The Brachistochrone Problem: the path of

the shortest time

Dhiman Roy

HL Mathematics

“The laws of nature are but the mathematical thoughts of God.” - Euclid

0
1 Introduction:
Isaac Newton is arguably the most brilliant man to walk on this planet, and he himself believed so
too. However, Newton’s arrogance was not baseless, as his contributions in the field of
mathematics and physics were tantamount. His invention of calculus and discovery of the laws of
gravity and motion was no ordinary feat.

When a famous mathematician, Johann Bernoulli, posed the brachistochrone problem to Newton,
he solved it in one night, mailed the solution anonymously and left a note with the solution, saying
he does not like the be pestered about mathematical things by foreigners.

The brachistochrone problem questions which path an initially stationary object should take to
reach from point A to point B, where point B is lower than point A, for the fastest descent while
only relying on gravity as a means of acceleration. The brachistochrone problem has many
applications in the real world – the path of shortest possible time a skier or glider should take to get
from point A to point B, or to design a roller coaster.

When I first came upon this question of fastest possible descent, I stated what seemed almost
obvious: “a straight path between the two points”. However, I could not have been more incorrect.
Therefore, the focus of this investigation is to find the brachistochrone curve, through an
experiment, and the analysis of the results. There are many websites and videos deriving the
cycloidal path (brachistochrone formulae) using partial derivatives. However, I found it difficult to
understand. Therefore, I decided to tackle the brachistochrone problem using simple physics
principles and calculus that I had previously learnt. In the future, I am hopeful I will be able to solve
the same problem using partial derivatives (which might be a quicker, more efficient).

The question is: what is the path of least time, and how long will the time taken to
reach the end of the path be?

The experiment will be carried out to show that the brachistochrone curve, or the curve of shortest
time, is the fastest path an object can take. This will be accomplished using calculus and Snell’s Law
and Fermat’s Principle from physics, that describes the path taken by light in different mediums.
The experiment carried out will prove that the cycloidal curve is in fact faster than the linear path.
However, since the effect of friction will slow down the object travelling across the path, I will also
use calculus to predict the time that the object will take to move from a point A to point B, in the
absence of friction.

The aim of this exploration is to calculate the actual and literature values for the time of descent of
the table tennis ball, down the linear and cycloidal curve. To calculate the actual value, an
experiment will be conducted, and the results of the experiment will be produced and graphed
using a scientific program. The literature value will be calculated by first proving that a cycloidal
path is the fastest path and using calculus to manipulate the parametric equations of the cycloid to
make time the subject. Achieving this aim will be relatively simple as I only need to talk about a few
variables, while finding ways to subtract out unwanted variables.

1
2 Experiment:
2.1 Procedure:

1. A cardboard was cutout to make 2 pathways, for a table tennis ball, of width 1.5 cm and the
distance between the end point B and the starting point (A) of 50 cm, as shown in figure (1). If Point
A, the starting point is considered as (0,0), then the co-ordinates for point B are approximately (30,
-40) in cm, as shown in figure (2). Point A and B will be referred to throughout the exploration as
the starting point of the decent and the ending point of the descent respectively.

Point A

Point B

Figure (1): Homemade cycloid and linear

path
2. A camera was positioned perpendicular to the paths, A 2 meters away and a ruler was
used as a reference for the distance the ball travelled while a stopwatch was used as a reference to
time, as shown in figure (3).

3. A video was recorded of the ball rolling down the 2 paths with a stopwatch besides the paths. A
scientific program called LoggerPro was used to analyze the movement of the ball, as shown in
figure (3). The center of the ball was mapped out for each frame since the ball was dropped. An
origin was laid out and the points the ball travelled were mapped out relative to the origin (the 2
yellow lines intersection). B
Figure (2): Coordinates of Points A and B

x-displacement
Origin

y-displacement

Stopwatch

Ball

Figure (3): Using LoggerPro

2
 2.2 Results:
The graphs in figure (4) and figure (5) show the displacement of the center of the ball as a function
of time. Table (1) and table (2) show the data collected using the LoggerPro program, and this data
will allow us to graph the ball’s displacement and understand its motion. LoggerPro calculates the
time difference between each frame, and the distance from the ball from the origin. Furthermore,
it graphs the line of best fit. The final x-axis displacement should be approximately 0.4 meters and
the y-axis displacement should be -0.3 meters. The ball took approximately 0.433 and 0.534
seconds to reach point B from point A using the cycloidal and linear path respectively. Figure (4)
and figure (5) below are used to model the x and y-displacement, which will be expressed in
parametric equations in terms of time t as the third variable. The data might be limited due to the
lack of data points, as the frames per second (FPS) of the camera I used
(iPhone 5s) was too low. A better camera (such as a Canon) with more FPS
would produce more data points, and leading to stronger data.

