Forming:
Architectural design of the key building types
This week is the first week of Forming, the next step in the building design. We will develop ways of going from space -> shape. This is a process that can be done on many ways, both manual and computational. This weeks objective is to check these processes and try to get a basic map. The bubble diagram is the input for these processes, as we say this configuration of the rooms is optimal or at least close to the most functional layout of the building. In week 2 we created the expected flowchart for this week. Concluding, this week should be about reading into masonry techniques, some preparations for the brick making and going from space to shape.
Masonry
To eventually make the translation to a buildable adobe building, we need to look into masonry. Eventually this turned out to be a bit of a chicken-and-egg-problem, as the structures
should influence the grid, but the grid should also influence the structures chosen. We decided that the shape, and therefore the grid, is the more important one. Nevertheless we
also looked into different kinds of vaults, domes, bricks and sections.
The main conclusions about roofs can be found in the following picture:
Fig. 1: Possibilities of vaults/domes
We realised that to achieve a builable building we have to read into these vaults and how they are constructed. Its not only about the end result, but it should also hold mid-construction.
Reading into this is also very important.
During one of our sessions we also discussed the possibility of digging into the ground to build. It has some good features, shown in Fig 2.
Fig. 1: Possibilities of vaults/domes
Shape
To go from space to shape several paths can be taken to get to a shape and mesh. Several discussions have taken place to get to a basic floorplan and how to get there. This will show some experimenting we did this week to get into designing the hammam.
The first approach we tried is a literal translation of the bubblediagram into a shape. Circles are a great start for compression-only buildings and therefore might work really well. Then we want to add some qualities in the spaces, such as corridors, free height and optical connections. To do this, we divide the plot in a grid with a lot of points and for each point calculate the "importance" of each point of being built. For example, the centre of the circle should have 100% of the height, a point on the edge of the circle has a lower importance and can be built with maybe a 20% height. Points close to a roomconnection should have higher importance, to allow for corridors to exist. When giving the freeheight as an input (Fig 3) all the points in the box are also considered "more important". Overall added (with some factors) this gives a heatmap with importances, seen in Fig 4. This is a potential approach to get an interesting floorplan which might be good for compression-only building. To show the potential we plotted our bubblediagram and moved the points up with x^2 (an approach to cartesian curve, only for presentation purposes) to show how the hammam might look. It was an interesting point of looking at it, but we quickly realized it was never buildable. With only inexperienced builders, we have to come up with something more understandable.
Fig. 3: Importance of the bubble diagram
Fig. 4: Importance of the hallways
Fig. 6: Importance of the free height
Fig. 6: Added imporances to an organic shape
Fig. 6: An example of our bubblediagram, where imporance is heightened under x^2
We realised that this shape is not optimal, for the reason that unskilled people have to build this. We decided to not
futher develop computational ways of getting to a floorplan, and leave computational interventions for later on. However this experiment was far from useless.
It gave an iteresting view on how the whole building might look like and gave a better look towards the garden and the open spaces.
The next path we took was with the Magnetizing floor plan generator, a plugin that translates bubble diagram into spaces within a boundary. It sounds really interesting to use but had some flaws that
made the plugin not suitable for our project. The main flaw is the use of corridors, we planned out 3 "hallways" to function as the main places within the building (the entrance, the garden and the
hammam hall), but implementing this was super hard. Secondly, the maximum rooms allowed in this plugin is 15, where we have more rooms. Lastly, the plugin doesn't give perfect, usable floorplans,
but merely a more interesting to look at floorplan, that is equal to the bubblediagram. We tried the plugin with the boundary and that gave around 20 different configurations. These gave very interesting
insights in the building and how we should (and or could) configurate the building.
Fig. 7: Possible configurations by the Magnetizing floor plan tool
Fig. 8: 4 combinations, from which 2 were rendered without corridors
What can we learn from these 2 computational approaches? First of all these gave optional designs of our Hammam, which is always a nice start. But we did not do these approaches with this reasoning.
The first approach was really experimental and might not create an easy-to-build building, but it made us think about how you could design with compression-only building. The result was not usable, but
it gave a clear view on how the garden could be used. The building would become a bit of 2 buildings in one with a garden inbetween. The second thing we saw was that having a lot of solo "iglos" could give a
nice view, but it's very messy and this might not reflect original Hammam well.
From the magnetizing tool we learned what was a logical by computer to divide the building. This was done with rectangles, from which we all realised it was not conform the topology. Also, when looking at the figure 7, all of the configurations have a corridor around the garden. This might be an interesting feature of the building that we'd like to use in the building.
So concluding:
- The garden should be the glue inbetween 2 functions
- Look into roofing and don't create 15 different "iglos"
- Create a corridor around the garden
Manual shape
We decided it made more sense to create the shape by hand, as we explored some computational approaches which didn't gave the results we hoped. This opened the discussion on how we'd develop the grid. From the more classical hammam we saw that round shapes or polygons were commonly used. (For reference check this link) Also, it would have a kind of elegance that fit well with the hammam where we like to bring some luxury back in the lives of the habitants. From a structural point of view, polygons are quite allright, as they consist out of triangles. This would make it buildable and suitable for compression-only building. So from the topology we'd prefer to have an polygonal grid, more specific a hexogonal grid. The main issue is shown in Fig 9. Rectangular grids allow for distribution of horizontal forces and to cancel out with the next arch on the grid. This is not really possible with the hexogonal grid. Because of this, from a structural point of view, we'd prefer the rectangular grid. Also; its way easier to design using arches and vaults, seen in the masonry paragraph.
Fig. 9: The use of more arches (Joffroy, 1994)
After discussing we decided that topology should be more important and we should try to get it structural buildable, while it doesn't get to complicated. This allows for it to achieve elegance in the design, having a challenge and having hexagons as "our thing". On the other hand, if on the end of next week, we dont have it working, we should retry with rectangles. For this it is important to work parralel on the 3d shape and on the dynamic relaxation.
Brick making
This week also contained the brick making. We have a seperate page about this.
See this page
Next steps
We edited the flowchart to be more fitting of this weeks progress. This week we saw two ways of configuration that were done computational. We also researched a lot into masonry, which will influence the shape. For now, the forming process looked like this:
Fig. 10: Bubble Diagram of the program of requirements
As previously written, next week is an important one that should determine if a hexogonal grid is reachable, but more important buidable. Developing a grid by hand and structural solutions are key in this process. For next week, we should develop a shape that we will go with before wednesday. This will be our deadline, that we should stick to.