No description available.

Abrasive waterjet (AWJ) cutting is one of the most recently developed manufacturing technologies. It is superior to many other cutting techniques in processing various materials, particularly in processing difficult-to-cut materials. This technology is being increasingly used in various industries. However, its cutting capability in terms of the depth of jet penetration and kerf quality is the major obstruction limiting its further applications. More work is required to fully understand the cutting process and cutting mechanism, and to optimise cutting performance. This thesis presents a comprehensive study on the controlled nozzle oscillation technique aiming at increasing the cutting performance in AWJ machining. In order to understand the current state and development in AWJ cutting, an extensive literature review is carried out. It has found that the reported studies on controlled nozzle oscillation cutting are primarily about the use of large oscillation angles of 10 degrees or more. Nozzle oscillation in the cutting plane with such large oscillation angles results in theoretical geometrical errors on the component profile in contouring. No published attempt has been found on the study of oscillation cutting under small angles although it is a common application in practice. Particularly, there is no reported research on the integration of nozzle oscillation technique into AWJ multipass cutting, which is expected to significantly enhance the cutting performance. An experimental investigation is first undertaken to study the major cutting performance measures in AWJ single pass cutting of an 87% alumina ceramic with controlled nozzle oscillation at small angles. The trends and characteristics of cutting performance quantities with respect to the process parameters as well as the science behind which nozzle oscillation affects the cutting performance have been analysed. It has been shown that as with oscillation cutting at large angles, oscillation at small angles can have an equally significant impact on the cutting performance. When the optimum cutting parameters are used for both nozzle oscillation and normal cutting, the former can statistically increase the depth of cut by 23% and smooth depth of cut by 30.8%, and reduce kerf surface roughness by 11.7% and kerf taper by 54%. It has also been found that if the cutting parameters are not selected properly, nozzle oscillation can reduce some major cutting performance measures. In order to correctly select the process parameters and to optimise the cutting process, the mathematical models for major cutting performance measures have then been developed. The predictive models for the depth of cut in both normal cutting and oscillation cutting are developed by using a dimensional analysis technique. Mathematical models for other major cutting performance measures are also developed with the aid of empirical approach. These mathematical models are verified both qualitatively and quantitatively based on the experimental data. The assessment reveals that the developed models conform well to the experimental results and can provide an effective means for the optimum selection of process variables in AWJ cutting with nozzle oscillation. A further experimental investigation of AWJ cutting of alumina ceramics is carried out in order to study the application of AWJ oscillation technique in multipass cutting. While high nozzle traverse speed with multipass can achieve overall better cutting performance than low traverse speed with single pass in the same elapsed time, it has been found that the different combination of nozzle traverse speed with the number of passes significantly affects cutting process. Optimum combination of nozzle traverse speed with the number of passes is determined to achieve maximum depth of cut. It has also demonstrated that the multipass cutting with low nozzle traverse speed in the first pass and a comparatively high traverse speed for the following passes is a sensible choice for a small kerf taper requirement. When nozzle oscillation is incorporated into multipass cutting, it can greatly increase the depth of cut and reduce kerf taper. The predictive models for the depth of cut in both multipass normal cutting and multipass oscillation cutting are finally developed. With the help of dimensional analysis, the models of the incremental cutting depth for individual pass are derived based on the developed depth of cut models for single pass cutting. The models of depth of cut for a multipass cutting operation are then established by the sum of the incremental cutting depth from each pass. A numerical analysis has verified the models and demonstrated the adequacy of the models' predictions. The models provide an essential basis for the development of optimization strategies for the effective use of the AWJ cutting technology when the multipass cutting technique is used with controlled nozzle oscillation.