2.3 Linear Path:

Figure (4):x and y displacement for a linear path

B
Displ
ace x (t )
ment
in X
(m) A

Displ
acem y (t)
ent in
Y (m) B

Time (s)
The functions for the x and y displacements (from table (1)) are:

(2.2.1) x ( t )=29.4−12.66 t +1.363 t 2 , { 4.668 ≤t ≤5.202 }

(2.2.2) y ( t ) =−21.85+9.403 t−1.012t 2 , { 4.668 ≤t ≤ 5.202 }

Where t=4.668 is when the ball is released at point A and t=5.202 is when the ball reaches point
B. Functions y ( t ) and x ( t ) above are calculated through the LoggerPro program and have a
correlation coefficient of 1.000 and 0.9999 respectively. To change the equations, so that the ball is
released at t=0, we can substitute all the t with ( t+ 4.668 ) . This will give us the equations:

(2.2.3) s x ( t )=1.363 t 2 +0.064968 t+ 0.003195 , {0 ≤ t ≤ 0.534 }

(2.2.4) s y ( t ) =−1.012t 2−0.045032 t−0.008503 , { 0 ≤t ≤ 0.534 }

Where s x and s y is the displacement in the x and y-axis respectively.

3
Table (1): linear X-axis
and Y-axis displacement
 2.4 Cycloid Path:
Figure (5): x and y displacement for a cycloidal path

B
Displ
x (t )
acem
ent in
X (m)
A

Displ
acem
ent
y (t)
in Y B
(m)

Time (s)
The functions for the x and y displacements (from table (2)) are:
Table (2): linear X-axis
2
(2.3.1) x ( t )=25.42−15.77 t+2.446 t , {3.217 ≤ t ≤ 3.650 } and Y-axis displacement
(2.3.2) y ( t ) =−430.8+381.2 t−112.1 t 2+10.96 t 3 , {3.217 ≤ t ≤ 3.650 }

Similar to the linear displacement function, we can substitute t with ( t+ 3.217 ) to get:

(2.3.3) s x ( t )=2.446 t 2−0.032436 t+ 0.001782, { 0 ≤t ≤0.433 }

(2.3.4) s y ( t ) =10.96 t 3−6.32504 t 2 +0.226646 t +0.279015 , { 0≤ t ≤ 0.433 }

The correlation coefficient for the functions s x ( t ) and s y ( t ) is 0.9987 and 0.9998, respectively. Since,
the time taken to move from point A to point B, was very similar between the 2 paths, possibly due
to friction. I would have liked to carry out an experiment including a circular path and a quadratic
path, however due to time constraints, I was unable to do so. I would also like to have used a
variety of balls, however, I only had table tennis balls available. A more accurate cycloidal path may
increase the time gap, however due to various limitations already stated, data produced may be
weak. I wanted to calculate the theoretical time that will be taken by the table tennis ball from
point A to point B if there was no friction or air resistance was present. To do this, we must first
understand what a cycloid is and prove that a cycloidal curve is the path of fastest time.

3 Understanding Parametric Equations and Cycloids:

To analyze the 2 curves, we first need to understand what parametric 1 equations are. Parametric
equations are used to define more complex curves such as a circle or cycloid. Since it is difficult to
define a circular function in terms of either variable x or y in a Cartesian plane, we can use
parametric functions to define variables x and y in terms of a third variable called the parameter.

1
Paul’s Online Notes: https://tutorial.math.lamar.edu/classes/calcii/parametriceqn.aspx

4
 The equation for a circle can be written as:

(3.0.1) x 2+ y 2=r 2

However, if written in the parametric form, we can use 2 equations to define the function of a
circle, where both x and y are a function of t, which can be seen in Figure (6).

(3.0.2) x=x ( t )=r cos t and y= y ( t )=r sin t

¿
x (t)=t−sin t
y (t)=1−cos t
t

r
t

Figure (6): circles parametric equation Figure (7): A cycloidal path

5
A cycloid2 is a curve that is traced by a point P on the circumference of the circle, when rolling
(without slipping) along the x-axis3, as shown in Figure (7). By rolling the circle on the x-axis at an
angle t, under the no-slip conditions, from the bottom on the circle in contact with the x-axis, we
can change the parametric equations for point P to:

(3.0.3) x ( t )=−r sin t and y ( t ) =r cos t

If the circle rolls towards the positive side of the x-axis, the x-coordinates of the center of the circle
can be given by the distance travelled using the arc length equation:

