International Review of Mechanical Engineering (I.RE.M.E.), Vol. 5, N.

7
November 2011

A Study on Two Plate and Three Plate Mold

of Ultra Thin Plates in Minimizing Warpage Issue

M. Fathullah1, Z. Shayfull1, S. Sharif2, N. A. Shuaib1, S. M. Nasir1

Abstract – The demand of ultra thin plastic parts in global industries is increasing today. Ultra
thin plastic parts are widely produced using injection molding processes and it is understood that
warpage issue is regularly found in producing these parts. This situation happens due to lacking
of knowledge in selecting the best injection molding parameters to produce parts with minimal
warpage. In addition, designers also have difficulties to control the parameters since the 0.3mm
thickness of the ultra thin parts is too hard to maintain. Therefore this study is performed
purposely to determine the best parameters can be selected in manufacturing ultra thin plates.
Two types of gating systems which are side gate for 2-plate mold and pin point gate for 3-plate
mold are tested. The results are obtained using Taguchi Method and Analysis of Variance
(ANOVA) and run through simulation software. Both parameters are then compared with each
other in recommending molding designers which is the best to be applied at mold design stage.
Results from this research recommend using pin-point gate for 3-plate mold using the outlined
parameter setting in order to obtain the minimal warpage of ultra thin wall plastic plates.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Thin Plate, Injection Molding, Warpage, Taguchi Method, ANOVA

I. Introduction A recent research made by J.G. Kovács and B. Sikló

Injection molding process is considered as one of the
most popular and significant processes in mass
production of plastic components especially those parts
to be produced with intricate geometries. It is well
known that the quality of the injection molded parts
highly relies on the design of mold including its
processing parameters as well as type of materials used
[1]-[2] and this means injection molding processes are so
much involved since 33% of plastics are transformed
into parts needed by means of injection molding
processes [3]. Common defects of injection molded parts
such as warpage, shrinkage, weld line and sink marks
normally occur due to unfavourable process conditions
which give great impacts on the quality of the parts to be
produced.
S.H Tang [3] stated that increasing demand for thinner
and thinner molded plastics parts has led to the
appearance of warpage on parts due to the reduction of
wall thickness where the process becomes more difficult
to be controlled and more complicated. There have been
researches in finding the optimal parameters of injection
molding process of plastics and its properties. For an
example, Sadabadi and Ghasemi [4] as well as B.
Ozcelik et al. [5] have performed investigations on the
influences of injection parameters including mold wall
temperature, injection flow rate, packing pressure which
could have effects on the defects phenomenon such as
warpage and weld line of injection molded plastics.
[1] where V-top specimens are made with a special mold
designed with variable inserts and it was found that
warpage is mostly affected by fibre content and mold
temperatures but insignificant in view of holding
pressure when using polypropylene along with glass
fibre. For ultra thin parts in injection molding processes
which have less than 0.3mm thickness needed to be
produced, they have huge potentialities to be applied on
Micro-electro-mechanical systems (MEMS) but the
injection molding process becomes more complicated
with thinner part thickness compared to ordinary thin
parts [6]. Therefore as far as warpage defect is
concerned, an ultra thin plate with 120 mm × 50 mm and
0.3 mm dimension is to be analyzed. An experiment on
warpage analysis on an ultra thin part has been done by
S. H. Tang et al. [3] which is only limited to two plate
mold design and it is found that by using 9 trials in
Taguchi method, the best parameters that can reduce the
warpage occurrences on injection molded parts are melt
temperature (240°C), filling time (0.5 s), packing
pressure (90%) and packing time (0.6 s). The most
significant factor is melt temperature whereas the filling
time is considered not significant. Therefore the research
is extended in this paper to explore and identify the best
parameter setting and to compare which setting is better
between side gate with 2 plate mold and pin point gate
with 3-plate mold. Side gating system using two plate
mold design is by far the most basic type of mold can be
fabricated [7]. It is considered as the simplest and lowest

Manuscript received and revised October 2011, accepted November 2011 Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

