An experimental study of phase transitions of long-chain n-alkanes induced by the effect of interfaces is described. ¶ The phase behaviour of long-chain n-alkanes (carbon number 14, 16, 17, 18) adsorbed at isolated mica surfaces and confined between two mica surfaces has been studied in the vicinity of and down to several degrees below the bulk melting points, Tm. Using the Surface Force Apparatus we have measured the thickness of alkane films adsorbed from vapour (0.97 [equal to or greater-than] p/p[subscript o] [equal to or greater-than] 0.997), studied capillary condensation transition, subsequent growth of capillary condensates between two surfaces, and phase transitions in both the adsorbed films and the condensates. By measuring the growth rate of the capillary condensates we have identified a transition in the lateral mobility of molecules in the adsorbed films on isolated mica surfaces. This transition to greater mobility occurs slightly above Tm for n-hexadecane, n-heptadecane and n-octadecane but several degrees below Tm for n-tetradecane, and is accompanied by a change in wetting behaviour and a measurable decrease in adsorbed film thickness for n-heptadecane and n-octadecane. Capillary condensates that form below Tm remain liquid, but may freeze if the degree of confinement is reduced by separation of the mica surfaces. An increase in the area of the liquid-vapour interface relative to that of the liquid-mica interface facilitates freezing in the case of the long-chain alkanes, which show surface freezing at the liquid-vapour interface. ¶ Although thermodynamic properties of the surface freezing transition have been rather well documented, the kinetics involved in formation of such ordered monolayers has so far received very little attention. We studied the surface tension of n-octadecane as a function of temperature in the vicinity of Tm, using the static Wilhelmy plate and the dynamic maximum bubble pressure methods. The two methods give different results on cooling paths, where nucleation of the surface ordered phase is involved, but agree on heating paths, where both methods measure properties of the equilibrium surface phase. On cooling paths, the surface of bubbles may supercool below the equilibrium surface freezing temperature. The onset of surface freezing is marked by a sharp drop in the surface tension. The transition is accompanied by an increased stability of the films resulting in longer bubble lifetimes at the liquid surface, which suggests that the mechanical properties of the surfaces change from liquid-like to solid-like. Our results suggest occurrence of supercooling of the monolayer itself.

<p>Aqueous solubility is a key parameter influencing the bioavailability of drugs and drug candidates. In this thesis computational models for the prediction of aqueous drug solubility were explored. High quality experimental solubility data for drugs were generated using a standardised protocol and models were developed using multivariate data analysis tools and calculated molecular descriptors. In addition, structural features associated with either solid-state limited or solvation limited solubility of drugs were identified.</p><p>Solvation, as represented by the octanol-water partition coefficient (log<i>P</i>), was found to be the dominant factor limiting the solubility of drugs, with solid-state properties being the second most important limiting factor.</p><p>The relationship between the chemical structure of drugs and the strength of their crystal lattice was studied for a dataset displaying log<i>P</i>-independent solubility. Large, rigid and flat molecules with an extended ring-structure and a large number of conjugated π-bonds were found to be more likely to have their solubility limited by a strong crystal lattice than were small, spherically shaped molecules with flexible side-chains.</p><p>Finally, the relationship between chemical structure and drug solvation was studied using computer simulated values of the free energy of hydration. Drugs exhibiting poor hydration were found to be large and flexible, to have low polarisability and few hydrogen bond acceptors and donors.</p><p>The relationship between the structural features of drugs and their aqueous solubility discussed in this thesis provide new rules-of-thumb that could guide decision-making in early drug discovery.</p>

Aqueous solubility is a key parameter influencing the bioavailability of drugs and drug candidates. In this thesis computational models for the prediction of aqueous drug solubility were explored. High quality experimental solubility data for drugs were generated using a standardised protocol and models were developed using multivariate data analysis tools and calculated molecular descriptors. In addition, structural features associated with either solid-state limited or solvation limited solubility of drugs were identified. Solvation, as represented by the octanol-water partition coefficient (logP), was found to be the dominant factor limiting the solubility of drugs, with solid-state properties being the second most important limiting factor. The relationship between the chemical structure of drugs and the strength of their crystal lattice was studied for a dataset displaying logP-independent solubility. Large, rigid and flat molecules with an extended ring-structure and a large number of conjugated π-bonds were found to be more likely to have their solubility limited by a strong crystal lattice than were small, spherically shaped molecules with flexible side-chains. Finally, the relationship between chemical structure and drug solvation was studied using computer simulated values of the free energy of hydration. Drugs exhibiting poor hydration were found to be large and flexible, to have low polarisability and few hydrogen bond acceptors and donors. The relationship between the structural features of drugs and their aqueous solubility discussed in this thesis provide new rules-of-thumb that could guide decision-making in early drug discovery.

