Available online at www.sciencedirect.
com
Available online at www.sciencedirect.com
Available online at www.sciencedirect.com
ScienceDirect
Procedia Computer Science 00 (2023) 000–000
www.elsevier.com/locate/procedia
Procedia
Procedia Computer
Computer Science
Science 00 (2023)
218 (2023) 000–000
2541–2549
www.elsevier.com/locate/procedia
International Conference on Machine Learning and Data Engineering
International Conference on Machine Learning and Data Engineering
Deployment of Fractals through CNC Devices
Deployment of Fractals through CNC Devices
Deepsikha Bhattacharjeea , Sukanya Nagb , Dr. Suman Deba,b,∗
a b 799046, India
Deepsikhaab National
Bhattacharjee , Sukanya
Institute of Technology Agartala,Nag , Dr. Suman Deba,b,∗
Agartala
National Institute of Technology Agartala, Agartala 799046, India
a National Institute of Technology Agartala, Agartala 799046, India
b National Institute of Technology Agartala, Agartala 799046, India
Abstract
The concept of fractal propagation, through the deployment of computer numerical controlled devices, has been discussed in this
Abstract
paper. For experimentation purposes, four of the most well known fractal patterns namely, Sierpinski gasket, Barnsley fern, Fractal
The
tree concept
and Koch of snowflake
fractal propagation,
curve havethrough the deployment
been chosen. This work of shows
computer the numerical controlled devices,
practical reproduction has mentioned
of before been discussed in this
fractals on
paper. For experimentation purposes, four of the most well known fractal patterns namely, Sierpinski gasket,
a physical media by deploying a 3-axis CNC plotter, as a part of the hardware, after proper python implementation as a part of Barnsley fern, Fractal
tree and KochThe
the software. snowflake curveand
mathematics have been chosen.
algorithm used inThis
thework showsof
production thethepractical
fractals reproduction of before
have been discussed in mentioned fractals
detail and the on
method
aofphysical media by deploying a 3-axis CNC plotter, as a part of the hardware, after proper python implementation
deployment involves vectorisation of the fractals into a set of coordinates which are then passed on to the 3 axis CNC plotter as a part of
the software. The mathematics and algorithm used in the production of the fractals have been discussed
plotter for production of the final output on a larger drawing surface. This paper aims at combining the individual concepts ofin detail and the method
of deployment
fractal involves
propagation vectorisation
and robotics in orderof the fractals tangible
to produce into a setfractal
of coordinates
illustrations.which are then passed on to the 3 axis CNC plotter
plotter for production of the final output on a larger drawing surface. This paper aims at combining the individual concepts of
fractal propagation and robotics in order to produce tangible fractal illustrations.
© 2023 The Authors. Published by Elsevier B.V.
©This is an
2023 Theopen accessPublished
Authors. article under the CC BY-NC-ND
by Elsevier B.V. license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
© 2023 The under
Peer-review Authors. Published by
responsibility of Elsevier
the B.V. committee of the International Conference on Machine Learning and Data Engi-
This is an open access article under the scientific
CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
This is an open
neering.
Peer-review underaccess article under
responsibility of thethe CC BY-NC-ND
scientific committeelicense (http://creativecommons.org/licenses/by-nc-nd/4.0/)
of the International Conference on Machine Learning and Data Engineering
Peer-review under responsibility of the scientific committee of the International Conference on Machine Learning and Data Engi-
neering.
