A Systems Approach to the Service Life Prediction Problem for Coating Systems

The conventional and reliability-based service life methodologies for coating systems are compared with respect to their predictive abilities. It was concluded from this comparison that the scientific merit of several of the underlying premises of the conventional methodology are suspect. Specifically, the premise which draws into question the merit of the conventional methodology is the belief that the weather repeats itself over some time scale. Unlike the conventional methodology, the reliability-based methodology has a strong scientific basis and has had an outstanding record in predicting the service life of a wide variety of materials, components, and systems. The application of this methodology of coating systems, however, will require dramatic changes in the way that the industry views its service life prediction problem. Specifically, major changes will be required in 1) the missions and objectives assigned to the primary sources of service life data; 2) the characterization of the unaged coating system and coating constituents; 3) the characterization of the exposure environment and coating system degradation: and 4) the collection, analysis, storage, and retrieval of experimental data. Application and implementation of the reliability-based methodology to the coating service life prediction problem is discussed.