Model-based structural health monitoring of naval ship hulls

The present paper reports on results from an ongoing research program at Cornell University aimed at employing model-based structural health monitoring techniques within new and existing naval hull structures. The techniques discussed involve the solution of inverse problems, formulated using both optimization-based and Bayesian approaches. The forward modeling capability is handled using a specially developed hull structural analysis tool, CU-BEN, while the solution of the inverse problem is handled using stochastic search methods that are part of a dedicated inverse solution algorithm “toolbox,” CU-PSST. Results from the application of these tools to problems of detecting section loss in hull plating due to corrosion, and isolating damaged framing due to an internal blast, are discussed.