CLASSROOM

Vijay A Singh, Rajesh B

Khaparde and S R Pathare The Mechanical Black Box
Homi Bhabha Centre for Sci- A Challenge from the 35th International Physics Olympiad
ence Education (TIFR)
VN Purav Marg, Mankhurd
Mumbai 400 088, India. This article describes an interesting challenge con-
Email: cerning a mechanical black box that was posed as
[email protected]
a ¯ve hour experimental examination in the 35th
International Physics Olympiad1 held in Pohang,
Korea from July 15-23 2004. It marked India's
seventh foray into this exciting event where sev-
1
See articles listed in Suggested
enty one nations participated. As a leader of
Reading for details on IPhO.
the Indian team at Korea, one of us (RBK) was
privileged to be in the thick of action. Our per-
formance was a success and we secured one gold,
two silver and two bronze medals.
The Problem
The problem is presented in an abbreviated form to
make it suitable for presentation as an article. The so-
lution is discussed in a formal fashion without getting
into distracting `numerics'. The readers are however en-
Keywords couraged to set up the experiment and then they will
Black box, mechanical, sprin-
perforce get involved in the detailed numerics.
gs, rotation, centre of mass,
moment of inertia, work–energy
The mechanical `black box' consists of a solid spherical
theorem, Olympiad.
ball attached to two springs ¯xed in a cylindrical tube
as shown in Figure (1). The tube is sealed with two
identical end caps. The purpose of the experiment is to
determine the mass m of the ball and the spring con-
Figure 1. The mechanical stants k1 and k 2 of the springs without of course open-
‘black box’ in its equilib- ing the black box. Here we discuss the determination
rium horizontal position.
The box consists of a rigid
cylindrical tube. It encloses
a spherical ball of unknown
mass m which is attached
to two soft springs of un-
known spring constants k 1
and k2.

RESONANCE  April 2005 75

 CLASSROOM

of m by the method suggested by the Scienti¯c Com-

mittee of the 35th International Physics Olympiad. We
however discovered that the spring constants could be
determined by cleverly employing the results used to de-
termine m. We could thus dispense with the elaborate
observational exercise suggested by the Committee to
determine the spring constants. We present this `short-
cut' method here.
In the analysis of the experiment the following aspects
must be kept in mind:

² The cylindrical tube is rigid and homogeneous.

² The lengths of the springs when they are not ex-
tended can be ignored. The springs are soft.
² The diameter of the ball (2.2 cm) is slightly smaller
than the inner diameter of the tube (2.3 cm).
² There is ¯nite friction between the ball and the
inner walls of the tube.
Figure 2. The rotation of
² The energy dissipation due to the kinetic friction
the black box is accom-
plished by mounting it on a
between the ball and inner walls of the tube can
rotating stub. A thread be ignored.
wound around the stub ² The pulleys (see Figure 2) are frictionless. We also
passes over an arrange- assume that their rotational kinetic energy can be
ment of pulleys. As the ignored.
known mass m 0 descends,
the black box rotates. The experimental determination of the ball mass m and
the spring constants k1 and k2 proceeds in several stages:

76 RESONANCE  April 2005

 CLASSROOM

Product of the mass and the position of the ball:

Let l be the position of the centre of the ball relative to
that of the tube when the black box lies horizontally as
shown in Figure 1. Obtain the product ml.
Slow rotation: Figure 2 depicts the black box being
¯xed onto a rotating stub. One end of a string is wound
around the stub while the other end goes over a pulley
and has a mass m0 attached to it. As m0 falls under
gravity the black box rotates. We ¯rst consider slow
rotation regime. In this regime friction dominates and
the ball does not slip from its static equilibrium position.
Relate the speed v of the attached mass m0 to the height
h of the fall.
Fast rotation: When the black box rotates su±ciently
fast, friction is overcome, and the ball moves to the end
of the tube. It stays at this far end since the springs are
soft. Once again relate v to h.
Determination of mass of the ball: Determine the
mass m of the ball inside the black box.
Determination of the spring constants: Deter-
mine the two spring constants k1 and k2 . Hint: Employ
the results of `fast rotation'.
The Solution
Product of the mass and the position of the ball:
Let lCM be the distance between the centre of the tube
and the centre of mass (CM) of the entire assembly of
tube, springs, and the ball (see Figure 1). Then using
the lever rule

ml = (M + m)lCM : (1)

Here M is the mass of the assembly without the ball.

