Thesis (MEng)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: The building industry is responsible for a substantial contribution to pollution. The production of building materials, as well as the operation and maintenance of structures leads to large amounts of carbon-dioxide (CO2) being release in the atmosphere. The use of renewable resources and construction materials is just one of the ways in which the carbon footprint of the building industry can be reduced. Sisal fibre is one such renewable material. Sisal fibre is a natural fibre from the Agave Sisalana plant. The possibility of incorporating sisal fibre in a cement-based matrix to replace conventional steel and synthetic fibres has been brought to the attention of researchers. Sisal fibre has a high tensile strength in excess of polypropylene fibre and comparable to PVA fibre. Sisal fibre consists mainly of cellulose, hemi-cellulose and lignin. The disadvantage of incorporating sisal fibre in a cement-based matrix is the degradation of the composite. Sisal fibres tend to degrade in an alkaline environment due to changes in the morphology of the fibre. The pore water in a cement base matrix is highly alkaline which leads to the degradation of the fibres and reduced strength of the composite over time. Sisal fibre reinforced cement-based composites (SFRCC) were investigated to evaluate the durability of the composites. Two chemical treatments, alkaline treatment and acetylation, were performed on the fibre at different concentrations to improve the resistance of the fibre to alkaline attack. Alkaline treatment was performed by using sodium hydroxide (NaOH), while acetylation was performed by using acetic acid or acetic anhydride. Single fibre pull-out (SFP) tests were performed to evaluate the influence of chemical treatment on fibre strength, to study the fibre-matrix interaction and to determine a critical fibre length. A matrix consisting of ordinary Portland cement (OPC), sand and water were used for the SFP tests. This matrix, as well as alternative matrices containing fly ash (FA) and condensed silica fume (CSF) as supplementary cementitious material, were reinforced with 1% sisal fibre (by volume) cut to a length of 20 mm. The OPC matrix was reinforced with untreated- and treated fibre while the alternative matrices were reinforced with untreated fibre. Alternative matrices containing varying fibre volumes and lengths were also produced. Three-point bending- (indirect), direct tensile- and compression tests were performed on specimens at an age of 28 days to determine the strength of the matrix. The remainder of the specimens were subjected to ageing by extended curing in water at 24˚C and 70˚C respectively and by alternate cycles of wetting and drying, after which it was tested at an age of 90 days from production to evaluate the durability of the fibre. An increase in fibre volume led to a decrease in compressive strength and peak tensile strength. The optimum fibre length at a volume of 1% was 20 mm for which the highest compression strength was recorded. The combination of alkali treatment and acetylation was the most effective treatment condition, followed by alkali treatment at low concentrations of sodium hydroxide. At higher concentrations of sodium hydroxide, a significant reduction in strength was recorded. The addition of supplementary cementitious materials also proved to be effective in mitigating degradation, especially in the cases where CSF was used. FA proved to be less effective in reducing the alkalinity of the matrix. However, the use of FA as fine filler resulted in higher strengths. Specimens manufactured by extrusion did not have superior mechanical properties to cast specimens. The conclusion was made that the use of sisal fibre in a cement-based matrix is effective in providing ductile failure. Chemical treatment and the addition of supplementary cementitious materials did improve the durability of the specimens, although degradation still took place. / AFRIKAANSE OPSOMMING: Die boubedryf is verantwoordelik vir 'n aansienlike bydrae tot