A comparison of 200 kN magneto-rheological damper models for use in real-time hybridsimulation pretesting

Abstract

Control devices can be used to dissipate the energy of a civil structure subjected todynamic loading, such as earthquake, wave and wind excitation, thus reducing structuraldamage and preventing failure. The magneto-rheological (MR) fluid damper isa promising device for use in civil structures due to its mechanical simplicity,inherent stability, high dynamic range, large temperature operating range, robustperformance, and low power requirements. The MR damper is intrinsically nonlinearand rate dependent. Thus a challenging aspect of applying this technology is thedevelopment of accurate models to describe the behavior of such dampers forcontrol design and evaluation purposes. In particular, a new type of experimentaltesting called real-time hybrid simulation (RTHS) combines numerical simulationwith laboratory testing of physical components. As with any laboratory testing,safety is of critical importance. For RTHS in particular the feedback and dynamicinteraction of physical and numerical components can result in potentially unstablebehavior. For safety purposes, it is desired to conduct pretest simulations wherethe physical specimen is replaced with an appropriate numerical model yet thenumerical RTHS component is left unchanged. These pretest simulations require aMR damper model that can exhibit stability and convergence at larger fixedintegration time steps, and provide computational efficiency, speed of calculation, andaccuracy during pretest verification of the experimental setup. Several models for MRdampers have been proposed, including the hyperbolic tangent, Bouc–Wen, viscousplus Dahl and algebraic models. This paper examines the relative performance offour MR damper models of large-scale 200 kN MR dampers as needed for pretestsimulations of RTHS. Experimental tests are conducted on two large-scale MRdampers located at two RTHS test facilities at the Smart Structures TechnologyLaboratory at the University of Illinois at Urbana Champaign and the LehighUniversity Network for Earthquake Engineering Simulation facility. It is shown thateach of the MR damper models examined has relative merits and the ultimateselection of the particular model is dependent on the specific RTHS being tested.