Passive and semi-active damping of base-excited structures
It has become commonplace for discrete passive or active elements to be added to Structures to mitigate against vibration. More recently, semi-active damping has been a focus of research. Semi- active systems are attractive due to their performance, low cost, power consumption and control stability....
| Main Author: | |
|---|---|
| Published: |
University of Bristol
2011
|
| Subjects: | |
| Online Access: | http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.618724 |
| Summary: | It has become commonplace for discrete passive or active elements to be added to Structures to mitigate against vibration. More recently, semi-active damping has been a focus of research. Semi- active systems are attractive due to their performance, low cost, power consumption and control stability. In this thesis we consider how passive and semi-active damping systems may he designed to mitigate against. vibration in base-excited structures. initially we consider the base isolation of a single degree-of-freedom system. The optimality of the common sky-hook switching controller is assessed and its control form is generalised. Numerical and theoretical studies identify the optimal form of switching control and a linear switching surface controller is proposed which is found to achieve performance close to optimal. We propose the new design methodology of quasi-active damping . The motivation for this method is to approach the levels of performance obtainable using fully active systems whilst retaining the desirable attribute of semi- active systems, A quasiactive suspension design is proposed and validated in simulation. It is found that this system call produce a quasi-active region in the frequency response of very low displacement transmissibility. This study is then extended to consider how quasi-active systems should be designed specifically for use with magnetorheological dampers. The influence of external passive damping on 2:1 internal resonance in stay cables is examined. An efficient, low-order damped nonlinear cable model is derived and used to numerically generate stability boundaries. These are used to assess how damping should be designed to mitigate against this excitation mechanism. We then study the optimality of dipped-optimal LQR control in the semi-active clamping of cables. We propose the addition of all extra term to the cost function to encourage the control output. To satisfy the semi-active constraint. Finally we study the feasibility of real-time dynamic substructuring as a technique for the experimental testing of damped stay cables. |
|---|