The suspended cohesive sediments, mostly clay minerals with negative surface charges, usually interact through an array of intermolecular/surface forces (e.g., Coulomb force, van der Waals attractions, and hydrogen bonding) with waterborne organic matter (e.g., exopolymers) and dissolved salts, resulting in fractal, tenuous, and hybrid organic-inorganic flocs or marine snow. A pilot and systematic laboratory and theoretical work have been conducted towards a synthetic nanoscale understanding of the flocculation kinetics and mechanics of the clay-exopolymer micro-sized flocs. This systematic study has achieved two impressive objectives: (1) the effect of exopolymer polarity and concentration on the particle size kinetics of clay-exopolymer mixtures has been investigated, based on extensive flocculation experiments involving four clay minerals (i.e., kaolinite, illite, Na-montmorillonite, and Ca-montmorillonite), three exopolymers (i.e., xanthan, guar, chitosan), six exopolymer to clay weight ratios (i.e., 0, 1, 2, 5, 10, and 20 wt.%), and three hydrodynamic conditions (i.e., laminar, transitional, and turbulent flows); (2) the floc mechanical properties affected by salinity and exopolymers have also been investigated by means of extensive floc nanocompression testing involving the four aforementioned clay minerals, four neutral exopolymer (guar) to clay weight ratios (i.e., 0, 2, 10, and 20 wt.%), and four salinities (i.e., 0, 2, 10, and 30 PSU). Results from flocculation experiments indicate that both the pure clay and clay-exopolymer mixture suspensions possess discrete particle groups consisting of primary particles, flocculi, microflocs, and macroflocs that exhibit subordinate unimodal lognormal distributions. There exists a critical exopolymer to clay ratio that can lead to a maximal or minimal fraction of microflocs or macroflocs. Meanwhile, the highly scattered floc mechanical properties led to the theoretical finding that both flocs elasticity and yield shear strength satisfy Weibulls Law. These findings are of great importance to geotechnical engineering.

Louisianas saltmarshes, one of the most productive wetlands in the nation, are undergoing severe erosion due to hydrodynamic forces (storm surge) and sea level rise. The erosional behavior of coastal saltmarshes, which possess cohesive sediments as their integral components, are very difficult to analyze and understand. The erosional resistance of cohesive sediments determines the stability and sustainability of coastal wetlands. This erosional resistance is expressed as an erosion threshold or critical shear stress, which depends on various soil properties (e.g. water content, root content, organic matter content, and pore water salinity) affecting saltmarsh erodibility integratively. A cohesive strength meter was deployed to measure the critical shear stress in Bay Jimmy (Barataria Bay), an oil spill site, and Terrebonne Bay, a normal saltmarsh. Results show that erosion threshold is positively correlated with organic matter and root contents. In fact, both roots and organic matters assist with the aggregation of mineral particles through bridging effect. However, when the organic matter content exceeds 10%, the critical shear stress decreases with organic matter content, as observed for the Terrebonne Bay sediments. The erosion threshold of the sediment surface of Bay Jimmy was found to be less than that of Terrebonne Bay, which may be attributed to the lower pore water salinity and possibly reduced inter particle binding due to residual crude oil retained by the cohesive sediments. However, the higher root content at depth in Bay Jimmy ensured greater stability than that found in Terrebonne Bay. Therefore the possible critical mechanism of erosions in Bay Jimmy and Terrebonne Bay are surface erosion and undercut erosion respectively.