(3.0.4) x=rt

where r is the radius and t is the radians the circle has travelled. Since we know the distance
travelled by the center of the circle in equation (3.0.4) and the x-coordinates of point P relative to
the center of the circle in equation (3.0.3), we can calculate the x values of the cycloidal curve by:

(3.0.5) x ( t )=rt +(−r sin t)=r ¿

Similarly, the y-coordinate of the center of the circle will always be equal to:

(3.0.6) y=r

The y-coordinate of point P relative to the center of the circle is given by equation (3.0.6),
therefore, the y values are:

(3.0.7) y ( t ) =r−r cos t=r (1−cos ¿t) ¿

Equations (3.0.5) and (3.0.7) are the 2 parametric equations for the cycloidal curve.

4 Solving for the Brachistochrone Problem:

Snell’s Law4 states that the sine of the angles of incidence times the refractive index is always
constant. Therefore, we can write that:

(4.0.1) n1 sin θ1=n 2 sin θ2=n3 sin θ 3=...=n n sin θn  ni sin θi=nn sin θ n

Where n1 and n n is the refractive index of medium 1 and medium n respectively. θ1 and θn are the
angles of incidence independent to the normal respectively. The normal is the line perpendicular to
the medium, as shown in figure (8). As light enters from one medium, such as air, into another
denser medium, such as water, the speed of light changes causing light to travel at a different

2
Cycloid: https://en.wikipedia.org/wiki/Cycloid
3
The Cycloid: http://mathonline.wikidot.com/the-cycloid
4
The Law of Refraction: https://www.math.ubc.ca/~cass/courses/m309-01a/chu/Fundamentals/snell.htm

6
angle. The refractive index for any medium is calculated by dividing the speed of light in a
respective medium by the speed of light in a vacuum:

c0 c c c c
(4.0.2) n1 = ,n 2= 0 , n3= 0 , … , nn= 0  ni = 0
c1 c2 c3 cn ci

Where n1 is the refractive index of the first medium, c 0 is the speed of light in a vacuum and c i is the
speed of light in medium i. This can be seen in the Figure (8) below, where light enters water from
air. Since the speed of light is slower in water than in air, the refractive index of water is greater
than air (using equation 4.0.2). This results in the angle of refraction (θ2 ¿ being smaller than the
angle of incidence (θ1 ¿.
Normal

Less dense
Air =

Water = More dense

Figure (9): movement of light

Furthermore, Fermat’s
Figure (8): Snell’s Law 5 tells us that light takes
principle the path
through thatmediums
multiple requires the least amount of
time. Therefore, if an object takes the path of least time, it can be said to follow Fermat’s principle.
From figure (9), we can see the path any object would take if it were following Fermat’s Principle
through successively less dense mediums ( n 1> n2> n3 ). Due to this change in density, Snell’s Law can
be applied to find the brachistochrone curve. If there were infinite mediums of infinitesimally small
width, the object passing through would form a smooth curve. This curve would be the path of least
time. To find the equation for the path of least time we will use the Snell’s Law equations above.

c0
Therefore, substituting equation (4.0.2) ni = into equation (4.0.1) ni sin θi=nn sin θ n :
ci

c 0 sin θ1 c 0 sinθ 2 c 0 sin θ3 c sin θn

(4.0.3) = = =…= 0
c1 c2 c3 cn

sin θ1 sin θ2 sin θ3 sin θ n

(4.0.4) = = =…=
c1 c2 c3 cn

sin θi sin θ n
(4.0.5) =
ci cn

This tells us the ratio between the size of the angle of incidence and the speed of light is equal in all
mediums, 1 to n. As the object passes through less dense mediums, the angle of incidence will

5
Fermat’s Principle: http://electron9.phys.utk.edu/optics421/modules/m1/Fermat's%20principle.htm

7
 π
increase. Since the angle of incidence with respect to the normal must be in the domain [ 0 , ¿, the
2
π
angle of incidence will approach radians if the object passes infinitely many mediums.
2

π
(4.0.6) lim (¿ θ n)= ¿
n→∞ 2

Therefore, we can simplify the equation above (4.0.4) to:

π
sin
(4.0.7) sin θi sin θn 1 2
=lim ( )= =
ci n →∞ cn cn cn

ci
(4.0.8) ∴ sin θi=
cn

Equation (4.0.8) tells the incidence angle,θi , in every medium i that it goes through. the incidence
angle(θi ) is given in terms of speed in medium i and final speed in the final medium n. Now we can
use energy equations from physics to understand the movement of the ball in the system. All we
must do is make velocity v the subject of the equations. The object is to roll down from point A to
point B, with no friction present and gravity as its only means of acceleration. If we assume that no
energy is lost due to the law of conservation of energy6 and the ball consists of only 2 types of
energy (kinetic and gravitational potential energy), we can write that:

