I present physical shear-box experiments investigating the relationship between geometrical properties of shear zones and mechanical properties of deformed rocks. Experimental methodology is also examined critically and new materials for analogue modelling of shear localization are presented. First, I tested experimentally whether meaningful rheological information can be deduced from finite geometrical shear zone data. The results predict characteristic geometrical responses for certain end-member materials. However, it will be difficult to constrain such responses in the field. In the second part physical controls on deformation in the shear box are analysed for Newtonian and power-law fluids and an elastoviscoplastic strain-softening material. Since models always represent simplifications of the natural problem, it is essential to understand fully the physics of a given simulation. I show that displacement boundary conditions, model geometry, and rheology control shear zone geometry. Practical applications of the shear box for modelling natural shear localization and limitations of isothermal physical models with displacement boundary conditions in general are discussed. In the third part, new data on the rheology of highly-filled silicone polymers are introduced. Since dynamic similarity must be satisfied in analogue models to permit scaled, quantitative simulations of deformation processes, the choice of suitable rock analogues is critical for physical experiments. In particular, we address the problem of designing power-law fluids to model rocks deforming by dislocation creep. We found that highly-filled polymers have complex rheologies. Hence, such materials must be used with care in analogue modelling and only for certain experimental stress-strain rate conditions. Finally, I investigated whether fault network geometry and topography of brittle strike-slip faults are influenced by the degree of compaction of the host rock. Analogue shear experiments with loose and dense sand imply that the degree of sediment compaction may be a governing factor in the evolution of fault network structure and topography along strike-slip faults in sedimentary basins. Therefore, models of strike-slip faults should consider potential volume changes of deformed rocks.

Certain strains of hydrophobic bacteria are known to play critical roles in petroleum-related applications. The aim of this study was to investigate how hydrophobic bacteria in their stationary phase could adsorb onto the hexadecane-water interface and alter its mechanical properties. The two strains of bacteria used in forming the interfacial films were Acinetobacter venetianus RAG-1 (a Gram-negative bacterium) and Rhodococcus erythropolis 20S-E1-c (Gram-positive). Experiments at two different length scales (millimetre and micrometre) were conducted and the results were compared. In addition, a simple flow experiment was designed in a constricted channel and the results were related to the intrinsic mechanical properties of bacteria-adsorbed films. On the millimetre scale, using the pendant drop technique, the film interfacial tension was monitored as the surface area was made to undergo changes. Under static conditions, both types of bacteria showed no significant effect on the interfacial tension. When subjected to transient excitations, the two bacterial films exhibited qualitatively similar, yet quantitative distinct rheological properties (including film elasticities and relaxation times). Under continuous reduction of surface area, the RAG-1 system showed a “paper-like” interface, while the interface of the 20S-E1-c system was “soap film-like.” These macroscopic observations could be explained by the surface ultrastructures of the two cell strains. On the micrometre scale, using the micropipette technique, colloidal stability of the bacteria-coated oil droplets was examined through direct-contact experiments. Both types of bacteria were seen to function as effective stabilizers. In addition, the adsorbed bacteria also interacted with one another at the interface, giving rise to higher order 2-D rheological properties. A technique of directly probing the mechanical properties of the emulsion drop surfaces revealed that (a) the films behaved as purely elastic sheets, and (b) with a reduction in cell concentration in the aqueous phase, less oil was emulsified, but the elastic moduli of the adsorbed films remained unchanged. These results are in contrast to the above millimetre-scale study. Therefore the rheological properties of these bacteria-adsorbed films appear to be length scale-dependent. An oil displacement experiment was designed to investigate the flow behaviour of micron-scale emulsion drops in a constricted channel. The qualitative results can be correlated with the interfacial rheological properties and may have valuable relevance to the study of multiphase flow through constricted channels in porous rocks (e.g. in MEOR operations). / Chemical Engineering

