c 2005 Nonlinear Phenomena in Complex Systems

Beyond the pole-barn paradox: How the pole is caught

J.P. van der Weele1 and J.H. Snoeijer2,3
Mathematics Department, Division of Applied Analysis, University of Patras, 26500 Patras, Greece 2 Laboratoire de Physique et Me canique des Milieux He te roge `nes, ESPCI, 10 rue Vauquelin, 75005 Paris, France 3 Present address: School of Mathematics, University of Bristol, University Walk, Bristol BS8 1TW, United Kingdom E-mail: weele@math.upatras.gr; Jacco.Snoeijer@bristol.ac.uk Hundred years after Einstein formulated the special theory of relativity [1], we review one of its famous paradoxes: the length contraction paradox, about a farmer who wants to catch a pole in a short barn. The original paradox describes the situation in which the velocity of the pole with respect to the barn is constant, but we go beyond that and focus upon the actual catch in which the pole is brought to a standstill. This natural follow-up question, rarely addressed in textbooks, turns out to have a very surprising outcome. Key words: Special relativity, paradox, Lorentz contraction, shock wave PACS numbers: 03.30.+p, 01.40.-d, 62.50.+p
1

Introduction

One of the most popular paradoxes of Special Relativity, appearing in almost every course on the subject, is the pole-barn paradox [2, 3, 4, 5, 6, 7]. It can be stated as follows: A farmer, who owns a barn 4 m long ( 0 ) and a ladder of length 5 m (L0 ), would like to t the ladder inside the barn. To accomplish this, he plans to use the Lorentz contraction, i.e., the relativistic eect that objects shrink when moving with constant velocity with respect to an observer. The farmer tells his son to take the ladder and run into the barn with speed v , as in Fig. 1, since then the poles length will be reduced (in the rest frame of the barn) to: v . (1) c The square root term is the famous Lorentz contraction factor and c = 3 108 m/s denotes the speed of light. In order to make the pole t in the barn, the farmer requires L0 0 , or equivalently: L0 = L0 1 2 , with = 1
0 2

FIG. 1. The pole-barn paradox in the farmers frame of reference (top), and in the sons frame of reference (bottom).

L0

(2) 1

Nonlinear Phenomena in Complex Systems, 4:1 (2005) 1 - 2

J.P. van der Weele and J.H. Snoeijer: Beyond the pole-barn paradox

For the given values of 0 and L0 this means that the son should run with speed 0.6. Now, the rst basic principle of special relativity is that the laws of physics are the same for two observers who move with a constant velocity with respect to one another [1]. Thus in the sons frame of reference it is the barn that is shortened (Fig. 1 bottom) and the pole does not t. That is the paradox, and we briey review its resolution in Section 2, which holds as long as the speed of the pole with respect to the barn (v ) remains constant. The actual catching of the pole, in which it is brought to rest, is discussed much more rarely [4, 5, 6]. Yet it is a natural follow-up question to the paradox and it illustrates the fact that the outcome of an experiment must be the same for all observers, even if the views on how this outcome is reached may (and will) dier from one inertial frame to the other. In Section 3 we therefore address the problem of bringing the ladder to a standstill. Intriguingly, it turns out that the speed required for a successful catch is much smaller than the one suggested by Eq. (2). Finally, in Section 4 we summarize our results and put the catch into its proper physical context.

FIG. 2. The situation at t = t = 0, in the reference frames of the farmer (top) and the son (bottom). The back door at the right-hand side has been included to allow the pole to pass through the barn uninterruptedly with constant velocity.

Resolution of the paradox

The paradox disappears as soon as one analyzes the situation in the relativistically correct way, recognizing that it is not only a question of length (space) but also of time. That is, we have to work in space-time and we have to consider events, an event being anything that happens at a given point in space-time. In the present case there are two relevant events: (A) the front of the ladder hits the back wall, and (B ) the rear end of the ladder enters the door. According to the farmer event B comes rst (so there is an instant when the ladder is completely

inside the barn and can be caught) but according to the son event A comes rst. This is conrmed by a straightforward inspection of Fig. 2, where the views of the farmer and the son are put side by side. The clocks have been set such that t = t = 0 when the front of the ladder enters the barn (event zero). We have also introduced a back door at the right-hand side of the barn, so the pole can pass through the barn uninterruptedly with constant velocity, which is necessary for the present analysis. Event A: In the farmers frame of reference, the front of the ladder goes through the back door at time tA =
0

(3)

while according to the son this happens at [8] tA =

0

1 2 . v

(4)

