Experimental Study on the Behaviour of a Magnetorheological (MR) Damper and Evaluation of Numerical Models
Abstract
MR dampers are smart dampers that improve structural performance through vibration control. Vibration of tall or flexible structures and structural elements induced by dynamic loads such as wind and earthquakes can cause severe damage to the structure and discomfort for individuals. Therefore, dampers are commonly installed to increase the energy dissipation capacity and mitigate large amplitude vibrations. MR dampers are gaining recognition because of the unique rheological properties offered by the magnetically sensitive MR fluid. Changes to the current level, along with the frequency and amplitude of vibration, affect the force-displacement and force-velocity hysteresis relationship and thus the amount of damping provided by a MR damper. One method of analyzing the hysteresis behaviour is conducting experimental studies that involves varying input parameters to analyze the output response. A limitation of experimental testing is the range of input parameters that can be tested, and the difficulty of controlling and reproducing the experiment. Therefore, numerical models are often used to analyze the output parameters of the hysteresis response of a MR damper. Due to the nonlinear hysteresis behaviour of MR dampers, one of the challenges is the selection of a suitable numerical model for analysis and design. The accuracy of a numerical model is also affected by its complexity level. However, complex numerical models can be computationally expensive requiring a significant amount of computational resource, and thus limiting their practical application. As a result, choosing an appropriate numerical model for MR damper analysis and design requires careful consideration. An experimental study has been conducted in the current thesis to investigate the behaviour of a MR damper and evaluate four existing numerical models i.e., the Bingham, Bouc-Wen, modified Bouc-Wen, and modified viscous Dahl models. Data obtained from the experimental study are used to identify the parametric behaviour and the most suitable model among those discussed in describing the MR damper behaviour. The experimental results and damper behaviour predicted by numerical models are compared through force-displacement and force-velocity hysteresis results, including the trends, the amount of dissipated energy and the equivalent linear viscous damping coefficient of the MR damper. The merits and shortcomings of the numerical models are discussed.