The fundamentals of the photopolymerization process are not well understood. As a result, issues affecting the cure speed and overall quality of the final product (shape, size, and surface finish) of photopolymerization impose significant limitations on applications that require fast processing and high spatial resolution. To address this issue, microrheology was employed to perform in-situ monitoring of the liquid-to-gel transition during free-radical photopolymerization. Photosensitive acrylate and hydrogel resins were exposed to ultraviolet light, while the Brownian motion of micrometer sized, inert fluorescent tracer particles was tracked via optical videomicroscopy. Statistical analysis of particle motion yielded the rheological properties of the embedding medium as a function of time and location, thereby relating UV exposure to the progress of polymerization and gelation. The microrheological setup enabled a detailed study of three-dimensional gelation profiles; other experimental parameters that were initially varied include photoinitiator concentration, monomer composition, and light intensity. Significant changes in gelation time were observed with varying UV intensity and UV penetration depth into the sample. In addition, oxygen inhibition was found to significantly impact the cure speed of monomeric resins. The preliminary results were used to test the accuracy of the energy threshold model, which is often used to empirically predict the outcome of photopolymerization reactions. By using lithographic masks to generate well-defined UV illumination patterns with characteristic dimensions of tens of micrometers, it could be shown unambiguously that the diffusion of oxygen, an inhibitor, plays a critical role in the polymerization reaction. The experiments are in excellent agreement with a simple two-step model of oxygen consumption followed by polymerization. The use of high-speed electronic shutters in the UV light path enabled us to control the illumination time of the samples with high precision. Microrheological analysis could be used to reconstruct three-dimensional profiles of partially polymerized samples. Traditional photorheometry is not capable of resolving the evolution of sample rheology with such spatial resolution. In addition, experiments with pulsed illumination were used to quantify the role of dark reactions due to residual free radicals after termination of UV illumination.

Delivery of a desired ‘active’ compound (for example, starch (as an energy substrate)) to the gastrointestinal (GI) tract is most easily achieved by oral administration. Unfortunately, the efficacy of most actives is greatly reduced due to the aggressive nature of digestive enzymes and processes which occur in this environment. A commonly applied strategy to prevent deactivation of the active prior to absorption at the target site is to encapsulate the active in another ‘sacrificial’ or non-degradable polymer matrix. Traditionally, the active and matrix is processed into a microparticle format for easy oral delivery (dispersed in a liquid or paste). However, established encapsulation methods which rely on bulk-phase processing to produce these microparticles (e.g. emulsification) are far from ideal as they lack control over the final microparticle size, size distribution, composition and shape. The lack of control in the physical properties of the resultant microparticles in turn results in an inherent lack of control over the kinetics of release of the active at the target site. In contrast, recent advances in microfluidic device fabrication and methodology development have firmly proven that these new generation devices can produce monodisperse droplets and microparticles in a continuous, controllable and predictable manner. Their potential as a processing tool for the production of highly tailored microparticles for targeted delivery, however, remains to be fully explored. Both the physical and chemical (physicochemical) properties of microparticles made from a single polymer system may be altered by the deposition of one or more additional polymer layers onto the microparticle surface (for example, alternating layers of oppositely charged polyelectrolytes to produce core-shell like particles), and this method has proven to be favorable with regards to retarding the release of active compounds. However, this addition of alternate layers of oppositely charged polyelectrolytes (so called Layer-by-Layer (LbL) deposition or assembly) does increase the number of processing steps the particles must undergo prior to storage or delivery. Further, the overall effectiveness of this additional processing is still highly dependent on the properties of the original (core) microparticles. In this thesis, a microfluidic technique was developed to encapsulate starch granules in alginate-based microparticles. Using this continuous technique, the size of the microparticles produced were shown to be monodisperse and reproducible. The developed microfluidic device included a drop formation section, followed by a gelation region and a transfer section, where the particles made on-chip are transferred from the carrier oil phase to an aqueous phase prior to collection. The microparticles collected from this microfluidic device were found to be stable for several weeks and in stark contrast to particles produced via a standard bulk emulsification routes, no aggregation was observed over this time frame. The release profile of glucose (as a result of starch hydrolysation) from microparticles produced using both a standard bulk emulsification method and the developed microfluidic-based method were compared. It was found that the monodisperse particles produced using the microfluidic method showed significantly more retardation to release compared to the glucose release profile from bulk-processed particles. This retardation effect was more pronounced when a thin layer of an oppositely charged polyelectrolyte (chitosan) was adsorbed onto the negatively charged surface (alginate is an anionic polyelectrolyte) of the microfluidic-processed microparticle. The microfluidic device developed within this thesis and the resulting tailored microparticles thus show significant potential with regards to offering a new generation of microparticle delivery systems with highly deterministic delivery over extended lifetimes.

