Optimization of 3D Printing Process Parameters on Tensile Strength
of ABS Filament Material Product using Taguchi Method
Farizi Rachman, Bayu Wiro Kurniawan and Mega Tri Yoningtias
Design and Manufacture Engineering, Shipbuilding Institute of Polytechnic Surabaya, Surabaya, Indonesia
Keywords: 3D Printing, Tensile Strength, Optimization, Design of Experiment, Taguchi.
Abstract: The world of the manufacturing industry continues to develop from time to time and one of the technologies
that support the manufacturing industry in the era of the industrial revolution 4.0 is 3D printing technology.
It is necessary to improve the quality of 3D printing products, especially in their mechanical properties,
namely tensile strength. Low tensile strength can cause the product to experience mechanical failure such as
yielding or fracture, it is necessary to combine the right and optimal process parameters in the 3D printing
process. The test specimens were designed using Fusion 360 software according to the ASTM D638-14 type
IV standard. The material used is ABS filament. The process parameters used are layer height 0.1 mm, 0.15
mm, 0.2 mm, infill pattern with grid, line, cross pattern and nozzle temperature 230°C, 240°C, 250°C. The
Design of Experiment used the Taguchi method in the form of an orthogonal matrix L
9
(3
4
) with three
replications. From the experimental results, it is found that the most optimal combination of process
parameters is layer height at level 3 with a value of 0.2 mm, infill pattern at level 2 with line pattern and
nozzle temperature at level 1 with a value of 230°C.
1 INTRODUCTION
3D printing is one type of additive manufacturing,
where a filament is processed layer by layer which is
controlled by a computer and produces a three-
dimensional product (Riza, Budiyantoro and
Nugroho, 2020). 3D printing brings many benefits to
engineering design, product development, and
production processes (Nguyen, Huynh, Nguyen,
Tran, 2020). The potential market for 3D printing
globally in 2025 is predicted to reach US$ 230-550
billion spread across several manufacturing industry
sectors, including consumer goods (US$ 100-300
billion), production of medical equipment and
transportation (US$ 100-200 billion), and moulding
production (US$ 30-50 billion). The presence of 3D
printing technology has created new businesses in the
creative industry sector, some examples of which are:
Spuni, printing products for baby needs, such as
tablespoons whose shape is adjusted to the shape of
the baby's mouth. Technologia Humana 3D, prints a
replica of the baby's fetus in three-dimensional form
after doing an ultrasound for pregnant women. Make
Eyewear, which can print the necessary glasses and
supplies. Shapeways, printing models for prototypes
ordered from around the world through online
channels (Kusuma, 2016). The use of 3D printing has
also increased (Nicholson, 2014) in other business
environments such as the automotive industry and
aerospace engineering. Spare parts, for example, are
being manufactured in the automotive and aerospace
industries leading to increased economies of scale.
3D printing is changing the way industrial production
lines work, causing analysts to consider the advent of
3D printers as the second industrial revolution
(Mpofu, Mawere and Mukosera, 2014). There are
several technologies used in the rapid prototyping
(RP) process, namely Stereolithography (SLA),
Selective Laser Sintering (SLS), Fused Deposition
Modeling (FDM), Laminated Object Manufacturing
(LOM), 3-D Printing (3DM), and others
(Nancharaiah, 2011). Fused Deposition Modeling
(FDM) is a process of melting, extrusion, and
deposition of thermoplastic filaments among various
types of additive manufacturing (AM) techniques
(Gebisa and Lemu, 2019).
More and more 3D printing products are made for
various functions, it is necessary to improve the
quality of these products, one of which is the
mechanical properties in the form of tensile strength
in 3D printing materials because low tensile strength
can cause the product to experience yielding or
284
Rachman, F., Kurniawan, B. and Yoningtias, M.
Optimization of 3D Printing Process Parameters on Tensile Strength of ABS Filament Material Product using Taguchi Method.
