Thesis (MSc)--Stellenbosch University, 2011. / ENGLISH ABSTRACT: Needle-electrospinning is an uncomplicated and highly versatile nanofiber (fiber diameter of 50 to 500 nm) production technique. Nevertheless the process can only produce 0.01 to 1.0 g/h/m2 of nanofibers, unpractical for large-scale productions. Bubble-electrospinning, in the presence of surfactants, is a novel nanofiber mass-production technique developed at Stellenbosch University.[1] The technique is similar to needle-electrospinning only that the surface area of a bubble surpasses that of a solution droplet, making it possible for multiple jets to form on the bubble surface at high field strengths. Thus far little research has been done on the influence of solution properties on the bubble-electrospinning technique. During electrospinning the solution experiences three competing forces, namely, surface tension (contracting force), charge repulsion (expanding force), and viscosity (resistance to flow). The first aim of this study was to obtain better understanding on the influence of three significant solution properties (viscosity, conductivity and surface tension) on bubble-electrospinning in terms of bubble lifetime, bubble size, average number of jets and the resultant fibers. The solution properties were varied using a range of polymer and surfactant concentrations. A second aim was to obtain better understanding on the comparison of the bubble-electrospinning process between two polymer solutions, namely Polyvinyl alcohol (PVOH) solutions containing sodium lauryl ether sulphate (SLES) surfactant, and Polyacrylonitrile (PAN) solutions containing silicone surfactant. Results indicated that the solution viscosity and conductivity increased with increasing polymer concentrations for both polymer solutions. In addition, both the solution surface tensions were not influenced by polymer concentration. With regards to bubble-electrospinning of PVOH solutions, results indicated that the average number of jets per bubble was influenced by the polymer concentration. Regarding PAN solutions, bubble lifetime and the average number of jets was influenced by polymer concentration. Results indicated that the solution viscosity increased and surface tension decreased with increasing surfactant concentration for both polymer solutions. PVOH solution conductivity increased whilst PAN solution conductivity decreased with increasing surfactant concentrations. With regards to bubble-electrospinning of PVOH solutions, the bubble lifetime and bubble size was significantly influenced by the SLES concentration. Regarding PAN solutions, the silicone surfactant concentration had no significant effect on the bubble-electrospinning process. Overall, PVOH fiber diameters decreased with increasing surfactant concentration. There was no common trend between the bubble-electrospinning of PVOH and PAN solutions in relation to their solution properties. It was concluded that solution viscosity, conductivity and surface tension are not the only significant contributing parameters to the bubble-electrospinning process. / AFRIKAANSE OPSOMMING: Die naald-elektrospinproses is 'n eenvoudige, hoogsaanpasbare tegniek wat gebruik word vir die maak van nanovesels. Nanovesels het tipies 'n deursnee van 50nm tot 500nm. Ongelukkig is dit onprakties vir grootskaalse produksie omdat die uitset daarvan beperk is tot 0.01 tot 1.0 g/h/m2. Die borrel-elektrospinproses, waar elektrospinstrale gespin word vanaf die oppervlak van borrels op die oppervlak van die spinoplossing en waar die borrels gestabiliseer is m.b.v. sepe, is 'n nuwe tegniek wat ontwikkel is by die Universiteit van Stellenbosch. [1]. Die tegniek is soortgelyk aan die naald-elektrospinproses in dié sin dat die elektrospinstraal vorm vanaf 'n gelaaide halfsfeervormige oppervlak in die spinoplossing, maar die aansienlik groter oppervlakarea van die borrel in die borrel-elektrospinproses maak dit moontlik om verskeie elektrospinstrale gelyktydig op die oppervlak van die borrel te onderhou. Dit lei tot baie hoër doeltreffendheid in die saamgroeppering van die strale en gevolglik tot hoër