Thèse d'exercice : Pharmacie : Université de Nantes : 2003. / Bibliogr. f. 132-135 [49 réf.].

No description available.

Heat penetration data were obtained for a canned food model (gelatinized waxy maize starch) during processing in an agitating retort under various conditions to evaluate the effect of process variables on the heat penetration parameters (f$ sb{ rm h}$ and j$ sb{ rm ch}$) and resulting process lethality (F$ sb{ rm o}$) and cook value (C$ sb{ rm o}$). Overall heat transfer coefficient (U$ sb{ rm o}$), Hunter L, a and b values were determined. The study indicated that f$ sb{ rm h}$, j$ sb{ rm ch}$, F$ sb{ rm o}$, C$ sb{ rm o}$ and U$ sb{ rm o}$ were influenced (p $<$ 0.05) by all process variables except can headspace. Retention percentage of thiamin, ascorbic acid and chlorophyll at each process condition was estimated using the heat penetration data and documented D and z values for each quality factor. Results confirm better retention of quality factors at higher temperatures and higher rotation speeds. / Rheological changes associated with the starch were evaluated in relation to different factors and were well described by the Power law model. Apparent viscosity of the processed product was related to the cook value of the process. / Fresh vegetables were canned and subjected to end-over-end rotation. Heating times were adjusted to give approximately the same process lethality (F$ sb{ rm o}$ = 10 minutes). Processing under higher temperatures and higher rotation speeds resulted in better retention (p $<$ 0.01) of vegetable texture, while only temperature (higher the better) had a desirable effect on color. / Buffered aqueous mixtures of ascorbic acid and thiamin were sealed in glass vials and cans. At each temperature, rotation speed did not have a significant effect on the retention of either vitamin. Processing at 110$ sp circ$C yielded significantly lower retention (p $<$ 0.01) of both vitamins than at 120$ sp circ$C and 130$ sp circ$C. Estimated ascorbic acid and thiamin retention using experimental time-temperature data and published kinetic parameters showed fairly good agreement with experimental data.