Keywords: Fractal Geometry; CNC Machines; 3-axis CNC Plotters; Vectors
Keywords: Fractal Geometry; CNC Machines; 3-axis CNC Plotters; Vectors
1. Introduction
1. Introduction
Fractals are self-similar geometric figures that are often irregular, fragmented, fractured or loosely connected in
appearance. A geometric construction created by infinitely performing a specific operation, shape or object results in
Fractals are self-similar geometric figures that are often irregular, fragmented, fractured or loosely connected in
the formation of a fractal. This paper combines the concept of fractal geometry with CNC devices in order to propose
appearance. A geometric construction created by infinitely performing a specific operation, shape or object results in
a method of creation of fractal illustrations[1] by deploying a 3-axis CNC plotter. For experimentation purposes, four
the formation of a fractal. This paper combines the concept of fractal geometry with CNC devices in order to propose
Iterated Function System(IFS)[2] fractals have been chosen, namely Sierpinski Gasket, Koch Snowflake Curve[3],
a method of creation of fractal illustrations[1] by deploying a 3-axis CNC plotter. For experimentation purposes, four
Barnsley Fern[4] and Fractal Tree[5]. The reason behind the selection being:
Iterated Function System(IFS)[2] fractals have been chosen, namely Sierpinski Gasket, Koch Snowflake Curve[3],
Barnsley Fern[4] and Fractal Tree[5]. The reason behind the selection being:
• These are popularly used structures
• Being deterministic in nature, they can be intuitively analysed.
• These are popularly used structures
• Being deterministic in nature, they can be intuitively analysed.
∗ Dr. Suman Deb. Tel.: +91-9436459622
E-mail address: [email protected]
∗ Dr. Suman Deb. Tel.: +91-9436459622
E-mail address:
1877-0509 [email protected]
© 2023 The Authors. Published by Elsevier B.V.
This is an open
1877-0509 access
© 2023 Thearticle under
Authors. the CC BY-NC-ND
Published license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
by Elsevier B.V.
1877-0509 ©under
Peer-review 2023responsibility
The Authors.of Published by Elsevier
the scientific B.V.of the International Conference on Machine Learning and Data Engineering.
committee
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-review under responsibility of the scientific committee of the International Conference on Machine Learning and
Peer-review under responsibility of the scientific committee of the International Conference on Machine Learning and Data Engineering.
Data Engineering
10.1016/j.procs.2023.01.228
2542 Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549
2 Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000
Although these are not complex structures but their study allows an insight into the characteristics of fractal fronts [6].
• Sierpinski Gasket : The Sierpinski gasket(or triangle) is a fractal attractive fixed set consisting of an overall
structure of an equilateral triangle that is recursively subdivided into other smaller similar triangles. The final
structure is a set of all the points that remain once the entire process has been repeated infinitely often.
• Koch Snowflake : (also referred as the Koch curve, Koch star, or Koch island) is a mathematical curve with
a six-fold symmetry which resembles a natural snowflake. It is self-similar in nature as it comprises of three
self-similar parts, each of which is comprised of four non-overlapping parts that are exact scaled-down versions
of the entire thing. It follows that each of the four parts itself consists of four parts again that are also the
scaled-down versions of the entire curve.
• Barnsley Fern : is a pattern generated mathematically that can be created by iterating over a large number
of times on four mathematical equations, introduced by Michael Barnsley, known as Iterated Function System
(IFS). Being an IFS fractal pattern, it exhibits the characteristics of self-similarity, i.e. reproducible at any
magnification or reduction. It is also infinitely complex in nature, as well as a chaotic fractal.
• Fractal Tree : A recursive symmetric binary branching where a trunk of unit length splits into two branches of
length R and make a Q angle with the direction of their parent. This process keeps on repeating a set certain
number of times in order to give the final output.
There are several existing methods that can be applied for generating fractal design patterns, like Menger sponge[7],
Hypercomplex manifold, Brownian tree, Brownian motion[8], Decomposition, L-systems[9], Lyapunov fractals, New-
ton fractals[10, 11], Pickover stalks and Strange attractors. Among all these, we choose the L-System method[12],
which can be applied easily in programming using the implementation of turtle geometry[13] interpretation. Accord-
ing to the mentioned reasoning, this research is intended for developing customized application software for CNC
machines which can be used to create design patterns using fractal geometry [14] automatically based on input from
users.
Since the combination of the fields of embedded systems and fractal geometry have been proposed in this paper, the
predecessors of this amalgamation have been addressed in the Related Works section. The construction of the XY
plotter along with the experimentation details have been discussed in the later sections.