besoedeling. Die produksie van boumateriale, sowel as die bedryf en instandhouding van strukture lei tot groot hoeveelhede koolstof dioksied (CO2) wat in die atmosfeer vrygestel word. Die gebruik van hernubare hulpbronne en boumateriale is maar net een van die maniere waarop die koolstof voetspoor van die boubedryf verminder kan word. Sisal vesels is 'n voorbeeld van 'n hernubare materiaal. Sisal vesel is 'n natuurlike vesel afkomstig vanaf die Agave Sisalana plant. Die moontlikheid om sisal vesels in 'n sement gebasseerde matriks te gebruik om konvensionele staal en sintetiese vesels te vervang, is tot die aandag van navorsers gebring. Sisal vesel het 'n hoër treksterkte as polipropileen vesels en die treksterkte vergelyk goed met die van PVA vesels. Sisal vesel bestaan hoofsaaklik uit sellulose, hemi-sellulose en lignien. Die nadeel verbonde aan die gebruik van sisal vesels in 'n sement gebasseerde matriks is die degradasie van die komposiet. Sisal vesels is geneig om af te breek in 'n alkaliese omgewing as gevolg van veranderinge wat in die morfologie van die vesel plaasvind. Die water in die porieë van 'n sement gebasseerde matriks is hoogs alkalies wat lei daartoe dat die vesel afgebreek word en die sterkte van die komposiet afneem oor tyd. Sisal vesel versterkte sement gebasseerde komposiete is ondersoek om die duursaamheid van die komposiete te evalueer. Twee chemiese behandelings, alkaliese behandeling en asetilering, is uitgevoer op die vesels teen verskillende konsentrasies om die weerstand van die vesels teen alkaliese aanslag te verbeter. Alkaliese behandeling was uitgevoer met natrium-hidroksied (NaOH) terwyl asetilering met asynsuur en asynsuurhidried uitgevoer is. Enkel vesel uittrek toetse is uitgevoer om die invloed van chemiese behandeling op veselsterkte te evalueer, om die vesel/matriks interaksie te bestudeer en om die kritiese vesellengte te bepaal. 'n Matriks wat uit gewone Portland sement (OPC), sand en water bestaan, is gebruik vir die enkel vesel uittrek toetse. Dieselfde matriks, sowel as alternatiewe matrikse wat vliegas (FA) en gekondenseerde silika dampe (CSF) as aanvullende sementagtige materiaal bevat, is versterk met 1% vesel (by volume) wat 20 mm lank gesny is. Die OPC matriks was versterk met onbehandelde- en behandelde vesels, terwyl die alternatiewe matrikse met onbehandelde vesels versterk is. Matrikse wat wisselende vesel volumes en lengtes bevat het is ook vervaardig. Drie-punt buigtoetse (indirek), direkte trek toetse en druktoetse is uitgevoer op proefstukke teen 'n ouderdom van 28 dae om die sterkte van die matriks te bepaal. Die oorblywende proefstukke is onderwerp aan veroudering deur verlengde nabehandeling in water teen 24˚C en 70˚C onderskeidelik en deur afwissilende siklusse van nat- en droogmaak waarna dit op 'n ouderdom van 90 dae vanaf vervaardiging getoets is om die duursaamheid van die matriks te evalueer. 'n Toename in vesel volume het tot 'n afname in druksterkte en piek treksterkte gelei. Die optimum vesel lengte teen 'n volume van 1% was 20 mm, waarvoor die hoogste druksterkte opgeteken is. Die kombinasie van alkaliese behandeling en asetilering was die mees effektiewe behandeling, gevolg deur alkaliese behandeling by lae konsentrasies natrium-hidroksied. Vir hoë konsentrasies natrium-hidroksied is 'n aansienlike afname in sterkte opgeteken. Die toevoeging van aanvullende sementagtige materiale was ook effektief om die degradadering van die vesels te verminder, veral in die gevalle waar CSF gebruik is. FA was minder effektief om die alkaliniteit van die matriks te verminder. Die gebruik van FA as fyn vuller het nietemin hoër sterkte tot gevolg gehad. Proefstukke wat deur ekstrusie vervaardig is, het nie beter meganiese eienskappe gehad as proefstukke wat gegiet is nie. Daar is tot die gevolgtrekking gekom dat sisal vesel in 'n sement gebasseerde matriks wel effektief is om 'n duktiele falingsmode te voorsien. Chemiese behandeling en die toevoeging van aanvullende sementagtige materiale het die duursaamheid van die proefstukke verbeter, alhoewel degradering steeds plaasgevind het.

