INTERNAL ASSESSMENT

HOW TO TILE WITH PENTAGONS?

An Investigation into Penrose Tiling

Subject: Mathematics Analysis and Approaches HL

Page count: 19 (including bibliography)

1 Introduction

Tiling or tessellation is a way of arranging identical shapes (prototiles) so that they

cover an entire area without any overlap or gaps. Commonly, this is a practice in
construction, notably on floors, and pavements (Schwartz, 2007).

The only way for a set of identical prototiles to tile the plane is so that a sum of the
sizes of their interior angles of the prototiles equals to exactly 360°. In other words,

+¿ ¿
k ×θ=360 ° , k ∈ Z

This cannot be just any angle; for example, 144° does not meet the requirement, as
then, according to the formula, k would not be an integer. It is important to note that 144° is
the size of an interior angle of a regular decathlon. From this we can reason that not all
regular polygon shapes can be used as prototiles.

To determine which polygons can be used, we can use the formula for the size of
interior angles of n-sided regular shapes:

360 ° +¿¿
θ=180°− ,∈ Z
n

In order to know which polygon shapes can be used when tiling, we have to express k
in terms of n. This way, we can quickly check if an n-sided shape, after we plug in the value
of n, results in a positive integer value of k. If it does, the n-sided shape can be used as a
prototile.

We get rid of the fraction by multiplying both sides by n:

n ×θ=n ×180 °−360 °

We express n:
360 °
n=
180 °−α

2
Lastly, we replace α with k:
360 °
n=
360 °
180 °−
k

We simplify and express k:

2n
k= ,{n , k }∈ Z +¿¿
n−2

Which tells us which n-sided regular shapes can be used during the process of tiling;
notably, triangles, squares, and hexagons, as seen on diagram 1, and 2:

Diagram 1: using triangle shapes in Diagram 2: using square shapes in

tessellations tessellations
(note, this also includes hexagons as six triangles
make a hexagon)

But not notably, pentagons. This is clear both from the formula;
2× 5 10 + ¿¿
= ∉Z
5−2 3

And through visual representation on the following page;

3
 Since we don’t get a positive integer value of k if we
plug in (n=5) into the formula

2n
k=
n−2

it would only make sense that identical regular

pentagons couldn’t be used in tessellations, but it is
even clearer when we look at diagram 3. Note the red
pentagon that is clearly overlapping pentagon AHIJK.

Diagram 3: attempt at using regular

pentagon shapes in tessellations

This intrigued me since I have seen tiles involving pentagons everywhere; from my
nation’s capital’s main street to my middle school’s playground. Interestingly, after some
investigation, I noted that on different tiling, similar shapes are used to compensate for the
inability of pentagons to cover a surface completely without overlap. After looking a bit more
into it, I have stumbled upon the concept of “Penrose tiling”; a way to use smaller (imagine
cut-out) pieces from a pentagram to fill out the empty spaces tiling with pentagrams would
otherwise leave.

Throughout this internal assessment assignment, I am going to investigate Penrose

Tiling, show some already existing tessellation patterns involving it, and create a final
product using the knowledge acquired.

My goal for this assignment is exactly that; to better understand the complexities of
Penrose tiling and create a shape made out of those tessellations, effectively learning the
basics on how to tile using Penrose tiling.

4
2 Theory on Penrose Tiling
2.1 Introduction to Penrose Tiling
Penrose tiling, discovered by Roger Penrose is, simply put, a form of tiling in which
one uses parts of a pentagon in order to tile with them. This process results in aperiodic tiling;
no matter the direction and size of a translation of segments of any size, we can never get the
same pattern twice (Gardner, 1997).
There are different types of Penrose tiling; the “Original Penrose tiling”, “Kite and
Darts tiling”, and “Rhombus tiling”, each being based around taking different sections of a
pentagon to cover the area (Shultz, 2006).
I will only be discussing the “Kites and Darts” tiling in depth as I found it the most
interesting and wish to cover the most amount of information surrounding it; however before
that, I will also briefly show what the Original Penrose tiling looks like as it will later be
useful when creating my own shape.

2.2 The original Penrose Tiling

The Original Penrose tiling, or Star Penrose tiling was the first to be created. It uses
a
two pentagons of different sizes, their sides being in a ratio of =2 ×sin ⁡(72° ). This will
b
make a lot more sense in a bit (Schwartz, 2007).
To achieve this tiling, we first have to draw the larger of the two pentagons and its
diagonals (Diagram 4). This gives us the star shape from which this tiling gets its name.
From this we use only the star segment as we can then attach the smaller pentagons to it by
its sides (Diagram 5).