High Purity Quartz (HPQ; quartz containing less than 50 ppm trace elements) is of increasing economic significance due to its use in certain high-tech components (computer chip/semiconductor manufacture) and in green technologies (silicon wafer production). Current HPQ deposits (hydrothermal veins/leuco-granites/alaskites) are rare and volumetrically small. Unless significant new deposits are found, increasing demand will raise its prices, elevating the strategic nature of this limited commodity. The large volumes and simple mineralogy of pegmatites and the high chemical purity of their constituents make them an attractive target for HPQ. PhD studies are being carried out on quartz from the Evje-Iveland pegmatite field of the Bamble-Evje pegmatite cluster, southern Norway. The area was targeted due to its well constrained geological setting and previously identified potential for HPQ. The aim of the investigation is to develop exploration tools for HPQ by determining the genetic history of the pegmatites and mode of HPQ formation. The study is focussing on 7 pegmatites and their country rocks. Each shows typical pegmatite zonation, with quartz/feldspar intergrowths at the margins, a massive quartz core and a variety of accessory (including REE-bearing) phases. The proximal Høvringsvatnet granite was previously suggested to have supplied late-stage, highly fractionated melts to form the pegmatites. However, from their trace element systematics (no relationship was observed between trace element content and degree of fractionation in each pegmatite body), and a difference in U/Pb age of approximately 70 Ma, the pegmatites cannot be related to the granites. From field evidence (corroborated by geochemical modelling) the pegmatites formed by ‘in situ’ anatexis of country rocks; some locally, some from distal sources. Some pegmatites contain brecciated feldspar and replacement quartz. From LA-ICP-MS analyses, hydrothermal quartz, compared with magmatic quartz, typically contains lower quantities of trace elements. Hydrothermal material shows relatively elevated levels of Al and Li, low Ge and a complete absence of Ti, indicating relatively low temperature hydrothermal formation. Different quartz domains (from SEM-CL imaging) show distinct δ18O values; late stage low trace element zones show values consistent with meteorically derived fluids. In situ LA-ICP-MS studies will provide further information about the characteristics of the fluids which have replaced/refined magmatic quartz to form HPQ. This beneficiation process is a potential mechanism for the generation of economically significant HPQ deposits.

The common manufacturing process of prosthetic sockets is usually a time- and labor consuming activity. This project’s purpose was to look for alternative manufacturing methods that could speed up the process and enhance the experience for the patient for example make some personal design or make the socket lighter. The main goal was to investigate which properties could be achieved by applying Electron Beam Melting as an alternative manufacturing process for prosthetic sockets by applying an earlier developed methodology. An investigation of earlier scientific works with the keywords (additive manufacturing, free form fabrication, orthopedic, prosthetic sockets and rapid manufacturing) was done as well as gathering knowledge how to operate and handle the machines necessary to carry out the project. An updated version of the methodology was developed where the design was verified using finite element analysis. With the updated version the methodology contained nine steps, which in short was as follows. First apprehend an inner socket from an orthopedic clinic with a pattern drawn up on it, the pattern is then transferred to a computer environment and manipulated to the desired shape and thickness. A compressive strength test, both virtual and experimental, was designed by a modified version of the ISO-10328 standard and the virtual design was verified before the socket was manufactured in the Electron Beam Melting machine. The manufactured socket was tested in the experimental set-up to verify the virtual one. The result was a personal designed socket of Ti6Al4V including the male pyramid for connection and a suspension system, which consisted of an inner socket and a one-way valve. It was concluded that Electron Beam Melting could be used as an alternative manufacturing process of prosthetic sockets.

