Wood is an interesting alternative to fossil fuels. It is CO2-neutral and widely available. However it is a difficult fuel to handle which features a low energy content. Thus technologies for wood thermal conversion need to be improved. This work concerns the development of a comprehensive two-dimensional mathematical model describing the pyrolysis of large wood particles and its implementation in a Fortran program. The model has been continuously tested and improved by experimental results obtained in a reactor for single particle pyrolysis (SPAR) at the Division of Physical Chemistry at Göteborg University. The first part of the thesis (Paper I) presents a kinetic study of the pyrolysis of large wood particles, based on experiments carried out in the SPAR. Three pyrolysis kinetic schemes were selected for later inclusion in a model featuring heat and mass transfer. Paper II concerns the addition of a sub-model for heat and mass transfer to the three kinetic schemes. The resulting model for large wood particles has been tested against experiments in the SPAR. A scheme based on two competing reactions developed from experiments at low temperature pyrolysis in the SPAR was found to perform well but its empirical nature limits its validity to the experimental conditions of the SPAR. A scheme from the literature based on TGA experiments appeared promising, especially when planning to enhance it with secondary reactions. Paper III deals with the development of shrinkage models for 2D cylindrical particles. The predicted mass loss, size variation and surface temperature were tested against experiments carried out in the SPAR. The shrinkage does not a?ect the pyrolysis rate or the surface temperature in the conditions prevailing in the SPAR. Paper IV investigates the influence of different shrinkage models and the geometry on the heating rate of a shrinking particle. Shrinkage influences the heating rate positively by increasing the conductive heat flow and negatively by decreasing the surface area of the particle. Therefore the net effect of shrinkage on the heating rate depends on the particle geometry and the location of shrinkage. Paper V studies three di?erent models for wood drying under pyrolysis conditions. The predicted surface temperature and global drying rate were compared with experimental results from pyrolysis experiments of wet particles in the SPAR. A model based on a first order kinetic evaporation rate was found to be the most interesting because of the quality of the prediction of the drying rate and the ease of implementation. / QC 20100624

The large amount of used polypropylene baler twine generated from the agricultural community may provide a low-cost, environmentally friendly source of fibre reinforcement that can be used to improve the properties of concrete. However, the performance of such fibres for the application has not yet been explored. The effectiveness of using small amounts of chopped baler twine to control the restrained plastic shrinkage cracking of portland cement mortar was investigated in this study. To determine the influence of baler twine fibre type, length and volume fraction on their performance, two types of baler twine ( one composed of strands with circular cross section, the other composed of flat band shape strands) in two lengths (19 mm and 38 mm) and three volume fractions (0.05%, 0.1%, and 0.3%) were evaluated. To compare the performance of baler twine fibre with that of other commercially available synthetic fibres, fibrillated polypropylene fibres at equal lengths and volume fractions was investigated.<p>The restrained plastic shrinkage tests were carried out by subjecting the fibre-reinforced mortar specimens, cast on rough substrate bases, to a wind speed of 2.6 m/s, and relative humidity less than 3% at 35 °C for 22 hours. To evaluate the effectiveness of the fibres, the crack numbers were recorded, and the maximum crack width and total crack area on the surface of each specimen were measured using an image analysis technique. Unrestrained plastic shrinkage tests were also conducted in which fibre-reinforced mortar specimens without the substrate bases were tested under the same environmental conditions.<p>Test results indicate that both types of baler twine are capable of controlling restrained plastic shrinkage cracking to some extent, but are not as effective as fibrillated polypropylene. The baler twine composed of band shape strands performed better than the one composed of strands with circular cross section. Compared with plain specimens, the total crack area was reduced by 95.3, 77.5 and 38.7% when 0.3% volume fraction of 38 mm fibrillated polypropylene, band shape baler twine and circular baler twine fibres, respectively, were added. Similar reductions in maximum crack width were observed. Fibre length did not significantly influence cracking behaviour. Free plastic shrinkage was significantly reduced only when long fibre lengths (38 mm) and high volume fractions (0.3%) were used.

