<p>An Integrated Coastal Engineering model (ICEM) has been developed, which models the waves, the coastal circulation and the rates of bed level changes simultaneously. The wave model simulates the combined effect of refraction, diffraction, shoaling, reflection and breaking and the circulation model simulates the wave driven currents in the coastal zone. Finally, the morphological model simulates the sediment transport loads and the morphological changes after solving the combined wave current boundary layer. The waves, currents and sediment transport loads are recalculated for every so-called morphological time step. The ICEM is capable of modeling both the short and long term effects of coastal structures such as groins, detached breakwaters, seawalls, etc. The ICEM is calibrated and verified using field data in the Mediterranean coast of Egypt. The model is applied to similar areas in Crete, where the wave climate and bathymetric data are available. The best protection scheme is defined for each area. From the engineering point of view, this research provides an effective design tool for coastal protection projects. It also can provide an assessment on the effectiveness of the existing coastal protection projects.</p> / Doctor of Philosophy (PhD)

<p>Structural frames designed or proportioned to resist seismic forces, must possess adequate ductility to redistribute internal forces or have needed energy absorbing capability. This research investigation touches on a number of aspects that deal with or relate to the above characteristics, while the principal aim is to assess and compare the rotation capacity and energy absorption of locally web stiffened beams with unstiffened beams.</p> <p>An approach which allows for the prediction of the initiation of local buckling in the design of wide-flanged beams under moment gradient, is first presented. The method described represents a refined moment-rotation model that includes the effects of strain hardening. This same approach helps define more accurately the appropriate slenderness limits of a beam's plate-elements in relation to the required rotation capacity at maximum moment.</p> <p>The study then addresses the interaction effects of a steel member's plate slenderness values and its lateral slenderness on rotation capacity at ultimate deformation. It is shown that members with slenderness values close to the limits specified by codes of practice may not be able to redistribute moments adequately under seismic loading.</p> <p>To further ascertain and assess the ductility and energy dissipation capabilities of W-shaped beams, a series of test specimens subjected to monotonic and quasi-static cyclic loading was conducted. The specimens were meant to represent beams in ductile moment resisting frames undergoing cyclic lateral loads. Of direct relevance to seismically designed moment resisting steel frames, the experimental results of this research effort clearly highlight the superiority of herring-bone style stiffened specimens over unstiffened specimens. To provide further supporting evidence, as well as insights into the behaviour of herring-bone stiffeners, an inelastic large deformation analytical study was undertaken using cubic-quadratic shell finite elements. This work allowed a parametric study to be undertaken on the effects of stiffener thickness on strength and ductility properties. Based on these, and the experimental results, preliminary design guidelines have been proposed. Another important objective was to assess and document the strength and energy deterioration occurring under conditions of low cycle fatigue and which involve local buckling.</p> <p>The results of a series of W-shaped test specimens subjected to fatigue type of loading under constant amplitude are presented. This work has permitted strength and energy deterioration and damage models to be developed for the W-shape steel beams. A generalized model which uses plate slenderness values together with lateral slenderness is proposed for predicting the rate in strength deterioration per reversal and cumulative damage after a given number of reversals.</p> <p>The research investigation concludes with a review of some of the concepts used in damage assessment; simple damage parameters such as ductility ratio and realistic mathematical models reflecting the deterioration of steel beams due to maximum response, and, dissipated energy are then discussed. Damage models are then presented that combine maximum response with repeated effects in low-cycle fatigue loading. The proposed models calibrated through the use of cyclic tests on steel beams, are then used to yield deterministic parameters that predict adequate ductility values for steel beams under cyclic loading. This phase of the work was completed by having the damage models incorporated into a non-linear dynamic analysis of a sample building and ends with the recommendation of employing a new "adequate ductility" design parameter.</p> / Doctor of Philosophy (PhD)

