How do different heights affect the bounce numbers of a table tennis ball

after it is dropped?
Introduction
When a ball is raised to a height above ground level, it gains potential energy.
Releasing this ball will transform the potential energy into kinetic energy. Contacting the
ground creates an impulse causing the ball to bounce back up. By simply referring to
this theory as well as intuition, there must be a relationship between the height released
and the number of bounces by a ball.
Research Question:
How do different heights affect the number of bounces of a table tennis ball after it
is dropped until it comes to rest?
Hypothesis
If the table tennis ball is released at a higher height, higher speed leads to stronger
impulse, then the more bounces a ball will have before coming to rest.
Independent variable:

Height
The height will vary as 0.5m, 0.9m, 1.3m, and 1.7m, increasing by 0.4m each time. Under
each variation there will be five measurements taken to reduce random error.
Dependent variable:

Number of bounces
There will be 2 people counting the number of bounces and the final result for each trial
will be the average of the two numbers.
Controlled variables:

Table Tennis Balls

Instead of using different types of balls, table tennis balls are used exclusively
throughout the experiment

Bounciness
To further control the experiment, all the table tennis balls are the same brand (Yaping),
and also the same model (AS7); they are also brand new to ensure the table tennis balls
would not potentially lose its bounce from time.

Method
To ensure no extra downward force is exerted on the table tennis ball when dropped, the
ball is held by a clamp from the horizontal sides. By only squeezing the clamp to release
the ball ensures no downward force is provided.

Temperature
Knowing temperature will affect the hardness of the plastic, the entire experiment is
carried inside the house with a constant temperature of 22 , according to the value on

the central heating system.

Environment
The experiment takes place in a garage, thus the balls are all bounces up from a concrete
floor. The doors and windows are closed, so there are no extra air resistance affecting
the experiment. The reason this is taken into account is because table tennis balls are
very light, and is easily affected by wind.

Counters
There are two counters recording the data. The same two people are helping
throughout the experiment. Final data will be taken from averaging the two results.
Apparatus
- Measuring tape
- Six Yuping brand Model AS7 Table Tennis Balls
- Clamp

Figure 1: And illustration of the method devised for this experiment.

Procedure
1. Locate the designated height with a meter tape.
2. Hold the bouncing ball with a clamp on the horizontal side of the ball.
3. Release and the counters should start counting.
4. Note: the order of the heights should be randomized in order to avoid fatigue as well
as pre-cognitive assumptions by the counters after 2-3 trials of one variation. For
example, one counter may start to assume the data according to prior trials instead
of focusing recording the present one.
5. Repeat the steps until all 25 trials are finished.

Raw Data Table

Released Height (m) 0.05m

Number of bounces by Counter 1

( 2 bounces)

Number of bounces by Counter

2 ( 2 bounces)

0.50m

10

0.50m

10

0.50m

0.50m

0.50m

10

0.90m

12

11

0.90m

14

13

0.90m

12

10

0.90m

14

11

0.90m

13

12

1.30m

15

16

1.30m

16

15

1.30m

14

15

1.30m

15

16

1.30m

16

15

1.70m

11

12

1.70m

10

13

1.70m

13

12

1.70m

12

11

1.70m

13

14

2.10m

11

12

2.10m

12

2.10m

10

11

2.10m

11

12

2.10m

10

11

Table 1: Quantitative data for the bouncing table tennis ball

-

Uncertainty in height due to the possibility of fluctuation when performing the

experiment, because there has to be enough empty space for the ball to bounce. The
ball is slowly moved away from the meter tape to let drop, creating uncertainty.
Uncertainty in bounces due to human limitation. Even though I have used two
counters to decrease the random error, however, there are still likely to be error.

Qualitative Data
- The sound of the first bounce becomes different after reaching 1.7m. It made more
like a breaking sound, while the other bounces are more like a normal, light table
tennis ball sound.

Processed Data
Released Height (m) 0.05m

Average bounces ( 2 bounces)

0.50m

0.50m

9.5

0.50m

0.50m

8.5

0.50m

0.90m

11.5

0.90m

13.5

0.90m

11

0.90m

12.5

0.90m

12.5

1.30m

15.5

1.30m

15.5

1.30m

14.5

1.30m

15.5

1.30m

15.5

1.70m

11.5

1.70m

11.5

1.70m

12.5

1.70m

11.5

1.70m

13.5

2.10m

11.5

2.10m

11.5

2.10m

10.5

2.10m

11.5

2.10m

10.5

Table 2: Average number of bounces for each trial

Sample Calculation 1
Average bounces =

# 1: 2

8:10

= 9 bounces

Uncertainty unchanged because all the maximum number of bounces subtracting the
minimum number of bounces and dividing by 2 is smaller or equal to the given
uncertainty 2 bounces.

