Michael Pecht and Patrick McCluskey
CALCE Electronic Products and Systems Consortium
University of Maryland
College Park, MD 20742
Significant improvements in the reliability of electronics can be achieved by retrieving fielded assemblies and examining them to determine the root cause of any failures and the extent of degradation which has occurred over time. These root cause failure analyses employ a number of non-destructive and destructive analytical techniques to identify the locations at which failures occur, characterize the modes in which failures occur, and provide clues as to the fundamental mechanical, electrical, chemical, and electrochemical mechanisms by which failures occur. This information highlights the weak links in the electronic system which can be modified to provide increased reliability. Because of the importance of failure analysis and degradation assessments in reliability growth, this chapter will focus primarily on describing the tools and techniques for systematically conducting such an assessment.
The identification of the damage mechanisms which are most likely to cause failure and the sites at which the damage is the most pronounced in fielded assemblies also permits the development of a focused accelerated test program. This focused accelerated test program can be used to qualify new assemblies, thereby eliminating those assemblies most likely to fail in the field, to determine the expected life of a new assembly, or to determine the remaining life of an assembly which has been stored or used for an extended period of time. The benefits of the focused accelerated test program are that it allows the proper test stresses (e.g., temperature, relative humidity, temperature cycling) and the levels of those stresses to be selected so as to cause failure by the observed mechanisms in the shortest time without altering the failure mechanism. This is preferable to choosing a random set of test loads and levels, or subjecting the assemblies to a set of "one size fits all" standard tests prescribed by decades-old military and commercial standards, which are often inaccurate, improperly applied, unnecessarily restrictive, and do not address the actual failure mechanisms occurring in the application environment. In addition, the failure distribution in the test can be converted to a failure distribution in the intended use environment using the acceleration factors calculated by the PoF models.
article is available to CALCE Consortium Members.