INTRODUCTION
Fiberreinforced thermoplastics have been gaining popularity as substitutions
for many of the weight critical components in the aerospace and automotive industries
(Mallick, 1993). There exists an abundance of fiberreinforced
thermoplastics that exhibit material properties such as strength and modulus
those are either comparable to or better than traditional metallic materials
(Zampaloni et al., 2004). These materials can
exhibit the same strength properties as sheet steel, but at a fraction of the
weight (Dweib and O'Bradaigh, 1998). In response to
driving forces requiring weight reduction in lightweight structures, a stamp
forming process has been investigated using Glass Fiber (GF), Carbon Fiber (CF)
and hybrid carbonglass fiber fabrics impregnated with polyamide (PA) (Wakeman
et al., 2005).
Since the combination of thermoset composites through structural adhesive is
both timeconsuming and laborintensive (Stavrov and Bersee,
2005), which generally involves the dry forming of fibrous woven materials
(Boisse et al., 2011; Cao
et al., 2008; Peng and Rehman, 2011) performs
postmold resin infiltration and cure process. Thermoplastic composites with
fabric reinforcement seem more suitable for producing and can be processed rapidly
from intermediate materials using a melting and curing process. They offer the
potential advantage because they are molded in massproduction more easily than
are reinforced thermosets. Furthermore, the thermoplastics do not undergo any
additional chemical reaction during cure, which theoretically makes the process
simpler and faster (Greco et al., 2007; TrudelBoucher
et al., 2006).
Considered here is the stamp forming of carbon fiber/polyamide6 (CF/PA6)
composites that is analogous to match die sheet metal stamping. This study describes
an experimental study on the development of a twodimensional stampforming
method for the processing of CF/PA6 composites with thermoplastic matrices,
in which two different angle molds (90°, 120°) and a mechanical press
are used. Splitplot designs are used in the experiments of Taguchi L_{16}
orthogonal array. The present study will focus on the processing conditions
(factors), e.g., the thermoforming temperature, mold temperature, mold holddown
pressure and time, required to give highquality rightangle parts are established.
Through this experiment and data analysis, can determine the optimal process
condition and establish quality characteristics of the shortbeam strength (SBS)
prediction mode. Stamping process by different angles, these results could be
used as a further understanding about the relationships between the forming
angle and strength. It also pointed out that, the estimated equation calculated
from the stamping results of the two mold angle, when the preheating temperature
is higher, obviously, can produce higher shear strength.
MATERIALS AND METHODS
Experiment design: The present study focuses on the processing conditions,
e.g., forming temperature, mold temperature, mold pressure and time. Parameters
selection and their level described as Table 1. L_{16}
orthogonal array is used in splitplot designs, allocation and trial sequence
shown as Table 2. According to the degree of difficulty on
level change, Taguchi's orthogonal array applied to splitplot design, the levels
of the factors are expressed as 1's, 2's and 3's. 1's factors defined as the
level is the most difficult to change, 2's and 3's factors defined as difficult
and less difficult respectively. The groups correspond to plots in splitplot
design. The whole plot can be allocated to group 1, subplots to group 2, subsubplots
to group 3 and subsubsubplots to group 4. Column 1 allocated to group 1;column
2, 3, column 4, 5, 6, 7 and column 8, 9, 10, 11, 12, 13, 14, 15 allocated to
group 2, 3, 4, respectively. 1's, 2's and 3's factors could be allocated to
each group, which make the level of each factor are treated repeatedly 8 times.
Experiment materials and processing: CF/PA6 composite sheets are used
in the present experiment, provided by the Applied Fiber System Inc., USA. The
billet, 914x914x1 mm (lengthxwidthxthickness), should be cut into the size of
10x7x1 mm held for forming. Thirty two pieces is needed and stacked 2 pieces
into one pairs.
Stamp forming process was carried out as shown in Table 1.
Each heated pairs were positioned between the top and bottom mould (Fig.
1) halves in a press and closing the heated mould halves as quickly as possible.
Table 1: 
Description of parameters selection and the degree of difficulty
on level change 

1's, 2's and 3's factors defined as the degree of difficulty
on level change 
Table 2: 
L_{16} Orthogonal array allocation and trial sequence 

A: Mold angle, B: Forming temperature, C: Mold temperature,
D: Holddown pressure, E: Pressure time, e1: First order error, e2: Second
order error, e3: Second order error 

