The main objective of this work is to establish an analytical model to evaluate the rate of corrosion in a horizontal pipe downstream of an orifice under flow-accelerated corrosion (FAC). FAC is a serious issue in nuclear and fossil power plants. In this work, an experimental setup was built to observe the effect of the flow on corrosion inside a tube. The experiments confirmed that the flow inside the tube caused more corrosion. However, accurate experimental data from literature has been selected and correlated by dimensional analysis, the modelling method of repeating variables and the Buckingham Pi theorem. It was found that the Sh number and the relative distance from the orifice are the main dimensionless parameters influencing FAC downstream of an orifice. The maximum value of the FAC rate could be well-predicted for the OR of 0.25, while the location of the maximum FAC rate could be well predicted for the OR of 0.5. The maximum FAC rate occurs between 2D to 4D downstream of the orifice and increases with a decreasing OR. This work could be useful for professionals in industry and researchers in the field and could be the starting point for a new way of evaluating the FAC rate downstream of a flow’s singularity. / Dissertation (MSc)--University of Pretoria, 2017. / Mechanical and Aeronautical Engineering / MSc / Unrestricted

No description available.

No description available.

Flow accelerated corrosion is a piping degradation mechanism that results in pipe wall thinning due to the dissolution of the magnetite oxide layer on carbon steel surfaces to the bulk flow. The rate limiting process of flow accelerated corrosion in piping system is the diffusion-controlled mass transfer. The surface roughness develops due to the mass transfer and can subsequently have a significant effect on the mass transfer. The naturally developing surface roughness in many dissolving surfaces, including carbon steel pipes, is a densely packed array of saucer shaped depression called scallops, which can have several length scales. Heretofore, the developing roughness on soluble walls has not been quantified, mainly due to the lack of a reliable measurement methodology. The overall objective of this research is to investigate the developing roughness and the corresponding mass transfer on soluble walls in different piping geometries. A wall dissolving method using gypsum test sections dissolving to water in a closed flow loop was used to mimic the mass transfer in carbon steel pipes due to a similar Schmidt number of 1200. A novel non-destructive measurement technique using X-ray CT scans was developed to measure the development of surface roughness and the corresponding mass transfer. The method was validated by performing experiments using straight pipe test sections and comparing against traditional measurements method using ultrasonic sensors, coordinate measurement machine and laser scans. The time evolution of surface roughness and the corresponding mass transfer were measured in pipe test sections at Reynolds number of 50,000, 100,000 and 200,000. The roughness scallops were observed to initiate locally and then develop until the surface is spatially saturated. The surface roughness was characterized by the RMS height, peak-to-valley height, integral length scale, density and spacing of the scallops. Two time periods of roughness development were identified: an initial period of slower growth in the roughness height followed by a relatively higher growth rate that corresponded to the period before and after the surface saturates with the scallops. The mass transfer enhancement due to the roughness in each of these time periods was also found to be different, with a higher increase in the first period followed by a slower increase once the streamwise spacing was approximately constant. Both the height and spacing of the roughness elements was found to affect the mass transfer enhancement. A new correlation is proposed for the mass transfer enhancement as a function of the height-to-spacing ratio of roughness, with a weak dependence on Reynolds number. The measurement methodology was extended to study the mass transfer and developing roughness in a complex S-shaped back to back bend at Reynolds number of 200,000. The mass transfer in bend geometry can be enhanced by both the local flow due to the geometry effect and the developing roughness. Two high mass transfer regions were identified: at the intrados of the first and second bends. The height-to-spacing ratio of the roughness was found to increase more rapidly in these high mass transfer regions. An additional one-time experiment was performed at a Reynolds number of 300,000. A higher surface roughness with smaller values of spacing-to-height ratio was found in the regions with high mass transfer. / Thesis / Doctor of Philosophy (PhD)

