<p>This thesis presents a comprehensive approach to numerical modelling of the nonlinear behaviour of structural masonry. Masonry is a heterogenous material, which displays orthotropic symmetry. The anisotropy effects are described here by incorporating a set of distribution functions specifying the directional dependence of material properties.</p> <p>In the first part, a macroscopic failure criterion for structural masonry is proposed. This criterion is derived within the framework of the critical plane approach. First, a general discussion is provided examining the performance of this framework within the context of both classical linear and nonlinear criteria. Subsequently, a bi-linear form of failure criterion for strucutral masonry is proposed. Extensive numerical study is performed examining the behaviour in biaxial compression-tension and compression-compression regimes for different orientations of the sample relative to the loading direction. The results are compared with the available experimental data.</p> <p>In the next part, the results of a 3D seismic analysis of the masonry walls of a power substation building - typical of those constructed in the elastic range, assuming orthotropic material properties, and the admissibility of the stress field is assessed based on the proposed failure criterion. A numerical study is performed examining the effect of different reinforcement strategies.</p> <p>The focus of the last part of this thesis is on the description of progressive failure in structural masonry. A continuum formulation is developed here which is applicable to a representative volume comprised of a large number of units interspersed by mortar joints. The framework defining the conditions at failure when employing the critical plane approach is extended to model the inelastic deformation process. This is accomplised by incorporating a multi-laminate approach in which the average response is derived from sliding/separation characteristics along a set of randomly distributed planes. The localized deformation is descrived by considering a structured medium comprising the intact masonry intercepted by a distinct macrocrack. Extensive numerical simulations are performed examining the response of brickwork in compression/tension regimes at different orientations of the bed joints relative to the loading direction. A boundary-value problem which involves an inelastic finite element analysis of a bearing masonry wall subjected to in-plane loading is also studied.</p> / Doctor of Philosophy (PhD)

Master of Engineering (ME)

<p>Traffic engineers rely heavily upon the fundamental relationships between the variables speed, flow and capacity which characterize the traffic stream. Accurate knowledge of the operational behaviour of roadway traffic inevitably enhances the reliability of planning and design. Because of the importance of basic relationships to traffic engineering, this study was undertaken. The study involves investigating the important relations between speed, flow and capacity currently existent for uninterrupted flow on multilane highways in Ontario. The results are compared with information established by both the Highway Capacity Manual (HCM) and Polytechnic Institute of New York (PINY). Comparisons are made of 5-minute and 15-minute hourly flow rates, and calculated truck equivalents with those currently recommended. In addition, the impact of adverse weather conditions on fundamental speed-flow relationships and how occupancy (density) relates to speed and flow were investigated.</p> <p>Results of the study indicate that there is a significant difference between relations presently in use and those existing on roadway facilities. In particular, capacity was found to be in excess of 2000 passenger cars per hour and speeds were noted as being higher than established values.</p> / Master of Engineering (ME)

<p>A research program into the investigation of the behaviour of double chord HSS Warren trusses is presented. The experimental results of eleven tests on five different types are reported; these include two Back-to-Back trusses, two Standard trusses and a Bolted type. One of the Back-to-Back trusses employed stiffening plates to reinforce gapped connections while the other had web members fully overlapped. The two Standard trusses had different eccentricities depending on whether the ends of the web members were square cut or angle cut. Gusset plates and tie plates were used to stiffen the connections for the Bolted truss. Retreats after reparations were undertaken in the event of a localized joint or member failure so that maximum information could be obtained from the program.</p> <p>An analytical model has been developed and incorporated into an existing plane frame program for analysis of the double chord HSS trusses. Three types of yield mechanisms that are accounted for are plastic hinge formation at the end of a member, member failure due to plastic limit load and yielding of the spring for modelling a connection.</p> <p>The experimental and analytical results are then compared thus confirming the validity of the analytical model. Finally, conclusions and recommendations are outlined for the analysis, design and feasibility of double chord Warren trusses with hollow structural sections.</p> / Master of Engineering (ME)

<p>This thesis studies the effects that nonproportional damping has on the vibrational characteristics of two and three level systems. In additions, error levels in response prediction are estimated for two simplified time history procedures and a response spectrum technique. It is shown that nonproportional damping creates significant changes in the damped frequency, damping ratio and mode shape of each mode of vibration of the system relative to those found for a similar system exhibiting proportional damping. The study of the prediction of the maximum system response to seismic basic motions demonstrates that there exist three regions where different dynamic analysis procedures should be used. A physically uncoupled analysis should be used for large mass ratios. A complex modal analysis should be used if the mass ratio is of an intermediate value. Diagrams are supplied which allow one to readily determine which of the three methods is the most appropriate for a design.</p> / Doctor of Philosophy (PhD)

