As part of a research program at McGill University, involving the testing of eight full-scale deep beams, this research studies the response of four full-scale deep beams that were designed in accordance with the strut-and-tie model provisions of the CSA Standard A23.3 (2004). All of the deep beams had a centre-to-centre spacing of 2000 mm between the two supports, 1000 mm in height, and 350 mm in thickness. They were simply supported with an application of a single point load at midspan. The total length of the four beams varies depending on the type of anchorage being used. The four anchorage details that were considered in this study include a straight development length, a standard 90 degree hook, friction-welded 1.5 in. (38 mm) diameter circular headed bars and friction-welded 2 in. (50 mm) diameter headed bars. The dimensions of the loading pad and the bearing pads were 200 mm (and 300 mm) and 100 mm, respectively. All beams were reinforced with two layers of five 15M bars forming the tension tie reinforcement. El-Jorf (2006) carried out the testing and analysis of the other four full-scale deep beams in this overall research program. / This research project demonstrated that providing friction-welded circular headed bars capable of developing the full yield strength at the head results in shorter beams and improved response compared to straight bar embedment details. Providing a lead-in length as well as friction-welded circular headed bars, with reduced head size, provides improved ductility compared to the straight bar embedment and compared to the hooked anchorage. This research also shows that benefits of confinement pressures at the supports can increase the bar stress although the provided development length is below code requirements.

The buckling behaviour of fibre reinforced polymer (FRP) sheet pile panels subjected to a uniform lateral pressure was investigated. Based on the previous full scale tests by Shanmugan in year 2003 (Shanmugan, 2004), the critical load at buckling initiation was first determined through experimental data analysis, and the theoretical modeling was then followed in an attempt to predict the buckling initiation and understanding the failure mechanism. The behavior of the panels loaded in upright position and inverted position was studied. / The local buckling of the compressive flanges was monitored by the strain measurements, which demonstrated that when tested in upright position, the panel failed immediately after local buckling of compressive flange, and when tested in inverted position, the panels could be able to carry the load into post buckling region. The stresses and corresponding axial forces at buckling were calculated by the classical beam flexure formula but taking into consideration the reduction of flexure rigidity and neutral axis shifting. The axial force calculated from the beam flexure formula was comparable with that from stain gauge measurements. The axial force was not uniformly distributed along the width of the compressive flange at upright position and was about zero at the free edge. When tested in inverted position, the neutral axis distance and the flexure rigidity kept almost as a constant. The sheet pile panels were with a uniform axial force along the width of the compressive flange. / An analytical modeling was performed to predict the buckling initiation. The buckling of the panel was simplified as the buckling of the compressive flange with various boundary conditions. The differential equation of the compressive flange was established based on the assumption that the flange was subjected to an in-plane axial force and an out-of-plane lateral pressure simultaneously. It was found that the lateral pressure did not have direct effect on the critical load. It was the compressive axial force that determined the local buckling of the flange. Kollar's explicit expressions were also applied but only valid for long plate loaded by uniform axial force. / The buckling load obtained by solving the differential equation for the inverted panel compared well with that from the experimental results. However, for the flange in a pile at an upright position, the theoretical prediction was far less than the experimental value which might be attributed to the non uniform axial force on the flange. Energy method was applied to estimate the range of the buckling load of a plate loaded by a linearly distributed axial force. The upper bound value was obtained from fixed boundary condition and the lower bound from simply supported assumption. The experimental result was found in between the two bounds and was in favour of the lower bound as a conservative estimation of critical load for upright panel.

The objective of this thesis was to determine, by means of discrete element simulations of ice floes, information that could be used for continuum modelling of broken ice covers on waterways. The ice floes or 'particles' were modelled as random-sized circular disks. The internal angles of friction for cohesionless assemblies of particles under compression were determined using molecular dynamics type simulations of biaxial tests. Constant confining pressures as well as constant displacement rates were used for the biaxial tests. Stress-strain curves were obtained for a range of confining pressures. The yield envelopes were found to be represented by a Mohr-Coulomb failure criterion. The effect of the interparticle friction coefficient on the global angle of friction was examined. It was found that the internal angle of friction was essentially independent of the interparticle friction angle. Furthermore, assemblies with different particle size distributions were used in order to investigate the effects of this parameter on the internal angle of friction. It was found that as the particle size distribution is made wider, the magnitude of the internal angle of friction increases. Angle of repose simulations were performed in order to verify the internal angles of friction results. It is found that the angles of repose measured are in rough agreement with, but slightly higher than, the values found for the internal angles of friction in the biaxial tests. Finally, biaxial tests were performed on assemblies in which bonds were introduced between the particles in order to simulate an intact ice cover. Results were consistent with a cohesive Mohr-Coulomb material.