<p>A new shell element for the analysis of thin and thick plate and shell structures has been formulated along with the compatible beam element. The new elements are referred to as the consistent shell element and the consistent beam element. By combining the consistent shell element and the consistent beam element a finite element model for the analysis of reinforced concrete structures has been formulated. The scope of the study is broadened further through the modification of the new elements to allow for the analysis of laminated fiber-reinforced composite beams and shells.</p> <p>Many of the currently available beam and shell elements exhibit spurious variations of the transverse shear stress(es). To obtain improved responses when thin beams or shells are analysed the reduced integration technique has typically been employed. This approach is not acceptable since the reduced integration technique cannot be applied with complete confidence. In the present study it is found that the unsatisfactory behaviour of these elements is due to an inconsistency in their formulation. In the formulation of the new elements a consistent formulation has been ensured. The new elements behave very well in the analysis of both thin and thick beams and shells and contain no spurious zero energy modes. In addition, they provide a quadratic variation of the transverse shear stress(es), and a cubic variation of the displacement(s) through their thicknesses. Therefore, the shear correction factor k, which is usually required to correct for the assumption of constant transverse shear strain through the thickness is not required. Both elements include material non-linearity. Special attention has been given to the efficient implementation of the consistent shell element by employing a sub-matrix formulation in conjunction with a modified frontal solution algorithm. The numerical results show the new elements to be highly accurate and computationally efficient.</p> <p>The reinforced concrete finite element model employs a rational elasto-plastic constitutive relationship for concrete, discrete bar elements for modelling of reinforcement, joint elements for bond slip between concrete and reinforcement, beam elements for supporting girders and shear connector elements along the concrete/girder interface. The numerical results show that the model accurately predicts the behaviour of reinforced concrete slabs, including punching shear failure under point loads. The constitutive model employed has been found to be reasonably objective with respect to refinement of the finite element mesh.</p> <p>The consistent laminated beam element and the consistent laminated shell element, have been formulated for the analysis of laminated fiber-reinforced composites. Special attention has been given to the approximation of stresses through the thickness of the laminate because of their importance in predicting delamination failures. This has been achieved by allowing the transverse shear strain(s) to be discontinuous at the interface of two layers while still maintaining continuity of the global displacements across the interface. The numerical results show that the elements provide very accurate predictions of stresses through the thickness of both thin and thick laminates.</p> / Doctor of Philosophy (PhD)

<p>Combined sewer systems are designed to bypass excess flows directly to the receiving water during rainstorm events. This is commonly achieved by means of diversion structures incorporating side spillways, oblique weirs or orifices.</p> <p>The simulation of diversion devices by the EXTRAN Block of the Stormwater Management Model is examined. Results for the weir algorithm were not satisfactory. Therefore, a new procedure is written to compute the overflow and the water surface profile along the side weir. The hydraulics of flow over side weirs is reviewed in detail. The momentum approach is used in computing the water surface profile. The coefficient of discharge is estimated from a newly proposed relationship. A method for predicting the formation of a hydraulic jump along the weir section is derived from the momentum balance between the upstream and downstream side of the jump.</p> <p>A computer program called OVRFLO3 is developed to compute the overflow and the water surface profile. Validation tests show good agreement with observations.</p> <p>Finally, OVRFLO3 is adjusted to fit the SWMM package. The new Block is called SIDWEIR.</p> / Master of Engineering (ME)

<p>The finite element method of simulating tunnel excavations in squeezing ground requires experimental data on multiaxial creep relationships and related characteristics of rock and concrete as input. A servo-controlled test frame, for applying multiaxial stress states to cubical specimens through three independent actuators, was developed. This test system allows the simulation of typical stress levels and stress paths involved during rock excavation, with continuous monitoring of loads and displacements through a data acquisition system. The overall equipment development required considerable design and assembly effort. A testing programme for concrete specimens was completed to evaluate the equipment. Finally, conclusion and recommendations were made which provide the necessary information and guidance for future testing to develop multiaxial creep relationships for rocks known to have significant time-dependent behaviour.</p> / Master of Engineering (ME)

