<p>This work presents the analytic and experimental development of active vibration control of large flexible space structures (LFSS) using smart materials. Two basic configurations were studied: flexible manipulators, and truss structures, which encompass most of the flexible structures in space applications. The dynamics of LFSS are characterized by their high order and the significant presence of lightly damped, closely spaced low frequency modes. In space applications, space structures are required to perform precision trajectory tracking and attitude regulation, tasks that introduce disturbance torques and forces that may excite vibrational modes in the flexible parts of the structures, degrading their performance. To solve this problem, the use of piezoelectric materials coupled to structural members, to form smart structural members, is proposed for the implementation of active control techniques. For the flexible manipulator, the use of shaped piezoelectric sensors is presented. Piezoelectric sensors can be shaped to provide state feedback, which can be used as part of a control law to compensate vibrations induced by flexible degrees of freedom. A robust H∞ state feedback control law is obtained. The control law is implemented using the shaped sensor to stabilize the flexible manipulator. Simulation and experimental results in a single link flexible manipulator confirm the effectiveness of the proposed approach. The design of a truss structure and the control for active damping of vibration is presented, taking advantage of the use of piezoelectric actuators and sensors. Three control techniques are tested: negative velocity feedback, LQG and H∞ control. Simulations and experiments are performed on the closed loop to assess the relative merits of each control technique. Results show that the controllers increase the damping of the structure noticeably. The robust H∞ controller provides the better performance of the control techniques presented, even in the presence of higher order modes and parametric uncertainties not accounted for in the control design process.</p> / Doctor of Philosophy (PhD)

<p>The flow of non-Newtonian fluids in both concentric and eccentric annuli was investigated in this thesis. The model for generalized Bingham (Herschel-Bulkley) fluids was used in the studies, which included fully developed flow, entrance flow, start-up flow and pulsating flow in a concentric annulus and start-up flow in an eccentric annulus. A set of mathematical formulations has been developed for fully developed flow of generalized Bingham fluids in a concentric annulus. Velocity profiles are presented by using a numerical scheme to solve the equations. The position of the unsheared plug in the annulus may be determined by the solutions. The equations of motion for entrance flow and unsteady flow of generalized Bingham fluids in a concentric annulus have been derived with a group of dimensionless variables. A control volume based finite difference technique was used to solve the governing equations. The effects of generalized Bingham number Pl, flow index n and radius ratio s on velocity profiles and pressure drop in the annulus are presented. Velocity profiles of start-up flow of generalized Bingham fluids in an eccentric annulus were obtained from finite difference solutions of the equation of motion after transformation into bipolar coordinates. The effects of eccentricity were also considered.</p> / Doctor of Philosophy (PhD)

