TRƯỜNG ĐẠI HỌC CÔNG NGHIỆP HÀ NỘI HANOI UNIVERSITY OF INDUSTRY Tập san SINH VIÊN NGHIÊN CỨU KHOA HỌC Số 14 ● 2024 227
STUDY TO OPTIMIZE THE INJECTION MOLDING PAREMETERS TO
REDUCE SHRINKAGE OF PLASTIC PRODUCTS USING TAGUCHI METHOD
NGHIÊN CỨU TỐI ƯU CÁC THÔNG SỐ ÉP PHUN NHẰM GIẢM THIỂU CO NGÓT
SẢN PHẨM NHỰA BẰNG PHƯƠNG PHÁP TAGUCHI
Lê Văn Dũng1,*, Nguyễn Mạnh Hùng1,
Nguyễn Trọng Hoàng1, Nguyễn Quang Huy1, Đào Ngọc Hoành2
1Lớp KTKM 01 - K15, Trường Cơ khí - Ô tô, Trường Đại học Công nghiệp Hà Nội
2Trường Cơ khí - Ô tô, Trường Đại học Công nghiệp Hà Nội
*Email: manhdung2311ch@gmail.com
ABSTRACT
This study focuses on optimizing parameters in the plastic injection molding process, aiming to minimize the
occurrence of shrinkage in the product. By employing the Taguchi method and an L27 orthogonal array, the study focuses
on altering and analyzing the impacts of five crucial parameters: cooling temperature, melt temperature, cooling time,
packing time, and packing pressure. The entire injection molding process is simulated using Moldflow Plastic Insight
software. Subsequently, Minitab software is utilized to synthesize and analyze the collected data. Analysis of variance
(ANOVA) results are also employed to evaluate the extent of influence of each parameter on the shrinkage phenomenon
of the product. The results indicate that holding packing time (accounting for 59.02%) and melt temperature (accounting
for 20.2%) are the two factors with the greatest impact on product shrinkage, providing directions for the company to
enhance its product quality.
Keywords: Injection molding, Taguchi method, ANOVA, Shrinkage, Optimization.
TÓM TẮT
Nghiên cứu này tập trung vào việc tối ưu hóa các thông số trong quy trình ép phun nhựa nhằm giảm thiểu sự co ngót
của sản phẩm. Bằng cách sử dụng phương pháp Taguchi và mảng trực giao L27, nghiên cứu này tập trung vào việc thay
đổi và phân tích tác động của năm thông số quan trọng: nhiệt độ làm mát, nhiệt độ nóng chảy, thời gian làm mát, thời gian
nén và áp suất nén. Toàn bộ quá trình ép phun được mô phỏng bằng phần mềm Moldflow Plastic Insight. Sau đó, phần
mềm Minitab được sử dụng để tổng hợp và phân tích dữ liệu thu thập được. Kết quả phân tích phương sai (ANOVA) cũng
được sử dụng để đánh giá mức độ ảnh hưởng của từng thông số đến hiện tượng co ngót của sản phẩm. Kết quả cho thấy
thời gian nén (chiếm 59,02%) và nhiệt độ nóng chảy (chiếm 20,2%) là hai yếu tố có tác động lớn nhất đến co ngót sản
phẩm, cung cấp định hướng cho công ty để nâng cao chất lượng sản phẩm.
Từ khóa: Ép phun, phương pháp Taguchi, ANOVA, co ngót, tối ưu hóa.
1. INTRODUCTION
In the mold manufacturing and plastic injection molding
industry, the injection molding process plays a crucial role
in producing copious quantities of products with high
precision and uniformity [1-3]. One of the biggest
challenges in the manufacturing process is managing and
controlling shrinkage in plastic products, which affects their
precise dimensions, aesthetic appeal, and final appearance.
This not only impacts the overall quality of the products but
also causes significant setbacks in terms of time and
finances for manufacturers, affecting their reputation [4-5].