nanoveseluitsette. Tot dusver is daar weinig navorsing aangaande die invloed van oplossingseienskappe op die borrel- elektrospintegniek gedoen. Tydens die elektrospinproses ervaar die oplossing drie kompeterende kragte, naamlik: oppervlakspanning (sametrekkende krag), elektrostatiese afstoting (afstotende krag) en viskositeit (vertragende effek op vloei van die oplossing). Die hoofdoelwit van hierdie navorsing was om 'n beter begrip te kry van die invloed van drie gemete oplossingswaardes, d.w.s. viskositeit, elektriese geleidingsvermoë en oppervlakspanning op die borrel-elektrospinproses. Die impak van hierde waardes is spesifiek geëvalueer in terme van borrellewensduur, borrelgrootte, gemiddelde hoeveelheid elektrospinstrale per borrel en die morfologie van die vesels wat in die proses gevorm is. Die tweede doelwit van die studie was om 'n vergelyking te tref tussen die mees optimale oplossingswaardes in die borrel-elektrospinproses van twee baie uiteenlopende polimeerspinoplossings, naamlik polivinielalkohol (PVOH), met natrium dodesieletersulfaat (SLES) as die borrelstabiliserende seep en poliakrilonitriel (PAN) oplossing, met 'n silikoonseep as die borrelstabiliserende seep. Resultate het getoon dat die viskositeit en elektriese geleidingsvermoë toeneem met toename in polimeerkonsentrasie vir beide PVOH- en PAN-oplossings. Verder is oppervlakspanning in beide gevalle nie beduidend beïnvloed deur die polimeerkonsentrasie nie. In die geval van die borrel-elektrospin van die PVOH-oplossings het resultate daarop gedui dat die gemiddelde aantal elektrospinstrale per borrel moontlik beïnvloed kon word deur die polimeerkonsentrasie. In die geval van borrel-elektrospin van PAN-oplossing is bevind dat polimeerkonsentrasie die borrelleeftyd en die gemiddelde aantal elektrospinstrale per borrel beïnvloed. Resultate het ook getoon dat die viskositeit vermeerder en die oppervlakspanning afneem met toename in die konsentrasie van die sepe in beide die polimeeroplossings. Die PVOH-oplossing se elektriese geleidingsvermoë het vermeerder terwyl dit verminder in die geval van die PAN-oplossings met 'n toename in die seepkonsentrasie. Tydens borrel-elektrospin van die PVOH-oplossings is beide borrelleeftyd en borrelgrootte beduidend beïnvloed deur die SLES konsentrasie. By die borrel-elektrospin van PAN-oplossings het die silikoonseepkonsentrasie nie 'n beduidende invloed gehad op die borrelleeftyd en borrelgrootte nie. Oor die algemeen het die gemiddelde PVOH veseldeursnee afgeneem met toename in seepkonsentrasie. Geen algemene tendens kon waargeneem word tussen die optimale oplossingswaardes vir borrel-elektrospin van die PVOH- en die PAN-oplossings onderskeidelik nie. Die gevolgtrekking is dat die viskositeit, elektriese geleidingsvermoë en oppervlakspanning nie die enigste beduidende waardes is wat bepaal of die borrel-elektrospinproses sal werk vir 'n spesifieke polimeeroplossing nie.

The gas assisted injection moulding process is an important extension to conventional injection moulding. Gas assist can be applied in a number of ways, but here the penetration of a gas bubble through the polymer melt is of interest. A 3D fi nite element implementation of a pseudo concentration method is employed to simulate the primary penetration of the gas bubble. The wall thickness prediction is an important result since the extent of bubble penetration is sensitive to the remaining melt fraction. A number of methods for experimental measurement are developed to measure characteristics of the gas assisted injection moulding process dynamics and product. Key process variables, on an industrial gas-assist machine, were measured and analysed, leading to an empirical model for wall thickness prediction. Gas delay time and injection velocity are shown to be most influential in controlling residual wall thickness. Simulation results are evaluated against the empirical model. The trends observed, for simulation and experiment, in wall thickness after changes in process variable settings are found to agree qualitatively. The wall thickness prediction is found to be within 10% of the experimentally obtained measurements.