<p>This thesis is devoted to investigating adaptive neurocontrol of nonlinear systems with uncertain or unknown dynamic models. Novel theoretical synthesis and analysis of neurocontrol systems have been conducted, and applied to the control of flexible joint robots with experimental tests. The contributions of this thesis fall into the following three areas: (1) neural networks, (2) adaptive neurocontrol and (3) control of flexible joint robots.</p> <p>The aim of my research in the neural network area is to search for fast and global convergent learning algorithms with reduced computation burden. The localized neural networks with competitive lateral inhibitory cells were introduced. The developed extended Kalman filtering algorithm with UD factorization can make the localized polynomial networks and localized pi-sigma networks possess fast learning convergence and less computation. The multi-step localized adaptive learning algorithm was derived for RBF networks which leads to about 10 fold improvement in the speed of learning convergence. New neural network models of nonlinear systems were introduced to facilitate neurocontroller design.</p> <p>In the adaptive neurocontrol area, theoretical issues of the existing backprop-based adaptive neurocontrol schemes were first clarified. Then new direct and indirect adaptive neurocontrol schemes, with better performance, were proposed. It is noticed that the system stability of many existing neurocontrol schemes cannot be proved. In addition, few stability-based adaptive neurocontrol schemes are available and can only be applied to feedback linearizable nonlinear systems. The thesis provides two major contributions to the stability-based adaptive neurocontrol approach. The first contribution is extending the classical self-tuning control methodologies for linear systems to the self-tuning neurocontrol of nonlinear systems by using localized neural networks. This extension greatly enriches the neurocontrol algorithms with guaranteed system stability. The second contribution is proposing the variable index control approach, which is of great significance in the control field, and applying it to derive new stable robust adaptive neurocontrol schemes. Those new schemes possess inherent robustness to system model uncertainty, which is not required to satisfy any matching condition. They do not impose any growth condition and infinite differentiability assumption on the system nonlinearity. They can also be applied to nonlinear systems which are not feedback-linearizable.</p> <p>As applications and extensions of the above theory, three different robust adaptive neurocontrol schemes for general flexible joint robots were derived with proven system stability. All three schemes are able to incorporate a priori information about the robot dynamics into the neurocontroller design to simplify the neural network design. No acceleration and jerk signals are required in these control laws. Moreover, arbitrary joint stiffness is allowed in the control algorithms.</p> <p>To demonstrate the feasibility of the proposed learning algorithms and adaptive neurocontrol schemes, intensive computer simulations were conducted based on different nonlinear systems and functions. Different types of adaptive tracking problems and regulation problems were considered. Furthermore, the proposed adaptive neurocontrol schemes were experimentally tested using an existing experimental flexible joint robot. Both the simulation and experimental results confirm the practicability of the proposed schemes.</p> <p>The thesis concludes that the neurocontrol approach, along with the development of neural computers and large scale parallel distributed processors, is capable of solving the complex control problem of nonlinear systems with uncertain or unknown dynamic models.</p> / Doctor of Philosophy (PhD)

<p>A corotational finite element formulation for flexible multibody systems which are subjected to frictional impact is developed in this thesis. The formulation can predict the motion of the system, the contact forces, the velocities, the accelerations, the duration of impact and the associated deformations.</p> <p>First, a corotational finite element formulation is developed for the dynamic analysis of flexible multibody systems without impact. A numerical algorithm is developed along the lines of the incremental-iterative method of the Newmark direct integration and Newton-Raphson methods.</p> <p>Frictional impact is then included in the formulation. The prediction of contact establishment and separation is achieved using an event predictor. Point impact is assumed and Coulomb's friction law is used to model the friction forces.</p> <p>Two multibody-oriented approaches are used to model the frictional impact. The first approach is based on a modified momentum balance model. An energy-based method is developed to resolve the problem of energy mismatch which arises with the use of Newton's impact law or Poisson's hypothesis. The concept of the coefficient of restitution is used and a new technique is developed to calculate the contact forces in some special cases. In general, it is assumed that multiple impulses occur during the contact period. An automatic time stepping algorithm is developed for numerical solutions.</p> <p>The second approach is based on the Lagrange multiplier method. The model exactly satisfies the geometric compatibility conditions during contact. It also allows the direct evaluation of the contact forces. Both sliding and sticking modes are considered. The proposed scheme overcomes the problem of high dimensionality in the traditional Lagrange multiplier models in structural dynamics.</p> <p>The applicability and accuracy of the formulation and the numerical technique are demonstrated. Simulation of various mechanical systems, which are subjected to impact loads, are presented.</p> / Doctor of Philosophy (PhD)