(4.0.9) inital energy=K Ea +GPE ❑a=K Eb +GPE ❑b =final energy

The equation for kinetic and gravitational potential energy (from the IB physics data booklet) is:

1
(4.0.10) KE= m v 2
2

(4.0.11) GPE=mgh
Where KE is kinetic energy, GPE is gravitational potential energy 7, m is mass, g is acceleration
caused by earth’s gravity, y is the height on the Cartesian plain and v is velocity. Therefore, we can
write the initial energy and the energy at any point throughout descent as:

1 1
(4.0.12) m v 2a +mg y a= m v 2i + mg y i
2 2

Where i refers to the energy at any point i. Since the object is initially stationary, we can equate the
initial kinetic energy to 0 joules and divide by mass m on the left- and right-hand sides to get:

6
"Law Of Conservation Of Energy - Energy Education". Energyeducation.Ca, 2021,
https://energyeducation.ca/encyclopedia/Law_of_conservation_of_energy. Accessed 18 Feb 2021.
7
Ibdocuments.Com, 2021, https://www.ibdocuments.com/IB%20DOCUMENTS/Data%20and%20Formula
%20Booklets/Physics/Data%20Booklet%202016%20-%20English.pdf. Accessed 18 Feb 2021.

8
 1
(4.0.13) g y a= v 2i + g y i
2

1 2
(4.0.14) v =g y a −g y i=g ( y a− y i )
2 i

(4.0.15) vi =√ 2 g ( y a− y i )

ci
Let us substitute velocity of light c with velocity v, in equation (4.0.8) [sin θi= ]
cn

(4.0.16) vi =c i and v n=c n

c i v i √ 2 g ( y a− y i ) √ ( y a − y i )
∴ sin θi= = = =
c n v n √ 2 g ( y a− y n ) √ ( y a − y n )

To simplify equation (4.0.16), we can assume the initial point A to be at the origin (0,0) on the
Cartesian plane. If the initial height is 0, the final height must be less than 0 (negative) as gravity is
the only force acting on the object, causing it to accelerate downwards. Therefore:

(4.0.17) y a=0

(4.0.18) y i < y a , ∀ i=Z +¿¿

+ ¿¿

y i ∈Z −¿, ∀ i=Z ¿
(4.0.19)

(4.0.20) − y i=| y i|= y ∈ Z+ ¿¿

For simplicity’s sake, − y i will be written as y and − y n will be written as r as − y i and − y n are
positive numbers; y i∧ y n are is the y-axis co-ordinate of the object at a point i and point B,
respectively. As stated earlier, point B is the end point of the curve and point i is any point along
the path. Equation (4.0.16) can be simplified as:

(4.0.21) sin θi=

√ ( 0− y ) = √( − y ) = √ y
i i

√ ( 0− y ) √( − y ) √ r
n n
Let us imagine a medium i as shown in Figure (10), where θi is the angle of refraction, ds are the
total displacement, dx is the displacement in the x-direction and dy is the displacement in the y-
direction.

θi
ds Figure (10): Medium i
dy

dx
9
Through Pythagoras’s Theorem, we know that:

(4.0.22) ds=√ ¿ ¿

To calculate the parametric equation for y, we must simply make state y in terms of θ. This can be
done by manipulating equation (4.0.21) sin θi=
√ y and using the right-angle triangle above:
√r
dx dx
(4.0.23) sin θi= =
ds √ ¿¿ ¿

y y
(4.0.24) dx= √ × ds= √ × √ (dx)2 +(dy )2
√r √r
(4.0.25) ¿

We can rewrite this equation as:

(4.0.26) (1− yr )(dx) =( r−r y )( d x ) = yr (dy )

2 2 2

(4.0.27) ¿

y
(4.0.28) dx=
√ r− y
dy

dx y
(4.0.28)
dy
=
√
r−y

Using Figure 10 above, we can say that:

dx y
(4.0.29) tan ( θ i )=
dy √
=
r−y

sin2 θi y
(4.0.30) 2
=
cos θi r− y
r sin θi− y sin2 θi= y cos2 θi
2
(4.0.31)

(4.0.32) y¿

(4.0.33) y=r ¿

This can be written as,

r
(4.0.34) y= ¿
2

10
Therefore:
r
y= ( 1−cos 2 θi )
2

Now to calculate the parametric equation for x .