Flow boiling in microchannels, while very promising as a cooling technology in electronics thermal management, is still a subject being explored that requires further investigation. Before applying this technology for high heat flux computer chip cooling, challenging issues such as fully understanding boiling mechanisms in confined spaces, extending and stabilising the nucleate boiling regime, suppressing flow boiling instabilities, maintaining uniform flow distribution among microchannels, have to be addressed. If flow boiling is to be used as a thermal management method for high heat flux electronics it is necessary to understand the behaviour of a non-uniform heat distribution, which is typically the case observed in a real operating computer chip. In this study, flow boiling of deionised water in a silicon microchannel heat sink under uniform and non-uniform heating has been investigated with particular attention to flow boiling instabilities. An experimental system was designed and constructed to carry out the experimental investigations. The experimental heat sink consisting of forty parallel rectangular microchannels with 194 μm hydraulic diameter together with integrated inlet and outlet manifold was fabricated on a silicon wafer using inductive coupled plasma dry etching, in conjunction with photolithographic techniques. A design with integrated temperature sensors made from a thin nickel film allows local temperature measurements with a much faster response time and smaller thermal resistance as compared to temperature measurements using thermocouples. The integrated heater was designed to enable either uniform or non-uniform heating (hotspot investigation) with a low thermal resistance between the heater and the channels. Numerical simulations for single phase flow in adiabatic conditions were used to assist the design of the manifold geometry in the microchannels heat sink. Microfabricated temperature sensors were used together with simultaneous high speed imaging in order to obtain a better insight related to temperature fluctuations caused by two-phase flow instabilities under uniform and non-uniform heating. Two types of two-phase instabilities with flow reversal were identified and classified into flow stability maps. The effect of inlet water temperature on flow boiling instabilities was experimentally studied, with the influence of different subcooling conditions on the magnitude of temperatures as well as the influence on temperature uniformity over the heat sink being assessed. The effect of various hotspot locations on flow boiling instabilities has been investigated, with hotspots located in different positions along the heat sink. Bubble growth and departure size have been experimentally investigated. The results of this study demonstrate that bubble growth in microchannels is different from that in macroscale channels. Furthermore, the effects of bubble dynamics on flow instabilities and heat transfer coefficient have been investigated and discussed.

Approved for public release, distribution is unlimited / When a burst of air is produced in water, the result can be a toroidal bubble. This thesis is concerned with experimental investigations of three acoustical properties of toroidal bubbles: (i) propagation through high-intensity noise, (ii) emission, and (iii) scattering. In (i), an attempt to observe a recent prediction of the acoustic drag on a bubble is described, which is analogous to the Einstein-Hopf effect for an oscillating electric dipole in a fluctuating electromagnetic field. No effect was observed, which may be due to insufficient amplitude of the noise. In (ii), observations of acoustic emissions of volume oscillations of toroidal bubbles are reported. Surprisingly, the emission occurs primarily during the formation of a bubble, and is weak in the case of very smooth toroidal bubbles. In (iii), we describe an experiment to observe the effect of a toroidal bubble on an incident sound field. In addition to the acoustical investigations, we describe the construction of a large hallway apparatus for further investigations and for hands-on use by the public. The tank has cross section 2 feet by 2 feet and height 6 feet, and the parameters of reservoir pressure and time between air bursts are adjustable by the observer. / Lieutenant, United States Navy

Si le phénomène bullaire reste le primum movens à l’origine de l’obstruction vasculaire lors de la décompression, les plaquettes sanguines jouent un rôle déterminant dans la physiopathologie de l’accident de décompression (ADD). L’objectif de ces travaux était d’étudier les mécanismes à l’origine de cette agrégation plaquettaire induit par le phénomène bullaire. La première étape a permis de valider un modèle animal d’ADD chez le rat et de confirmer que le degré d’agrégation constituait un indice fiable de sévérité de l’ADD. La seconde a eu pour but d’étudier différents marqueurs spécifiques de l’activation plaquettaire. La troisième étape a étudié les effets de différents anti-agrégants plaquettaires administrés avant l’exposition. Les résultats confirment que le clopidogrel, un inhibiteur spécifique des récepteurs à l’ADP, a un effet protecteur sur le risque de survenue et la sévérité des ADD en réduisant l’importance de l’agrégation plaquettaire. Ces résultats sont en faveur d’un mécanisme d’agrégation ADP-dépendant conséquence des interactions entre les plaquettes et les bulles circulantes. La génération de thrombine, un autre puissant agoniste plaquettaire, interviendrait dans la genèse d’un état pro-thrombotique loco-régional conséquence des lésions induites par le passage des bulles au contact de l’endothélium vasculaire. Enfin, les résultats montrent le rôle joué par les micro-particules dans la diffusion à distance de cet état pro-thrombotique. Des études à venir devront confirmer l’intérêt d’une utilisation du clopidogrel dans le traitement de l’ADD ainsi que le rôle de l’oxyde nitric, de la prostacycline PGI2 et du shear-stress dans l’effet protecteur des vibrations et de l’exercice