<p>Conical steel shells are fairly widely used as elevated water tanks. However, the current code of practise in North America for the design of such reservoir structures provides an obsolete method for ascertaining their adequacy to resist hydrostatic loadings. Moreover, there are no provisions available for handling liquid-filled conical tanks subjected to seismic forces. The lack of appropriate design methods could not have been demonstrated more vividly when in December of 1990, an elevated conical water tower failed by buckling when being filled for the first time. The steel vessel, located in Fredericton, New Brunswick, is claimed to have "exploded" by eyewitnesses.</p> <p>The work of this thesis, then, was motivated by this failure. It involves non-linear stability analysis of liquid-filled conical steel vessels possessing geometric imperfections and residual stresses, and which can be subjected to hydrostatic and seismic loading. To achieve this, a finite element formulation is developed based on a consistent shell element which is free from spurious shear modes known to exist in the isoparametric shell elements. The consistent shell element employed also exhibits excellent performance in the analysis of plates and shells in the small displacement range. This element is extended to include both geometric and material non-linearities as well as non-linear dynamic analysis. The non-linear finite element model developed is general and can be applied to any thin or thick shell problem. Numerical testing of the non-linear model through static and dynamic analysis of different plate and shell problems indicates the continued excellent performance of the consistent shell element in the non-linear range.</p> <p>Hydrostatically loaded conical steel vessels are modelled using the consistent shell element. Static stability analyses of conical shells with different geometric imperfection patterns are undertaken and the results indicate that the presence of axisymmetric imperfections leads to the lowest limit load for the structure. The sensitivity of the hydrostatically loaded conical vessels to geometric imperfections and residual stresses is investigated by considering three cases: (i) analysis of perfect vessels, (ii) same as case (i) but with axisymmetric geometric imperfections of the order of the thickness of the shell, (iii) same as case (ii) but with the addition of residual stresses due to welding. The results from these analyses indicate that the liquid-filled conical shells are significantly sensitive to geometric imperfections, and that yielding precedes elastic buckling for tanks having practical dimensions.</p> <p>The non-linear dynamic (stability) analysis of elevated liquid-filled conical vessels subjected to both horizontal and vertical accelerations, but free from rocking motion, is then considered. The boundary integral method is used to formulate the fluid added-mass matrix resulting from the impulsive component of the hydrodynamic pressure. This is added to the mass matrix of the shell structure to perform free vibration as well as nonlinear time history analyses for elevated liquid-filled conical tanks treated as either perfect or axisymmetrically imperfect. Tanks with different dimensions and imperfection levels are subjected to an appropriately scaled real input ground motion. Some of these elevated structures exhibit inelastic behaviour and generally develop a localized buckle near the bottom of the vessel which leads to the overall instability of the structure. In general, time history analyses indicate that liquid-filled conical tanks, often possessing apparently adequate safety factors under hydrostatic loading, may not be safe under seismic loading. Therefore, a proper modelling procedure along with time dependent analysis must be followed in order to design such tanks safely. The finite element model developed in this thesis is a means provided for such a purpose.</p> / Doctor of Philosophy (PhD)

<p>A Frustum Confining Vessel (FCV) was developed by McMaster University and Berminghammer Foundation Equipment Ltd., which provides an environment for the testing of reduced-scale piles (physical model). This vessel is intended to produce stress distributions within sand specimens, which resemble field conditions, however at a smaller geometric scale. This thesis presents the findings of the experimental and analytical investigations conducted on the FCV device. A technique for measuring normal stresses in dry sand is developed, and the stresses and displacements measured at specific locations within the sand specimens are used to calibrate the finite element model. Finite element simulations are used to evaluate different aspects of the responses, which cannot be directly measured in the experiments. The fundamentals of dimensional analysis are reviewed and a set of primitive variables for the pile-soil system is presented. Using the Buckingham-π theorem, a derived set of dimensionless groups is thereby proposed for the study of pile-soil interaction. The scaling factors necessary for the extrapolation of results from model to prototype conditions are obtained via similarity analysis. It is suggested that the lack of gravity scaling in the FCV device does not introduce significant distortions in the physical model provided that stress distributions, particularly horizontal stresses, are properly controlled by FCV loading. In relation to physical modeling of piles, a criterion for the acceptance of testing conditions is established, and the suitability of the current device is assessed in terms of the mentioned criterion. It is found that the current device does not completely meet the acceptance criteria. The improvement on testing condition is thereby sought by redesigning the vessel's shape. Alternative shapes for the vessel are investigated by means of finite element modeling. A redesigned shape, which offers optimal stress conditions for the purposes of physical modeling, is presented.</p> / Doctor of Philosophy (PhD)

