CALENDER

An algorithm for on-line detection of damage to structures caused by ground shaking is presented. Real-time sampled response data are processed by templates in the form of ID (for structural identification) neural networks, which differentiate damage according to similarity of the response to those encapsulated in the templates. Numerical examples based on a simple 2-story steel-frame building are used to illustrate the proceedings and to underscore the limitations of the method. The challenges of on-line damage detection are discussed in detail to promote better understanding of how the proposed algorithm has evolved and, in particular, why neural networks are used. Widespread application of the algorithm, and damage detection in general, depends on the establishment of an adequate ID networks database, a far more daunting task in practice than in the theoretical setting of the paper.