5
 Diagram 4: The larger pentagon with Diagram 5: Smaller pentagons attached to
drawn-in diagonals. the pentagram

The purpose of this is that regular star polygons have acute interior angles of the size
36° . Since the interior angle of a pentagon is equal to 72° , and 2 ×72 °+36 ° =180° , we can
now fill out the entire 360° ; there is no gap or overlap.

In the event of there still being smaller

gaps to fill in, we can always use
smaller pieces of the pentagram to do
so. Two examples and where to get
them from are shown in Diagram 6 on
the left.

Diagram 6: Segments of the pentagram that can

be used as tiling

In practice, this looks like Diagram 7.

6
Note the two highlighted areas in the tiling
are a smaller segment already highlighted in
Diagram 6 (purple) and a segment consisting
of two smaller segments from Diagram 6 put
together (red).

2.3 Kites and Darts tiling Diagram 7: The segments in practice

2.3.1 Introduction to Kites and Darts tiling

The construction of the Kites and Darts Penrose tiling starts similarly to the one of
Star Penrose tiling. First, we draw a regular pentagon and draw its diagonals. Then we
connect the intersect of two diagonals, as seen in Diagram 8 (Shultz, 2006).

Diagram 8: The construction of the rhombus

This gives us a rhombus. In itself, that could already be used for tiling as if we sum up
the size of its acute (72 °) and obtuse angle (108 °), we get 180 °. The application of this
notion can be seen on Diagram 9 (Shultz, 2006):

Diagram 9: tiling with the rhombus

7
 However, this rhombus also gives us the ability to produce aperiodic tiling. This is
done with using another intersect of two diagonals (Shultz, 2006). We connect that intersect
with one of the points of the original pentagon as seen on Diagram 10:

Diagram 10: Connected intercepts

With this, we get two new shapes; the dart (purple) and the kite (red). When
investigating the interior angles of the two new shapes (see visual proof below), we would
find that they have the sizes of 36° , 72° , 144° , 216° . There are ways to sum these up to 360,
which proves we might be able to tile with them.
Visual proof: the size of interior angles of kites and darts (see Diagram 11-14)

Since the segments v and u are part of the same regular

pentagon, they are of the same size. This makes the
triangle TUV an isosceles triangle, and the angles β and γ
be the same size.
We know that the size of angle α is 108° , as it is part of a
regular pentagon. The sum of all angles in any triangle is
T equal to 180° . Since 180° - 108° =72° , the sum of β and γ
Diagram 11: The Visual Proof 1 must equal to 72° . Therefore, the size of them both
individually is 36° . The same reasoning is applied to
get the value of δ . This gives us the first interior angle
of the dart shape.
Next, see Diagram 12, we know that the size of
an acute interior angle of a pentagram is 36° . The size
of ε is the sum of γ and that acute interior angle of the
pentagram, therefore it is equal to 72° . ζ is of the same
size as the kite is horizontally symmetric.

Diagram 12: Visual Proof 2

8
 Lastly, see Diagram 13, the same reasoning applied for ε can be applied for η. As the
sum of all interior angles in a quadrilateral shape is 360° , we can determine the size of κ to be
144° . This also gives us the reflex interior angle, λ , of the dart (360° - 144° = 216° ).
Altogether, this gives us the angles shown in Diagram 14.

2.3.2 The Rules surrounding

Diagram Penrose
13: Visual Proof tiling
3 Diagram 14: The angles
In order for this tiling to be aperiodic, and to not result in simple rhombic tiling, there
Penrose alongside with John Corton Conway devised a set of rules as to how to tile with the
two shapes (Shultz, 2006). They labeled the vertices of the tiles as heads (H) and tails (T)
(see Diagram 15 and 16). The rule states that when tiling, only vertices with the same
labelling can coincide (Shultz, 2006).