No description available.

No description available.

Ce travail traite de l'optimisation de la formulation et de la durabilité des micro-bétons de résine. Ces matériaux peuvent être utilisés en substitution aux bétons cimentaires pour la réalisation d'éléments structuraux exposés à des agressions chimiques et/ou climatiques. Ils résultent de l'association entre un polymère, le plus souvent thermodurcissable, et des granulats. Le béton qui en résulte est connu sous la dénomination anglo-saxonne “ Polymer Concrete (PC) ”. On le désignera béton à matrice organique “ BMO ” également connu sous l'appellation de “ Béton Résineux ”. Ces matériaux souffrent de leur coût élevé qui constitue un frein quant à leur développement. Aussi, la première partie de ce travail a consisté à étudier le comportement mécanique et physique de ces matériaux formulés avec des fractions massiques en polymère jouant le rôle de liant qui varient entre 5 et 13%. Le polymère est un époxyde de type DGEBA réticulée avec une diamine aliphatique. L'objectif est de trouver une formulation de micro béton résineux (MBR) économique alliant résistance mécanique, durabilité et coût. On montre qu'une fraction massique de 9% de polymère «est un optimum. Au-delà de ce pourcentage les caractéristiques mécaniques et physiques des MBR ne varient plus et en dessous elles diminuent. Afin de diminuer encore le pourcentage d'époxyde, des fillers calcaires ont été incorporés dans la formulation des MBR. On montre qu'une formulation avec une fraction massique en époxyde de 7% et un dosage en fillers de 10% et de 83% en granulats conduit à un MBR ayant des propriétés physiques et des résistances mécaniques plus importantes que celles du MBR formulé avec 9% d'époxyde. La résistance des MBR à base d'époxyde optimaux avec incorporation de fillers calcaires et sans aux attaques chimiques et au gel/dégel modéré a été étudiée et comparée à celle d'un micro béton hydraulique (MBH). Les résultats ont montré que les deux formulations de MBR ont une durabilité meilleure que celle de MBH. / X

This study combined research available through literature with extensive experimental studies and substantial physical modelling to estimate the remaining ultimate life of large offshore reinforced concrete structures. Although much research has focussed on concrete degradation due to chloride ingress, corrosion of permanently submerged concrete is regarded as negligible due to the long-assumed apparent worst case of tidal or splash zone exposure. Around 350 specimens were tested with a further 200 exposed for further testing by future research groups. Specimens ranged in size from standard cubes to various beam lengths up to 1.5 metres, allowing for material and structural properties to be assessed. My original contribution to knowledge in the sector enhances the fundamental understanding of corrosion in subsea concrete, challenging the generally held belief of negligible corrosion. Results and modelling provides an improved ability to ultimately estimate the longevity of fully submerged offshore reinforced concrete. Throughout this thesis, the results from experimental works, carried out as a direct result of the lack of data or information in literature, are reported, assessed and then utilised to provide updated ultimate life estimations. With the current offshore concrete structures currently coming to the end of their service life, and the likelihood of further offshore development using concrete for the renewables sector, understanding the long-term degradation is vital in determining the most effective decommissioning and derogation options. The research carried out directly provides detailed information of the likely time-to-failure, allowing for an informed decision to be made on operational and decommissioning plans. Experimental work was carried out over four main phases; corrosion initiation due to bulk diffusion of chlorides (Phase I), corrosion propagation in low oxygen environments (Phase II), corrosion in statically and dynamically cracked sections (Phase III) and structural response of heavily corroded individual and lapped bar sections (Phase IV). Phase I work shows a marked difference between submerged exposures to seawater as opposed to NaCl solution, the unsuitability for accelerated testing with seawater and the likelihood of rapid initiation in offshore structures. Further experimental works through Phases II and III found that although exposed to low oxygen concentrations, reinforcement corrosion continued at significant rates. A variation between anode sizes on the reinforcement is noted, but critically the cross sectional area of the steel was still reduced, albeit in fewer locations. Corrosive products were visibly different, with fewer expansive products, if any, present. Additionally, this study further highlights the importance of cracking on corrosion, currently ignored by recent model codes, such as the fib Model Code 2010, up to 0.2mm crack width. A linear relationship was found between crack width and corrosion rates, with cracking above 0.1mm considered significant. The loss of cross sectional area due to propagation was determined for the given environment, and consequently further studies were initiated in an attempt to determine the relationship between this corrosion propagation and the reduced serviceability or ultimate life of concrete beams. Serviceability, defined by beam stiffness, was reduced due to bond loss along reinforcement. Most importantly, however, results prove that the loss of cross sectional area to be the critical influence on loss of ultimate life. Initial estimates on the remaining ultimate life of the large offshore structures support early rough work that the structures would last centuries. This thesis, however, has shown this is due to the ability of concrete structures with such large volumes of steel to continue to ultimately withstand loading at high corrosion percentages and not due to negligible corrosion, or long initiation periods, commonly suggested in submerged, low oxygen environments.