The model-free analysis of NMR relaxation data, which is widely used for the study of protein dynamics, consists of the separation of the Brownian rotational diffusion from internal motions relative to the diffusion frame and the description of these internal motions by amplitude and timescale. Through parametric restriction and the addition of the Rex parameter a number of model-free models can be constructed. The model-free problem is often solved by initially estimating the diffusion tensor. The model-free models are then optimised and the best model is selected. Finally, the global model of all diffusion and model-free parameters is optimised. These steps are repeated until convergence. This thesis will investigate all aspects of the model-free data analysis chain. (For complete abstract open document)

Resumo: Esta pesquisa tem como objetivo buscar compreender como ocorre a participação dos alunos, e quais elementos sociais e pedagógicos tornam-se presentes ao se abordar questões ambientais, na perspectiva do ensino-aprendizagem da Modelagem Matemática. A pesquisa foi desenvolvida numa abordagem qualitativa com alunos do Ensino Fundamental e Médio, ambas escolas públicas do município de Rio Claro, SP. A ação pedagógica, desenvolvida na perspectiva de Modelagem Matemática, envolveu os temas Água, Lixo, Energia Elétrica e Desmatamento, com a confecção de alguns modelos, dentre os quais, previsão do crescimento da população, do consumo de água, da produção de lixo e da energia elétrica do município de Rio Claro. As questões ambientais vislumbraram um campo rico de aplicações, permitindo integrar a experiência dos alunos com a Matemática e possibilitando aos mesmos levantarem dados e informações, formularem, resolverem e decidirem questões de interesse, ao mesmo tempo que passavam a ter consciência em relação aos problemas ambientais. No Ensino Fundamental, além da Matemática e de outras ciências do contexto escolar, os alunos trabalharam com tecnologia informática. As discussões geradas e o contato com os vários órgãos responsáveis por dados e informações contribuíram para que os alunos compreendessem a necessidade da conservação dos recursos naturais com os quais interagem, introduzindo novos hábitos e atitudes em relação a eles, visualizando a Matemática como um instrumento de análise e interpretação da realidade. Os dados desta pesquisa revelam que é possível a escola desenvolver atividades matemáticas relacionadas com a temática ambiental, proporcionando um ensino da Matemática mais envolvente e significativo para os alunos. / Abstract: This research has the objective of understanding how students' participation takes place, and which social and pedagogic elements are present, when environmental matters are approached from the standpoint of teaching and learning of Mathematical Modelling. The research was developed through a qualitative approach with students from Elementary and Junior High Schools, in two Public Schools in the town of Rio Claro, SP, Brazil. The pedagogic action, developed under the perspective of Mathematical Modelling included topics such as: Water, Waste, Electric Power and Deforestation, with the execution of certain models, from among which: population growth forecast; water consumption; waste and electric power generation in the town of Rio Claro. Environmental subjects discern a field that's rich in applications, which allow for the integration between students' experiences and Mathematics, letting them collect data and information, formulate, resolve and decide on questions of interest, concurrent with a gain of awareness towards environmental problems. In Elementary School, in addition to Mathematics and other sciences within the scholar context, students worked with Information Technology. Discussions that were created and the contact with several agencies responsible for the data, helped the students understand the needs for conservation of natural resources with which they interact, introducing new habits and demeanor in relation to them, perceiving Mathematics as a tool for analysis and interpretation of reality. Data in this research show that it's possible to develop mathematical activities related to environmental subjects, offering the students a more involving and significant Mathematics teaching. / Orientador: Maria Lúcia Lorenzetti Wodewotzki / Coorientador: Antonio Carlos Carrera de Souza / Banca: Ademir Donizeti Caldeira / Banca: Alexandrina Monteiro / Banca: Miriam Godoy Penteado / Banca: João Frederico da Costa Azevedo Meyer / Doutor