Corrosion of reinforcing steel embedded in concrete due to the presence of moisture, aggressive chemicals, inadequate cover, and other factors can lead to deterioration that substantially reduces the strength and serviceability of the affected structure. Accounting for corrosion degradation is critical for evaluation and assessment of the load carrying capacity of existing reinforced concrete (RC) structures. However, little is known about the relationship between high strain rate blast loading and the degradation effects that govern corrosion damaged structures such as concrete cover cracking, reduction in reinforcement areas, and deterioration of bond between concrete and steel. Ten identical RC beams were constructed and tested, half under blast loading conditions produced using the Virginia Tech Shock Tube Research Facility and the other half under quasi-static loading. The blast tests were conducted to investigate how increasing blast pressure and impulse affect the global displacement response and damage modes of beams subjected to blast loads. The quasi-static tests were performed to establish fundamental data on the load-deflection characteristics of corroded RC beams. One beam from each testing group served as a control specimen and was not corroded while the remaining beams were subjected to varying levels of corrosion (5%, 10%, 15%, and 20%) of the longitudinal reinforcement along the midspan region. The specimens were corroded using an accelerated corrosion technique in a tank of 3% sodium chloride solution and a constant electrical current, creating a controlled environment for varying levels of corrosion. An analytical model was also created using a single degree of freedom (SDOF) approach which predicted the performance of corroded RC beams under blast loading. The results of the quasi-static tests revealed that as corrosion levels increased, the load to cause yielding decreased, the yield displacements decreased, and failure occurred earlier for all specimens. This was accompanied by increased damage to the concrete cover and the addition of longitudinal corrosion induced cracking. For the blast loaded specimens, the results demonstrated that the maximum displacements and residual displacements increased beyond the expected response limits for corrosion levels greater than 5%, but at corrosion levels less than 5% there was no significant change in displacements. Damage levels increased by one or more categories with the introduction of even small levels of corrosion of less than 5%. At corrosion levels greater than 5%, before loading was applied, the specimens exhibited moderate damage due to the introduction of corrosion induced cracking. After loading, the specimens sustained hazardous damage at progressively lower blast volumes. The failure mode changed from ductile to sudden and brittle failure at corrosion levels greater than 5% but remained ductile with flexural failures at low corrosion levels below 5%. The experimental results could be predicted with a high level of accuracy using the SDOF approach, provided that the degraded strength of corroded concrete cover, degraded mechanical properties of corroded steel, length of the corroded region, and determination of either uniform or pitting corrosion are accounted for. Overall, the introduction of corrosion to an RC beam subjected to blast loading resulted in decreased strength and ductility across all specimens but with most disastrous effects occurring at corrosion levels of 5% or greater. A recommendation is made to adjust the response limits in ASCE/SEI 59 to account for corrosion in RC beams. / Master of Science / The threat of blast loads, resulting from either terrorist attacks or accidental explosions, poses a significant threat to the structural integrity of buildings, life safety of occupants, and the functionality of the structure. Corrosion of reinforcing steel embedded in concrete, due to the presence of moisture, aggressive chemicals, and other factors, can lead to deterioration that substantially weakens the affected structure. Accounting for corrosion degradation is critical for evaluation and assessment of the strength of existing reinforced concrete structures. However, little is known about the effects of blast loading on the adverse nature that governs corrosion damaged structures. Ten identical reinforced concrete beams were constructed and tested, half under blast loading and the other half under quasi-static loading. The blast loaded beams were subjected to a series of increasing blast volumes until failure was reached. Five identical beams were tested under quasi-static loading to provide a baseline comparison against the blast loaded beams. One beam from each testing group served as a control specimen and was not corroded while the remaining beams were subjected to varying levels of corrosion of the steel reinforcement. An analytical model was also created to predict the performance of corroded reinforced concrete beams under blast loading. The results of the study showed that as corrosion levels increased, the displacements increased beyond the expected response limits. Damage levels became increasingly more severe with the introduction of corrosion at all levels. The behavior changed from ductile to brittle at corrosion levels greater than 5% but remained ductile with flexural failures at corrosion levels below 5%. Overall, the introduction of corrosion to a concrete beam subjected to blast loading resulted in decreased strength and ductility across all specimens but with most disastrous effects occurring at corrosion levels of 5% or greater. A recommendation is made to adjust the response the limits in the code to account for corrosion in reinforced concrete beams.