The CM is easily determined by balancing the assembly
on a knife edge or by suspending and balancing the as-
sembly horizontally using a thread. A simple balance is

RESONANCE  April 2005 77

 CLASSROOM

su±cient to determine (M + m). In the experiment it

was found that

ml = 296 g-cm: (2)

Slow rotation: The change in kinetic energy of the

system is
1 h i
¢Ks = m0 v2 + I! 2 + m(l2 + 2r 2=5)!2
2
1 h i
= m0 + I=R2 + m(l2 + 2r 2=5)=R2 v2: (3)
2
Here R is the radius of the rotating stub and I the e®ec-
tive moment of inertia (MI) of the whole system except
the ball. The radius of the ball is r. We have used the
parallel axis theorem for the MI of the ball. For no slip-
ping v = !R. As mentioned earlier we have ignored the
rotational kinetic energy of the pulley.
The mass falls through a height h and from the work-
energy theorem (or equivalently the principle of conser-
vation of mechanical energy) we have

¢Ug = m0gh
¢Ug = ¢K s
h = As v2 ; where
[m0 + I=R2 + m(l2 + 2r 2=5)=R2 ]
As = : (4)
2m0g
A close approximation to the instantaneous speed of the
ball v is obtained by a combination of photogate and
timer which measures the passage time of m0 across the
photogate after its fall through h. A plot of h versus v2
in the slow rotation regime ( v2 < 300 cm2-s¡2 ) is linear
as shown in Figure 3.
Fast rotation: In this regime static friction is over-
come. Since the springs are soft the ball goes to one end
of the tube and stays put. Its position with respect to

78 RESONANCE  April 2005

 CLASSROOM

the centre of the tube is L=2 ¡ r where r is its radius.

The change in kinetic energy is
1h i
¢Kf = m0 + I=R2 + m((L=2 ¡ r)2 + 2r2 =5)=R2 v 2:
2
(5)
The change in elastic energy is (Figure 1)

¢Ue =
1h i
k2(L ¡ 2r)2 ¡ k1 (L=2 ¡ l ¡ r)2 ¡ k2(L=2 + l ¡ r)2 :
2
(6)
As mentioned before we ignore the rotational kinetic
energy of the pulley as well as the work done against
kinetic friction. From the work-energy theorem, ¢Ks =
¢Ug + ¢Ue . This yields

h = Af v2 + B; (7)

where
1 h i
Af = m0 + I=R2 + m((L=2 ¡ r)2 + 2r2=5)=R2
2m0g
(8)
B = ¢Ue=2m0g =
1 h i
k2 (L ¡ 2r)2 ¡ k1(L=2 ¡ l ¡ r)2 ¡ k2(L=2 + l ¡ r)2 :
2m0 g
(9)
2
The plot of h versus v is once again a straight line but
this time with a ¯nite intercept on the h axis. There
is a distinct discontinuity when one crosses over from a
regime where static friction dominates to one where it
does not. Figure 3 depicts this. The crossover speed is
given by v2 = (h2 ¡ h1 ¡ B)=(Af ¡ As ).
Determination of mass of the ball: Employing
equations (4) and (8) we have
m
Af ¡ As = [(L=2 ¡ r)2 ¡ l2]: (10)
2m0gR 2

RESONANCE  April 2005 79

 CLASSROOM

Figure 3. The plot of the

‘instantaneous’ speed v 2 of
the mass m0 versus dis-
tance of its descent h. Note
that there are two clearly
defined regimes: ‘slow’ and
‘fast’ rotations. The cross-
over which occurs at v2
~330 cm 2–s–2 is reminis-
cent of a phase transition.
An understanding of this
plot is the key to unravel-
ing the mysteries of the
black box.

One can measure the length of the tube L, and the radius
of the ball is given as r = 1.1 cm. One obtains (L=2 ¡
r)2 = 338.6 cm2 . Further simple measurements yield
2m0gR2 = 756 £ 103 g - cm3 s¡2 . The unknown quantity
ml2 may be written using equation (2) as

ml2 = [(m + M )lCM ]2 =m

= (296)2=m:

Finally the slopes in Figure 3 yield Af ¡ As = 0.026 s2-

cm¡1. Inserting these numerical values in equation (10)
we obtain a quadratic equation for m.

