my testing node.
Same as module 4 the parameters are updated the metrics are evaluated and reported.
Here the custom nodes evaluate their respective metrics and all the data is complied into a list for each combination of parameters.
Here are the results from my optimization with the best option highlighted in yellow, the second in green, and the third in red. These were picked based on the score in the right-hand column. The score was a grade based on the weighted parameters. I divided each value by the maximum of all the options and that was the grade for that option for that category. Each category is weighted just like our grades in school except the grade for cost is negative. Each grade is multiplied by the respective weight and then added up except the weighted grade is subtracted giving us the total score. Profit is the primary concern so it was weighted at 60%, sun score or directness makes the building more attractive to renters so it was weighted the second highest at 30%, spatial efficiency is important but not as crucial as the other factors so it was weighted at 20%. The options were also penalized for their cost because that represents the risk the developer has to assume and the opportunity cost of not investing elsewhere so it is weighted as a negative 10%.
here is an example of the formula used to score each option.