dx dx dy
(4.0.35) = ×
dθ dy dθ

dy d
(4.0.36) = ¿
dθ dθ

By substituting equations (4.0.36) and (4.0.29) into equation (4.0.35), we can write:

sin2 θi
(4.0.37)
dx
dθ
=
y
√
r− y
× 2r sin θi cos θi =
√
cos 2 θi
× 2r sin θi cos θi

sinθ
¿ ×2 r sin θ i cos θ i=2 r sin 2 θi
cosθ

Similar to equation (4.0.34), we can rewrite equation (4.0.37) as:

dx
(4.0.38) =2 r sin2 θ i=r ( 1−cos 2θi )
dθ

(4.0.39) x=∫ dx=∫ r ( 1−cos 2 θi ) dθ

rsin2 θi
x=∫ r −r ( cos 2 θi ) dθ=r θ i−
2

Therefore, the parametric formula for the path should be the following:

r r
(4.0.40) x= ( 2 θi−sin 2 θi ) , y= (1−cos 2 θi)
2 2

Where r is the radius and θ is the angle in radians with respect to the x axis. Using the derivatives of
the cycloid parametric equations above, we can calculate the literature time taken to descend the
path, in the absence of friction.

5 Finding the travel time:

5.1 Brachistochrone Path:

Since we now know the parametric equations of a brachistochrone curve, as have proved that the
cycloidal curve is indeed the path of shortest time, let us use it to find the theoretical time taken for
the table tennis ball to move from point A to point B in the presence of no friction. Using calculus as

11
a mathematical approach to carry out this exploration allows us to make assumptions about the
initial and final location of the ball on the cycloidal path. However, I read a paper on solving the
brachistochrone using partial derivatives, and the paper achieved this goal in under 4 pages! This
leads me to think that partial derivatives is a more suited mathematical approach to this problem.

To calculate the time to descent, let us use the energy equations from section 4 again. Due to the
law of conservation of energy, we know that the initial energy and the final energy must be equal,
as shown in equation (4.0.9). Since the object starts from instantaneous rest, the kinetic energy is 0
joules initially. If we assume that all the gravitational potential energy the ball initially had is
converted into kinetic energy, we can write that:

1
(5.1.1) m gy = mv 2
2

1
(5.1.2) gy= v 2
2

ds
(5.1.3) v= =√ 2 gy
dt

Now we must make time t the subject of the equation, since that is what we are trying to find.

ds ds
(5.1.4) dt= =
v √ 2 gy

(5.1.5) T =∫ dt

Where T is time taken to travel, and dt is a small change in variable t in the parametric equations
for the cycloid curve. In section 4, the parametric equations are written in terms of angle θ;
however, in this section, angle θ will instead be written as angle t.

Using equation (4.0.22)ds=√ ¿ ¿, we can rewrite equation (5.1.4) as:

dy 2 dy 2
(5.1.6)
dt=
√ dx 2+ dy 2 =
√ 2 gy
√ (
dx 2 1+

√ 2 gy
dx) √(
2
=
1+
dx
√2 gy
2 )
dx

dy 2 dy 2
(5.1.7)
xb

∫ dt=∫
xa
xb

xa
√(1+ 2
dx
√ 2 gy
)
dx=
1
∫
x

√2 g x
b

a
√( 1+ 2
dx
√y
) dx

Where x aand x bare the x coordinates for point A and point B on the cycloidal path, respectively.
Now we need the differential equations of the parametric cycloid equations (3.0.5) and (3.0.7).

dx
(5.1.8) x (t)=r ( t−sin t )  =r (1−cos t)
dt
12
 dy
(5.1.9) y (t)=r ( 1−cos t ) =r sin t dy=r sin t dt
dt

dy dy dt 1 sin t
(5.1.10) = × =( r sin t ) =
dx dt dx r ( 1−cos t ) 1−cos t

If we square both sides, and write the derivate in terms of cos t we get:

(5.1.11) (dy )2
¿¿

By rewriting equation (3.0.7) y ( t ) =r (1−cos ¿t) ¿, we can write cos t in terms of y.

y
(5.1.12) cos t=1−
r

y
1+1−
(dy )2 1+ cos t r 2r − y y
(5.1.13)
(dx )2
=
1−cos t
=
y
1−(1− )
r
=
y
 dx=
2 r− y
dy
√
dy 2
Substituting equation (5.1.13) into equation (5.1.6)
dt=
√( 1+
dx 2
√ 2 gy
dx
)
results in:

