polymers

Article
Optimization of Injection-Molding Process for Thin-Walled
Polypropylene Part Using Artificial Neural Network and
Taguchi Techniques
Mehdi Moayyedian, Ali Dinc and Ali Mamedov *

College of Engineering and Technology, American University of the Middle East, Kuwait;
mehdi.moayyedian@aum.edu.kw (M.M.); ali.dinc@aum.edu.kw (A.D.)
* Correspondence: ali.mamedov@aum.edu.kw; Tel.: +965-2225-1400

Abstract: Plastics are commonly used engineering materials, and the injection-molding process is
well known as an efficient and economic manufacturing technique for producing plastic parts with
various shapes and complex geometries. However, there are certain manufacturing defects related
to the injection-molding process, such as short shot, shrinkage, and warpage. This research aims to
find optimum process parameters for high-quality end products with minimum defect possibility.
The Artificial Neural Network and Taguchi Techniques are used to find a set of optimal process
parameters. The Analytic Hierarchy Process is used to calculate the weight of each defect in the





 proposed thin-walled part. The Finite Element Analysis (FEA) using SolidWorks plastics is used to
simulate the injection-molding process for polypropylene parts and validate the proposed optimal
Citation: Moayyedian, M.; Dinc, A.;
set of process parameters. Results showed the best end-product quality was achieved at a filling
Mamedov, A. Optimization of
time of 1 s, cooling time of 3 s, pressure-holding time of 3 s, and melt temperature of 230 ◦ C. The
Injection-Molding Process for
Thin-Walled Polypropylene Part
end-product quality was mostly influenced by filling time, followed by the pressure-holding time. It
Using Artificial Neural Network and was found that the margin of error for the proposed optimization methods was 1.5%, resulting from
Taguchi Techniques. Polymers 2021, any uncontrollable parameters affecting the injection-molding process.
13, 4158. https://doi.org/10.3390/
polym13234158 Keywords: injection molding; shrinkage; warpage; short shot; Taguchi; artificial neural network

Academic Editors: Jorge Manuel

Mercado-Colmenero, Cristina
Martín-Doñate and Miguel Ángel 1. Introduction
Rubio-Paramio
Plastics offer a wide range of advanced mechanical properties such as high strength-to-
weight ratio, flexibility, corrosion resistance, transparency, etc., which make them irreplace-
Received: 13 November 2021
able materials in various engineering fields such as automobile and aerospace industries,
Accepted: 25 November 2021
Published: 28 November 2021
electronics, and biomedical industries. For instance, plastics are found in a range of applica-
tions, such as in aircraft windshields, automobile windows, medical and dental equipment,
food and beverage packaging, and petrol-chemical storage. Injection-molding processes
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
are used for manufacturing plastic end-products for most of the aforementioned appli-
published maps and institutional affil-
cations. Injection-molding processes can be described in three simple phases: (1) filling
iations. phase, where the molten polymer is injected into a mold cavity with the desired shape; (2)
packaging phase, where high packing pressure is applied to ensure proper filling of the
mold cavity; and (3) cooling phase, where the temperature of the mold is decreased, and
the polymer solidifies. The quality of an end-product in injection molding is the result of a
complex combination of the process parameters, part/mold design, and material used. In
Copyright: © 2021 by the authors.
this paper, three common defects which reduce the quality of end-products are evaluated:
Licensee MDPI, Basel, Switzerland.
This article is an open access article
(1) short shot, (2) shrinkage rate, and (3) warpage. Warpage is defined as a serious defect in
distributed under the terms and
injection-molded parts, especially the thin-walled products [1]. Many researchers tried to
conditions of the Creative Commons minimize warpage defects using different techniques. Kurtaran and Erzurumlu [2] used
Attribution (CC BY) license (https:// response surface methodology and genetic algorithm to achieve the minimum warpage.
creativecommons.org/licenses/by/ Gao et al. [3] used the Kriging surrogate model trained by numerical simulation to acquire
4.0/). more stable product quality compared to traditional methods. Kim and Lee [4] used the

Polymers 2021, 13, 4158. https://doi.org/10.3390/polym13234158 https://www.mdpi.com/journal/polymers

Polymers 2021, 13, 4158 2 of 13

modified complex method and obtained warpage reduction over 70% by controlling wall
thickness and process parameters. Yin et al. [5] used backpropagation neural network
modeling for precise prediction of warpage in plastic parts. Injection molding shrinkage
deals with dimensional differences between a molded part and the cavity. The shrink-
age behavior of a molded plastic part plays an important role in determining the final
dimensions of the part [6]. Lotti et al. [7] used an Artificial Neural Networks approach
to predict the shrinkage of injection-molded plastic plaques. Tang et al. [8] and Hassan
et al. [9] studied a cooling system design in terms of cooling channel size and location for
multi-cavity injection molds to ensure uniform solidification inside a mold cavity, which
would prevent shrinkage. A short shot is the incomplete filling of a mold cavity, which
results in the production of an incomplete part. In general, a short shot occurs when
insufficient material is injected into the mold or flow freezes before the mold cavity is fully
filled [10]. It is caused by different factors such as the wrong plastic material selection,
incorrect processing parameters, incorrect mold design, and part design [11]. Moayyedian
et al. [12] mentioned that the cross-sectional shape of a gate or runner leads to short shots at
the filling stage. The influence of runner/gate design on the quality of an injected part was
also investigated by Tsai [13], who placed a rectangular flow restrictor within the tertiary
runner of a precision optical lens mold to achieve uniform melt temperature distribution in
the runner channel and reduce the thermal residual stress and warpage of injection-molded
parts. Shen et al. [14] investigated optimal gate design for thin-walled injection molding
and noticed that gate design affects the shear rate, which in turn increases the material’s
temperature. The higher temperature can reduce the viscosity of melted plastic so that
the melted plastic can fill into the cavity easily. Kim et al. [15] used numerical analysis
to investigate polymer flow patterns for different gate locations, and results showed that
wrong positioning of the gate prevented flow to the other side of the part and resulted in
short shots.
All presented research tells us that if molding process parameters can be adjusted
in an intelligent way, the quality and mechanical performance of the end-product can be
improved. Different from previously presented studies that analyze particular defects, this
paper presents a novel approach to the quality evaluation of the injected part. In this paper,
the Artificial Neural Network and Taguchi Techniques are used to find a set of optimal
process parameters that will result in a part with minimum possible short shot, shrinkage
rate, and warpage. The Analytic Hierarchy Process is used to calculate the weight of each
defect in the proposed thin-walled part design. The Taguchi method is used to find an
optimal set of five different geometric and process parameters in three different levels
that will result in the highest end-product quality. Finite Element Analysis (FEA) using
SolidWorks plastics is used to simulate injection-molding process experiments and validate
a proposed optimal set of process parameters.

