摘要
Product packaging processes often encounter issues, such as warping and deformation caused by the mismatch between the coefficients of thermal expansion of materials. Considerable amounts of epoxy molding compound (EMC) are used in packages. EMC exhibits viscoelastic properties and cure shrinkage, which can substantially affect the stress and warpage of packages. Mathematical models of cure shrinkage at isothermal and variable temperatures have not been comparatively discussed. Accordingly, this study aims to clarify the differences between the cure shrinkage during EMC curing estimated by the isothermal and modified two-domain Tait models. A fitting flow and calculation procedure for cure shrinkage was established to analyze the influence of these models on structural warpage. This study involved the experimental measurements of EMC material properties by using the cure reaction kinetics, isothermal and variable-temperature cure shrinkage, and generalized Maxwell models and degree of cure at gel point. Structural warpage was estimated by developing a process-oriented finite element method simulation and validating simulation results against experimental results. Analysis revealed good agreement between the warpage estimated by the two cure shrinkage models and experimental warpage. Compared with those of the other model, the results of the two-domain modified Tait model better matched the experimental results, with a deviation of approximately 1 %, because this model considered cure shrinkage that precisely corresponds to the actual process. Furthermore, the stress relaxation of viscoelastic behavior was validated through a parameterized post-curing process. Results indicated that although warpage occurrence reduced by extending the post-curing process time, it eventually converged to a certain value.