Polylactic acid (PLA) is one of the most promising bio-compostable and biodegradable thermoplastic made from renewable sources. PLA, is typically obtained by polymerising lactide monomer. The technique mainly used for ring opening polymerization (ROP) of Lactide is based on metallic/bimetallic catalyst (Sn, Zn, and Al) or other organic catalysts in suitable solvent. However, the PLA synthesized using such catalysts may contain trace elements of the catalyst, which may be toxic. In this work, reactive extrusion experiments using stannous octoate Sn(Oct)2 and tri-phenyl phosphine (PPh)3 were considered to perform ROP of lactide monomer using ultrasound as an alternative energy (AE) source for activating and/or boosting the polymerization. Mathematical model of ROP of lactide, was developed to estimate the impact of reaction kinetics and AE source on the polymerization process. Ludovic® software, a commercial code was used. It was adopted for the simulation of continuous reactive extrusion of PLA. Results from experiments and simulations were compared to validate the simulation methodology. Results indicate that the application of AE source in reaction process boost the PLA formation rate. Result obtained through Ludovic simulation and experiments were validated. It was shown that there is a case for reducing the residence time distribution (RTD) in Ludovic due to the ‘liquid’ monomer flow in the extruder. This change in the parameters resulted in validation of the simulation. However, it was concluded that the assumption would have to be established by doing further validations. The simulation model includes the details of kinetics of reactions involved with in the process and helps to upscale the reaction output. This work also estimates the usefulness and drawbacks of using different catalysts as well as effect of alternative energies and future aspects for PLA production.

Submitted by Gabriel Figueiredo Puerta (gfpuerta@gmail.com) on 2018-11-12T17:24:33Z No. of bitstreams: 1 diss_Gabriel_Final_real_oficial_.pdf: 2623476 bytes, checksum: a1018496f3944da6d7825e6856e593ac (MD5) / Approved for entry into archive by Cristina Alexandra de Godoy null (cristina@adm.feis.unesp.br) on 2018-11-13T16:56:42Z (GMT) No. of bitstreams: 1 puerta_gf_me_ilha.pdf: 2623476 bytes, checksum: a1018496f3944da6d7825e6856e593ac (MD5) / Made available in DSpace on 2018-11-13T16:56:42Z (GMT). No. of bitstreams: 1 puerta_gf_me_ilha.pdf: 2623476 bytes, checksum: a1018496f3944da6d7825e6856e593ac (MD5) Previous issue date: 2018-08-23 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Neste trabalho, apresenta-se um modelo matemático especializado para a otimização do problema da restauração de redes de distribuição de energia elétrica radiais quando existem poucas chaves manobráveis. A existência de um número reduzido de chaves seccionadoras na rede de distribuição é uma característica típica de sistemas reais. O problema da restauração de redes de distribuição consiste em obter um plano de restauração que restabeleça o máximo possível de cargas presentes na parcela do sistema que foi desenergizada após a ocorrência de uma falta permanente. O plano de restauração deve ser obtido o mais rápido possível. O modelo matemático proposto possui como objetivo maximizar o restabelecimento de energia àqueles setores de carga que estão desenergizados e minimizar o número total de chaveamentos necessários para que a restauração aconteça. O modelo matemático resultante é de programação cônica de segunda ordem inteira mista e é resolvido através de solucionadores comerciais que se baseiam em técnicas de otimização clássicas. Os testes considerando faltas permanentes foram simulados em sistemas de distribuição testes de 53 e 84 barras. Os sistemas foram adaptados e transportados para a problemática do trabalho e os resultados adquiridos qualificam a robustez e a eficácia do modelo matemático especializado proposto para a resolução do problema. / This work proposes an specialized mathematical model for the optimization of the restoration problem in radial electric power distribution systems with few manageable switches. The existence of a reduced number of switches in the distribution systems it’s a typical characteristic of real systems. The restoration problem in radial electric power distribution systems is to obtain a restoration plan that reestablishes the maximum amount of