M. S.

published_or_final_version / Mechanical Engineering / Doctoral / Doctor of Philosophy

In this research, the effects of gas composition, elapsed time of reaction and temperature on scale formation and descaling of low carbon steel were investigated and results were discussed from the viewpoint of the phase composition of the scales, oxidation rates, oxidation mechanisms, adhesion, fracture mechanics, porosity and residual scale. / The phase composition and morphology of scales grown under conditions similar to those of reheating furnaces were analyzed. Low carbon steel was oxidized over the temperature range 1000-1250°C in gas mixtures of O 2-CO2-H2O-N2, O2-H 2O-N2 and O2-CO2-N2. The mole fraction of each phase, wustite (FeO), magnetite (Fe3O 4) and hematite (Fe2O3) was determined by the direct comparison method Two types of scales were observed. The first type was a crystalline scale with an irregular outer surface composed mostly of wustite, and a negligible amount of magnetite. The second type was the classical three-layer scale composed of wustite, magnetite and hematite. In general, the experiments showed that the furnace atmosphere, oxidation time and temperature influence the phase composition of the scales. / Low carbon steel was oxidized in air over the temperature range 600-1200°C for 120 s to approximate the formation of secondary and tertiary scale in hot rolling. The mole fraction of wustite, magnetite and hematite was determined by the direct comparison method The phase composition of the scales changed with temperature and time. During the initial 30 s of oxidation, wustite was the predominant phase in the temperature range 800-1200°C, and as oxidation proceeded, the percentages of magnetite and hematite increased. In addition, the texture of the scales was investigated by orientation imaging microscopy (OIM); it was found that temperature influences the texture of the scales. The experiments indicated that 850°C is the ideal temperature for the finishing mill in order to reduce surface defects and work roll wear. / The adhesion of scales formed in air on low carbon steel in continuous heating and isothermal conditions was investigated with a four-point bending test. The separation (crack) always occurred inside the scale indicating that the strength of the scale is lower than the strength of the scale/steel interface. It was found that scale adhesion is related to scale porosity, blister formation and stresses acting in the scale. A spallation process was observed when cooling from 800°C to room temperature. The microscopic observations revealed that spallation followed route 1 "strong interface and weak oxide." / Low carbon steel was oxidized over the temperature range 1050-1250°C in O2-CO2H2O-N2 gas mixtures in order to study the hydraulic descaling process. The oxidation times were 15-120 min. and the scales were 130-2000 mum thick. The experimental parameters were chosen to approximate scale formation under conditions similar to those of reheating furnaces. In the hydraulic descaling tests, two modes of scale removal were observed. In the first mode, observed in classical three-layer scales that developed an inner porous layer with low or medium porosity, the horizontal undercutting occurred at the boundary of the inner porous layer and dense scale. The second mode was observed in classical three-layer scales that developed an inner porous layer with high porosity and in crystalline scales. In the second mode, the horizontal undercutting occurred at the first plane of large pores relative to the scale/steel interface. In general, the experiments showed that scale morphology controlled the removability of scale.

Thesis (Ph. D.)--University of New South Wales, 2002. / Also available online.

Thesis (Ph.D.)--Ohio University, June, 2003. / Title from PDF t.p. Includes bibliographical references (leaves 158-164)

Quantitative chemical and scanning electron microscopical techniques have been employed to investigate the deoxidation kinetics and changes in oxidation product morphology in low carbon steel melts. The techniques have been used to study the deoxidation processes associated with aluminium, titanium, silicon, zirconium and a calcium-aluminium alloy. After the addition of the deoxidant, the total oxygen concentrations of all melts rapidly decreased corresponding with a decrease in the size and number of inclusions observed. This continued to a plateau level of total oxygen concentration and mean inclusion diameter. Samples removed from the melts prior to deoxidation were found to contain globular MnO-FeO inclusions. It was discovered that the morphological sequence for single element deoxidants involved a progressive evolution from liquid globular to solid spherical inclusions followed by polyhedral, dendritic and coralline morphologies. Finally, sintered agglomerates were formed when inclusion clusters collapsed. The extent to which the oxidation products went down the sequence depended on: the dissolution characteristics of the deoxidant; the thermodynamic affinity of the deoxidant for oxygen in the melt; the inclusion/melt interfacial energy characteristics; the refractoriness of the oxidation products and intermediate compounds; and the degree of turbulence experienced by the melt. Explanations have been postulated which elucidate the behaviour of the different deoxidants, as not all displayed the whole morphological sequence. Silicon deoxidation produced spherical silicates, whereas the zirconia inclusions were either spherical or dendritic and the titanium oxidation products had spherical or polyhedral morphologies. Aluminium exhibited all morphologies in the sequence. Deoxidation with the calcium-aluminium alloy was found to have preceded by a two stage process. The initial stage was dominated by the formation of aluminium rich solid oxides followed by the progressive reduction by calcium, resulting in an adhesive liquid calcium-aluminate surface coating. The role of refractory crucible as a collecting surface for the capture and removal of deoxidation products from the melt was investigated, which confirmed that the inclusions were generally incorporated into the low melting point matrix phases. Turbulence also increased the probability that emergence would take place at these capture sites.