This thesis reports on the research outcome of corrosion mechanism and corrosion rate of mild steel in different environments (saline, alkaline solutions and concrete media) using potentiodynamic polarization technique. The study also included the effect of corrosion on bond strength between reinforcing steel and concrete using pull-out test. Corrosion of mild steel and 316L stainless steel with different surface conditions in 1, 3 and 5% saline (NaCl + Distilled water) was investigated. Specimens ground with 200 and 600 grit silicon carbide grinding paper as well as 1μm surface finish (polished with 1μm diamond paste) were tested. In case of mild steel specimens, reduction in surface roughness caused increase in corrosion rate, while in 316L stainless steel corrosion rate decreased as the surface roughness improved. Metallographic examination of corroded specimens confirmed breakdown of passive region due to pitting corrosion. Corrosion of mild steel was also investigated in alkaline solution (saturated calcium hydroxide, pH =12.5) contaminated with 1, 3 and 5% saline. A series of corrosion experiments were also conducted to examine the efficiency of various concentrations of calcium nitrite (CN) on corrosion behaviour of both as-received and polished mild steel in alkaline solution containing 3% saline after 1 hour and 28 days of exposure. Corrosion rate was higher for the as-received than polished mild steel surface under the same testing conditions in NaCl alkaline solution with and without nitrites due to the effect of surface roughness. Morphology investigation of mild steel specimens in alkaline solution ii containing chlorides and nitrites showed localized pits even at nitrite concentration equal to chloride concentration. Corrosion of steel bars embedded in concrete having compressive strengths of 20, 30 and 46MPa was also investigated. The effect of 2 and 4% CN by weight of cement on corrosion behaviour of steel bar in low and high concrete strengths specimens were also studied. All reinforced concrete specimens were immersed in 3% saline solution for three different periods of 1, 7 and 15 days. In order to accelerate the chemical reactions, an external current of 0.4A was applied. Corrosion rate was measured by retrieving electrochemical information from polarization tests. Pull-out tests of reinforced concrete specimens were then conducted to assess the corroded steel/concrete bond characteristics. Experimental results showed that corrosion rate of steel bars and bond strength were dependent on concrete strength, amount of CN and acceleration corrosion period. As concrete strength increased from 20 to 46MPa, corrosion rate of embedded steel decreased. First day of corrosion acceleration showed a slight increase in steel/concrete bond strength, whereas severe corrosion due to 7 and 15 days corrosion acceleration significantly reduced steel/concrete bond strength. Addition of only 2% CN did not give corrosion protection for steel reinforcement in concrete with 20MPa strength at long time of exposure. However, the combination of good quality concrete and addition of CN appear to be a desirable approach to reduce the effect of chloride induced corrosion of steel reinforcement. At less time of exposure, specimens without CN showed higher bond strength in both concrete mixes than those with CN. After 7 days of corrosion acceleration, the higher concentration of CN gave higher bond strength in both concrete mixes. The same trend was observed at 15 days of corrosion acceleration except for the specimen with 20MPa compressive strength and 2% CN which recorded the highest deterioration in bond strength.

Steel corrosion is one of the major distress mechanisms that causes the deterioration of reinforced concrete structures around the world. It is an electrochemical reaction between the reinforcing steel and the surrounding concrete that produces a mass loss of the metal. Through the process of corrosion in reinforced concrete, iron ions get oxidized to form corrosion products (CP). Although multiple experiments and studies have been developed to understand the rheological behavior of corrosion products, this topic stays inconclusive. This work aims to characterize corrosion products at micro-scale in order to trace the progress of the formation of rust, to determine its nature and to analyse its rheological behavior in reinforced concrete. An experimental procedure to produce CP in the laboratory is also presented in this research. In addition, material characterization methods have been used to identify the iron oxide phases present in CP, determine their viscosity and rheological behavior and to study how CP flows in a porous media. In order to identify the different stages in the corrosion process, the CP was analysed at 2, 4, 6 and 8 weeks. The experiments identified four phases of iron oxide for each period. Furthermore, it was found that CP behaves as a shear-thinning slurry and as a result, its viscosity decreases with the applied shear rate. In addition, the damage caused by CP on concrete depends on the w/c ratio of the concrete mix and the exposure time to a corroding environment. The rebar mass loss results show that CP is formed in layers around the rebar, and the flow of each CP layer can differ.

No description available.