physique sur le risque de survenue de l’ADD en réduisant l’agrégation plaquettaire et en optimisant la cinétique d’élimination des bulles circulantes. / If bubble-induced mechanical obstruction of vessels is the central event during decompression, blood platelets play a key role in the pathogenesis of decompression sickness (DCS). But bubble-induced platelet aggregation (BIPA) mechanisms are unknown. In a previous study, we highlighted a post-dive decrease in platelet count in divers. The aims of this study was therefore to validate an experimental model of DCD in rat to clarify relationship between blood platelet and bubble formation, investigate platelet activation by measuring different platelet markers, and to study the effects of different antithrombotic pre-treatments. First, the results clearly indicate that clopidogrel, a powerful ADP-inhibitor, has a protective effect on decompression risk improving significantly DCS outcome and reducing DCS severity. The results point to the predominant involvement of the ADP release in the BIPA mechanism. Second, we showed the participation of thrombin generation, a powerful platelet agonist, in the thrombotic event. If local BIPA seems to be the direct consequence of bubble-blood component interactions in vascular bed, the regional thrombotic event could be the consequence of bubble-induced vessel wall injury with a key role played by micro-particles in the general extend. Further human studies should be aimed at demonstrating that clopidogrel can offer a benefit as an adjuvant in DCS treatment and the role played by nitric oxyd and prostacyclin in the protective effect of vibrations and physical exercise on DCS risk reducing bubble-induced platelet aggregation

This dissertation delivers a new computational framework for the automatic generation of geometric feature patterns for industrial design and architectural facades on free-form surfaces. Such patterns include holes in a speaker grill, showerhead holes, protrusions on ceramics or bumpy textures on a panel. These patterns play a key role in making a designed object aesthetically pleasing as well as functional. Computer Aided Design (CAD) systems currently offer tools for generating simple patterns, such as uniformly spaced rectangular or radial patterns. However, they are not applicable to more general cases required in industrial design, including arbitrarily shaped target geometry and graded feature sizes. These tools are limited in several ways: (1) They cannot be applied to free-form geometries used in industrial design, (2) Patterning of these features happens within a single working plane and is not applicable to highly curved surfaces, and (3) Created features lack anisotropy and spatial variations, such as changes in the size and orientation of geometric features within a given region. This thesis proposes a new method of taking input for a target region along with sizing metrics. It will generate feature patterns automatically in three steps: (1) packing isotropic or anisotropic cells tightly in a target region, (2) scaling features according to the specified sizing metrics, and (3) adding features on the base geometry. This approach automatically generates complex patterns that conform to the boundary of any specified region. User input of a small number of geometric features (called “seed features”) of desired size and orientation in preferred locations also can be specified within the target domain. These geometric seed features are then transformed into tensors and used as boundary conditions to generate a Riemannian metric tensor field. A form of the Laplace heat equation is used to generate the field over the target domain, subject to specified boundary conditions. The field represents the anisotropic pattern of the geometric features. The system is implemented as a plugin module in a commercial CAD package to add geometric features to the target region of the model using two set operations, union and subtraction. This method facilitates the creation of a complex pattern of hundreds of geometric features in minutes. All the features are accessible from the CAD system and can be manipulated individually if required by the user. This allows the industrial designer or architect to explore more alternatives by avoiding the tedious and time-consuming manual generation of these geometric patterns.

Ultrafine bubbles, often referred to as nanobubbles, have been used in various applications from environmental remediation to medicine. Even though the technology to generate ultrafine bubbles has been around for many years, the full potential of its applications has not been completely studied. This project seeks to study the use of ultrafine bubble technology for water treatment in combination with ozone gas. A factorial design experiment was chosen to test the effects of ultrafine bubbles on the concentration of an indicator organism, E. coli, in water as well as their effects on ozone gas being injected into water. Ozone gas or nitrogen gas was injected into water contaminated with E. coli as either ultrafine bubbles or fine bubbles as treatments for up to 60 minutes. Ultrafine bubbles were found to not have any significant effect on the concentration of E. coli in water. However, ultrafine bubbles did provide benefits when used in conjunction with ozone gas that regular, fine bubbles did not provide. The benefits included allowing the concentration of dissolved ozone in the water to decrease at a slower rate as well as allowing more ozone to dissolve into water at a higher rate than conventional methods of bubbling in ozone. While in this particular set of experiments the concentration of dissolved ozone in water didn't surpass 2 mg/L, which didn't allow for rapid disinfection and treatment of water, it is believed that with a more powerful ozone generator better results can be achieved. This project demonstrates the benefits and potential of injecting ozone gas as ultrafine bubbles into water as a way to effectively and efficiently disinfect and treat water.

