Loading conditions in a machine structure usually cause the contact pressure at the joints to take the form of a distribution, which in turn causes thermal contact resistance to be position-dependent also. / In the experiments described in this thesis, two thin-plate specimens of steel under plane-stress loading conditions generating contact pressure distributions of various profiles at the interface, were subjected to a thermal field. Temperature measurements served as reference for the finite element modelling which, through consecutive iterations, provided the values for the thermal contact resistance distributions. Combined mechanical contact pressure and thermal contact stress distributions were considered at the interface. / The function representing the relationship between thermal contact resistance and contact pressure for various distributions was defined using the least squares method. It was revealed that although this relationship can be expressed by the single function for the whole experimental range, the deviations experienced for different slopes and forms of distributions (convex and concave), particularly noticeable for steep slopes at high contact pressure levels, could indicate the effect of macro-constriction resistance, however small its values according to the theoretical calculations might be.

An experimental apparatus was designed and constructed for the study of laminar mixed convection heat transfer in vertical, horizontal and inclined tubes. The working fluid was distilled water, with bulk temperatures in the range of 8$ sp circ$C to 31$ sp circ$C. / An innovative design allows, for the first time, flow visualization over the entire heated portion of the test section. The key element of this design is a thin, electrically conductive gold-film heater suitably attached to the outside surface of a plexiglas pipe: the gold film is approximately 80% transparent to electromagnetic radiation in the visible wavelength band. This test section was mounted inside a transparent vacuum chamber to insulate it from the environment. A dye injection technique was used to visualize the mixed-convection flow patterns. The apparatus was also designed and instrumented to allow the measurement of both circumferential and axial temperature variations over the heated tube. / The flow-visualization results revealed the following: (i) a steady recirculating flow pattern, followed by laminar flow instability in vertical tubes; (ii) steady spiralling flow patterns in inclined and horizontal tubes, that confirmed earlier numerical predictions. The temperature results agreed qualitatively with earlier published experimental and numerical data. Local and overall Nusselt numbers can be calculated using the data presented, but this is not within the scope of this thesis.

The present thesis deals with the successful introduction of a second-moment closure turbulence model into a computer program using the Finite Element Method to solve the Navier-Stokes equations. The implementation presented has the advantage of using an equal interpolation for all the variables. It is also very economical in terms of the amount of memory required from the computer, since a fully decoupled formulation has been adopted, along with an iterative solver which permits to store in memory only the non-zero coefficients of the linear system of equations to be solved. Specialized elements are used to avoid resolving the near-wall region of the flow. The apparent viscosity concept is derived for the finite element formulation, along with a correction factor which permits a better representation of the Reynolds stresses. The RSM is compared to the older $k - epsilon$ model in two test cases where experimental data was available. The conclusion drawn from this work is that the RSM is able to reproduce more phenomenon occurring in turbulent flows than the $k - epsilon$ model. It is thought that the $k - epsilon$ model will gradually be supplanted by more complex models, as more computing power become available.

A model of a parallel mechanism called the Double Tetrahedron was constructed. The model displayed more mobility than predicted. The increased mobility of the model was attributed to extra rotary joints inadvertently created by the design of the mechanism. / The mobility and workspace of this newly created mechanism, dubbed Tetradon, are examined qualitatively. The movement of Tetradon is described by the rigid body rotation and translation of one tetrahedron relative to another. A method to calculate Tetradon's joint coordinates based on this rotation and translation is presented. The underconstrained solution for Tetradon and the constrained solution for a modified version of Tetradon are given. / With a view toward applying Tetradon as a positioning mechanism, three different ways in which to calculate the direct kinematics of Tetradon are presented. / Complementing the direct kinematic solutions, a solution to the inverse kinematics of Tetradon is presented. A Newton-Gauss approximation scheme is applied to a set of constrained objective functions. The objective functions are used to maintain the intersection of the six edge pairs in the mechanism.