Thesis (MSc)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: Entrainment and assimilation of xenolithic material during kimberlite ascent is considered to be important in shaping the chemistry of the magma and fuelling magma ascent by driving CO2 exsolution. Previous, but as yet unpublished experimental work from Stellenbosch University has demonstrated that orthopyroxene has a key role in this. Orthopyroxene is a very rare xenocrystic constituent of kimberlite but makes up a considerable fraction of the entrained xenolithic material. The initial study used a natural kimberlite composition (ADF1) doped with a peridotite mineral suite (by weight); 88 % ADF1 5% olivine, 5% orthopyroxene and 2% garnet-spinel intergrowth as a starting composition. The subsequent high PT experiments (1100 to 1300°C and 2.0 to 3.5GPa) established that equilibrium orthopyroxene is stable at 1100°C above 2.5GPa, at 1200°C above 2.5GPa and at 1300°C between 2.0 and 3.5GPa. At lower pressures orthopyroxene is completely digested by the experimental melt by the reaction; Mg2Si2O6 (opx) = Mg2SiO4 (ol) + SiO2 (in liquid). In contrast, clinopyroxene is a common phase in kimberlite and often occurs as more than one generation of crystals. Xenocrystic clinopyroxene is dominated by diopside compositions. However, rare omphacite is sometimes also inherited from an eclogite source. The Omphacite, like orthopyroxene, displays textural evidence of severe disequilibrium and may also contribute to the evolution of kimberlitic melt. Thus, a second study produced experiments on the ADF1 kimberlite material at upper mantle PT conditions (1100 to 1300°C and 2.0 to 4.0GPa) as well as an omphacite doped starting material (ADF1+O). These experiments examine the behaviour of pyroxene in kimberlite magma including the influence this may have on magma buoyancy. Within this PT range omphacitic clinopyroxene breaks-down via complex multipart reactions. At 1100°C and 2.0GPa reaction textures around remnant omphacite suggest that omphacite melts incongruently in a complex reaction similar to: Omp + Melt = Ap + Cr-diop + SiO2-enriched Melt. At 1300°C omphacite melts completely and is perceived to produce peritectic Cr-diopside, calcium-rich olivine, carbonate in the melt as well as enrich the melt in SiO2. The melts produced by both the ADF1+O and ADF1 compositions at 1300°C and 4.0GPa are reduced in SiO2 content and have increased TiO2, Cr2O3, Al2O3, MnO, CaO, K2O and P2O5 compared to their respective starting compositions. However, significantly higher proportions of Ca, Na and Fe observed within the ADF1+O melt is a direct consequence of omphacite melting. The ADF1+O starting composition produced equilibrium orthopyroxene above 1100°C and 4.0GPa as well as at 1300°C above 2.0GPa. At lower pressure the orthopyroxene melts incongruently to form peritectic olivine and more silica-rich melt compositions. This digestion favours CO2 exsolution. The effect of orthopyroxene melting can be seen in the melt compositions produced by the peridotite doped starting material (ADF1+P) of the initial study. At 1300°C and 2.0GPa, ADF1+P produced a siliceous melt (37.0 wt.% SiO2) enriched in Al and alkalis compared to the starting ADF1+P composition. This behaviour is directly attributed to xenocrystic orthopyroxene melting at high temperature. In contrast, at the same PT the original kimberlite (ADF1) composition produces a melt with 28.9 wt.