Sustainable treatment of chlorinated ethanes and ethenes contaminated groundwater using vertical flow engineered wetland systems were investigated in microcosm and column studies. Experiments on environmental and biogeochemical factors that affect system performance were conducted, and a numerical model involving advection, sorption, and sequential biodegradation was developed to describe the fate and transport of the contaminants of concern in the treatment wetland bed. 1,1-dichloroethane (1,1-DCA) and cis-1,2-dichloroethene (cis-1,2-DCE) were used as the chemicals of interest. The presence of cis-1,2-DCE inhibited dechlorination of 1,1-DCA but cis-1,2-DCE dechlorination was not affected by the presence of 1,1-DCA. Simulation runs showed that the treatment bed sizing was controlled by the 1,1-DCA dechlorination kinetics. Bioaugmentation and biostimulation amendments lead to higher dechlorination rates of both cis-1,2-DCE and 1,1-DCA. Studies conducted with different amounts of peat and sand mixtures to investigate long term effects of organic carbon depletion in engineered wetland systems showed complete biodegradation of 1,1-DCA in all soil mixtures with no significant difference in the rate constants. However, simulation runs showed larger bed size requirement for the lowest amount of peat soil used (5% peat) compared to the other peat soils (25%, 50%, and 100%), and no significant difference in the treatment bed size between the 25%, 50%, and 100% peat soils. Complete biodegradation of cis-1,2-DCE and 1,1-DCA was observed in treatment systems incubated at 10oC, 15oC, and 25oC. However, reduced temperatures resulted in lower dechlorination rates. Maintaining the soil and groundwater pH of an engineered wetland system to near neutral pH by applying alkaline solution was observed to be necessary for biodegradation to occur. The potential for plant assisted remediation of 1,1-DCA through the root system of Scirpus americanus indicated possible plant uptake and enhanced system performance. Microbial analysis of the treatment media using quantitative polymerase chain reaction (qPCR) technique, confirmed the presence of Dehalobacter sp. and Dehalococcoides sp. as well as the functional genes bvcA and tcrA reductase known to mediate the biodegradation of chlorinated ethanes and ethenes to non-toxic end products.

A continuous-flow system composed of two anaerobic bioreactor (ABR) columns (based on engineered wetland systems) connected in series was tested to observe the ability of such a system to treat multiple volatile organic compounds (VOCs) possessing suspected inhibitory effects on the dechlorination of other compounds with cis-1,2-dichloroethylene (cDCE) and 1,1-dichloroethane (1,1-DCA) as the main test chemicals. Following the establishment of cDCE and 1,1-DCA dechlorinating ability at baseline conditions, system perturbations of breakthrough and starvation were conducted to measure the response and recovery of the system dechlorination. In all column experiments, it was observed cDCE dechlorination inhibited the dechlorination of 1,1-DCA and that 1,1-DCA dechlorination proceeded only when cDCE had been degraded to ethene. The largest 1,1-DCA dechlorination was observed near the effluent of the second ABR due to conditions favoring a higher population of 1,1-DCA degrading bacteria. Core samples taken near the influent and effluent zones of the second ABR were tested on the ability to dechlorinate cDCE and 1,1-DCA to investigate the spatial distribution of dechlorinating bacteria as a function of depth and to determine the possible mechanisms behind the inhibition of 1,1-DCA degradation. From the core samples, a relationship between 1,1-DCA dechlorination and methanogenesis was observed which suggested the inhibition of 1,1-DCA observed in the column experiments may have been a result of hydrogen drawdown caused by cDCE dechlorination. Degradation rate constants of cDCE were found to be equal for both cores suggesting a heterogenic distribution of cDCE-degrading bacteria throughout the column, and that the bacteria responsible for cDCE dechlorination were resistant to adverse conditions.

Identification of unknown pollution sources is essential to environmental protection and emergency response. A review of recent publications in source identification revealed that there are very limited numbers of research in modeling methods for rivers. Whats more, the majority of these attempts were to find the source strength and release time, while only a few of them discussed how to identify source locations. Comparisons of these works indicated that a combination of biological, mathematical and geographical method could effectively identify unknown source area(s), which was a more practical trial in a watershed. This thesis presents a watershed-based modeling approach to identification of critical source area. The new approach involves (1) identification of pollution source in rivers using a moment-based method and (2) identification of critical source area in a watershed using a hydrograph-based method and high-resolution radar rainfall data. In terms of the moment-based method, the first two moment equations are derived through the Laplace transform of the Variable Residence Time (VART) model. The first moment is used to determine the source location, while the second moment can be employed to estimate the total mass of released pollutant. The two moment equations are tested using conservative tracer injection data collected from 23 reaches of five rivers in Louisiana, USA, ranging from about 3km to 300 km. Results showed that the first moment equation is able to predict the pollution source location with a percent error of less than 18% in general. The predicted total mass has a larger percent error, but a correction could be added to reduce the error significantly. Additionally, the moment-based method can be applied to identify the source location of reactive pollutants, provided that the special and temporal concentrations are recorded in downstream stations. In terms of the hydrograph-based method, observed hydrographs corresponding to pollution events can be utilized to identify the critical source area in a watershed. The time of concentration could provide a unique fingerprint for each subbasin in the watershed. The observation of abnormally high bacterial levels along with high resolution radar rainfall data can be used to match the most possible storm events and thus the critical source area.