Orientador: Rubens Maciel Filho / Dissertação (mestrado) - Universidade Estadual de Campinas. Faculdade de Engenharia Quimica / Made available in DSpace on 2018-07-26T21:05:08Z (GMT). No. of bitstreams: 1 Aguiar_HelenaCristinaI.L_M.pdf: 4620932 bytes, checksum: 599c90f6c71f6effafd2b6b0e96021e6 (MD5) Previous issue date: 2000 / Resumo: Estudos demonstram que a expectativa de crescimento do consumo de papel e baixo custo de produção no Brasil pode tornar a indústria de celulose e papel do país a maior geradora de negócios neste ramo internacionalmente. O processo de produção de celulose é complexo, e apesar de muito estudado ainda há muitas lacunas que precisam ser preenchidas. Por outro lado, as redes neurais são uma solução para a modelagem de processos cujo conhecimento é incompleto ou cuja complexidade dificulta a produção de bons resultados através da modelagem determinística. O objetivo principal desta tese foi a criação de um modelo para a predição do grau de polpação da madeira a partir de dados industriais, utilizando técnicas de modelagem diferentes. O trabalho explorou principalmente a modelagem através de redes neurais, mas também se dedicou à aplicação dos dados industriais a um modelo determinístico e a sua posterior combinação com uma rede neural para o desenvolvimento de um modelo híbrido. A discussão sobre os recursos necessários para a utilização de cada técnica, bem como a comparação entre as diferentes metodologias, suas vantagens e desvantagens, também são assuntos deste trabalho. A Aracruz Celulose SA, a maior fábrica do Brasil, cedeu os dados para o desenvolvimento do modelo. A avaliação de vários modelos encontrados na literatura selecionou o modelo determinístico a ser utilizado. O modelo neural apresenta uma rede "feedforward" treinada com o algoritmo de retropropagação. Inicialmente, os dados industriais foram avaliados e apenas aqueles que mostraram trazer informação relevante foram utilizados. A estrutura e parâmetros da rede foram otirnizados de modo a melhorar a performance do modelo. A combinação do resultado do modelo determinístico com o modelo neural formaram o modelo híbrido. A fábrica não pode ceder o esperado número de conjuntos de dados para o treinamento da rede neural. No entanto, a qualidade dos dados e o rigor com que foram obtidos, conjuntamente com um cuidadoso trabalho de seleção desses dados possibilitaram a busca de soluções para o sucesso do modelo. A alternativa encontrada foi a criação de um novo conjunto de dados, obtidos através de uma curva de correlação dos dados industriais. Apesar do pequeno conjunto de treinamento, a rede neural produziu resultados satisfatórios, onde o erro entre os valores esperados e preditos foram menores que o erro experimental para a determinação de lignina remanescente. O modelo determinístico foi capaz de reproduzir a taxa de deslignificação da celulose no tempo, o que determina o grau de polpação, e portanto foi considerado adequado para a utilização no modelo híbrido. A rede híbrida gerou resultados um pouco melhores que a rede pura, e seu treinamento foi significativamente mais rápido. Um modelo de computação rápida e que reproduza os resultados esperados pode ser implementado em um sistema supervisor ou de controle para funcionar "on-line". Os resultados apresentados mostram que tanto as redes neurais puras quanto os modelos luoridos, quando bem treinados e otimizados, atendem a esses requisitos. Mostram também que é possfvel utilizar variáveis normais de processo, tomando mais viável o desenvolvimento de modelos específicos / Abstract: Studies show that the expected growth of paper consumption and low production cost in Brazil may turn its pulp and paper industry into the greatest business generator in this field worldwide. The process to produce pulp is complex, and although it has been extensively researched, there are still many questions to be answered. On the other hand, neural networks may be the solution for processes whose knowledge is incomplete or whose complexity makes it harder to obtain good results with first principIe models. The main goal of this thesis was the development of a model to predict pulping degree using industrial data, with different modeling techniques. It explored mainly neural networks technique, but also did it apply the data into a first principIe model, which was later used for the development of a hybrid mode!. The discussion about the resources required for each technique, as well as the different methods, their advantages and limitations, is part of this work. Aracruz Celulose SA, the largest mill in Brazil, provided the data for the model development. The evaluation of published models set the criteria for the selection of the deterministic model to be used. The neural model presents a feedforward network trained with backpropagation algorithm. After an evaluation of the industrial data, only the variables that showed to carry relevant information were used. The network structure and its parameters were optimized in order to improve model performance. The combination of the result obtained from the deterministic model with a neural network composed the hybrid model. The mill could not provide the expected amount of training data. However, the quality of the data and the rigorous data collection procedure, as well as the careful data evaluation made it possible to search for solutions for a successful model. The alternative was the production of a new data set formed with points of the correlation curve obtained from industrial data. The difference between the expected and predicted values was lower than the experimental error inherent to the lab test for determination of remaining lignin. The fIrst principIe model was able to reproduce delignification rate, which determines the pulping degree, and therefore was considered adequate to be used in the hybrid model. The hybrid network results were slightly better than the ones obtained with the pure net and its training was appreciably faster. A model which is able to reproduce expected results in a timely fashion can be implemented in a control or supervisory system that works on-line. The results showed that when well trained and optimized, both, the pure nets and the hybrid models, fulfill these requirements. They also show that it is possible to use normal process variables, making it more feasible to develop customized models / Mestrado / Desenvolvimento de Processos Químicos / Mestre em Engenharia Química

Orientadores: Enrique Ortega Rodriguez, Pilar Rodriguez Massaguer / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia de Alimentos / Made available in DSpace on 2018-07-25T10:18:50Z (GMT). No. of bitstreams: 1 Frangipani_MarialuciOliveira_M.pdf: 16516753 bytes, checksum: a9b95e19bcc5e8d398a68ef021137390 (MD5) Previous issue date: 1999 / Resumo: Este trabalho mostra a viabilidade de novas técnicas de se monitorar um processo de esterilização em autoclave para refeições prontas em bandejas plásticas semi-rígidas. Nesta pesquisa, utilizou-se um programa de uso industrial para aquisição de dados (Genie, v.3.0 - Advantech, E.U.A) e o adaptamos para o uso com as interfaces PCl-818H e PClD-789D (Advantech, EU.A). O sistema adaptado demonstra em tempo real as temperaturas no processo, assim como a curva de letalidade acumulada durante todo o processo. O novo sistema foi comparado em sua performance ao sistema já em uso no laboratório de Termobacteriologia da FEA, UNICAMP, o sistema OORIC de aquisição de dados, mostrando ser tão eficiente quanto o mesmo, porém, mais versátil / Abstract: The present dissertation provides insights for monitoring an autoclave sterilization process for pre-cooked food products, placed in plastic semirigid trays. For this purpose, we used a software (Genie, v.3.0 - Advantech, EU.A) connected to two data-acquisition systems, PCl-818H e PClO-l890 (Advantech, EU.A), so that we could perform real-time data acquisition and visualization of both temperature profile and accomplished lethality curve during the entire process. The new system proved to be comparable with the OORIC data acquisition system in use at the Thermobacterilogy laboratory of the Food Engineering School at UNICAMP. Besides being comparable, it showed to be more versatile as well / Mestrado / Mestre em Engenharia de Alimentos