DOI: 10.5220/0010944200003260
In Proceedings of the 4th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2021), pages 284-290
ISBN: 978-989-758-615-6; ISSN: 2975-8246
Copyright
c
2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
fracture. Tensile testing is carried out to determine the
amount of tensile strength that can be resisted by the
material or product resulting from 3D printing.
The method that is often used for optimizing
process parameters is the Taguchi method. The
Taguchi method aims to optimize process parameters,
improve product quality and reduce production costs
so that a perfect/robust product is produced against
the noise factor (Soejanto, 2008). The primary
purpose of the experimental design technique is to
understand the interactions among the parameters,
which could help in the optimization of experimental
parameters and provide a statistical model
(Montgomery, 2014).
This research was conducted to obtain the optimal
combination of process parameters on the FDM 3D
printing machine, such as layer height, infill pattern,
and nozzle temperature using ABS (Acrylonitrile
Butadiene Styrene) filament material on tensile
strength and will be analyzed using the Design of
Experiment (DoE), namely the Taguchi method.
Parameters that have the most influence on tensile
strength will be identified.
2 LITERATURE REVIEW
2.1 3D Printing
3D printing or additive manufacturing is a technique
of printing products in three-dimensions form by
adding computer material layer by layer which is
controlled by a computer according to design made
using CAD (Computer-Aided Design) software. 3D
printing is the process of creating 3D solid objects
from almost any shape from a digital model (Mpofu,
Mawere and Mukosera, 2014). A typical 3D printer
contains a nozzle printing that can handle one or more
feed materials that can be moved in three dimensions
(x, y, z). The section is made layer by layer, guided
by 3D CAD models and stereolithography
(Abeykoon, Sri-Amphorn and Fernando, 2020).
2.2 FDM (Fused Deposition Modeling)
Fused Deposition Modeling (FDM) is a technique
that is usually used for 3D printing and additive
manufacturing technology to print prototypes or
products. Fused Deposition Modeling (FDM)
technology uses a hot extruder where the filament or
thermoplastic material is preheated in the extruder
until it melts and is extruded through a nozzle and
then moved to make layers to form the desired
product (Rinanto and Sutopo, 2017).
2.3 ABS (Acrylonitrile Butadiene
Styrene)
ABS is an engineering plastic whose portion
butadiene is evenly distributed over a matrix
acrylonitrile-styrene. ABS has excellent toughness,
good dimensional stability, easy processing
capability, chemical resistance, and low cost
(Cabezas Arribas, 2017). ABS is impact resistant, has
relatively high heat resistance, and is extruded at an
ambient temperature of 210°C. When printing using
ABS, a heated printing mat covered with polyimide
tape regulated at about 110°C is almost necessary.
Otherwise, the parts may warp when cold and may
even break into larger molded materials. ABS is a
widely used material by the 3D community printer
individual, although the recent increase in the
availability of PLA has changed that (J., 2012).
2.4 Tensile Strength
Tensile properties are often used as a determinant of
properties and as an indication of polymer strength.
Tensile properties measure the ability of a material to
withstand separate tensile forces, and prolong
deformation before breaking occurs. Tensile test
results data such as tensile strength, elongation and
modulus of elasticity can be used to select polymeric
materials as specific applications from a large group
(Divyathej, Varun and Rajeev, 2016).
2.5 Design of Experiment (DoE)
Design of Experiment is a joint evaluation of two or
more factors (parameters) on their ability to influence
the average or variability of the combined results of a
certain product or process characteristics. To achieve
this effectively and statistically, the levels of the
control factors are varied, the results of certain
combinations of tests are observed, and the complete
result set is analyzed to determine which factors are
influencing and the degree to which they are
favourable and if the levels increase or decrease will
result in further improvements (Soejanto, 2009).
2.6 Taguchi Method
Dr. Genichi Taguchi was the one who first introduced
the quality engineering methodology in the 1960s,
which until now is known as the Taguchi method
(Soejanto, 2008). Taguchi design provides a robust
and efficient method for designing processes that
consistently perform optimally under a wide range of
conditions. Strategically designed experiments
Optimization of 3D Printing Process Parameters on Tensile Strength of ABS Filament Material Product using Taguchi Method
285
should be used to determine the optimal design,
exposing the process to various levels of design
parameters (Semioshkina and Voigt, 2006). The
target of the Taguchi method is to improve product
quality, by looking for variables that affect quality,
then separating them into control factors and
uncontrollable factors (noise).