<p>Modern tube and shell heat exchangers designed with the existing standards may be susceptible to damage as a result of excessive tube vibrations caused by the shell-side flow. The present study was undertaken to further our understanding of the liquid cross-flow-induced vibrations of a tube array of contemporary interest.</p> <p>An experimental facility to study these phenomena was designed and built. The experiments were conducted using this specially designed water tunnel having a 305 x 305 mm working section. The tube bundle was a parallel-triangular array with pitch to diameter ratio of 1.375. The array was 6 rows deep with 5 tubes in each row. Ten tubes at the centre of the array were designed to be flexibly mounted. The natural frequency and damping of these tubes could be controlled over a range of values. The rest of the tubes were rigidly fixed in the test section.</p> <p>The critical velocity for instability of tubes located in the third and fourth rows was established and the stability for the array was determined. The present results indicate that extrapolation of the stability boundary proposed from results obtained in wind tunnels, for arrays of the same geometry yield unconservative predictions. A study of the flow velocity power spectra indicate that no flow periodicity could be detected in the flow before the critical velocity was reached. Detuning of adjacent tubes was proven to have little effect on the stability threshold of a tube located in the third row.</p> <p>The results of the present study were compared with data in the published literature. It is seen that the present results lie below the majority of the proposed stability boundaries.</p> / Master of Engineering (ME)

<p>Particulate metal matrix composites (PMMCs) are being widely used in the aerospace and automotive industry due to their favourable properties, mainly high specific strength and wear resistance. However, machining of particulate metal matrix composites presents a great challenge to the industry as the reinforcing particles easily abrade most of the common cutting tool materials. Polycrystalline diamond (PCD) tools appear to be the most practical type of tool material for machining PMMCs. The first part of the presented research concentrates on experimentally identifying the effect of the various cutting parameters on the chip formation mechanism, which in turn influences the tool wear. Increasing the feed rate had a beneficial effect on the tool wear due to abrasion. This is attributed to the reduced contact between the tool and the abrading reinforcing particles in the chip. The experimental work also involved an evaluation of the damage introduced into the workpiece due to machining. This damage is in the form of crushed particles, cracks and voids around the particles. The surface finish of the work-piece is a reflection of the tool's wear state. The second part of the research involved building up numerical models for the cutting tool and workpiece. Computer simulation of the machining process can potentially reduce the number of experimental iterations needed to determine the optimum cutting parameters, which are the ones that produce the minimum tool wear and least workpiece damage with minimum cost. The model stress results were verified through scanning electron microscopic observations as well as transmission electron microscopic analysis. The finite element models revealed that the area beneath the machined surface experiences high tensile stresses, which cause void formation around the reinforcing particles. These voids join up to form cracks, which have serious implications on the fatigue life of the machined part. Further research into the changes introduced in the microstructure of the chip and machined work-piece is recommended, as this could help in fully understanding the stress state beneath the machined surface.</p> / Doctor of Philosophy (PhD)

This thesis outlines a modelling approach and compensation strategy for understanding and improving the accuracy of five-axis machining. The kinematic model of the machine was based on Homogeneous Transformation Matrices (HTM), which used small angle approximations to form shape and joint transformations. HTMs were chosen in this form because they allowed the kinematic model to include the geometric, kinematic, and thermal errors of the machine. These sources of error result in a significant loss of accuracy in the production of dies, molds, aircraft parts, and many other critical industrial components. The individual error terms in the HTM's formulation are difficult to measure on an assembled machine. Therefore, a test procedure was designed to isolate the error terms and an allocation strategy, using the kinematic model, was developed to associate the error measured in the work space to the error in each of the main components of the machine. The kinematic model of the machine was also used in calculating the compensation values to capture the interaction of the error components and all of the axis motions. The compensation strategy was implemented by adjusting each axis in the motion code with offset values based on the expected error. The motion code was modified just before being sent to the machine tool for processing. Due to the difficulties in measuring the error continuously on-line, empirical models were used to estimate the errors. The empirical models related temperature distribution and various machine conditions to the error in the machine tool's components. The empirical models took advantage of a priori information in the form of known physical relationships which described the deformation of the structures under thermal loading. A multiple linear regression model was chosen as the empirical model for implementation in the compensation strategy. It was chosen because of its simplicity, robustness, and ease of implementation. Through the recasting of the linear model's parameters, nonlinearities could also be included in the model. The regression model was found to be able to interpolate within the experimental data set. Through the inclusion of physical relationships in the development of the model, it was found to provide reasonable estimates when extrapolating. A neural network model was also considered, but was not implemented in the final compensation strategy because of the model's limitations when extrapolating beyond the training set conditions. A means of updating the error model on-line was included in the compensation strategy using measurements made from reference surfaces in the work space. This was done to account for any unmodelled variation not captured during off-line testing. Also, this includes any slowly varying changes which can occur to the machine over time due to wear and structural material instability. Realistic cutting tests using light finishing cuts, high spindle speed and axis feed were conducted. These machining conditions are typical of the mold and die industry. The cutting tests showed that a reduction of dimensional error from 0.082 mm to 0.012 mm was possible using the proposed compensation strategy. / Doctor of Philosophy (PhD)