2. Related Work
The fields of embedded systems and fractal geometry have been previously explored but this paper attempts to com-
bine these two fields in order to produce tangible real-time outputs. We have formerly worked on an XY plotter[15]
with the plotter functioning as a miniature version of the 3 axis CNC plotter where an approach for efficient vectorisa-
tion of bitmap images was introduced. The research work proposed an algorithm which involved edge detection using
a CNN model followed by vectorisation. The output results were comparable to an initial rough sketch of a human
artist. The aim of the work was to add to ordinary plotter functionality and enhance the capabilities of CNC devices
in the field of spatial robotic artwork. The concept of fractal geometry[16] deals with study of irregular objects which
do not abide by the laws of Euclidean geometry. The study of fractals[17, 18] allow better understanding of every-
day objects[19] and accordingly, the shapes of the real world. These shapes having infinite perimeter but in a finite
area require a separate methodology for projecting them in a physical surface with desired accuracy and efficiency.
This work deals with creation of fractal illustrations on such physical media with the assistance of CNC devices and
vectorisation techniques.
3. CNC Device
CNC machine is a machine tool which has the ability to automatically form a pattern of computer-assisted design.
For the implementation of fractal artwork in this paper, an 3-axis CNC plotter has been constructed. The circuit
consists of the Arduino UNO board which is further supported by the CNC shield. The connection with the motors is
made using motor drivers which are attached to the CNC shield. Two types on drivers namely, A4988 and DRV8825
are used for experimentation purposes. The choice is purposefully made to test the machine compatibility with both
Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549 2543
Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000 3
the drivers and also to have a backup in case of an emergency. Heat sinks are also added to the drivers in order to rule
out the possibility of damage due to overheating. At a time, a total of three drivers are required to achieve plotter func-
tionality. This is because each driver has the capacity to run exactly one stepper motor. It is important to mention here
that only a single type of driver was used in every experimentation. A combination of A4988 and DRV8825 is not used.
Arduino UNO ATmega328P based microcontroller board having 14 digital input/output pins
CNC shield Uses open-source firmware on Arduino to control 4 stepper motors
A4988 motor drivers Has built-in translator for easy operation.
DRV8825 motor drivers Consists of two H-bridges and can operate a bipolar stepper motor or two DC motors
NEMA17 motors These motors provide the required axis motion
Fan For cooling purposes
SG90 mini servo motor For up and down motion of drawing pen
Glass slab and pens Formation of the final drawing
The most important step before complete motor functionality is achieved is micro stepping. Each axis has 3 jumpers
that can be set to configure the micro stepping for the plug-in driver board. ”High” indicates that a jumper is inserted
and ”Low” indicates that no jumper is inserted. After micro stepping is done, the drivers can be placed on the CNC
shield with proper orientation. Since the Y-axis gets duplicated, the second driver for the one of the Y-axes is placed
in place of the fourth axis (labelled as A in the CNC shield) and a jumpers are placed along the Y-axis marking in the
shield in order to account for the proper replication. Another jumper is also added to the Enable(EN) on the shield
in order to enable the shield. Additionally, it has been observed that the Arduino UNO tends to get heated up during
specific current or voltage settings. To avoid any damage due to overheating, a fan is also attached to the bottom of the
Arduino-CNC shield combination. The next step involves attaching the servo motor to the shaft or the gondola which
is placed on top of the fifth movable beam. The servo allows for the up and down motion of the pen ensuring a clean
output.
The end stops are added in the last step. The soft limits and hard limits for the end stop switches have to be set through
the G-code sender. This has been explained in the next section.
4. Experimentation using equations for Fractal Generation
4.1. Sierpinski Gasket
Sierpinski Gasket consists of an overall structure of an equilateral triangle which is recursively subdivided into
smaller triangles. Every time the iterated function system is applied to an equilateral triangle, similar triangles are
formed within the previously existing triangle. Additionally, these new triangles have perimeters that is exactly equal
to 1/2 the perimeter of the original triangle and an area equal to 1/4 times the area of the original triangle. Thereby
indicating that the overall structure comprises of a finite, measurable area which is equal to the area of the original
triangle whereas the perimeter tends to extend to infinity as the IFS can be repeated infinite number of times. Hence
the basic concept of fractal geometry[20] can be observed from the Sierpinski Gasket.
