Health monitoring of structures under ambient vibrations using semiactive devices

Structural health monitoring (SHM) is the process of monitoring structural health and identifying damage existence, severity and location. Clear needs for SHM exist for various types of civil structures. Yet, the dominant method for monitoring the health of civil structures is manual visual inspection. Global vibration-based SHM techniques have been studied, but no approach has been well established and accepted due to the limitations of ambient excitation for most civil structures and the small sensitivity of global vibration characteristics to damage. One approach that may alleviate some of the SHM difficulties for civil structures is using variable stiffness and damping devices (VSDDs) to improve damage estimates. In addition to providing near optimal structural control strategies for vibration mitigation, these low-power and fail-safe devices can provide parametric changes to increase global vibration measurement sensitivity to damage.; This dissertation proposes using VSDDs in structures to improve SHM, and demonstrates the benefits analytically and experimentally in contrast with conventional passive structures. A two degree-of-freedom (2DOF) bridge structure model and two shear building models, are used as test beds to study the VSDD approach with several identification algorithms. Using multiple channels of data from multiple VSDD configurations, a least-squares error formulation is used to estimate unknown structural parameters. The improvements in identification are even more effective when adding higher effective levels of stiffness or damping to a structural system; the resulting VSDD forces are small due to low ambient excitation. A 2DOF shear building laboratory structure is excited at the base using a small shaking table and its parameters are identified; using VSDDs gives parameter estimates that have better means and smaller variations than the conventional structure approach. Finally, a controlled approach is introduced for SHM. A state feedback gain matrix is chosen to minimize a cost function of parameter identification error and control effort. This controlled approach is applied to the 2DOF bridge model where an ARMAX (Auto Regressive Moving Average eXogenous disturbance) model is assumed to represent the controlled system.