The packaging community relies on finite element method (FEM) models to predict stresses that develop within microelectronic packages. However, FEM model predictions do not always accurately predict or explain package failures. Material heterogeneity and non-perfect geometries are not typically accounted for in such models and there has been debate about whether the physical models used in standard FEM models best describe complex viscoelastic materials such as underfills. We present preliminary results from a method to directly measure local stresses in a viscoelastic model underfill using carbon nanotubes as sensors with the aim of developing a robust experimental method that can be used to validate or supplement FEM model predictions.
Read full abstract