A novel non-linear perturbation theory based on the characteristic equations for structural dynamic systems is developed, which can provide an exact relationship between the perturbation of structural parameters and the perturbation of modal parameters. Then, depending on information about different types of the measured vibration modal data available, a system of governing equations based on the developed theory is derived, which can be utilised for general applications, such as eigendata modification, model updating, and damage identification, suitable for all types of structures, including framed structures and continua. A number of computational procedures based on the derived non-linear governing equations are presented for structural damage identification, which can be suitable for various cases of the measured vibration modal data available, such as only natural frequencies, complete mode shapes, locally complete mode shapes, and incomplete mode shapes. The effectiveness and convergence performance for the proposed approaches are demonstrated by various numerical examples, and the sensitivities of many factors to inverse predictions of structural damage are also investigated. The results for different types of structures, either framed structures or continua, indicate that the proposed approaches can be successful in not only predicting the location of damage but also in determining the extent of structural damage, while at the same time information about only a limited amount of the measured modal data is required. Furthermore, it is found that the proposed approaches are capable of providing information on the exact expanded damaged mode shapes, even if a very limited DOF's readings are available. Structural modelling problems, which have to be considered in structural analysis and damage identification, are discussed. It is shown that structural damage can be identified correctly from the proposed approaches using information about different types of the measured modal data, regardless of different structural models considered and different types of elements used. Therefore, a suitable structural model can be selected in order to properly identify structural damage depending on the available information about modal data

A study has been made of the forces generated on a submerged horizontal cylinder, held remote from the channel bed, due to the passage of a solitary wave. The horizontal cylinder was positioned with its major axis parallel to the direction of the wave crest. The vertical and in-line forces were measured using a force balance. The horizontal fluid velocities were measured using a Hot Film Anemometer (HFA). The drag and inertia components of the force are separated in the analysis by three different methods and correlated with Reynolds, Keulegan & Carpenter and Inverson's numbers. This study investigates a basic wave I structure interaction using Morison's equation. This supposes that the total force on a submerged object is the simple superposition of two well established force components, namely the drag and inertia components. History effects are removed by using the fluid environment generated by the solitary wave, which is unidirectional accelerated motion in still water. Measurement and validation of the fluid flow characteristics are made, to ensure a high degree of certainty for the results. The objective is to establish a relation by which the coefficients used in Morison's equation can be accurately determined from basic wave characteristics.

The modern trends towards economy and the use of high strength materials have resulted in long spans and slender floors of low frequencies. These frequencies may be within the range of the first few harmonics of daily life human activities. Though the problem of resonance with walking vibrations, an activity most common on all floors, is unlikely, high amplitude or persistent vibrations due to these low-level excitations may cause alarm to building occupants. There may also be some problems with the most sensitive equipment. These uncomfortable vibrations are a serviceability limit state problem and can only be avoided by ensuring a high floor fundamental natural frequency and damping. There is a need, therefore, for a method to accurately predict the fundamental natural frequency and damping of these floors and to ensure that they are high enough to avoid any resonance or perceptibility problems. Available analytical formulae for the estimation of fundamental natural frequency are not directly applicable to actual floors due to various assumptions. The only method that may be reliably used for static or dynamic analyses is the finite element method because it can conveniently model the three dimensional nature of structures and account for the various boundary conditions and material properties. The research reported in this thesis consists of measuring fundamental natural frequencies and corresponding damping of a range of actual floors. The experimental frequencies have then been compared with those results which are based on the analytical formulae and finite element method. The analytical methods suitable for various categories of floors have been identified. A new linear-elastic single panel or beam finite element model, correlated with the experimental results, has been developed for the accurate estimation of the fundamental natural frequency of these floors. The correct boundary conditions for various categories of floors have been identified. The single-degree-of-freedom (SDOF) formula for the estimation of fundamental natural frequency using static deflections has been modified for the floors tested. This modified SDOF formula can be used for convenient hand calculations by the consultants and designers who want a quick estimation of fundamental natural frequency due to time and cost limitations. The formula may also be used to limit static deflections and, therefore, design loads for any choice of a minimum fundamental natural frequency. Also, new limits on span/depth ratios for flat slabs and span limits for double-T beam floors have been suggested. Similarly, minimum fundamental natural frequencies, damping ratios and maximum static deflections have been suggested for the floors tested. The single panel or beam model may also be used for various parametric studies, both for static and dynamic analyses.