The study of flow of water through porous medium is important in geotechnical and geoenvironmental engineering. For determining the in situ hydraulic conductivity characteristics of porous media, one of the most convenient and popular methods is the cased borehole technique. In this method, water is allowed to enter the excavated region of the cased borehole inserted into the ground. The rate of rise of water in the casing is then used to estimate the hydraulic conductivity characteristics of the porous medium in the vicinity of the entry point. When the porous medium has isotropic hydraulic conductivity characteristics, the "intake shape factor", which influences the flow rate is solely dependent on the geometrical arrangement of the intake region. When the intake region is located in a porous medium with transversely isotropic hydraulic conductivity characteristics, the flow rate is influenced by both the geometrical characteristics of the intake region and the mismatch in the directional hydraulic conductivity. The objective of the thesis is to investigate certain aspects of water flow in porous geomaterials that display hydraulic anisotropy. The thesis also characterizes the flow rate for entry points that are located in soils with transversely isotropic hydraulic conductivity. / This study presents the application of computational procedures, based on finite element techniques, for the determination of the intake characteristics of a cylindrical intake located in a hydraulically transversely isotropic porous medium. Numerical results presented, illustrate the situations where the separate hydraulic conductivities can be estimated by a suitable alteration in the geometrical characteristics of the cylindrical intake. An approximate relationship is developed for estimation of the intake shape factor of cylindrical entry regions without a repeat of detailed finite element computations.

A methodology to evaluate structural reliability of existing bridge structures using advanced reliability FORM/SORM (First Order Reliability Method/Second Order Reliability Method) methods is presented. The methodology is applied to evaluate the structural reliability of bridge steel members subject to compression, tension, shear and bending. The steel truss members, floor beam and stringer of a typical span located on the Jacques-Cartier Bridge in Montreal are evaluated using this methodology. Data for the evaluation was obtained from a comprehensive inspection and sampling program on the bridge involving tests on 74 steel coupons. The results of the evaluation are then compared to the results of the evaluation using the load and resistance factors method and the mean load method outlined in the draft version of the CHBDC, 1998, (Canadian Highway Bridge Design Code). / In this application, the advanced reliability analysis indicates that the estimation of the reliability of bridge members, in their current condition, using the loads and resistance factors method specified in the code may be too conservative. This situation was encountered in the case of the floor beams of the analysed span, when the failure mode considered was shear. The advanced reliability analysis indicates also in this application, that the evaluation of the reliability of bridge members, using the mean load method specified in the code may not be conservative. This situation was encountered in the case of the truss members of the analysed span, when the failure mode considered was compression. / The evaluation of bridge structures using advanced structural reliability methods can be used to obtain more accurate estimates of their reliability. This information can be used to develop better, monitoring, inspection, maintenance, and rehabilitation strategies for bridge members.

Influence of Detailing on Response of Dapped End Beams Abstract This research investigated the behaviour of disturbed regions in dapped end beams. In this experimental research programme, two dapped end beams with similar geometry were designed and tested. Each had a span of 3 m, but each of the four different dapped ends utilized a different reinforcing detail. The main parameters studied in this research were the anchorage of the hanger reinforcement and the flexural reinforcement. The dapped ends were detailed in accordance with the provisions of the 2004 CSA Standard A23.3, as well as the 1971 and 1999 PCI Design Handbooks. Strut-and-tie models were used for the design and strength predictions. The results from the experimental programme are compared with strut-and-tie model predictions. It was concluded that: (1) proper anchorage is essential; (2) the anchorage and the details of hanger steel and longitudinal reinforcement have a great influence on shear capacity and ductility; (3) design using strut-and-tie models for dapped end beams provides a conservative approach; and (4) the early versions of the 1971 and 1999 PCI Design Handbooks gave poor design and detailing requirements. / Influence de la modélisation sur le comportement de poutres à extrémités emboutiesRésuméCette étude a porté sur le comportement de poutres à extrémités embouties sous contrainte. Dans le cadre de ce programme de recherche expérimentale, deux poutres à extrémités embouties ayant une géométrie similaire ont été conçues et mises à l’essai. Chaque poutre avait une portée de trois mètres, mais l’armature de chacune des quatre extrémités embouties était différente. Les principaux paramètres étudiés dans le cadre de cette recherche étaient l’ancrage principal l’armature de suspente et l’armature de flexion. Les extrémités embouties ont été modélisées conformément aux dispositions de la norme ACNOR A23.3 2004 ainsi que des manuels de conception du PCI. Le modèle de contrefiches et de tirants a servi aux prédictions de conception et de résistance.Les résultats du programme expérimental sont comparés aux prédictions du modèle de contrefiches et de tirants. Il est conclu que : 1. un ancrage adéquat est essentiel; 2. l’ancrage et les détails de l’acier de suspension et l’armature longitudinale ont une grande influence sur la résistance au cisaillement et la ductilité; 3. la conception utilisant des modèles de contrefiches et de tirants pour des poutres à extrémités embouties représente une approche conservatrice; et 4. les exigences en matière de conception et de modélisation enchâssées dans les versions antérieures (1971 et 1999) des manuels de conception du PCI étaient peu contraignantes.