present loads in the system portion that has been de-energized after the occurrence of a permanent fault. The elaboration of the restoration plan must be as soon as possible. The proposed mathematical model aims at maximizing the power re-establishment to the de-energized load sectors and at minimizing the total amount of operated switches that are required for the restoration happen. The resulting mathematical model is a mixed-integer second-order conic programming problem and it’s solved through commercial solves that use classical optimization techniques as base. The tests considering permanents faults were simulated in 84 and 53-bus distribution systems. The systems were adapted and transported to the work’s problematic and the acquired results show the robustness and the effectiveness of the proposed specialized mathematical model for the problem resolution. / CAPES: 001 / FAPESP: 2015/21972-6

Orientador: Rubén Augusto Romero Lázaro / Resumo: Neste trabalho, apresenta-se um modelo matemático especializado para a otimização do problema da restauração de redes de distribuição de energia elétrica radiais quando existem poucas chaves manobráveis. A existência de um número reduzido de chaves seccionadoras na rede de distribuição é uma característica típica de sistemas reais. O problema da restauração de redes de distribuição consiste em obter um plano de restauração que restabeleça o máximo possível de cargas presentes na parcela do sistema que foi desenergizada após a ocorrência de uma falta permanente. O plano de restauração deve ser obtido o mais rápido possível. O modelo matemático proposto possui como objetivo maximizar o restabelecimento de energia àqueles setores de carga que estão desenergizados e minimizar o número total de chaveamentos necessários para que a restauração aconteça. O modelo matemático resultante é de programação cônica de segunda ordem inteira mista e é resolvido através de solucionadores comerciais que se baseiam em técnicas de otimização clássicas. Os testes considerando faltas permanentes foram simulados em sistemas de distribuição testes de 53 e 84 barras. Os sistemas foram adaptados e transportados para a problemática do trabalho e os resultados adquiridos qualificam a robustez e a eficácia do modelo matemático especializado proposto para a resolução do problema. / Abstract: This work proposes an specialized mathematical model for the optimization of the restoration problem in radial electric power distribution systems with few manageable switches. The existence of a reduced number of switches in the distribution systems it’s a typical characteristic of real systems. The restoration problem in radial electric power distribution systems is to obtain a restoration plan that reestablishes the maximum amount of present loads in the system portion that has been de-energized after the occurrence of a permanent fault. The elaboration of the restoration plan must be as soon as possible. The proposed mathematical model aims at maximizing the power re-establishment to the de-energized load sectors and at minimizing the total amount of operated switches that are required for the restoration happen. The resulting mathematical model is a mixed-integer second-order conic programming problem and it’s solved through commercial solves that use classical optimization techniques as base. The tests considering permanents faults were simulated in 84 and 53-bus distribution systems. The systems were adapted and transported to the work’s problematic and the acquired results show the robustness and the effectiveness of the proposed specialized mathematical model for the problem resolution. / Mestre

Orientador: Prof. Dr. Magno Enrique Mendoza Meza / Dissertação (mestrado) - Universidade Federal do ABC, Programa de Pós-Graduação em Engenharia Mecânica, 2017. / Este trabalho apresenta (i) a implementação de uma mão robótica humana produzida por impressora 3D com a caracterização de sensores flexíveis e de força e os respectivos algoritmos de operação; (ii) a modelagem cinemática e desenvolvimento de seu simulador; (iii) a modelagem dinâmica e seu respectivo simulador e (iv) simulações numéricas da dinâmica de contato entre a ponta do dedo da mão e um objeto. A aplicação da tecnologia de impressão 3D é crescente na medicina. Para criar proteses mais naturais adiciona-se o conhecimento de outras areas como robotica e sistemas de controle. Os sensores flexíveis sao necessarios para adquirir o sinal de referência