No description available.

No description available.

La phase de démarrage est une étape critique pour les opérateurs de digesteurs anaérobies en raison des instabilités rencontrées, qui peuvent entraîner une réduction du flux de déchets à traiter, voire un arrêt complet du digesteur, nécessitant un redémarrage. Ces problèmes entraînent des pertes économiques importantes dues à l'accumulation de déchets organiques et à des coûts d'exploitation non optimaux de l'usine de biogaz (chauffage, moteurs, etc.). Ce projet de thèse vise à développer des approches pour comprendre les dynamiques de la phase de démarrage afin de permettre une application efficace, l'élaboration de protocoles standardisés et une prise de décision éclairée. Pour ce faire, la thèse a été divisée en deux phases : 1) la Phase 1 est conçue pour définir les principes fondamentaux menant à des phases de démarrage réussies, et 2) la Phase 2 vise à établir des lignes directrices pour la prise de décision afin d'optimiser la sélection et la mise en œuvre de l'inoculum dans les digesteurs anaérobies. Dans la phase 1, les principes de base de la digestion anaérobie sont définis au chapitre 1, suivis des particularités de la phase de démarrage au chapitre 2. Ainsi, les défis uniques liés à la phase de démarrage ont été soulignés et l'importance de l'inoculum au cours de la planification du démarrage a été mise en évidence. La plupart des données expérimentales examinées sont ensuite normalisées et compilées dans une base de données au chapitre 3. Cela a permis d'appliquer des méthodes statistiques et d'apprentissage automatique pour saisir les corrélations cachées entre les paramètres étudiés et la durée du démarrage. La base de données développée comprenait 90 % de cas de démarrage réussis et 10 % de cas d'échec, résumant efficacement les conditions conduisant à des cas de démarrage réussis. L'étude de l'effet des paramètres d'exploitation sur la durée de la phase de démarrage a mis en évidence l'importance du contrôle du régime d'alimentation pendant la phase de démarrage en maintenant un temps de rétention hydraulique (TRH) suffisamment bas, même si le contenu organique de substrat est élevé, grâce à une gestion optimale du taux de charge organique (TCO) sans dilution de substrat. Cependant, la base de données développée n'a pas pu capturer l'impact de l'inoculum sur la durée de démarrage en raison du manque de détails rapportés dans la littérature. Par conséquent, la deuxième phase se concentre sur le développement de lignes directrices pour une meilleure sélection et mise en œuvre de l'inoculum dans les digesteurs anaérobies. Dans un premier temps, les indicateurs pertinents de la performance de l'inoculum liés à la température et aux caractéristiques sont identifiés au chapitre 4 par le biais d'expériences en batch. Les concentrations initiales d'ammoniac total, d'alcalinité, d'acides gras volatiles (AGV) et la température de l'inoculum sont les principales caractéristiques reflétant la diversité bactérienne et la cinétique de l'inoculum. Le chapitre 5 étudie l'impact de la stabilisation de l'inoculum et de la réponse à des événements d'alimentation extrêmes sur la phase de démarrage au cours d'une expérience semi-continue de 115 jours. Lors d'événements d'alimentation extrêmes, le digestat thermophile stabilisé a montré une plus grande sensibilité aux charges organiques élevées, alors que le réacteur ensemencé avec du digestat thermophile frais a été plus influencé par la suspension de l'alimentation. Une fois que les deux digesteurs ont atteint le régime permanent, des différences mineures ont été observées entre les deux réacteurs. Enfin, le chapitre 6 met en évidence l'effet du choix de l'inoculum sur les performances économiques des systèmes de digestion anaérobie. Il a été constaté que la matière digérée (digestat) utilisée comme inoculum conduit à une meilleure performance des réacteurs pendant la phase de démarrage que la matière non digérée (boues municipales). En effet, le coût de production d'un mètre cube de méthane et de traitement d'un kilogramme de solides volatiles (SV) dans le digesteur inoculé avec du digestat était respectivement 18 et trois fois moins élevé que dans le digesteur inoculé avec des boues municipales diluées. Tout au long de ce projet de doctorat, les fondamentaux et les principes de la phase de démarrage sont présentés, suivis par le développement de lignes directrices pour la prise de décision concernant la sélection et la mise en œuvre de l'inoculum. / The start-up phase is a critical step for anaerobic digester