The deformation and fracture behaviour of a rubber-modified epoxy, and that of-its unmodified counterpart has been studied. The materials have been examined both in 'bulk' and when used as a thin adhesive layer for bonding steel adherends. Values of fracture toughness, the type of crack growth and the resultant fracture surfaces have been studied over a wide range of temperatures, displacement rates and specimen thicknesses. Both systems exhibit essentially the same types of crack growth behaviour. However, the values of fracture toughness for the rubber-modified epoxy were almost always significantly higher than those for the unmodified epoxy. A mechanism based on cavitation of the rubber particles and shear yielding of the matrix is proposed. This accounts for the increased fracture toughness and other alterations in property in the rubber-modified material. The deformation and fracture data are used to calculate values of the crack opening displacement. The rate/temperature dependence of this, together with other correlations, indicate that the extent of blunting at the crack tip governs the toughness of the epoxy materials and controls the type of crack propagation observed. A quantitative expression is presented which describes the variation of fracture toughness with both temperature and rate. Two parameters from this expression are shown to be material constants and to provide a unique failure criterion. The fracture behaviour of the rubber-modified epoxy as an adhesive between mild-steel adherends has also been examined. Values of adhesive fracture energy, type of crack growth and resultant fracture surface appearance have been studied as a function of adhesive bond thickness, joint width and displacement rate. The bond thickness particularly is shown to have a pronounced effect on adhesive fracture energy with a maximum occurring at a specific bond thickness. This relationship is discussed in terms of a plastic zone restriction/adhesive layer constraint model. This enables the bulk properties of the adhesive to be used to semi-quantitatively explain the dependence of Glc upon bond thickness.

Photonic crystal fibre (PCF), a new kind of optical fibre, has many air-holes in their cross-section and has potential applications to new optical communication systems. The main objective of this research is the modelling of photonic crystal fibre to identify the fundamental and higher order quasi-TE and TM modes with square, ,rectangular and circular air holes in a square and hexagonal matrix, by using a rigorous full-vectorial H-field based finite element method (FEM). Besides the modal solutions of the effective indices, mode field profiles, spot sizes, modal hybridness, polarization beat length and group velocity dispersion values for equal and unequal air holes; research was carried out to optimize and design highly birefringent PCF. The variation of modal birefringence is shown through the effect of hole diameters, air hole arrangement, structural asymmetry, operating wavelength, and pitch-distance. Birefringence was enhanced by breaking the structural symmetry and this was verified by using unequal air holes. The diameter of two air holes and four air holes in the first ring was changed to break the rotational symmetry and a comparison between the two designs is made in this work. In this work, highly birefringent PCF is designed with higher operating wavelength, larger d2/A value, lower pitch length for a given structural asymmetry. It is identified that birefringence value increases rapidly when d2 is much larger than d. At lower pitch value, one of the highest birefringence values reported so far at wavelertgth of 1.55 J.Jm for an asymmetric PCF using circular air holes. A single polarization guide PCF structure is also achieved. In this study, it has been identified that for fixed d/A and d2/A value, as operating wavelength is increased, birefringence increases significantly. It can also be identified that for higher d/A values, birefringence changes rapidly with A as their corresponding cutoff condition also approaches. One important validation of this work is the existence of modal birefringence for PCF with six-fold rotational symmetry. It is shown that birefringence value of a simple PCF incorporating circular holes but of different diameters is high compared to polarization maintaining Panda or Bow-tie fibres. This research also aims to investigate the modal leakage losses of PCF, by using a semi-vectorial beam propagation method (BPM) based on the versatile FEM. The robust perfectly matched layer (PML) boundary condition has been introduced to the modal solution approach. The effects of d2/A, operating wavelength and number of air holes have been thoroughly detailed and explained. In this study, it has been identified that the confinement loss decreases significantly with the increased number of rings, lower operating wavelength and lower d2/A value. For special case, PCF with large spot-size provides higher leakage loss.