1189
 M. Fathullah, Z. Shayfull, S. Sharif, N. A. Shuaib, S. M. Nasir

cost to get constructed compared to other types of molds
whereas pin point gate using three plate molds comprises
of three mold sections that move relative to each other
where each section consists of one or more plates. This
mold provides better freedom with respect to gating
locations and the feeds layout [8].
II. Experimental Method
According to the study on two plate mold analysis, the
best parameter setting for this experimentation is
identified using Taguchi method [3]. Hence, as to follow
the same process, determination of factors that can
influence on warpage, appropriate factor levels as well as
the orthogonal array are based on Taguchi method is
essential in determining the most favorable set of
parameters to produce minimal result of warpage.
However, a few steps are still needed to be considered
which are:
a. Selection of Factors,
b. Orthogonal Array and Factor Levels,
c. Material used,
d. General Assumptions,
e. Simulation.
a. Selection of Factors
As what have been regularly assumed in studies
performed in previous researches [9]-[11], mold
temperature is again ignored due to an assumption of the
presence of ambient. The factors taken into
considerations are only melt temperature, packing time,
mold temperature, filling time, and packing pressure.
limited with assumptions that:
1. Only the effects of the 5 factors are taken into
considerations and other factors are neglected.
2. Core and cavity molds are assumed to be at the same
and uniform temperature.
3. The temperature surrounding the cooling channels is
assumed constant.
4. The ejection temperature is set at 158°C and the
temperature of the environment is 27°C.
5. The liquid coolant is assumed as pure water.
e. Simulation and Design
The experiment is done using Autodesk Moldflow
Insight 2011 simulation softwares that helps simulating
the parameters set. The parts to be simulated are divided
into 8522 pieces of surface triangles with 4326 nodes for
both ultra thin plates. The parts meshes and the design of
the cooling channel can be seen in Fig. 1. Signal-to-noise
(S/N) ratio is then calculated according to Table I and the
warpage result with respect to z-direction is acquired to
identify the best setting parameters experimented. Then,
the percentage of contribution is calculated to determine
which of the factors have significant effects towards
warpage issues.
The z-deflections data for both side gate for 2-plate
and pin point gate for 3-plate mold as shown in Table II
are obtained from the simulation process represent the
warpage values which is to be studied using Analysis of
Variance (ANOVA) where level of confidence is set at
0.05.

b. Selection of Orthogonal Array and Factor Levels

Four levels of each factor are selected in an L16 45
array where the selection of the array is because of its
suitability for five factors to be analyzed with four levels.
The four-different levels to be considered are filling
time, packing pressure and packing time and based on
the same values used by S. H. Tang [3] and the levels of
mold and melt temperature is chosen based on the
thermal properties of Techno ABS 500 as well as
Fig. 1. Cooling channel design
recommended processing value by Autodesk Moldflow for side gate using 2-plate mold
Insight 2011 software.

c. Material
Material applied is Techno ABS 500 which is a
copolymer and generally applied in manufacturing
automotive exterior parts. It also has good stiffness and
can be retained at elevated temperatures.

d. General Assumptions
There are some more external factors influencing the
results of injected parts and among them are materials’
types, shape and dimensions of parts, selections of mold
materials, position of coolant runners and the use of
liquid coolant [12]-[13]. However, in order to simplify Fig. 2. Cooling channel design for pin point gate
by using 3-plate mold
the experiment, the scope of this study needs to be

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Mechanical Engineering, Vol. 5, N. 7

1190
 M. Fathullah, Z. Shayfull, S. Sharif, N. A. Shuaib, S. M. Nasir

The results are then brought back to be compared with
the results obtained from the S/N ratio and in the end the
percentage of contribution is determined to recognize
which selected factors highly influence the warpage of
ultra thin wall parts produced.
Using the control factors shown in Table I and by the
help of Autodesk Moldflow Insight 2011, the total of 32
trials are run for both 2-plate and 3-plate molds of ultra
thin wall parts. Fig. 3 and Fig. 4 show the z-deflections
measured in the experiments respectively.

