An examination of the applicability of the DNP metric on first level reliability assessments in underfilled electronic packages

The notion of using the distance from neutral point (DNP)¹ as a metric to assess the criticality of first level package thermomechanical stresses is examined in this paper. This metric has been applied by assuming that testing of electronic package assemblies with largest applicable DNP is required to infer the quality of similar packages with smaller die dimensions. Through a review of complex thermomechanical analyses available in the literature together with finite element studies performed herein for representative package geometries, this simple yet commonly applied approach is shown to be an overly conservative methodology for packages with first level underfill. For the representative cases considered, the local edge stresses and strains presumed to cause first level failure are shown to increase with DNP to a point where they no longer increase. A new methodology is presented based on this information for estimating test vehicle configurations needed to confidently qualify similar packages with larger die sizes.