% SiO2 and high Ca and Mg contents. Overall, with an increase in pressure the melts become enriched in alkalis and Al2O3 as a direct result of xenocrystic pyroxene melting. In addition, increased pressure allows for a greater solubility of CO2 within the melt. This results in a lower SiO2 melt content and the increased stabilization of equilibrium silica-rich mineral phases (i.e. olivine and equilibrium orthopyroxene). Within the peridotite doped static system (unpublished) the mineral separates with an average crystal size of 115μm ±10μm were all effectively digested in less than 48hours. Similarly, the omphacite doped experiments consumed the 150μm (±10μm) xenocrysts in under 24 hours. Thus, it is suggested that xenocrystic pyroxene is unstable in these experimental kimberlitic melt compositions and is likely to be efficiently assimilated in less than 24 hours. These experimental melts most likely resemble those of natural systems under upper mantle PT conditions. Therefore, pyroxene melting increases the silica content of the melt which in turn drives CO2 exsolution and ascent. / AFRIKAANSE OPSOMMING: Meevoering en assimilasie van xenolitiese materiaal gedurenende kimberliet bestyging is beskou as belangrik in verband met die vorming van die chemie van die magma, en bevorder magma bestyging deur die aandrywing van CO2 ontmenging. Vorige, maar ongepubliseerde eksperimentele werk vanaf Stellenbosch Universiteit het gedemonstreer dat ortopirokseen ‘n sleutelrol hierin het, omrede ortopirokseen ‘n baie skaars xenokristiese bestanddeel van kimberliet is maar ‘n aansienlike fraksie van die meevoerde xenolitiese materiaal moet opmaak. Hierdie studie het ‘n natuurlike primere kimberliet komposisie (ADFI) gedoop met ‘n peridotiet mineraal reeks (per gewig); 88 % ADF1 5% olivien, 5% ortopirokseen en 2% granaat-spinel ingroeiing as begin komposisie gebruik. Die daaropvolgende hoë DT eksperimente (1100 tot 1300°C en 2.0 tot 3.5GPa) het vasgestel dat ewewigsortopirokseen stabiel is teen 1100°C bo 2.5GPa, 1200°C bo 2.5GPa en teen 1300°C vanaf 2.0 tot 3.5Gpa. Teen laer druk word ortopirokseen geheel verteer deur die eksperimentele smelting volgens die reaksie Mg2Si2O6 (opx) = Mg2SiO4 (ol) + SiO2 (in vloeistof). In kontras hiermee is clinopirokseen algemeen in kiemberliet en kom dikwels voor as meer as een generasie se kristalle. Diopsiet komposisies domineer xenokristiese klinopirokseen. Seldsame omfasiet is tog somtyds ook geërf vanaf ‘n eklogiet bron. Die omfasiet, soos ortopirokseen, vertoon teksturuele bewys van ernstige disekwilibrium en mag ook bydra tot die evolusie van kimberlitiese smelt. Dus was daar addisionele eksperimente uitgevoer op die ADF1 kimberliet material teen hoër mantel DT kondisies (1100 tot 1300°C en 2.0 tot 4.0GPa), asook ‘n begin materiaal gedoop met omfasiet (ADF1+O). Hierdie eksperimente ondersoek die gedrag van pirokseen in kiemberliet magma, asook die invloed wat dit sal hê op die dryfvermoë van die magma. Binne hierdie DT reeks breek omfasitiese klinopirokseen af via komplekse multideel reaksie prosesse. Teen 1100°C en 2.0Gpa stel reaksie teksture rondom die oorblywende omfasiet voor dat omfasiet ongelykvormig smelt deur ‘n komplekse reaksie soortgelyk aan: Omp + Smelt = Ap + Cr-diop + SiO2-verrykde Smelt. Teen 1300°C smelt omfasiet volkome en is waargeneem om peritektiese Cr-diopsiet, kalsiumryke olivien en kalsiet te produseer, sowel as dat dit die smelt verryk in SiO2. Die smeltings geprodiseer deur die ADF1+O en ADF1 massa komposisies teen 1300°C en 4.0GPa is verlaag in SiO2 inhoud en bevat verhoogde TiO2, Cr2O3, Al2O3, MnO, CaO, K2O en P2O5 in