2 2

(5.1.14)
dt=T =
1
∫
xb

√2 g xa
√ (√
1+
2 r− y

√y
y ) dx= 1
∫
√2 g y
yb

a
√ (√
1+
2 r− y

√y
y )×
√ y
2r− y
dy

2r
¿
1
∫
√2 g y
yb
√ √
√y
y
×
y
2r− y
dy=
1
∫
y
2r
×
1
b

√ 2 g y y √ 2 r− y √
dy=
1
∫
y
2r
√2 g y y ( 2 r− y )
dy
b

√
a a a

The limits of the definite integral changes from x-displacement x a∧x b ¿ y -displacements y a∧ y b as
we integrate dt with respect to y. If we substitute y with the parametric equation of a cycloid, we
get:

xb xb
1 2r 1 2r
(5.1.15) T= ∫
√2 g x y (a
2 √
r− y )
dy= ∫
√2 g x r ( 1−cost ) (2 √
r−r ( 1−cost ))
a
dy

xb xb
1 2r 1 2r
¿ ∫ √
√2 g x ( r−rcost ) (r +rcost )
a
dy= ∫
√2 g x r −r 2 cos2 t
2
a
dy
√
13
 xb xb
1 2r 1 √ 2r dy
¿ ∫ 2
√
√2 g x r sin t
a
dy= ∫
√ 2 g x r sin t
a

Using equation (5.1.9) dy =r sint dt , we can calculate T as an integral with respect to t.

tb tb
1 √ 2 r r sin t dt =¿ 1 √ 2r dt ¿
(5.1.16) T= ∫ ∫
√ 2 g t r sin t
a
√2 g t a

The limits of the definite integral changes to be expressed in the terms of variable t instead of
variable y. If we allowt a=0 (since the curve starts at an angle of 0 radians with respect to the
normal) and t b=t , then we can simplify the equation to:

t √r
(5.1.17) T=
√g
Where t and r are angle and radius variables for the parametric equation of the cycloid curve and g
is the acceleration caused by gravity. Using this equation, we can calculate the theoretical time that
the table tennis ball should have taken down the cardboard path.

Using the measurements of the cardboard slope from section 2, the brachistochrone curve must
pass through points A (0, 0) and B (0.4, -0.3) (units in meters). Since point B has negative y
coordinates, we can graph the cycloid with a positive y coordinate and reflect it across the x-axis.
Therefore, the parametric equations are:

(5.1.18) 0.4=r ¿

Now we have simultaneous equations, which we will solve by equating r in terms of t.

0.4 0.3
(5.1.19) r= =
t−sin t 1−cos t

(5.1.20) 0.4−0.4 cos t=0.3 t−0.3 sin t

r

Figure (11): solving

simultaneous equations

14 r =0.4−0.4 cos t

r =0.3 t−0.3 sin t

t
The easiest way to calculate t was to graph the individual functions (using desmos8 graphing
calculator) and find the intersect. Using a graphing calculator in Figure (11), we know that t=2.88.
Now we can solve for r:

0.3
(5.1.21) r= =0.153
1−cos 2.88

Now we can plot the brachistochrone curve using the t and r values calculated above. In Figure (12)
below, we can see what a graphical cycloidal curve, that was used for the experiment in section 2,
would look like.
Figure (12): graphing a cycloid
y-axis

x ( t )=0.153¿
y ( t ) =0.153 ¿

x-axis

x ( t )=0.153¿
y ( t ) =−0.153 ¿

Using the values of t and r of the cycloidal curve in figure(12), we can calculate the time of descent:

t √ r 2.88 √ 0.153
(5.1.21) T= = =0.360 seconds.
√g √ 9.81
Therefore, we can conclude that the literature value for the time taken to completely descend the
cycloidal path would be 0.360 second s. We must now find the time taken to completely descend a
linear path.

5.2 Linear Path:

Similarly, we can estimate the time the ball would take to travel the linear path using simple
kinematics. Let us model the linear cardboard path as shown in Figure (13) below (not to scale):

A( 0,0) Ball

Overall acceleration Figure (13): kinematics

0.3 gravity
8
"Desmos | Beautiful, Free Math". Desmos.Com, 2021, https://www.desmos.com/.

15
 B(0.4 ,−0.3)
0.4

Since we know that:

0.4
(5.2.1) sin θ=
√ 0.4 2+ 0.32
0.4
(5.2.2) θ=arcsin
(√
0.4 2+ 0.32 )
=¿ 0.927 rad ¿

Furthermore, we know that the overall acceleration is caused by the summation of all the forces
acting on the ball. Since, we assume there if no friction, gravity is the only forces acting on the ball,
where gravity in acceleration towards the center of the earth. We can redraw the forces, as shown
in Figure (14), by using the components of gravity9 that are parallel and perpendicular to the slope.
The forces are as follows (not drawn to scale):

