Experimental validation of optimal sensor placement algorithms for structural health monitoring

While most efforts in structural health monitoring have been focused on the development of appropriate sensor technologies and signal processing techniques, formal methods for how to design sensor configurations (number, type, and location) have received much less attention. This presentation will present a design methodology based upon the concept of observability from the fields of dynamic systems theory and engineering design optimization. The design technique uses a dynamical model (such as a finite-element model) of the structure as a basis for the development of measures of the ability to detect and localize damage while the optimization techniques provide a framework for studying design tradeoffs such as global coverage versus localized damage sensitivity versus sensor configuration, etc. Application of this approach to a prototypical wing spar structure is presented with experimental results.