"November 2004" / Bibliography: p.167-179. / xi, 333 p. : ill., photos (col.) ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Focuses on behavioural models of masonry walls with a view to improving their accuracy and extending their application. Results include a numerical model and mathematical expressions capable of predicting the key stages of the non-linear load-deflection behaviour of walls subjected to vertical bending and axial loading; new mathematical expressions for horizontal and diagonal bending moment capacities that are dimensionally consistent and account for the beneficial effects of compressive stress; and. Experimental test data for masonry sections subjected to horizontal and diagonal bending, which were used in the development and verification of the new mathematical expressions. / Thesis (Ph.D.)--University of Adelaide, School of Civil and Environmental Engineering, 2004

The concept of load analogy is used in the elastic and elastic-plastic analysis of isotropic circular and annular plates undergoing moderately large deflection. The effects of the nonlinear terms of lateral displacement and the plastic strains are considered as additional fictitious lateral loads, edge moments, and in-plane forces acting on the plate. The solution of an elastic or elastic-plastic Von Karman type plate is hence reduced to a set of two equivalent elastic plate problems with small displacements, namely, a plane problem in elasticity and a linear elastic plate bending problem. The method of finite element is employed to solve the plane stress problem. The large deflection solutions are then obtained by utilizing the solutions of the linear bending problems through an iterative numerical scheme. The flow theory of plasticity incorporating a Von Mises layer yield criterion and the Prandtl-Reuss associated flow rule for strain hardening materials is employed in this approach.

This work investigates the potential of reducing stresses in the region of stress concentration by applied induced strains. A thin 30 inch square plate with a 1 inch diameter circular hole under uniaxial load was used for this investigation. This investigation considered first an ideal case with a few physical limitations for the purpose of probing the limits of active stress reduction. Applied induced strains were applied over the region A ≤ r ≤ 1.5A, where A is the radius of the hole. It was found that the axisymmetric applied induced strains could reduce the stress concentration factor (SCF) from 3 to 2. With non-axisymmetric applied induced strain distribution the SCF could be reduced to 1.45. Numerical optimizations based on finite element simulations were also carried out for a composite plate with a hole and similar reductions in stress concentration factors were obtained. Next, a more realistic case, consisting of bonded and embedded piezoelectric actuators was considered. It was found that partial-thickness actuators produce large radial stresses which erase any benefits associated with axisymmetric actuation. With non-axisymmetric actuation, the actuators with present technology limitations were found to be effective in reducing stress concentration factor for a fatigue load case. However, due to repetitive nature of the load, the energy expenditure may be large. For an extreme load case, actuators with present technology limitations were not very effective in reducing SCF. Actuators were needed to be placed over a larger area to achieve a larger reduction in SCF. Also, passive stiffening was found to be more effective in reducing stress concentration than bonded actuators with present technology limitations. / Ph. D.

One of the major problems with surface-mounted or embedded induced strain actuator (ISA) patches are the considerably high stress gradients introduced at the edges of the actuator patches when an electric field is applied. These excessive stress gradients initiate debonding of the actuators from the substrate, thus affecting the mechanical reliability of the structure. This thesis is begun by investigating existing theoretical models of induced strain actuated structures, and will later use these to compare with the finite element analysis. The finite element analysis is used to explore the stress concentrations located at the edges of the actuators and begins by refining the mesh areas of the same structure focusing in on the ends of the ISA’s. This preliminary analysis is conducted on a structural configuration with a perfectly bonded actuator and proceeds to one with a finite bonding layer. After completion of the mesh refinement investigation several modifications in the design and implementation of the induced strain actuators are examined to reduce the stress concentrations at the edges of the actuators. In the finite element analysis two separate modeling considerations are examined: 1) The actuator is perfectly-bonded to the substrate. 2) A finite adhesive layer is incorporated between the actuator and the substrate. With each modeling consideration several design modifications are considered in this thesis including employing partial electrodes on the induced strain actuator surface regions instead of fully electroded surfaces, examining an actuator with a chamfered end, and using caps to reduce the stress concentrations and possibly increase the performance of the structure by allowing the induced strain actuators to utilize their piezoelectric strain coefficient in the thickness direction, d₃₃. The design modifications and different modeling techniques help to alleviate the critical stresses in the structure while gaining a better understanding of causes them. / Master of Science

No description available.

No description available.

The first main goal of this research is to develop comprehensive analytical and numerical models for the stress distribution around an inclined cased wellbore by considering all wellbore processes and to amend erroneous models of most previous work. The second main goal is to apply the developed models to explain near wellbore phenomena such as cement failure and sand production. To achieve these goals, this work checked the eligibility of using simple elastic approaches for the system by using a poroelastic undrained condition and a steady state condition for stresses induced by wellbore temperature variation. It utilized the generalized plane strain to compensate for the limitation of the plane strain which most previous work had used. In addition, this research developed comprehensive models to improve previous work by using superposing principles. For applying the developed models to cement failure, Mogi-Coulomb criterion for shear failure instead of Mohr-Coulomb and Drucker-Prager criteria was used to properly consider the intermediate stress. Additionally, ABAQUSr was utilized for numerical models with the "model change" option to simulate and combine all individual wellbore processes while MATLABr was used for analytical models. For predicting sand production, fully coupled poroelastic solutions for an inclined open wellbore were modified to obtain the stress distribution around a perforation tunnel after perforating. Then, modified Lade failure criterion was used to calculate the critical drawdown when sand production occurs, that is, when the perforation tunnel starts failure. This research obtained the following results. For developing models, the analytical models improved the previous research. However, the numerical results under a vertical tectonic stress showed discrepancies because of the difference between the generalized plane strain and numerical models. For cement failure, Young's modulus of cement, wellbore pressure and wellbore temperature variation could affect shear failure more significantly than the other factors. The numerical results showed closer to the failure envelopes than the analytical results. For predicting sand production, well completion affected sand production near wellbore and the critical drawdown converged to asymptotic values. In addition, perforating along the minimum horizontal stress direction was most preferable in a vertical cased wellbore under a normal stress regime. / text