<p>A large number of low and medium-rise buildings have steel moment resisting frames as the primary lateral load resisting system. During the past few decades, much confidence has been placed on this type of structural system for resisting seismic loads. However, after recent earthquakes (e.g. Northridge, California, in 1994, and Kobe, Japan, in 1995) the confidence in this system was reduced as a result of various types of damage that moment resisting steel frames suffered. This resulted in a recognition of the need to evaluate the performance criteria on which current provisions are based. While there have not been any major casualties dueu to earthquakes in Canada during the past few decades, in fact there is an actual seismic hazard which affects significant regions of the country, for example, the cities of Victoria, Vancouver, Quebec City, Montreal, and Ottawa. The design peak ground motions in such regions are moderate in comparison with those in California or Japan, however the uncertainties associated with estimating the expected ground motions are such that twice or three times the seismic design level motions are likely to occur. The main objectives of this research study are: (i) to evaluate the seismic level of protection afforded to steel moment resisting frame building structures designed in accordance with the current Canadian provisions (i.e. NBCC (1995) and CAN/CSA-S16. 1-94), and (ii) to investigate the effect of the different design philosophies and seismic hazard design levels on the inelastic dynamic response of multi-storey steel frame structures. Six storey office buildings located in regions of high, intermediate, and low seismic hazard, and a ten storey office building located in a region of intermediate seismic hazard are designed in accordance with the current Canadian provisions using three design philosophies, namely strong-column weak-beam (SCWB), weak-column strong-beam (WCSB), and strong-column weak-panel zone (SCWP). The scope of the research program includes: (a) modelling of the structural elements; (b) nonlinear push over static analyses/inelastic dynamic analyses, and (c) evaluation of the damage potential associated with each design. In the study analytical models are modified and incorporated into the PC-ANSR computer program in order to perform the inelastic dynamic analyses of the frames. The inelastic models take into account the spreading of inelastic deformations in beam-column elements, connection flexibility and panel zone deformations. A cyclic model for the panel zone element is developed and introduced into PC-ANSR. The performance of the frames is evaluated both statically using monotonically increasing lateral load (nonlinear push over static analyses), and dynamically by subjecting the inelastic model to an ensemble of actual strong ground motion records (time-history analyses). The main ensemble of time-histories used in the study consists of twelve earthquake ground motion records selected on the basis of Newmark-Hall design spectra amplification factors. An additional ensemble of time-histories (twelve records) is selected based on the uniform hazard spectrum for Vancouver which describes the new seismic hazard information given by the Geological Survey of Canada. The additional ensemble is used to investigate the implication of the new seismic hazard information on the performance of the six storey frames in the intermediate seismic hazard region. The results of the inelastic dynamic analyses are presented in terms of statistical measures of the maximum response parameters determined during the time-history analyses. Also, the results of the nonlinear push over analyses are presented and compared with those of the dynamic analyses. The performance expectations of the frames are evaluated in order to assess both the overall level of protection provided to the frames and the preferred design philosophy. It is concluded from the analyses that in high and intermediate seismic hazard regions, a well-designed and detailed ductile moment resisting frame (i.e. SCWB or SCWP) can withstand ground motions of twice the design level with a very little likelihood of collapse, while an ill-conditioned designed frame (i.e. WCSB) may develop a collapse mechanism even at the design level excitation. In regions of low seismic hazard activity, the three frame design types perform satisfactorily, and can withstand twice the design level excitations with only a moderate amount of damage.</p> / Doctor of Philosophy (PhD)