By applying the Taguchi method, this study aims to
evaluate the influence of critical injection emolding
parameters such as cooling temperature, melt temperature,
cooling time, packing time, and packing pressure to control
and reduce the effects of shrinkage during the
manufacturing process [6-7]. The Taguchi method is an
effective approach for studying the simultaneous effects of
multiple variables and conserving the number of
experiments needed to determine optimal levels for each
factor under consideration [8-10]. Additionally, in this
study, we combine the use of Moldflow Plastic Insight
software to simulate the injection molding process with
input parameters from Taguchi method. The strength of this
TRƯỜNG ĐẠI HỌC CÔNG NGHIỆP HÀ NỘI HANOI UNIVERSITY OF INDUSTRY Tập san SINH VIÊN NGHIÊN CỨU KHOA HỌC Số 14 ● 2024 228method is further enhanced when combined with analysis
of variance (ANOVA) to identify the individual effects of
each parameter and their respective impacts on shrinkage
errors [11, 12]. Signal-to-noise ratio (S/N) analysis is also
used as a method to identify the optimal injection molding
processing parameters to minimize shrinkage [13, 14]. The
results of the analysis and optimization of these parameters
can help the company mitigate this drawback, instead of
relying on traditional trial-and-error methods that often
consume considerable time and resources [15, 16].
Moreover, there have been several successful studies
applying the Taguchi method to optimize parameters to
reduce final product defects [17-19].
2. STUDY SET-UP
2.1. Analysis model
Fig. 1. 3D Model for Analysis
Fig. 2. 2D Drawing of the Model
In this study, the 3D model of the product used for
research was designed using NX software with a shape like
Fig 1. The product has a diameter of 30mm, a height of
33mm, and a thickness of 1.3mm. Additionally, the model
includes a sprue, runner, and gate as shown in Figure 1. The
dimensions of these features are presented in the 2D
drawing in Figure2. After completing the design, the model
is imported into Autodesk Moldflow Insight for analysis.
The meshing process is conducted using a triangular mesh
with a size of 0.87mm, resulting in 183,436 triangular
elements and 91,895 nodes. Following meshing, the plastic
flow channel system is displayed in Fig 3(a), and the colling
system is displayed in Fig 3(b) with a cooling channel
diameter of 8mm.
(a)
(b)
Fig. 3. Post-meshing runner system (a); post-meshing cooling
system (b)
2.2. Material and injection molding machine
specifications
The material used in this study is Polypropylene (PP),
a semi-crystalline thermoplastic polymer known for its
outstanding properties such as high rigidity, heat resistance,
and chemical resistance. It is also recognized for its
lightweight, excellent impact resistance, good tensile
strength, water resistance, and inertness to most solvents at
room temperature. These properties make PP an ideal
TRƯỜNG ĐẠI HỌC CÔNG NGHIỆP HÀ NỘI HANOI UNIVERSITY OF INDUSTRY Tập san SINH VIÊN NGHIÊN CỨU KHOA HỌC Số 14 ● 2024 229choice for various applications ranging from packaging,
household items, to automotive and medical industries.
Detailed specifications of the material are presented in
Table 1.
Table 1. Physical properties of PP plastic
Properties Unit Value
Glass Transition Temperature °C 122
Thermal Conductivity W/(m-°C) 0.274
Specific Heat Capacity J/(kg-°C) 2047
Elastic Modulus MPa 2387
Poisson’s ration 0.375
Shear Modulus MPa 907
Melt Density g/cm3 1.2309
Solid Density g/cm3 1.3717
The Sumitomo SE100EV-C250 plastic injection
molding machine was used in this study.
Fig. 4. The Sumitomo SE100EV-C250 plastic injection molding
machine
2.3. The Taguchi method
This study utilizes the Taguchi method to pinpoint the
optimal injection molding parameters [20]. The input
parameters used in this study are cooling temperature, melt
temperature, colling time, packing time, and packing
pressure, with each set at three levels. Firstly, the cooling
temperature, denoted as “A”, is set at three levels: 10, 25
and 45 degrees Celsius. Secondly, the melt temperature,
denoted as “B”, is set at three levels: 215, 225 and 235
degrees Celsius. Thirdly, the cooling time, denoted as “C”,
is set at three levels: 10, 20 and 30 seconds. Fourthly, the
holding time, denoted as “D”, is set at three levels: 3, 4 and
5 seconds. Finally, the packing pressure, also known as the
compression pressure, labeled as “E” set at three levels: 60,
70 and 80MPa.
Table 2. Input parameters and levels used in the study
Injection molding
parameter
Level
1
Level
2
Level
3
A: Cooling temperature(
℃
) 10 25 40
B: Melt tempurature (
℃
) 215 225 235
C: Cooling time (s) 10 20 30
D: Packing time (s) 3 4 5
E: Packing pressure (MPa) 60 70 80
The experiments are conducted based on the Orthogonal
L27 array in Minitab software, with five input parameters,
each having three levels. In this model, the “smaller is
better” criterion is applied to the signal-to-noise ratio (S/N)
to minimize shrinkage. The S/N value is calculated using
equation, where “y” represents the results of the
experiments (in this case, the volume of shrinkage), and “n”
is the number pf trials in an individual experiment (in this
study, it is 3).