para o posicionamento dos dedos por meio de uma luva na qual estes sensores encontram-se fixados. Os de forca sao usados para controlar a m~o com base na forca que se deseja aplicar ao segurar um objeto, seu comportamento foi estudado e a curva de cada sensor obtido experimentalmente. O microcontrolador Arduino® é responsavel pela aquisição de dados dos sensores e pelo acionamento dos servomotores que movimentam os dedos. A modelagem cinematica da mao robotica foi realizada com base na convenção de Denavit-Hartenberg e a modelagem dinamica com uso das equações de Euler-Lagrange a partir das funções de energia. Os simuladores foram criados com uso de Matlab/Simulink®, os quais permitem verificar os modelos cinematico e dinamico obtidos, todavia, dada a complexidade matematica envolvida na modelagem foi criado um programa para realizar os passos matematicos; fornecer o codigo para os simuladores criados; e um breve relatorio com as equações resultantes para rápida verficação, ou com os resultados das simulções numericas. O modelo com restriçãoo ao movimento é obtido adicionando multiplicadores de Lagrange à função Lagrangiana de energia de acordo com as equações de restrição ao movimento, esses multiplicadores fornecem a força de contato. O instante da colisão é modelado para a correta execução da dinâmica dos corpos acoplados. Dois modelos foram obtidos (i) um modelo simplificado que conserva o momento linear entre a a ultima articulaçõa e a barreira e que atualiza as velocidades das outras articulações como uma fração das velocidades que antecedem o impacto; (ii) um modelo completo, o qual considera-se todos os torques e distancias envolvidas para a atualização das velocidades. Os resultados mostraram que o modelo simplificado é capaz de obter resultados proximos do modelo completo dependendo do ajuste do parametro que define a fração de velocidades. Testes utilizando uma simplificação das não linearidades para angulos e velocidades pequenos se mostraram satisfatorios somente quando o sistema nao entrar em contato com a barreira. / This works presents (i) the implementation of a robotic hand made by a 3D-Printer with the characterization of flexible and force sensors and the respective algorithms of operation, (ii) the kinematic modeling and its simulator development, (iii) the dynamic modeling and its respective simulator and (iv) numerical simulations of contacts dynamics between the fingertip of the hand with an object. Applications of this technology are increasing worldwide in medicine. To create more natural prostheses it is added knowledge from other areas such as robotics and control systems. Flexible sensors are needed to acquire the reference signal to move each finger from a glove in which these sensors are assembled. Force sensors are used to control the hand reading the applied force when it is desired to hold an object. The microcontroller Arduino® is responsible for acquiring data from the sensors and actuation of servomotors that move the fingers. Kinematic modeling of the prosthesis was based on Denavit-Hartenberg convention and dynamic modeling with use of the Euler-Lagrange from energy functions. The simulators were created with use of Matlab/Simulink®, they allow to verify the kinematic and dynamic models obtained, however, given the mathematical complexity involved in modeling, it was created a program to perform mathematical steps; to provide the code to simulator's blocks; and a short report with the resulting equations for simple checking, or with simulations results. The move constrained model is obtained adding Lagrange multipliers to the Lagrangian energy function accordingly to the restriction equations, these multipliers provide the contact force. The collision instant is modeled to the correct execution of the coupled bodies dynamics. Two models was made (i) one simplifoed model holding only the linear momentum conservation between the last articulation and the barrier,which updates other velocities as a fraction of the velocity before impact; (ii) one complete model, that evaluates every torque and distance of the bodies to update each velocities. The results show the simplified model is able to achieve similar results of the complete model depending of the adjustment of the parameter that denes the fraction of velocities. Some tests with a simplification of nonlinearities for small angles and velocities was made, although results were only satisfactory when no contact happens.