2. Proposed Methodology
2.1. Problem Description
There are different internal and external defects in injection-molding technology that
evaluate the quality of injected parts, such as sink mark, shrinkage, warpage, weld line,
and short shot. In this paper, three common defects which reduce the quality of injected
parts were chosen: short shot possibility, shrinkage rate, and warpage. The possibility of
having the selected defects is related to different geometrical and process parameters.

2.2. Weight Calculation for the Selected Defects via AHP

The initial weight of each plastic defect was calculated via the Analytic Hierarchy
Process, as shown in Table 1. Short shot had the highest weight followed by warpage and
shrinkage, respectively.
Polymers 2021, 13, 4158 3 of 13

Table 1. Calculation of initial weights.

S1 S2 S3 Initial Weight
Step weight 0.5 0.2 0.3
Short shot 1 0.5
Shrinkage rate 1 0.2
warpage 1 0.3

2.3. Taguchi Orthogonal Array

The Taguchi method has been employed over the years to improve products and man-
ufacturing processes. It is a powerful and effective method to solve the quality problems
of products [16,17]. The objective of this paper is to combine the Taguchi method with
simulation tools, namely SolidWorks plastics, to reduce the percentage of different internal
and external defects in injection molding. Five different geometric and process parameters
in three different levels are selected, as shown in Table 2. Selecting the parameters is
based on the literature review, with a high percentage of contribution through the injection
process for the evaluation of the selected defects. Additionally, based on the number of
parameters and number of levels, an L18 orthogonal array is selected, as shown in Table 3.

Table 2. Geometric and process parameters in three levels.

Parameters Level 1 Level 2 Level 3

Gate design, A 1 2 -
Filling time, B (s) 0.2 0.6 1
Part cooling time, C (s) 3 3.9 5
Pressure holding time, D (s) 1 2 3
Melt temperature, E (◦ C) 200 230 280

Table 3. L18 orthogonal array.

Experiment A B C D E
1 1 1 1 1 1
2 1 1 2 2 2
3 1 1 3 3 3
4 1 2 1 1 2
5 1 2 2 2 3
6 1 2 3 3 1
7 1 3 1 2 1
8 1 3 2 3 2
9 1 3 3 1 3
10 2 1 1 3 3
11 2 1 2 1 1
12 2 1 3 2 2
13 2 2 1 2 3
14 2 2 2 3 1
15 2 2 3 1 2
16 2 3 1 3 2
17 2 3 2 1 3
18 2 3 3 2 1

3. Simulation
Two circular parts with 100 mm diameter and 1 mm thickness are designed using
SolidWorks, as shown in Figure 1. Sprue, runner, and gate have also been calculated and
designed with reference to the geometry, the dimension of the selected design, and the
selected material. To evaluate the selected defects in a critical condition, 1 mm thickness
and round shape parts are selected to avoid having any extraneous variables such as the
effect of corners or busses on the flow of molten plastic through the injection process. Since
 3. Simulation
3. Simulation
Two circular parts with 100 mm diameter and 1 mm thickness are designed using
Two circular
SolidWorks, parts with
as shown 100 mm
in Figure diameter
1. Sprue, andand
runner, 1 mmgatethickness
have alsoare designed
been using
calculated and
SolidWorks,
designed with as shown in Figure
reference to the1.geometry,
Sprue, runner, and gate have
the dimension of thealso been calculated
selected design, andand the
designed with reference to the geometry, the dimension of the selected
selected material. To evaluate the selected defects in a critical condition, 1 mm thickness design, and the
Polymers 2021, 13, 4158 4 of 13
selected material.
and round shapeTopartsevaluate the selected
are selected defects
to avoid in aany
having critical condition,
extraneous 1 mm thickness
variables such as the
and round
effect shape parts
of corners are selected
or busses to avoid
on the flow having
of molten any through
plastic extraneous variablesprocess.
the injection such as Since
the
effect of corners
the gate or busses
type leads on the
to short flow
shot andofshrinkage,
molten plastic
two through
differentthe injection
gates process.
have been Sinceas
selected
the
the gate
gatetype
shown type
in leads
leads
Figure toto
1, short
shortshot
namely shotand
round and shrinkage,
shrinkage,
gate two
twodifferent
and modified different
edge gategates
gates have
havebeen
[11]. beenselected
selectedasas
shownininFigure
shown Figure1,1,namely
namelyround roundgate
gateand
andmodified
modifiededgeedgegate
gate[11].
[11].