<p>Continuous hydrology packages such as Hydrologic Simulation Program--Fortran (HSPF) and Version 3 of the Storm Water Management Model (SWMM3) require extensive computing time even for certain cases of event modelling. In their original sequential, time-sharing, computing schemes, continuous modelling applications required a prohibitively expensive overall turn-around time and an overwhelming amount of manual Input/Output (I/O) time series (TS) data management effort. By decentralizing the processing, avoiding a time-sharing computing environment, and finding ways to keep track of I/O TS data as part of the processing, these problems can be overcome.</p> <p>Computational Hydraulics Group Time Series Manager (CHGTSM), a Database Management System (DBMS), was developed by the present author to provide easy access to TS data, independent of details of storage. CHGTSM is based on an unconventional data access technique, also developed by the author, that handles variable resolution continuous TS records. The CHG Time Series Store (CHGTSS) is prepared by the CHGTSM. CHGTSM is applied succesfully. In a case study, CHGTSM saved 88% in hard disk storage compared to the raw database. CHGTSM serves event as well as continuous modelling applications. A manual for CHGTSM is also written by the present author.</p> <p>CHGTSM can be used to distribute the database to the nodes of a linked configuration of microcomputers. Distributed processing (DISP) of centralized data is made possible by CHG Distributed Data Processing Software (CHGDPS). CHGDPS, developed by the present author, is a shell around CHGTSM which includes security provisions and node synchronization assistance. The simulation of a DISP application for continuous hydrologic modelling in a Local Area Network (LAN) improved the computing efficiency by 54% compared to sequential processing. CHGDPS will allow true concurrency only through a batch application.</p> <p>CHGDPS synchronization assistance relies on timing relations of the computational modules that are accessing the system. A program measurement (PM) technique was adopted, enhanced and applied by the present author to obtain the timing relations. Such DISP control helps the user to optimize overall system processing. For example, CHGDPS may decide (or help decide) a shift from one computational approach to another for considerations of time, space, numerical stability, and accuracy.</p> <p>The DISP scheme is flexible and can be implemented on any linked configuration of computers e.g. LAN' s, distributed data gathering and real time control (RTC) devices. The whole system, called the Computational Hydrology Work Group System (CHWGS) is assessed as an environment for continuous (or event) hydrologic modelling, distributed data gathering and distributed RTC.</p> / Doctor of Philosophy (PhD)

<p>In order to assess the structural charateristics of a pavement-subgrade system, non-destructive, in-situ tests together with backcalculation procedures are widely used. Traditionally, the analytical models adopted for this proccess are deterministic, however, in reality, the quantities involved in the problem may be random variables. Neglecting the variable nature of the system parameters, e.g., highway material properties, may affect the reliability of the pavement response prediction. On the other hand, inverse solutions to pavement problems are often ill-conditioned and sensitive to the input parameters. Past experience has shown that the estimated values of a blackcalculated parameter by different agencies may vary by several orders of magnitude, representing a high level of uncertainty in the estimated paramter. Unless the uncertainty is quantified, practitioners are forced to resort to higher safety factors, which is neither economical nor always conservative. The present study investigates, rigorously, the behavior of a pavement-subgrade system from a stochastic point of view, and addresses the sensitivity of response variation to variations in layer properties. The results of a forward analysis are utilized to establish a relation between input and output statistical moments in order to interpret the pavement deflection data stochastically. The proposed framework in this research allows one to quantify the uncertainty level in backcalculated system parameters. It also provides a tool to infer the accuracy of the pavement performance prediction based on mechanistic models. For the purpose of introducing the stochastic approach, the perturbation technique is applied to an idealized, two-layered, pavement-subgrade system for the case of: (a) a static solution based on Odemark definition of equivalent layer thickness; and (b) a frequency domain solution to a single degree of freedom (SDOF) system using an impedance function. The methodology is then extended to a stochastic finite element framework in order to analyze boundary-valued problems of more complex geometry and distribution of material properties. The perturbation method is a mean-based, second-moment analysis for the second-order accurate expected value, and first-order accurate cross-covariance function. For the dynamic analysis, viscoelastic response of the pavement is obtained by using the periodic-load analysis approach and Fourier synthesis. Based on the results of the simulations, it is demonstrated that, the sensitivity of surface deflections is significantly higher to the subgrade properties than those of the surface and base layers, both in a static and a dynamic analysis. Consequently, it is concluded that, the low dominant frequency of the falling weight deflectometer (FWD) load limits the capability of this test in characterizing surface layer properties. Using the concept of coefficient matrix, it is illustrated that, the low sensitivity of deflections to surface layer properties can be interpreted as a high level of uncertainty in the estimated pavement moduli in a backcalculation exercise. It is indicated that uncertainties in backcalculated parameters often result in an unacceptable pavement performance prediction. Moreover, the physical behavior of the layers are identified by finding the contribution of each layer to the total deflection of the system using the notation of contribution ratio.</p> / Doctor of Philosophy (PhD)

