Suggested Geometric Finite Element Model

The choice of model type is most crucial to the computational expense involved in performing the second-level reliability calculations. While some authors (Ref 31, 48) have had success with 3D octants and 3D octants combined with sub-models, these analyses are typically limited to smaller packages with depopulation and approximately 200 to 500 solder joints (or I/O). A significant amount of interest exists, however, in the reliability of larger packages, such as those with 1500 and greater I/O. 

Due to their physical size, these large devices are more prone to early failures than the lower I/O packages. Constructing a 3D octant model of a 2000 I/O package with 2 planes of reflection symmetry is a costly computational proposition. The 3D slice approach easily captures the physi- 

A 256 Input/Output PBGA package is analyzed in the next section. Consideration is paid to presenting specific characteristics of the modeling strategy, as well as discussing the physical response in the context of the material properties. 

---

Based on the original rock mechanical parameters, the author suggested that the similar material could be simplified into the sandstone, shale and coal three physical finite element plate (PFESA) to simulate the geological structure of the practical engineering. This geomechanical model test adopted different plates of five kinds of dimensions and their geometric dimensions and mechanical parameters are shown in Table 2. 

---