This Thesis investigates theoretically and experimentally the dynamical behaviour and the stability of a cylindrical shell coaxially located in a rigid cylindrical pipe and subjected to internal or annular flow. / In the theoretical study, the shell could be clamped or pinned at both ends and the fluid flow in the inner shell and the annulus is assumed to be viscous and incompressible. The fluid forces consist of two parts: (i) steady viscous forces which determined using turbulent fully-developed theory; (ii) unsteady viscous forces which are derived by means of linearized Navier-Stokes equations. Shell motions are described by Flugge's modified shell equations. Two methods of solution are employed to formulate the problem. (1) Fourier transform techniques. (2) Travelling wave solutions. / The objectives are to investigate the effects of unsteady viscous forces as the dynamical behaviour and stability of the system in the presence and absence of steady forces. / Calculations have been conducted with a steel shell conveying water with different gap-to-radius ratios g/a$ sb{ rm i}$ = 1/10 and 1-100. / First, the system is subjected to unsteady viscous forces only. It is found that, for internal flow and annular flow for g/a$ sb{ rm i}$ = 1/10, the effects of viscosity on the stability of the system are insignificant; however, for the smaller gap (g/a$ sb{ rm i}$ = 1/100), those effects are more pronounced, rendering the system more stable. When both steady and unsteady viscous forces are applied, the results are quite different from the previous case. For internal flow, the system becomes more annular flow, the system loses stability at much lower velocities for both gap-systems. / In the experimental study, the flow is only annular. The shell could be clamped at both ends or clamped at one end and free at the other. For the clamped-clamped shell, the system loses stability by divergence (buckling) as predicted by linear theory. However, coupled-mode flutter was never observed experimentally. Clamped-free shells, on the other hand, lose stability by flutter.

The necessity of modelling multibody systems with flexible links in high-speed operations and space structures has become apparent. In this thesis, a general formulation for the simulation of multibody systems with multiple kinematic loops and flexible links of arbitrary shapes is developed. Parallel manipulators, space structures, multi-armed manipulators, and cooperating serial manipulators are examples of these systems. / Finite element method is used for discretization of the flexible links, while the Lagrange formulation is used to derive the equations of motion of the uncoupled links. The kinematic constraints equations are generated next by using the natural orthogonal complement (NOC) of the twist-constraint matrix. Here, the formulation of the problem is obtained both in joint and in Cartesian spaces. Using the NOC, the constraint forces are eliminated from the equations of motion to obtain the governing equations of the system in minimum coordinates. Moreover, the formulation incorporates geometric nonlinearities in the elastic displacements, which can be very crucial in large rigid-body motions. / A simulation environment is developed to perform the procedures underlying the above formulations for different types of robotic manipulators with kinematic loops and flexible links. To highlight the link flexibility effect, the governing equations of motion are used in the simulation of the aforementioned systems to compare the results obtained with the rigid and the flexible-link models.

The migration phenomenon of a neutrally buoyant solid particle suspended in an arbitrary flowing fluid of zero Reynolds number bounded by a plane solid wall has been analytically studied for the situation where the influence of boundary conditions is important. In particular, it is shown that for flows near a solid wall, particles can move in such a manner so as to cause large spatial variations of particle concentration in the suspension. / Similar techniques are used to obtain a theoretical formulation for the non-deformable spherical droplet suspended in the prescribed flowing fluid bounded by the solid plane boundary. Special cases such as flows containing a vortex near the solid boundary or flows which are periodic along the wall are particularly examined. / Experiments have been performed to verify the validity of the theory for the solid particles and non-deformable fluid drops. Experimental results are also compared to theoretical conclusions.

