I will be posting to this weblog throughout the year-long design period to document my design process, ideas and more. Check back often to see how I'm doing, and add comments to the entries if you'd like.

Archive for the ‘Scripting’ Category

Week 5

I started this week by creating a large section through my site showing conceptually how the elevated rail, the Milwaukee Innovation Center and the wetland are positioned. Below is the overall section and details:

Preliminary Wetland Section Overall

Preliminary Wetland Section Overall

Preliminary Wetland Section Detail East

Preliminary Wetland Section Detail East

Preliminary Wetland Section Detail West

Preliminary Wetland Section Detail West

I was pretty satisfied how it turned out. I placed the plants according to how much water they prefer: lily pads with deep water; cattails, wild rice and switch grass medium; ash trees and taller grasses moist soil; and ash trees liking fairly dry soil.

The wood platforms seen throughout the section are comprised of two pylons and a cross member. I started working on a script last week that generated the angles and elevations within a series of parameters, but the more I learn with Kyle’s Digital Design class I continue to improve the script. The next generation pictured below takes points supplied by the user (by clicking) to generate the walkway. It still needs some improvements, but it is getting closer:

Articulated Walkway Overhead

Articulated Walkway Overhead

Articulated Walkway Axon

Articulated Walkway Axon

It has become increasingly obvious that what I really required was a polar script. MicroStation allows users to enter rectilinear values (X,Y,Z) or polar values (Angle, Distance) but scripting thus far only has dealt with XYZ. I first developed a polar function, which takes a start point, an angle and a distance and uses trigonometry to determine the position of the end point and returns that XYZ value. The beauty is that I can define an angle range from -10° to 10° and have it draw a zig-zag line with variation, but overall is fairly straight.

I used the principle to create a series of mound shapes for my wetland terrain. I was inspired by Stoss LU (see last week). The script draws 6 points and 6 lines, with points 1-3 between -20° and 20° and points 4-6 draw in reverse, from 160° to 200° (the first set + 180°). The last line connects point 6 to the start at point 1. The result are 6-sided trapezoid shapes. A check function had to be added to determine if any of the lines were crossing and if they did it would reject the points and start over. Twisted shapes would not work. Finally, two points are extended from the middle of the shape to form the top points of the mound. All of these points and then connected with triangular planes to form a 3D object.

Below are a few examples of the shapes generated:

Mounds v1 Sample

Mounds v1 Sample

I then chained them along a zig-zag line as described above using the start point of the next as the mid point of the previous to make a snake:

Mounds v1 Chain

Mounds v1 Chain

The lines sometimes intersect and/or cross which makes nice areas between the landforms for water to collect. Below is a gallery with additional images from the first script:

I then re-wrote the script and the polar function to accommodate spherical geometry, meaning that I can sent a polar vector into any place in space, not just along a plane. The polar function was more complex, but made the actual mound-generating script much easier: I never had to check to see if lines crossed because they were all radiating from a central point. Each point got a 60° piece of the 360° pie and was able to randomly place a point within that area. The left and right pieces were supplied with a longer vector, making the overall shape longer length-wise that height-wise. The height points were also based a the same center point, yielding a very similar mound shape, but is much more efficient and easier to modify:

Mounds v2 Sample

Mounds v2 Sample

I used the polar function again to generate a path framework for the mounds to follow, but instead made them radiate from a single point instead of being parallel. The mound determines the angle of the line used to make the framework, and orients itself along that axis so the long ends are always parallel to the path. Here are some images of the second version script and the radiating and branching effect:

Combined with the existing terrain and the branching script (with some tweaking) generated a system which I am really happy with. Below are two perspectives:

Mounds v2 Axon

Mounds v2 Axon

Mounds v2 Perspective

Mounds v2 Perspective

Next week: back to building design.


Week 4

I began this week by creating a new site model, again at 1″=60′ which is pretty much the largest I can make it and have it remain manageable and cheap to make. This time I carved removed the area where the wetland was to go and filled it in with modeling clay. Again I covered it with polyurethane so that I can easily move the clay around without destroying the paper base. The same building study was teleported to this model and worked into the site. Below are some photos of the new model:

This first scheme was inspired by the work of Stoss Landscape Urbanism (StossLU) and their Riverside Park proposal for New Bedford, Massachusetts. More info here: http://www.stoss.net/riverside.html

I also did some more site sketches and master planning:

Revised Site Sketch Overall

Revised Site Sketch Overall

Revised Site Sketch Wetland 1

Revised Site Sketch Wetland 1

Revised Site Sketch Wetland 2

Revised Site Sketch Wetland 2

I focused mostly on the wetland area and postponed the building designs for now. I did the initial study in clay with the model, and then moved into scripting a elevated walkway to transverse the wetland. Essentially the walkway is composted of two pylons embedded several feet into the soil, sometimes above the water, sometimes below the water line. Spanning between these two pylons is a cross beam, and secondary crossbeams connect each pylon unit. The script then randomly (within certain parameters) changes the angle of the pylons. As the pylons spread apart, the area between them increases yielding a wider space along the walkway to serve as a gathering area. Conversely when they are bent together, the area becomes narrower and less likely to be used as a gathering space.

Walkway v1 3D

Walkway v1 3D

Walkway v1 Side

Walkway v1 Side

Walkway v1 Top

Walkway v1 Top

The above images are some preliminary script outputs using Bentley MicroStation. The next stage in the script process will be to respond to user input and site conditions and generate the appropriate configuration.

Finally, I had my pre-thesis book reprinted in hard cover with Lulu self-publishing service. They turned out really nice and were only around $30 (though there is a 2-week lead time). Below are some images: