3.1. Experiments of foamer analysis

The altering of foaming agent will affect not only the microcellular structure of products but also injection defect, such as warpage, volume shrinkage and sink mark depth so on. The molding quality is directly related to the dimension precision and mechanical properties of products, so there are close relationship between the foaming agent and product quality. Nitrogen and carbon dioxide are the two kinds of gas foaming agent which are commonly used in microcellular foam injection molding. In this section, CO2 and N2 were selected as study object. While the foaming agent was being altered, the other processing parameters were kept in a fixed level. The influence of foaming agent on microcellular injection molding and injection defects was  investigated  by conducting CAE experiments.

3.2. Factorial analysis

There are three kinds of factors affect the value of warpage in microcellular injection molding, such as material, product design, and manufacturing processing. In this study, the injection defects were reduced by controlling the processing parameters. However, there are many parameters in injection processing, it is difficult to find out the influence of the parameters on warpage at the same time. Fractional factorial design can be used to investigate the effects of two or more factors, it is more efficient than one-factor-at-a-time experiments. In the present study, the following parameters are selected as subject investigated: flow rate, V/P switch-over, mold temperature, melt temperature, coolant temperature, coolant 5H\QROGV QXPEHU, volume filled at start of foaming, initial bubble radius, number of cells per volume, initial gas concentration. In this section, L32  array are employed to investigate the

influence of the selected parameters on warpage. The processing parameters and their levels selected are indicated in Table 2.

Table 2. The processing parameters and their levels.

4. Results and analysis

4.1. Analysis of foamer selected

In this section, the CAE experiments with different foamer were carried out in MoldFlow. The main results were recorded as shown in Fig. 3. Fig. 3(a) shows the different warpage results of key nodes on the FE model, which were collected from the CAE analysis by using different foaming agent. Nodes 67-73 are on the continuous area in the middle of the tensile spline, node 70 is in the central of the FE model. It is observed that, the warpage changed trends under different foaming agent are coincident, which can be summarized as follow: warpage of central product is minimum, and warpage increases gradually from the center to both ends of the tensile spline. It also can be found that warpage of each sample region under N2 are less than under CO2 . The reason related to the solubility of a gas in the melt, the solubility of CO2 is higher than N2 under the same temperature and pressure condition. The viscosity of the melt including N2 is smaller than the melt including CO2 , so the melt including N2 has better filling property, which will reduce the residual stress in injection molding.

The sample can be pided into three parts along the thickness direction, which are surface layer, interlayer and core layer. The elasticity modulus and bubble radius of different layers under various foaming agents are shown in Fig. 3(b/c). It is observed that, there are highest elastic modulus and smallest bubble radius in surface layer, and there are lowest elastic modulus and largest bubble radius in core layer. Elastic modulus changing trend with the thickness direction is opposite to bubble radius, the reason is that the effective stress area under tensile loading is diminishing with the increasing of bubble radius. It also can be found that, the bubble radius of interlayer and core layer under N2 are smaller than under CO2 , and the radius of surface layer under N2 is larger than under CO2 .