Event B: According to the farmer, the back end of

Nonlinear Phenomena in Complex Systems Vol. 4, No. 1, 2005

J.P. van der Weele and J.H. Snoeijer: Beyond the pole-barn paradox the ladder passes through the front door at L0 1 2 v and according to the son at tB = tB = (5)

L0 . (6) v It is easily shown from Eqs. (3)-(6) that at sufciently large velocities (v 0.6c for 0 /L0 = 4/5) the order of A and B is not the same for the farmer and the son; e.g., v = 0.7c gives tA = 5.71/c, tB = 5.10/c, tA = 4.08/c, and tB = 7.14/c. So according to the farmer event B comes rst (tA > tB ), while for his son event A comes rst (tA < tB ). This resolves the paradox: There is no contradiction at all, only a dierent perspective in space-time for the farmer and his son. The fact that an event does not occur simultaneously for observers moving with respect to one another is a common feature of special relativity, and if two events are not causally connected (as is the case for A and B , their separation in space-time being such that |x| > c|t|) even the order in which they occur may be reversed. This makes it possible that in the sons frame of reference (with the shortened barn and the normalsized ladder) the ladder will always stick out from one side or the other, while in the farmers frame (normal-sized barn, shortened ladder) the ladder is completely inside the barn between tB and tA .

FIG. 3. The catch in the sons frame of reference: a race between shock wave and barn door towards the son. The situation is sketched at t = tA , when the ladder rst hits the back wall.

3
3.1

Beyond the paradox: The pole is brought to rest

In the reference frame of the son

Now we come to the actual catching of the ladder. Let us assume that the farmer closes the barn at some moment between tB and tA , when the ladder (in his reference frame) is inside the barn [9]. Bringing the ladder to a standstill requires a tremendous deceleration, inducing considerable internal stresses in the ladder [2], starting when its front end hits the wall and comes to an abrupt stop [10] (Fig. 3): A shock wave carries the information of this collision through the successive parts of the ladder, agitating them violently; the part to the left of the

shock remains at constant velocity until the shock wave passes. Assuming that the ladder survives the shock, at the end of the deceleration process (when the entire ladder has come to rest with respect to the barn), it attains its rest length L0 again. The ladder will of course not t nicely: it will bend, or break, or poke through the door. But the fact is that it is caught. How can this result be explained from the point of view of the son? The crucial point to note here is that the seeming diculty (i.e., that according to the son the ladder is never completely inside the barn) only holds as long as the relative speed between ladder and barn is constant. And this is not the case anymore. Indeed, during the deceleration process even the dierent parts of the ladder itself are in motion with respect to each other, so the ladder temporarily does not represent one single inertial frame, but a range of (accelerated) frames. The length of the ladder during its deceleration is therefore no longer dened by the Lorentz contraction Eq. (1). There is no such thing as a rigid body in special relativity when speed changes. Let us take up the series of events for the son at tA [Eq. (4)] when the front of the ladder hits the back wall, as sketched in Fig. 3. At this instant the son at the rear end of the ladder has no way

Nonlinear Phenomena in Complex Systems Vol. 4, No. 1, 2005

J.P. van der Weele and J.H. Snoeijer: Beyond the pole-barn paradox

of knowing yet that the front has already hit the wall. A shock wave is just starting to come towards him, at a speed w, which cannot exceed c (information cannot travel faster than the speed of light) and typically will be much smaller. The part of the ladder that has already been aected by the shock wave will be in turmoil, but for the son - until the wave reaches him - the situation is unchanged, with the door of the barn coming towards him at speed v . If v is large enough, the door reaches him before the shock wave does so, and the catch will be successful. The door reaches him at t = tA + L0
0

1 2 v

(7)

and the shock wave at t = tA + L0 . w (8)

So we require, for the catch to succeed, L0

0

FIG. 4. The velocity of the pole (v/c) required for a successful catch, as a function of the shock speed w/c, for two representative values of the barn-to-pole ratio f = 0 /L0 [Eq. (11)]. The required velocity gently curves upwards from the dashed line v/c = (1 f )w/c and ends [at w/c = 1] at the value v/c = (1 f 2 )/(1 + f 2 ), cf. Eq. (12). Realistic shock speeds of several km/s lie towards the extreme left of the diagram.

