Master of Science / Department of Civil Engineering / Hayder A. Rasheed / Damage detection and structural health monitoring are topics that have been receiving increased attention from researchers around the world. A structure can accumulate damage during its service life, which in turn can impair the structure’s safety. Currently, visual inspection is performed by experienced personnel in order to evaluate damage in structures. This approach is affected by the constraints of time and availability of qualified personnel. This study aims to facilitate damage evaluation and detection in concrete bridge girders without the need for visual inspection while minimizing field measurements. Simply-supported beams with different geometric, material and cracking parameters (cracks’ depth, width and location) were modeled in three phases using Abaqus finite element analysis software in order to obtain stiffness values at specified nodes. In the first two phases, beams were modeled using beam elements. Phase I included beams with a single crack, while phase II included beams with up to two cracks. For phase III, beams with a single crack were modeled using plane stress elements. The resulting damage databases from the three phases were then used to train two types of Artificial Neural Networks (ANNs). The first network type (ANNf) solves the forward problem of providing a health index parameter based on the predicted stiffness values. The second network type (ANNi) solves the inverse problem of predicting the most probable cracking pattern, where a unique analytical solution is not attainable. In phase I, beams with 3, 5, 7 and 9 stiffness nodes and a single crack were modeled. For the forward problem, ANNIf had the geometric, material and cracking parameters as inputs and stiffness values as outputs. This network provided excellent prediction accuracy measures (R2 > 99%). For the inverse problem, ANNIi had the geometric and material parameters as well as stiffness values as inputs and the cracking parameters as outputs. Better prediction accuracy measures were achieved when more stiffness nodes were utilized in the ANN modeling process. It was also observed that decreasing the number of required outputs immensely improved the quality of predictions provided by the ANN. This network provided less accurate predictions (R2 = 68%) compared to ANNIf, however, ANNIi still provided reasonable results, considering the non-uniqueness of this problem’s solution. In phase II, beams with 9 stiffness nodes and two cracks were modeled following the same procedure. ANNIIf provided excellent results (R2 > 99%) while ANNIIi had less accurate (R2 = 65%) but still reasonable predictions. Finally, in phase III, simple span beams with 3, 5, 7 and 9 stiffness nodes and a single crack were modeled using plane stress elements. ANNIIIf (R2 > 99%) provided excellent results while ANNIIIi had less accurate (R2 = 65%) but still reasonable predictions. Predictions in this phase were very accurate for the crack depth and location parameters (R2 = 97% and 99%, respectively). Further inspection showed that ANNIIIi provided more accurate predictions when compared with ANNIi. Overall, the obtained results were reasonable and showed good agreement with the actual values. This indicates that using ANNs is an excellent approach to damage evaluation, and a viable approach to obtain the, analytically unattainable, solution of the inverse damage detection problem.

Doctor of Philosophy / Department of Civil Engineering / Kyle A. Riding / The cement industry is an important component in the quest to reduce global greenhouse gas emissions because of vast amounts of cement used annually. Incorporating supplementary cementitious materials (SCMs) into concrete is one alternative to reduce cement production and thereby reduce greenhouse gas emissions. This study investigated three types of agricultural residues, namely corn stover, wheat straw, and rice straw, in addition to bioethanol byproducts as potential resources for SCM production for concrete applications. Pretreatments, commonly used in bioethanol production, were used to improve pozzolanic reactivity of corn stover ash (CSA), wheat straw ash (WSA), and rice straw ash (RSA) in cementitious systems. In the first part of this research, the impact of distilled water and dilute hydrochloric acid pretreatments on pozzolanic reactivity of WSA, RSA, and CSA were studied. Results showed that pretreatments, particularly dilute acid, improved pozzolanic properties of CSA, WSA, and RSA by removing potassium and phosphorous from the biomass prior to ashing. In addition, WSA and RSA were shown to have similar pozzolanic reactivity to that of silica fume. In the second part of this study, suitability of high lignin residue (HLR), a bioethanol byproduct, for SCM production was investigated. It was shown that burning high lignin residue produces HLR ash that is very reactive in cementitious materials and can be used as a reactive SCM in concrete. The impact of each step in the production of bioethanol on the quality of bioethanol byproduct for subsequent burning and use in concrete was also studied. Sodium hydroxide and sulfuric acid pretreatments and enzymatic hydrolysis were used. Results revealed that sodium hydroxide pretreatment of the biomass have negative impact on biomass ash properties for concrete use because sodium hydroxide pretreatment did not remove phosphorous and other crystalline phases out of the biomass. However, sulfuric acid pretreatment of biomass greatly improved ash properties. It was also shown that enzymatic hydrolysis could have beneficial impact on ash properties because, during enzymatic hydrolysis, some phosphorous was leached out of the biomass.