<p>A waste form can be prepared by mixing a hydraulic cement and, if needed, a bulking agent with an aqueous waste to cause it to solidify. A mechanistic leaching model was developed based on describing the chemistry of the waste form-Ieachant system and the flow regime of the leachant and assuming that transport takes place via diffusive exchanges through the waste form-Ieachant Interface. This model was successful in predicting Ieaching from simpIe waste matrices and in identifying important containment mechanisms effective in the more complex matrices.</p> <p>The cement-based waste forms studied had porosities ranging from 40 to 60%. Portland cement provided acid neutralization capacity to maintain the high pH environment where the waste form is stable; a typical waste form contains enough cement to neutralize between 2000 to 3000 times its volume of a pH 3 leachant.</p> <p>In a mild environment, leaching was controlled by the diffusion of the soluble fraction of a contaminant present in the connected pores of the matrices. Immobile species instantaneously solubilized to maintain chemical equilibrium between the soluble and insoluble fractions. In tests conducted over a period of almost 2 years, less than 1% of the initial amount of cadmium, chromium and lead contained in a specimen leached out.</p> <p>In an acidic environment, the leaching rates were limited by the availability of acid to dissolve the matrix. The leaching process was similar to a surface corrosion process. However, since waste matrices are not completely soluble in a mild acid, a leached layer develops, eventually protecting the core of the waste form from direct contact with aggressive groundwaters. Contaminants solubilized at the leaching front are subjected to concentration gradients that force them to diffuse both inward and outward. Only a fraction of the total concentration therefore leaches out.</p> <p>The mechanistic knowledge developed through the experimental programme and the modelling effort was used to develop eight long term leaching scenarios covering a wide range of waste form and groundwater conditions. In addition to allowing the prediction of leaching rates for periods of up to 100 years, these scenarios were useful in developing recommendations to prepare more efficient waste forms and design better landfills.</p> / Doctor of Philosophy (PhD)

<p>Pollution in the coastal zones of the Great Lakes has become more serious in recent years. This is due to increased use of coastal water as a result of population and industrial growth. A substantial portion of the contaminants that enters a lake do so from the shoreline via discharges from sewer overflows, industrial outfaIIs and runoff. Such discharges contain particulates and other materials of density greater than that of lake water. Many heavy metals with toxic components are present in these fractions. The dynamic behavior of these particIes in the coastal and offshore waters is thus of great importance. The principal removal processes for these materials are transport and particle settling. An understanding of the characteristics of nearshore currents, diffusion and temperature patterns is essentiaI to determine their effect on removal processes, and in turn, on coastal biological and chemical processes. This study is limited to the physical fluid mechanics of coastal zones.</p> <p>The structure of the nearshore flow in the vicinity of Cleveland, Ohio is analyzed in detaiI in this study. The impact of Cleveland, one of the largest urban and industrial agglomerations on the shoreline of Lake Erie, in terms of additional loading is thought to be considerable. A computer program (ADVDIFF) was developed to calculate the mean flow, horizontal turbulent length and time scales, horizontal diffusivities and kinetic energy. ADVDIFF uses filtering techniques, spectral analyses and statistical analyses. Five episodes representing three different flow regimes which may exist in the coastal zone were chosen for special analyses.</p> <p>To generate the coastal currents, a rigid-lid, channel-type model with fine grid size in the coastal zone was used. A model originally developed by Simons (1983) was modified to incIude nonlinear acceIeration terms and two different forms of the vertical eddy viscosity. Also, a two dimensional x-y model developed by Simons and Lam (1982) was modified and used to expIain some of the observations. Both new modeIs (ERCH, ONELAY) were verified, calibrated and applied to Lake Erie.</p> <p>A computer program (SEDTRAN) was developed to predict the inflow sediment concentration distribution within the coastaI zone. SEDTRAN soIves numericaIIy the three dimensionaI time-dependent mass transport equation including the settIing term. The modeI uses the currents and diffusivities computed by ERCH and ONELAY and the statistical analyses, respectively. SEDTRAN was verified using several test examples, and partiaIIy vaIidated using the avaiIable data set. The modeI was applied to many cases of settIing activity that may take place in the coastaI zone. The results were used to define a representative influence zone for a pollutant source at CIeveIand.</p> / Doctor of Philosophy (PhD)