Scaling factor of the Sierpinski triangle is 2. The dimension of the Gasket is greater than 1 and less than 2, indicating
that it lies dimentionally between a line and a plane. It is given as:
log3
D= = 1.5849 (1)
log2
4.2. Koch Snowflake
The progression for the area of the snowflake converges to 8/5 times the area of the original triangle, while the
progression for the snowflake’s perimeter diverges to infinity. Consequently, the snowflake has a finite area bounded
2544 Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549
4 Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000
Fig. 1. (a) first picture; (b) second picture.
by an infinitely long line. The scaling factor at each stage is three in case of a snowflake curve. The fractal dimension,
given by D, denotes the power to which 3 must be raised to produce 4- i.e.
3D = 4 (2)
Fractal dimension is a key property and an indicator of the complexity of a given figure. Hence, the similarity dimen-
sion of the snowflake curve can be calculated by :
log4
D= = 1.26186 (3)
log3
Fig. 2. (a) first picture; (b) second picture.
Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549 2545
Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000 5
4.3. Barnsley Fern
The Fractal Dimension of the Barnsley Fern cannot be calculated by conventional means, and is estimated to be
about 1.45. Barnsley’s fern uses four affine transformations. The formula for one transformation is the following:
ab x e
f (x) = + (4)
cd y f
where, the letters have the following value :
Table 1. Values:
a b c d e f p PART
0 0 0 0.16 0 0 0.01 Stem
0.85 0.04 -0.04 0.85 0 1.60 0.85 Small Leaflet
0.20 -0.26 0.23 0.22 0 1.60 0.07 Large Leaflet(Left)
-0.15 0.28 0.26 0.24 0 0.44 0.07 Large Leaflet(Right)
In the table above, the columns ”a” through ”f” are the coefficients of the equation, whereas ”p” represents the proba-
bility factor. Hence, the four equations are :
f1 :
xn+1 =0 (5)
yn+1 = 0.16yn
f2 :
xn+1 = 0.85xn + 0.04yn (6)
yn+1 = −0.04xn + 0.85yn + 1.6
f3 :
xn+1 = 0.2xn − 0.26yn (7)
yn+1 = 0.23xn + 0.22yn + 1.6
f4 :
xn+1 = −0.15xn + 0.28yn (8)
yn+1 = 0.26xn + 0.24yn + 0.44
With the help of the above equations, the fern fractal is generated as shown in Figure 3.
4.4. Fractal Tree
A fractal tree, also known as a Symmetric Binary Tree[21], exhibits the concept of self-similarity. Each branch is a
smaller version of the main trunk of the tree. The primary idea in constructing fractal trees is to have a base object and
to then create smaller, similar objects protruding from that initial object. The angle, length and other features of these
children can be randomized for a more naturalistic look. The method used is a recursive symmetric binary branching
where a trunk of unit length splits into two branches of length R and make a θ angle with the direction of their parent.
This process keeps on repeating a set certain number of times in order to give the final output.
2546 Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549
6 Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000
Fig. 3. (a) Barnsley Fern; (b) Code snippet
The two branches are obtained from the trunk by scaling by a factor of r, rotating counterclockwise by θ (left) and
θ (right), then translating to the top of the trunk. The third function needed is the identity. This keeps the branches
already drawn in their current locations while the first two functions add the new branches.
rcos(θ) −rsin(θ) 0
f1 (x) = x+ (9)
rsin(θ) rcos(θ) 1
rcos(θ) rsin(θ) 0
f2 (x) = x+ (10)
−rsin(θ) rcos(θ) 1
10
f3 (x) = x (11)
01
If the scaling factor r is too small, the branches of the tree will be self-avoiding, while if r is too large the branches
will overlap.
