Two falling-chain demonstrations based on Einstein’s equivalence principle

Chun Wa Wong,∗ Seo Ho Youn, and Kosuke Yasui

Department of Physics and Astronomy, University of California, Los Angeles, CA 90095-1547
(Dated: September 22, 2006)
Simple demonstrations based on the equivalence principle are given of how a folded chain and a
horizontal flat chain fall down when one chain end is fixed to a rigid support.
arXiv:physics/0609219v1 [physics.class-ph] 25 Sep 2006

We describe here how Einstein’s equivalence principle other words, the falling end falls freely because it is part
can be used to show how the falling ends of two flexible of a freely falling horizontal chain segment. The rea-
chains fall when one chain end in each chain is held fixed son why the horizontal chain segment itself falls freely
to a rigid support. The first of these falling chains is a is because the chain is flexible and bends smoothly and
folded chain that has its two ends initially close together, without kink to the original horizontal shape towards the
with an initial horizontal separation ∆x = 0. Kucharski1 falling end. This bend starts from the supported end of
has given a one-dimensional continuum model of this the chain and travels outward towards the falling end
falling chain. He finds from its Lagrangian equation of as the chain falls. The chain tension begins at the sup-
motion that its energy is conserved and is concentrated ported end and follows the direction of the chain itself,
in the falling arm. The falling arm is thus forced to fall becoming horizontal beyond the bend. It can thus only
faster than g, the acceleration due to gravity. The falling pull the horizontal chain segment towards the supported
time turns out to be 85% of that for free fall, making end without affecting the horizontal segment’s accelera-
it relatively straightforward to confirm the theoretical tion in the vertical direction. Finally, the smoothness of
description.2 A description of this falling chain based the chain at the bend is the result of Newton’s third law
directly on energy conservation can be found in many that states that action and reaction forces must be equal
places,3 including Ref. 4–6. in strength and opposite in direction. This is possible
Calkin and March5 have obtained experimental confir- only when the chain has no kink. If there were a kink,
mation of energy conservation of the falling folded chain the tangent direction of the chain would change discon-
by measuring its tension at the support as a function of tinuously across the kink to give a reaction force that is
the vertical falling distance. They show that their experi- not opposite in direction to the action, in violation of
mental result can be reproduced closely by the Kucharski Newton’s third law.
one-dimensional continuum model, except near the end Historically, the nature of these falling chains have
of fall when the last link in the chain turns over as a been clarified by using high-speed photography,8,9 sensor
rigid body.3 The recent history of this falling chain prob- and electronic circuitry,5 and video capture7 . We shall
lem has been reviewed by Wong and Yasui3 who point demonstrate the motion of these falling chains using only
out that energy conservation arises from the fact that a chain and a smooth horizontal surface provided by a
links are transferred to the stationary arm at the bend of desk or table, with the result interpreted by the equiva-
the flexible chain by elastic collisions, not inelastic ones. lence principle of Einstein.
The falling chain end falls faster than g because that part The principle of mass equivalence, proposed by Ein-
of the chain immediately next to the falling end is below stein in 1907,10 states that the gravitational and iner-
it and pulls it down further. Wong and Yasui are able tial masses of an object are always equal. This means
to isolate the downward chain tension involved with the that the acceleration acquired by a freely falling observer
help of the Lagrangian formulation of mechanics. in a gravitational field is just the acceleration g due to
Recently, Tomaszewski, Pieranski and Geminard7 have gravity. A freely falling observer then sees no gravity.11
re-confirmed the faster than g fall of the folded chain both The sudden realization of this connection between iner-
by experiment and by numerical simulation. They have tia and weight was for Einstein the “happiest thought”
also extended their study to falling chains with nonzero of his life.11
initial separations ∆x. For the limiting case of ∆x ≈ L The weightlessness of a freely falling observer is beau-
(the chain length), where the chain is initially stretched tifully demonstrated by a toy made by Eric Rogers of
taut as an almost flat catenary, they find that the chain Princeton University and given to Einstein as a puzzle
end falls (almost) freely on release, with an acceleration on the occasion of Einstein’s 76th and last birthday in
very close to g. Their video capture pictures show clearly 1955.12 A cup with a hollow tube at its bottom, looking
that the falling end is part of a freely falling horizontal like a champagne glass with a long stem, is placed at the
chain segment that shrinks in length as the chain falls. end of a broomstick. A long soft spring is attached to
A simple way to “see” that the falling end is falling the inside base of the hollow tube. The soft spring is
freely is to note that the chain near the falling end is (al- stretched at the free end by a long thread that connects
most) horizontal so that there is little mass next to and it to a metal ball outside the cup so that the ball hangs
below the falling end to drag it down further, and no over the lip of the cup, as shown schematically in Fig. 1.
mass next to and above the falling end to pull it up. In The problem is to find a foolproof way to move the hang-
 2

States). These light chains will not scratch your table if

you do it only a few times.
The demonstration of a falling flat chain is just as easy.
Arrange the ball chain on the table in a straight line. Lift
the chain by the end representing the support, and pull
it suddenly in a horizontal direction perpendicular to the

FIG. 1: The cup, soft spring and hanging metal ball in the
toy made by Eric Rogers for Einstein’s 76th birthday.

