Predicting crack growth and fatigue lives of QFN solder joints using a multiscale fracture model

Dennis Chan, Dhruv Bhate, Ganesh Subbarayan, Jeff Zhao, Darvin Edwards

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The hierarchal fracture process model is a theory derived from continuum thermodynamics and information theory. This single parameter model provides a non-empirical approach able to predict crack growth for solder interconnects regardless of package or solder interconnect geometry. The model relates inelastic dissipations to the probability of failure at a given material point. The hierarchal fracture process model is used to predict crack growth and fatigue lives in a package with atypical solder interconnect geometry: a quad-flat-no-lead (QFN) package. The fracture model uses the material parameter calibrated for Sn3.8wt%0.7wt%Cu solder in a previous study of wafer-level CSP packages. Finite element models of these packages are created in ABAQUS with the solder interconnects given full creep and plastic properties. The results obtained from the hierarchal fracture process model are validated against the experimentally obtained fatigue lives and cross-sectional images of the crack path.

Original languageEnglish (US)
Title of host publicationProceedings of the ASME InterPack Conference 2009, IPACK2009
Pages207-213
Number of pages7
DOIs
StatePublished - Jun 30 2010
Externally publishedYes
Event2009 ASME InterPack Conference, IPACK2009 - San Francisco, CA, United States
Duration: Jul 19 2009Jul 23 2009

Publication series

NameProceedings of the ASME InterPack Conference 2009, IPACK2009
Volume2

Other

Other2009 ASME InterPack Conference, IPACK2009
Country/TerritoryUnited States
CitySan Francisco, CA
Period7/19/097/23/09

ASJC Scopus subject areas

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Predicting crack growth and fatigue lives of QFN solder joints using a multiscale fracture model'. Together they form a unique fingerprint.

Cite this