Civil & Environmental Engineering Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/2753
Browse
5 results
Search Results
Item ANALYTICAL STUDY OF THE BEHAVIOR OF COMPOSITE DOVETAIL METAL DECKING FLOOR SYSTEMS FOR THE DEVELOPMENT OF PRACTICAL DESIGN GUIDELINES(2022) Pase, Tara; Phillips, Brian M; Fu, Chung C; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Prior testing and industry practice have shown composite metal floor systems – floors systems constructed from concrete composite with metal decking – behave stiffer than the current state-of-the-practice simplified calculations and estimations predict. Specifically, the dovetail decking does not have the quantity of available research and universal design guidance compared to the more common trapezoidal composite decking; this lack of a more accurate design standard has made the calculation of the non-linear stiffness behavior of the dovetail composite deck floor systems under small to intermediate strains inaccurate, and therefore limits its use for long-span configurations where deflection limits (i.e., serviceability limits) control the design.The objective of this research is to create an analytical model of the flexural behavior of re-entrant dovetail composite decking floor systems for service (i.e., deflection) and strength (i.e., ultimate capacity) limit states and to understand the unique mechanical behavior of the slab system. By creating a more accurate analytical model of the flexural behavior of the dovetail composite deck system, a robust design guide table for engineering use is developed for a multitude of variables typically seen in construction, including but not limited to: various loads, deck gauges (thicknesses), concrete strength, concrete depths, etc. The flexural behavior of the composite metal deck is modeled based on its material properties and the following base assumptions: the composite slab is in pure bending; plane sections remain plane and are orthogonal to the neutral axis; the laws of static equilibrium apply; and loads are assumed to be static. Application of this composite theory to determine the moment-curvature relationship using a numerical strain-compatibility computer-based solver is compared against physical tests to validate and calibrate the theoretical assumptions that make up the basis for the calculations. The resulting flexural behavior derived from this numerical strain-compatibility method has a multitude of uses including but not limited to: the derivation of a robust design table for a number of decking gauges, common slab thickness values, concrete strength ranges, and so forth; and variable stiffness properties for use in simplified finite element plate models. The real-world purpose for this new numerical strain-compatibility model is to provide robust design guidance and engineering resources for practicing structural engineers, without a time-consuming and expensive finite element model. With the numerical strain-compatibility analysis, an engineer can accurately analyze and specify composite concrete slabs in building projects without being limited by shorter spans or thicker slabs due to inaccuracies in deflection calculations.Item QUANTITATIVE ANALYSIS OF MICROCRACKS IN CONCRETE FROM DELAYED ETTRINGITE FORMATION BY IMAGE PROCESSING OF LASER SHEAROGRAPHY IMAGES(2021) Sharma , Shivam; Amde M., Amde Dr.; Livingston, Richard Dr.; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The objective of this research was to determine which image processing algorithms were most effective in quantifying the microcrack distribution in concrete in laser shearography images. The motivation is the need for a nondestructive method to measure the development of damage in concrete due to expansive stresses generated by delayed ettringite formation (D.E.F.). This produces networks of fine microcracks. These may not be visible to the human eye or detectable by conventional imaging techniques. However, laser shearography provides a means to visualize them at very early stages of growth. This nondestructive method generates a first derivative image of the surface topography. This can make cracks less than 1 µm visible. These can in turn be quantified by automated image processing algorithms. These crack pattern images can then be analyzed to obtain sets of statistics that can be tracked over time for investigating the D.E.F. process which can provide insights into the crack propagation mechanisms. This research thus concerned the application of automated image processing to a set of laser shearography images of concrete prisms where D.E.F. damage had been induced by an accelerated test method. Four prisms were involved. Two were treated to accelerate the D.E.F. rate of development. The other two were controls. They were imaged periodically at roughly month intervals by laser shearography for up to 200 days. A commercial automated image processing software, ImagePro was used. Two approaches were tried to identify the cracks in each image: manually or by an automated macro. Once a crack was identified, its track was traced by an auto tracing algorithm. It was found that the macro generated too many artifacts, so the manual method was used. The results showed significant differences between the control prisms and the treated ones in terms of crack numbers and their length. Over time the cracks in the treated specimens tended to grow longer but fewer in number as the individual cracks joined up. In the controls, the cracks tended to disappear with time. This may be because they were only superficial in the first place and then were covered by a thin surface layer of calcium hydroxide that precipitated from the lime water bath. The conclusion is that it is feasible to apply automated imaging techniques to quantify damage due to D.E.F. However, a disadvantage of using the existing commercial software was that it produced crack tracks that were only one pixel wide. Thus, it was not possible to measure actual crack width and their changes with expansion. Another issue is that it is designed for images of surfaces in real space whereas the laser shearography image is the first derivative of the surface topography. This contains information that could be used to estimate crack widths. It could also be used to automate the detection of cracks, possibly by the application of artificial intelligence techniques.Item COMPARISON OF NEUTRON NON DESTRUCTIVE METHOD AND CONVENTIONAL CHEMICAL METHOD FOR CHLORIDE MEASUREMENT IN CONCRETE(2019) Sridhar, Preethi; Amde, Amde M; Livingston, Richard; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The presence of chloride in concrete is a critical issue raising concerns