Hindawi

International Journal of Biomaterials

Volume 2022, Article ID 4541450, 7 pages
https://doi.org/10.1155/2022/4541450

Research Article
Implementation of Taguchi and Genetic Algorithm Techniques for
Prediction of Optimal Part Dimensions for Polymeric
Biocomposites in Fused Deposition Modeling

Raman Kumar ,1 Jasgurpreet Singh Chohan ,1 Sandeep Singh ,2 Shubham Sharma ,3

Yadvinder Singh,3 and S. Rajkumar 4
1
Department of Mechanical Engineering, Chandigarh University, Gharuan 140413, India
2
Department of Civil Engineering, Chandigarh University, Gharuan 140413, India
3
Department of Mechanical Engineering, I.K. Gujral Punjab Technical University, Jalandhar-Kapurthala Highway,
VPO Ibban 144603, India
4
Department of Mechanical Engineering, Faculty of Manufacturing, Institute of Technology, Hawassa University,
Awasa, Ethiopia

Correspondence should be addressed to Shubham Sharma; shubham543sharma@gmail.com and S. Rajkumar; ccetraj@gmail.com

Received 6 November 2021; Revised 1 January 2022; Accepted 6 January 2022; Published 31 January 2022

Academic Editor: Nicholas Dunne

Copyright © 2022 Raman Kumar et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Additive manufacturing has gained popularity among material scientists, researchers, industries, and end users due to the flexible,
low cost, and simple manufacturing process. Among number of techniques, fused deposition modeling (FDM) is the most
recognized technology due to easy operation, lower environmental degradation, and portable apparatus. Despite numerous
advantages, the limitations of this technique are poor surface finish, dimensional accuracy, and mechanical strength, which must
be improved. The present study focuses on the implementation of the genetic algorithm and Taguchi techniques to achieve
minimum dimensional variability of FDM parts especially for polymeric biocomposites. The output has been measured using
standard testing techniques followed by Taguchi and genetic algorithm analyses. Four response variables were measured and were
converted into single variable with combination of different weightages of each response. Maximum weightage was given to width
of FDM polymeric biocomposite parts which may play critical role in biomedical and aerospace applications. The advanced
optimization and production techniques have yielded promising results which have been validated by advanced algorithms. It was
found that layer thickness and orientation angle were significant parameters which influenced the dimensional accuracy whereas
best fitness value was 0.377.

1. Introduction fused deposition modeling (FDM) is the most adapted and

utilized technique due to lower installation cost and ease of
Additive manufacturing technologies manufacture the part operation [6]. The step-by-step procedure of manufacturing
through layer-by-layer strategy as opposite to conventional is shown in Figure 1.
subtractive manufacturing techniques [1]. The major ad- The apparatus of FDM contains extrusion head, nozzle,
vantage of these advanced manufacturing techniques over platform, motors, and microcontroller, which controls the
traditional manufacturing techniques is digitalization of the whole operation [7, 8]. The schematic of the FDM process
process which receives input form computer-generated along with major components is shown in Figure 2. As one
product designs [2–4]. The rapid production and custom- layer is actually deposited, build platform moves downwards
ization of parts with low cost and lower tooling requirements (in Z direction), and subsequent level of material is actually
also add to the advantages of these manufacturing strategies deposited, and the process is actually repeated till the desired
[5]. Out of numerous additive manufacturing techniques, part is actually attained [9–11]. At times, another filament of
2 International Journal of Biomaterials

Part STL File Tool Path Post

Slicing Fabrication
Designing Conversion Generation Processing

Figure 1: Step-by-step procedure of the FDM process.