The calculation of degrees of freedom is carried
out to calculate the minimum number of experiments
that must be carried out to investigate the observed
factors. The form of the orthogonal matrix equation
in determining the number of experiments to be
observed is as follows:
=
(
)
(
)
(1)
=(−1)
(2)
The equation of degrees of freedom to find out the
total of degrees of freedom is as follows:
=(−1)
(3
)
Orthogonal matrices are used to analyze
experimental data and are used to design efficient
experiments to determine the minimum number of
experiments that can provide as much information as
possible on all the factors that affect the parameters.
The general form of an orthogonal matrix is as
follows: (Soejanto, 2008)
(
)
(4)
With:
L = Latin square design
a = Many experiment / lines
b = Many level
c = Many factor / column
The calculation of the S/N ratio is used to
determine the factors that have contributed to an
experiment. There are several types of quality
characteristics of the S/N ratio, which are as follows:
(Soejanto, 2008)
1 Smaller is Better
The measurement of the characteristics of smaller is
better, the ideal target is zero and the measurement
when the value is lower, the quality will be better.
Examples: roughness, number of failed products,
defects, time efficiency, and others. The S/N value for
the smaller is better characteristic type according to
equation 5 is as follows:
⁄
=−10log
∑
(5
)
2 Larger is Better
The measurement of the characteristics of larger is
better is that when the expected output is greater in
value, the better the quality. Example: the number of
production, the number of sales, tensile strength, and
others. The S/N value for the larger is better
characteristic type according to equation 6 is as
follows:
⁄
= −10
∑
(6
)
3 Nominal is Better
The measurement of the nominal is better
characteristic is usually assigned a certain nominal
value by the user (user-defined). The closer to the
nominal value that has been set, the better the quality
will be. Example: dimensions, weight, and others.
The value of S/N for the type of nominal is better
characteristic according to equation 7 is as follows:
⁄
= −10log
(7
)
With:
=
∑
=
∑
(
−)
n = Amount of data
= Observation response data to – i
The purpose of statistical Analysis of Variance
(ANOVA) is to investigate which design parameters
significantly affect the response variable of a test.
ANOVA is used to analyze experimental data that
occurs from the calculation of degrees of freedom
(df), the number of squares (sum of square, SS), the
middle square (mean of square, MS), and F
value
(Soejanto, 2009).
The confidence interval factor calculation was
used for the treatment conditions at the time of the
experiment. The confidence interval for optimal
conditions can be calculated using the following
equation:
=±
;
;
×
(8
)
μ
−CI
≤μ
≤μ
+CI
= Number of effective observations
=
(9
)
iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science
286
3 METODOLOGY AND
IMPLEMENTATION
The test specimen is designed according to the ASTM
D638-14 type IV standard using Autodesk Fusion 360
software and then the design file is exported in STL
format so that the file can be opened and parameters
set using the Ultimaker Cura software. This
experiment was carried out by combining the levels
of each predetermined process parameter, namely
layer height, infill pattern, and nozzle temperature.
The following are the levels for each process
parameter used:
Table 1: Free Variables and Control Levels.
Parameters
Layer
Height
(mm)
Infill
Pattern
Nozzle
Temperature
(°C)
Level 1 0,1 Grid 230
Level 2 0,15 Line 240
Level 3 0,2 Cross 250
In the calculation of the degrees of freedom
obtained a value of 6. The orthogonal array according
to this experiment is L
9
(3
4
). Furthermore, the
calculation of the total degrees of freedom of the
orthogonal array matrix using equation 3 obtained the
total degrees of freedom is 8. By the standard
orthogonal matrix, the value is greater than or equal
to the calculation of the degrees of freedom of the
experiment that has been applied, and then the
experiment is declared feasible. Table 2 is an
experimental design using Minitab19 software. The
following is Table 2:
Table 2: Experiment Design.
