Friction-based control of building complex for seismic mitigation : numerical and experimental studies

Abstract

This thesis mainly focuses on developing coupled control for the seismic protection of building complex, which is a grouped building of main and podium structures, and understanding the control performance under passive and semi-active control strategies using friction dampers. Numerical and experimental methods are employed to identify the effectiveness in this research. The control effectiveness of passive coupled control for the building complex is first investigated experimentally. To this end, an experimental model of a building complex and laboratory-scale friction damper are designed and manufactured. Cyclic tests of the friction damper are also performed to identify its dynamic performance. Seismic behaviors of the building complex in three different coupling configurations are examined under several historical earthquake records. The coupling configurations considered include: uncoupled, rigid-coupled, and damper-coupled. In the rigid-coupled case, which is the current construction scheme, it is physically identified that exaggerated vibrations of interstory drifts and absolute accelerations are associated with the main building when it is uncoupled to the podium structure. The responses of both buildings are, in contrast, moderately reduced in the damper-coupled case while acceleration reduction is relatively sensitive to the design level of friction force. To study the potential effectiveness of passive coupled control with consideration of building configuration and damper deployment, numerical studies were then undertaken. A mathematical model of the coupled building system with passive friction dampers as joint control devices is first established, and the adequacy of the model was verified by comparison with the experimental results. The parameters that influence the design of optimal friction force and control efficiency include type of earthquake excitation, the mass and height difference between the structures, as well as the damper implementation. In addition to the parametric study of the damper-coupled system, the exaggerated seismic responses of the rigid-coupled building system that were observed in the experiment are analytically investigated using modal analysis to understand the underlying reasons. The study is continued by introducing semi-active control to pursue increased vibration reduction of the building complex. There are two different classes of control algorithms considered in the numerical study, which are global- and local-feedback controllers. The significance of investigating local-feedback control is that the requirement on sensing instrumentation in terms of quantity is less and is therefore practically appealing. To implement global-feedback control which employs LQG control with acceleration feedback, a modified clipped algorithm is developed for the variable friction damper to realize semi-active force based on the command active force. Three local feedback controllers are examined, one of which is developed from existing controllers to allow effective design of maximum damper capacity. Parametric studies for the coupled building system covering the parameters of height and mass differences and damper deployment are conducted to provide comprehensive evaluation of control performance. Experimental validation of the numerical findings regarding the semi-active control of building complex is also carried out. To establish a closed-loop control of the complex building, substantial works focus on the development of a laboratory-scale variable friction damper. A new leverage-based piezoelectric actuator is employed to provide higher sensitive adjustment of the friction force. Key features of the actuator and variable friction damper are both identified through a series of characterization tests. Based on the identified results, an operating scheme for real-time control of the damper is proposed. Effectiveness of the developed operating scheme is also assessed by forming a closed-loop system. A force-feedback controller is accordingly developed to improve precision of friction force replication of the command force. The characterized variable friction damper is subsequently implemented on the test structures. Experimental validation of the semi-active control strategy is evaluated through a series of shaking table tests. Both local- and global-feedback control algorithms are successfully realized, providing semi-active control of the structures and demonstrating consistent findings observed in the numerical study. This thesis finally investigates the design details of a variable friction that is valuable for the prototype design of damper. Particular attention is placed on evaluating the dynamic performance and force regulation ability of the damper configured with multiple friction surfaces and force amplifying unit to provide larger friction force. Two variable friction dampers are accordingly designed and manufactured to respectively attain these two mechanical configurations. By carrying out cyclic tests to the dampers under constant and varying input voltage, it is demonstrated that increasing friction surfaces is an effective approach to increase friction force of the damper. Most importantly, all the potential characteristics identified in the laboratory-scale damper are retained regardless of the friction surface size and thus the developed operating scheme is valid. This generally confirms the potential application of the variable friction damper for civil structures.