and 0° resulted in higher tensile strength in direction parallel to

Rollers deposition of filament during the FFF process. On the other
hand, positive air gap resulted in smooth surface, which also
Build material
improves the shore D hardness. Gao et al. [24] added poly-
Heated Extrusion head
ethylene glycol with different concentrations inside polylactic
acid for strength enhancement. It was noticed that bond
Y strength between the layers has significantly increased, whereas
X mechanical anisotropy was reduced. The intermolecular dif-
Nozzle fusion and entanglement at bond location were found to be the
Part Semi molten bead
most possible reason for strength enhancement. In another
study [25], the fracture toughness of continuous carbon fiber
reinforced nylon composite by varying the printing speed, bed
temperature, and nozzle temperature. It was observed that
fracture toughness reduces with printing speed increases,
whereas an improvement has been noticed with an increase in
Build platform nozzle and bed temperature. When compared to their tradi-
Z (Base) tional equivalents, FFF-fabricated polymer components have
weak and anisotropic mechanical characteristics [26].
Many researchers have implemented the advanced op-
timization tools, artificial intelligence, and machine learning
Figure 2: Schematic and components of FDM. approached. Xue et al. [27] developed a variational
autoencoder based upon machine learning and Bayesian
washable material is actually utilized to allow for overhanging optimization for designing a 3D printed prototype with
part that is very easily washed away after fabrication [12]. customized macroscopic elastic properties. Goh et al. [28]
FFF supplies the personalized products with minimum implemented the neural network technique for exploring the
lead time and manufacturing, cost but the physical strength relationship between process parameters and mechanical
of part is surely a situation of interest for researchers as strength of PolyJet 3D printed parts. Finally, the genetic
extensive variation in physical properties is actually expe- algorithm was used to identify optimum design conditions
rienced because of perturbation in design [13–15]. Addi- to attain desired shore D hardness. Another study reported
tionally, issues related to lower physical strength of FFF parts [29] that the hierarchical machine learning was imple-
might impede the usability of the products for particular mented on 3D printed silicone elastomer using freeform
programs [16, 17]. Hence, there is surely a necessity of reversible embedding, which is difficult due to the need to
intelligent optimization tools for maximization and pre- deposit a Newtonian prepolymer liquid phase within a
diction of physical strength of FFF parts. There are many Bingham plastic support bath. The printed speed was in-
input parameters of FFF technology, which have a con- creased more than twice using this optimization tool,
siderable effect on tensile strength, compressive strength, whereas mechanical strength was not compromised.
flexural, and impact strength of FFF part [18–20]. Even- Despite several advantages and potential applications, the
though many research studies have been carried out for major challenge faced by machine learning and artificial in-
optimization of process parameters of FFF, recent research telligence in 3D printing are data acquisition, computational
has focused on development of advanced mathematical tools cost, and standards for qualification [30]. Furthermore, in the
and hybrid algorithms which may enhance and forecast the field of bioprinting, machine learning can be used for opti-
physical strength of FFF parts [21–26]. Next section dis- mization of process parameters, minimization of dimensional
cusses about recent literature on impact of FFF process variability in implants, manufacturing fault detection, and
parameters on mechanical stability and implementation of estimation of morphological properties of materials [31].
sophisticated and hybrid algorithms employed for optimi-
zation of process parameters of FFF technology [23]. 3. Experimentation
2. Literature Review 3.1. Planning of Work. The secondary data have been used
for analysis through Taguchi and genetic algorithm pro-
There are numerous process parameters of FFF technology cesses. Five parameters have been used with three levels
which have a significant impact of surface quality, mechanical each, while four dimensions are measured as given in
strength, and hardness of fabricated parts. Raster angles of 90° Table 1. The data of initial and final dimensions of width,
International Journal of Biomaterials 3

Table 1: Input and output parameters used for analysis.

Factors Responses
Exp.
No. Layer thickness Orientation angle Raster angle Raster width Air gap Mod
(mm) A (°) B (°) C (mm) D (mm) E W � 0.7ΔW + 0.1ΔL + 0.1ΔT + 0.1ΔD
1 0.127 0 0 0.4064 0 0.816457
2 0.127 15 0 0.4564 0.004 0.806392
3 0.127 30 0 0.5064 0.008 0.951515
4 0.127 0 30 0.4564 0.004 0.883827
5 0.127 15 30 0.5064 0.008 0.851524
6 0.127 30 30 0.4064 0 0.686596
7 0.127 0 60 0.5064 0.008 0.791827
8 0.127 15 60 0.4064 0 0.979389
9 0.127 30 60 0.4564 0.004 0.928935
10 0.178 0 0 0.4564 0.008 0.507068
11 0.178 15 0 0.5064 0 1.072302
12 0.178 30 0 0.4064 0.004 0.906102
13 0.178 0 30 0.5064 0 0.731976
14 0.178 15 30 0.4064 0.004 0.899691
15 0.178 30 30 0.4564 0.008 0.819047
16 0.178 0 60 0.4064 0.004 0.66351
17 0.178 15 60 0.4564 0.008 0.999945
18 0.178 30 60 0.5064 0 0.727309
19 0.254 0 0 0.5064 0.004 0.849293
20 0.254 15 0 0.4064 0.008 1.35796
21 0.254 30 0 0.4564 0 1.223476
22 0.254 0 30 0.4064 0.008 1.096522
23 0.254 15 30 0.4564 0 1.433019
24 0.254 30 30 0.5064 0.004 0.962855
25 0.254 0 60 0.4564 0 0.956191
26 0.254 15 60 0.5064 0.004 1.244108
27 0.254 30 60 0.4064 0.008 1.148151

