Thesis (M.Eng.Sc.)--University of Adelaide, School of Mechanical Engineering, 2005. / Includes bibliographical references (leaves 131-137). Also available electronically as part of the Australian Digital Theses Program.

Thesis (M.Eng.Sc.)--University of Adelaide, School of Mechanical Engineering, 2005. / Title from screen page; viewed 16 Aug. 2005. Includes bibliographical references (leaves 131-137). Also available in print format.

Thesis (M.Eng.)--University of Wollongong, 2006. / Typescript. Includes bibliographical references: leaf 101-106.

Lightweight designs in engineering applications give rise to flexible structures with extremely low internal damping. Vibrations of these flexible structures due to an unwanted excitation of system resonances may lead to high cycle fatigue failure and noise propagation. A common method to suppress the vibrations is to increase the damping of the system using one of the classical control techniques i.e. passive, active, and/or hybrid. Passive techniques are those control systems that are simply integrated into the structures with no need of external power source for their operations, like viscoelastic damping, piezoelectric and electromagnetic shunt damping, tuned mass damper, etc. However, the control performance of these systems, in terms of the damping ratio and the robustness to uncertainties, is highly limited to the system properties. For example, viscoelastic damping may not perform well at low frequencies and the performance of shunt damping is dependent on the electromechanical coupling between the structure and the transducer. To overcome the limitations associated with passive controls, it has been proposed to use active control systems, which are less sensitive to the system's parameters, to improve the control performance. It requires an integration of sensors and actuators with a feedback loop containing control laws. However, the high requirement of the external power source is not favorable for engineering applications where energy efficiency is the key parameter. The combination of active and passive strategies, known as hybrid control systems, can provide a fail-safe configuration with a high control performance and low power consumption. The price to pay for such configurations is the complexity of the design. This doctoral thesis first investigates the conceptual designs of all kinds of classical control systems for a simplified mechanical system. They include 1) the passive shunt using an electromagnetic transducer, 2) the active control system using positive and negative feedback, and 3) the hybrid electromagnetic shunt damper using both an active voltage source as well as an active current source. The next part of this thesis is focused on bladed structures as real-life applications which highly require vibration control due to their low internal damping. Because of practical reasons, piezoelectric transducers are used for the application of control systems. The finite element model of the structure is made first without piezoelectric patches to optimize the best locations of piezoelectric patches. Then, the model is updated with the piezoelectric patches to numerically simulate different control strategies. The experiments are performed to validate the numerical designs. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished

We tackle the problem of identifying the structural dynamics of the twin tails of the F-15 fighter plane. The objective is to first investigate and identify the different possible attractors that coexist for the same operating parameters. Second is to develop a model that simulates the experimentally determined dynamics. Third is to suppress the high-amplitude vibrations of the tails due to either principal parametric or external excitations. To understand the dynamical characteristics of the twin-tails, the model is excited parametrically. For the same excitation amplitude and frequency, five different responses are observed depending on the initial conditions. The coexisting five responses are the result of the nonlinearities. After the experimental identification of the system, we develop a model to capture the dynamics realized in the experiment. We devise a nonlinear control law based on cubic velocity feedback to suppress the response of the model to a principal parametric excitation. The performance of the control law is studied by comparing the open- and closed-loop responses of the system. Furthermore, we conduct experiments to verify the theoretical analysis. The theoretical and experimental findings indicate that the control law not only leads to effective vibration suppression, but also to effective bifurcation control. We investigate the design of a neural-network-based adaptive control system for active vibration suppression of the model when subjected to a parametric excitation. First, an emulator neural network was trained to represent the structure and thus used to predict the future responses of the model. Second, a neurocontroller is developed to determine the necessary control action. The computer-simulation studies show great promise for artificial neural networks to control the model vibrations caused by parametric excitations. We investigate the use of four different control strategies to suppress high-amplitude responses of the F-15 fighter to a primary resonance excitation. The control strategies are linear velocity feedback, nonlinear velocity feedback, positive position feedback, and saturation-based control. For each case, we conduct bifurcation analyses for the open- and closed-loop responses of the system and investigate theoretically the performance of the different control strategies. We also calculate the instantaneous power requirements of each control law. The experimental results agree with the theoretical findings. / Ph. D.