1 2 v

L0 . w

(9)

This can also be written as 1 f 1 2 v/w (with f = 0 /L0 the barn-to-pole ratio), or equivalently 1 /(w/c) f 1 2 . Squaring both sides this gives a quadratic inequality for : 2 1 + f 2 2 + (1 f 2 ) 0. 2 (w/c) w/c (10)

Taking the equality sign, this is an ordinary quadratic equation, whose two solutions dene the bounds of the interval of -values for which Eq. (10) holds. For our problem (which is to nd the minimally required value of ) we are interested in the lower bound of the interval, i.e., 1 1 (1 f 2 ){1 + (w/c)2 f 2 } w . (11) {1 + (w/c)2 f 2 } c

is seen to follow a straight line (dashed in Fig. 4) from which it curves upward for growing w/c. This is conrmed by expansion of Eq. (11): (1 2 3 5 f )(w/c) + 1 2 f (1 f )(w/c) + O (w/c) . The condition (11) is considerably less demand1 f 2 , which ing than that of Eq. (2), was based upon the erroneous notion that the pole should t into the barn while moving at uniform speed v . Even in the worst-case scenario w = c, which makes the catch as hard as possible, the new condition requires a velocity that is not nearly as large [Eq. (11) with w/c = 1]: 1 f2 . 1 + f2 (12)

This is the condition for a successful catch, written in dimensionless form. Figure 4 shows the required velocity of the pole as a function of the shock speed w/c for two dierent values of f . For small w/c the required velocity

For f = 0 /L0 = 4/5 the uniform-speed reasoning would have 3/5 = 0.60, whereas Eq. (12) shows that only = 9/41 0.22 is really needed to accomplish the catch. For realistic shock wave speeds (w c) the required velocity will be much smaller. Shock waves through a solid typically have speeds in the order of several km per second [11]: Taking w = 10 km/s

Nonlinear Phenomena in Complex Systems Vol. 4, No. 1, 2005

J.P. van der Weele and J.H. Snoeijer: Beyond the pole-barn paradox (i.e., w/c = 3.33 105 ), the required velocity according to Eq. (11) is only 6.67 106 , or v 2 km/s. So not only is the catch successful according to the son, but the required speed is much less than foreseen by the farmer based on Eq. (2). This of course means that the analysis in the frame of reference of the farmer can be sharpened. That is the topic of the next subsection.

3.2

In the reference frame of the farmer

In order to make the farmers criterion agree with that of the son, we take up the series of events at tA [Eq. (3)] when the ladder hits the wall (just as in the previous subsection, but now from the point of view of the farmer). At tA the shock wave starts to travel towards the rear end of the ladder, but this end (unaware of the oncoming shock) is still travelling at speed v to the right and keeps doing so until it meets the shock wave, see Fig. 5. Thus, the ladder keeps shrinking until at the meeting point, and that is the moment the farmer should close the door. Obviously, the ladder does not yet have to t inside the barn in the constant-velocity situation (as the farmer had argued initially) and that is why in Fig. 5 we have sketched a pole that is much less contracted than in Figs. 1(top) and 2(top). The shock wave travels at a speed w in the sons frame of reference. Its speed in the farmers frame of reference is, by the well-known velocity addition rule [12, 13]: wv . (13) wf = 1 vw/c2 The ladder will be entirely in the barn after L0 1 2 0 (14) v and the shock wave arrives at the ladders rear end at 2 0 0 (1 vw/c ) = tA + . (15) t = tA + w wv This now yields the following condition: t = tA +
2 L0 1 2 0 0 (1 vw/c ) (16) v wv which, after some rewriting, proves to be exactly the same as the sons criterion Eq. (11). [14]

FIG. 5. The catch in the farmers frame of reference: a race between shock wave and son towards the door of the barn. Analogously as in Fig. 3, the situation is sketched at t = tA , when the ladder just hits the back wall. Note that the son might be inside or outside the barn, depending on the value of (= v/c); here we have chosen to depict the situation for a relatively small value of , in agreement with the catch condition Eq. (11).

Conclusion

In conclusion, the ladder is caught if v is suciently large to outrun the shockwave: The required velocity, Eq. (11), is much less than one would expect by naively requiring that the pole should t into the barn while moving at uniform speed [i.e., Eq. (2)]. In fact, the catch is completely beyond the paradox and is accomplished not thanks to the Lorentz contraction, but thanks to the fact that shock waves have a nite speed of propagation. There is hardly anything relativistic about this: The situation may be compared to a squash ball that is attened during its impact with a wall. At the right moment, it can be caught in a box that is much shorter than the rest diameter of the ball. As mentioned in Section 3.1, the typical speed of shock waves through a solid is in the order of 10 km per second, i.e., w c. Accordingly, the required velocity of the pole is well approximated by the rst term in the expansion of Eq. (11): v (1 f )w . (17)