operators due to the instabilities encountered, which can lead to a reduced flow of waste to be treated or even a complete shutdown of the digester, necessitating a restart. These issues result in significant economic losses due to the accumulation of organic waste and non-optimal operating costs of the biogas plant (heating, motors, etc.). This PhD project aims to develop knowledge-based approaches to understand the dynamics of the start-up phase for efficient application, the development of standardized protocols, and informed decision-making. To achieve this, the thesis has been divided into two phases: 1) the first phase is designed to define the fundamentals leading to successful start-up phases, then 2) the second phase targets the establishment of decision-making guidelines for optimizing the selection and implementation of the inoculum in anaerobic digesters. In Phase 1, the basic principles of anaerobic digestion are defined in Chapter 1, followed by the particularities of the start-up phase in Chapter 2. As such, the unique challenges related to the start-up phase have been outlined and the importance of the inoculum during the start-up planning has been highlighted. Most of the reviewed experimental data are then standardized and compiled into a database in Chapter 3. This allowed for applying statistical and machine learning methods to capture hidden correlations between the studied parameters and start-up duration. The developed database included 90 % successful and 10 % failed start-up cases, effectively summarizing conditions leading to successful start-up cases. The study of the impact of operating parameters on the start-up phase duration highlighted the importance of controlling the feeding regime during the start-up phase by maintaining sufficiently low hydraulic retention time (HRT), even if the organic content of the feed is high, through optimal organic loading rate (OLR) handling without feed dilution. However, the developed database couldn't capture the impact of the inoculum on the start-up duration due to insufficient reported details in the literature. Therefore, the second phase focuses on developing guidelines for the enhanced selection and implementation of the inoculum in anaerobic digesters. Initially, relevant indicators of inoculum performance related to temperature and characteristics are identified in Chapter 4 through batch experiments. Initial concentrations of total ammonia, alkalinity, VFA, and inoculum temperature were the main characteristics reflecting inoculum bacterial diversity and kinetics. Chapter 5 studies the impact of inoculum stabilization and response to extreme feeding events on the start-up phase during a 115-day semi-continuous experiment. Under extreme feeding events, the stabilized thermophilic digestate showed greater sensitivity to high organic loads, whereas the reactor seeded with fresh thermophilic digestate was more affected by feed suspension. Once both digesters reached the steady-state, minor differences were observed between the two reactors. Finally, Chapter 6 highlights the impact of inoculum choice on the economic performance of anaerobic digestion systems. It has been found that digested material (digestate) used as inoculum leads to a better performance of reactors during the start-up phase than undigested material (municipal sludge). Indeed, the cost of producing one cubic meter of methane and processing one kilogram of volatile solids (VS) in the digester inoculated with digestate was 18 and three times cheaper, respectively, than in the digester inoculated with diluted municipal sludge. Throughout this PhD project, the fundamentals and principles of the start-up phase are presented, followed by the development of decision-making guidelines for inoculum selection and implementation.

Th. doct.-ing.--Sci. des matér.--Toulouse--I.N.P., 1983. N°: 263.

A high-temperature, high-pressure autoclave has been adapted for use with a high-field nuclear magnetic resonance spectrometer. The autoclave has an internal pressure range of atmospheric pressure to 1.5 kbar and a temperature range of 273 K to 1900 K. The autoclave is usable in a high field (8 T) magnet with a room temperature, 76.4 mm bore. The autoclave was tested using assorted nuclear species with resonant frequencies ranging from 57 to 70 MHz at pressures ranging from atmospheric pressure to 1220 bar and temperatures ranging from 273 K to 448 K. Previously, the autoclave was used in conjunction with an iron magnet at temperatures to 1900 K and pressures to 1.5 kbar. / Graduation date: 1997