TABLE I
CONTROL FACTORS AND LEVELS FOR SIDE GATE USING 2-PLATE MOLD AND PIN POINT GATE USING 3-PLATE MOLD
Control Factor
Factor A Factor B Factor C Factor D Factor E
Trial No.
Packing time
Mold Temp. (°C) Melt Temp. (°C) Filling time (s) Packing Pressure (MPa)
(s)
1 40 220 0.1 60 0.1
2 40 230 0.2 70 0.2
3 40 240 0.3 80 0.3
4 40 250 0.4 90 0.4
5 50 220 0.2 80 0.4
6 50 230 0.1 90 0.3
7 50 240 0.4 60 0.2
8 50 250 0.3 70 0.1
9 60 220 0.3 90 0.2
10 60 230 0.4 80 0.1
11 60 240 0.1 70 0.4
12 60 250 0.2 60 0.3
13 70 220 0.4 70 0.3
14 70 230 0.3 60 0.4
15 70 240 0.2 90 0.1
16 70 250 0.1 80 0.2

TABLE II
SUMMARY OF RESULTS OF
Z-DEFLECTIONS AND S/N VALUES FOR 2 AND 3 PLATE MOLDS
2 PLATE MOLD 3 PLATE MOLD
Control Factor
Trial No.
z-deflections S/N z-deflections S/N
A B C D E
1 40 220 0.1 60 0.1 0.0058 44.7314 0.0058 44.7314
2 40 230 0.2 70 0.2 0.0050 46.0206 0.0047 46.5580
3 40 240 0.3 80 0.3 0.0043 47.3306 0.0041 47.7443
4 40 250 0.4 90 0.4 0.0037 48.6360 0.0026 51.7005
5 50 220 0.2 80 0.4 0.0044 47.1309 0.0039 48.1787
6 50 230 0.1 90 0.3 0.0043 47.3306 0.0021 53.5556
7 50 240 0.4 60 0.2 0.0055 45.1927 0.0047 46.5580
8 50 250 0.3 70 0.1 0.0055 45.1927 0.0053 45.5145
9 60 220 0.3 90 0.2 0.0046 46.7448 0.0043 47.3306
10 60 230 0.4 80 0.1 0.0059 44.5830 0.0050 46.0206
11 60 240 0.1 70 0.4 0.0045 46.9357 0.0045 46.9357
12 60 250 0.2 60 0.3 0.0050 46.0206 0.0029 50.7520
13 70 220 0.4 70 0.3 0.0054 45.3521 0.0044 47.1309
14 70 230 0.3 60 0.4 0.0052 45.6799 0.0044 47.1309
15 70 240 0.2 90 0.1 0.0055 45.1927 0.0056 45.0362
16 70 250 0.1 80 0.2 0.0053 45.5145 0.0055 45.1927

Fig. 3. Z-deflections for side gate using 2-plate experiment Fig. 4. Z-deflections for pin point gate using 3-plate experiment

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Mechanical Engineering, Vol. 5, N. 7

1191
 M. Fathullah, Z. Shayfull, S. Sharif, N. A. Shuaib, S. M. Nasir

III. Result and Discussion MSD = Mean Square Deviation,

y= Observations
Since the smaller warpage value, the better quality it n= No. of tests in a trial.
would be, hence in determining the S/N ratio of the
warpage, Taguchi has properly provided an equation in The data are also evaluated using Analysis of
calculating S/N ratio for this situation. The equation to Variance (ANOVA) technique where the percentage
obtain the values of S/N is shown based on this case is: contribution for all factors is determined by putting side
by side with the relative variance calculated. The
S / N - 10 log MSD examples of S/N ratio calculations for factor A are
1 n 2 shown below and the results of S/N ratio for the part are
where MSD y
i 1 i listed in Table III.
n

TABLE III
THE RESPONSE TABLE OF S/N RATIO
2 PLATE MOLD 2 PLATE MOLD
Level
A B C D E A B C D E
1 46.6797 45.9898 46.1281 45.4062 44.9250 47.6836 46.8429 47.6039 47.2931 45.3257
2 46.2118 45.9035 46.0912 45.8753 45.8682 48.4517 48.3163 47.6313 46.5348 46.4099
3 46.0710 46.1630 46.2370 46.1398 46.5085 47.7598 46.5686 46.9301 46.7841 49.7957
4 45.4348 46.3409 46.1081 46.9760 47.0956 46.1227 48.2900 48.7578 49.4058 48.4865
Diff. 1.2448 0.4374 0.1458 1.5699 2.1707 2.3290 1.7477 1.8277 2.8709 4.4700

44.7314 + 46.0206 + on the graphs shown in Figs. 5 and 6.