The large amount of used polypropylene baler twine generated from the agricultural community may provide a low-cost, environmentally friendly source of fibre reinforcement that can be used to improve the properties of concrete. However, the performance of such fibres for the application has not yet been explored. The effectiveness of using small amounts of chopped baler twine to control the restrained plastic shrinkage cracking of portland cement mortar was investigated in this study. To determine the influence of baler twine fibre type, length and volume fraction on their performance, two types of baler twine ( one composed of strands with circular cross section, the other composed of flat band shape strands) in two lengths (19 mm and 38 mm) and three volume fractions (0.05%, 0.1%, and 0.3%) were evaluated. To compare the performance of baler twine fibre with that of other commercially available synthetic fibres, fibrillated polypropylene fibres at equal lengths and volume fractions was investigated.<p>The restrained plastic shrinkage tests were carried out by subjecting the fibre-reinforced mortar specimens, cast on rough substrate bases, to a wind speed of 2.6 m/s, and relative humidity less than 3% at 35 °C for 22 hours. To evaluate the effectiveness of the fibres, the crack numbers were recorded, and the maximum crack width and total crack area on the surface of each specimen were measured using an image analysis technique. Unrestrained plastic shrinkage tests were also conducted in which fibre-reinforced mortar specimens without the substrate bases were tested under the same environmental conditions.<p>Test results indicate that both types of baler twine are capable of controlling restrained plastic shrinkage cracking to some extent, but are not as effective as fibrillated polypropylene. The baler twine composed of band shape strands performed better than the one composed of strands with circular cross section. Compared with plain specimens, the total crack area was reduced by 95.3, 77.5 and 38.7% when 0.3% volume fraction of 38 mm fibrillated polypropylene, band shape baler twine and circular baler twine fibres, respectively, were added. Similar reductions in maximum crack width were observed. Fibre length did not significantly influence cracking behaviour. Free plastic shrinkage was significantly reduced only when long fibre lengths (38 mm) and high volume fractions (0.3%) were used.

Specifying a prior distribution for the large number of parameters in the linear statistical model is a difficult step in the Bayesian approach to the design and analysis of experiments. Here we address this difficulty by proposing the use of functional priors and then by working out important details for three and higher level experiments. One of the challenges presented by higher level experiments is that a factor can be either qualitative or quantitative. We propose appropriate correlation functions and coding schemes so that the prior distribution is simple and the results easily interpretable. The prior incorporates well known experimental design principles such as effect hierarchy and effect heredity, which helps to automatically resolve the aliasing problems experienced in fractional designs. The second part of the thesis focuses on the analysis of optimization experiments. Not uncommon are designed experiments with their primary purpose being to determine optimal settings for all of the factors in some predetermined set. Here we distinguish between the two concepts of statistical significance and practical significance. We perform estimation via an empirical Bayes data analysis methodology that has been detailed in the recent literature. But then propose an alternative to the usual next step in determining optimal factor level settings. Instead of implementing variable or model selection techniques, we propose an objective function that assists in our goal of finding the ideal settings for all factors over which we experimented. The usefulness of the new approach is illustrated through the analysis of some real experiments as well as simulation.

The present research seeks to study the decrease in diffusivity rate as relative humidity (RH) decreases and modeling drying shrinkage of hardened cement paste as a poroviscoelastic respose. Thin cement paste strips of 0.4 and 0.5 w/c at age 3 and 7 days were measured for mass loss and shrinkage at small RH steps in an environmental chamber at constant temperature. Non-linear drying diffusion rate of hardened cement was modeled with the use of Fick's second law of diffusion by assuming linearity of diffusion rate over short drops of ambient relative humidity. Techniques to determine drying isotherms prior to full equilibration of mass loss, as well as converting mass loss into concentration of water vapor were developed. Using the measured water vapor diffusivity, drying shrinkage strain was modeled by the theory of poroviscoelasticity. This approach was validated by determining viscoelastic properties from uniaxial creep tests considering the effect of aging by the solidification theory. A change in drying diffusion rate at different RH was observed in the 0.4 and 0.5 w/c pastes at different ages. Drying diffusion rate decreases as RH drops. This can be attributed to a change in diffusion mechanisms in the porous media at smaller pore radius. Shrinkage modeling with an average diffusion coefficient and with determined viscoelastic parameters from creep tests agreed well compared to the shrinkage data from experiments, indicating that drying shrinkage of cement paste may be considered as a poroviscoelastic reponse.