Lors de la fusion du verre, de nombreuses petites bulles de gaz sont produites. L’affinage du verre a pour objectif de faire disparaitre ces bulles par l’ajout d’espèces réactives contribuant à la résorption des bulles ou à une augmentation de leur taille. La modélisation de l’hydrodynamique et des transferts nécessite l’étude des couplages entre convection, diffusion et réaction. Une approche locale à l’échelle de la bulle (simulation directe du transfert réactif et de l’écoulement) est utilisée pour déterminer le transfert interfacial. Des mesures de la propagation de fronts d’oxydation dans la fonte ont permis de préciser certaines propriétés physiques des espèces réactives. L’ensemble de cette analyse multi-échelles a contribué à l’élaboration d’un modèle de simulation d’un nuage de bulles / Many bubbles are generated during glass production. Due to the high viscosity of molten glass, their rising velocity is extremely low. The refining step consists in adding reactive agents to improve the glass quality. Bubble release is enhanced by chemical reaction (iron and sulfate oxidation-reduction) which will favor shrinkage or growth of bubbles through interfacial mass transfer. Better understanding of bubble cloud behavior in molten glass requires studying the interplay between convection, diffusion and chemical reactions. The direct numerical simulation of the flow and reactive mass transfer provided new insights on modeling interfacial bubble gas fluxes. The acceleration factor has been determined for simple reversible reactions in order to validate the simulation tool. Different Péclet and Damkhöler numbers have been tested to map all the different regimes (diffusion, convection and reaction). Together with those simulations, we have carried out series of experiments in molten glass : propagation of oxidation fronts. At different temperatures and for various glass compositions, we have determined physical properties of reactive species. A theoretical model of reactive transport for instantaneous reactions has been proposed to interpret experimental data. The core of this multi-scale analysis contributed to elaborating an Euler- Lagrange model to simulate bubble clouds in reactive media. This model has been applied to specific processes related to glass industry and can easily be extended to any reactive bubbly flows

The purpose of this paper is to examine the banking industry of Nepal from 2005 to 2010 to track the causes of banking crisis of 2011 using theories of macro-economics and finance as a conceptual starting point. In 2011, several commercial and development banks faced severe liquidity crisis that caused panic in the general public. Banks lost large amount of money in their loan and investment portfolios, which compelled the Government of Nepal to inject liquidity in the market. In the recent past years leading to the banking crisis of 2011, there was rapid change in the size and activity of banking industry. This paper analyzes changes that occurred in the financial market in the period of 2005 to 2010 that triggered the banking crisis. The roles of remittance, credit expansion and asset bubble have been analyzed in terms of their connection to the liquidity crisis.

O glicerol pode ser aproveitado em processos biotecnológico como substrato, para a obtenção de diversos produtos, dentre eles os óleos microbianos. Desta forma, o presente estudo teve como objetivo avaliar a obtenção de lipídeos pela levedura Rhodotorula glutinis a partir de glicerol. Na etapa inicial deste estudo foram realizados ensaios em frascos agitados de 250 mL, contendo 50 mL de meio, segundo um planejamento experimental 24, com face centrada e repetições no ponto central, no qual foram avaliadas as influências das variáveis concentração de substrato (40 a 200 g/L), da razão carbono/nitrogênio (20:1 a 100:1), pH (5 a 7) e concentração de inóculo (1 a 5 g/L), sobre a produção de lipídeos. Verificou-se nesta etapa que, dentro da região avaliada, a concentração de substrato, o pH e a razão carbono/nitrogênio (C/N), apresentaram efeitos estatisticamente significativos sobre a produção de lipídeos. Dentre estas variáveis, a concentração de substrato e o pH apresentaram comportamento quadrático, com pontos de máximo acúmulo de lipídeos próximos a 140 g/L e pH 6,5, respectivamente. Quanto a razão C/N, esta variável mostrou um efeito positivo sobre o acúmulo de lipídeos, ou seja, dentro a região avaliada, o aumento da razão C/N levou a um aumento do acúmulo de lipídeos pela levedura. Nos cultivos realizados nas condições determinadas