Figure 1. The 3D design of plastic part with sprue, runner and two types of gate system.
Figure 1. The 3D design of plastic part with sprue, runner and two types of gate system.
Figure 1. The 3D design of plastic part with sprue, runner and two types of gate system.
For the
For theflowflowanalysis,
analysis,SolidWorks
SolidWorksplastic plasticisisapplied,
applied,and andFinite
FiniteElement
ElementAnalysis
Analysis
(FEA)For the flow
(FEA)isisconducted analysis,
conductedwith withshellSolidWorks
shell(triangle)
(triangle) plastic
mesh
mesh is applied,
with
with element
element and Finite
thickness
thickness Element
of 1ofmm,
1 mm, Analysis
as shown
as shown in
(FEA) is 2.
in Figure
Figure conducted with shell
2. Polypropylene
Polypropylene (P.P.)(triangle) meshselected
(P.P.) material
material was with
waselement
for thethickness
selected for the of
analyses. 1 mm,
glassas
analyses.
The shown
The glass
transition
intemperature
Figure 2.temperature
transition Polypropylene of the (P.P.) material
◦(T iswas selected for the isanalyses. The glass4.
of the material (T gmaterial
) is 135 C,g)and 135
the°C, and the
viscosity viscosity
model model is
presented presented
in Table
transition
The meshtemperature
in Table The meshof
4.refinement the material
isrefinement
implemented (Twith
g) is 135 °C, with
elementand
is implemented sizethe
of viscosity
element
0.3 mm size model
for 0.3 is
ofsprue mm presented
and for sprue
runner
inand
Table
and 4.
runner The mesh
and 0.2 refinement
mm for the is implemented
gate. To avoid with
having element
any size
extraneous
0.2 mm for the gate. To avoid having any extraneous variable affecting the result for of 0.3 mm
variablefor sprue
affecting
andtherunner
the result for
selected and 0.2selected
the
plastic mm for plastic
defects, the gate.
one of the To
defects, avoid
input one having any
of the input
parameters, extraneous
needs tovariable
parameters,
which which
be affecting
needs
set through tothe
be
the result for the
set throughprocess,
simulation selected
the simulation plastic
is mold process, defects,
temperature. one
is mold of the input
temperature.
Hence, parameters,
the moldHence, which
the mold
temperature needs
◦ to beofis
is 50temperature
C as one
set50through
the °C as one
constants. theof simulation
themesh
The process,
constants.
details areis
The meshmold temperature.
details
tabulated in are Hence,
tabulated
Table 5 in the in the mold
Table
following: 5 intemperature
the following: is
50 °C as one of the constants. The mesh details are tabulated in Table 5 in the following:

Figure2.2.Finite
Figure Finiteelement
elementanalysis
analysisfor
for3D
3Dpart
partdesign.
design.
Figure 2. Finite element analysis for 3D part design.
Table 4. Modified cross-model for viscosity.

D1 (Pa) D2 (K) D3 A1 A2 (K) τ (Pa) n

4.44489 × 1014 263.15 0 32.7 52.6 26,260 0.272
Polymers 2021, 13, 4158 5 of 13

Table 5. FEA simulation and mesh parameters.

Triangle Total Node Total Mesh Size Mesh Size

Mesh Meshing
Size of Surface Elements of for Runner for Gate
Type Method
(mm) Mesh Surface Mesh (mm) (mm)
Shell Mesh Manual 2.76 5624 11,244 0.3 0.2

3.1. Experimental Setup

In this paper, polypropylene was chosen as the injected material for the injection
of two circular plates. Material characteristics are listed in Table 4. For the fabrication
process, computer numerical control (CNC) milling machine, grinding machine, and
drilling machine are used to produce the main components of mold tools, namely top
clamping plate, core and cavity plates, side plates, and bottom clamping plate. The selected
injection machine was the Poolad-Bch series with maximum inlet pressure of 100 MPa. The
details of the process parameters are presented below in Table 6.

Table 6. Process parameters.

Melt Max Melt Min Melt Mold Melt Flow Max Shear
Temperature Temperature Temperature Temperature Rate Stress
230 ◦ C 280 ◦ C 200 ◦ C 50 ◦ C 20 cm3 /10 min 250 kPa

A two-plate mold with two cavities and one parting line with runner, gate, and sprue
but without ejector system is chosen, and the selected material for the fabrication of core
and cavity is steel CK45 with surface hardness 56 HRC.

3.2. Simulation Results

The analysis of short shot possibility is implemented using SolidWorks plastic. Short
shots happen far from the gate locations or on thin wall products. They also happen
as a result of insufficient venting [18]. In analyzing the short shot possibility (the ratio
of simulated inlet pressure to maximum inlet pressure), different factors are taken into
consideration to diagnose short shot before it occurs [19]. The minimum level of short shot
possibility is related to experiment number 17, as shown in Figure 3a, and the maximum
short shot possibility is related to experiment number 1, as shown in Figure 3b. Any
increase in filling time, part cooling time, and melt temperature will decrease the short shot
possibility, as shown in experiment number 17. In contrast, in experiment number 1, filling
Polymers 2021, 13, x FOR PEER REVIEW6 of 14
time, part cooling time, and melt temperature are at their minimum levels, which leads to
a high level of short shot possibility.

