CHAPTER 10

PHYSICAL TREATMENTS OF SOME MATHEMATICAL PROBLEMS

Mathematics is the study of the relationships between numbers, quantities and space, whereas
physics is a science that deals with laws of nature. For a long time these two disciplines have been
closely related like twins, as evidenced by the fact that they exert strong influence on each other
through their course of progress. On one hand, mathematics provides powerful tools for fuelling the
advancement of physics. On the other hand, some physical laws offer fresh perspectives for solving
mathematical problems in ingenious ways.

In this note we would like to focus on how physics can be applied to solve challenging
mathematical problems neatly. Hopefully the reader can appreciate the elegance of the physical
treatments and how well mathematics is consistent with the laws of nature.

It should be noted, however, that most of the physical arguments presented in this note are
intuitive in nature and must not be deemed the correct way of problem solving. Rather they had better
be seen as illuminating ways of predicting answers or seeking the right lines of attack. The reader is
therefore advised to refer to the rigorous mathematical solutions of each example in this note.

1. Center of Mass

Center of mass is one of the important notions in statics. It is also a useful tool in tackling
geometric problems.

Center of mass is, by definition, the position of the object concerned through which its weight acts,
no matter how the object is positioned. In this section, we are more interested in center of mass of
discrete particles (i.e. point masses, which have mass but no size).

For two particles of masses m1 and m 2 placed at A and B respectively, their center of mass lies
on AB and divides AB in the ratio m2 : m1 . In general, the center of mass of particles of masses m1 ,
m2 , …, mk at ( x1 , y1 ), ( x2 , y2 ), …, ( xk , yk ) is ( xM , yM ) where
k k

∑ mi x i ∑m y i i
xM = i =1
k
and y M = i =1
k
.
∑m
i =1
i ∑m
i =1
i

Page 1 of 11
 Mathematical Database

¾ We use A(a) to denote the particle of mass a at A.

Example 10-1.1
In ∆ABC (Figure 1), D and E are points on BC and AC respectively such that BD:DC = 1:1 and AE:EC
= 2:1. AD cuts BE at F. Find AF:FD.
A

B D C

Figure 1

Solution.
Put three particles of masses 1, 2 and 2 at A, B and C respectively so that D is the center of mass of
{B(2), C(2)} and E is the center of mass of {A(1), C(2)}. Then the whole system {A(1), B(2), C(2)} can
be reduced to {A(1), D(4)}. It follows that its center of mass lies on AD. Similarly we have the center
of mass on BE. So, F is the center of mass of the whole system. Consider {A(1), D(4)}. We conclude
that AF:FD = 4:1

Example 10-1.2 (Ceva’s Theorem)

Prove that in ∆ABC (Figure 2), if D, E and F are on BC, AC and AB respectively such that AD, BE and
AF BD CE
CF are concurrent, then =1.
FB DC EA
A

a f

E
F

b O e

B c D d C

Figure 2

Page 2 of 11
 Mathematical Database

Solution.
Put three particles of masses be, ea and bf at A, B and C respectively. Then F and E are the centers of
mass of {A(be), B(ea)} and {A(be), C(bf)} respectively. Also, the intersection of CF and BE is in fact
the center of mass of {A(be), B(ea), C(bf)}. Denote this point by O. Therefore, it can be seen that the
center of mass of {B(ea), C(bf)} is the intersection of AO and BC, i.e. D.
ea d ace
∴ = ⇒ = 1,
bf c bdf
or in other words
AF BD CE
= 1.
FB DC EA

Center of mass can be used to solve algebraic problems, as illustrated in the following examples

Example 10-1.3
n
Evaluate ∑ k (a + k − 1) .
k =1

Solution.
Dissect the equilateral triangle ABC into smaller congruent equilateral triangles as shown in Figure 3.

n=5

A
x
a a+n-1

Figure 3

Page 3 of 11
 Mathematical Database

Put a particle of unit mass at each vertex of the smaller equilateral triangle. Let the x-coordinate of the
particle at A be a and those of the particles on BC be a + n – 1. Intuitively, we assert that the center of
mass of the configuration coincides with the centroid of the triangle ABC. Considering the x-coordinate
of the center of mass of the whole system, we have
n

∑ k (a + k − 1) 2
k =1
n
= a + (n − 1)
∑k
3
k =1
n
n(n + 1) 2
⇒ ∑ k (a + k − 1) = [a + (n − 1)] .
k =1 2 3
Putting a = 1, we obtain the well-known formula of sum of squares of the first n positive integers:
n
n(n + 1)(2n + 1)
∑k
k =1
2
=
6
.

