Well, it’s done. A week ago today I finished the installation of line -> plane -> volume. and two minutes later the opening started! Good timing. This week I have enjoyed eating lunches in the gallery with the sculpture, relaxing, and catching up on all the little things I was able to justify putting off while installing. One of those little things happens to be this post. Last time I explained what I had been doing on the computer in preparation for this installation. I finally finished that work about three weeks ago – two weeks before the opening. That Tuesday the steel I had ordered arrived and I moved into the construction phase, the topic of this post. The order was for thirteen lengths of 3/16″ x 1″ steel bars, and three lengths of 1/8″ x 1″ steel angle iron. I had sent a list of lengths which I had printed out from the model in Mathematica, accurate down to one-sixteenth inch. The first thing to do was to measure each piece to check the lengths and figure out which piece was which. Unfortunately, because of a mistake I had noticed over the weekend, I had to trim half of the beams slightly shorter than the numbers I had sent (good thing I didn’t need them to be longer!). Enter machine number one!
The horizontal band saw in the metal shop and I became good friends. Trimming the beams to size was pretty straightforward. Clamp the beam in the vise at the right place, turn on the saw, and let the arm swing down, bringing the blade in contact with the steel. Slick. Once I had beams of the right lengths I marked locations for all of the notches which would hold the string (652, to be exact). The strings are set up so that there is a constant angle of 1° between neighboring lines. This means that if a beam is closer to the center sphere, the notches are closer together and vice-versa. Thus, no two notches are the same distance apart, and on each beam the spacing between notches changes along the length of the beam.
To make the process a little more straightforward I had Mathematica print out the locations of all 652 notches relative to one end of each beam. Armed with a tape measure and 89.3FM I could then run down the list and mark each notch. To cut these marks I set up a sawhorse underneath the arm of the band saw to keep it from cutting all the way through. After the first hundred or so I got into a pretty good rhythm.
The notches were the hard part. After the notches the only thing that made the process slow was that there were sixteen of the beams. I wanted the screws attaching the beams to the walls to be fairly evenly spaced, but their exact location didn’t really matter. I made marks every three feet and then used a drill press, which made drilling through steel a lot easier. I also countersunk the holes so that the screw heads wouldn’t be sticking out. A little polish on the notches with a pneumatic grinder (This to keep the rough steel from snapping the string. That would be bad.) and the beams were ready for paint.
Thus ended week one. Sunday, February 13th was the last day of the group apprentice show (The one in which I displayed the cube.) and so Monday morning I was able to move into the gallery. In addition to getting to start to actually install the work, the next few days were important because I was finally able to begin verifying that the computer model was correct. Certainly I had cut, notched, drilled, and painted all the beams, but all of that was done to measurements from the model; I had no verification that those measurements were correct. Now, actually driving screws into the walls, I could start to see whether points were really where I thought they were. The big fear here was that something would turn out to have been incorrectly measured or calculated and one of the beams would end up over a window or colliding with a pillar or security camera.
Actually I was worried most about hanging the ceiling beams so that’s where I started. In the beginning of January I had Facilities come over and put an array of hooks into the ceiling in preparation. To each end of each of the three ceiling beams I connected four wires to the ceiling and then tightened them so that the beams hung level at just the right spot. That’s easy to say, but it took about three hours to hang each one. Invaluable in this process was the loan of a laser range finder, the coolest gadget of the project. Point it at a wall, press a button, and it gives you the distance to the wall, accurate to one-eighth inch. While up in the lift I could figure out exactly where I was in space with just three clicks (pointing at the floor and the two closest walls). As I had hoped, eight wires on a beam were sufficient to remove almost all motion, keeping the beam steady when bumped or pulled on. I know it’s simple physics, but it gets me every time, that something hanging in midair by tiny wires would be that stable.
In the background of the wider shot of the ceiling above you can see some black lines on the wall to the right and a beam on the left, far wall. These were the next step. Using the range finder I marked out locations for the start and end of all the beams, put in push pins, and tied taught strings where each beam would go. This was necessary because, especially in the longer beams, the middle tends to sag when unsupported. When driving screws into the wall I had a guide line so that the beam was straight along its entire length. Starting with putting up the beams themselves, the rest of the project required more than one set of hands. I was blessed with more than enough help during the final week, from willing friends, girlfriend, sibling, and parents. After five months of solitary planning it was great to be able to share the process with other people.
After the beams were all in place it was the moment of truth. Thankfully my numbers had been correct and all of the beams seemed to be in the right spots, but it was still hard to tell whether all of the string would end up at the right angles. The idea was that each pair of beams would lie in the same plane so that the string between them would also all lie in one plane. It wouldn’t take much of an error though for one beam to be slightly crooked, throwing the string in between into a hyperbolic paraboloid. Luckily, however, the numbers were right and this did not happen. If you stand in the right place, each of the sections of string collapses into a single thick line. (Well, there’s one mistake: one of the ceiling beams was pulled out of place, so that section is a paraboloid. As I was reminded more than once, only Allah is perfect.)
The stringing became more of a logistical than a mathematical puzzle. How can you be efficient when half of the notches are twenty feet off the ground, the others are forty feet away on the other side of the room, and there are forty notches? As you can see in the photos, we came up with a few systems which made the work go pretty fast. This was of course the exciting part. Each string confirmed the numbers and made real the equations I had been staring at for months. To help document the excitement I mounted a webcam in the corner of the gallery and wrote a script to take a photo every thirty seconds. 1,278 frames / 639 minutes /10.5 hours later we tightened the last string, tied it, and cut it off! These frames created the following video, which was on display at the opening.
The only thing left to do is to document well with photographs before it all comes down on Saturday. It was a short run, but absolutely worth it. As someone thoughtfully observed last semester, the installation verges on performance art. More than half the fun was putting it up little by little while people watched and wondered what it would look like. Being in the space and seeing it finished is cool, but it’s even better to have seen it going up or to know the back story of how it was constructed.