Experiment
Free Variables
A B C
1 1 1 1
2 1 2 2
3 1 3 3
4 2 1 2
5 2 2 3
6 2 3 1
7 3 1 3
8 3 2 1
9 3 3 2
The study was conducted by printing test specimens
of ABS filament material using an Ender 3 Pro brand
3D printer by a predetermined experimental design,
namely the orthogonal array matrix nine times and
replicated three times. Furthermore, the results of the
specimen print were carried out with a tensile test
using a Hung Ta brand tensile testing machine Type
HT-2402 to get the maximum tensile strength value.
The research data were calculated and analyze using
the Taguchi method. F
value
is used for hypothesis
testing by comparing F
value
for each factor with F
table
.
If F
value
> F
table
, then H
0
is rejected, and if F
value
< F
table
,
then H
0
is accepted.
• Test F
value
on layer height parameter
o H
0
: there is no effect of layer height on
tensile strength
o H
1
: there is an effect of layer height on tensile
strength
• Test F
value
on infill pattern parameter
o H
0
: there is no effect of infill pattern on
tensile strength
o H
1
: there is an effect of infill pattern on
tensile strength
• Test F
value
on nozzle temperature parameter
o H
0
: there is no effect of nozzle temperature
on tensile strength
o H
1
: there is an effect of nozzle temperature
on tensile strength
4 RESULT AND DISCUSSION
The maximum tensile test results data on each
specimen can be seen in Table 3. After the specimen
has fractured, the detailed tensile test results data
(Test Report) will come out and can be viewed on the
U.T.M. Testing Program software. Table 3 can be
seen below.
Table 3: Experiment Results Data.
Combinations
Ultimate Tensile Strength (N/mm²)
Replication
1
Replication
2
Replication
3
1 16.11 16.36 16.53
2 16.55 16.98 16.49
3 13.29 13.53 13.30
4 17.25 17.30 17.50
5 17.60 17.91 17.53
6 15.25 15.97 15.59
7 19.00 19.32 19.04
8 19.53 20.01 19.62
9 16.97 17.12 17.02
4.1 Value for Means and S/N Ratios
Tensile strength has the characteristic of larger is
better, the greater the value of the tensile strength, the
better because the stronger it is. The following are the
Optimization of 3D Printing Process Parameters on Tensile Strength of ABS Filament Material Product using Taguchi Method
287
results of the calculation of the means and the S/N
ratios. Table 4 can be seen below.
Table 4: Experiment Results Data.
Combi
nations
Free Variables
S/N
Ratios
Means
LH
(mm)
IP
NT
(°C)
1 0.1 Grid 230 24.260 16.333
2 0.1 Line 240 24.438 16.673
3 0.1 Cross 250 22.524 13.373
4 0.15 Grid 240 24.785 17.350
5 0.15 Line 250 24.949 17.680
6 0.15 Cross 230 23.860 15.603
7 0.2 Grid 250 25.629 19.120
8 0.2 Line 230 25.897 19.720
9 0.2 Cross 240 24.628 17.037
4.2 ANOVA (Analysis of Variance)
The results of the calculation of the ANOVA tensile
strength for means and S/N ratios can be seen in
Tables 5 and 6. The calculations were carried out
manually using Microsoft Excel and validated using
Minitab19 software. The confidence interval used is
= 0.05. Table 5 and 6 can be seen below.
Table 5: ANOVA Mean Results Data.
Source DF SS
Contribution
%
MS F
value
LH 2 15.086 53.624 7.543 97.182
IP 2 12.520 44.504 6.260 80.655
NT 2 0.371 1.320 0.186 2.392
Error 2 0.155 0.552 0.078
Total 8 28.132
Based on Table 5, it can be seen that the layer
height and infill pattern parameters have a significant
effect, while nozzle temperature does not have a
significant effect on the tensile strength response. The
layer height parameter gives a contribution
percentage of 53.624%, the infill pattern is 44.504%
and the nozzle temperature is 1.320%.