length, diameter, and thickness have been used to convert FFF process parameters, i.e., layer thickness, orientation
into single response with different weightages. angle, raster angle, raster thickness, and air gap has been
The maximum weightage of 70% is given to width (W), studied on dimensional accuracy of parts. The genetic al-
while equal weightage of 10% is given to length (L), diameter gorithm approach has been implemented on to calculate
(D), and thickness (T). The equation used for conversion is Mod W, which is the output of four different dimensional
as follows: accuracy parameters with different weightages.
The output in form Mod W consists weightages given
Mod W � 0.7ΔW + 0.1ΔL + 0.1ΔT + 0.1Δ D. (1)
to different response variables which have been initially
evaluated using Taguchi analysis. Figure 3 shows the mean
4. Results and Discussion and SN ratio graphs of output and defines the relationship
between input and response. It must be noted that layer
The genetic algorithm is a method based on natural se- thickness is the most prominent parameter followed by
lection, the mechanism that drives biological evolution, orientation angle. The layer thickness of 0.178 mm yielded
for addressing both limited and unconstrained optimi- the maximum value of SN ratio which signifies better
zation problems. A population of individual solutions is dimensional stability. Furthermore, the orientation angle
repeatedly modified by the genetic algorithm. At each of 0° was optimum for attaining better dimensional ac-
phase, the genetic algorithm chooses parents at random curacy. In case of air gap, the SN ratio is maximum at
from the current population and utilizes them to generate 0.004 mm, whereas it is reduced by maximum and min-
the following generation’s children. The population imum values of air gap, i.e., 0 mm and 0.008 mm, re-
“evolves” toward an ideal solution over generations. The spectively. The impact of raster angle and raster width is
genetic algorithm may be used to handle a number of minimum on SN ration of dimensional accuracy. The SN
optimization problems that are not well suited for tra- ratio is maximum at 0° and 60° raster angle settings,
ditional optimization techniques, such as issues with whereas 0.4064 raster width yielded better dimensional
discontinuous, nondifferentiable, stochastic, or highly stability.
nonlinear objective functions. The evolutionary algo- The significance value and rank of each parameter are
rithm can be used to solve issues involving mixed integer given in Table 2, as derived from Taguchi analysis.
programming, in which certain components must be The equation has been generated for Mod W using re-
integer-valued. In the present study, the impact of five gression analysis and described as
4 International Journal of Biomaterials

Main Effects Plot for Means

Data Means
A B C D E
1.15

1.10

1.05
Mean of Means

1.00

0.95

0.90

0.85

0.80

0.127 0.178 0.254 0 15 30 0 30 60 0.4064 0.4564 0.5064 0.000 0.004 0.008

(a)
Main Effects Plot for SN ratios
Data Means
A B C D E
2.0

1.5

1.0
Mean of SN ratios

0.5

0.0

-0.5

-1.0

0.127 0.178 0.254 0 15 30 0 30 60 0.4064 0.4564 0.5064 0.000 0.004 0.008

Signal-to-noise: Smaller is better

(b)

Figure 3: (a) Mean plot. (b) SN ratio plot.

Table 2: Response table for means.

Level A B C D E
1 0.8552 0.8107 0.9434 0.9505 0.9585
2 0.8141 1.0716 0.9295 0.9509 0.9050
3 1.1413 0.9282 0.9377 0.9092 0.9471
Delta 0.3272 0.2609 0.0139 0.0417 0.0536
Rank 1 2 5 4 3
International Journal of Biomaterials 5

Table 3: ANONA analysis of response parameters.