Master of Science / Department of Civil Engineering / Mustaque Hossain / Reclaimed Asphalt Pavement (RAP) is a useful alternative to virgin aggregates in hot-mix asphalt (HMA) as it reduces cost, conserves energy, and enables reuse of existing asphalt pavement. However, use of higher percentage of RAP sometimes leads to drier mixes that are often susceptible to early cracking. In this study, cracking resistance of Superpave mixtures with varying asphalt and RAP contents were investigated. HMA specimens were prepared based on Superpave mix design criteria for 12.5-mm (1/2-inch) nominal maximum aggregate size (NMAS). Specimens were compacted using the Superpave gyratory compactor. Static and repeated semi-circular bending (SCB) tests and Texas overlay tests (OT) (TEX-248-F) were performed in order to evaluate cracking resistance of Superpave mixtures containing three different asphalt contents (5.2%, 4.9%, and 4.6%) and three RAP percentages (20%, 30%, and 40%) from two distinct sources. Results from both crack tests showed that, with decreased asphalt content, cracking propensity increases. In general, higher percentage of RAP decreases cracking resistance. Statistical analysis of the results indicated a strong positive correlation between the asphalt film thickness and the number of load cycles before failure. Comparison of mean test results suggested that the Texas overlay test could do better evaluation of cracking resistance than the R-SCB test. This study was limited to mixtures with two sources of RAP. Because of such limitations and conflicting results from these RAP sources, a general conclusion regarding the minimum binder and maximum RAP contents without compromising cracking resistance could not be made. However, separate conclusions were drawn depending upon the characteristics of the RAP source.

Master of Science / Department of Civil Engineering / Kyle A. Riding / Air void clustering around coarse aggregate in concrete has been identified as a potential source of low strengths in concrete mixes by several Departments of Transportation around the country. Research was carried out to (1) develop a quantitative measure of air void clustering around aggregates, (2) investigate whether air void clustering can be reproduced in a laboratory environment, (3) determine if air void clustering can blamed for lower compressive strengths in concrete mixes, (4) and identify potential factors that may cause clustering. Five types of coarse aggregate and five different air entraining agents were included in the laboratory study to see if aggregate type or chemical composition of air entraining agent directly relates to air void clustering. A total of 65 mixes were made, implementing the frequently used technique of retempering that has been previously associated with air void clustering around aggregates. Compressive strength specimens as well as samples for hardened void analysis were made. Compressive strength at 7 and 28 days was determined and the automated hardened void analysis (including a new method of clustering evaluation) was performed on all samples. It was found that it is possible to reproduce air void clustering in laboratory conditions. However, the results have shown that retempering does not always cause air void clustering. It was also observed that air void clustering is not responsible for a decrease in compressive strength of retempered concrete as neither aggregate type nor chemical composition of air entraining agent had a significant impact on severity of void clustering around coarse aggregate particles. It was also found that the total air content and an inhomogeneous microstructure and not air void clustering were responsible for lower strengths.

Master of Science / Department of Civil Engineering / Sunanda Dissanayake / Although extensive research has been conducted on urban freeway capacity estimation methods, minimal research has been carried out for rural highway sections, especially sections within work zones. This study filled that void for rural highways in Kansas. This study estimated capacity of rural highway work zones in Kansas. Six work zone locations were selected. An average of six days’ worth of field data was collected, from mid-October 2013 to late November 2013, at each of these work zone sites. Two capacity estimation methods were utilized, including the Maximum Observed 15-minute Flow Rate Method and the Platooning Method divided into 15-minute intervals. The Maximum Observed 15-minute Flow Rate Method provided an average capacity of 1469 passenger cars per hour per lane (pcphpl) with a standard deviation of 141 pcphpl, while the Platooning Method provided a maximum average capacity of 1195 pcphpl and a standard deviation of 28 pcphpl. Based on observed data and analysis carried out in this study, the recommended capacity to be used is 1500 pcphpl when designing work zones for rural highways in Kansas. This research provides the proposed standard value of rural highway work zone capacities so engineers and city planners can effectively mitigate congestion that would have otherwise occurred due to impeding construction/maintenance.