pelo modelo para maximizar o acúmulo de lipídeos foram alcançadas concentrações de células de 30 ± 1 g/L e lipídios de 15 ± 3 g/L, em 200 h de cultivo. Na segunda etapa deste estudo foi avaliada a ampliação de escala dos cultivos da levedura de frascos agitados para biorreatores de agitação pneumática do tipo coluna de bolhas (CB) e airlift (AL), com volumes de 0,5 e 1,8 L. Os cultivos em biorreatores foram realizados empregando-se as condições otimizadas na etapa anterior deste trabalho. De modo geral, os cultivos realizados em biorreatores de bancada aprestaram concentração de células (15 a 21 g/L) e de lipídeos (5 a 9 g/L), inferiores aos observados em frascos agitados (30 g/L de células e 15 g/L de lipídeos). Tal resultado pode estar relacionado a condição de disponibilidade de oxigênio uma vez que o coeficiente volumétrico de transferência de oxigênio (kLa) para os cultivos em frascos agitados (kLa 49 h-1) foi superior ao alcançado em biorreatores (kLa ente 20 e 30 h-1). Nesta etapa, verificou-se ainda que, os biorreatores do tipo CB possibilitaram alcançar uma concentração de lipídeos (8 a 9 g/L) superior à obtida nos reatores AL (5 a 7 g/L), além de proporcionar uma condição de mistura mais eficiente. Quanto a composição do óleo microbiano (OM) extraído da biomassa celular ao fim do cultivo, verificou-se elevados teores dos ácidos graxos palmítico (C16:0), esteárico (C18:0), oleico (C18:1) e linoleico (C18:2), os quais corresponderam a cerca de 95% de sua composição. A proporção de ácidos graxos de dezesseis e dezoito carbonos do óleo microbiano assemelha-se a encontrada no óleo de soja (cerca de 94% de C16 e C18), o que possibilita o emprego deste óleo para finalidades semelhantes às do óleo de soja, como por exemplo, produção de biodiesel. / Glycerol can be used in biotechnological processes as a substrate to obtain various products, among them microbial oils. In this way, the present study aims to evaluate the lipids production by the yeast Rhodotorula glutinis from glycerol. In the initial stage of this study, experiments were performed in 250 mL shaken flasks, containing 50 mL of medium, according to an experimental design 24, face centered and repetitions at the central point, in which the substrate concentration (40 to 200 g/L), carbon/nitrogen ratio (20:1 to 100:1), pH (5 to 7) and inoculum concentration (1 to 5 g/L) effects on lipid production were evaluated. It was verified that, within the evaluated region, the substrate concentration, pH and carbon/nitrogen ratio (C/N) had statistically significant effects on lipid production. Among these variables, the substrate concentration and pH presented a quadratic behavior, with maximum lipids accumulation points close to 140 g/L and pH 6.5, respectively. The C/N ratio presented a positive effect on the lipid accumulation, that is, within the region evaluated, the increase in the C/N ratio led to an increase in the lipid accumulation by yeast. Cultures performed under conditions determined by the model to maximize lipid accumulation reached cell concentrations of 30 ± 1 g/L and lipids of 15 ± 3 g/L in 200 h of culture. In the second stage of this study, the scale-up of the yeast shake flasks cultures for bubble column (CB) and airlift (AL) pneumatic agitation bioreactors, with volumes of 0.5 and 1.8 L, were evaluated. Cultures in bioreactors were performed using the optimized conditions in the previous stage of this work. In general, cultures in bioreactors presented cells concentrations (15 to 21 g / L) and lipids (5 to 9 g/L) lower than those observed in shaker flasks (30 g/L of cells and 15 g/L of lipid). This result may be related to the oxygen availability condition since the volumetric oxygen transfer coefficient (kLa) for cultures in shaker flasks (kLa 49 h-1) was higher than in bioreactors (kLa 20 and 30 h-1). In this stage, it was also verified that CB-type bioreactors achieve a lipid concentration (8 to 9 g/L) higher than that obtained in AL reactors (5 to 7 g/L), besides providing more efficient mixing conditions. About the composition of the microbial oil (MO), extracted from the cell biomass at the end of the cultivation, presented high levels of palmitic (C16: 0), stearic (C18: 0), oleic (C18:1) and linoleic (C18:2) fatty acids, which corresponded to about 95% of its composition. The proportion of microbial oil fatty acids of sixteen and eighteen carbons resembles that found in soybean oil (about 94% C16 and C18), which makes it possible to use this oil for similar purposes as soybean oil, such as biodiesel production, for example.