OpenWorldLab

Loading interactive view…

Induction and circuits

A magnet falling through a copper pipe

Drop a strong magnet down a copper pipe and it drifts down as if through honey, although copper is not magnetic. As it falls it changes the flux through each slice of the wall, and the eddy currents that flow round the pipe hold it up.

Before you look

Predict first: Once the magnet in the copper pipe has stopped speeding up, how does its speed compare with the free fall in the plastic pipe?

What you are looking at

Two neodymium magnets, 12 mm across, dropped at the same moment into a copper pipe on the left and a plastic one on the right. The clock runs at a fifth of real time.

The plastic one drops straight through. The copper one sinks slowly, and the wall around it lights up: rings of current running one way above the magnet and the other way below.

The physics

Each thin slice of wall, height dz, is a ring of resistance 2πa/(σw dz)2\pi a/(\sigma w\,dz) with an EMF v dΦ/dzv\,d\Phi/dz round it. Adding up the heat over all slices gives a drag force proportional to the speed, F=kvF = kv, with

k=2πaσw∫−∞∞Bρ(a,z)2 dzk = 2\pi a\sigma w\int_{-\infty}^{\infty} B_\rho(a, z)^2\,dz

where aa is the pipe’s radius, ww its wall and σ\sigma copper’s conductivity. The magnet stops speeding up when kv=mgkv = mg:

vt=mgkv_t = \frac{mg}{k}

For a magnet small compared with the pipe this becomes MacLatchy’s formula k=45μ02m2σw/1024a4k = 45\mu_0^2 m^2\sigma w/1024a^4. This magnet nearly fills the pipe, so the lab uses the integral; the point-dipole value would be about a quarter too slow.

Going deeper

Vector calculus and Maxwell’s equations

The approximation behind the formula is that the eddy currents’ own field is negligible: the magnetic Reynolds number μ0σwv\mu_0\sigma w v is about 0.005 here. A magnet falling hundreds of times faster would push the eddy currents’ field out ahead of it, and the drag would stop growing with speed, as in a magnetic brake.

What to measure

Terminal speed against the thin-wall eddy-current result

  1. Watch the copper pipe’s magnet once it is well inside: its pill reads its speed, steady at about 5.8 cm/s.
  2. Compare with the plastic pipe’s magnet as it leaves the bottom: about 1.63 m/s, √(2gh).
  3. The copper pipe’s pill gives the biggest eddy current, a few amperes.

What the model leaves out

  • The pipe is a stack of rings 1.2 mm high, each carrying a uniform current round the axis: exact in direction for a magnet on the axis, and right while the skin depth is thicker than the wall.
  • Coils, rings and magnets are filament loops, each standing for a bundle of turns or a slice of a magnet’s surface current. Very close to a loop the field is a little lumpy; a few loop spacings away it is exact.
  • The magnet slides straight down the middle; a real one rubs on the wall and tumbles a little.

Things to try

  • Change the copper pipe to aluminium: its conductivity is lower, so the magnet falls about 1.6 times faster.
  • Make the wall twice as thick: the drag doubles and the speed halves.
  • Halve the magnet’s remanence BrB_r (how strongly it is magnetised): the drag goes as Br2B_r^2 while the weight stays the same, so it falls four times faster.

References

  • C. S. MacLatchy, P. Backman and L. Bogan (1993), A quantitative magnetic braking experiment, American Journal of Physics 61, 1096.
  • D. J. Griffiths (2017), Introduction to Electrodynamics, 4th edition, Cambridge University Press.
  • E. M. Purcell and D. J. Morin (2013), Electricity and Magnetism, 3rd edition, Cambridge University Press.

More experiments, and how the lab works, are in the lab’s overview.