Prediction of Corrosion Protection System Aging by Theory Based Data Mining

Corrosion damage to aging aircraft is the highest maintenance cost for the U.S. Air Force, and decreases the time that aircraft are available for missions. In order to continue using the aircraft beyond their design life, while staying within the budgetconstraints of the U.S. Air Force, maintenance costs must be decreased. This requires the capability to predict both degradation of the corrosion protection systems, as well as subsequent corrosion damage to aircraft structures, so that early action can betaken to avoid damage. This Phase I SBIR program address the critical need for prediction of corrosion protection system aging and corrosion damage. The proposed approach is to enhance existing theoretical models for protection system breakdown andcorrosion mechanisms, with sophisticated data mining processes to validate, refine or change those models. In Phase I, we will demonstrate the power of data mining to model one aircraft component and its associated corrosion protection system, specificallya coated lap splice joint, and validate the models with limited experimental data. In Phase II, we would extend this approach to a wide range of protection systems and structures on aircraft, and develop broad models based on a combination of theory anddata mining.The anticipated results of the Phase I & II efforts are the development and commercialization of a model that predicts the lifetime and performance of a corrosion protection system for aging aircraft components, and the onset and propagation ofcorrosion resulting from coating degradation, for both military and commercial applications.