S/N =−10log (∑()/) (1)
Analysis of variance (ANOVA) is then used to
determine the extent of influence of these parameters on the
process. This helps identify which parameters have the
greatest impact on the outcome, thereby allowing for more
effective optimization of the process.
3. RESULTS AND ANALYSIS
After using the Minitab software and based on the input
parameters designed in Table 3, the research conducted 27
entirely different simulation cases to serve optimization
purposes. All obtained results were simulated on Moldflow
Plastic Insight software and presented in Table 3. The
results of the shrinkage volume and S/N of the 27
simulation runs varied, providing the input variables
influenced the outcome.
Table 3. Orthogonal array L27 (3) with input parameters,
results, and Signal-to-Noise (S/N) ratios
Run
A B C D E Shrinkage
(%) S/N
1 10 215 10 3 60 7.828 -17.8730
2 10 215 10 3 70 7.834 -17.8797
3 10 215 10 3 80 7.639 -17.6607
4 10 225 20 4 60 7.571 -17.5831
5 10 225 20 4 70 7.078 -16.9982
6 10 225 20 4 80 6.648 -16.5438
7 10 235 30 5 60 7.795 -17.8363
8 10 235 30 5 70 7.472 -17.4687
9 10 235 30 5 80 6.975 -16.8709
10 25 215 20 5 60 7.023 -16.9305
11 25 215 20 5 70 6.649 -16.4551
12 25 215 20 5 80 6.128 -15.7464
13 25 225 30 3 60 8.605 -18.6950
14 25 225 30 3 70 8.49 -18.5782
15 25 225 30 3 80 8.46 -18.5474
16 25 235 10 4 60 7.511 -17.5140
17 25 235 10 4 70 6.925 -16.8084
18 25 235 10 4 80 6.568 -16.3487
19 40 215 30 4 60 6.681 -16.4968
20 40 215 30 4 70 5.987 -15.5442
TRƯỜNG ĐẠI HỌC CÔNG NGHIỆP HÀ NỘI HANOI UNIVERSITY OF INDUSTRY Tập san SINH VIÊN NGHIÊN CỨU KHOA HỌC Số 14 ● 2024 23021 40 215 30 4 80 5.238 -14.3833
22 40 225 10 5 60 7.365 -17.3435
23 40 225 10 5 70 6.718 -16.5448
24 40 225 10 5 80 5.916 -15.4406
25 40 235 20 3 60 9.579 -19.6255
26 40 235 20 3 70 9.503 -19.5572
27 40 235 20 3 80 9.439 -19.4985
3.1. Signal-to-Noise (S/N) ratios analysis
Table 4. Response table for S/N ratios of shrinkage
Level Injection molding parameter
A B C D E
1 -17.40 -16.55 -17.05 -18.66 -17.77
2 -17.29 -17.35 -17.65 -16.46 -17.31
3 -17.16 -17.95 -17.16 -16.74 -16.77
Delta 0.24 1.40 0.60 2.20 0.99
Rank 5 2 4 1 3
Table 4 presents the results of analyzing the impact of
input parameters (A, B, C, D) on the shrinkage volume,
based on the S/N values. The ranking in this table is
determined based on the Delta value. Delta is the difference
between the largest and smallest S/N values for each input
parameter. This difference represents the level of the
shrinkage volume variation when changing the parameters
from one level to another. More specifically, the higher the
delta value, the greater the impact on the shrinkage volume
and vice versa. Therefore, as per Table 6, parameter D with
a Delta of 2.20 has the greatest impact on the shrinkage
volume (ranked 1), while parameter A with a Delta of 0.24
has the least impact (ranked 5). Parameters B, C, and E,
with corresponding Delta values of 1.40, 0.60, and 0.99,
rank 2, 4, and 3 respectively.
Fig. 5. Main effects plot for S/N ratios of shrinkage
This table is also used to plot the S/N analysis chart
displayed in Fig. 5. It shows that four variables, namely
plastic flow temperature, cooling time, holding pressure
time, and packing pressure, have a considerable influence
on the S/N value. In the chart, the x-axis, which is the
optimal level of the combined variables. Here, they are A3
(cooling temperature 40 ℃), B1 (melt temperature 215 ℃),
C2 (cooling time 20s), D2 (packing time 4s), and E4
(packing pressure 80MPa).