vergelyking met die onderskeie begin komposisies. Aansienlike hoër proporsies van Ca, Na en Fe is egter waargeneem in die ADF1+O smelt en is ‘n direkte gevolg van die smelting van omfasiet. Die ADF1+O begin samestelling het ewewigsortopirokseen bo 1100°C en 4.0Gpa geproduseer en massa teen 1300°C en 2.0 tot 4.0GPa. Teen laer druk smelt hierdie pirokseen inkongruent om peritektiese olivien en meer silika-ryke smelt samestellings te vorm, en ontmeng CO2. Die effek van ortopirokseen smelting kan aanskou word in die smelt samestellings wat produseer is deur die begin materiaal wat gedoop is in peridotiet (ADF1+P), in die oorspronklike studie. Teen 1300°C en 2.0GPa het ADF1+P ‘n silikahoudende smelt (37.0 wt.% SiO2) produseer wat verryk is in Al en alkalies in vergelyking met die ADF1+P massa samestelling. Hierdie gedrag is direk toegeskryf aan die xenokristiese ortopirokseen wat smelt teen hoë temperatuur. In kontras hiermee, teen dieselfde DT kondisies produseer die oorspronklike kiemberliet (ADF1) massa ‘n smelt met 28.86 gewigspersentasie SiO2 en hoë Ca en Mg inhoud. In die algeheel word die smeltings verryk in alkalies en Al2O3 teen verhoogde druk as ‘n derekte gevolg van xenokristiese pirokseen smelting. Verder laat verhoogde druk toe vir hoër oplosbaarheid van CO2 in die smelt, wat lei tot laer SiO2 inhoud en ‘n toename in stabilisering van ewewigs silika-ryke mineraal fases (dws. olivien en ewewigsortopirokseen). In die peridotiet gedoopde statiese sisteem (ongepubliseerd), was die mineraal skeiding met ‘n gemiddelde kristal grootte van 115μm ±10μm almal effektief verteer in minder as 48 ure. Soortgelyk hieraan het die omfasiet gedoopde eksperimente die 150μm (±10μm) sade onder 24 ure verteer. Dus stel dit voor dat xenokristiese pirokseen in naatuurlike sisteme onstabiel is in kiemberlietiese smelt samestellings en sal waarskynlik geassimileer wees in miner as 24 ure en ‘n meer silica-ryke kiemberlietiese smelt samestelling produseer terwyl dit CO2, ontmenging en bestyging aandryf.

Research focused on the selective laser melting (SLM) of stainless steels and aluminium alloys. For steels, the possibility of creating a magnetically graded material was demonstrated as well as the ability to improve consolidation with austenitic and martensitic stainless steel powder mixtures. Stainless Steel/CoCr hybrid samples were also manufactured and tested to investigate the advantages of functionally graded materials (FGMs). Al alloy research began with examining the requirements for successful Al alloy consolidation in SLM and through experimentation it was found that Al alloys with good welding properties were the best choice: pure Al was found to be completely unsuitable. 6061 Al alloy was then used as a base material to manufacture Al-Cu alloy samples. Single layer SLM samples were produced first, which resulted in recognised Al-Cu microstructures forming. Multilayer Al alloy SLM research resulted in the discovery of the theorised ability to manufacture Al-Cu alloy parts with a nanocrystalline Al matrix with dispersed Al2Cu quasicrystals, resulting in a material comparable to a metal matrix composite that showed excellent corrosion resistance and compressive strength. Finally, a demonstration part was made to test the capability of the SLM process producing an aerospace type geometry using a customised Al alloy. Observations during manufacture and post process analysis showed that Al alloys were susceptible to changes in mechanical properties due to the geometry of the manufactured part.

