Modular Block Studies
Problem
The Modular block studies compiled in this post are an outworking of a tangent in form making during grad school, my first "Random Iterating Script." The designs it produces are loosely restrained and the results, though some are formally interesting have neither the appearance of consistency, nor do they look orderly. So I'd like to unpack two ideas motivating this foray into blocks and modular design:
- Self-similar modules are not a hinderance to curvilinear formal geometry, but rather the inevitable de facto discrete architectural rationalizers of it.
- It's time we dissociated our perception of chaos from our understanding of chaos. Similarly, order is only order within a single system and crossover between systems is the rule rather than the exception. (Look for my upcoming article titled "A Diatribe on Architecture's Perception of the Order/Chaos Dichotomy")
See Modular Block Studies - Selected Iterations for a brief sampling of outputs from this iterative system.
Modular Block Studies - Selected Iterations
As you can see some are aggregations with semblances of order some appear completely random, some have many blocks and some have just a few, some incredibly dense, and some are linear or loosely distributed. These are all the product of the same algorithm, albeit one with lots of looseness and chance built in. Think 100 different chefs all following the same recipe, but rolling dice to determine the number of cups of flour added, and number of minutes left to rise, etc. some might have a cake, some a cookie, some the burnt stuff that oozed onto the bottom of the oven...
Process
These Modular Block Studies are produced by iterations of a three step process:
- Generate a series of parts with a set of connection logic for each.
- Select a series of connections from a set of all possible connection logic with which to attempt aggregating.
- Randomly select the parts from step one, and connect them with the rules selected in step two, and place them in a composition one part at a time without overlapping.
See the following diagram:
Now watch a few of the following iterations aggregate one module at a time...
Modular Block Studies - Selected Iteration #1 - aggregating
Looking at Modular Block Studies - Selected Iteration #1 - Aggregating: We see the rotation input is close to zero, and the parts have a mix in orientations and their connections. The rules employed are numerous enough to allow for the aggregation to grow well in most dimensions. This will be an iteration that informs future script designs.
Modular Block Studies - Selected Iteration #2 - aggregating
Looking atModular Block Studies - Selected Iteration #2 - Aggregating: We see one or two parts and connections led to a repeating iteration that produces a spiral. Here the appearance of randomness doesn’t exist because at each step the aggregator has only one or two identical options for aggregation so it kept selecting the same connections, producing a consistent spiral.
Modular Block Studies - Selected Iteration #3 - aggregating
Looking at Modular Block Studies - Selected Iteration #3 - Aggregating: We see that this one is drawing from such a large pool of possible parts and connections that it takes on the random appearance of branches of lightning. This is what I mean about the appearance of chaos. What looks chaotic may still just be a few elements repeatedly juxtaposed in just a handful of ways.
Conclusion
Ultimately, this particular script is loose, allowing for a lot of variety. This is a good early approach, but inevitably a selection process will have to narrow this script down to engage with a particular design problem of interest. At the same time, some system for architecturalizing this purely formal exercise so far will have to inform scale and consequentially what these elements are.
Summary
This is the working mode of formal generation that will eventually collide with the process in the Tropic of Charlie article titled "Algorithmic Representation." This script employs some iteration mechanics learned from Daniel Pruske during courses at SCI-Arc, and employ's the plug-in Wasp by Andrea Rossi that was used heavily for my partner John Chan and I's thesis, and its in dialog with Modular trajectory of architecture at large. This work is particularly geared toward the design process, in light of both new methodology around construction: modular, discrete, "divide et impera", and computation: AI, this notion of "codified sensibility," and computational design.