Delivery of a desired ‘active’ compound (for example, starch (as an energy substrate)) to the gastrointestinal (GI) tract is most easily achieved by oral administration. Unfortunately, the efficacy of most actives is greatly reduced due to the aggressive nature of digestive enzymes and processes which occur in this environment. A commonly applied strategy to prevent deactivation of the active prior to absorption at the target site is to encapsulate the active in another ‘sacrificial’ or non-degradable polymer matrix. Traditionally, the active and matrix is processed into a microparticle format for easy oral delivery (dispersed in a liquid or paste). However, established encapsulation methods which rely on bulk-phase processing to produce these microparticles (e.g. emulsification) are far from ideal as they lack control over the final microparticle size, size distribution, composition and shape. The lack of control in the physical properties of the resultant microparticles in turn results in an inherent lack of control over the kinetics of release of the active at the target site. In contrast, recent advances in microfluidic device fabrication and methodology development have firmly proven that these new generation devices can produce monodisperse droplets and microparticles in a continuous, controllable and predictable manner. Their potential as a processing tool for the production of highly tailored microparticles for targeted delivery, however, remains to be fully explored. Both the physical and chemical (physicochemical) properties of microparticles made from a single polymer system may be altered by the deposition of one or more additional polymer layers onto the microparticle surface (for example, alternating layers of oppositely charged polyelectrolytes to produce core-shell like particles), and this method has proven to be favorable with regards to retarding the release of active compounds. However, this addition of alternate layers of oppositely charged polyelectrolytes (so called Layer-by-Layer (LbL) deposition or assembly) does increase the number of processing steps the particles must undergo prior to storage or delivery. Further, the overall effectiveness of this additional processing is still highly dependent on the properties of the original (core) microparticles. In this thesis, a microfluidic technique was developed to encapsulate starch granules in alginate-based microparticles. Using this continuous technique, the size of the microparticles produced were shown to be monodisperse and reproducible. The developed microfluidic device included a drop formation section, followed by a gelation region and a transfer section, where the particles made on-chip are transferred from the carrier oil phase to an aqueous phase prior to collection. The microparticles collected from this microfluidic device were found to be stable for several weeks and in stark contrast to particles produced via a standard bulk emulsification routes, no aggregation was observed over this time frame. The release profile of glucose (as a result of starch hydrolysation) from microparticles produced using both a standard bulk emulsification method and the developed microfluidic-based method were compared. It was found that the monodisperse particles produced using the microfluidic method showed significantly more retardation to release compared to the glucose release profile from bulk-processed particles. This retardation effect was more pronounced when a thin layer of an oppositely charged polyelectrolyte (chitosan) was adsorbed onto the negatively charged surface (alginate is an anionic polyelectrolyte) of the microfluidic-processed microparticle. The microfluidic device developed within this thesis and the resulting tailored microparticles thus show significant potential with regards to offering a new generation of microparticle delivery systems with highly deterministic delivery over extended lifetimes.