Master of Science / Department of Civil Engineering / Natalie Mladenov / In the United States approximately 37% of the 4 trillion kWh of electricity is generated annually by combusting coal (USEPA, 2013). The abundance of coal, ease of storage, and transportation makes it affordable at a global scale (Ghose, 2009). However, the flue gas produced by combusting coal affects human health and the environment (USEPA, 2013). To comply with federal regulations coal-fired power plants have been implementing sulfur dioxide scrubbing systems such as flue gas desulfurization (FGD) systems (Alvarez-Ayuso et al., 2006). Although FGD systems have proven to reduce atmospheric emissions they create wastewater containing harmful pollutants. Constructed wetlands are increasingly being employed for the removal of these toxic trace elements from FGD wastewater. In this study the effectiveness of using a constructed wetland treatment system was explored as a possible remediation technology to treat FGD wastewater from a coal-fired power plant in Kansas. To simulate constructed wetlands, a continuous flow-through column experiment was conducted with undiluted FGD wastewater and surface sediment from a power plant in Kansas. To optimize the performance of a CWTS the following hypotheses were tested: 1) decreasing the flow rate improves the performance of the treatment wetlands due to an increase in reaction time, 2) the introduction of microbial cultures (inoculum) will increase the retention capacity of the columns since constructed wetlands improve water quality through biological process, 3) the introduction of a labile carbon source will improve the retention capacity of the columns since microorganisms require an electron donor to perform life functions such as cell maintenance and synthesis. Although the FGD wastewater collected possessed a negligible concentration of arsenic, the mobilization of arsenic has been observed in reducing sediments of wetland environments. Therefore, constructed wetlands may also represent an environment where the mobilization of arsenic is possible. This led us to test the following hypothesis: 4) Reducing environments will cause arsenic desorption and dissolution causing the mobilization of arsenic. As far as removal of the constituents of concern (arsenic, selenium, nitrate, and sulfate) in the column experiments, only sulfate removal increased as a result of decreasing the flow rate by half (1/2Q). In addition, sulfate-S exhibited greater removal as a result of adding organic carbon to the FGD solution when compared to the control (at 1/2Q). Moderate selenium removal was observed; over 60% of selenium in the influent was found to accumulate in the soil. By contrast, arsenic concentrations increased in the effluent of the 1/2Q columns, most likely by dissolution and release of sorbed arsenic. When compared to the control (at 1/2Q), arsenic dissolution decreased as a result of adding inoculum to the columns. Dissolved arsenic concentrations in the effluent of columns with FGD solution amended with organic carbon reached 168 mg/L. These results suggest that native Kansas soils placed in a constructed wetland configuration and amended with labile carbon do possess an environment where the mobilization of arsenic is possible.

Master of Science / Department of Civil Engineering / Kyle Riding / The durability of partial depth concrete repair is directly related to the bond strength between the repair material and existing concrete. The wait time effects of cementitous grouts, epoxy, acrylic latex, and polyvinyl acetate bonding agents were observed on bond strength. Three rapid repair materials were as a comparison to bond strength, as well as concrete samples with no bonding agents having dry conditions and saturated surface dry moisture condition. The bonding agents and rapid repair materials were tested in a controlled laboratory environment. Bond strength loss with wait times of 0, 2, 5, 10, and 30 minutes were observed when bonding agents were applied. The laboratory samples were loaded using a direct shear test. Field tests were performed using the same repair materials and bonding agents. When the agents were applied in the field the wait times used were 0, 15, 30, and 45 minutes. 7 day and 5 month pull off tensile tests were performed during the field experiment. The data from both experiments show that when using cement grout bonding agents the high bond strength can be obtained when the repair material is applied within 15 minutes of application of the cement grout, and after 15 minutes bond loss can be expected. Wait time didn't have a significant effect on epoxy and acrylic latex bonding agents as long as they were placed before setting. The polyvinyl acetate agent and repair materials can develop high bond strength in laboratory settings, but when used in the field the bond strengths experience loss. When not using bonding agents in a repair, adequate bond strength can be obtained when using saturated surface dry condition.