Made available in DSpace on 2014-06-11T19:31:42Z (GMT). No. of bitstreams: 0 Previous issue date: 2004-12-20Bitstream added on 2014-06-13T20:22:45Z : No. of bitstreams: 1 jacobini_or_dr_rcla.pdf: 1226882 bytes, checksum: 334ebeb4b436a3efc244c87f112026fc (MD5) / Com esta pesquisa busquei analisar as possibilidades de crescimento político dos estudantes, quando a modelagem matemática é adotada como estratégia de ensino-aprendizagem. Considero essas possibilidades como reflexões que se identificam com questionamentos, críticas, ações e transformações. Ao inserir este estudo no contexto da Educação Matemática Crítica baseei minha pesquisa bibliográfica em autores que se identificam com esse movimento. Adotei na pesquisa uma abordagem qualitativa e para a coleta de dados organizei três ambientes de aprendizagem, denominados cenários para investigação, com estudantes voluntários das séries iniciais dos Cursos de Matemática e de Engenharia de Computação, ambos da PUC-Campinas. A análise dos dados indica que o crescimento político dos atores nos cenários associa-se, de um lado, com a conscientização política resultante da sua atuação em investigações diretamente relacionadas com os temas dos projetos de modelagem, nas discussões que acompanham os resultados obtidos e no debate sobre o alcance desses resultados e sobre as conseqüências sociais do trabalho realizado. Do outro lado, esse crescimento associa-se com uma ação política que se concretiza por meio do envolvimento do estudante com a comunidade. Concluo este estudo indicando que o processo de crescimento político dos estudantes deve ser pensado como uma forma de alfabetização matemática, estreitamente relacionada com o núcleo de uma literacia matemática voltada para mudanças sociais. / With this research I intended to analyze the possibilities of the students' political growth when the mathematical modelling is adopted as teaching-learning strategy. I consider those possibilities as reflections that identify with questioning, critics, actions and transformations. When inserting this study in the context of the Critical Mathematics Education I based my bibliographical research on authors that identify with this movement. The research was performed within a qualitative approach, starting from the construction of three learning environment, denominated scenarios to investigation, organized with voluntary students from the initial series of the courses of Mathematics and Engineering of Computation, both of PUC-Campinas University. The analysis of the data indicates that the actors' political growth in the sceneries associates, on a side, with the political awareness as a result of their involvement in investigations directly related with the themes of the modelling projects, in the discussions that accompany the obtained results and in the debate about the reach of those results, and about the social consequences of the accomplished work. On the other hand, that growth associates with a political action that is rendered through the student's involvement with the community. I conclude this study indicating that the process of the students' political growth should be thought as a form of mathematical literacy, strictly related with the nucleus of a mathematical literacy focused on social changes.

Made available in DSpace on 2016-06-02T19:56:55Z (GMT). No. of bitstreams: 1 6081.pdf: 3692615 bytes, checksum: c275b7cf71f1d4575f5deeff0b9c56cc (MD5) Previous issue date: 2014-03-18 / Universidade Federal de Sao Carlos / Ethanol is a renewable fuel produced from biomass. Starting in 1975, Brazil began a national program to promote ethanol production in large scale, the ProAlcool Programme. In the harvest 2012/13 it was produced 23.64 billion of litters of ethanol, from that 9.85 billion was in anhydrous form. The anhydrous ethanol is mainly used as a gasoline additive. This mix reduces the dependence on fossil fuels, improves the engine performance and reduces the emission of toxic gases. Anydrous ethanol can be produced by azeotropic distillation (cyclohexane), by extractive distillation (mono ethylene glycol) or by molecular sieves technology (zeolytes). The objective of this work was the construction and operation of a distillation apparatus (in a bench scale) with the purpose of evaluating the extractive distillation of hydroalcoholic solutions of ethanol employing glycerol as solvent, in substitution of mono ethylene glycol. The substitution is motivated by several factors indicated by the literature like the biodiesel production growth, whose glycerol is a byproduct; the low toxicity; the superior capacity to promote the relative volatility increase of water-ethanol solution; and