Example 10-1.4 (Chebyshev’s Inequality)

If 0 < a1 < a 2 < ... < a n , 0 < b1 < b2 < ... < bn , then

n∑ k =1 ak bk >
n
(∑ n
k =1
ak )( ∑ n
k =1 k )
b .

Solution.
Let A1 , A2 ,..., An be collinear and their coordinates on the line be a1 , a2 ,..., an respectively. Put at
M
, where M = ∑ k =1 bk . The center of mass of the system is
n
A1 , A2 ,..., An identical particles of mass
n
1 n
∑ ak .
n k =1
Consider another system obtained by putting particle of mass bk at Ak . Then its center of mass is
1 n
∑ ak bk .
M k =1
Now that b1 < M
n < bn , there exists i such that b1 < Mn < bi but bi +1 < M
n . The first i particles of the first
system are heavier than those of the second system. At the same time the total masses of the two
systems are the same. Thus the center of mass of the first system should be ‘biased’ to the left.
1 n 1 n
∴ ∑
n k =1
ak <
M
∑a b
k =1
k k ,

i.e.

n∑ k =1 ak bk >
n
(∑ n
k =1
ak )( ∑ n
k =1 k )
b .

Page 4 of 11
 Mathematical Database

2. Systems of Forces

Theorem 10-2.1 (Lami’s Theorem)

If three coplanar forces F1 , F2 and F3 are in equilibrium, their lines of action are concurrent and
F1 F F
= 2 = 3
sin α sin β sin γ
where α , β and γ are the angles between the lines of action of F2 and F3 , F1 and F3 , and,
F2 and F1 respectively.

Theorem 10-2.2
If the resultant of a system of forces passes through A, B, C, …, then A, B, C, … are collinear.

The two theorems above can be used to solve geometric problems involving concurrence and
collinearity.

Example 10-2.1 (Ceva’s Theorem)

sin α sin β sin γ
In ∆ABC (Figure 4), = 1 if and only if AD, BE and CF are concurrent.
sin α ' sin β ' sin γ '

γ
-c β' b

E
F
-b
c O

α α'
γ' β
B C
a D -a

Figure 4

Solution.
‘If’ part: let two forces a and c act through BC and BA respectively such that their resultant acts
through BO. Then
| c | sin α =| a | sin γ ' ,
i.e.

Page 5 of 11
 Mathematical Database

sin α | a |
= .
sin γ ' | c |
Similarly, let − c and b act through AB and AC respectively such that their resultant acts through
AO. Then
sin γ | −c | | c |
= = .
sin β ' | b | | b |
Let − a and − b act through CB and CA respectively. Now, since the system of forces is in
equilibrium, the resultant − a + (−b ) must act through O. Then
sin β | −b | | b |
= = .
sin α ' | − a | | a |

sin α sin β sin γ | a | | b | | c |

= = 1.
sin γ ' sin α ' sin β ' | c | | a | | b |

‘Only if’ part: We can always find the system of forces mentioned above such that

sin α | a | sin β | b | sin γ |c |

= , = , = .
sin γ ' | c | sin α ' | a | sin β ' | b |

Since the system is in equilibrium, the lines of action of the three resultants are concurrent.

Example 10-2.2
In ∆ABC (Figure 5), the exterior angle bisector of A and the interior angle bisectors of B and C cut
BC, AC and AB at D, E and F respectively. Prove that D, E and F are collinear.

F E
c

B D
a C

Figure 5

Solution.
Let three forces a , b and c of equal magnitude act through BC, AC and BA respectively. We can

Page 6 of 11
 Mathematical Database

see that the resultant a + c acts through BE. Moreover, b acts through AE, so a + b + c acts
through E. Similarly, we can prove that a + b + c also acts through D and F. Thus D, E and F are
collinear.

Example 10-2.3
Four directed lines, no three of which are concurrent, intersect each other at six points called vertices.
Two vertices are said to be opposite if they are not on the same line (so there are altogether three pairs
of opposite vertices). For each pair of opposite vertices, a pair of angle bisectors is drawn as shown in
Figure 6 and their intersection point is found. Prove that these three intersection points are collinear.

Figure 6
The four directed lines are represented by the four thin lines with arrows, while the six vertices are the unlabelled points.
The dashed lines are the angle bisectors, while A, B and C are the three (collinear) intersection points.