Calcium aluminate cement (CAC) is a rapid hardening cementitious material often used in niche concrete repairs where high early-age strength and robust durability are required. This research project characterized the implications of the additions of various mineral and chemical admixtures to plain CAC to mitigate strength reductions associated with conversion, an inevitable strength reduction associated with the densification of metastable hydrates (CAH10 and C2AH8) to stable hydrates (C3AH6 and AH3). The effect of these admixtures on early-age strength development, volume change, and the correlation to macro-scale performance were reported in this dissertation. Various mixtures of CAC were investigated including: pure CAC, binary blends of CAC with fly ash (Class C) or CaCO3, and ternary blends of CAC with slag and silica fume. Characterization of the influence of these admixtures on hydration was completed using x-ray diffraction, isothermal calorimetry, and chemical shrinkage. Investigations on the implications of early-age volume change were conducted for autogenous deformation. In addition to laboratory testing, the final phase of the project was to correlate and elucidate the data generated in the laboratory to real-world field performance. Field trials were conducted to evaluate and monitor the behavior of CAC systems and investigate the link between laboratory generated research and actual large scale behavior. / text

Prestressed concrete construction relies on the application of compressive stresses to concrete elements. The prestressing force is typically applied through the tensioning of strands that react against the concrete and induce compression in the concrete. Loss of prestress is the decrease of this pre-applied stress. The conservative estimation of the prestress losses is imperative to prevent undesired cracking of the prestressed element under service loads. A large fraction of the prestress losses is a consequence of concrete deformations. This fraction of the losses can be identified as strain-related losses, and these occur due to instantaneous elastic shortening, and time-dependent creep and shrinkage. Creep and shrinkage of concrete depend on many factors that are extremely variable within concrete structures. The time-dependent behavior of concrete is not well-understood, but recent findings in the topics of concrete creep and shrinkage provide a better understanding of the underlying mechanisms affecting the nature of these two phenomena. However, current design practices and prestress loss estimation methods do not reflect the state-of-the-art knowledge regarding creep and shrinkage. The main objective of this dissertation was the study and estimation of strain-related prestress losses in simply supported pretensioned bridge girders. Simply supported pretensioned girders are widely designed, produced and frequently used in bridge construction. Due to this common use, pretensioned concrete bridge girders has become fairly standardized elements, which results in a reduced variability in the behavior of pretensioned bridge girders, as compare to that of less standardized concrete structures. Hence, a simplified method was calibrated to estimate prestress losses within pretensioned girders to an adequate level of accuracy. To achieve an acceptable accuracy experimental data from the monitoring of pretensioned simply supported girders was used for the calibration of the method. The accuracy of this simplified method is comparable to that achievable using more elaborate methods developed for generic concrete structures. / text

In prestressed precast segmental concrete bridges, conventional longitudinal reinforcement serves only as shrinkage crack controllers. The presence of this reinforcement, however, has restricted the ability to reduce the cross-section of the segments when high strength concrete is used because of the minimum dimensions required to accomodate the reinforcement. Research on fibre reinforced concrete (FRC) indicated that the addition of steel fibres to concrete significantly improved the tensile behaviour and the crack control characteristics of the concrete. This research investigates the feasibility of fibres to replace the conventional shrinkage reinforcement, allowing for the design of thinner and lighter structures with comparable or better crack control characteristics. Extensive work was conducted to investigate the effectiveness of hooked-end steel fibres to control cracks. Seven types of material tests were performed: uniaxial tension test, cylinder compression test, modulus of rupture test, splitting test, free and autogenous shrinkage test, and restrained shrinkage test. In addition, ten 890×890×70 mm concrete panels were tested under in-plane pure-shear loading using the Panel Element Tester. The parameters of study were the fibre volume content (0.5%, 1.0%, and 1.5%), the concrete compressive strength (50 and 80 MPa), and the fibre geometry and tensile strength. In addition to the experimental study, a model was developed to investigate the behaviour of a 1D restrained FRC member subjected to shrinkage. The experimental results indicated that the addition of fibres significantly improved the behaviour of the concrete, particularly the crack control characteristics, the post-peak compressive response, the post-cracking tensile response, the toughness, and the ductility of the concrete. The results also indicated that steel fibres were as effective as conventional reinforcement in controlling shrinkage cracking, provided that sufficient fibre volume content was added to the concrete. For example, in order to achieve a maximum crack width of 0.35 mm, a minimum fibre content of 0.9% and 1.1% should be provided for 50 MPa FRC containing high aspect ratio fibres and low aspect ratio fibres, respectively. In addition, the results indicated the importance of fibre content and fibre aspect ratio on the effectiveness of fibre reinforcement.

