In the Table 4, the model “F-value” of 39。3 indicates that the model is considered to be statistically significant。 “P- value” less than 0。05 indicate the model terms are significant。 In this manner, all the single terms (A, B, C, D, E), interac- tion terms (AB, AD, BC, BD, BE) and quadratic terms (B2, C2) were found to be significant model terms。 The  other terms which P-value > 0。05 are not significant model terms which have little effect on the response variables in the de- sign space。

In the Table 5, the model “F-value” of 396。31 also implies that the response quadratic model is very significant。 There is only a 0。01 chance that a “model F-value” this large could occur due to noise。 Based on the identification that the factors with “P-value” bigger than 0。05 are insignificant terms, the

Figure 8。 Predictive values and numerical values: (a) clamping force, (b) warpage。

Figure 9。 History plot of responses during optimization process using NSGAII: (a) history optimization of clamping force,

(b) history optimization of warpage。

single terms of A, B, C and D, the interaction term of BD and the quadratic terms of A2, B2, C2, D2 are significant model terms for the clamping force。

Figure 8 presents a comparison among predicted and nu- merical experimental values for desirability。 Predicted values were in agreement with the numerical data。 R-squared clamp- ing force and warpage were 0。99 and 0。99, respectively, indi- cating highly-accurate results for the regression models。 The developed models could thus be applied in the optimization process。 The optimization process will be described in the next step。

4。 Optimization results

Process parameters such as mold temperature, melt tem- perature, packing time, packing pressure and cooling time have complex effects on objective functions。 An objective for setting clamping force values was that equality constraint。 In this study, warpage values acted as inequality constraints。 Through   practical   conditions,   acceptable   warpage values

were set as being smaller than the initial value。 The optimiza- tion problem is described based on the following expressions:

Find � =