Fig. 4. (a) Fractal Tree; (b) Code Snippet
Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549 2547
Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000 7
5. Fractal Deployment by sending G-Code to CNC Machine
The creation of a G-code file is the most crucial step in the entire process because the CNC device can not process
a simple SVG file. G-codes are understood by plotters, printers, laser cutters and other such CNC devices in order to
project an ordinary image onto a physical 2D or 3D media. The plotter uses the G-code files to control the movement
of the stepper motors, servo motors and all other hardware components. The factors like timing, duration, speed etc.
of the hardware parts depend upon the G-code.
We design a G-Code interpreter so that the CNC machine can interpret the G-code and produce the instructions for
controlling the stepper motors directly instead of using G-code sender softwares like Universal Gcode Sender(UGS),
SourceRabbit Gcode Sender, etc.
The Arduino[22] will need to:
• Setup: It has to start listening to the serial connection, followed by setting up the stepper motors and then detect
whoever is listening that it is ready for accepting more instructions.
• Loop forever: It waits for a message, then reads it. This is followed by interpreting the meaning, acting on that
meaning and finally telling whoever is listening that it is ready for further instructions.
The Gcode Commands required:
Table 2. Codes:
Commands Meaning
G00 [X(number)] [Y(number)] [F(number)] Absolute mode: Move in a line to (X,Y) at speed F
G01 [X(number)] [Y(number)] [F(number)] Relative Mode: Move (X,Y) amount at speed F
G04 P(number) Do nothing for P seconds
G90 absolute mode
G91 relative mode
G92 [X(number)] [Y(number)] change logical position
M18 turn off power to motors
M100 print out instructions for the human
M114 report position and feed rate
M03 Spindle ON – clockwise
M04 Spindle ON – counterclockwise
M05 Spindle stop
We have prepared the code for setup, loop as well as interpretation of the commands and further response. Hence, in
around 300 lines of code, we have successfully made a simple CNC machine Gcode interpreter that handles six G
commands and three M commands.
6. Result analysis
On feeding the G-Code to the 3 axis CNC plotter, it was able to accurately plot out the fractal illustration. The
image sections with higher noise presence facilitated creation of small individual vectors which impacted the smooth
movement of the plotter. The presence of noise further added to the completion time of the plotting process. This can
be solved by implementing a pre-processing step to the output image generated after python implementation. A noise
removal algorithm has been found to be beneficial in this regards. Figure 5 below shows the example of a sample
figure made using the CNC plotter.
Conclusion
The paper aims at creating efficient medium for tangible fractal illustrations and further enhancing plotter function-
ality. Fractal patterns can be recreated on physical media through CNC deployment. These patterns are also beneficial
2548 Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549
8 Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000
Fig. 5. Sample output
in conceptual design stage which is generally more expensive in terms of time and cost. With the help of GRBL
firmware and Arduino, we have been able to easily set up and get our 3-axis CNC plotter up and running. This re-
search is intended for CNC machines like the 3-axis CNC plotter which can be used to illustrate design patterns using
fractal geometry automatically based on input from users. Hence this approach demonstrates promising results and
has the potential to further contribute greatly in overcoming the targeted working limitations of computer numerical
controlled devices.
References
[1] Losa, Gabriele & Ristanovic, Dusan & Ristanovic, Dejan & Zaletel, Ivan & Beltraminelli, Stefano. (2016). From Fractal Geometry to Fractal
Analysis. Applied Mathematics. 7 (4): 346-354. 10.4236/am.2016.74032.
[2] József Vass (2014). On the Geometry of IFS Fractals and its Applications. UWSpace. http://hdl.handle.net/10012/8200
[3] Campos, Daniel & Fort, Joaquim & Méndez, Vicenç. (2007). Propagation through fractal media: The Sierpinski gasket and the Koch curve.
EPL (Europhysics Letters). 68. 769. 10.1209/epl/i2004-10284-4.