Problems involving mechanical behavior of materials with microstructure are receiving an increasing amount of attention in the literature. First of all, it can be attributed to the fact that a number of recent experiments shows a significant discrepancy between results of the classical theory of elasticity and the actual behavior of materials for which microstructure is known to be significant (e. g. synthetic polymers, human bones). Second, materials, for which microstructure contributes significantly in the overall deformation of a whole body, are becoming more and more important for applications in different areas of modern day mechanics, physics and engineering. <br /><br /> Since the classical theory is not adequate for modeling the elastic behavior of such materials, a new theory, which allows us to incorporate microstructure into a classical model, should be used. <br /><br /> The foundations of a theory allowing to account for the effect of material microstructure were developed in the beginning of the twentieth century and is known as the theory of Cosserat (micropolar, asymmetric) elasticity. For the last forty years significant results have been accomplished leading to a better understanding of processes occurring in Cosserat continuum. In particular, significant progress has been achieved in the investigation of three-dimensional problems of micropolar elasticity, plane and anti-plane problems, bending of micropolar plates. These problems can be effectively solved in a very elegant manner using the boundary integral equation method. <br /><br /> However, the boundary integral equation method imposes significant restrictions on properties of boundaries of domains under consideration. In particular, it requires that the boundary be represented by a twice differentiable curve which makes it impossible to apply the method for domains with reduced boundary smoothness or domains containing cuts or cracks. Therefore, the rigorous treatment of boundary value problems of Cosserat elasticity for domains with irregular boundaries has remained untouched until today. <br /><br /> A mathematically sophisticated, but very effective approach which allows to overcome the difficulty relating to the boundary requirement consists of the formulation of the corresponding boundary value problems in terms of the distributional setting in Sobolev spaces. In this case the appropriate weak solution may be found in terms of the corresponding integral potentials which perfectly works for domains with reduced boundary smoothness. <br /><br /> The objective of this work is to develop such a method that allows us to describe and solve the boundary value problems of plane Cosserat elasticity for domains with non-smooth boundaries and for domains weakened by cracks. We illustrate the method by establishing the analytical solutions for boundary value problems of plane Cosserat elasticity, which plays an important role as a pilot problem for the investigation of more challenging problems of three-dimensional theory of micropolar elasticity. We show that the analytical solutions derived in this work may be successfully approximated numerically using the boundary element method and that these solutions can be extremely important for applications in engineering science. <br /><br /> One of the important applied problems considered herein is the problem of stress distribution around a crack in a human bone. The bone is modeled under assumptions of plane Cosserat elasticity and the solution derived on the basis of the method developed in this dissertation shows that material microstructure does indeed have a significant effect on stress distribution around a crack.

The research presented in this thesis focused on the derivation of experimental fragility curves for windborne debris (WBD) impact risk assessment of building envelope components (BECs) with ductile behavior (in particular, aluminum storm panels) within the performance-based hurricane engineering (PBHE) framework. Using a pneumatic wind cannon, rod-type WBDs were fired at aluminum storm panels to represent real-life WBD impact hazard. The experimental data from testing were used to derive the probability of failure relative to specific damage measures (DMs) versus its corresponding interaction parameter (IP). These experimentally derived probabilities were compared with results that are available in the literature and were obtained from finite element (FE) analyses. It was found that: (1) the impact kinetic energy of rod-type missiles is a sufficient IP for BECs with ductile behavior subjected to WBD impact; (2) the performance of aluminum storm panels (particularly in terms of probability of penetration) is strongly dependent on the details of the panels installation; (3) the numerical results available in the literature regarding the fragility curves of BECs with ductile behavior are qualitatively representative of the behavior of aluminum storm panels subject to WBD impact; and (4) careful modeling of the actual mechanical behavior of the panels boundary conditions is necessary for accurate numerical evaluation of the fragility curves of BECs with ductile behavior. It is noteworthy that accurate fragility curves are essential in the development of a general probabilistic performance-based engineering framework for mitigation of WBD impact hazard.