The contribution of the clamping force in the technique used in this study to stretch-bend low carbon steel samples was investigated to support the subsequent changes in the microstructure and properties of the formed material with regard to parent material. Although plastic deformation by cold working is known to induce texture or preferred orientation to the grains of a formed material while decreasing its ductility and increasing the strength, as well as inducing residual stress, it is not known how the different directions (rolling, transverse and oblique/45°) of the sheet steel will respond to this stretch bending technique. The first part of the research work involved a thorough literature review on sheet metal forming processes and their effect on the formed material with interest on the above mentioned directions of the sheet. It became clear from the literature overview that cold working of a material will induce strain-hardening which varies with the magnitude of cold work, resulting in changes in the strength and ductility of the material. Besides, when plastic deformation is not uniform (e.g., tensile and compressive) throughout the entire cross section of the formed part, residual stresses remain in the material with the grains been elongated along the direction of the maximum strain. The main parameters that were considered and controlled in this study are as follows: strain experienced / stress induced into the form sample, the sample direction, the stroke length and the clamping torque, the generated radius of curvature. The chapters that follow the literature review, deal with the set-up of the different equipment used in this study, the specimen preparation as well as the recording, the calculation and interpretation of the results. It was found that the stress magnitude that generated the different radii of curvature (120 mm, 150 mm and 185 mm) was between 1 percent & 13 percent higher than the parent material’s yield strength with the lower stress been associated to the smaller radius of curvature and the higher stress to the higher radius. The stress induced into the sample during forming was not only proportional to the stroke length but also to the distance between the punch’s tip and the sample and the sample to the die’s nadir. The clamping torque adopted was restricted to the manual capacity of the operator who used a preset torque wrench to fasten the plate sample into the jig. Plate samples of low carbon steel were cut to angles of 0°, 45°, and 90° to the rolling direction of the sheet material and stretch-bent on a single-action mechanical press to 120 mm, 150 mm, and 185 mm radii of curvature. The preliminary results indicate that stretch-bent samples had increased hardness to the parent plate, in particular below the surface layers up to around 1.1 mm depth. Since there is a well established relationship between hardness, yield and tensile strengths for steel, the yield and tensile strengths of the formed material were estimated using the Nobre et al [34] incremental relation, which relates the linearity between relative increments of hardness and yield strength. Changes were not noticeable at the microstructural level of the formed samples. Meanwhile, samples from which higher plastic deformation stress values were calculated were not those absorbing higher impact energy when Charpy specimen cut from plate and stretch-bent samples were tested. The maximum relieved residual stress in the parent material was predominantly compressive and represents in magnitude approximately 12 percent (average for the three directions) of its original yield strength. In the stretch-bent samples, the relieved residual stress was compressive in the outer curved section with a magnitude about 50 percent of the parent material yield strength and tensile in the inner curved section with a magnitude approximately 25 percent of the parent material yield strength.

