STUDIO ARC300: THIRD YEAR DESIGN
University of Kansas, School of Architecture and Urban Design
Nils Gore, Assistant Professor

Back to Home | Back to Project Descriptions

Jennifer Morris

Creating Volumes Through Voids

Voids are used often in architectural forms to create spaces and imply the completion of masses. When an area is left empty, the eye often makes the assumption that a form should be there and completes the object even though it may be incomplete.

Through the process of making several cardboard boxes, I tested and developed the concept of leaving voids around planes to imply the boundaries of a cubic volume. The cube began as six planes, the opposite planes connected by two smaller planes. This produced a solid cube in the middle of the volume, allowing the six exterior planes to “float” and appear to lack edges. I again achieved this concept by forming three pairs of planes, which interlocked. Each plane consisted of two “extensions” which moved beyond the main intersected plane, creating “void” implied boundaries, opposite that of the first box. The third box I attempted consisted of 1” wide loops and 1” wide spaces, which interchanged through all three axises. Two of the three axises I offset, generating many different voids, but never really encompassing or implying the cubic volume.

The next round of boxes expanded the idea of just a void implying a cube, to layers of voids creating space. In order to fulfill this principle, I combined the idea of the interlocking box with that of identical sides. The identical sides all had notches cut out of them. Once formed, the void spaces created an exterior void cubic “frame” as well as an interior one.

At that point, I feel I derailed from my original concept and pursued a trajectory that dead-ended. The next two boxes continued to explore the multi-layer void concept, but both failed in implying the cube; One left too many “layers of voids,” while the other had structural pieces, which had to be added in order to maintain the basic cubic form. This taught me that to imply forms through voids, you must have a certain amount of manipulated masses to begin with. If the form becomes to “spacey,” there will be nothing to contain the void and imply something more. The problem seemed to occur because I began focusing on the interior pieces instead of the exterior.

After realizing this, I backpedaled to my original box with the floating panels and built off this. The next two boxes achieved the goal I sought and also integrated the interior structure. This equalized the importance of the exterior and interior masses, creating many levels of void layers, which also used asymmetry. The first of the two, I fabricated by cutting out six different plane sizes. Opposite planes had symmetrical squares cut from them and recessed into the main void. Four additional “strips” connected the opposing planes, allowing the recessed cubic pieces to create other volumes through voids. Once all six planes were connected, the structure and interior planes as well as the exterior planes implied many layers of space. The other “cube” implied itself through the thickness of its planes. The smaller the plane, the more layers it had and the larger the plane, the fewer layers it had. ‘L’ shaped joints of many layers held the six pieces together. Although these joints protruded from the interior into the exterior void, they completed other void layers and spaces, which in turn, helped define the exterior cubic volume.

Overall, I found that through the use of voids, it is possible to create multiple layers of implied space. There must, however, be enough mass present that the void does not become the overwhelming characteristic of the cube and causes it to “disappear.” I also found that by using masses on the exterior and manipulating the interior space it was easier to define the cube, as opposed to defining the interior space and manipulating the exterior space.

A Whirligig of Risk

What is a whirligig? The first thing that comes to my mind are the wooden birds that look as though they have four wings, and spin when the wind blows. This is a type of whirligig, but upon further investigation, I discovered that a whirligig can be almost any series of things that move or spin due to wind. There are many ways to create a whirligig; the use of junkyard parts, pristine new parts, or handmade pieces, along with anything else one might find can be used. Certainty of risk and certainty of workmanship are the options Pye explains we have when building anything. It is the difference between a handmade piece and a machine made product.

Pye’s idea of workmanship of risk versus workmanship of certainty played a large role in the making of my whirligig. All of the parts, except for the springs, were made by hand. When I began bending wire to form gears I tried to measure or eyeball the points which needed to bend, but found that it produced uneven segments, which made the gear look warped and out of balance. I created a jig made of four nails placed so I could wrap the wire around it easily, and then rotate and invert the wire and bend it again. This worked well, but still caused problems because I could create a wire that had identical bending intervals, but there was no way to create a circular gear from my bent wire. I constructed a second jig, which formed a circular nail template. Using my first jig to make the even intervals, I then fit the bent wire to my template to form a perfectly circular gear. By having it on the template, I thought it would be easy to solder gears together, slip them off, and produce them quickly and easily. Trying to remove it from the jig after soldering, however, was quite a task. It didn’t slip off the jig easily and every time I finally removed it from the jig, the gear would break because of the forces acting on the wire. I thought that by adding tension wires to force the sides of the gear together, the gear would remain intact, instead of flying apart. Unfortunately, the pressures were still too strong and continued to break the welds. I ended up using a small vitamin bottle to wrap the jigged wire around to form the circle. I then wrapped tape about the pieces to keep them in place while soldering the joint. This was upsetting because it left me with less than perfect circular gears. The point of certainty I sought had not been reached. The risk I encountered by having to solder the gears unattached to the jig caused varying shapes of gears, but only in the soldered area. All of the gears, however, because of this method, turned out very similarly.