ing metal ball at the end of the long thread into the cup. FIG. 2: Pulling on chains placed on a horizontal surface: (a)
the folded chain, and (b) the straight chain.
The old man solved the problem “at once”.12 Holding
the toy high by the end of the broomstick, he allowed the
toy to fall freely along the length of the long broomstick.
The outside mass became weightless. The soft spring length of the chain on the table, as shown in Fig. 2(b).
then retracted, pulling the ball into the cup, to the old You will see that the chain on the table is pulled along
man’s great delight.13 This story is behind the title “An its length towards the bend of the chain, but it will not
Old Man’s Toy” of Zee’s book on gravity.14 move in the direction of the pull or opposite to it, thereby
Let us now return to our falling folded chain. If its showing that the falling end of the original chain falls
falling end falls freely, it will remain at rest relative to freely. In both demonstrations, the equivalence principle
a freely falling observer. We now propose to show what allows the observer to effectively fall freely in space so
this freely falling observer actually sees by putting a light that the relative motion of the falling chain end can be
folded ball or beaded chain on a smooth horizontal ta- seen more clearly by the unaided eye.
ble top made of glass, marble, steel, Formica or polished These demontrations are most effective when the pull
wood where the friction is likely to be relatively small. at the “support” end is of middling strength, fast enough
Hold one chain end by the hand above the table. This to make friction unimportant but slow enough to make
chain end represents the original chain end at the fixed the motion to be demonstrated clearly seen by the eye.
support. It is next made to move relative to the freely Of course, our pull is unlikely to generate a constant ac-
falling observer by being pulled suddenly away from the celeration or a value necessarily close to g. The demon-
fold of the chain in a horizontal direction along the length strations work even with variable accelerations of any
of the folded chain, as shown in Fig. 2(a). You the ob- convenient strength. This is indeed one of their charms.
server will see that the free end still on the table will The falling flat chain can also be demonstrated directly
not remain at rest on the table, as it should if it were but perhaps not as persuasively in the following way:
falling freely in a falling folded chain. You will find in- A demonstrator stretches the chain horizontally before
stead that it will move in the direction opposite to the letting one end fall down. The observer should stand
pulling direction, thereby showing that the free end of some distance away from the demonstrator in order to
the falling folded chain is pulled by the chain next to and follow the fall readily with the eye. It is possible to see a
below it to fall faster than g. The pull and the motion falling horizontal chain segment before it finally merges
to be observed are sketched in the figure. Marking the into the rotating arm at the support end. One can also
initial position of the chain end on the table with a small see that the kinetic energy of the rotating arm is supplied
object placed next to it will help to define its subsequent by the energy carried in by the chain links transferred
motion. We recommend using about a meter length of a from the falling horizontal chain segment. However, the
light ball chain like those used to operate overhead fans. observer cannot tell for sure that the falling horizontal
(It is sometimes called a No. 6 ball chain in the United chain segment is falling freely.

2
∗
Electronic address: cwong@physics.ucla.edu Experimental confirmation of Kucharski’s observation in
1
W. Kucharski, “Zur Kinetik dehungsloser Seile mit Knick- early years has been mentioned in other papers, and we
stellen,” Ing-Archiv 12, 109–123 (1941). remember seeing in an old book on mechanics a photograph
 3

10
of a falling folded chain falling faster than a small freely A. Einstein, “Über das Relativitätprinzip und die aus dem-
falling ball. However, we are not able to locate any specific selben gezogenen Folgerungen”, Jahrb. Radioakt. 4, 411
reference on these observations. (1907). Reprinted in J. Stachel et al., Ref. 15, Vol. II, pp.
3
C.W. Wong and K. Yasui, “Falling chains”, Am. J. Phys. 433–484.
11
74, 490–496 (2006); arXiv: physics/0508005. A. Einstein, “Grundgedanken und Methoden der Rela-
4
G. Hamel, Theoretische Mechanik, Grundlehren der Math- tivitätstheorie, in ihrer Entwicklung dargestellt”, in J.
ematischen Wissenschaften, Bd LVII (Springer-Verlag, Stachel et al., Ref. 15, Vol. VII, pp. 245–278. The phrase
Berlin, 1949), Aufgabe 100, pp. 643–645. “der Glücklichste Gedanke meines Lebens” appears on p.
5
M.G. Calkin and R.H. March, “The dynamics of a falling 265.
12
chain: I”, Am. J. Phys. 57, 154–157 (1989). E.M. Rogers, “The equivalence principle demonstrated”,
6
S. T. Thornton and J. B. Marion, Classical Dynamics in A.P. Frenchin, Ref. 16, p.131.
13
(Brooks/Cole, Belmont, CA, 2004), 5th ed., pp. 333–335. I.B. Cohen, “The solution”, in A.P. French, Ref. 16, p.132.
7 14
W. Tomaszewski, P. Pieranski and J.-C. Geminard, “The A. Zee, An Old Man’s Toy (Macmillan, New York, 1989),
motion of the freely falling chain tip” Am. J. Phys. 74, pp. 3–4, 10–11.
15
776–783 (2006); arKiv: physics/0510060. J. Stachel et al. (ed.), The Collected Papers of Albert Ein-
8
M. Schagerl, A. Steindl, W. Steiner and H. Troger, “On stein (Princeton University Press, Princeton, 1989).
16
the paradox of the free falling folded chain”, Acta Mech. A.P. French (ed.) Einstein, a centennial volume (Harvard
125, 155–168 (1997). University Press, Cambridge, MA, 1979).
9
<http://math.arizona.edu/~ura/034/Taft.Jefferson/
midterm.pdf>.