Table 6: ANOVA S/N Ratio Results Data.
Source DF SS
Contribution
%
MS F
value
LH 2 4.055 51.478 2.027 39.455
IP 2 3.560 45.205 1.780 34.647
NT 2 0.159 2.013 0.079 1.543
Error 2 0.103 1.305 0.051
Total 8 7.876
Based on the Table 6, it can be seen that the layer
height and infill pattern parameters have a significant
effect, while nozzle temperature does not have a
significant effect on the tensile strength response.
F
table
for the S/N ratio for each factor is F
(0.05;2;2)
=
19.00. The layer height parameter gives a
contribution percentage of 51.478%, the infill pattern
is 45.205% and the nozzle temperature is 2.013%.
4.3 Optimal Parameters
Calculation of optimal parameters is used to
determine the most optimal level of process
parameters on the tensile strength response. The
greater the averages value of each level, the greater
its contribution to the tensile strength response
variable. The following are the results of calculating
the optimal parameters for the mean and S/N ratio
which can be seen in tables 7 and 8.
Table 7: Optimal Parameters for Mean.
Level
Layer
Height
Infill
Pattern
Nozzle
Temperature
1 15.460 17.601 17.219
2 16.878 18.024 17.020
3 18.626 15.338 16.724
Delta 3.166 2.687 0.494
Rank 1 2 3
Table 7 shows that the optimal parameters for the
mean to tensile strength response are layer height at
level 3 with a value of 0.2 mm, infill pattern at level
2 with line pattern, and nozzle temperature at level 1
with a value of 230°C. Figure 1 is the main effects
plot for means according to the minitab19 software.
Figure 1: Main Effects Plot for Mean.
From the means graphs, the best combination of
layer height of 0.2 mm, infill pattern of line pattern,
and nozzle temperature of 230°C for high tensile
strength.
iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science
288
Table 8: Optimal Parameters for S/N Ratio.
Level
Layer
Height
Infill
Pattern
Nozzle
Temperature
1 23.741 24.892 24.672
2 24.531 25.094 24.617
3 25.384 23.670 24.367
Delta 1.644 1.424 0.305
Rank 1 2 3
Table 8 shows that the optimal parameters for the
S/N ratio to the tensile strength response are layer
height at level 3 with a value of 0.2 mm, infill pattern
at level 2 with line pattern and nozzle temperature at
level 1 with a value of 230°C. Figure 2 is a main effect
plot for S/N ratio according to the minitab19
software.
Figure 2: Main Effect Plot for S/N Ratio.
From the S/N ratios graphs, the best combination
of layer height of 0.2 mm, infill pattern of line pattern
and nozzle temperature of 230°C for high tensile
strength.
4.4 Confidence Interval
The following is the Optimal Condition Confidence
Interval Prediction for Mean Value.
1 Optimal Condition Prediction
=+
(
−
)
+
(
−
)
+
(
−
)
=16.988+
(
18.626−16.988
)
+
(
18.024−16.988
)
+
(
17.219−16.988
)
=19.893
2 Confidence Interval
=
=3.8571
=±
.;;
×.
.
=±0.6103
Optimal confidence interval for mean value
−
≤
≤
+
19.893−0.6103≤
≤19.893+0.6103
.≤
≤.
The Following is Optimal Condition Confidence
Interval Prediction for S/N Ratio Value.
1 Optimal Condition Prediction
=
+
(
−
)
+
(
−
)
+
(
−
)
= 24.552+
(
25.384−24.552
)
+
(
25.094−24.552
)
+
(
24.672−
24.552
)
=26.047
2 Confidence Interval
=
=3.8571
=±
.;;
×.
.
=±0.4966
Optimal confidence interval for S/N Ratio value
−
≤
≤
+
26.047−0.4966≤
≤26.047+0.4966
.≤
≤.
Based on the results of the calculation of the
confidence interval, the minimum confidence value is
19.283 and the maximum confidence value is 20.504
for the means, while for the S/N ratios, the minimum
confidence value is 25.550 and the maximum
confidence value is 26.543.