Although concrete is widely considered a very durable material, if conditions are such, it can be vulnerable to deterioration and early distress development. Alkali-Silica Reaction (ASR) is a major durability problem in concrete structures. It is a chemical reaction between the reactive silica existent in some types of rocks and alkali hydroxides in the concrete pore water. The product of this reaction is a gel that is hygroscopic in nature. When the gel absorbs moisture, it swells leading to tensile stresses in concrete. When those stresses exceed the tensile strength of concrete, cracks occur. The main objective of this study was to address a method of testing concrete materials as a combination to assist engineers to effectively mitigate ASR in concrete. The research approach involved capturing the combined effects of concrete materials (water cement ratio, porosity, supplementary cementitious materials, etc.) through a method of testing to allow the formulation of mixture combinations resistant to ASR leading to an increase in the life span of concrete structures. To achieve this objective, a comprehensive study on different types of aggregates of different reactivity was conducted to formulate a robust approach that takes into account the factors affecting ASR; such as, temperature, moisture, calcium concentration and alkalinity. A kinetic model was proposed to determine aggregate ASR characteristics which were calculated using the System Identification Method. Analysis of the results validates that ASR is a thermally activated process and therefore, the reactivity of an aggregate can be characterized in terms of its activation energy (Ea) using the Arrhenius equation. Statistical analysis was conducted to determine that the test protocol is highly repeatable and reliable. To relate the effect of material combinations to field performance, concrete samples with different w/cm?s and fly ash contents using selective aggregates were tested at different alkalinities. To combine aggregate and concrete characteristics, two models were proposed and combined. The first model predicts the Ea of the aggregate at levels of alkalinity similar to field conditions. The second model, generated using the Juarez- Badillo transform, connects the ultimate expansion of the concrete and aggregate, the water cement ratio, and the fly ash content to the Ea of the rock. The proposed models were validated through laboratory tests. To develop concrete mixtures highly resistant to ASR, a sequence of steps to determine threshold total alkali in concrete were presented with examples. It is expected that the knowledge gained through this work will assist government agencies, contractors, and material engineers, to select the optimum mixture combinations that fits best their needs or type of applications, and predict their effects on the concrete performance in the field.

Polymer coated sheet metal components find diverse applications in many industries. The manufacturing of the components generally involves forming of sheet metal into the desired shape and coating of the formed part with organic coating. An alternative manufacturing route is to coat the sheet metal first before forming. The change in the manufacturing sequence can potentially improve cost and reduce environmental impact. This approach, however, requires the coating to survive the deformation process. Thus, the effect of plastic deformation on coating adhesion is of primary interest to many engineers and researchers. This research aims at developing a methodology to predict the adhesion of coating after metal forming processes. A pull-off apparatus that measures the coating pull-off stress was used to indicate the coating adhesion strength. Several types of specimen were designed to obtain uniaxial tension, biaxial tension, and tension-compression deformation modes on pre-coated sheet by using a uniaxial tensile tester. Experimental results from two selected polymer coated sheet metals show that coating adhesion was affected by plastic deformation. An analytical model based on a virtual interface crack concept was developed to indicate the adhesion potential of the coating-substrate interface. From interfacial fracture mechanics, the initial adhesion potential is defined as the energy release rate characterized by the virtual interface crack and the initial pull-off stress. The analytical model was used to predict coating adhesion loss after deformation in uniaxial tension mode. The analytical model predictions agreed well with experimental results. Finite element analysis tool was applied to simulate more complex deformation modes in stamping of coated sheet meals. The stress field near the interface crack tip was used to calculate the energy release rate and predict the adhesion loss under different deformation modes. The predictions obtained from numerical method are also in good agreements with the experimental results in biaxial tension and tension-compression modes. The research has led to a better understanding of the effects of plastic deformation on coating adhesion. The developed adhesion test methods can be used to generate useful information on coating durability for diverse practical use. It is also expected that the results of the research will facilitate the development of better polymer coated sheet metal to be used in sheet metal forming processes.