This thesis reports on the research outcome of corrosion mechanism and corrosion rate of mild steel in different environments (saline, alkaline solutions and concrete media) using potentiodynamic polarization technique. The study also included the effect of corrosion on bond strength between reinforcing steel and concrete using pull-out test. Corrosion of mild steel and 316L stainless steel with different surface conditions in 1, 3 and 5% saline (NaCl + Distilled water) was investigated. Specimens ground with 200 and 600 grit silicon carbide grinding paper as well as 1¿m surface finish (polished with 1¿m diamond paste) were tested. In case of mild steel specimens, reduction in surface roughness caused increase in corrosion rate, while in 316L stainless steel corrosion rate decreased as the surface roughness improved. Metallographic examination of corroded specimens confirmed breakdown of passive region due to pitting corrosion. Corrosion of mild steel was also investigated in alkaline solution (saturated calcium hydroxide, pH =12.5) contaminated with 1, 3 and 5% saline. A series of corrosion experiments were also conducted to examine the efficiency of various concentrations of calcium nitrite (CN) on corrosion behaviour of both as-received and polished mild steel in alkaline solution containing 3% saline after 1 hour and 28 days of exposure. Corrosion rate was higher for the as-received than polished mild steel surface under the same testing conditions in NaCl alkaline solution with and without nitrites due to the effect of surface roughness. Morphology investigation of mild steel specimens in alkaline solution ii containing chlorides and nitrites showed localized pits even at nitrite concentration equal to chloride concentration. Corrosion of steel bars embedded in concrete having compressive strengths of 20, 30 and 46MPa was also investigated. The effect of 2 and 4% CN by weight of cement on corrosion behaviour of steel bar in low and high concrete strengths specimens were also studied. All reinforced concrete specimens were immersed in 3% saline solution for three different periods of 1, 7 and 15 days. In order to accelerate the chemical reactions, an external current of 0.4A was applied. Corrosion rate was measured by retrieving electrochemical information from polarization tests. Pull-out tests of reinforced concrete specimens were then conducted to assess the corroded steel/concrete bond characteristics. Experimental results showed that corrosion rate of steel bars and bond strength were dependent on concrete strength, amount of CN and acceleration corrosion period. As concrete strength increased from 20 to 46MPa, corrosion rate of embedded steel decreased. First day of corrosion acceleration showed a slight increase in steel/concrete bond strength, whereas severe corrosion due to 7 and 15 days corrosion acceleration significantly reduced steel/concrete bond strength. Addition of only 2% CN did not give corrosion protection for steel reinforcement in concrete with 20MPa strength at long time of exposure. However, the combination of good quality concrete and addition of CN appear to be a desirable approach to reduce the effect of chloride induced corrosion of steel reinforcement. At less time of exposure, specimens without CN showed higher bond strength in both concrete mixes than those with CN. After 7 days of corrosion acceleration, the higher concentration of CN gave higher bond strength in both concrete mixes. The same trend was observed at 15 days of corrosion acceleration except for the specimen with 20MPa compressive strength and 2% CN which recorded the highest deterioration in bond strength.

<p>Flow accelerated corrosion(FAC) results in wall thinning of pipes, tubes or vessels from exposure to flow due to corrosion. If FAC is not detected, it can lead to sudden failure of piping components. Orifices are used in piping systems to monitor and control the flow. Flow separation and reattachment downstream of an orifice can enhance the mass transfer of the pipe wall. In this thesis, the effect of Reynolds numbers and surface roughness on the mass transfer rate downstream of an orifice was investigated. A dissolving wall method was used to measure the wall mass transfer. The test sections were cast from gypsum with water as the working fluid. Multiple destructive tests were performed for different test times in a 2.5 cm diameter flow loop, and the wear topology measured by a laser scanner to obtain the progression of wear with time over the pipe surface. None-destructive tests were performed in a 20 cm diameter flow loop using test section with an inner gypsum lining, and measured online at selected locations using an ultrasonic method. Experiments were performed at Reynolds numbers of 80000, 140000 and 200000 in the 2.5 cm diameter flow loop, and at 180,000 in the 20 cm diameter flow loop with an orifice to pipe diameter ratio of 0.5. The results show that different surface roughness patterns are developed at different Reynolds numbers from the initially smooth surfaces. The different surface roughness patterns have a significantly different effect on the mass transfer rate downstream of an orifice. A larger population of scallops developed from the smooth pipe surface, as the Reynolds number was increased, which enhanced the mass transfer rate. The mass transfer rate in the 20 cm diameter test section was much smaller than in the 2.5 cm diameter test section at a similar Reynolds number. The pattern of the roughness in the 20 cm diameter test section was formed as isolated roughness which is similar to the roughness pattern in 2.5 cm diameter test section at much lower Reynolds number.</p> / Master of Applied Science (MASc)