5 CONCLUSION
The Taguchi method can be used to find the optimal
combination of process parameters for one response
variable. Optimization aims to improve product
quality from better machining.
From the results of the study, it can be seen that
the 3D printing process parameters that affect the
tensile strength of ABS filament material products are
layer height and infill pattern with the results of the
hypothesis that F
value
is greater than F
table
, while the
nozzle temperature parameter does not have a
significant effect on the tensile strength response
variable with the results of the hypothesis F
value
less
than F
table
. The optimal combination of process
parameters to produce a product with the highest
tensile strength is layer height at level 3 with a value
Optimization of 3D Printing Process Parameters on Tensile Strength of ABS Filament Material Product using Taguchi Method
289
of 0.2 mm, infill pattern at level 2 with a line pattern,
nozzle temperature at level 1 with a value of 230°C.
REFERENCES
Abeykoon, C., Sri-Amphorn, P. and Fernando, A. (2020)
‘Optimization of fused deposition modeling parameters
for improved PLA and ABS 3D printed structures’,
International Journal of Lightweight Materials and
Manufacture, 3(3), pp. 284–297.
Cabezas Arribas, S. (2017) ‘Characteritation of 3D Printing
Parameters’, Bachelor Thesis.
Divyathej, M. V., Varun, M. and Rajeev, P. (2016)
‘Analysis of mechanical behavior of 3D printed ABS
parts by experiments’, International Journal of
Scientific & Engineering Research, 7(3), pp. 116–124.
Gebisa, A. W. and Lemu, H. G. (2019) ‘Influence of 3D
printing FDM process parameters on tensile property of
ultem 9085’, Procedia Manufacturing, 30, pp. 331–
338.
J., G. (2012) ‘3-D Printers’, in 3-D Printers for Libraries,
pp. 23–31.
Kusuma, I. E. (2016) ‘Business Model Development Based
on 3D Printer Technology with Product Service System
(Pss) Approach’, THESIS - TI142307.
Montgomery, D. C. (2014) Design and Analysis of
Experiments. 8th edn. New York: John Wiley and Sons.
Mpofu, T. P., Mawere, C. and Mukosera, M. (2014) ‘The
Impact and Application of 3D Printing Technology’,
International Journal of Science and Research (IJSR),
3(6), pp. 2148–2152.
Nancharaiah, T. (2011) ‘Optimization of Process
Parameters in FDM Process Using Design of
Experiments’, International Journal on Emerging
Technologies 2(1):, 2(1), pp. 100–102.
Nguyen, V. H. et al. (2020) ‘Single and multi-objective
optimisation of processing parameters for fused
deposition modelling in 3D printing technology’,
International Journal of Automotive and Mechanical
Engineering, 17(1), pp. 7542–7551.
Nicholson, D. (2014) ‘3D Printing-Approaching Critical
Mass In An $8 Billion Industry’, Forbes.com.
industry/?sh=7c89a0181075.
Rinanto, A. and Sutopo, W. (2017) ‘Development of Rapid
Prototyping Technology : Study Literature’, Journal of
Metris 18 (2017) 105 - 112. ISSN: 1411 - 3287,
18(January 2017), pp. 105–112.
Riza, E. I., Budiyantoro, C. and Nugroho, A. W. (2020)
‘Improving The Flexural Strength of PETG 3D Printer
Material Products by Optimizing Process Parameters
Using The Taguchi Method’, Machine Media :
Mechanical Engineering Magazine Vol. 21, No.2, July
2020, pp. 66–75.
Semioshkina, N. and Voigt, G. (2006) ‘An overview on
Taguchi Method’, Journal Of Radiation Research, 47
Suppl A(2), pp. A95–A100.
Soejanto, I. (2008) Quality Engineering : Experiment with
Taguchi Technique. Two. Publisher Humaniora
Foundation.
Soejanto, I. (2009)
Design Experiment with Taguchi
Method. Yogyakarta, Indonesia : Science House.
iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science
290
