The first cause of trouble was the realization that we would have to construct our own centrifuge. Such a thing is not lightly done, however, as we soon discovered. We wanted something that would spin fast enough to produce a high g-force, because we didn't expect much ion deflection under low gravity. Examining our available options, we decided to try to turn an ordinary blender into the highly scientific sedimentation facilitating apparatus that we needed.
Unfortunately, the blender that was available to us was not manufactured with scientific uses in mind. In fact, we found it very difficult to attach anything to the rotating knob except the manufactured pieces that came with the blender. We were hoping not to destroy the blender in performing our experiment, so we made an effort to work with the blender's design to satisfy our requirements. Here is a rough sketch of what we had to work with.
Essentially, all we needed was a stable, rotating platform to which we could attach our container of the plates and the solution. Thanks to a helpful suggestion, we decided we didn't need to have the battery rotating with the platform, which made things simpler for us.
Our first attempt was quite crude, and was met with the predictable results. As we didn't have the resources to order specially shaped parts, we had to make due with rather cheap materials that either fit or could be transformed into the desired shape. The most difficult obstacle was finding someway to make our platform spin, which meant attaching something to the blender's spinning knob without destroying it. We thought we found a solution, and designed our first setup around that solution. In the end, our construction was too weak, and could not withstand the violent rotation of the blender. We abandoned our design and tried to come up with a more sturdy solution.
We thought we had a solution. We happened to get hold of a piece that fit rather well on the knob of the blender, and we were able to make attach the platform we needed. But again, our materials were not very durable, and this was the cause of our eventual failure. But the real problem lay in the instability of the centrifuge. When spinning around at the very high speed that blenders do, any sort of wobble in our platform would be magnified greatly. This is not a problem for blenders in their normal mode of operation, because it is only a small set of blades that spins around, inside a sturdily built supporting structure. With specially designed centrifuges, the rotating rack is very precisely symmetric, and the structure that surrounds this rack prevents any vibration that might result from an unbalanced load. In our case, we could not build a supporting structure, and thus there was nothing to keep the platform from vibrating fiercely. We had tested the centrifuge successfully for short durations, of 10 seconds or less. When we attempted to let it run for a prolonged amount of time however, a piece cracked, and with it all our efforts and hopes.
With these unexpected problems, and rather poor budgeting of time, we were at an impasse. There was no experiment, as far as we were concerned, without a centrifuge. After having invested so much time into our first two attempts, it was unlikely that we could have constructed another, and an operational one, in the time we were left with. We had expected the actual gathering of data to take a considerable amount of time anyway, and we decided that, with the time contraints we unfortunately put ourselves under, our proposed experiment was unfeasible.