In prestressed precast segmental concrete bridges, conventional longitudinal reinforcement serves only as shrinkage crack controllers. The presence of this reinforcement, however, has restricted the ability to reduce the cross-section of the segments when high strength concrete is used because of the minimum dimensions required to accomodate the reinforcement. Research on fibre reinforced concrete (FRC) indicated that the addition of steel fibres to concrete significantly improved the tensile behaviour and the crack control characteristics of the concrete. This research investigates the feasibility of fibres to replace the conventional shrinkage reinforcement, allowing for the design of thinner and lighter structures with comparable or better crack control characteristics. Extensive work was conducted to investigate the effectiveness of hooked-end steel fibres to control cracks. Seven types of material tests were performed: uniaxial tension test, cylinder compression test, modulus of rupture test, splitting test, free and autogenous shrinkage test, and restrained shrinkage test. In addition, ten 890×890×70 mm concrete panels were tested under in-plane pure-shear loading using the Panel Element Tester. The parameters of study were the fibre volume content (0.5%, 1.0%, and 1.5%), the concrete compressive strength (50 and 80 MPa), and the fibre geometry and tensile strength. In addition to the experimental study, a model was developed to investigate the behaviour of a 1D restrained FRC member subjected to shrinkage. The experimental results indicated that the addition of fibres significantly improved the behaviour of the concrete, particularly the crack control characteristics, the post-peak compressive response, the post-cracking tensile response, the toughness, and the ductility of the concrete. The results also indicated that steel fibres were as effective as conventional reinforcement in controlling shrinkage cracking, provided that sufficient fibre volume content was added to the concrete. For example, in order to achieve a maximum crack width of 0.35 mm, a minimum fibre content of 0.9% and 1.1% should be provided for 50 MPa FRC containing high aspect ratio fibres and low aspect ratio fibres, respectively. In addition, the results indicated the importance of fibre content and fibre aspect ratio on the effectiveness of fibre reinforcement.

Self-service may increase customer satisfaction. But by allowing customers to be in close contact with goods before they purchase, exposes the retailer to possible shrinkage in the form of dishonest customers and staff. Suppliers contribute to the shrinkage problem by under-supplying or falsifying invoices. Errors in the capture of inventory information on the part of management and employees are further contributors to possible shrinkage. The combination of these three factors contributes to the phenomenon of 'retail shrinkage' and is directly responsible for eroding the profits of retailers. At present the South African retail market is highly competitive. It is almost saturated and retailers depend on sales volume and lower profit margins to remain in business. To increase profitability, South African retailers are pursuing re-branding and foreign market expansion strategies. This is aggravated by a higher shrinkage rate which in turn increases prices. Robust Shrinkage reduction leads to lower prices, better availability and safer shopping for consumers, greater sales and guaranteed shelf space for manufacturers and reduced losses and increased profitability for retailers. How this can be achieved is the challenging question for retailers. Bearing this in mind, this research is carried in two selected stores, from Woolworths and Checkers. The findings reveal that the above listed shrinkage problems are of great concern to the stores and they have put in place various loss reduction solutions. In this situation the list of solutions is diverse and 'prescribing' a solution is not the main concern. The challenge is how to systematically identify critical and solution seeking causes. From this research it was clear that neither of the two stores has a systematic Shrinkage related data gathering, recording and analysis tools. They are not collaborating with other firms in similar businesses. And worst of all they do not have a strategy to reduce loss because they do not seem to consider loss reduction as a core business activity. Finally, a strategic or policy deployment approach to loss reduction is recommended for the two stores. This includes shrinkage awareness, planning, shrinkage cause analysis, implementation and evaluation. Based on the strategic approach, specific solutions to the specific problems at the two stores are also provided. / Thesis (MBA)-University of KwaZulu-Natal, Durban, 2004.