Figure3.3.Injected
Figure Injectedparts
partswith
with(a)
(a)minimum
minimumpossibility
possibilityofofshort
shortshot
shotand
and(b)
(b)maximum
maximumpossibility
possibilityofof
shortshot.
short shot.

The second analysis was shrinkage analysis. The difference between the linear
dimensions of the cavity and the injected parts at room temperature will evaluate the
shrinkage rate [20]. Experiment number 14 represents the minimum shrinkage rate, and
experiment number 5 represents the highest shrinkage rate, as shown in Figure 4a,b,
respectively. With reference to the simulation result, it can be concluded that when the
Polymers 2021, 13, 4158 6 of 13
Figure3.3.Injected
Figure Injectedparts
partswith
with(a)
(a)minimum
minimumpossibility
possibilityofofshort
shortshot
shotand
and(b)
(b)maximum
maximumpossibility
possibilityofof
short shot.
short shot.

Thesecond
The
The secondanalysis
analysis
analysis was
was
was shrinkage
shrinkage
shrinkage analysis.
analysis.
analysis. TheTheThedifference
difference
difference between
between
between thedimen-
the
the linear linear
linear
dimensions
dimensions
sions of the ofof the cavity
the cavity
cavity and
and theand the injected
the injected
injected parts
parts at parts
room at at room temperature
room temperature
temperature will
will the
will evaluate evaluate
evaluate the
the
shrinkage
shrinkage
shrinkage rate [20]. Experiment
rate [20]. Experiment
rate [20]. Experiment number 14numbernumber 14 represents
14 represents
represents the minimum the minimum
the shrinkage
minimum rate, shrinkage
shrinkage rate, and
rate, and
and experiment
experiment
experiment
number number5the
number
5 represents 5represents
represents thehighest
the
highest shrinkage highest shrinkage
rate,shrinkage
as shown in rate,
rate, asasshown
Figure shown
4a,b, ininFigure
Figure4a,b,
respectively. 4a,b,
With
respectively.
reference With
to With
respectively. the referenceresult,
simulation
reference totothe
thesimulation
itsimulation result,ititcan
can be concluded
result, canbe
that beconcluded
whenconcluded
the melt that
that whenthe
temperature
when the
melt temperature
increases,
melt increases,
the shrinkage
temperature the
ratethe
increases, shrinkage
increases. rate increases.
shrinkage rate increases.

Figure 4. Injectedparts
Figure parts with(a)
(a) minimumshrinkage
shrinkage and(b)
(b) maximumshrinkage.
shrinkage.
Figure4.4.Injected
Injected partswith
with (a)minimum
minimum shrinkageand
and (b)maximum
maximum shrinkage.

Thelast
The
The lastdefect
last defectanalysis
defect analysisfor
analysis forquality
qualitypurposes
purposesisisis
purposes warpage,
warpage,
warpage, which
which
which refers
refers
refers tototo adistortion
distortion
a adistortion of
ofof
the the
the original
original
original design
design of of
design of the
thethe injected
injected parts
injected parts because
because
parts of different
of different
because shrinkage
shrinkage
of different rates
ratesrates
shrinkage in
in differentdifferent
parts
in different
parts
of
parts ofthetheinjected
theofinjected injected part
part [18].
part With[18].
[18]. Withreference
reference
With reference totothe
thesimulation
to the simulation simulation result,the
result, theresult,
minimum thewarpage
minimum
minimum is
warpage
related
warpage toisis relatedtotoexperiment
experiment
related 6,experiment 6,6,and
and the maximum andthe
the maximum
warpage
maximum warpage
is related
warpage isisrelated
relatedto9,
to experiment toexperiment
experiment
as shown in
9,asasshown
9,Figure shown
5a,b, in inFigure
Figure5a,b,
respectively. 5a,b, respectively.
Hence, any increase
respectively. Hence, anytemperature
in melt
Hence, any increaseininmelt
increase melt temperature
and temperature
filling and
time based
and
filling
on Tabletime3 based
will on
result Table
in an 3 will
increaseresult
in thein an increase
warpage in the
percentage. warpage
filling time based on Table 3 will result in an increase in the warpage percentage. percentage.

Figure5.5.
Figure
Figure 5.Injected
Injectedparts
Injected partswith
parts with(a)
with (a)minimum
(a) minimumwarpage
minimum warpageand
warpage and(b)
and (b)maximum
(b) maximumwarpage.
maximum warpage.
warpage.

Based on the L18 orthogonal array of Taguchi 18, experiments have been conducted
with different settings using SolidWorks plastics, and the defect values were tabulated, as
shown in Table 7. Maximum and minimum values for each defect type are highlighted
together with corresponding experiment numbers.

Table 7. Defects determination of 18 experiments based on SolidWorks plastics results.

Trial Number Short Shot Shrinkage Rate Warpage

1 75.57 16.45 2.43
2 64.72 18.24 1.73
3 52.65 21.08 1.54
4 55.34 18.24 2.64
5 44.8 21.08 1.94
6 64.57 16.43 1.08
7 62.4 16.43 1.3
Polymers 2021, 13, 4158 7 of 13

Table 7. Cont.

Trial Number Short Shot Shrinkage Rate Warpage

8 53.49 18.23 1.25
9 43.27 21.04 3.25
10 72.27 16.44 2.48
11 72.08 16.44 2.48
12 62.17 18.24 1.77
13 43.81 21.08 1.83
14 62.72 16.43 1.08
15 53.98 18.24 2.78
16 52.76 18.23 2.78
17 42.86 21.04 2.78
18 61.18 16.43 2.78

Weight calculation for the selected defects is implemented, as shown in Table 8, and
the sum of the defect’s value for individual experiments has been calculated. In a similar
way, normalized maximum and minimum values for each defect type are highlighted
together with corresponding experiment numbers.