3.2. Simulation verification
After applying the optimized parameter set obtained
from the S/N chart, a simulation was conducted using
Moldflow Plastic Insight software, resulting in the outcome
shown in Fig. 6.
(a)
(b)
Fig. 6. Simulation results of inner surface (a) and outer surface
(b) of the component
3.3. Analysis of variance (ANOVA)
The results of the ANOVA analysis are shown in Table
5, which are used to evaluate the significance of the five
input factors: cooling temperature (A), melt temperature
(B), cooling time (C), packing time (D), and packing
pressure (E) on the shrinkage volume. The table presents
degrees of the freedom, total squares, mean squares,
F-value, P-value, and the percentage of the influence.
Usually, the selection of the factor is based on the P-value.
If P < 0.05, it can be concluded the results are not due to
randomness, and there is over a 95% chance that it is due to
the factor being evaluated. Thus, according to Table 9,
factors B, C, D, E significantly affect the shrinkage volume
while factor A does not. The percentage contributions on the
321
-16.5
-17.0
-17.5
-18.0
-18.5
321 321 321 321
A
Mean of SN ratios
B
C
D
E
Main Effects Plot for SN ratios
Data Means
Signal-to-noise: Smaller is better
TRƯỜNG ĐẠI HỌC CÔNG NGHIỆP HÀ NỘI HANOI UNIVERSITY OF INDUSTRY Tập san SINH VIÊN NGHIÊN CỨU KHOA HỌC Số 14 ● 2024 231factors in this research are as follows: packing time
contributes the highest proportion compared to other
parameters, about 59.02%. Melt temperature contributes
the second highest, with around 20.2%. Packing pressure
contributes about 10.21%, cooling time contributes about
4.25% and finally, cooling teamperature contributes only
about 0.61%.
Table 5. The ANOVA results
Source
DF
Seq SS
Adj
MS
F-
Value
P-
Value
Contributio
n (%)
A 2 0.2669
0.1334
0.395
0.95 0.61
B 2 8.8280
4.4140
0
.
000
32.56
20.20
C 2 1.8583
0.9292
0.007
6.85 4.25
D 2 25.7901
12.8951
0.000
95.12
59.02
E 2 4.4609
2.2304
0.000
16.45
10.21
Sai số
16
2.1691
0.1356
-- -- 5.71
Tổng 26
43.6964
-- -- -- 100
3.4. The actual results
After using the Taguchi method to obtain the optimal
parameters, the research conducted injection molding of the
output product. The product is displayed in Fig. 7 with a
reasonable shrinkage level. In addition to reducing the
shrinkage phenomenon, the use of optimal parameters helps
the company save both material and injection molding time.
Fig. 7. Actual Product
4. CONCLUTIONS
The study focused on identifying optimal injection
molding parameters to reduce product shrinkage. From the
Taguchi analysis results, the optimal injection molding
parameters proposed are: cooling time temperature (40℃).
Melt teamperature (215℃), cooling time (20s), paking time
(4s), and packing pressure (80MPa). From the ANOVA
analysis table, we observe that packing time (59.02%) and
melt tempeture (20.20%) have the largest influence and
determination on the occcurrence of shrinkage. Optimizing
these parameters has proven highly effective in minimizing
shinkage, rather than rely on traditional trial-and-error
methods which often consume significant time and
resources. Additionally, it helps the company enhance the
plastic injection molding process, including: increasing
productivity, quality, and reliability of plastic products.
REFERENKES
[1]. B. KC., O. Faruk.,J.A.M. Agnelli., A.L. Leao., J. Tjong a, M. Sain, 2016. Sisal-
glass fiber hybrid biocomposite:
Optimization of injection molding parameters using Taguchi method for reducing shrinkage, vol. 83, ScienceDirect,. 152-
159.
[2]. Sigit Yoewono Martowibowo & Agung Kaswadi , 2017.
Optimization and Simulation of Plastic Injection Process
using Genetic Algorithm and Moldflow, vol. 30, Chinese Journal of Mechanical Engineering, 398-406.
[3]. Mustafa Kurt, O. Saban Kamber, Yusuf Kaynak, Gurcan Atakok, Oguz Girit, 2009.
Experimental investigation of
plastic injection molding: Assessment of the effects of cavity pressure and mold temperature on the quality of the final
products,
vol. 30, ScienceDirect, 3217-3224.