Prognostics Health Management and Damage Relationships of Lead-Free Components in Thermal Cycling Harsh Environments

Abstract

There has been a growing interest in the area of health monitoring of electronics in order to predict the failure and provide warnings. Existing Prognostics Health Management methodologies are very reactive in nature and unable to estimate the residual life. The presented methodology helps in evaluating the reliability of the component in its actual life cycle environment. Also, the developed techniques are applied to the pre-failure space of the electronics system, in which no macro-indicator of failure such as crack or delamination exists. In this thesis, the prognostics health management and damage relationships for lead-free devices under thermal cycling harsh environments have been studied. The presented methodology eliminates the need for estimation of prior-damage. The procedures have been developed using leading indicators of failure like solder micro-structural coarsening. Failure data has been collected for 95.5Sn4.0Ag0.5Cu lead-free solder alloy joints in area array components mounted on FR-4 printed circuit boards. A technique for the determination of prior damage history has been presented using non-linear least squares based interrogation techniques. The presented method uses the Levenberg-Marquardt Algorithm. The utilized test vehicles include various area-array packaging architectures which are subjected to cyclic thermo-mechanical stresses over the range of -40º C to 125 º C. In addition, the developed procedure has been further extended to various kinds of lead-free alloys like SAC105, SAC305 and SAC307 which are subjected to thermal cyclic load from -55ºC to 125ºC. Damage proxies have been investigated and correlated to solder-interconnect damage. A mathematical relationship has been developed to compute the residual life