This thesis presents a method based on the velocity singularities for the analysis of oscillating airfoils. The method of velocity singularities has been originally developed by Mateescu for the analysis of the steady flows past airfoils. This method makes use of special singularities associated to the leading edge and ridges that directly represent the complex perturbation velocity. / Closed form solutions were derived for the pressure distributions and for the aerodynamic forces and moments acting on the oscillating airfoils with or without oscillating ailerons. / The solutions obtained with the velocity singularity method for steady flows past airfoils were found to be in very good agreement with the exact solutions obtained by conformal transformation in the case of thin Joukowski airfoils, as well as with the previous solutions for the case of flexible-membrane airfoils obtained by Mateescu & Newman, Nielsen and Thwaites. / For unsteady flows, this method has been validated for airfoils executing oscillatory motions in translation and pitching rotation and for airfoils with ailerons executing oscillatory rotations. The pressure distributions and the aerodynamic forces and moments obtained with the present method were found to be in excellent agreement with the previous solutions obtained by Theodorsen and Postel and Leppert. / The method has then been extended to the unsteady flows past airfoils executing flexural harmonic oscillations. Closed form solutions were also derived for the pressure distributions and the aerodynamic forces and moments acting on an airfoil with an aileron executing flexural oscillations. No comparisons were presented for the cases of flexural oscillations since there are no previous results known. / The present approach displayed an excellent accuracy and efficiency in all problems studied.

Our interest is in designing reduced degree-of-freedom articulated mechanical assistive devices capable of tracing desired end-effector paths in the plane with the ability to be easily reconfigured for different sets of tasks. / We examine the use of a manipulator configuration called Single Degree-of-freedom Coupled Serial Chains (SDCSC). Such SDCSC configurations combine the motion flexibility due to multiple articulations with the simplicity of reduced-degree-of-freedom control and actuation due to the presence of hardware constraints and offer a low-cost alternative to expensive robotic manipulators. Our interest is in designing manipulators, based on such an SDCSC configuration, whose end-effector can match or closely approximate a set of positions along a desired planar path in the so-called Path Following problem. / Past efforts examined the design of such manipulators for open-loop curves with limited applicability to closed-loop curves. To this end, we examine the development of a novel method for kinematic synthesis of such SDCSC mechanisms for such planar closed-loop path-following tasks. / We also examine some of the issues related to converting the mathematical model into the physically realizable design of a reconfigurable prototype based on the SDCSC configuration. (Abstract shortened by UMI.)

The propagation of non-ideal detonations arising from friction, heat transfer and reactions steps involving a competition between exothermic and endothermic reactions (pathological detonations) has been studied theoretically by including source terms in the 1-D conservation equations of momentum and energy. To predict the steady-state non-ideal detonation velocity, the detailed structure of the detonation has been considered. The Generalized C-J Criterion has been used to seek a singularity-free solution from the whole spectrum of possible solutions to the differential equations for the structure. / For pathological detonations, the steady-state analysis predicts a detonation velocity in excess of the ideal C-J value in the H2 --Cl2 mixture, in agreement with experimental observations. Unsteady pathological detonation calculations with a simplified two-rate law model have also been carried out. The resulting asymptotically stable detonations are found to be in agreement with the steady-state predictions. / For non-ideal detonations due to friction, the Generalized C-J criterion is found to break down for very low detonation velocities. An alternative criterion based on matching the detonation wave with the back boundary condition is used instead. A continuous spectrum of steady-state solutions has been found for detonation velocities ranging from the ideal C-J value down to that of a sonic wave. For activation energies above some critical value, multiple steady-state solutions have been found for a given friction factor, and various detonation regimes have been defined. / An unsteady analysis of the transient development of non-ideal detonations due to friction has been carried out to determine whether the solutions from the steady-state analysis can be approached asymptotically. Friction and heat transfer are found to increase the instability of the detonation wave. Oscillatory and even chaotic detonations were observed for values of the activation energies corresponding to stable detonations in absence of source terms (ideal detonation). Moreover, the transient analysis has revealed that in the case of multiple steady-state solutions, only that with the highest detonation velocity could be approached asymptotically. The transient results of detonation with friction and heat transfer have been found to be in qualitative agreement with the experimental gaseous detonations propagating in porous media and in obstacle fields.