Source DF Seq SS Seq MS F value P value Percentage contribution
A 2 56.100 23.0502 12.37 0.001 54.06
B 2 36.797 13.3986 7.19 0.006 38.35
C 2 0.063 0.0315 0.02 0.983 0.06
D 2 0.420 0.2100 0.11 0.894 0.40
E 2 0.571 0.2855 0.15 0.859 0.55
Error 16 9.804 1.8627 9.44
Total 26 103.755

Best: 0.37773 Mean: 0.377876

1
Fitness value

0.5

0
0 50 100 150 200 250 300 350 400 450 500
Generation
Best fitness
Mean fitness
Current Best Individual
Current best individual

0.5

0
1 2 3 4 5
Number of variables (5)

Fitness Scaling
10
Expectation

0
0.377 0.378 0.379 0.38 0.381 0.382 0.383 0.384 0.385
Raw scores
Figure 4: Fitness scaling and best values predicted by the genetic algorithm.

Mod W � −3.58 + 0.0A + 0.0096B − 0.0064C + 20.0 D − 71.4E

+ 40.2A ∗ A − 0.000898B ∗ B + 0.000012C ∗ C − 17.9 D ∗ D
− 295E ∗ E − 0.0121A ∗ B − 0.0302A ∗ C − 27.7A ∗ D (2)
+ 155A ∗ E − 0.000100B ∗ C + 0.0478B ∗ D
+ 1.167B ∗ E + 0.0232C ∗ D + 0.520C ∗ E.

It can be observed that parameter A, i.e., layer thickness means as shown in Figure 4. The charts are plotted between
has the maximum impact in dimensional accuracy followed fitness value vs. generation, current best value vs. variable, and
by orientation angle. The analysis using ANOVA has been expectations vs. raw sores. The results yielded by the genetic
carried and is given in Table 3. In the present study, the algorithm optimized and predicted the results with higher
R-squared value is 71.28%. accuracy as compared to conventional optimization tech-
Similar results have been observed after ANOVA niques. It was predicted that optimum parameter settings
analysis which indicates that maximum contribution of would be 0.127, 0, 0, 0.4064, and 0.008 for layer thickness,
54.06% and 38.35% of layer thickness and orientation angle orientation angle, raster angle, raster width, and air gap, re-
has been measured. The analysis using the genetic algorithm spectively, with objective function value of 0.377730056.
has been performed, and charts are derived which show the The efficacy of the genetic algorithm is validated as
fitness scaling, current best value, and overall best values and previous studies have found similar results, but time and
6 International Journal of Biomaterials