Table 8. Normalized defect based on initial weight calculation.

Trial Number Short Shot Shrinkage Rate Warpage Sum

1 0.500 0.001 0.188 0.69
2 0.334 0.078 0.090 0.50
3 0.150 0.200 0.064 0.41
4 0.191 0.078 0.217 0.49
5 0.030 0.200 0.119 0.35
6 0.332 0.000 0.000 0.33
7 0.299 0.000 0.031 0.33
8 0.162 0.077 0.024 0.26
9 0.006 0.198 0.300 0.50
10 0.450 0.000 0.194 0.64
11 0.447 0.000 0.194 0.64
12 0.295 0.078 0.096 0.47
13 0.015 0.200 0.104 0.32
14 0.304 0.000 0.000 0.30
15 0.170 0.078 0.236 0.48
16 0.151 0.077 0.236 0.46
17 0.000 0.198 0.236 0.43
18 0.280 0.000 0.236 0.52

Since the objective of this study is to minimize different defects in injection molding,
the smaller-the better-quality characteristic has been chosen, which is defined by Equations
(1) and (2) [21]:
S
= −10 log( MSD ) (1)
N
n
1
MSD =
N ∑ y2i (2)
i =1

where yi is the total value of the selected defects for different experiments and N is the total
number of data points. Signal-to-noise ratio calculation has been conducted and tabulated
in Table 9. The next step was to determine the response table of Taguchi to find the most
significant parameters from the selected parameters and their optimum levels.
With reference to Table 10, the optimum level is the highest value of each parameter.
Hence, the best combination is gate type at level 1, filling time at level 2, cooling time at
level 2, pressure-holding time at level 3, and melt temperature at level 3.
Polymers 2021, 13, 4158 8 of 13

Table 9. Signal-to-noise ratio for the smaller-the-better quality characteristics.

Experiment A B C D E S/N
1 1 1 1 1 1 3.24
2 1 1 2 2 2 5.98
3 1 1 3 3 3 7.67
4 1 2 1 1 2 6.28
5 1 2 2 2 3 9.14
6 1 2 3 3 1 9.58
7 1 3 1 2 1 9.65
8 1 3 2 3 2 11.58
9 1 3 3 1 3 5.94
10 2 1 1 3 3 3.82
11 2 1 2 1 1 3.86
12 2 1 3 2 2 6.58
13 2 2 1 2 3 9.93
14 2 2 2 3 1 10.35
15 2 2 3 1 2 6.30
16 2 3 1 3 2 6.65
17 2 3 2 1 3 7.24
18 2 3 3 2 1 5.74

Table 10. Response table of Taguchi.

Cooling Pressure Melt

Level Gate Type Filling Time
Time Holding Time Temperature
L1 7.67 5.19 6.60 5.48 7.07
L2 6.72 8.60 8.03 7.84 7.23
L3 NA 7.80 6.97 8.28 7.29
Difference 0.95 3.41 1.43 2.80 0.22

The next step was to run the simulation based on the optimum level to evaluate the
individual defect values and the sum of the selected defects. Based on the simulation
Polymers 2021, 13, x FOR PEER REVIEW9 of 14
results as shown in Figure 6, the optimum defects values are tabulated in Table 11, as
shown in the following:

Figure
Figure6.6.Defect
Defectvalues based
values on the
based on optimum level level
the optimum (a) short shot, (b)
(a) short shrinkage,
shot, (c) warpage.
(b) shrinkage, (c) warpage.

With reference to Table 11, running the simulation with optimum level of the
minimum defects rate shows that the used optimization tool gives good results for the
injection-molding process. The proposed methodology was experimentally validated by
Polymers 2021, 13, 4158 9 of 13

Table 11. Simulation result based on the optimum level of the selected parameters from Taguchi.

Short Shot Volume Shrinkage Warpage Sum

Weight 0.5 0.2 0.3
Defects 45.03 20.94 1.26
Normalized 0.033 0.192 0.025 0.2499

With reference to Table 11, running the simulation with optimum level of the mini-
mum defects rate shows that the used optimization tool gives good results for the injection-
molding process. The proposed methodology was experimentally validated by Moayye-
dian [21].
The final step was to apply analysis of variance to determine the percentage of con-
tribution for individual parameters. The percentage of contribution can be calculated as
follows [17]:
1. Degree of freedom: The total degree of freedom (d f T ), the degree of freedom of
factor A (d f A ), and the degree of freedom for error variance (d f E ) are as follows:

d f T = ( N − 1) (3)

d f A = ( K A − 1) (4)
 
d f E = d f T − ∑ d f f actor (5)

where N is the total number of experiments.

2. Sum of squares: The sum of the square of factor A (SS A ), the total sum of square
(SST ) and the sum of the square for error variance (SSE ) are calculated as follows:
 2
∑iN=1 xi
!
KA
A2i
SS A = ∑ n Ai
−
N
(6)
i =1

 2
N ∑iN=1 xi
SST = ∑ xi 2 − N
(7)
i =1
 
SSE = SST − ∑ SS f actor (8)

where xi is a value at level (1, 2, . . . N), n Ai is the number of levels and Ai is a value at level
i of factor A.
3. Percentage contribution: the percentage contribution of factor A is calculated using
the following Equation:

SS A
PC A = × 100% (9)
SST
The percentage of contribution for the selected factors is tabulated in Table 12 in the
following:

Table 12. Analysis of variance.