The primary purpose of this study was to determine what factors influence the integration of educational technology as perceived by Texas teachers. The secondary purpose was to examine the relationships between the determinant factors. This study answered the following questions: 1) Are there significant differences in teachers’ perception of school technology and readiness across grade level and subject area? 2) Are there significant differences in teacher-principal technology readiness congruence across school percentage of economically disadvantaged students? 3) How do teachers’ perceived levels of technology readiness predict student mastery of Technology Applications (TA) TEKS? To address the research questions quantitative procedures were followed to investigate whether significant relationships existed among dependent variables, school technology and readiness, and teacher-principal school technology and readiness congruence, and the independent variables (a) grade level, (b) subject area, and (c) percentage of economically disadvantaged students. Data analysis indicated significant differences in teacher school technology and readiness perceptions by grade level and subject area, and significant differences in teacher-principal school technology and readiness congruence by school percentage of economically disadvantaged students. Using path analysis a theorized Texas School Technology and Readiness Effects Model was validated. The findings were that (a) teachers in higher grade levels and more technical subject areas perceived their school technology and readiness at significantly higher levels; (b) as the percentage of economically disadvantaged student increased in a school, teachers perceived their school technology readiness at lower levels and were less congruent between their perceived school technology and readiness and their principals’ ratings of the teachers’ school technology and readiness; and (c) Leadership Administration and Instructional Support, followed by Infrastructure for Technology, Educator Preparation and Development, and Instruction Practice were the main drivers for student mastery of Technology Applications (TA) related Texas Essential Knowledge and Skills (TEKS). / text

No description available.

Rheology involves the study and evaluation of the flow and permanent deformation of time-and temperature-dependent materials, such as bitumen, that are stressed through the application of a force. The fundamental rheological properties of bituminous materials including bitumen are normally measured using a dynamic shear rheometer (DSR), from low to high temperatures. DSR is a powerful tool to measure elastic, viscoelastic and viscous properties of binders over a wide range of temperatures and frequencies, provided the tests are conducted in the linear viscoelastic region. Therefore, the study of bitumen rheology is crucial since its reflects the overall performance of a flexible pavement. However, it is well known that the DSR also has limitations, where the measurements are exposed to compliance (testing) errors particularly at low temperatures and/or high frequencies. In addition, conducting laboratory tests are known to be laborious, time consuming and require skilled personnel. Therefore, this research is conducted to elucidate a better understanding of the rheological properties and modelling procedures of bitumens and bituminous binders. Various materials such as unmodified bitumens, polymer-modified bitumens (PMBs) and bitumen-filler mastics, unaged and aged samples, are used in this study. An extensive literature review was undertaken to identify reliable models that can be considered as a valuable alternative tool to describe or fit the rheological properties of bitumen. These properties are commonly presented in terms of complex modulus and phase angle master curves, together with the determination of shift factor values at a particular reference temperature. In general, the complex modulus and phase angle master curves can be modelled using different techniques; nomographs, mathematical equations and mechanical models. However, the nomographs have become obsolete in recent years and tended to be replaced by the two latter models. Those models are able to satisfactorily describe the rheological properties of unmodified bitumen. However, the observations suggest a lack of agreement between measured and predicted rheological properties for binders that contain a phase transition, such as found for highly crystalline bitumen, structured bitumen with high asphaltenes content and highly modified bitumen. An attempt was made to evaluate the validity of several mathematical equations and mechanical element approach using unaged and aged unmodified bitumens and PMBs database. It is observed that the Sigmoidal, Generalised Logistic Sigmoidal, Christensen and Anderson (CA), and Christensen, Anderson and Marasteanu (CAM) Models are able to satisfactorily describe the rheological properties of unmodified bitumens. Nevertheless, they suffer from the same drawbacks where the presence of highly EVA semi-crystalline and SBS elastomeric structures render breakdowns in the complex modulus master curves. Similar discrepancies are observed when one of the mechanical models (the 2S2P1D Model) is used. To construct the master curves, different shifting methods are available. It is found that a numerical shift produced the best fit between measured and modelled data, followed by the Laboratoire Central des Ponts et Chaussées (LCPC) approach, William, Landel and Ferry (WLF), Modified Kaelble, Viscosity Temperature Susceptibility (VTS), Arrhenius and Log-Linear methods. A temperature range from 10 to 75oC is used in this study. It is worth mentioning that most of the methods are empirical and might not be applicable for all materials. Finally, the phase angle master curves must also not be neglected to yield a complete rheological properties of binders. The statistical analysis between measured and modelled data shows that the Fractional Model yielded the best correlation for a temperature range from10 to 75oC, followed by the Al-Qadi and Co-workers, CAM, CA and Kramers-Kronig relationships. An anomaly is observed between measured and descriptive data of the Kramers-Kronig relationship particularly at high frequencies and/or low temperatures. The Fractional Model is not considered suitable for practical purposes due to the high number of coefficients that need to be solved.

