Real-time structural monitoring and damage detection by acceleration and

The objective in real-time structural monitoring is to track the changes in the dynamic characteristics of the structure and to detect and locate damage after an extreme event. Damage detection typically involves analysis of acceleration data to search for changes in the structure’s natural frequencies. Data from instrumented structures have clearly shown that changes in natural frequencies are not always a reliable indicator of damage because various environmental factors can change frequencies without damage. Similarly, the inter-story drift, if calculated from acceleration records, is also not a reliable damage indicator because the errors generated by the noise in the records are exponentially amplified during the calculations. This paper presents new approaches to structural monitoring, damage detection, and data analysis. We propose that damage can be detected more accurately if we investigate, not only the changes in natural frequencies, but also the changes in the propagation characteristics of seismic waves in the structure and any permanent changes in the structure’s configuration. Since permanent deformations cannot be calculated accurately from acceleration data, we propose using GPS sensors. A 17-story, moment-resisting steel frame building at the UCLA campus in Los Angeles, California is currently being used to test these new concepts. The building is being monitored in real time with a very dense downhole-surface-structural array of accelerometers, plus three GPS sensors on the roof. We are developing adaptive signal processing and system identification tools to process and analyze the data in real time, and statistical decision-making tools to identify the changes in the building due to real damage from those due to environmental factors