Structural Prognosis Using Intelligent Distributed Systems

Detection of the onset of microstructural damage (e.g. cracks, corrosion, and delamination or debonding) and the correlation of this damage with predicted failure is not currently available in any kind of on-board system. A system of structuralprognostics using neural nets based on dynamic data could result in sizeable cost savings as well as greatly enhanced operating safety of aging aircraft, but to our knowledge has not been attempted before. We propose to develop novel artificial neuralnetworks to process and learn the structural mechanics of aging aircraft, and from this training to be able to predict needs for maintenance based on the progression of detected damage.Aging military fleets are now being asked to serve for much longer thantheir original life expectancy, and this technology could enable safe and cost-effective service life extension. In addition, this technology can be applied to commercial and advanced military aircraft, as well as to a host of complex mechanical systemsfrom bridges to ships.