Figure (14): gravity and acceleration Figure (15): simple trigonometry

0.927 rad

Ball Gravity Parallel

Slope
Total gravity
=9.81 m/s2
Gravity Parallel = acceleration Gravity Perpendicular
Gravity perpendicular

Total Gravity

Therefore, using figure (15) and simple trigonometry, we know that:

(5.2.3) Gravity parallel ¿ a=TG × cos( 0.927¿)¿

Where a is the acceleration and TG is the total gravity. Therefore, we know that acceleration is
5.89 m/s2. The equation for time taken in physics (from the IB data booklet) is as follows:

1
(5.2.4) s=ut+ a T 2
2

9
Khan Academy: https://www.khanacademy.org/science/high-school-physics/two-dimensional-motion-2/inclined-
planes/v/inclined-plane-force-components

16
Where u is the initial velocity (which equals to 0 as the ball is initially stationary), s is the distance
travelled from point A to point B and T is the time taken for the ball to reach point B from point A.
If we make T the subject of the equation we can write:

2s 2(0.5)
(5.2.5) T=
√ √
a
=
5.89
=0.412 seconds

Table ( 3 ) :comparing literature∧actual values ¿ time of descent

Time taken to move from

Linear Path Brachistochrone Path
point A to point B
Theoretical (absence of
0.412 0.360
friction) (seconds)
Actual (presence of friction)
0.534 0.433
(seconds)
Difference in actual and
0.122 0.073
theoretical (seconds)
Percent difference between
the theoretical and actual time
(%) 22.9% 16.9%
|t heoritical−actual|
( ×100 ¿
actual

I believe the difference between actual and theoretical time taken is due to imperfections along the
path, which may lead to a less smooth movement and greater friction. Furthermore, the deviation
in results due to air resistance must be small, as the table tennis ball has a small, smooth surface
area. Air resistance increases and surface area and the roughness of the surface increases. It is
interesting to see how the time of descent varies between the 2 curves. However, I find it intriguing
that the actual and theoretical time taken varies to different degrees.

6 Conclusion:
6.1 Analysis:

Interestingly, the brachistochrone curve is the path of shortest time, both, in the absence and
presence of friction. However, what I found even more surprising is the fact that the degree to
which friction slowed down the ball on the brachistochrone curve (16.9%) was smaller than the
degree to which friction slowed down the ball on the linear path (22.9%). I would have believed
that friction would slow down the ball on the brachistochrone curve to a greater extent than the
linear path, as the ball is travelling a longer distance (linear path is the shortest distance) and at a
higher speed (since the ball reaches point B first). In physics, it is known that the faster an object
travels, the greater the effects of air resistance. For some reason, that does not seem to be the case
for this experiment.

17
However, it is extremely explicit that the brachistochrone curve is faster, as presented by
theoretical and actual data. In the cycloidal path from figure (5), we can see that the initial rate of
change of displacement in y was large before the gradient became less steep. For the linear path,
the instantaneous rate of change of displacement increased as time increased, which can be seen
from the increasing gradient. I believe it is due to the greater initial steepness of the cycloidal path
that allowed the ball to accelerate faster and reach end point B quicker.

6.2 Limitations:

The degree to which the cycloidal cardboard curve is accurate is unknown as cardboard is difficult
to cut and may leave some uneven edges which will slow the ball down. For the experiment
conducted in section 2, the cycloidal curve was cut out 2 times (initially with cardstock). However,
cardstock was though to cut through and produced an uneven, rough curve, which resulted in
anomalous and inaccurate data. Cardboard was easier to cut through and did not bend during the
experiment. If I had more time and resources, the degree to which the cycloidal curve is accurate
could have been increased. I should have conducted the experiment not only with different balls,
but also a greater variety of curves. However, time constraints only allowed me time to cut out 2
cardboard paths, which were thin; this resulted in only table tennis balls being able to balance on it.

In addition, the ball may not have been in contact with the paths the whole time as it may have
bounced. This might have drastically slowed down the ball to an unknown degree. The final
coordinates of the ball at point B where not exactly (0.4, -0.3) and this may have been the effect of
the ball bouncing slightly on the path. Nonetheless, the experiment has many strengths which are
discussed in the next section.

6.3 Strengths:

It is impossible to completely remove friction as a variable, in any experiment, however, I do

believe friction played a very small role in mine. The surface of the table tennis ball and the
cardboard is extremely smooth, and the fact the ball rolls, should decreases the friction. The small
size of the table tennis was also a strength as the air resistance would be larger on larger balls (such
as a tennis ball).

