Raumplan Simulation
Plan of Volumes
Zhuorui Li / Shouzhong Lin / Yuqiu Yao
My architecture is not conceived in plans, but in spaces . I do not design floor 4plans, facades, sections. I design spaces. For me, there is no ground floor, first floor etc…. For me, there are only contiguous, continual spaces, rooms, anterooms, terraces etc.
- Adolf Loos
Instead of considering the layout of functions in terms of plan, as in traditional housing, Adolf Loos considers the functions as separate blocks and thinks about the connections between these separate blocks, and uses the transportation space to connect these blocks. The vertical distribution progresses from the public rooms on the groundfloor to the more private on the upper floors, a typical arrangement in Loos houses, and is structured like an interior promenade starting in the entrance room and continuing to the hall with a high jump of 1.2 m. There are two systems usable to vertically cross the house from the basement to the top: a continuous “traditional” staircase which follows the south facade, connected to functions like the kitchen, and a non-continuous one which provides the access to the different rooms from the stepped platforms.
Methodology
The process consists of the following steps: collecting parameters and constraints, generating a bubble diagram, allocating spaces, and transforming bubbles into boxes.
Step 1
The inputs for the final model would mostly be user preferences. Therefore, to generate the dataset, it would be necessary to set some parameters and constraints that would ultimately be set by the user. These parameters include the desired rooms, their size, the connections between them, and the height.
# Inputs
area = x # room area
height = y # room height
radius = z # radius of each sphere
# Calculate the volume of the room
volume = area * height
# Calculate the volume of a sphere with radius 0.5
sphere_volume = (4 / 3) * math.pi * radius**3
# Calculate how many spheres fit into the volume of the room
sphere_count = math.ceil(volume / sphere_volume)
# Output
a = sphere_count
This input room information is transformed algorithmically into equal-volume spheres of customizable radius, and the number of spheres is counted.
The information about the center of the sphere is converted into a point cloud. The points are first categorized, and then the strength of attraction between like-summed and unlike spheres is defined based on the correlation between the functions. The stronger the functional correlation the stronger the interaction force.
Step 2 Simulation
Set the rigid deformation value based on the previous blob data. Make the spheres attract each other but not merge. And set the bounding box according to the building volume.
All of the above is then implanted into the kangaroo Mechanical Simulation Operator for point cloud generation.
Step 3 Visualization
We use the Schröder House as a case study to try out the system and get information about the area, height and relevance of each feature into our system.
The first step in the visualization is to further transform the generated point cloud into spheres and color them according to different functions. The second step is to place the mesh at different heights according to the bounding box placed and the required building height, grab the closest mesh to the sphere and color it and generate the cube.
The final result is a functional block that can be used as a reference for architectural design.
Practical Application