The objective of this project was to evaluate and model the effects of moisture, temperature, and combined hygrothermal aging on the durability of a graphite/epoxy woven composite material system. Imposed environmental and aging conditions were considered to be representative of service conditions for the engine of an advanced subsonic aircraft for which the composite system is a candidate material. The study was designed such that the results could be used in a residual strength based life prediction approach that accounted for both the mechanical fatigue and environmental conditions. Damage mechanisms and failure modes were determined through fatigue testing, residual strength testing, and nondestructive evaluation. The experimental data generally revealed little effect of environment on strength degradation during fatigue despite notable differences in damage accumulation processes. Modeling efforts were concentrated on initial stiffness, moisture uptake, and residual strength prediction, where the results from the first two efforts were intended to generate inputs for the life prediction. The Ishikawa and Chou fiber undulation and bridging model [22] was shown to provide an accurate stiffness prediction and was subsequently used in parametric studies to determine the effect of weave architecture and geometry. A moisture uptake model developed by Roy [16] for laminates containing single direction cracks was extended to predict moisture uptake in laminates containing cracks in directions parallel and transverse to the loading direction. The life prediction approach was based on ideas developed by Reifsnider and colleagues [36,37,43]. The intention in this case was to use the critical element paradigm to predict the combined effects of alternating environmental (temperature and moisture) conditions imposed during fatigue. Since experimental results indicated that temperature and moisture did not significantly affect the strength and life of the material, a successful life prediction analysis was performed as a function of only fatigue stress level and cycles. / Master of Science

This thesis presents the experimental and analytical results to investigate the effect of corrosion on the mechanical properties of reinforcing bars and concrete beams, with particular reference to their ductility. In the experimental works, specimens were electrochemically corroded, before they were loaded to failure. In the finite element analysis, the corrosion of reinforcement was modelled as either internal pressure or radial expansion around corroded bars. The study indicates that the amount of corrosion to cause cracking at the bar and concrete surfaces almost linearly increased with the bar diameter and ratio of cover to diameter, respectively. No matter whether concrete cover c increased or bar distance S decreased, once the ratio of S / c became less than 2.5, corrosion cracks first propagated internally between the bars and caused delamination. Although corrosion did not alter the shape of force-extension curves substantially, it decreased bar strength and, especially, ductility greatly. Furthermore, although the reductions of strengths were identical, the ductility of bars corroded in concrete decreased more rapidly than that of bare corroded bars. Corrosion decreased beam strength and altered its ductility and failure mode. When the cracking of compressive concrete or the reduction of tensile bar area dominated beam response, corrosion increased beam ductility and caused a beam to fail in a less brittle and even ductile manner. When the deterioration of bond strength or the reduction of steel ductility controlled beam behaviour, however, corrosion decreased beam ductility and led the beam to fail in a less ductile and even brittle manner. There is a concern regarding the ductility of reinforcing bars and under-reinforced beams if the amount of corrosion exceeds 100/0, since bar ultimate strain decreased below the minimum requirements prescribed in the Model Code 90 for situations requiring high ductility.

Concrete repair materials are being widely used to restore and extend the service life of structures. While most cement-based repair materials are compatible with concrete structures, their durability properties do not attract much attention which it deserves from researchers. Since repair materials can deteriorate like conventional concrete, the search for reliable, long-lasting concrete repair materials is becoming more intensive. Amongst other factors, concrete permeability and chloride diffusivity within concrete are believed to play a major role in determining the durability and success of the repair. These two parameters determine the penetration rate of aggressive substances into concrete and how fast degradation could take place. A number of test methods have been proposed to study these two factors, and the commonly used test methods are water penetration, surface/bulk electrical resistivity, rapid chloride permeability (RCP), and half-cell potential. However, the relationship between each durability test method and their correlation with compressive strength measurement have not been fully understood. So, in this study, we aim for using multiple testing techniques, destructive and non-destructive, to evaluate the durability of concrete repair materials as well as correlating different test methods. Three types of commercially available cement-based materials are tested and evaluated, and results have indicated that cementitious concrete mortar (termed as Mix M) amongst others has the best durability performance which means low water permeability, high resistivity, and compressive strength. Whereas, the flexural performance of Mix M still needs some improvement in terms of flexural strength and flexural toughness. For various durability testing methods, surface resistivity is found to have a strong linear relation and a polynomial relation to bulk resistivity and water permeability respectively. No relationship is established between concrete resistivity and compressive strength, though high-strength concrete tends to have a high resistivity in our study. RCP test results do not correlate well with resistivity measurements, which requires further study to overcome its heating and binding effect when measurements are being taken. Half-cell potential method is used for validating test results but it reveals no difference for materials with different permeability and resistivity. A model is proposed to counteract temperature’s effect while calculating the coefficient of diffusion, which indicates the concrete to resist chloride diffusion. It is found that this model can shift the RCP measurement slightly closer to its theoretical prediction but the difference between them is still large. Therefore, further research is required for acquiring more raw data from RCP measurements as the regression analysis input. In addition, a more comprehensive model that involves more correction factors for binding effects, etc., is also needed. / Graduate / 2020-04-30