A thorough review of soil-water energy theories and soil-moisture measurement techniques are presented. Monitoring of piezometric head and soil-moisture potential and content at 3 vertical profiles was conducted along a cross section through Waite Amulet tailings, Quebec. Results characterized the subsurface hydrology during the fall of 1990, a recharge period, which shed light on the hydrologic controls of tailings oxidation. / Infiltration rates, wetting front velocities, hydraulic conductivities, and field capacity were determined. The presence of a tree cover increases infiltration by a factor of three. Recharge peaks reflected the conversion of tension-saturated to pressure-saturated porewaters. And, a water balance showed soil-moisture storage was 53% of precipitation, saturated-zone recharge 31%, evapotranspiration 12%, and runoff 4%. / The reliability of time-domain reflectometry to predict gravimetric water content of oxidized tailings was 88%. Heat dissipation sensors detected small dial fluctuations in response to evapotranspiration. Gypsum blocks proved valuable in detecting wetting fronts.

Elastic as well as plastic buckling of circular cylindrical shells filled with a core material is analyzed under axial compressive loading. A practical example of this situation is the buckling of concrete filled steel tubular (CFT) columns used widely in high-rise buildings. The theoretical problem is modeled as the bifurcation buckling of a perfect "infinitely" long circular cylindrical shell under uniform compression, constrained by a one-way (tension-less) foundation. An important and useful novelty is that the shell material is allowed to undergo strain-hardening plasticity before buckling. For simplicity, the core material is assumed to remain elastic. The approach is analytical. The governing equations are solved exactly to obtain buckling loads, and wavelengths in contact and nocontact regions. The theoretical results, when applied to CFT columns, are found to be in very good agreement with the experimental buckling loads of other researchers. / Le flambage élastique ainsi que plastique de coquilles cylindriques remplis d'un matériel sont analysé sous le chargement de la compression axiale. Un exemple pratique de cette situation est le flambage de colonnes tubulaires en acier remplis de béton (CFT) qui sont largement utilisées dans les immeubles de grande hauteur. Le problème théorique est modélisé comme le flambage par bifurcation d'une coquille parfaite cylindrique de longueur "infinie" sous la compression uniforme, en présence d'une contrainte à sens unique. Une nouveauté importante et utile est que le matériel de coquille est sollicitée dans le domaine post-élastique avant de flambement. Pour simplifier, le matériel remplis est supposé rester élastique. L'approche est analytique. Les équations régissants sont résolus exactement à obtenir les charges de flambage, et les longueurs d'onde en contact et sans contact régions. Les résultats théoriques, aux applications de CFT colonnes, se trouvent en très bon accord avec des charges de flambage d'autres chercheurs.

Full-scale exterior beam-column-slab sub-assemblages were tested under reversed cyclic loading to investigate the role or the spandrel beam in the overall response. Results from specimens having different spandrel beam sizes and different amounts of torsional reinforcement in the spandrel beams provided a better understanding of their behaviour. The test specimens were heavily instrumented to enable detailed strain measurements in the slab bars, the joint region and the spandrel beam. These strains together with the crack pattern provide some insight into the flow of forces from the slab and spandrel beam into the joint region. The tests show that the effective width of the slab contributing to the negative bending of the main beam is affected by the torsional yielding of the spandrel beam. However, after this yielding, a different force mechanism, involving concrete compressive struts, and tension ties in the slab and spandrel beam bars, provides means of transferring additional forces from the slab bars to the joint region. It is important to consider the effect of these forces since they increase the negative moment capacity of the beams, and hence, they may affect the hierarchy of yielding between the columns and the beams. This effect may in turn alter the failure mode of the structure. / Tentative design proposals, accounting for the flow of forces through the spandrel beam, are presented. The application of these proposals are illustrated for a number of different specimens having varying spandrel beam dimensions.

The concept of soil buffer capacity was used in this study to investigate the capacity of soil to attenuate heavy metals when acid is involved in the soil system. The buffer capacity of soil in this study was found to depend mainly on carbonate content and cation exchange capacity (C.E.C.) of soils. The magnitude of buffer capacity followed the order: illite $>$ montmorillonite $>$ natural clay soil $ gg$ kaolinite. / The study of heavy metal retention in soils was performed both by soil suspension test and soil column test. The results showed that as soils received increasing amounts of acid, high amounts of heavy metals (particularly Pb and Cu) could still be retained if the soils had a high enough buffer capacity to resist a change in pH such that it will not drop to $ Cu > Zn > Cd.$ The order changed to $Pb > Cd > Zn > Cu$ when soils were at low soil solution pH. Relative mobility of heavy metals found from the soil column test followed the order $Pb < Cu < Zn leq Cd.$ / The relation of soil buffer capacity and heavy metal retention and movement in the clay soils found from this study revealed that the soil buffer capacity was a parameter that can be used in the prediction and prevention of heavy metal migration in soil. The soil buffer capacity is recommended as a parameter that should be included in the determination of soil properties particularly for the purpose of land application and disposal of wastes with leachates that could contain heavy metals. (Abstract shortened by UMI.)