[4] Neff, Cassondra, ”Fractal Ferns” (2021). Academic Excellence Showcase Proceedings. 281. https://digitalcommons.wou.edu/aes/281
[5] Pons, Bernat Espigulé. Unfolding Symmetric Fractal Trees, Proceedings of Bridges 2013: Mathematics, Music, Art, Architecture, Culture,
Edited by George Hart and Reza Sarhangi, pp 295-302.
[6] S.C. Soo, K.M. Yu, Rapid Prototyping Using Fractal Geometry, 424-431 (2001)
[7] Mayama, Hiroyuki, and Kaoru Tsujii. ”Menger sponge-like fractal body created by a novel template method.” The Journal of chemical physics
125.12 (2006): 124706.
[8] Mandelbrot, Benoit and Michael Frame. ”The canopy and shortest path of a self-contacting fractal tree,” The Mathematical Intelligencer, vol.
21, No. 2 (1999), 18-27.
[9] Prusinkiewicz, Przemyslaw. ”Graphical applications of L-systems.” In Proceedings of graphics interface, vol. 86, no. 86, pp. 247-253. 1986.
[10] Alves FRV, Machado Vieira RP. The Newton Fractal’s Leonardo Sequence Study with the Google Colab. INT ELECT J MATH ED. 2020;15(2),
em0575. https://doi.org/10.29333/iejme/6440
[11] Burton, Aaron. ”Newton’s method and fractals.” (2009).
[12] Development of Fractal Pattern Making Application using L-System for Enhanced Machine Controller, Alexander A S Gunawan, Jimmy
Linggarjati, Yandi Wijaya. EPJ Web of Conferences 68 00032 (2014)
[13] Ron Goldman, Scott Schaefer, Tao Ju, Turtle geometry in computer graphics and computer-aided design, Computer-Aided Design, Volume 36,
Issue 14, 2004, Pages 1471-1482, ISSN 0010-4485, https://doi.org/10.1016/j.cad.2003.10.005.
[14] Edger, G. (2008) Measure, Topology and Fractal Geometry. 2nd Edition, Springer, New York. http://dx.doi.org/10.1007/978-0-387-74749-1
[15] S. Nag, D. Bhattacharjee, A. Bhaumik and S. Deb, ”Generating Vectors from Images using Multi-Stage Edge Detection for
Robotic Artwork,” 2021 International Conference on Computational Performance Evaluation (ComPE), 2021, pp. 651-656, doi:
10.1109/ComPE53109.2021.9752031.
Deepsikha Bhattacharjee et al. / Procedia Computer Science 218 (2023) 2541–2549 2549
Dr. Suman Deb / Procedia Computer Science 00 (2023) 000–000 9
[16] Fouché, W.L. (2009). Fractals Generated by Algorithmically Random Brownian Motion. In: Ambos-Spies, K., Löwe, B., Merkle, W. (eds)
Mathematical Theory and Computational Practice. CiE 2009. Lecture Notes in Computer Science, vol 5635. Springer, Berlin, Heidelberg.
https://doi.org/10.1007/978-3-642-03073-4 22
[17] Demmie, Paul & Ostoja-Starzewski, Martin. (2011). Waves in Fractal Media. Journal of Elasticity. 104. 187-204. 10.1007/s10659-011-9333-6.
[18] Sorrell, Steve (2009) ”The Rebound Effect: definition and estimation”, in Joanne Evans and Lester Hunt (eds) International Handbook on the
Economics of Energy, Cheltenham, Edward Elgar
[19] Adam K. Glaser, Ye Chen, and Jonathan T. C. Liu, ”Fractal propagation method enables realistic optical microscopy simulations in biological
tissues,” Optica 3, 861-869 (2016)
[20] Bovill, Carl. (2000). Fractal geometry as design aid. Journal for Geometry and Graphics Volume. 4. 71-78.
[21] Mendler, Nick and Vincent J. Matsko. ”Symmetric Binary Trees with Branching Ratios Larger than 1,” Proceedings of Bridges 2017: Mathe-
matics, Art, Music, Architecture, Education, Culture, 507–510.
[22] P. Smid, CNC Programming Handbook 2nd Edition, New York: Industrial Press Inc (2003)