Table 4: Genetic algorithm parameters. processed parts using fuzzy logic,” Journal for Manufacturing
Science & Production, vol. 13, no. 3, pp. 183–197, 2013.
Solver Gamultiobj-multiobjective optimization
[3] F. Górski, W. Kuczko, and R. Wichniarek, “Influence of
Population vector Double vector process parameters on dimensional accuracy of parts man-
Population size 50 ufactured using fused deposition modelling technology,”
Creation function Constraint dependent Advances in Science and Technology Research Journal, vol. 7,
Crossover function Intermediate no. 19, pp. 27–35, 2013.
Scaling function Tournament [4] M. N. Sudin, S. A. Shamsudin, and M. A. Abdullah, “Effect of
part features on dimensional accuracy of FDM model,” APRN
Journal of Engineering and Applied Sciences, vol. 11,
efforts done to achieve the results were more along with pp. 8067–8072, 2016.
errors in calculations. The parameters used for the genetic [5] J. S. Chohan, R. Singh, K. S. Boparai, R. Penna, and
algorithm are given in Table 4. F. Fraternali, “Dimensional accuracy analysis of coupled fused
The dimensional accuracy of FDM-printed polymeric deposition modeling and vapour smoothing operations for
biocomposite parts plays the vital role as medical implants biomedical applications,” Composites Part B: Engineering,
are installed in human body, and dimensional variability vol. 117, pp. 138–149, 2017.
may cause postsurgery complications [30]. [6] M. S. Alsoufi and A. E. Elsayed, “Surface roughness quality
and dimensional accuracy—a comprehensive analysis of
100% infill printed parts fabricated by a personal/desktop
5. Conclusions cost-effective FDM 3D printer,” Materials Sciences and Ap-
plications, vol. 9, no. 1, pp. 11–40, 2018.
Fused deposition modeling (FDM) of polymeric bio- [7] P. Kumar, R. Singh, and I. P. S. Ahuja, “Investigations on
composites has been extensively used for rapid tooling, dimensional accuracy of the components prepared by hybrid
medical implants, aerospace components, engine parts, and investment casting,” Journal of Manufacturing Processes,
household equipment. The process parameters of FDM have vol. 20, pp. 525–533, 2015.
significant impact on surface roughness and dimensional [8] P. K. Gurrala and S. P. Regalla, “DOE based parametric study
variability which must be minimized to increase the usability of volumetric change of FDM parts,” Procedia Materials
and applicability of this technique. The dimensional accu- Science, vol. 6, pp. 354–360, 2014.
racy of FDM-printed polymeric biocomposite parts plays a [9] M. Kaveh, M. Badrossamay, E. Foroozmehr, and
vital role as medical implants are installed in human body, A. Hemasian Etefagh, “Optimization of the printing pa-
and dimensional variability may cause postsurgery com- rameters affecting dimensional accuracy and internal
plications. Also, in some situations, the weightage given in cavity for HIPS material used in fused deposition modeling
every dimension of part is not same; hence, there is need of processes,” Journal of Materials Processing Technology,
the advanced optimization technique which could provide vol. 226, pp. 280–286, 2015.
[10] S. Saqib and J. Urbanic, “An experimental study to determine
accurate results for complex situations. In the present study,
geometric and dimensional accuracy impact factors for fused
unequal significance was given to response variable, and a deposition modelled parts,” in Enabling Manufacturing
combined response factor has been optimized using the Competitiveness and Economic Sustainability, pp. 293–298,
genetic algorithm. The results achieved were better as Springer, Berlin, Germany, 2012.
compared to Taguchi analysis and lesser time taken for [11] B. Singh, R. Kumar, and J. S. Chohan, “Polymer matrix
finding the best value. The predicted results were validated composites in 3D printing: a state of art review,” Materials
with 99.12% accuracy which indicated the improved efficacy Today: Proceedings, vol. 33, 2020.
of the genetic algorithm. [12] G. Papazetis and G.-C. Vosniakos, “Feature-based process
parameter variation in continuous paths to improve di-
mensional accuracy in three-dimensional printing via ma-
Data Availability terial extrusion,” Proceedings of the Institution of Mechanical
The data used to support the findings of this study are Engineers, Part B: Journal of Engineering Manufacture,
vol. 233, no. 12, pp. 2241–2250, 2019.
available from the corresponding author upon request.
[13] A. N. M. A. Tomal, T. Saleh, and M. R. Khan, “Improvement
of dimensional accuracy of 3-D printed parts using an ad-
Conflicts of Interest ditive/subtractive based hybrid prototyping approach,” IOP
Conference Series Materials Science and Engineering, vol. 260,
The authors declare that they have no conflicts of interest. Article ID 012031, 2017.
[14] Y. R. Kumar, “An application of Taguchi’s technique to
References improve the accuracy of rapid prototyped FDM parts,” In-
ternational Journal of Materials Engineering Innovation,
[1] R. C. Pennington, N. L. Hoekstra, and J. L. Newcomer, vol. 3, no. 3-4, pp. 228–246, 2012.
“Significant factors in the dimensional accuracy of fused [15] J. S. Chohan, R. Singh, and K. S. Boparai, “Effect of process
deposition modelling,” Proceedings of the Institution of Me- parameters of fused deposition modeling and vapour
chanical Engineers, Part E: Journal of Process Mechanical smoothing on surface properties of ABS replicas for
Engineering, vol. 219, no. 1, pp. 89–92, 2005. biomedical applications,” in Additive Manufacturing of
[2] R. K. Sahu, S. S. Mahapatra, and A. K. Sood, “A study on Emerging Materials, pp. 227–249, Springer, Berlin,
dimensional accuracy of fused deposition modeling (FDM) Germany, 2019.
International Journal of Biomaterials 7