Fig. 1: 
Perspective view in the upper right and the other three for
the 3D projection view of bottom mold 

Fig. 2: 
Schematic diagram of the stamp forming process 
While cooling down to mould temperature the matrix solidified. After the press
was opened, the finished laminate can be removed. Transport of the pairs from
the heating source to the press and the thermoforming operation itself must
be carried out speedily to allow for completion of the pressing operation before
the thermoplastic falls below its recrystallization temperature. Otherwise
fabric shearing will be impeded. Processing diagram, shown as Fig.
2.
After the abovementioned process, 16 solidified laminates were completed.
The Shortbeam Strength (SBS) tests are conducted on model4206 testing machine,
made by INSTRON Ltd. USA. According to the ASTM standard, the test specimens
for the SBS test were rectangular shaped and with the following nominal dimensions:
1x0.25 inch. From the maximum force (P, kg), the interlaminar shear strength
(F, kg mm^{2}) was calculated as follows Eq. 1:
where, w and t are the width and thickness of the specimen, respectively. A
minimum of three specimens per condition was tested. According to the reference
standard (ASTM, 2006), the SBS test is suited for comparative
testing of composite materials, provided that failures occur consistently, that
is in the same mode. The SBS test results are shown in Table 3.
RESULTS AND DISCUSSION
In the present experiment, mold angle is a qualitative factor, which may be
one of the specifications of the product, not the process parameters. Therefore,
this study needs to explore these two angles to build their response equation,
to serve as the basis to explore the processing of different angles or shapes
in the future. Next, the experiment uses the SBS as the response variable, comparing
the two forming process under the different mold angles, to discuss the optimum
parameter model.
Model of the mold angle 90degree: The data below, resulted from measuring
the SBS under mold angle 90°. Table 4 is extracted by
using the data in Table 2, 3, i.e., the
data in the column 3 (factor B) of Table 2 is allocated as
Table 1.
The analysis of variance (ANOVA) is the statistical treatment most commonly
applied to see which process factors significantly affect the process responses.
The variance is the amount of deviation degree, which is obtained by finding
the sum of squares of Fvalue between a set of data and actual value. The Sig.
means the significance, which is calculated from the experimental results. Then,
it was compared to the given critical significance level 0.05 representing significance
levels at a 95% confidence interval. If the Sig. calculated is less than 0.05,
it is an indication that the statistical test is significant at the confidence
level selected. If not, it indicates that the statistical test is not significant.
The results of ANOVA for the Taguchi in splitplot method experiment are tabulated
in Table 5.
The test statistic is the F value of 10.993 and 7.152. Using an α of 0.05,
the results of the test have met the significance level (F_{0.05(1(A5)}
= 6.608). The F test is a test to determine the overall significance of the
model and not just of one individual coefficient. There are two classes of effects
that this research is interested in: Main Effects and Interactions. It shows
that both main effects, B (Forming Temperature) and C (Mold Temperature), are
significant at the p<0.05 levels, yet the interaction is not. Therefore,
the best combination of factor levels can be determined in accordance with the
main effect. The best combination of factor levels can display by the main effects
plot for SBS, as shown in Fig. 3.
Table 3: 
Shortbeam strength test results by L_{16} orthogonal
array experiment 

