Reliability estimation for large-area solder joints using explicit modeling of damage

Research output: Contribution to journalArticlepeer-review

20 Scopus citations


The issues of available cyclic fatigue models in life prediction of large-area solder joints using finite-element analysis (FEA) are discussed. In this paper, a new FEA approach called successive initiation (SI) is modified and introduced in conjunction with energy partitioning (E-P) damage model to resolve some of the issues with available damage models such as geometry and scale dependency and provide a solution to large-area solder joints. This new technique models damage explicitly, meaning that it separates initiation from propagation by monitoring the plastic and creep damage at the tip of the crack successively. The SI technique could be easily used with continuous loadings of different types and frequencies. The modeling approach is then implemented on a power device with large-area solder. Sensitivity study is conducted with the help of the experiment to determine the right initiation threshold for smooth crack initiation and propagation. The results of modeling are then compared with available experimental data for the same power device. The comparison shows that using the damage model constants generated for small solder joints such as ball-grid array or chip-scale package could significantly overpredict the life of larger area solders. New E-P damage model constants for large-area solder joints are obtained and presented by calibrating the modeling to the experiment.

Original languageEnglish (US)
Article number4472168
Pages (from-to)375-386
Number of pages12
JournalIEEE Transactions on Device and Materials Reliability
Issue number2
StatePublished - Jun 2008
Externally publishedYes

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Safety, Risk, Reliability and Quality
  • Electrical and Electronic Engineering


Dive into the research topics of 'Reliability estimation for large-area solder joints using explicit modeling of damage'. Together they form a unique fingerprint.

Cite this