+ 47.3306 48.6360 TABLE IV
Level 1 46.6797 BEST COMBINATION SETTING
4
Side gate using 2-plate mold Pin point gate using 3-plate mold
Factors Values Factors Values
48.1787 + 53.5556 + E 0.4s E 0.3s
+ 46.558 + 45.5145 D 90MPa D 90MPa
Level 2 46.2118 A 40°C A 50°C
4 B 250°C C 0.4s
C 0.3s B 230°C
47.3306 46.02066 +
+ 46.9357 50.7520
Level 3 46.071
4

47.1309 47.1309 +
+ 45.0362 45.1927
Level 4 45.4348
4

Referring to the response table as calculated in Table

III, the best arrangement of parameters can be identified
by deciding on the highest difference value of S/N ratio
Fig. 5. All factors with 5 levels vs S/N values for side gate
from each factor. Therefore, for side gate using 2-plate using 2-plate mold
mold, the sequence of significance level starts with
Factor E, followed by Factor D. Factor A is at the third
place whereas Factor B and C are positioned at fourth
and fifth respectively. Likewise for pin point gate using
3-plate mold, the sequence is nearly the same except the
fourth and the fifth which are Factor C and B
respectively. In this case, the most significant factor that
has an effect on warpage for ultra thin plate is cavity
temperature (A) followed by packing time (D), core
temperature (B) and packing pressure (C). Looking at the
highest number of S/N values on every factor has led to
new finding of the best parameter setting to be
combined. Table IV shows the summary of best Fig. 6. All factors with 5 levels vs S/N values for pin point gate using
combinations of parameter. The result can also be seen 3-plate mold

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Mechanical Engineering, Vol. 5, N. 7

1192
 M. Fathullah, Z. Shayfull, S. Sharif, N. A. Shuaib, S. M. Nasir

In ANOVA calculations, the degree of freedoms for Se ST SA SB SC SD

all factors needs to be obtained first. As an example of 7 7
calculations, the degree of freedom for 2-plate mold is 8.7 10 1.2 10
shown below; 6 7 6
5.93 10 - 1.1 10 1.51 10
Total degree of freedom, f: 6
3.32 10
fT N -1 0
16 - 1 15
The values of variances for all factors are then
For Factor A: calculated.
fA kA - 1 The example of calculating variance is shown below:
4 -1 3 For Factor A:
where kA is the number of level factor A. 7
SA 8.7 10 7
VA 2.90625 10
For Error, fe: fA 3

fe fT fA fB fC fD fE For Variance Error, Ve :

15 -( 3 3 3 3 3 ) 0
Se 0
Ve 0
Sum of squares for all factors is then calculated and fe 0
the example of calculating sum of squares is shown
below; F-ratio, F for all factors is calculated afterwards and
the example of calculation is shown below:
Sum of squares, S:
For Factor A:
2
za1 za 2 ... zaN vA
ST za12 za22 ... zaN 2 FA
N ve
2 2
ST 0.0058 0. 0050 ... 0.00532
FA ,FB ,FC and FD cannot be determined as ve 0 .
2
0.0058 0.0050 ... 0.0053 Last but not least, Percentage Contribution, PA for all
16 factors are calculated and the example of the calculation
ST 5.93 10 16 is shown as:

7
For Factor A: SA 8.7 10
PA 100% 100% 14.7043%
ST 5.93 10-6
2 2 2
A1 A3 za1 za 2 ... zaN
SA ... Looking at the percentage of contributions from the
kA k4 N experiment which can be seen from Table V, the results
2 again agree that both side gate using 2-plate mold and
0.0058 0.0050 0.0043 0.0037
SA pin point gate using 3-plate mold are most significantly
4
influenced by packing time which contribute 56.02% for
2
0.0044 0.0043 0.0045 0.0050 side gate for 2-plate mold and 59.92% for pin point gate
4 for 3-plate mold.
2 These are followed by packing pressure which
0.0054 0.0052 0.0055 0.0053 contributes 16.76% and 25.41% respectively while mold
4 temperature is the third significant factor in influencing
2 the result of warpage.
0.0058 0.0050 0.0043 0.0053
The fourth and fifth positions are melt temperature
4
and filling time but even though the sequence is in
SA 8.7 10 7 contrary situation with the result in Taguchi Method,
both factors are considered negligible since the total
For Error: contribution factors are small.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Mechanical Engineering, Vol. 5, N. 7

1193
 M. Fathullah, Z. Shayfull, S. Sharif, N. A. Shuaib, S. M. Nasir

Table IV has recommended the best combination gate of 2-plate mold in an injection molding process of
parameters to be used for both molds; side gate for 2- ultra thin plate parts.
plate mold and pin point gate for 3-plate mold.