We tackle the problem of suppressing high-amplitude vibrations of cantilever beams when subjected to either primary external or principal parametric resonances. Guided by results of previous investigations into the nonlinear dynamics of single- and multi-degree-of-freedom structures, we design mechatronic systems of sensors, actuators, and electronic devices and implement nonlinear active feedback control. In the case of external excitation, we devise two vibration absorbers based on either quadratic or cubic feedback. We conduct theoretical analyses and demonstrate that when a two-to-one (one-to-one) internal resonance condition is imposed between the plant and the quadratic (cubic) absorber, there exists a saturation phenomenon. When the plant is forced near its resonant frequency and the forcing amplitude exceeds a certain small threshold, the nonlinear coupling creates an energy-transfer mechanism that limits (saturates) the response of the plant. Our theoretical studies reveal that the cubic absorber creates regimes of high-amplitude quasiperiodic and chaotic responses, thereby limiting its utility. However, we show that superior results can be achieved when the natural frequency of the quadratic absorber is set equal to one-half the excitation frequency. Consequently, we apply the quadratic technique through a variety of linear and nonlinear actuators, sensors, and electronic devices. We design and build second-order analog circuits that emulate the quadratic absorber. Using a DC motor, piezoelectric ceramics, and Terfenol-D struts as actuators and potentiometers, strain gages, and accelerometers as sensors, we demonstrate successful single- and multi-mode vibration control. In order to realize a more versatile implementation of the control strategy, we resort to a digital signal processing (DSP) board. We compose a code in C and design a digital absorber by developing algorithms that, in addition to replacing the analog circuit, automatically detect the amplitude and frequency of oscillation of the plant and fine-tune the absorber parameters. We take advantage of the digital realization, implement a linear absorber, and compare the performance of the quadratic absorber with that of its linear counterpart. In the case of parametric excitation, we investigate two techniques. First, we explore application of the quadratic absorber. We prove theoretically and demonstrate experimentally that this control scheme is not reliable. Then, we propose an alternate approach. We devise a control law based on cubic velocity feedback. We conduct theoretical and experimental investigations and show that the latter strategy leads to effective vibration suppression and bifurcation control. / Ph. D.

In this thesis we use a new numerical code, based upon a mixed FEM-Runge-Kutta method, for the analysis and the design of plane 2-dimensional smart structures. We applied the developed code to the study of arbitrarily shaped piezo-electromechanical (PEM) plates. This code is based on a weak formulation of their governing equations as found in [18]. The optimal parameters needed to synthesize the appropriate electric networks are computed, and the overall performances of such plates are investigated. In particular, two examples are studied: firstly, a simple case is used to test the main features of the code; secondly, a more complex PEM plate is designed and analyzed by means of the proposed numerical approach. / Master of Science

This research reviews and extends the impedance-based structural health monitoring technique in order to detect and identify structural damage on various complex structures. The basic principle behind this technique is to apply high frequency structural excitations (typically higher than 30 kHz) through the surface-bonded piezoelectric transducers, and measure the impedance of structures by monitoring the current and voltage applied to the transducers. Changes in impedance indicate changes in the structure, which in turn can indicate that damage has occurred. Several case studies, including a pipeline structure, a composite reinforced aluminum plate, a precision part (gear), a quarter-scale bridge section, and a steel pipe header, demonstrate how this technique can be used to detect damage in real-time. A method to process impedance measurements to prevent significant temperature and boundary condition changes registering as damage has been developed and implemented. Furthermore, the feasibility of using the technique for high temperature structures and for condition monitoring of critical facilities subjected to a severe natural disaster has been investigated. While the impedance-based structural health monitoring technique indicates qualitatively that damage has occurred, more information on the nature of damage is necessary for remote structures. In this research, two different damage identification schemes have been combined with the impedance method in order to quantitatively assess the state of structures. One is based on a wave propagation modeling, and the other is the use of artificial neural networks. A newly developed wave propagation model has been developed and combined with the impedance method in order to estimate the severity of damage. Numerical and experimental investigations on 1-dimensional structures were presented to illustrate the effectiveness of the combined approach. Furthermore, to avoid the complexity introduced by conventional computational methods in high frequency ranges, multiple sets of artificial neural networks were integrated with the impedance-based health monitoring technique. By incorporating neural network features, the technique is able to detect damage in its early stage and to determine the severity of damage without prior knowledge of the model of structures. The dissertation concludes with experimental examples, investigations on a quarter-scale steel bridge section and a space truss structure, in order to verify the performance of the proposed methodology. / Ph. D.