<p>The feasibility of an alternative concept of a small scale hydroelectricity generation system which is suited to high volume, low head flows was investigated. The device consists of a cascade of hydrofoils which are mounted in an arrangement which resembles a conveyor belt. The investigation included the use of a computer model which simulated the water flow through the cascade. A sensitivity analysis was done using the model to establish the effects of varying the design variables which defined the system. On the basis of the analysis, a prototype was designed with the consideration of the material, configuration, safety and environmental requirements. A cost estimate and economic analysis were made of the prototype and subsequently the concept was evaluated.</p> / Master of Engineering (ME)

<p>Numerical analysis for turbulent buoyancy-driven flows shares many common topics with other computational fluid dynamics (CFD) fields. However, it has several special problems that must be dealt with. The major contribution of this thesis is the development of a new algorithm, SIMPLET, for buoyancy-driven flows. The essence of the SIMPLE method lies in its coupling between the momentum and continuity equations. Almost all the algorithms of the SIMPLE family are based on one precondition, that is, the corrected velocity is obtained from the corrected pressure only. However, in buoyancy-driven flows, there are two major forces driving the fluid movement: the force caused by the temperature gradients and the force caused by the pressure (including kinetic pressure) gradients. In this thesis, the effect of the temperature correction on the velocity correction is considered during the derivation of the pressure linked equation. A modification to the SIMPLE algorithm--SIMPLET--was proposed. The development of the SIMPLET is divided into two stages. The first version of SIMPLET was developed on the basis of the Boussinesq assumption. Since the temperature variation in the flow fields encountered in modern electronic equipment and other industrial facilities is large enough that the Boussinesq assumption is not acceptable, the second version of SIMPLET was developed to remove this restriction so that it can be used for general cases. Because large temperature variations invariably cause turbulence, the flows with appreciable length scales are nearly always turbulent. As a preview of the application of the SIMPLET algorithm to real industrial problems, this thesis investigates several cases of turbulent mixed convection flows in a cavity problem using the RNG turbulence model. The test cases show that the SIMPLET method can speed up the energy equation convergence rate because of its linkage between temperature and velocity. When the convergence rate of the energy equation becomes the determinant in reaching a solution, the advantage of the SIMPLET method will be prominent.</p> / Doctor of Philosophy (PhD)

<p>This thesis is devoted to applying Boolean methods to investigate more efficient methodologies for discrete optimization and expert systems; both are based on "binary decision".</p> <p>An efficient nonlinear 0-1 programming algorithm is proposed, which relies mainly on logic analysis applied to the prime implicants generated iteratively from the constraint system. A general method for design optimization with discrete variables is also developed, for which the basis is an accurate neighborhood search procedure based on Boolean operation.</p> <p>A new methodology for designing and implementing rule based expert systems using Boolean methods is proposed. This consists of a consensus based algorithm for converting a set of rules to a minimal Boolean form, together with a new control algorithm for rapidly minimizing the evidence set required for a solution. These algorithms have considerable potential for simplifying systems, and speeding up the execution, which would be highly desirable for real time systems where high speed is vital.</p> <p>A procedure for building an expert system on a VLSI chip has been presented. An Erasable Programmable Logic Device (EPLD) is used to "hard wire" the logic rules represented by Boolean expressions on a microchip. The result is an extremely fast system with considerable promise for control applications, and also in other systems where size and speed are important performance characteristics.</p> / Doctor of Philosophy (PhD)