The aim of semantic image understanding is to reveal the semantic meaning behind the image pixel. This thesis investigates problems related to semantic image understanding, and have made the following contributions. Our first contribution is to propose the usage of histogram matching in Multiple Kernel Learning. We treat the two-dimensional kernel matrix as an image and transfer the histogram matching algorithm in image processing to kernel matrix. Experiments on various computer vision and machine learning datasets have shown that our method can always boost the performance of state of the art MKL methods. Our second contribution is to advocate the segment-then-recognize strategy in pixel-level semantic image understanding. We have developed a new framework which tries to integrate semantic segmentation with low-level segmentation for proposing object consistent regions. We have also developed a novel method trying to integrate semantic segmentation with interactive segmentation. We found this segment-then-recognize strategy also works well on medical image data, where we designed a novel polar space random field model for proposing gland-like regions. In the realm of image-level semantic image understanding, our contribution is a novel way to utilize the random forest. Most of the previous works utilizing random forest store the posterior probabilities at each leaf node, and each random tree in the random forest is considered to be independent from each other. In contrast, we store the training samples instead of the posterior probabilities at each leaf node. We consider the random forest as a whole and propose the concept of semantic nearest neighbor and semantic similarity measure. Based on these two concepts, we devise novel methods for image annotation and image retrieval tasks.

This thesis presents work relating to the measurement of the nonlinear acoustoelastic response of several materials with respect to microstructure and fatigue life. The nonlinear acoustoelastic response measures the acoustoelastic coefficient of a material. During normal usage components are subject to stresses that while not sufficient to cause fracture cause fatigue, gradually weakening the component. Linear ultrasonic methods have been shown to be poor at detecting fatigue. However, there is evidence that the accumulation of fatigue damage gives the material a nonlinear elastic response that can be probed by ultrasound. A potential technique to monitor fatigue is produced by monitoring changes in the response of nonlinear ultrasonic techniques. Several methods of detecting material nonlinearity using acoustic waves have been proposed. In this thesis a system using the collinear mixing of ultrasonic waves is developed. By measuring the velocity change of a probe wave due to the induced stress from a pump wave, a measure of the nonlinearity is obtained. By using laser ultrasound techniques we gain the benefits of high spatial and temporal resolution. This is important when investigating the nonlinear response of a material as there is evidence that the microstructure affects the nonlinear response of a material. Single point measurements of the acoustoelastic coefficient are taken. In polycrystalline materials the measurement is found to vary with location. A technique is developed to measure the spatial variation of the acoustoelastic coefficient. A relationship between the acoustoelastic coefficient of a material and the underlying microstructure is found. The technique to monitor the spatial variation of the acoustoelastic coefficient is used to monitor the change in samples as they are fatigued. The acoustoelastic coefficient is found to change with fatigue by a larger proportion (9-30%) than linear velocity measurements (0.5-0.6%). Spatial variation of the acoustoelastic coefficient indicated a complex relationship between fatigue and the acoustoelastic coefficient of a material. The implications and further work required from the observed changes in the acoustoelastic coefficient with microstructure and fatigue are discussed.