A robust nonlinear observer, utilizing the sliding mode concept, is developed for the dynamic positioning of ships. The observer provides the estimates of linear velocities of the ship and bias from the slowly varying environmental loads. It also filters out wave frequency motion to avoid wear of actuators and excessive fuel consumption. Especially, the observer structure with a saturation function makes the proposed observer robust against neglected nonlinearties, disturbances and uncertainties. A direct adaptive neural network controller is developed for a model of an underwater vehicle. Radial basis neural network and multilayer neural network are used in the closed-loop to approximate the nonlinear vehicle dynamics. No prior off-line training phase and no explicit knowledge of the structure of the plant are required, and this scheme exploits the advantages of both neural network control and adaptive control. A control law and a stable on-line adaptive law are derived using the Lyapunov theory, and the convergence of the tracking error to zero and the boundedness of signals are guaranteed. Comparison of the results with different neural network architectures is made, and performance of the controller is demonstrated by computer simulations. The sliding mode observer is used to eliminate observation spillovers in the vibration control of flexible structures. It is common to build a state feedback controller and a state estimator based on the mathematical model of the system with a finite number of vibration modes, but this may cause control and observation spillover due to the residual (uncontrolled) modes. The performance of a sliding mode observer is compared with that of a conventional Kalman filter in order to demonstrate robustness and disturbance decoupling characteristics. Simulation and experimental results using the sliding mode observer are presented for the active vibration control of a cantilever beam using smart materials. / Ph. D.

Finite element analysis of laminated shells using a three-noded flat triangular shell element is presented. The flat shell element is obtained by combining the Discrete Kirchhoff Theory (DKT) plate bending element and a membrane element similar to the Allman element, but derived from the Linear Strain Triangular (LST) element. Though this combination has been employed in the literature for linear static analysis of laminated plates, the results presented are not adequate to ascertain that the element would perform well in the case of static and dynamic analysis of general shells. The element is first thoroughly tested for linear static analysis of laminated plates and shells and is extended for free vibration, thermal, and geometrically nonlinear analysis. The major drawback of the DKT plate bending element is that the transverse displacement is not explicitly defined within the interior of the element. Hence obtaining the consistent mass matrix or the derivatives of the transverse displacement that are required for forming the geometric stiffness matrix is not straight forward. This problem is alleviated by borrowing shape functions from other similar elements or using simple displacement fields. In the present research, free vibration analysis is performed both by using a lumped mass matrix and a so called consistent mass matrix, obtained by borrowing shape functions from an existing element, in order to compare the performance of the two methods. The geometrically nonlinear analysis is performed using an updated Lagrangian formulation employing Green strain and Second Piola-Kirchhoff (PK2) stress measures. A linear displacement field is used for the transverse displacement in order to compute the derivatives of the transverse displacement that are required to compute the geometric stiffness or the initial stress matrix. Several numerical examples are solved to demonstrate the accuracy of the formulation for both small and large rotation analysis of laminated plates and shells. The results are compared with those available in the existing literature and those obtained using the commercial finite element package ABAQUS and are found to be in good agreement. The element is employed for two main applications involving large flexible structures. The first application is the control of thermal deformations of a spherical mirror segment, which is a segment of a multi-segmented primary mirror used in a space telescope. The feasibility of controlling the surface distortions of the mirror segment due to arbitrary thermal fields, using discrete and distributed actuators, is studied. This kind of study was required for the design of a multi-segmented primary mirror of a next generation space telescope. The second application is the analysis of an inflatable structure, being considered by the US Army for housing vehicles and personnel. The tent structure is made up of membranes supported by arches stiffened by internal pressure. The updated Lagrangian formulation of the flat shell element has been developed primarily for the nonlinear analysis of the tent structure, since such a structure is expected to undergo large deformations and rotations under the action of environmental loads like the wind and snow loads. The wind load is modeled as a nonuniform pressure load and the snow load as lumped concentrated loads. Since the direction of the pressure load is assumed to be normal to the current configuration of the structure, it changes as the structure undergoes deformation. This is called the follower action. As a result, the pressure load is a function of the displacements and hence contributes to the tangent stiffness matrix in the case of geometrically nonlinear analysis. The thermal load also contributes to the system tangent stiffness matrix. In the case of the thermal load this contribution is similar to the initial stress matrix and hence no additional effort is required to compute this contribution. In the case of the pressure load, this contribution (called the pressure stiffness) is in general unsymmetric but can be systematically derived from the principle of virtual work. The follower effects of the pressure load have been included in the updated Lagrangian formulation of the flat shell element and have been validated using standard examples in the literature involving deformation-dependent pressure loads. The element can be used to obtain the nonlinear response of the tent structure under wind and snow loads. / Ph. D.