Application of load-dependent Ritz vectors to Bayesian probabilistic damage detection

Abstract

This paper demonstrates the possibility of incorporating load-dependent Ritz vectors as an alternative to modal parameters into a Bayesian probabilistic framework for detecting damages in a structure. Recent research has shown that it is possible to extract load-dependent Ritz vectors from vibration tests. This paper shows that load-dependent Ritz vectors have the following potential advantages for damage detection over modal vectors: (1) in general, load-dependent Ritz vectors are more sensitive to damage than the corresponding modal vectors; and (2) substructures of interest can be made more observable using the load-dependent Ritz vectors generated from particular load patterns. An eight-bay truss example and a five-story frame example, explicitly considering both modeling error and measurement noise, are presented to illustrate the applicability of the proposed approach.