TABLE V IV. Conclusion

ANOVA TABLE FOR SIDE GATE USING 2-PLATE AND PIN POINT GATE
USING 3-PLATE MOLD The best combination parameters in reducing warpage
issue of ultra thin plates for side gate using 2-plate mold
2-PLATE MOLD
and pin point gate using 3-plate mold have been
Factors f S V F P(%) identified based on Taguchi Method supported by
A 3 8.7 10
7
2.90625 10
7
- 14.070443 ANOVA analysis. The most significant factor is packing
B 3 7 8
- 2.05544
time followed by packing pressure with their influence of
1.2 10 4.0625 10
7 8
contributions are more than 70%. Melt temperature and
C 3 1.1 10 3.5625 10 - 1.802467
filling time are less significant with their percentage of
D 3 1.51 10
6
5.0229 10
7
- 25.41372 contributions towards warpage issue is not more than
E 3 3.32 10
6
1.10729 10
6
- 56.02403 10%. The best parameters for side gate of 2-plate mold
Pooled and pin point gate of 3-plate mold have been confirmed
0 0 0 - - and from the result, it is highly recommended for mold
Error
Total 15 5.93 10
6
100 designers to design their mold in producing ultra thin
plate using pin point gate of 3-plate mold using the best
3-PLATE MOLD parameters outlined in this paper.
Factors f S V F P(%)
6 7
A 3 2.18 10 7.275 10 - 12.77997
References
6 8
B 3 1.47 10 4.90833 10 - 8.622456
[1] J.G. Kovács B. Sikló, Test method development for deformation
7 7
C 3 3.3 10 1.0917 10 - 1.917728 analysis of injection moulded plastic parts, Polymer Testing 30
6 7 (2011) 543–547.
D 3 2.86 10 9.5417 10 - 16.76182
[2] Mirigul Altan, Reducing Shrinkage In Injection Moldings Via The
5 6 Taguchi, ANOVA And Neural Network Methods, Materials &
E 3 1.023 10 3.4108 10 - 59.91802
Pooled Design, Volume 31, Issue 1, January 2010, Pages 599-604.
0 0 0 - - [3] S.H. Tang, Y.J. Tan, S.M. Sapuan, S. Sulaiman, N. Ismail, R.
Error
Total 5
Samin, The Use Of Taguchi Method In The Design Of Plastic
15 1.708 10 100 Injection Mould for Reducing Warpage, Journal Of Materials
Processing Technology 182 (2007) 418–426.
[4] H. Sadabadi, M. Ghasemi, Effects of some injection molding
These results are then brought to be tested again in process parameters on fiber orientation tensor of short glass fiber
order to testify what would be the result of z-deflections polystyrene composites (SGF/PS), Journal of Reinforced
for both types of molds if the combined parameters are Plastics and Composites 26 (17) (2007) 1729–1741.
applied. The experiment is run and the result obtained is [5] Babur Ozcelik, Alper Ozbay and Erhan Demirbas, Influence of
injection parameters and mold materials on mechanical properties
summarized in Table VI. Table VI shows the results of of ABS in plastic injection molding, International
Z-deflections for both types of molds and it is Communications in Heat and Mass Transfer 37 (2010) 1359–
acknowledged that both results are so small. 1365.
[6] M.C. Song, Z. Liu, M.J. Wang, T.M. Yu, D.Y. Zhao, Research on
TABLE VI effects of injection process parameters on the molding process for
Z-D EFLECTIONS OF BOTH MOLDS WHEN BEST COMBINATION ultra-thin wall plastic parts, Journal of Materials Processing
SETTINGS ARE APPLIED Technology 187–188 (2007) 668–671.
[7] Tim A. Osswald, Lih-Sheng Turng, Paul J. Gramann, Injection
Pin point gate for 3-plate Molding Handbook, Edition 2, Hanser Verlag, 2008.
Side gate for 2-plate mold
mold [8] David Kazmer, Injection Mold Design Engineering, Hanser
Factors Values Factors Values Verlag, pp. 125, 2007.
Packing time 0.4s Packing time 0.3s [9] M.F. Ghazali, Z. Shayfull, ,N. A. Shuaib, S.M. Nasir,, M. Mat
Packing pressure 90MPa Packing pressure 90MPa Salleh, Injection Mould Analysis in Reducing Warpage of Nylon
Mold temperature 40°C Mold temperature 50°C PA66 Side Arms Using Taguchi Method and ANOVA,
Melt temperature 250°C Melt temperature 230°C International Journal of Basic & Applied Sciences IJBAS-IJENS
0.3s 0.4s Vol: 11 No: 01 pp. 87-92.
Filling Time Filling Time
[10] Z. Shayfull, N.A. Shuaib, M.F. Ghazali, S.M. Nasir, M. Azaman,
0.0019 Z- Deflections 0.0011
Z- Deflections (mm) N.A. Faris, Optimizing Length of Weld Line Formation in Thin
(mm) Plate by Taguchi Method and Analysis of Variance (ANOVA),
International Journals of Engineering & Technology, Vol: 11
Issue: 01, ISSN: 2077-1185 pp.132-137.
Z-deflections for side gate for 2-plate mold after using [11] N.A.Shuaib, M.F. Ghazali, Z. Shayfull, S.M. Nasir, Warpage
the best combination parameters are 0.0019mm while the Factors Effectiveness of a Thin Shallow Injection-Molded Part
result for pin point gate for 3-plate mold is 0.0011mm. using Taguchi Method, International Journals of Engineering &
Therefore as far as the magnitude of warpage is Technology, Vol: 11 Issue: 01, ISSN: 2077-1185 pp.182-187.
concerned, it is recommended for mold designers to use
pin point gate of 3-plate mold rather than applying side

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Mechanical Engineering, Vol. 5, N. 7

1194
 M. Fathullah, Z. Shayfull, S. Sharif, N. A. Shuaib, S. M. Nasir

Authors’ information
1
School of Manufacturing Engineering,
Universiti Malaysia Perlis, Malaysia.
2
Faculty of Mechanical Engineering,
Universiti Teknologi Malaysia,
Malaysia.

M. Fathullah works as a lecturer in Universiti

Malaysia Perlis. His research interest is in
injection molding process and manufacturing
systems engineering. He has obtained Bachelor
of Engineering (Manufacturing) in 2006 from
International Islamic University Malaysia
(IIUM), Malaysia and then graduated with MSc
in Manufacturing Systems Engineering from
Coventry University, United Kingdom in 2007. He is a member of a
research group under School of Manufacturing Engineering in UniMAP
that is actively doing research specifically in analyzing and optimizing
parameters of injection molding processes. Apart from academic
research, M. Fathullah also actively involves with Engineering
Professional Bodies in Malaysia such as Board of Engineers, Malaysia
(BEM) and Institution of Engineers, Malaysia (IEM) until today.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Mechanical Engineering, Vol. 5, N. 7

1195