338:6m2 ¡ 19655m ¡ 87616 = 0; (11)

where m is expressed in units of grams (g). The physi-

cally valid positive root yields m = 62 g.
Determination of the spring constants: When
the black box is in the horizontal position the force bal-
ance equation yields (see Figure 1)

k2 (L=2 + l) ¡ k1 (L=2 ¡ l) = 0: (12)

80 RESONANCE  April 2005

 CLASSROOM

Employing equation (9) we have

h i
k 2(L ¡ 2r)2 ¡ k1 (L=2 ¡ l ¡ r)2 ¡ k2 (L=2 + l ¡ r)2

= 2m0gB: (13)
Equations (12) and (13) constitute two linear simulta-
neous equations in the two unknowns k1 and k2. We
note that the numerical values of the remaining physi-
cal quantities in these two equations can be established.
From the intercept on the h axis in Figure 3 we have
B = 7.2 cm. The value of the `extension' l in the hori-
zontal position is estimated from equation (2) and from
the determination of the mass m = 62 g in the previous
subsection, namely

ml = 296 g-cm
l = 296=62
= 4:8 cm:

Note that the mass m0 = 100.4 g, the length of the

black box L = 39.0 cm, and the radius of the ball r
= 1.1 cm. The coe±cients and constants of the two
linear simultaneous equations in the two unknowns k1
and k2 are thus unambiguously speci¯ed. Solving these
we obtain k1 = 4:6 N/m and k 2 = 2.8 N/m.
Postscript
As mentioned earlier we discovered that the spring con-
stants could be determined by ingeniously employing the
results used to determine m. In the Olympiad problem
it was suggested that we measure the time period of
small oscillations of the black box suspending it from
either end. In e®ect, we treat it as a compound pendu-
lum. We note that it is not necessary to carry out this
elaborate observational exercise.
The black box experiment is uncommon at the higher
secondary school level. In the event it is posed, it tests

RESONANCE  April 2005 81

 CLASSROOM

Suggested Reading the student's knowledge of basic electronics. A box en-

[1] Vijay A Singh and
closes three to four electronic components (resistors, ca-
Manish Kapoor, The pacitors, diodes, etc). The student measures the voltage-
Magnetohydrodynamic current responses across the terminals and must identify
Generator: A Physics the components and the circuit enclosed by the black
Olympiad Problem 2001,
box. The mechanical black box experiment is a rarity.
Resonance, Vol.7, No.7,
pp. 68 - 75, 2002.
This is the ¯rst time in the thirty-¯ve year history of
[2] H C Pradhan, Report on the Physics Olympiad that the mechanical black box
the First National Physics has been posed as a full-°edged ¯ve hour experimental
Olympiad and the 29th problem.
International Physics Oly-
mpiad, Resonance, Vol .4, One can make an argument for including black box phys-
No.2, p. 104, 1999.
ics experiments at the school and university level. These
[3] Vijay A Singh and R M
Dharkar, The Interna-
experiments have a detective theme and can be excit-
tional Physics Olympiad - ing to students. Further they are in the same genre
1999, Physics News, Vol.30, as non-destructive testing. Oil logging and water table
pp. 60-64 , 1999. determination are examples. Non-invasive procedures
[4] Vijay A Singh, Ampere
such as ultrasound and MRI medical scans make similar
versus Biot-Savart, Reso-
nance, Vol.5, No.8, pp.84-
demands on us: given the sound or spin echoes, deter-
91, 2000. mine the underlying pathology. In fact a large num-
ber of research experiments, photoemission and Auger
spectroscopy to name just two, are simply sophisticated
versions of the black box experiment.
One Final Observation: One should be permitted
the luxury of guess{estimating the unknown variables
in any black box situation. Let us do so for the mass m
of the ball. We have been given the radius of the ball,
viz. 1.1 cm and also that kinetic friction is small. The
examination was conducted with around 200 students
at a given time. This required mass production of the
set up and it is reasonable to assume that the balls were
cast from a metal. The density of metals is around 10
g-cm¡3. From this data the mass is easily estimated to
be 56 g which is not too far from the experimental value
of 62 g!

82 RESONANCE  April 2005