Magnet falling down a copper pipe

From left to right: the thick-walled pipe that came with the magnet; The next three are thin-walled copper; one uncut, one slit along the length and one divided into about 55 rings. The magnet is in front.
At the Science Toy store (Science Toy Magic LLC, (505) 820-7264 500 Montezuma in Santa Fe) I bought this toy consisting of a 2-foot section of 3/4" copper pipe and a small but strong magnet. When you drop the magnet into the pipe, it   s l o w l y   descends and eventually falls out the bottom end of the pipe. Since copper is non-magnetic, this effect is of course caused by the eddy currents induced by the moving magnet: the currents in turn produce a magnetic field opposing that of the falling magnet, thereby slowing it down.

Now in the standard picture of how this works, the currents run around the tube, in loops perpendicular to the tube axis. Therefore if I make a slit in a tube all the way down its length, the currents wouldn't be able to go all the way around the tube anymore, and the effect would disappear. Conversely, if I divide the tube into many rings, currents would still be able to make the circuit, and a collection of rings would therefore be as good in producing the slowing-down effect as an uncut tube.

So I went to the hardware store, and bought three 2-foot lengths of copper pipe. These pipes had thinner walls than the one I bought in the toy store. One I left unmolested, one I cut down the length producing a 1.5mm gap, and the last one I cut into 1/2" sections (about 50-60 before I got tired). I slipped the rings onto a thin transparent plastic tube (rolled from a fluorescent-tube cover).

The expectation was that the transit time in the original thick-walled tube would be the longest, and in the thinner-walled tube a bit shorter. The rings would be about the same as the uncut tube, and time through the slit tube would be very short. Here are the results. Since I didn't have a stopwatch, I counted ticks on a metronome. This is very rough:
Tube type ticks
thick-walled 15
thin-walled 7
slit 5
rings 3
Surprise, surprise! Not what I expected.



When you look down the tube as the magnet descends, you can see it rotating and tumbling. I decided to restrict the orientation of the magnet, since would also make it easier to think about the problem. I rolled the magnet into a small paper tube that loosely fit inside the tubes, in one of two positions as shown in the figure. Again I timed the transit times with ticks of the metronome:
Tube type orientation
A B tumbling
thick-walled 9 14 15
thin-walled 4.5 6 7
slit 3.5 5 5
rings 3 8 3


The difference that the orientation makes for the rings is the most pronounced.

If there is anyone out there that has a good explanation for all this, I'd like to hear it.


Last update 3 Jan 2005 - Hubert van Hecke
back