Factor f SS PC (%)
A 1 0.009 3.59
B 2 0.109 42.8
C 2 0.016 6.15
D 2 0.069 26.66
E 2 0.002 0.89
pool error 8 0.05 20.6
Total 17 0.26 100
Polymers 2021, 13, 4158 10 of 13

By determining the optimum level and the significant parameters reducing the total
defects value, the next step was to evaluate the percentage of contribution, based on
Equations 3 to 9. The percentage of contribution for individual parameters can be achieved
by employing an ANOVA. The largest value of contribution indicates the most significant
factor affecting the system’s performance. It can be concluded that the filling time has the
highest percentage of contribution (42.8%), followed by pressure-holding time (26.6%).

4. Artificial Neural Network Model

Artificial Neural Network (ANN) is a modeling tool that has a particular ability to
Polymers 2021, 13, x FOR PEER REVIEW11 of 14
learn and generate functions from training series. ANNs establish the relationships between
inputs and outputs using particular transfer functions. In a series of training operations,
these are used to alter the values of the biases and weights. ANNs are made up of neurons,
up of are
which neurons, which processors.
small, linked are small, linked Weightedprocessors.
linkages Weighted
connect the linkages
neurons, connect
allowing the
neurons, allowing messages to travel between them. Each
messages to travel between them. Each neuron receives many inputs according to their neuron receives many inputs
according to
connection their connection
weights from otherweights neuronsfrom other neurons
and creates a singleand creates
output thatamay
single output that
propagate to
may propagate
several additionalto several[22,23].
neurons additional neurons [22,23].learning
The backpropagation The backpropagation
algorithm has been learning
the
algorithm
most widelyhas been the
employed most widely
approach employed
in engineering approach among
applications in engineering
the numerousapplications
types
ofamong
ANNs that the exist.
numerous types of ANNs that
The Levenberg–Marquardt exist. The training
backpropagation Levenberg–Marquardt
algorithm was
backpropagation
selected training algorithm was selected for this study.
for this study.
Two phases are
Two phases are involved involvedininthe theANN
ANNmodeling
modelingprocess.
process.TheThefirst
firstphase
phaseisistototrain
trainthe
the
network,and
network, andthe the second
second is toistest
to test it using
it using datawere
data that thatnot
were not utilized
utilized in the process.
in the training training
Itprocess. It is
is critical critical
that that the network
the network obtains all obtains
of theallinformation
of the information
it needsitto needs
learntoinlearn in the
the form
ofform of set.
a data a dataWhenset.theWhennetworkthe network
reads each reads each it
pattern, pattern,
utilizesitthe
utilizes
input thedatainput data to
to generate
an output, an
generate which is then
output, compared
which is thento the trainingtopattern.
compared If therepattern.
the training is a discrepancy,
If there the is a
connection
discrepancy, weights are adjusted
the connection in a way
weights that reduces
are adjusted in a the
wayerror.
that If the error
reduces theiserror.
still more
If the
than
errortheis maximum
still more acceptable
than the maximumtolerance after the network
acceptable has gone
tolerance afterthrough all of the
the network hasinput
gone
patterns,
through the all ofANNthe goes
inputthrough
patterns,allthe of ANN
the input
goespatterns
throughagainall ofuntil all of the
the input errorsagain
patterns are
within theofnecessary
until all the errorstolerance
are within [24].
the necessary tolerance [24].
ANN
ANNwas wasemployed
employedfor forthetheintended
intendedrange
rangeofoffourfourinputs
inputs(filling
(fillingtime,
time,part-cooling
part-cooling
time,
time,pressure-holding
pressure-holdingtime, time, and andmelt temp).
melt Figure
temp). 7 shows
Figure this model
7 shows with thewith
this model appro-
the
priate neuron numbers in the hidden and output layers.
appropriate neuron numbers in the hidden and output layers. A normalized and A normalized and weighted
output
weighted wasoutput
used,waswhich represents
used, the associated
which represents values of
the associated threeofoutputs
values (short (short
three outputs shot,
shrinkage rate, warpage) as given in Table 8 earlier. The backpropagation
shot, shrinkage rate, warpage) as given in Table 8 earlier. The backpropagation (BP) (BP) training
technique was used to
training technique was create
usedthis neuralthis
to create network
neural unit because
network it has
unit the capacity
because it has theto forecast
capacity
values in between
to forecast values learning
in between values and make
learning valuesinterpolations between learning-curves
and make interpolations between learning- data.
This was accomplished using the appropriate amount of network
curves data. This was accomplished using the appropriate amount of network layers and layers and neurons at
minimum error.
neurons at minimum error.

Figure7.7.Neural
Figure Neuralnetwork.
network.

Afterobtaining
After obtaining the
the ANN
ANN model,
model, itit was
was coupled
coupledwithwithananoptimization
optimizationalgorithm
algorithmin
the
in thecomputer
computercode.
code.The
Theobjective
objectivewaswas to the minimum
to obtain the minimum “output”
“output”value.value.After
After
convergenceofofoptimization
convergence optimization with
with thethe
ANNANN model,
model, the best
the best output
output valuevalue was obtained
was obtained with
with
the the following
following parameters:
parameters: fillingattime
filling time levelat
3, level 3, cooling
cooling time 1,
time at level atpressure-holding
level 1, pressure-
holding time at level 3, and melt temperature at level 2. The result calculated
time at level 3, and melt temperature at level 2. The result calculated by ANN optimization by ANN
optimization
was was to
0.2542. In order 0.2542.
validateIn the
order to validate
“predicted the scenario”,
best case “predictedanbest case scenario”,
additional SolidWorks an
additionalwas
simulation SolidWorks
performed simulation
to check thewasresult
performed
for thisto check the
particular result
case, for thisinparticular
as shown Figure 8.
case, as shown
SolidWorks in Figure
simulation gave 8. an
SolidWorks
output of simulation gave an
0.2461, as shown output
in Table 13,ofwhich
0.2461,
is as shown
close to thein
Table 13, which is close to the ANN prediction and better than all trials in Table 8. In other
words, SolidWorks simulation validated that the parameter configuration proposed by
the ANN results in the best output compared to all previous trials.
Polymers 2021, 13, 4158 11 of 13

