Modal parameter changes as function of drift in two shear wall buildings

Abstract

Seismic damage generally causes changes in the stiffness and energy dissipation characteristics of a building, which in turn affect its modal parameters and overall dynamics response It has also long been recognized that amplitude of the response, without damage, also produced changes in the apparent modal characteristics. Understanding and quantifying the modal variations are key for Seismic Structural Health Monitoring systems that want to distinguish between normal behavior and damage particularly if low-level damage detection is an objective. Two building structures that have been continuously monitoring and have registered together more than 1800 earthquake motions, most of them weak or medium but one event that generated maximum accelerations of 0.14 and 0.16 g at their base and up to 0.31 and 0.47 g on the structures causing damage. The database of events enables the identification of its modal properties and variations during both small, non-damaging events and a damaging event. Analysis shows that there is a clear amplitude dependent nonlinear response that can be captured by the variations of the periods as a function of building drift amplitude. These periods variation can be in the order of 20 % for a roof drift ratio of 0.1 % without any damage to these structures. A strong change in modal parameters can be observed in these structures after 0.1 % drift. After damage has occurred, the modal properties change as a function of drift with a different pattern when compared with the previous state of the structure. For the same drift, modal parameters in the damage state have a much larger mean (longer periods) and larger variation around the mean due to ambient conditions. The amplitude and sensitivity to drift of these changes are unique to each system and can be captured using non damaging event if a window base system identification technique is used. The amplitude variations that are not related to damage are determined and are later used on their respective autonomous structural health monitoring system.