Time- and Temperature-Dependent Thermo-Mechanical Modeling of a Packaging Molding Compound and its Effect on Packaging Process Stresses

Abstract

For reliable virtual thermo-mechanical prototyping of electronic packages appropriate descriptions of the mechanical behavior of the constituent materials are essential. In many packages molding compounds are used for encapsulation and underfill to provide environmental protection and/or to improve the package thermal mechanical reliability. Therefore, among others, the availability of appropriate constitutive models for various epoxy-molding compounds is one of the requirements for computational prototyping. As there is a large variability of available molding compounds, it is essential to be able to experimentally establish the model parameters in an efficient manner. Because of the implied simplicity, linear visco-elastic models combined with the time-temperature superposition theory are mostly used in thermo-mechanical simulations. Among the various experimental possibilities to efficiently establish the model parameter functions, in the present paper the use of unidirectional creep testing is worked out for a chosen molding compound. Here isothermal one-day creep experiments at different temperatures (ranging below and above the glass transition temperature of the compound) are performed. The tensile creep compliance and the time-dependent Poisson’s ratio of the material at different temperatures are successfully used to construct visco-elastic master curves. As the Poisson’s ratio shows a significant change during a creep or relaxation test, its effect in partly constraint situations (as in packages) will be evident. Therefore it is not reliable to approximate this variable using a constant value. Further, the visco-elastic model of the material is implemented in a finite element program and verified by means of a shear stress relaxation experiment and a creep experiment both under nonisothermal conditions. Moreover, the effect of the creep behavior of the molding compound on the packaging process stress field and its evolution is investigated. Substantial cost saving was realized by package design optimization based on the reliable prediction of the packaging process stresses.