Israel Zangwill, novelist, playwright, poet, and essayist, can be understood and appreciated best as an apologist whose chosen mission was to introduce the Jew to the English-speaking reader, a reader who had often see the word Jew on the pages of his literature but seldom had been able to meed an authentic specimen of the group in--or out--of print. This thesis will describe the works of Zangwill from an apologetic standpoint.

The ice melting and lateral cavity formation caused by in situ burning (ISB) of liquid fuels in ice-infested waters was studied in order to improve predictions on the removal efficiency of this oil spill mitigation method. For this purpose, several experimental studies were conducted to increase the fundamental understanding of the mechanisms that lead to ice melting and lateral cavity formation. The findings of the experimental studies provided the required knowledge to mathematically formulate the ice melting problem. Mathematical scaling analysis of ice melting during burning of oils in the vicinity of ice was performed to create a tool to estimate the extent of melting that occurs during ISB in ice-infested waters. A series of lab-scale experiments were designed to systematically investigate the ice melting problem. The first set of experiments were conducted in cylindrical shaped ice cavities with a 5.7 cm diameter. Burning of n-octane from ignition to natural extinction and the subsequent geometry change of the ice, fuel thickness, and fuel temperature were measured. The preliminary experimental observations showed that the melting of the ice walls was higher in areas where the fuel layer was in contact with ice compared with places of flame exposure. Based on these observations, a hypothesis that suggested the convective flows in the liquid fuel (driven mainly by surface tension and buoyancy) were contributing in melting of the ice was proposed to explain the origins of the lateral cavity. To evaluate this hypothesis, two dimensionless numbers (Marangoni and Rayleigh) were calculated as the indicators of the mechanisms of convection in the fuel layer. The comparison between the melting speed and these dimensionless numbers indicated surface tension driven flow was dominant while the role of buoyancy was negligible. In another set of experiments, Particle Image Velocimetry (PIV) was used to study the flow structure within the liquid-phase of n-octane pool fire bound on one side by an ice wall. Experiments were conducted in a square glass tray (9.6 cm × 9.6 cm × 5 cm) with a 3 cm thick ice wall placed on one side of the tray. Burning rate, flame height, and melting front velocity were measured to analyze the effect of heat feedback on melting of the ice. The melting rate of the ice increased from 0.6 cm/min for the first 50 seconds after ignition to 1 cm/min for the rest of burning period. Meanwhile, the measurement of the burning rates and flame heights showed two distinctive behaviors; a growth period from self-sustained ignition to the peak mass loss rate (first 50 seconds after ignition) followed by a steady phase from the peak of mass loss rate until the manual extinguishment. Similarly, the flow field measurements by a 2-dimensional PIV system indicated the existence of two different flow regimes. In the moments before ignition of the fuel, coupling of surface tension and buoyancy forces led to a combined one roll structure in the fuel. This was when a single large vortex was observed in the flow field. After ignition the flow field began transitioning toward an unstable flow regime (separated) with an increase in number of vortices around the ice wall. As the burning rate/flame height increased the velocity and evolving flow patterns enhanced the melting rate of the ice wall. Experimentally determined temperature contours showed that a hot zone with thickness of approximately 3 mm was present below the free surface, corresponding to the multi-roll location. The change in the flow field behavior was found to relate to the melting front velocity of ice. To further study the lateral cavity phenomena, a parametric experimental study on melting of ice adjacent to liquids exposed from above to various heat fluxes was conducted in order to understand the role of liquid properties in formation of cavities in ice. Multiple liquids with wide variety and range of thermophysical properties were used in order to identify the key influential properties on melting. The melting rate of the ice and penetration speed of the liquid in a transparent glass tray (70 mm × 70 mm × 45 mm) with a 20 mm thick ice wall (70 mm × 50 mm × 20 mm) was measured. The