Table 4: 
Shortbeam strength test records under mold angle 90° 

Table 5: 
Results of ANOVA for Taguchi in splitplot designs under mold
angle 90° 


Fig. 3(ab): 
Main effects for SBS in the case of mold angle 90° (a)
Main effects of B (forming temperature) and (b) Main effects of C (mold
temperature) 
As can be seen in the above plots, under the mold angle 90°. It appears
that the best result is obtained when each of the factors: B2 and C1 is at their
high level. Thus, the best combination of factors can be obtained, i.e., when
the mold angle is 90°, the best combination of the parameters will be B2
(263°C) and C1 (105°C). The following estimated equation (Eq.
2) is the model under mold angle 90° for the response variable SBS:
Model of the mold angle 120degree: The data below, resulted from measuring
the SBS under mold angle 120°. Table 6 is extracted by
the same way, similar as the paragraph of Mold Angle 90°. The results of
ANOVA for the experiment are tabulated in Table 7.
There are concluded that the main effects B (Forming Temperature), C (Mold
Temperature), D (Mold Pressure) and the interaction BC are significant. Therefore,
the best combination of factor levels can be determined in accordance with the
main effects and interaction effect simultaneously. This interaction plot confirms
the significance of BC interaction as stated below. Interaction occurs when
one factor does not produce the same effect on the response at different levels
of another factor. Therefore, if the lines of two factors are parallel, there
is no interaction. On the contrary, when the lines are far from being parallel,
the two factors are interacting. In the case of BC interaction, the Main Effects
Plot for SBS is shown in Fig. 4, the molding temperature is
in high level, there will be a higher SBS. Therefore, high level molding temperature
B2 (263°C), which is one of the best factors. Then, if fixed B1, there is
a higher SBS in C2 (105°C). Then, to observe the main effect of factor D,
referring to Fig. 4, the best factor mold pressure level is
D2, mold pressure 39 kg cm^{2}.
By the results of the above analysis, the best combination of the parameters
will be B2 (263°C). C2 (115°C) and D2 (39 kg cm^{2}). The following
estimated equation (Eq. 3) is the model under mold angle 120°
for the response variable SBS:
Confirmation experiment: In order to verify the prediction ability of
the estimated equations, the final step of the Taguchi method is the confirmation
experiments conducted for examining the quality characteristics. The model used
in the confirmation tests is defined with the main effect generated by the control
factors (processing conditions).
Table 6: 
Shortbeam strength test records under mold angle 120° 

Table 7: 
Results of ANOVA for Taguchi in splitplot designs under mold
angle 120° 


Fig. 4(ab): 
Main effects for SBS in the case of mold angle 120° (a)
Interaction plot between BxC, (b) Main effects of D (mold pressure) 
The optimal SBS obtained by taking into account the influential factors within
the evaluated best combination. Therefore, the predicted optimum SBS (Eq.
2 and 3) was calculated by considering individual effects
of the factors (A1, B2, C1, D2, E2) and (A2, B2, C2, D2, E2). The optimal SBS
(90°) was computed as 50.35 kg mm^{2} and another one (120°)
was 49.02 kg mm^{2}.
The confidence interval was employed to verify the quality characteristics
of the confirmation experiments. The confidence interval for the predicted optimal
values is calculated also. In this study, three confirmation experiments (r
= 3) were carried out to evaluate the performance of experimental trials for
SBS under optimal conditions. The confidence interval was calculated as (46.552,
54.123) and (47.312, 50.717).
Table 8: 
Comparison between confirmatory test results and theoretical
prediction under the best combination of the parameters 

With a 95% confidence level, the confirmatory test results were 48.67 and 48.16,
fell in the confidence interval. Therefore, the optimization for SBS was achieved
using the Taguchi method at a significance level of 0.05. The experiment condition
of the best combination of the parameters was done repeatedly. The data from
the confirmatory test results are shown in Table 8. These
tests have confirmed the prediction ability of estimated equation and accepted
the repetition of the experimental results.
CONCLUSION
In this study, stamping trials were performed with thermoplastic composite
materials. The forming conditions were carried out by the Taguchi method in
splitplot designs. Stamping by different angles, this result can be used as
a further understanding about the relationship between the forming angle and
strength. In the performed experimental trials using Taguchi L_{16}
orthogonal arrays, obtain two sets of best combination of the parameters. It
was found that the main effects of mold angle 90°, B (Forming Temperature)
and C (Mold Temperature), are significant at the p<0.05 levels, yet the interaction
is not. From their main effect plot, this study obtains the best combination
of the parameters, B2 (263°C) and C1 (105°C). Similarly, from the condition
of mold angle 120°, there get another best combination of the parameters,
B2 (263°C), C2 (115°C) and D2 (39 kg mm^{2}).
From both estimated equation, the shortbeam shear stresses measured at the
specimens from different mold angles were almost equal. By comparing specimens
shear results, it was observed that the forming temperature and mold temperature
were the dominating parameters, no matter what kinds of mold angle are. It appears
that temperature is the most decisive factor, especially the forming temperature.
Finally, for verifying the prediction ability of the estimated equations, the
confirmation experiments were conducted. The confirmatory test results, 48.67
and 48.16, fell in their confidence interval, respectively. It shown that the
prediction ability of estimated equation and accepted the repetition of the
experimental results have confirmed by the tests.