Solution.
Four forces a1 , a 2 , a3 and a 4 of equal magnitude are arranged along the directions of the four lines.
Considering their resultant and using similar arguments to those in the previous proof (more precisely,
the resultant of these four forces must pass through all of A, B and C), we get the desired conclusion.

3. Stable Equilibrium and Potential Energy

A system is said to be in stable equilibrium if, after the system is subject to slight disturbance,
equilibrium can be restored. As shown Figure 7, particle A is in stable equilibrium while particle B is in
unstable equilibrium.

Page 7 of 11
 Mathematical Database

A
Figure 7

There is an important physical law about stable equilibrium, which is stated below.

Theorem 10-3.1 (Dirichlet’s Principle)

If a system is in stable equilibrium, its potential energy is minimal.

Example 10-3.1
AX XB
A and B are on two sides of a line l. Find a point X on l such that + is minimal.
p q

Solution.
Arrange the two smooth, light fixed pulleys A and B and a light smooth horizontal wire l on the same
vertical plane (Figure 8). A small smooth ring X which can freely slide on l, is connected by two light
1
inextensible strings, which pass through A and B and hang two particles A’ and B’ of masses and
p
1
respectively. It can be shown experimentally that the equilibrium of the system is a stable one.
q
When the stable equilibrium is reached, the potential energy of A ' and B ' is minimal. Then
AA' BB' AX XB AX XB
+ is maximal and + is minimal. In other words, + attains its minimum
p q p q p q
when X is at the equilibrium position. In equilibrium, the horizontal components of tensions in the two
strings balance each other.
sin α sin β
∴ = .
p q
We can use the above equality to locate this equilibrium position.

Page 8 of 11
 Mathematical Database

X l

A'

B'

Figure 8

Example 10-3.2
There are respectively 40, 50 and 60 children in village A, B and C. The villagers propose that a school
be built for these children. Where should this school be located so that the total distance traveled by the
children from their own villages to the school is minimal?

Solution.
On a wooden board drill three holes representing the locations of the three villages A, B and C (Figure
9). Three identical light inextensible strings pass through A, B and C and hang three particles of masses
4, 5 and 6 at one end respectively, with the other ends being joined together on the board. Denote this
meeting point of the three strings by X. Release the system from rest and it will reach a stable
equilibrium. Since the potential energy of the three particles is minimal when in equilibrium, it can be
shown (why?) that 4XA + 5XB + 6XC is also minimal when X is at its equilibrium position, which is
therefore the required location of the school. By Lami’s Theorem, X should be a point such that
4 5 6
= = .
sin ∠BXC sin ∠AXC sin ∠AXB

Page 9 of 11
 Mathematical Database

X
C
B

Figure 9

¾ The Fermat Point O of ∆ABC (i.e. the point that the sum of its distance from the three vertices is
minimal) can be found by using three identical particles. By Lami’s Theorem, O should be such
that ∠BOC = ∠AOC = ∠AOB = 120 0 .

Example 10-3.3
Prove that the tangent of an ellipse at A makes equal angle with F1 A and F2 A , where F1 and F2
are the two foci of the ellipse.

Solution.
Fix the two ends of a light inextensible string at F1 and F2 on a vertical wall. A light smooth ring A
can freely slide through the string. The locus of the ring is an ellipse if it moves in such a way that the
string is held taut. (See the left of Figure 10) If the system is released from rest, it reaches a stable
equilibrium, when the potential energy of the ring is minimal. Thus the equilibrium position of the ring
is the lowest position of the ellipse. It follows that the tangent of the ellipse at A is a horizontal line. On
the other hand, the equilibrium of the ring requires the horizontal component of the tension of the string
to vanish. This can only be achieved when F1 A and F2 A make equal angle with the horizontal (note
that the tensions in F1 A and F2 A are equal in magnitude).

A
F2

F1 F2
F1

A
Figure 10

Page 10 of 11
 Mathematical Database

Alternative Solution.
Let a small smooth ring A freely slide through a wire in the shape of an ellipse. A light elastic string
passes through A, with its two ends fixed at the two foci of the ellipse F1 and F2 . (See the right of
figure 10) When the ring moves along the wire, the string does no work to the ring as the length of the
string remains unchanged.
Consider the tensions in F1 A and F2 A . Since they are equal in magnitude, their resultant bisects
∠F1 AF2 . Note that this resultant does no work to the string. Thus its line of action is always
perpendicular to the direction of movement of the ring, which is also the tangent of the ellipse at that
point.

Page 11 of 11