Nonlinear Phenomena in Complex Systems Vol. 4, No. 1, 2005

J.P. van der Weele and J.H. Snoeijer: Beyond the pole-barn paradox
[4] W. Rindler, Essential Relativity, Special, General, and Cosmological (Van Nostrand Reinhold Co., New York, 1969) 51-52; see also W. Rindler, Introduction to Special Relativity, 2nd ed. (Clarendon Press, Oxford, 1991) 26-27. [5] D.J. Griths, Introduction to Electrodynamics, 2nd ed. (Prentice Hall, Upper Saddle River, NJ, 1989) 458-459. [6] L. Sartori, Understanding Relativity (University of California Press, 1996) 173-184. [7] For a valuable list of relativistic paradoxes, see G.P. Sastry, Is length contraction really paradoxical?, Am. J. Phys. 55, 943-946 (1987). Also on the world wide web many references to these paradoxes can be found. [8] These times are related to each other via the Lorentz transformation tA = (tA vxA /c2 )/ 1 2 , with xA = 0 (and analogously for the times of event B). [9] Several other stopping scenarios, using a series of clamps, are treated in [6]. [10] A non-instantaneous stop of the front end can be treated along the lines of E.F. Taylor and A.P. French, Limitation on proper length in special relativity, Am. J. Phys. 51, 889-893 (1983). See also: H. Nikoli c, Relativistic contraction of an accelerated rod, Am. J. Phys. 67, 1007-1012 (1999). [11] The shock wave speed is related to the speed of sound waves (typically 4 km/s in a solid) via the so-called Hugoniot relation, vshock = vsound + Cup , where C is a material constant (typically 1.5) and up the local velocity of the displaced particles in the solid (also in the order of a few km per second). See e.g. H.C. Pant et al., Equation-of-state studies using laser-driven shock wave propagation through layered foil targets, Current Science 82, 149-157 (January 2002); R.S. Hixson, G.T. Gray, and D.B. Hayes, Shock compression techniques for developing multiphase equations of state, Los Alamos Science 28, 114-119 (2003). [12] This relation, being the relativistically correct version of the more intuitive Galilei addition formula wf = w v , ensures that the speed of light is the same for all observers, i.e., cf = (c v )/(1 v/c). In fact, this invariance of c is one of the two postulates on which Einstein based the theory of relativity; see [1]. [13] From Eq. (12) it may be inferred that w v should be positive, since the shock wave must travel in the proper direction (from the wall to the rear end

For pole-to barn ratios close to 1 this is well within experimental reach: If f = 0.995 the required velocity v will be just about 50 m/s. One might for instance shoot a 100-cm blunt arrow (with a characteristic speed of 70 m/s) into a 99.5-cm box equipped with a trap door. By closing the door at the proper moment, specied in the rst paragraph of Section 3.2, the catch should be fairly easy to accomplish. As a nal remark, it may be noted that an instantaneous transmission of the shock (wf ) would bring us back to the original uniform-speed condition Eq. (2). An innite shock wave velocity in the frame of the farmer would correspond to a nite (but superluminal) shock wave velocity in the sons frame: From the velocity addition rule, Eq. (13), one nds w = c2 /v . Inserted in the catch criterion (9) this superluminal shock speed yields 1 2 , i.e., precisely the original, much 0 /L0 too severe condition (2) formulated by the farmer. This underlines once again the central point of the paper: The catch condition is determined solely by the speed of the shock wave that carries the information of the oncoming standstill from one end of the pole to the other. Acknowledgements We gladly seize the opportunity to congratulate Prof. Dr. Siegfried Grossmann on the occasion of his 75th anniversary. We wish him many more years to come, in good health, and always with the admirable kindness and mastery of the subject that are so characteristic of him.

References
[1] A. Einstein, Zur Elektrodynamik bewegter K orper, Ann. Physik 17, 891-921 (1905); for the English translation see Albert Einstein et al., The Principle of Relativity (Dover, New York, 1924), a collection of the original papers that founded relativity theory. [2] E.M. Dewan, Stress eects due to Lorentz contraction, Am. J. Phys. 31, 383-386 (1963). [3] E.F. Taylor and J.A. Wheeler, Spacetime Physics (Freeman, San Francisco, 1966) 70; the pole and barn paradox is treated in the form of an exercise.

Nonlinear Phenomena in Complex Systems Vol. 4, No. 1, 2005

J.P. van der Weele and J.H. Snoeijer: Beyond the pole-barn paradox
of the pole) also in the farmers frame of reference. Stated dierently, the required velocity v of the pole never exceeds the shock wave speed w, in full agreement with the catch criterion [Eq. (11)] as can be seen clearly in Fig. 4. [14] It may be checked that the worst-case criterion Eq. (12) of the son is found if one takes w = c. In this case also the farmer observes the same shock speed wf = c (cf. footnote [12]), so the right hand side in Eq. (16) becomes 0 /c, reproducing the sons worst-case criterion Eq. (12).

Nonlinear Phenomena in Complex Systems Vol. 4, No. 1, 2005