This thesis presents a study on the shear retrofit of reinforced concrete (RC) beams with externally bonded fiber reinforced polymers (FRP), since it is very important for RC beam to have a shear strength that is higher than the flexural strength in order to ensure ductile flexural failure mode. The study proposes a new model to predict the FRP shear strength contribution for different modes of failure, i.e., bending, shear with FRP rupture, shear with FRP debonding, and mixed shear-flexure and various retrofit techniques, i.e., side-bonding, U-jacketing, and full wrapping. The proposed model is compared to other existing models for FRP shear strength contribution, which are available in the literature. This comparison is made in terms of model prediction capabilities for experimentally measured shear strength increments due to FRP retrofit, which are also taken from previous literature studies. It is observed that the proposed model is in overall in good agreement with the experimental data. Furthermore, the results of this study are used to formulate a general-purpose frame finite element (FE) to compute the load carrying capacity and predict the behavior of RC beams when retrofitted with externally bonded FRP in shear. The finite element is extended to model a two dimensional frame structure with strong columns and weak beams that are deficient in shear. It is found that the proposed frame FE captures well the increase in load carrying capacity of the frame structures.

Most vibration-based damage identification methods make use of measurements directly from bridge structures with attached sensors. However, the vehicle moving on the bridge can serve as both an active actuator and a response receiver. This dissertation aimed to develop new methodologies to eventually detect bridge damages such as scour using the dynamic response of the vehicle. To reach the final objective, a framework of damage identification was developed first, which gave a guideline on the three crucial steps for damage detection. An optimization method was proposed that combines the Genetic Algorithm (GA) and the First Order (FO) method. It has the advantages of the global and local algorithms and converges faster than the traditional method using any initial values. Secondly, a new methodology using the transmissibility of vehicle and bridge responses was developed to detect bridge damages. The transmissibility of a simplified vehicle-bridge coupled (VBC) system was analyzed theoretically and numerically to study the feasibility of this method. To obtain the transmissibility, two methods were proposed using two static vehicles on the bridge. Then, a tractor-trailer test system was designed to obtain reliable responses and extract bridge modal properties from the dynamic response of moving vehicles. The test vehicle consists of a tractor and two following trailers. The residual responses of the two trailers were used, which successfully eliminated the roughness and vehicle driving effect and extracted the bridge modal properties. This methodology was applied on a field bridge and revealed a good performance. Most previous studies of bridge scour focus on the scour causes instead of its consequences. Finally, in this dissertation the developed methodologies were applied to detect scour damage from the response of bridge and/or vehicles. The scour effect on a single pile was studied and methods of scour damage detections were proposed. A monitoring system using fiber optic sensors was designed and tested in the laboratory and is being applied to a field bridge. Finally, the scour effect on the response of the entire bridge and the traveling vehicle was also investigated under the bridge-vehicle-wave interaction, which in turn was used to detect the bridge scour.

UV disinfection performance was investigated under the influence of representative particle sources, including wastewater particles from secondary effluent in a wastewater treatment plant, river particles from surface water, floc particles from coagulated surface water, floc particles from coagulated process water in a drinking water treatment plant, and soil particles from runoff water (planned). Low-pressure (LP) and medium-pressure (MP) UV dose-response of spiked indicator bacteria <i>E. coli</i> was determined using a standard collimated beam apparatus with respect to different particle sources. Significant impacts of wastewater suspended solids (3. 13~4. 8 NTU) agree with the past studies on UV inactivation in secondary effluents. An average difference (statistical significance level of 5% or &alpha;=5%) of the log inactivation was 1. 21 for LP dose and 1. 18 for MP dose. In river water, the presence of surface water particles (12. 0~32. 4 NTU) had no influence on UV inactivation at all LP doses. However, when the floc particles were introduced through coagulation and flocculation, an average difference (&alpha;=5%) of the log inactivation was 1. 25 for LP doses and 1. 12 for MP doses in coagulated river water; an average difference (&alpha;=5%) of the log inactivation was 1. 10 for LP doses in coagulated process water. Chlorination was compared in parallel with UV inactivation in terms of particulate impacts. However, even floc-associated <i>E. coli</i> were too sensitive to carry out the chlorination experiment in the laboratory, indicating that chlorine seems more effective than UV irradiation on inactivation of particle-associated microorganisms. In addition, a comprehensive particle analysis supported the experimental results relevant to this study.