the inferior energetic consumption. The column was composed by three glass modules with five perforated plates each module. On the top it was used a glass coil condenser. As reboiler it was used a two litters round bottom boiling flask linked to a heating mantle. The feed of ethanol, solvent (glycerol) and the bottom product withdraw was made with pumps. The distillation system was modelled and simulated in the Aspen Plus software using the RadFrac distillation model and NRTL (non-random two-liquid model) thermodynamic model. Four simulations were made for different ethanol feed compositions (25, 35 and 45 wt %). Three simulations were made without the glycerine feed and one with solvent feed and 35% ethanol feed. The distillation column was operated in total reflux with 0.599 L/h ethanol feed flow and 0.341L/h glycerine feed flow. The systems without the solvent reached a mass purity of ethanol of 94.1% in distillate (45% of ethanol feed). The process with the solvent reached a purity of 99.8% in distillate. The four simulated conditions were experimentally evaluated. The experimental results confirmed that the use of glycerine feed promoted an improvement in separation, reaching distillate purity over 99%. Experiments without the glycerine reached compositions around 90%. Murphree efficiencies were determined in all experiments carried out. For the systems without the solvent the obtained efficiencies were: 45% (25% ethanol feed), 30% (35% ethanol feed), 35% (45% ethanol feed). The system with glycerine feed reached 70% efficiency. / O etanol é um combustível renovável que teve sua produção incentivada no Brasil a partir da implantação do Programa Nacional do Álcool em 1975. Na safra 2012/13 foram produzidos 23,64 bilhões de litros, destes 9,85 bilhões em sua forma anidra. Nesta composição é utilizado como aditivo na gasolina (25% em volume) o que contribui para redução da dependência dos combustíveis fósseis, melhora do desempenho dos motores e diminuição na emissão de gases tóxicos. A produção de etanol anidro pode ser realizada pela destilação azeotrópica (cicloexano), extrativa (monoetilenoglicol) ou por meio do emprego da tecnologia das peneiras moleculares (zeólitas). Este trabalho teve como objetivo a construção e operação de um aparelho de destilação (em escala de bancada) com o objetivo de avaliar a destilação extrativa de soluções hidroalcoólicas de etanol empregando glicerol como solvente em substituição ao monoetilenoglicol. Essa substituição é motivada por diversos fatores apontados pela literatura como o crescimento da produção de biodiesel, que possui como subproduto o glicerol; a menor toxicidade do glicerol em relação ao monoetilenoglicol; a maior capacidade em promover o aumento da volatilidade relativa da solução água-etanol; e o menor consumo energético. A coluna de destilação era composta de três módulos de vidro com cinco pratos perfurados em cada módulo. No topo foi utilizado um condensador de vidro tipo serpentina. Como refervedor foi utilizado balão de fundo redondo de dois litros acoplado a uma manta de aquecimento. A alimentação das correntes de etanol, solvente e a retirada do produto de fundo da coluna foram realizadas por bombas. O sistema de destilação foi simulado no aplicativo Aspen Plus utilizando o bloco RadFrac e o modelo termodinâmico NRTL (Non-random two-liquid model). Foram realizadas quatro simulações variando-se a composição de etanol na corrente de alimentação (25%, 35% e 45% m/m). Três simulações foram feitas sem a alimentação de glicerina e uma simulação com alimentação do solvente e alimentação de etanol de 35%. A simulação representou o processo em refluxo total com vazão de alimentação de etanol de 0,599 L/h e alimentação de glicerina com vazão de 0,341 L/h. As simulações mostraram que para os sistemas sem alimentação de glicerina a máxima pureza mássica de etanol obtida no destilado foi de 94,1% (alimentação com 45% de etanol). Para a operação com alimentação do solvente a pureza obtida no destilado atingiu 99,8%. As quatro condições simuladas foram avaliadas experimentalmente. Os resultados experimentais confirmaram que o uso da alimentação de glicerina promoveu uma melhora na separação, obtendo um destilado com pureza mássica de etanol superior a 99%. Os experimentos sem a presença do solvente atingiram concentrações próximas de 90% no destilado. A eficiência de Murphree foi determinada em cada uma das operações. Para os sistemas sem a presença de solvente, as eficiências obtidas foram de 45% (25% de etanol na alimentação), 30% (35% de etanol na alimentação) e 35% (45% de etanol na alimentação). Para o sistema com alimentação de glicerina a eficiência foi de 70%.