<p>For asphaltic concrete surface courses, the surface texture of the exposed coarse aggregate in the key factor influencing the skid resistance performance of the pavement. With time, this surface texture is altered by the dynamic, interactive processes of polishing, wear or abrasion and weathering. In this study, four types of aggregate which are available in Ontario - traprock, limestone, steel slag and blast furnace slag - were subjected to these processes in order to evaluate their surface texture characteristics. The study was completed in terms of both quantitative analyses of laboratory samples and qualitative analyses of field and laboratory samples. Two major aspects were surface texture evaluation using the scanning electron microscopy, and the simulation of weathering influences on skid resistance. Both chemical and x-ray diffraction analyses were completed to support the evaluation.</p> <p>From the study, two new concepts - positive and negative rejuvenating processes and Potential Rejuvenating Value (PRV) - were developed. Steel slag, with a relatively high PSV, low AAV and potential negative rejuvenation was found to be the most desirable aggregate in terms of skid resistance performance.</p> / Master of Engineering (ME)

<p>Short and long term capacities of slender concrete block masonry walls subject to eccentric axial loads were investigated. A finite element model was developed to include both the geometric and the material nonlinearities. To analyze large members, a combined macro-micro finite element model for the masonry assemblage was also developed by which the advantages of the micro material analysis, using the theory of plasticity, as well as the interaction among the different constituent materials were accounted for within the macro finite element model.</p> <p>Utilizing the finite element model, parametric studies were conducted to cover wide ranges of the parameters which affect behaviour. The geometric, loading, and the material parameters as well as their interactions were involved in this study.</p> <p>Three simplified methods were developed to provide some rational design approaches. The first method, called "Lumped Area Method", provides a closed form solution for the wall capacity. In the second method, Moment Magnifier Method, a rational method for determining the modulus of rigidity was empirically developed to be used in the calculation of the moment magnification factor. In the third method, Deflection method, the theory of structural mechanics was used to determine the secondary moments using a secant modulus of rigidity which was numerically developed.</p> <p>Experimental creep tests were carried out on ungrouted and grouted masonry prisms built with normal weight blocks, and ungrouted prisms built with light weight blocks. Creep data were collected for almost 400 days under two levels of sustained stresses. Creep functions were statistically developed and were incorporated in the finite element-creep analysis.</p> <p>Another set of parametric studies was conducted to investigate the long term capacities of masonry walls. Also, the Lumped Area Method was extended to account for the creep effect in predicting the long term capacity.</p> / Doctor of Philosophy (PhD)

<p>A mathematical model to predict the bed sediment and suspended sediment load as a function of time and distance along a one-dimensional quasi-steady receiving area is described. The model called TOTSED, uses the modified. Einstein approach for bed load calculations and the Einstein approach for suspended load calculations. The plug flow reactor is used for the inflow load (sediment pollutograph) calculations.</p> <p>The model was verified and calibrated using data obtained from a sampling program carried out in Chedoke Creek outfall channel in Coote's Paradise in Hamilton during summer 1980. Similar computations were carried out using the HEC-6 program. TOTSED predictions compare favourably with both HEC-6 and the observed data. TOTSED has been modified to be a new block of the SWWM program.</p> / Master of Engineering (ME)