Delivery of a desired ‘active’ compound (for example, starch (as an energy substrate)) to the gastrointestinal (GI) tract is most easily achieved by oral administration. Unfortunately, the efficacy of most actives is greatly reduced due to the aggressive nature of digestive enzymes and processes which occur in this environment. A commonly applied strategy to prevent deactivation of the active prior to absorption at the target site is to encapsulate the active in another ‘sacrificial’ or non-degradable polymer matrix. Traditionally, the active and matrix is processed into a microparticle format for easy oral delivery (dispersed in a liquid or paste). However, established encapsulation methods which rely on bulk-phase processing to produce these microparticles (e.g. emulsification) are far from ideal as they lack control over the final microparticle size, size distribution, composition and shape. The lack of control in the physical properties of the resultant microparticles in turn results in an inherent lack of control over the kinetics of release of the active at the target site. In contrast, recent advances in microfluidic device fabrication and methodology development have firmly proven that these new generation devices can produce monodisperse droplets and microparticles in a continuous, controllable and predictable manner. Their potential as a processing tool for the production of highly tailored microparticles for targeted delivery, however, remains to be fully explored. Both the physical and chemical (physicochemical) properties of microparticles made from a single polymer system may be altered by the deposition of one or more additional polymer layers onto the microparticle surface (for example, alternating layers of oppositely charged polyelectrolytes to produce core-shell like particles), and this method has proven to be favorable with regards to retarding the release of active compounds. However, this addition of alternate layers of oppositely charged polyelectrolytes (so called Layer-by-Layer (LbL) deposition or assembly) does increase the number of processing steps the particles must undergo prior to storage or delivery. Further, the overall effectiveness of this additional processing is still highly dependent on the properties of the original (core) microparticles. In this thesis, a microfluidic technique was developed to encapsulate starch granules in alginate-based microparticles. Using this continuous technique, the size of the microparticles produced were shown to be monodisperse and reproducible. The developed microfluidic device included a drop formation section, followed by a gelation region and a transfer section, where the particles made on-chip are transferred from the carrier oil phase to an aqueous phase prior to collection. The microparticles collected from this microfluidic device were found to be stable for several weeks and in stark contrast to particles produced via a standard bulk emulsification routes, no aggregation was observed over this time frame. The release profile of glucose (as a result of starch hydrolysation) from microparticles produced using both a standard bulk emulsification method and the developed microfluidic-based method were compared. It was found that the monodisperse particles produced using the microfluidic method showed significantly more retardation to release compared to the glucose release profile from bulk-processed particles. This retardation effect was more pronounced when a thin layer of an oppositely charged polyelectrolyte (chitosan) was adsorbed onto the negatively charged surface (alginate is an anionic polyelectrolyte) of the microfluidic-processed microparticle. The microfluidic device developed within this thesis and the resulting tailored microparticles thus show significant potential with regards to offering a new generation of microparticle delivery systems with highly deterministic delivery over extended lifetimes.

Thesis (Ph. D.)--University of Sydney, 2006. / Title from title screen (viewed 27 February 2007). Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy to the School of Aerospace, Mechanical and Mechatronic Engineering. Includes bibliographical references. Also issued in print.

Diss. (sammanfattning) Uppsala : Univ., 2001. / Härtill 6 uppsatser.

Dissertation (Ph. D.)--University of Akron, Dept. of Polymer Engineering, 2007. / "May, 2007." Title from electronic dissertation title page (viewed 04/07/2008) Advisor, Erol Sancaktar; Committee members, Avraam I. Isayev, Sadhan C. Jana, Darrell H. Reneker, Shing-Chung Wong; Department Chair, Sadhan C. Jana; Dean of the College, George R. Newkome; Dean of the Graduate School, George R. Newkome. Includes bibliographical references.

Thesis (M.Sc. (Tech.))--University of Waikato, 2006. / Title from PDF cover (viewed March 5, 2008). Includes bibliographical references (p. 107-110)