Doctor of Philosophy / Department of Civil Engineering / Hayder Rasheed / Recent applications in reinforced concrete columns, including strengthening and extreme loading events, necessitate the development of specialized nonlinear analysis methods to predict the confined interaction domain of axial force, shear, and bending moment in square and slightly rectangular concrete columns. Fiber-reinforced polymer (FRP) materials are commonly used in strengthening applications due to their superior properties such as high strength-to-weight ratio, high energy absorption and excellent corrosion resistance. FRP wrapping of concrete columns is done to enhance the ultimate strength due to the confinement effect, which is normally induced by steel ties. The existence of the two confinement systems changes the nature of the problem. Existing research focused on a single confinement system. Also, very limited research on rectangular sections was found in the literature. In this research, a model to estimate the combined behavior of the two systems in rectangular columns is proposed. The calculation of the effective lateral pressure is based on Lam and Teng model and Mander model for FRP wraps and steel ties, respectively. The proposed model introduces load eccentricity as a parameter that affects the compression zone size, and in turn the level of confinement engagement. Full confinement corresponds to zero eccentricity, while unconfined behavior corresponds to infinite eccentricity. The model then generates curves for eccentricities within these boundaries. The numerical approach developed has then been extended to account for shear interaction using the simplified modified compression field theory adopted by AASHTO LRFD Bridge Design Specifications 2014. Comparisons were then performed against experimental data and Response-2000, an analytical analysis tool based on AASHTO 1999 in order to validate the interaction domain generated. Finally, the developed models were implemented in the confined analysis software “KDOT Column Expert” to add FRP confinement effect and shear interaction.

Master of Science / Department of Civil Engineering / Hayder Rasheed / The importance of the analysis of circular columns to accurately predict their ultimate confined capacity under shear-flexure-axial force interaction domain is recognized in light of the extreme load event imposed by the current AASHTO LRFD specification. In this study, various procedures for computing the shear strength are reviewed. Then, the current procedure adopted by AASHTO LRFD 2014, based on the simplified modified compression field theory, is evaluated for non-presetressed circular concrete bridge piers. This evaluation is benchmarked against experimental data available in the literature and against Response 2000 freeware program that depicts interaction diagrams based on AASHTO 1999 requirements. Differences in results are discussed and future improvements are proposed. A new approach is presented to improve the accuracy of AASHTO LRFD calculations. The main parameters that control the cross section shear strength are discussed based on the experimental results and comparisons.

Doctor of Philosophy / Department of Civil Engineering / Mustaque A. Hossain / Tack coat is a thin layer of asphaltic material used to bind a newly-placed lift of hot-mix asphalt (HMA) pavement to a previously-placed lift or a new HMA overlay/inlay and existing pavement. The purpose of a tack coat is to ensure that a proper bond occurs so that traffic loads are carried by the entire HMA structure. Proper bonding exists when HMA layers act as a monolithic structure, transferring loads from one layer to the next. This depends on appropriate selection of tack coat material type and application rate, and is essential to prevent slippage failure and premature cracking in the wearing surface. This study focuses on development of a three-dimensional finite element (FE) model of HMA pavement structure in order to assess HMA interface bonding. The FE model was constructed using commercially available ABAQUS software to simulate an Accelerated Pavement Testing (APT) experiment conducted at the Civil Infrastructure Systems Laboratory (CISL) at Kansas State University. Mechanistic responses measured in the CISL experiment, such as localized longitudinal strain at the interface, were used to calibrate the FE model. Once calibrated, the model was used to predict mechanistic responses of the pavement structure by varying the tack coat property to reflect material characteristics of each application. The FE models successfully predicted longitudinal strains that corresponded to APT results.