During the experiment, the independent variable was the type of path the ball takes while the
dependent variable was the displacement of the ball. Since I used a scientific program (LoggerPro),
to measure the displacement, the data was considerable accurate, with 4 significant figures. All the
data points had a correlation of greater than or equal to 0.9987, which is an extremely strong
correlation. This allowed me to have high confidence in my data.

Furthermore, there were many variables that where controlled to prevent the generation of
anomalous or incorrect data. The same table tennis ball was used to prevent air resistance or
friction from varying in-between different experimental runs. Furthermore, the experiment was
carried out at the same location, as gravity differs throughout different locations around the world.

18
Since Singapore is at the equator, the acceleration due to gravity 10 is around 9.798 m/s 2; however, I
decided to use 9.81m/s 2 as it is the average value of gravity between the equator and the poles of
the earth. Since location is a constant variable, atmospheric pressure and wind will be constant,
which will lead to less inaccuracies in the data, and will produce reliable data 11.
6.4 Improvements and extensions:

To further improve the accuracy of my data, I would use a plastic to make the path, as the smooth
surface and even curves of the malleable material may yield better data. Furthermore, using
marbles or even smaller balls instead of table tennis balls may also yield better data, as a smaller
surface area will cause the air resistance to be smaller, allowing a faster time of descent.

This topic also has many extensions as the brachistochrone curve has many surprising properties,
such as a property called isochronism. Isochronism means that no matter where the ball is left on
the curve, it will reach the bottom of the curve at the same time. Furthermore, I would like to
understand why friction slowed the ball on the linear path to a greater extent than the ball on the
brachistochrone curve.

If I had more pages to carry out this exploration, I would have derived the displacement functions in
section 2 to calculate the instantaneous velocity of the ball at certain points and compared it to the
literature value of the instantaneous velocity. It would be interesting to see at which point the ball
has the greatest acceleration or velocity.

6.5 Conclusion:

Before this exploration, I always thought of math as difficult to apply to real life and physics as
somewhat uninteresting. I never really saw the connection between physics and math except math
being a means to solve physics equations. However, through this exploration, I have realized that
math can be easily applied to a great deal of problems, and problem that sound simple, can be
extraordinarily difficult to solve. However, through this exploration, I now have greater confidence
in my abilities in calculus, especially in kinematics.

The brachistochrone problem has many solutions. One of the simpler ones is the one I have used;
however, Isaac Newton solved the problem is a different way, using partial derivates. Although, I
tried to understand his method, I found it much too difficult, not only to understand but to explain.
Therefore, I decided to solve the mathematical problem using physics principle I had recently learnt
in physics class: Snell’s Law and Fermat’s Principle. However, I do hope that I will be able to
understand and solve this problem using partial derivates, sooner than later.

10
"Do I Weigh Less On The Equator Than At The North Pole?". Science Questions With Surprising Answers,
2021, https://wtamu.edu/~cbaird/sq/2014/01/07/do-i-weigh-less-on-the-equator-than-at-the-north-
pole/#:~:text=Using%20Newton's%20law%20of%20gravity,force%20also%20varies%20with%20latitude.
Accessed 5 Mar 2021.
11
Experiments, Validity. "Validity Vs Reliability Vs Accuracy In Physics Experiments". Matrix Education, 2021,
https://www.matrix.edu.au/the-beginners-guide-to-physics-practical-skills/physics-practical-skills-part-2-
validity-reliability-accuracy-experiments/.

19
7 Bibliography:
1. Admin. “Fermat's Principle - Observations & Applications: Laws of Reflections.” BYJUS, BYJU'S, 15
Oct. 2020, byjus.com/physics/fermats-principle/.
2. “Brachistochrone Curve.” Wikipedia, Wikimedia Foundation, 27 Dec. 2020,
en.wikipedia.org/wiki/Brachistochrone_curve.
3. “Brachistochrone Problem.” From Wolfram MathWorld,
mathworld.wolfram.com/BrachistochroneProblem.html.
4. “Law of Reflection.” Reflection and Fermat's Principle, hyperphysics.phy-
astr.gsu.edu/hbase/phyopt/Fermat.html.
5. “Parametric Equation of a Circle.” Parametric Equation of a Circle - Math Open Reference,
www.mathopenref.com/coordparamcircle.html.
6. “Parametric Equations, Polar Coordinates, and Vector-Valued Functions.” Khan Academy, Khan
Academy, www.khanacademy.org/math/ap-calculus-bc/bc-advanced-functions-new.
7. Partial Differential Equations. www.math.toronto.edu/ivrii/PDE-textbook/PDE-textbook.pdf.
8. Snell's Law -- The Law of Refraction, www.math.ubc.ca/~cass/courses/m309-
01a/chu/Fundamentals/snell.htm.

20
21