The crankshaft was much easier to work with since it didn’t require me to bend it into a circle. I made a jig that I could wrap the wire about, pull it off, and have the piece I sought. While pulling it off the jig, the crankshaft bent slightly, but was more or less useable. Due to the workmanship of risk, it seems highly unlikely that perfection can ever be reached. Imperfection was the most frustrating part of building my own pieces, but it also gave the end product character. While it may not be “perfectly built”, it carries a character (additions of pieces unneeded by a workmanship of certainty) that cannot be achieved by machine made parts.
If I were to rebuild my whirligig I would create different jigs to alleviate the problems that occurred while constructing my parts and reduce the imperfection of them. In the building of the crankshaft and gears, the wire would be wrapped around something other than nails. The head of the nails caught the wire as I removed the piece, which not only made removal difficult, but also affected the outcome of the piece by warping it. For the crankshaft, some type of tool that would allow me to remove the entire thing in one movement needs to be incorporated. This might help reduce the amount of bend that occurred. As for the gears, I believe that if I had made them slightly larger and allowed the teeth to be longer, the pressure found in the gears might be lessened. Unnecessary tension wires, time, and frustration would be eliminated.

I was very pleased with the workmanship I put into the frame. I constructed yet another jig to make sure that each connection I soldered was at a right angle. The pieces fit in perfectly; I slid a piece of wood on top of them, and then used a C-clamp to fix them in place. Pieces added to the frame were measured to fix exactly. In the process of creating the frame, I had to take into account the possibility of wracking etc., but small pieces to alleviate such situations were easily incorporated into its design. The difficulty in building the frame was minimal because of its orthogonal orientation, not circular, making the frame close to perfect. There are flaws in the gears, however, constituting “rough workmanship”. Since my whirligig was based on the certainty of risk, I had to “expect and intend the roughest of approximation[s],” as Pye states. Although rough, the gears all meld together, rotating each other. Precision would have helped smooth the movement between the gears, but by adding prongs to the gears, I ensured that each gear would entwine with the others, resulting in uninterrupted movement.

The propeller, created without the use of a jig, was probably one of the most accurate pieces in the entire product. I used a circle template as a basic form and then cut “fins” from the circle at 30-degree angles. I marked each fin to regulate all of the bends needed to form the propeller. The center was marked due to the crosses of the angle marks so, finding and driving a nail through the center was easy.

I chose to use brass because of its flexibility and soldering qualities. It may not have been the most cost efficient material to use, but the medium size whirligig did not need an enormous amount of brass. The softness of aluminum and difficulty and difficulty of soldering copper shied me from them. Having had to work with brass and knowing its properties, I felt the most comfortable using brass to create my whirligig.

After moving from risk to certainty, I still feel that more certainty could have been achieved. I’m sure if I continued making jigs, I would stumble upon a jig that would allow me to make any number of identical gears, possibly with varying teeth sizes. However precise you make the science of creating gears, you will never be able to reach absolute certainty just because of the fact that it is by the human hand these things are created. Human error creates differing parts and results in workmanship of risk. The only absolute way to achieve workmanship of certainty is to use a machine. Although it may still have flaws, there are fewer than that of a human.
I find it ironic that the only pieces of workmanship of certainty found in my whirligig play no role in its function. They serve the purpose of deception. The springs are the “ah-ha” because they seem to be the things causing the lower crankshaft to turn, when in fact it is the top crankshaft turning the gears, which rotate the lower crankshaft.

It is very disillusioning to believe that the perfect thing you’ve drawn on paper can ever be created, at least, by workmanship of risk. As both the designer and constructor, I felt that I could hold myself to a higher standard of construction than if I had passed the design off to a production worker. Since I put the time into the design, I should have the same passion for building it, but at the same time, when I find that things don’t work the way I’ve designed them, I become more easily discouraged. A worker could “make due” with what he had to work with or return to the designer and have them deal with the problem. As both, I had to figure out the bugs and how to solve the problems I incurred. Creating the whirligig proved frustrating at times. In the end, I found that creating everything by hand and tweaking it so that it worked properly was much more rewarding than had I bought “certainty” pieces that I knew would fit together perfectly.