Diagram 15: Labelled Kite Diagram 16: Labelled Dart

2.3.3 Introduction into tiling with Kites and Darts
Using the aforementioned rules as a guide, one could simply tile with placing the
tiling one after the other. This is what I did when first experimenting with Penrose tiling (see
Diagram 17)

9
 Diagram 17: My first attempt at Penrose Tiling

But even still, there is a possibility that even following these rules, a pattern may
produce a gap or an overlap (see Diagram 18)

Diagram 18: Gap and Overlap while following the rules

Because of this, there is a specific method to ensure that does not happen when tiling;
Inflation.
2.3.4 Specific method; Inflation
Inflation takes an existing tiling (a dart or a kite), replaces it with smaller tilings, and
inflates the end result. This ensures there is no gaps as all one does is section an already
existing shape into smaller ones (Schwartz, 2007).
There are multiple styles of inflation. The following are only a few that I ultimately
used in my own pattern.

1. Diagram 19: The ace

10
 2. Diagram 20: The Deuce:

3. Diagram 21: The Sun and Star:

2.3.5 A Characteristic of tiling with Kites and Darts: Aperiodicity

Previously, I have only stated that Penrose tiling is aperiodic, but have never proven
this assumption. I will be doing so now.
The reason why this proof is important for this IA is that one cannot create a new
pattern with simply translating a part; as you could with periodic tiles as the result would
likely lead to either a gap or an overlap.

To prove Penrose tiling is aperiodic, I will show that the ratio between the kites and
darts in a specific inflation is an irrational number (Gardner, 1997). I will not be proving why
that is so as it is a long process and will be using this as an assumption.

11
The following table shows the number of kites and darts in each of the first four inflations.

Number of Inflations
1 2 3 4
Dar
Type of Tile t Kite Dart Kite Dart Kite Dart Kite

Number of tiles (When inflating the dart) 1 1 2 3 5 8 13 21

Number of tiles when (When inflating the kite) 1 2 3 5 8 13 21 34

Table 1: a record of the number of kites and darts following each inflation
(Investigating the Effectiveness of Penrose Tilings as a Model of Quasicrystals (extended
essay))
One could notice that the number of both kites and darts for each successive inflaitons
seems to mimic the Fibonacci sequence (Weisstein). The proof for this is also long and I will
be using this as the second assumption. I believe this is just due to my internal assessment not
focusing on the math surrounding the Fibonacci sequence, but the application of the
knowledge that comes with knowing that Penrose tiling is aperiodic. I will however be
proving that the ratio of the successive terms in the Fibonacci sequence will approach the
golden ratio. Then, if I show that the golden ratio is an irrational number, this is enough to
show that Penrose tiling is aperiodic.

Proof 1: The ratio of successive terms in the Fibonacci sequence approaches the golden ratio
(Investigating the Effectiveness of Penrose Tilings as a Model of Quasicrystals (Extended
Essay))
We can represent the Fibonacci sequence using:

F a+1=F a + F a−1 (1)

The ratio of the successive terms in the Fibonacci sequence can be represented as:
F a +1
Fa

12
We are trying to prove this ratio approaches the golden ratio with later terms. Therefore, we
introduce a limit:
F a +1
lim
a→∞ Fa
Here, we can substitute in (1):
F a +1 F +F
lim = lim a a−1
a→∞ F a a →∞ Fa
F a +1 F a−1 (2)
lim = lim 1+ lim
a→∞ F a a →∞ a→ ∞ F a

We then introduce a new variable “g”:

F a+1
g= lim
a →∞ Fa
Fa F a−1 1
lim =¿ lim = ¿
a→∞ F a +1 a→ ∞ F a g
We then substitute “g” in (2):
1
g= lim 1+
a →∞ g

We know that alim

→∞
1=1

1
g=1+
g
All that is left to do is solve the equation for g:
2
g =g+1
g2−g−1=0
1±√5
g=
2
F a +1 1 ± √ 5
lim =
a→∞ Fa 2
This proves that the ratio of the successive terms in the Fibonacci sequence will
approach the golden ratio. All that is left to do is prove that the golden ratio is an irrational
number (unknown, Investigating the Effectiveness of Penrose Tilings as a Model of
Quasicrystals (Extended Essay)).

13
Proof 2: The golden ratio is an irrational number
I will be providing this proof through first proving that √ 5 is an irrational number
through the proof of contradiction.

Let us assume that √ 5 is rational. This would mean it could be expressed as a quotient
between two integers, simplified to the lowest terms.
h
√ 5=
i
i √ 5=h
2 2
i 5=h
Since h is an integer, h2 must have an even number of prime factors. The same goes
for i2. However, this shows a contradiction; If i2 has an even number of prime factors, and we
multiply it by a prime number, the result must have an odd number of prime factors.
Therefore, h2 cannot be the result. This proves that √ 5 is an irrational number.