melting front velocities obtained from experiments were then compared to surface flow velocities of liquids obtained through a scaling analysis of the surface flow to elucidate the influence of the various thermophysical properties of the liquids on ice melting. The surface velocity of the liquids correlated well to the melting front velocities of the ice which showed a clear relationship between the flow velocity and melting front velocity. As the final step of this work, to extend the findings of the experimental studies conducted herein to larger sizes comparable to realistic situations in the Arctic, an order of magnitude scaling analysis was performed to obtain the extent of ice melting. The scaling considered the heat feedback from the flame to fuel surface, the convective heat transfers toward the ice, and the melting energy continuity of ice. The existing experimental data on the size of lateral cavity were also collected and were correlated to the results of the scaling analysis using a nonlinear regression fitting technique. The mathematical correlation that was obtained by the scaling analysis can be used to predict the size of the lateral cavity for a given fuel, pool fire diameter, and burning time. This correlation will provide a predictive tool to estimate the size of a potential lateral cavity formed during ISB of a given spill scenario. In general, the ability to predict the ice melting caused by burning of spilled oil in ice-infested waters is of great practical importance for assessment of the response outcome. This would assist with quantifying the geometry change of the burning medium which in turn will define oil burning rate and extinction condition. Knowledge of burning behavior and extinction condition indicate the burned volume which can directly be used to define the removal effectiveness of ISB. Nevertheless, this analysis was conducted on a generic interaction of oil and ice and the specific details that are observed in actual application of ISB in ice-infested waters were neglected for simplicity. Extending the outcome of this study to more specific (scenario-based) oil-in-ice situation and improving the predictability of the melting correlation with large-scale experiments are the next steps to develop this work.

Meltwater accumulated on the Greenland Ice Sheet drains to glacier beds, discharging into fjords hundreds of meters below sea level. The injection of meltwater at depth generates an upwelling plume that entrains warm ocean water as it rises along the terminus, increasing submarine melt and driving a fjord-scale exchange flow. However, due to sparse ocean-glacier observations, we lack a process understanding of how plumes control fjord circulation and submarine melt. Combining numerical modeling, theory, and observations, this dissertation investigates near-glacier plume dynamics, the influence of glacier depth on plume structure and submarine melt, and the role of fjord-glacier geometry on circulation in tidewater glacier fjords. In Chapter II, I use buoyant plume theory and a nonhydrostatic, three-dimensional ocean–ice model to investigate the sensitivity of plume dynamics to subglacial discharge, turbulent diffusivity, and conduit geometry. Large discharges result in plumes with positive temperature and salinity anomalies in the upper water column. Fjord circulation is sensitive to conduit geometry; distributed subglacial discharge results in a stronger return flow of warm water toward the terminus. In Chapter III, I use buoyant plume theory, initialized with realistic ranges of subglacial discharge, glacier depth, and ocean stratification, to investigate how plume structure and submarine melt vary during summer months in 12 Greenland fjords. Grounding line depth is a primary control on plume-induced submarine melt: deep glaciers produce warm, salty subsurface plumes that undercut termini, and shallow glaciers produce cold, fresh surface-confined plumes that can overcut. Finally, in Chapter IV, I use regional-scale numerical ocean simulations to systematically evaluate how fjord circulation forced by subglacial plumes, tides, and wind stress depends on fjord width, glacier depth, and sill height. Glaciers grounded below sill depth can draw shelf waters over a shallow sill and into fjord basins with seasonal subglacial discharge; this process is independent of external shelf forcing. These results underscore the first-order effect that subglacial discharge and fjord-glacier geometry have in controlling fjord circulation and, thus, ocean heat flux to the ice. This dissertation includes previously published and co-authored material.