MODELLING AND CONTROL OF SEMI-ACTIVE FLUID VISCOUS DAMPERS IN STRUCTURAL VIBRATION OPTIMIZATION

In this original final project, called “Modelling and control of semi-active fluid viscous dampers in structural vibration optimization” the behaviour and usefulness of a device referred to as “variable orifice fluid viscous damper” is analysed in the context of improving the dynamic response of structures. Nowadays, in the civil engineering field, the presence of strategies to act on the behaviour of structures subjected to external inputs such as earthquakes or wind gusts is more and more often present. These strategies in literature are divided into three main approaches: passive control, semi-active control, and active control. Passive control of structures consists in exploiting devices that are fixated on a certain behaviour that cannot be modified in real-time it can be also considered a redesign of the structure parameters. Active control of structures, instead, exploits solutions that are capable of dynamically modifying the characteristics of the structure, requiring a relevant amount of energy to the actuators (which is particularly critical during an exceptional event such as an earthquake) that must be perfectly controlled to avoid the destabilization of the structure. Semi-active control of structures represents the compromise between the two previous approaches, it exploits devices that are capable of modifying their mechanical parameters in real-time, that do not need a relevant amount of power to work, and also are not capable of destabilizing a structure because they do not possess actuators. In this context, the specific “variable orifice fluid viscous damper” device was considered and studied starting from a literature survey that was fundamental to obtaining the necessary information for building a model that could integrate the device to analyse its effect. In the MATLAB/Simulink environment, the equation of motion of the systems was linearized and put into state-space form to create an LTI system for the simulations. After this step two main control algorithms were considered, the PID and the MPC controller. Many simulations were carried out considering various tuning of the controllers, different levels of approximation of the model, and also different sampling times, which is a fundamental parameter for the computational weight of the controller. After this step, the sensors, filters, and delay of the real components were added to the model to decrease even the approximation and move closer to a hypothetical real application of the device in a specific structure.