ANN prediction and better than all trials in Table 8. In other words, SolidWorks simulation
Polymers 2021, 13, x FOR PEER REVIEW12 of 14
validated
that the parameter configuration proposed by the ANN results in the best output
compared to all previous trials.

Figure 8. Defects value based on the optimum level (a) short shot, (b) shrinkage, (c) warpage.
Figure 8. Defects value based on the optimum level (a) short shot, (b) shrinkage, (c) warpage.

Table 13. Simulation result based on the optimum level of the selected parameters from ANN.
Table 13. Simulation result based on the optimum level of the selected parameters from ANN.
Short Shot Volume Shrinkage Warpage Sum
Short shot Volume Shrinkage Warpage Sum
Weight 0.5 0.2 0.3
Weight 0.5 0.2 0.3
Defects 54.66 17.96 1.01
Defects
Normalized 0.1803 54.66 0.0658 17.96 0.0 1.01 0.2461
Normalized 0.1803 0.0658 0.0 0.2461
The
Thecomparison
comparisonofofpredictions
predictionsby
bythe
theTaguchi
TaguchiMethod
Methodand andANN
ANNmodel
modelwas
wasmade
made
based
basedononthetheSolidWorks
SolidWorkssimulation
simulationresults
resultsof
of both
both cases.
cases. The
Thesimulation
simulationresult
resultfor
forthe
the
parameters predicted with the Taguchi method given in Table 11 was 0.2499,
parameters predicted with the Taguchi method given in Table 11 was 0.2499, and the and the
simulation
simulationresult
resultfor
forparameters
parameterspredicted
predictedwith
withthe
theANN
ANNmethod
methodgiven
givenininTable
Table13
13was
was
0.2461.
0.2461. It shows that the Taguchi Method and ANN model predictions were successfulin
It shows that the Taguchi Method and ANN model predictions were successful in
estimating
estimatingthethebest
bestcases
cases(minimum
(minimumoutput)
output)and
andwere
wereclose
closeininvalue.
value.The
Themargin
marginofoferror
error
was
wascalculated,
calculated,asasshown
shownin inEquation
Equation(10):
(10):

Taguchi method − ANN


Margin of error % = Taguchi⁡method−ANN
( × 100 (10)
(10)
Margin⁡of⁡error⁡%⁡ = ( ANN
ANN ) × 100
With reference to Equation (10), the margin of error for the Taguchi method and ANN
is equal toreference
With to Equation
1.5%, which (10),acceptable
is within the the margin of error
range for the Taguchi
in engineering fields. method and
ANN is equal to 1.5%, which is within the acceptable range in engineering fields.
5. Conclusions
5. Conclusions
The combination of simulation with DOE was a useful approach to find the significant
parameters that lead toof
The combination short shot, warpage,
simulation and shrinkage
with DOE of the approach
was a useful injected part. Different
to find the
processes and geometric parameters were selected for the proposed solution. Based
significant parameters that lead to short shot, warpage, and shrinkage of the injected part. on
the selected orthogonal array of Taguchi, 18 experiments were conducted via
Different processes and geometric parameters were selected for the proposed solution. SolidWorks
Based on the selected orthogonal array of Taguchi, 18 experiments were conducted via
SolidWorks plastics and the finite element method (FEM) to determine the optimum level
of the selected parameters to minimize different internal and external defects. The signal-
to-noise ratio (S/N) was an effective tool to determine the optimum level of each
Polymers 2021, 13, 4158 12 of 13

plastics and the finite element method (FEM) to determine the optimum level of the selected
parameters to minimize different internal and external defects. The signal-to-noise ratio
(S/N) was an effective tool to determine the optimum level of each parameter, and an
ANOVA was used for determining the percentage of contribution. It can be concluded
that filling time had the highest percentage of contribution (42.8%), followed by pressure-
holding time (26.6%).
The ANN was applied to determine the optimum levels of different parameters to
minimize the selected defects. The normalized output value based on the ANN model
and FEM simulations was 0.2542 and 0.2461, respectively, which were very close to each
other. The normalized simulation output values of the ANN model and Taguchi method
were 0.2461 and 0.2499, respectively. The margin of error percentage between the ANN
model and Taguchi method was equal to 1.5%, which demonstrated the robustness of the
proposed method and the compatibility of the selected tools. It can be concluded that the
predicted model with minimum defects had been selected, which was the ANN model. The
selected optimum model was to have filling time at 1 s, cooling time at 3 s, pressure-holding
time at 3 s, and melt temperature at 230 ◦ C. The optimum level of the selected parameters
based on the ANN model was very realistic, resulting in the lower temperature to avoid
having any other defects related to high temperature and lower part-cooling time to reduce
the injection time. Further research in this direction will provide more comprehensive
guidelines for designers by considering other processes and geometric parameters which
increase different defect rates in injection molding.

