Figure 5 displays the variation of the volumetric shrinkage with time in the Moldflow analyses. The simulation results of the volumetric shrinkage shown in Fig. 5 are selected from  27

Fig. 5   The results of volumetric shrinkage with time for FE analyses of 4, 10, 16, and 22

Fig. 6    The change of the clamp force with the   time

Fig. 7    The change of the recommended ram speed with the   stroke

FE analyses equally. From the results shown in Fig. 5, it can be

Fig. 8    The measurement of experiments

Here, Lcavity is the length of the mold  cavity;  Lproduct  is the target length of the molded plastic part of the DVD-ROM cover.  The  values  of  Lproduct  for  the   DVD-ROM  cover were accurately measured by a  profile  control  instrument (T-S  INDUSTRIES Inc, St. James,  MN).

seen that the values of the volumetric shrinkage versus time exhibit an unstable trend for FE analyses of 4, 10, 16, and 22. Furthermore, these analyses demonstrate a general variation for

Shrinkage ð%Þ¼ 

.. Lcavity  — Lproduct.

Lcavity

.

m 100

ðEq 19Þ

all of the 27 FE  analyses.

Figure 6 demonstrates the change of the clamp force in the plastic injection machine versus time for the plastic injection of the DVD-ROM cover. It can also be inferred that the clamp force increases regularly as the time increases to 3 s and  after 3 s, it remains constant. Figure 7 indicates the change of the recommended ram speed with the stroke, and it can be emphasized that the recommended ram speed increases regu- larly as the stroke increases and then decreases continuously.

5. The Measurement of Experiments

The experiments for four plastic injection moldings of the plastic part of the DVD-ROM cover and their measurements, as shown in Fig. 8, were carried out to control the simulated values obtained from the developed regression model with FE analyses obtained from Moldflow. The four injection moldings were taken from ten plastic molding experiments (4 parts   per

10 experiments). Five measurements were performed for a total of 40 experiments and each measurement was repeated at least three times. The formulation written in Eq 19  was utilized to control the simulated values with the FE    analyses.

Table 9 shows the results of the measured values, the simu- lated values, and FE analyses. In Table 9, it can be seen that the measured values match very well with the simulated val- ues and FE analyses. The APE was found to be 2.887% for the regression model (the simulated values) and 0.972% for the FE analyses.

6. Conclusion

In this study, an effective regression model was developed to determine the mathematical relationship between the plastic injection molding parameters and the volumetric shrinkage of the DVD-ROM cover. The orthogonal array of Taguchi (L27) designed by the process parameters of the mold temperature, melt temperature, injection pressure, injection time, and cooling time was created to run 27 FE analyses in Moldflow simulation program. ANOVA analysis was performed to identify the most significant process parameters and to check the adequacy of the regression model developed on the shrinkage. Based on the first-order and the second-order models for the shrinkage, the maximum error (ME) was found to be 3.895 and 3.994 %, respectively. Finally, the experiments consisting of four plastic

Table 9    The results of the  experiments

The errors (APE %)

Experiment no Measured values