The goal of this thesis is to probe the functionality of 2-mercaptobenzimidazole (MBI) as corrosion inhibitor of carbon-steel in both strong and weak aqueous acidic solutions (HCl and H2CO3). To achieve this target electrochemical techniques have been employed, in combination with substrate analysis. Concerning aqueous HCl media, results demonstrate that MBI is an effective corrosion inhibitor, functioning essentially equally well in 1 M, 0.1 M, and 0.01 M HCl concentrations. X-ray photoelectron spectra suggest that MBI is typically bound to the surface in two tautomeric forms (thione and thiol). Furthermore, these data indicate that substrate termination varies as a function of both HCl and MBI concentration, with the interface consisting of MBI bound to film-free carbon-steel on highly inhibited substrates. In further work, the impact of dissolved oxygen, solution temperature, and immersion time on MBI performance in HCl solutions has been assessed. The latter two parameters have considerable influence on MBI inhibition efficiency. More specifically, it was found that MBI decreases dramatically its inhibition efficiency between 60°C and 70°C in 1 M HCl, and also apparently work less well as substrate immersion time increases. As regards MBI performance in deaerated CO2-saturated NaCl (0.62 M) solution, results demonstrate that MBI effectively inhibits corrosion within the parameter space explored i.e. solution temperatures of 30°C and 55°C and total applied pressures (p(H2O) + p(CO2)) of 1 bar and 20 bar. The performance of MBI does not vary greatly for different combinations of these temperatures and pressures. Post immersion substrate characterisation with XRD and SEM indicate that no significant surface scaling occurs under these conditions.

The Florida Department of Transportation (FDOT) has been using supplementary cementitious materials while constructing steel reinforced concrete marine bridge structures for over three decades. It has been found from previous studies that such additions in concrete mix makes the concrete more durable. This research was conducted to better understand the corrosion propagation stage of steel rebar embedded in high performance concrete exposed to high humidity environment. Reinforced concrete samples that were made with binary mixes, and ternary mixes were considered. None of these concretes had any admixed chloride to start with. An accelerated chloride transport method was used to drive chloride ions into the concrete so that chlorides reached and exceed the chloride threshold at the rebar surface and hence the corrosion process initiated after a short period of time (within few days to few months). Once corrosion has initiated the corrosion propagation can be studied. Electrochemical measurements such as rebar potential measurements, Linear Polarization Resistance (LPR), Electrochemical Impedance Spectroscopy (EIS), and Galvanostatic Pulse (GP) measurements were taken at regular intervals (during and after the electro-migration process) to observe the corrosion propagation in each sample. During the propagation stage, reinforcement eventually reached negative potentials values (i.e., Ecorr≤ –0.200 Vsce) for all the samples. The corrected polarization resistance (Rc) was calculated by subtracting the concrete solution resistance from the apparent polarization resistance measured. The Rc values obtained from LPR and GP measurements were converted to corrosion current (as the corroding area is unknown), and these corrosion current values measured over time were used to obtain the calculated mass loss (using Faraday’s Law). A comparison was made of the calculated corrosion current obtained using the LPR and GP tests. A comparison of mass loss was also obtained from the values measured from LPR and GP tests. From the experimental results, it was observed that the corrosion current values were largely dependent on the length of solution reservoirs. For specimens cast with single rebar as well as three rebars, the most recent corrosion current values (measurements taken between July 2018 to October 2020) in general were larger for the rebars that are embedded in specimens prepared with SL mix, followed by specimens prepared with FA, T1, and T2 mixes respectively. The range of corrosion current values (most recent) were 0.8-33.8 μA for SL samples, 0.5-22.5 μA for FA samples, 0.8-14.8 μA for T1 samples, and 0.7-10.4 μA for T2 samples respectively. It was also found that the calculated mass loss values were larger for rebars that are embedded in specimens (single rebar and three rebars) prepared with SL mix, followed by specimens prepared with FA, T1, and T2 mixes respectively. The range of calculated mass loss values were 0.07-1.13 grams for SL samples, 0.06-0.62 grams for FA samples, 0.12-0.54 grams for T1 samples, and 0.06-0.40 grams for T2 samples respectively. A variety of corrosion related parameters (Ecorr, Rs, Rc, and Icorr) and calculated theoretical mass loss values observed, were due to the changing parameters such as concrete compositions, concrete cover thickness, rebar diameter, total ampere-hour applied, and reservoir size. The specimens showed no visual signs of corrosion such as cracks or corrosion products that reached the concrete surface. The actual size of the corroding sites was unknown as the specimens were not terminated for forensic analysis. The size of the corroding sites could affect how much corrosion products are required to crack the concrete. It is speculated that the corrosion products in liquid form penetrated the pore structure but did not build up enough to cause cracks. No cracks or corrosion bleed outs were observed within the monitored propagation period of approximately 1600 days. / Includes bibliography. / Dissertation (Ph.D.)--Florida Atlantic University, 2020. / FAU Electronic Theses and Dissertations Collection

No description available.