Project Number: C96-05

Reliability Assessment Methodology for Electronic Systems

Point of Contact: Dr. Michael Pecht email: pecht@eng.umd.edu Phone: (301) 405-5323 Fax: (301) 314-9269

Objectives

Develop a reliability assessment methodology (framework) for the performance of electronic systems. The proposed methods should possess the following attributes:

emphasize integration of physical parameters describing the design attributes, environments and usage rates
focus on failure prevention by monitoring progress at all levels of the design and development process.
be robust but allow refinement of estimate by taking advantage of a variety of data sources from multifunctional stress analysis
be usable by large and small companies alike by not being entirely dependent on the level of available resources

Background

Reliability assessment of electronics has traditionally been based on empirical failure- rate models (eg. MIL-HDBK-217) developed largely from curve fits of field-failure rate data. These field-failure data are often limited in terms of the number of failures in a given field environment, and determination of the actual cause of failure. In MIL-HDBK-217, crucial failure details like failure site and mechanism, load/environment history, materials, and geometry are not collected leading to:

lack of insight or control over the actual cause of failure
inability to address design and usage parameters that most influence reliability

Thus there is a need for a credible alternative to MIL-HDBK-217 for the evaluation of electronic systems.

Work Accomplished

A review of the existing reliability prediction methods with a view to developing potential new approaches is performed. Existing empirical failure rate techniques were studied with a view to investigate alternates. In the process a seies of related studies were performed.

Developed a criterion in which to evaluate reliability models.
Reviewed existing reliability models, including the reporting metrics, subject to evaluation criteria.
Proposed potential reliability assessment method(s) with emphasis on acceptable metrics. Proposed transforms that allow use of available analysis and industry test data.
Reviewed critical failure mechanisms and associated key characteristics (KC's) relative to hardware exposure to mechanical, thermal, chemical, and electrical stresses at LRU level, LRM or PWA level, and component levels.
Established methodologies that allow reduction of the number of KC's to those defining the life of the assembly. Established methods to address the variability of the KC's and their effect on the service life of the hardware.
Established methods to address probabilistic nature of overall durability life predictions.
Developed appropriate reliability metrics that include durability life data.
Foundation for a reliability assessment methodology, which focuses on failure prevention by monitoring progress at all levels of the design and development.