Author Contributions: Conceptualization, A.M. and M.M.; methodology, M.M. and A.D.; software,
M.M.; validation, A.D.; formal analysis, A.D. and M.M.; investigation, A.M.; writing—original
draft preparation, M.M., A.D. and A.M.; writing—review and editing, A.M.; visualization, A.M. All
authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Shi, H.; Xie, S.; Wang, X. A warpage optimization method for injection molding using artificial neural network with parametric
sampling evaluation strategy. Int. J. Adv. Manuf. Technol. 2013, 65, 343–353. [CrossRef]
2. Kurtaran, H.; Erzurumlu, T. Efficient warpage optimization of thin shell plastic parts using response surface methodology and
genetic algorithm. Int. J. Adv. Manuf. Technol. 2006, 27, 468–472. [CrossRef]
3. Gao, Y.; Turng, L.-S.; Wang, X. Adaptive geometry and process optimization for injection molding using the kriging surrogate
model trained by numerical simulation. Adv. Polym. Technol. 2008, 27, 1–16. [CrossRef]
4. Lee, B.H.; Kim, B.H. Optimization of Part Wall Thicknesses to Reduce Warpage of Injection-Molded Parts Based on The Modified
Complex Method. Polym. Technol. Eng. 1995, 34, 793–811. [CrossRef]
5. Yin, F.; Mao, H.; Hua, L.; Guo, W.; Shu, M. Back Propagation neural network modeling for warpage prediction and optimization
of plastic products during injection molding. Mater. Des. 2011, 32, 1844–1850. [CrossRef]
6. Hassan, H.; Regnier, N.; Pujos, C.; Arquis, E.; Defaye, G. Modeling the effect of cooling system on the shrinkage and temperature
of the polymer by injection molding. Appl. Therm. Eng. 2010, 30, 1547–1557. [CrossRef]
7. Lotti, C.; Ueki, M.M.; Bretas, R.E.S. Prediction of the shrinkage of injection molded iPP plaques using artificial neural networks. J.
Inject. Molding Technol. 2002, 6, 157–176.
8. Tang, L.Q.; Chassapis, C.; Manoochehri, S. Optimal cooling system design for multi-cavity injection molding. Finite Elements Anal.
Des. 1997, 26, 229–251. [CrossRef]
9. Hassan, H.; Regnier, N.; Lebot, C.; Pujos, C.; Defaye, G. Effect of cooling system on the polymer temperature and solidification
during injection molding. Appl. Therm. Eng. 2009, 29, 1786–1791. [CrossRef]
10. Moayyedian, M.; Abhary, K.; Marian, R. The analysis of short shot possibility in injection molding process. Int. J. Adv. Manuf.
Technol. 2017, 91, 3977–3989. [CrossRef]
Polymers 2021, 13, 4158 13 of 13

11. Moayyedian, M.; Abhary, K.; Marian, R. Gate design and filling process analysis of the cavity in injection molding process. Adv.
Manuf. 2016, 4, 123–133. [CrossRef]
12. Moayyedian, M.; Abhary, K.; Marian, R. Optimization of injection molding process based on fuzzy quality evaluation and Taguchi
experimental design. CIRP J. Manuf. Sci. Technol. 2018, 21, 150–160. [CrossRef]
13. Tsai, K.-M. Runner design to improve quality of plastic optical lens. Int. J. Adv. Manuf. Technol. 2013, 66, 523–536. [CrossRef]
14. Shen, Y.-K.; Wu, C.-W.; Yu, Y.-F.; Chung, H.-W. Analysis for optimal gate design of thin-walled injection molding. Int. Commun.
Heat Mass Transf. 2008, 35, 728–734. [CrossRef]
15. Kim, H.; Son, J.; Im, Y. Gate location design in injection molding of an automobile junction box with integral hinges. J. Mater.
Process. Technol. 2003, 140, 110–115. [CrossRef]
16. Oktem, H.; Erzurumlu, T.; Uzman, I. Application of Taguchi optimization technique in determining plastic injection molding
process parameters for a thin-shell part. Mater. Des. 2007, 28, 1271–1278. [CrossRef]
17. Yang, K.; El-Haik, B.S. Design for Six Sigma: A Roadmap for Product Development, 2nd ed.; McGraw-Hill Companies: New York, NY,
USA, 2009.
18. Goodship, V. Troubleshooting Injection Moulding; Smithers Rapra Press: Shropshire, UK, 2004; Volume 15.
19. Moayyedian, M.; Abhary, K.; Marian, R. The Analysis of Defects Prediction in Injection Molding. Int. J. Mech. Aerosp. Ind.
Mechatron. Manuf. Eng. 2016, 10, 1863–1866. [CrossRef]
20. Fisher, J.M. Handbook of Molded Part Shrinkage and Warpage, Plastics Design Library, 2nd ed.; William Andrew: Norwich, NY, USA,
2003.
21. Moayyedian, M. Intelligent Optimization of Mold Design and Process Parameters in Injection Molding; Springer: Singapore, 2019.
22. Chow, T.; Zhang, G.; Lin, Z.; Song, C. Global optimization of absorption chiller system by genetic algorithm and neural network.
Energy Build. 2002, 34, 103–109. [CrossRef]
23. Abidoye, L.; Das, D.B. Artificial neural network modeling of scale-dependent dynamic capillary pressure effects in two-phase
flow in porous media. J. Hydroinformatics 2014, 17, 446–461. [CrossRef]
24. Kalogirou, S. Applications of artificial neural-networks for energy systems. Appl. Energy 2000, 67, 17–35. [CrossRef]