[16] J. S. Chohan, R. Singh, and K. S. Boparai, “Multi response machine learning,” 3D Printing and Additive Manufacturing,
optimization and process capability analysis of fused filament vol. 6, no. 4, pp. 181–189, 2019.
fabrication and chemical vapor smoothing operations for [30] G. D. Goh, S. L. Sing, and W. Y. Yeong, “A review on machine
rapid casting of biomedical implants,” in Proceedings of the learning in 3D printing: applications, potential, and chal-
International Manufacturing Science and Engineering Con- lenges,” Artificial Intelligence Review, vol. 54, no. 1, pp. 63–94,
ference, American Society of Mechanical Engineers, Erie, PA, 2021.
USA, June 2019. [31] C. Yu and J. Jiang, “A perspective on using machine learning
[17] R. Kumar and H. Singh, “Exploring the success factors for in 3D bioprinting,” International Journal of Bioprinting, vol. 6,
examining the potential of manufacturing system output,” no. 1, p. 253, 2020.
Benchmarking: An International Journal, vol. 25, no. 4,
pp. 1171–1193, 2018.
[18] H. Singh and R. Kumar, “Hybrid methodology for measuring
the utilization of advanced manufacturing technologies using
AHP and TOPSIS,” Benchmarking: An International Journal,
vol. 20, no. 2, pp. 169–185, 2013.
[19] H. Singh and R. Kumar, “Measuring the utilization index
of advanced manufacturing technologies: a case study,”
IFAC Proceedings Volumes, vol. 46, no. 9, pp. 899–904,
2013.
[20] J. S. Chohan, R. Singh, and K. S. Boparai, “Post-processing
of ABS replicas with vapour smoothing for investment
casting applications,” Proceedings of the National Academy
of Sciences, India Section A: Physical Sciences, pp. 1–6,
2020.
[21] R. Kumar, J. S. Chohan, R. Goyal, and P. Chauhan, “Impact of
process parameters of resistance spot welding on mechanical
properties and micro hardness of stainless steel 304 weld-
ments,” International Journal of Structural Integrity, vol. 12,
no. 3, pp. 366–377, 2020.
[22] R. Kumar, R. Kumar, J. S. Rai, and N. S. Virk, “Analysis the
effects of process parameters in EN24 alloy steel during CNC
turning by using MADM,” International Journal of Innovative
Research in Science, Engineering and Technology, vol. 2,
pp. 1131–1145, 2013.
[23] N. K. Dixit, R. Srivastava, and R. Narain, “Comparison of two
different rapid prototyping system based on dimensional
performance using grey relational grade method,” Procedia
Technology, vol. 25, pp. 908–915, 2016.
[24] X. Gao, S. Qi, D. Zhang, Y. Su, and D. Wang, “The role of poly
(ethylene glycol) on crystallization, interlayer bond and
mechanical performance of polylactide parts fabricated by
fused filament fabrication,” Additive Manufacturing, vol. 35,
Article ID 101414, 2020.
[25] G. D. Goh and W. Y. Yeong, “Mode I interlaminar fracture
toughness of additively manufactured carbon fibre thermo-
plastic,” in Proceedings of the 3rd International Conference on
Progress in Additive Manufacturing, pp. 505–510, Singapore,
May 2018.
[26] X. Gao, S. Qi, X. Kuang, Y. Su, J. Li, and D. Wang, “Fused
filament fabrication of polymer materials: a review of inter-
layer bond,” Additive Manufacturing, vol. 37, Article ID
101658, 2021.
[27] T. Xue, T. J. Wallin, Y. Menguc, S. Adriaenssens, and
M. Chiaramonte, “Machine learning generative models for
automatic design of multi-material 3D printed composite
solids,” Extreme Mechanics Letters, vol. 41, Article ID 100992,
2020.
[28] G. D. Goh, S. L. Sing, Y. F. Lim et al., “Machine learning for
3D printed multi-materials tissue-mimicking anatomical
models,” Materials & Design, vol. 211, Article ID 110125,
2021.
[29] A. Menon, B. Póczos, A. W. Feinberg, and N. R. Washburn,
“Optimization of silicone 3D printing with hierarchical