These devices arose during the development of the Stroop Effect exhibit on the new Illusions Wall at Glasgow Science Centre. The picture on the left shows the exhibit. The user needs to be able to turn the board around to read what is on the other side, but we want the board to return automatically to its original position when the user has left. There is also a safety/misuse issue: some "users" will try to spin the board as fast as possible. We therefore need some method of slowing the board down rapidly, ideally without affecting the rotation of the board in normal use.
I designed devices to deal with both of these issues. They were both mounted on the shaft of the exhibit, out of sight in the exhibit housing below the board. The next sections describe the devices.
The auto-return mechanism
After considering various arrangements of rising hinges and off-vertical axes, we used a system of magnets to provide the return mechanism. The magnets were very powerful NIB magnets. The sketch below is a schematic representation of the device, with the north-seeking poles of the magnets shaded in red. The rotating arm is shown in the "wrong" position - the fixed magnets are repelling the magnets on the arm, and so the arm wants to rotate by 180° to a position where the magnets are attracting each other. This worked very well. Although it is possible to exactly centre the board in the "wrong" position, so that it won't return, you have to concentrate hard to do so. This is partly because of our choice of slowing mechanism for the board (see later).
Another reason why it's difficult to get the board stuck in the "wrong" position is that the repulsion of the magnets partly unloads the bearings (from which the shaft hangs), thus reducing friction exactly at the position where we want friction to be the least.
The damping mechanism
The aim here was to control violent spinning of the board. We couldn't use a friction device (resistance independent of speed) for two reasons. Firstly, a friction device would be subject to wear and so its characteristics might change with use. Secondly, a friction device would make the disc stiff to turn. This would not only hamper the normal user, but it would also interfere with the return mechanism: it would make it easier for the board to stick in the "wrong" position (see the discussion in the last section). What we needed was a resistive force that was low at low rotation speeds, and increased as speed rises. In other words, we needed damping, not friction.
After some experiments with fluid damping mechanisms, we went for electromagnetic damping instead. We fixed a thick aluminium disc to the axle, and arranged a number of powerful magnets in pairs close to the surface of the disc. The sketch below is a schematic representation of the system. In reality, there was more than one pair of magnets, and the magnets were closer to the disc. The magnets produced an electromagnetic damping effect that meant that the board would never do more than one-and-a-half revolutions no matter how hard you tried to spin it. However, when you turned the board in normal use, the damping was undetectable. In addition, the damping did not increase the likelihood that the board would get stuck in the "wrong" position, because at zero speed there is zero resistance. The strength of the damping could be fine-tuned by adjusting the distance by which the magnets overlapped the disc.
