The Black Ghosts

Basic Theory Behind Our Project

Our project employed a clever method of moving the fins. Instead of using motors or other standard actuators to control the position of the fins, we built our own actuators. Our actuators took up very little space and allowed the fins to be spaced under a quarter of an inch away from one another. The key idea for our actuators was the same as that for motors: a loop of wire with a current running through it will produce a force if it is oriented perpendicular to a magnetic field. We spaced an array of magnets close enough to one another to create a strong magnetic field. We then inserted a coil of wire between the magnets and ran a current through the coil:

The resultant force tended to move the coil to one side of the magnets for a positive current, and to the other side for a negative current. If we attach this coil to a long rod on a pivot, the rod will go through a certain angular displacement as the current in the coil is reversed.

However, to be useful, we also needed a way of sensing the position of the coil/rod assembly. We utilized the same principle that we used to actuate the rod. This time, we used a smaller coil of wire concentric with the rotor coil. This ìsenseî coil had a voltage induced in it proportional to the change in flux. In other words, the voltage was related to the velocity of the coil in the magnetic field. We took this signal and integrated it to obtain a voltage related to the position of the coil. This position measurement could then be fed back to a controller that controlled the position of each individual fin ray.

Magnet Holder Design

Our first task was to design a way of holding the magnets apart. The magnets were extremely strong, and spacing them apart from one another would prove difficult. Furthermore, we decided early on that we wanted the magnets to be spaced as close together as possible. There is a tradeoff between having a strong magnetic field and having a lot of turns in the rotor coil, both of which would give a larger force per unit of current flowing through the coil. The magnetic field gets stronger as the magnets are spaced closer together, but the number of windings you can squeeze in between the magnets decreases. If we decrease the space between the magnets by a factor of two, the number of coils we can fit also decreases by a factor of two, but the magnetic field strength increases by a factor of four. Therefore, we decided that we wanted the magnets as close together as was feasible, since this would yield a larger force per unit of current.

In our initial one-fin prototype, we sandwiched a piece of sheet metal between four magnets, and then held two of these sheet metal sandwiches apart from one another by a series of nuts and bolts running through the sheet metal. This was simple to make and allowed us to get going on the electrical design, but it did not allow us to space the magnets very close together. Furthermore, the sheet metal tended to deflect under the force of the magnetic attraction, and we felt that we would not be able to achieve align the magnets very precisely with the amount of bending that we observed.

After experimenting with several different ways of separating the magnets, we found a way of spacing the magnets that we felt was elegant in its simplicity. The basic idea was to position six bolts around the edges of the magnets, and sandwich the magnets in place with nuts tightened down on the bolts. We began by punching six holes in two pieces of sheet metal. The bolts passed through these holes, and the sheet metal kept the bolts spaced correctly. We then inserted a pair of magnets between the bolts, fastened down six nuts on these six bolts, inserted another pair of magnets on top of these nuts, fastened down another six nuts, and so on. This process could have been repeated to create an array of any size.

We chose to create an array of 8 pairs of magnets. This would allow us to have four fins controlled by the handyboard and three empty slots in between these fins to demonstrate that fins could be placed there if desired. This method aligned the magnets very precisely and held the whole structure in place very firmly (we accidentally dropped the assembly down a flight of stairs only to find that the magnets stayed snugly in their places). The picture of the completed design shows what this array looked like when it was fully assembled.

Fin Design

The next step in the design was to create a set of fins with a coil on one end that could slide back and forth between the magnets. Because we decided that the magnets would be spaced close together, the fins had to be extremely thin.

The first hurdle to overcome was the creation of the rotor and sense coils. Professor Peshkin had created some prototype coils using magnet wire wound on a cassette tape spool between two sidewalls made of overhead transparencies. Our goal was to eliminate the sidewalls so that we could devote the entire thickness to wire windings. This meant that the wire windings had to be self-supporting.

After experimenting with several different ways of creating such coils, we found a clever way to wind the coils that impregnated them with glue and eliminated the need for sidewalls. We built a spool that could be attached to the shaft of a motor David had lying around. The spool consisted of a top disk that could be screwed down to the bottom cylinder of the fixture. The disk was separated from the bottom cylinder by a few small washers. Before the top disk was screwed on, we put a small ring of Elmerís glue on the bottom cylinder. The magnet wire could then be wound into the small space between the top disk and the bottom cylinder under the power of a motor. As the magnet wire was wound onto the spool, it was coated with glue. Both the rotor coil and the sense coils were wound concentrically onto this spool. In the picture shown, the fixture has two spools instead of one, allowing two coils to be wound at once.

After about a half an hour, the spool could be unscrewed and the coil could be removed, excess glue peeled off, and set aside for further drying. In another hour, the glue had hardened and the coil was completely self-supporting.

However, we encountered significant difficulty when we tried to attach this coil to the rod that would create our fin ray. Attaching the rod without exceeding the small width dimension of the coil was extremely tricky. A glob of epoxy was too messy, did not meet dimensional specifications, and was not strong enough (we had several fins break that were made this way).

After many iterations, we finally came up with a design that provided enough strength to withstand normal operation of the fin array. We cut very thin strips of aluminum, bent them into loops that would fit snugly around the edge of the coil and extend along the rod for several inches. This design did not adding any width to the structure, but provided much more support for the coils. When the coil, rod, and support brace were assembled, a thin coating of low-viscosity epoxy was applied to the entire structure to seal the coils and permanently connect the rod to the coils.

Final Assembly

Finally, we had the parts to assemble into a completed fin ray. The individual rays had a pivot hole drilled in them. They were then mounted on a pivot rod running the length of the fin array and the coils were assembled in between the magnets. The wires coming from the coils were soldered to a section of ribbon cable that ran to our control circuitry. As a finishing touch, we sewed a membrane out of theraband that could be slid into place over our four fins to create a fin that could potentially move some water. The final assembled fin ray is shown in the picture of our final design.