in the construction industry as they promote corrosion of the steel reinforcements, drastically reducing the strength of the structure. The aim of this study is to compare the performance of a neutron-based nondestructive testing method, Prompt Gamma Activation Analysis (PGAA) against the destructive wet chemistry method ASTM C-1152 currently used to determine the chloride concentration in concrete. Two modes of PGAA operation were tested. One was to use PGAA with a slit collimator to measure the chlorides at 2 mm thick cross-section in intact samples. The other was a direct comparison with C-1152 to analyze powdered concrete samples. Concrete was prepared in four batches, in which three batches had added chloride -at nominally 0.2%, 0.1%, 0.01% by weight of cement and the fourth (control) batch has zero added. The PGAA analysis was done at the Cold Neutron PGAA station at NIST and the C1152 testing was done at the National Ready Mixed Concrete Association (NRMCA) laboratory. The intact samples were scanned at three different vertical positions. The PGAA method is capable of detecting Cl at levels corresponding to the corrosion threshold of 0.1-0.2% Cl by weight of cement. The minimum detectable limit for PGAA is below 0.02% Cl by weight of cement and approaches the Cl background contributed by the raw materials, in this case, the cement. The PGAA- measured chlorides concentrations showed excellent linearity after correction for the chloride content in the concrete raw materials, mainly the cement. For the powdered samples, the C1152 and PGAA results were in very good agreement. However, the PGAA data showed much less scatter with an uncertainty as low as 0.3%. The findings of this study indicate that PGAA is a feasible replacement for the C1152 method and since it can be done on intact specimens, it avoids the time-consuming steps of crushing, sieving and nitric acid extraction and can be more cost-effective.Item DETERMINATION OF DYNAMIC MODULI AND PERMANENT DEFORMATION OF MARYLAND ASPHALT MIXTURES USING AMPT(2017) Haider, Intikhab; Schwartz, Charles W; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Full implementation of the mechanistic-empirical pavement design guide (MEPDG) in Maryland requires Level 1 (measured) material properties to characterize asphalt mixtures commonly used in the state. Specifically, these proprieties are the dynamic modulus (DM) and the repeated load permanent deformation (RLPD) properties. To achieve this goal, 28 asphalt mixtures were collected from construction sites/asphalt plants and tested in the Maryland State Highway Administration Office of Materials Technology Asphalt Technology Division laboratory. The DM and RLPD testing was performed on all 28 asphalt mixtures following the AASHTO PP 60, AASHTO PP 61 and AASHTO TP 79 protocols. In addition to the 28 asphalt mixtures from Maryland, DM and RLPD data for 18 asphalt mixtures tested in NCHRP Project 9-30A were also included in parts of this study. In addition to developing a catalog of typical Level 1 material properties for common Maryland asphalt mixtures, this study produced several other important results and findings. These include: (1) The L-1 inputs (measured E* and G* and recalibrated coefficients of rut model, K1, K2, K3) consistently give lesser predicted distresses than L-3 inputs (predicted E* values, default G* values, and default coefficients of rut model) in MEPDG software. (2) The average percentage differences for each predicted distress at all levels of traffic are highest for L-1 versus L-3 inputs and lowest for L-1 versus L-1A (measured E* and G* data and default coefficients of rut model) inputs. (3) The recalibration of Witczak E* model removes the bias toward underprediction in the original Witczak model. The distresses predicted using L-3 (CWM-Calibrated Witczak Model based on Maryland mixes) inputs are closest to the distresses predicted using the measured L-1 inputs. (4) The total number of samples required for complete characterization of one asphalt mixtures as per AASHTO PP 61 and AASHTO TP 79 can be reduced from 12 to 3. The reduction in total specimen preparation (from 60 to 15 hours) and testing time (from 30 to 10 hours) represents substantial economies in structural characterization of asphalt mixtures and motivates state agencies to perform DM and RLPD testing on routine basis to develop performance based specification.Item Prediction of Permanent Deformation in Asphalt Concrete(2012) Carvalho, Regis Luis; Schwartz, Charles W; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Permanent deformation is a major distress in flexible pavements that leads to the development of rutting along the wheel path of heavily trafficked roads. Early detection of rutting is very important for preventive maintenance programs and design of rehabilitation strategies. Rutting by definition is the accumulated permanent deformation that remains after removal of the load. Rigorous modeling of permanent deformations using nonlinear finite element analysis based on the correct physical mechanism of residual deformations after removal of the load provides important insights into the rutting problem. This dissertation documents the study of permanent deformation in asphalt concrete in pavement structures using a fully mechanistic model based on Schapery's viscoelasticity and Perzyna's viscoplasticity theories. The model is calibrated and implemented in a 3D finite element commercial software package. Two calibration procedures are performed and discussed. Two immediate practical applications are shown and a simulation of full scale accelerated pavement test is performed. This research demonstrates that the Perzyna-HiSS viscoplastic model can be successfully calibrated using either research-grade creep and recovery tests or the more simple and production-oriented Flow Number test. The importance of induced shear stress reversals under a moving wheel load is documented. The 3D finite element simulation is then used to identify the fundamental differences on how rutting develops in different pavement structures in terms of the differences in the transverse profile and distribution of rutting within the layer. The analysis results are used to develop new pavement-specific depth functions for potential future incorporation into the AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG). Lastly, the 3D finite element model is used to predict rutting at one lane of the FHWA's full-scale Accelerated Load Facility experiment. After correction for some anomalies during the early loading cycles in the experiment, the predicted and measured rutting at the center of the wheel path were in good agreement.