Released Height (m) 0.05m

Average bounces ( 2 bounces)

0.50m

8.8

0.90m

12.2

1.30m

15.3

1.70m

12.1

2.10m

11.1

Table 3: Average number of bounces for each variation

Sample Calculation 2
Average bounces =

1: 2: 3: 4: 5

9:9.5:8:8.5:9

= 8.8 bounces

Released Height vs Average bounces

20
18

Average bounces (bounces)

16
14
12
10
8
6
4
2
0
0

0.5

1.5

2.5

Released Height (m)

Graph 1: Average bounces vs. Released height

From the graph, it shows a weak correlation between the best fit line and the data points.
The best fit line underestimates the bounces between 0.5m to 1.3m while overestimates
the bounces after 1.3 m till 2.1m. Therefore, the best way to illustrate the data seems to
be separating the data points into two graphs.

Best fit Line : 8.125x + 4.7875

Released Height vs Average bounces
(1)
Minimum Gradient: y = 3.5714x + 8.8357
20

Maximum gradient: y = 11.667x + 1.55

18
16

Average bounces (bounces)

14
12
10
8
6
4
2
0
0

0.2

0.4

0.6

0.8

1.2

1.4

1.6

Released Height

Graph 2: Average bounces vs. Released Height from 0.5m to 1.3m

Released Height vs. Average bounces

(2)
Best fit line y = -5.25x + 21.758
20

Maximum Gradient: y = -11.714x + 33.114

18

Minimum Gradient: y = -0.2222x + 13.578

16

Average bounces (bounces)

14
12
10
8
6
4
2
0
0

0.5

1.5
Released Height

Graph 3: Average bounces vs. Released Height from 1.3m to 2.1m

Graph and equations obtained by Excel 2010.

2.5

Sample Calculation 3
Line of best fit:
Slope= 8.125m m/bounce
Sample Calculation 4
Minimum gradient line
Selected data points
(0.55, 10.8)
(1.25, 13.3)
Slope =

13.3;10.8
1.25;0.55

= 3.5714 m/bounce

Mmaximum gradient line

Selected data points
(0.45,6.8)
(1.35,17.3)
Slope =

17.3;6.8
1.35;0.45

= 11.667 m/bounce

Sample Calculation 5
Uncertainty for gradient =

1
2

| | =

4.0478 m/bounce
Graph 2:
Gradient = 8.1254.0478m/bounce
Gradient =8.134.05m/bounce
Percentage uncertainty=

4.05
8.13

= 49.8%

The percentage uncertainty is 49.8%

Graph 3:
Gradient = -5.255.7459m/bounce
Gradient =-5.255.75 m/bounce
The percentage uncertainty is 110%

1
2

|11.667 3.5714| =

Conclusion
I decided to break down the one set of data points into making two graphs in the
end. Afterwards, each graph separately demonstrated strong correlation between the
best fit line and the data points, creating an excellent example of a linear model.
Graph 2 shows that for the height between 0.5 to 1.3 meters, every increase in meter
will lead to an increase of 8.125 bounces of the ball. Oppositely, for the released height
between 1.3 to 2.1 meters, every increase of 1 meter will lead to a decrease of 5.25
bounces f the ball.
The result of the experiment partially testifies my hypothesis which says the
increase of height will lead to an increase in bounces of the ball. However, the
experiment shows that after reaching around 1.3 meters, instead of increasing the
bounces, the number of bounces has actually decreased. The reason this has occurred I
have thought of two reasons. The first reason is because the higher the released point,
the air resistance would be bigger, and because the table tennis ball is very light, this
resistance must have affected the ball a lot; another reason is that the table tennis ball
may be damaged by releasing from a greater height, because I did hear a different sound
when the height is increased as the table tennis ball is first dropped. The ball instead of
using the energy to bounce back up, it might have been instead released into sound
energy and damaging the ball.
The uncertainty calculated has also been very huge, as big as 49.8% for graph 2 and
110% for graph 3. The error will be discussed in the Evaluation section. The range of
values can be disperse more and definitely more variations are needed. Because each
graph only has 3 datapoints, the uncertainty is huge and the reliability is also
questionable.
Evaluation
The chosen type of ball had been most likely affected by air resistance as the height
increases, accounts for a type of systematic error. There was obvious random error for
the counters when they count the bounces, affecting the accuracy of the data and
contributing to the uncertainty. The method could have been better at decreasing the
uncertainty for the released height, another random error from performance.
Improvements
Next time I will choose a different type of ball such as a tennis ball to do the
experiment. The greater mass of a tennis ball would decrease the effect of air resistance
as well as the material is much durable than a table tennis ball which is made of thin
plastic. Also, I would consider using a ExplorerX instead of two counters to record the
number of bounces, because it would be much more accurate. Last but not least, I would
come up with a better method at releasing the ball. For example, I would add a
horizontal stick attached to the vertical meter tape from the wall to increase in accuracy
for the release height, which would significantly decrease the uncertainty.
I think more variations should be performed for this experiment as well, after
dividing the graph only 3 data points where in each illustration, decreasing the
reliability of the data.