Since the sum of any rational number and an irrational number is always irrational,
this would mean 1 ± √ 5 would also be irrational. Since the quotient of an irrational number
1± √ 5
and a rational number is also always irrational, this would prove is irrational. Since
2
this is the ratio that the quotient of two consecutive terms in the Fibonacci sequence
approaches, and the Fibonacci sequence accurately describes the numbers of kites and darts
present in each consecutive deflation, we can assume that Penrose tiling is aperiodic.

Again, this gives me the information that when tiling on my own, I cannot simply
translate a shape in an already existing pattern to create a new pattern, as it will result in gaps
or overlaps.

3 The Creation of my Own Pattern

When creating my own pattern, I decided on using a combination of the patterns that I
have learned about. This is because there was no way of me simply translating the tiling in
the already existing patterns due to the aperiodic nature of them, but also because I wanted to
properly explore all the knowledge I have acquired both about Penrose tiling and the program
I was using.

14
3.1 The Process

I started off with a simple kite tiling (see diagram 22).

Diagram 22: The Kite

I then continued to inflate the kite into an ace (see diagram 23).

Diagram 23: The Ace Inflation

Next, instead of using the ace inflation, I inflated both the kites and darts using deuce
inflation (see Diagram 24).

Diagram 24: The Deuce Inflation

Because this is not standard for ace inflation, I had to introduce a new shape; the half
dart (see red highlighted section in Diagram 25).

15
 Diagram 25: Highlighted Half Dart

It was here that I noticed that there is a star-shape forming on the bottom part of the
pattern. Thus, I added two additional darts to complete it (see Diagram 26)

Diagram 26: The Star

Next, I again inflated all kites using ace-style inflation, and the darts in the bottom
star-shape using sun and star-style inflation (see Diagram 27).

Diagram 27: Ace Inflation 2

16
 It is here that I encountered a problem. The red-highlighted darts cannot be inflated in
the same style, as this would result in an overlap (see Diagram 28).

The way I went about the solution

Diagram was identifying
28: Highlighted Problemhow the dart itself is structured. I
Darts
found that it is composed of a isosceles triangle that can be found in the Original Penrose
Tiling and a rhombus that can later be again inflated into new kites and darts (see diagram 29
and 30).

Diagram 29: The angles within the problem darts Diagram 30: The shapes within the problem darts

Zoomed out, this gave me the final product (see Diagram 31).

17
 Diagram 31: The final result

Potentially, one could go on and inflate the structure further with both ace and deuce-
style inflation for most of the structure, as well as sun and star-style inflation for the now
three sun structures forming in the middle. Personally, I decided not to both due to the page
limit and the program crashing every time I attempted to reflect an object through a line.
Regardless, I am happy with the visual appeal of the final result.

4 Conclusion
Penrose tiling is a process of tiling involving smaller segments of a pentagon. These
segments are used as tiling material. There are multiple ways to utilize such tiling, most
notably “The Original”, the “Rhombus”, and “The Kites and Darts”, and even within each of
the three, there are different ways of tiling.
The Kites and Darts style of tiling involves a process of “inflation”. This looks like
using an already existing shape, dividing it into smaller components, and inflating the result.
This process guarantees no gaps and overlaps, provided one follows the rules given.
When combining the different styles of Penrose tiling, one will come to a point where
inflations doesn’t necessarily work. However, seeing there is no rule against it, one can
always use a wider variety of shapes from the pentagon to fill in the gaps that may arise
accordingly.

18
Bibliography
1:
Gardner, M. (1997). Penrose Tiles to Trapdoor Ciphers. Washington: Mathematical
Association of America.
2: Schwartz, R. (2007, November). Penrose Tiling Basics. Retrieved December 2020, from
https://www.math.brown.edu/reschwar/MFS/handout7.pdf
3: Shultz, K. (2006). Penrose Tilings. Retrieved December 2020, from MATH 7200:
http://jwilson.coe.uga.edu/emat6680fa05/schultz/penrose/penrose_main.html
4: unknown. (n.d.). Investigating the Effectiveness of Penrose Tilings as a Model of
Quasicrystals (extended essay). Retrieved from
http://www.mrbertman.com/EE/penrose.pdf
http://www.mrbertman.com/EE/penrose.pdf
5: Weisstein, P. C. (n.d.). Retrieved December 2020, from Wolfram Mathworld:
http://mathworld.wolfram.com/FibonacciNumber.html

All the diagrams were made my me.

19