[pt] Apresenta-se, neste trabalho, resultados de um extenso estudo experimental de laboratório, cujos objetivos eram: obter informações do comportamento tensão-deformação em condições não drenadas, e verificar a susceptibilidade à liquefação de um resíduo oriundo da lavra do itabirito silicoso da Mina de Fernandinho, que situa-se no Quadrilátero Ferrífero (Minas Gerais ), região com grande concentração de minério de ferro. Para realizar o estudo de susceptibilidade à liquefação do resíduo, foi necessário projetar e construir um equipamento que é constituído de: uma célual triaxial, servo-motores, válvulas para regulagem de pressão, um microcomputador AT486, conversor D/A, um sistema de aquisição de dados da National Instruments e alguns acessórios que foram desenvolvidos para facilitar a execução dos ensaios. Durante a fase de projeto e montagem deste equipamento foi implementado um programa na linguagem de programação C para gerenciar os ensaios. Este equipamento possibilitou a execução dos ensaios triaxiais cíclicos e monotônicos com trajetórias de tensões servo controladas. Na primeira fase deste estudo foram realizados ensaios de caracterização completa, análise mineralógica e de microscopia eletrônica. Após a caracterização física do material, passou-se ao estudo do comportamento tensão- deformação e resistência com a execução de ensaios triaxiais. Nesta fase foram executados ensaios monotônicos e cíclicos em corpos de prova adensados isotropicamente e anisotropicamente. Com as análises dos resultados e entendimento do comportamento tensão deformação deste material em condições de laboratório, concluiu-se que este é susceptível à liquefação devido ao comportamento colapsível e desenvolvimento elevado de poropressões. Finalmente, propõe-se uma nova metodologia experimental para estudar os mecanismos que levam o solo a desenvolver o fenômeno de liquefação, tendo em vista que as metodologias apresentadas na literatura corrente não são adequadas para estudar este fenômeno. / [en] This work presents results of na experimental research programme executed in the laboratory at PUC-Rio, Brazil. The aim of this research was investigated the stress-strain behaviour and the evaluation of the susceptibility of fine grained tailing material from Fernandinho Mine to liquefaction. This mine is at Quadrilatero Ferrífero (Minas Gerais - Brazil), region with one of the largest sources of iron ore in Brazil.In order to study the susceptibility to liquefaction of tailing material an servocontrolled triaxial device was developed. This device were composed by: triaxial cell, servo-motors,pressure regulators, microcomputer, one digital analogy convert of, one analogy digital convert developed in the laboratory at PUC-Rio and other accessories. During the development of the device, one software to control all the trajetories was implemented. Using this device some cyclic and monotonic triaxial tests were carried out.Complete characterization, mineralogical and eletronic microscope analysis were carried out in the preliminary steps of this research. After this, an extensive programme of triaxial tests were carried out in order to determine the failure envelope, stress-strain behaviour and liquifiction resitance of the tailing material.A colapsive behavior of the tested material was observed at small strain and stress conditions. Another important observation was the high level of poropressure development,suggesting that this material is susceptible to liquefaction in special conditions.Finally, based on the test results, a new methodology is proposed to investigate soils when submitted to undrained conditions.

Actuellement la production d’huile pyrolytique (ou bio-huile) est destinée à en faire un carburant pour les moyens de transport. Cependant, le liquide issu de la pyrolyse est de piètre qualité, il est nécessaire de faire une opération d’hydrodéoxygénation (HDO), très coûteuse et énergivore, pour aboutir à un produit ayant les spécificités d’un carburant. Une des idées proposées, plus économique, consiste à faire de la bio-huile une source de biosyngas (CO+H2) ou biohydrogène renouvelables via du vaporeformage (VR). Ce projet de maitrise étudie le reformage à la vapeur d’eau de deux bio-huiles (MemU et WOU) sans apport externe de vapeur sur un nouveau catalyseur à base de nickel, Ni-UGSO, développé par le GRTP-C à partir du résidu minier UGSO. Les expériences de reformage ont été réalisées à pression atmosphérique, dans un réacteur différentiel et pour une durée de 500 min en faisant varier la température (750-850 °C) et la vélocité spatiale (WHSV= 1.7-7.1 g/gcat/h) en plus d’un test longue durée à 105h. Des tests supplémentaires ont été réalisés aussi avec un catalyseur commercial à titre de comparaison en plus d’un test de régénérabilité. La caractérisation du catalyseur s’est faite par DRX, MEB-FEG, BET et TPR. Les résultats des tests de VR de l’huile MemU entre 750 et 850 °C à WHSV ~1.8 g/gcat/h montrent une bonne production de biosyngas avec une concentration entre 90-95% et une sélectivité en H2 entre 80-95%. Le VR de l’huile WOU dans les mêmes conditions a donné moins de biosyngas et de H2 en raison de la teneur élevée en eau de l’huile. Le catalyseur est resté actif pendant toute la durée des tests, la DRX et la MEB ne montrent aucune trace de carbone. Cependant à WHSV > 6 g/gcat/h du carbone filamenteux sur le catalyseur a été observé par MEB après le VR de l’huile MemU mais pas après le VR de l’huile WOU. La DRX a permis aussi de montrer qu’après le VR des huiles, les oxydes de Fe et Ni qui constituent le catalyseur se réduisent et se combinent pour donner du Ni métallique et des alliages Ni-Fe. Le test BET indique que le catalyseur a une surface spécifique, après activation, de 10 m2/g. La TPR montre qu’il y a plus d’espèces oxydées sur le Ni-UGSO après le VR de la bio-huile WOU qu’après le VR de la bio-huile MemU, d’où les faibles rendements en H2/biosyngas. Les tests de VR réalisés avec le catalyseur commercial montrent des résultats similaires que ceux réalisés avec Ni-UGSO à faible WHSV. Cependant à WHSV élevée le catalyseur commercial a été plus résiliant et plus performant du fait de sa grande surface spécifique. Le test de régénérabilité montre que Ni-UGSO ne peut que partiellement être régénéré et sa structure initiale n’est pas retrouvée Les résultats positifs confirment que la production de biosyngas/biohydrogène par VR de bio-huiles est viable techniquement dans une bioraffinerie. Le procédé est plus économique que l’HDO. De plus, l’huile pyrolytique se trouve être une bonne matière première pour le reformage car on a un bon rendement en biosyngas (ou H2). Le catalyseur Ni-UGSO développé par le GRTP-C a montré des performances similaires que celles de catalyseurs actuellement sur le marché mais nécessite d’être encore optimisé. / Abstract : Currently the production of pyrolysis oil (or bio-oil) is intended to be transformed to transportation fuel. However, the produced liquid is of bad quality and it needs a hydrodeoxygenation (HDO) process which is very expensive and lot of energy is consumed to obtain a final product with the right fuel specifications. One of the ideas proposed, more economical, consists on producing renewable biosyngas (CO+H2) or biohydrogen from biooil by steam reforming (SR). This master project study the steam reforming of two bio-oils (MemU and WOU) without external steam addition over a new nickel based catalyst, Ni-UGSO, developed by the GRTP-C from the mining residue UGSO. The reforming tests were carried out at atmospheric pressure in a differential reactor during 500 min varying the temperature (750- 850 °C) and the weigh hourly space velocity (WHSV= 1.7-7.1 g/gcat/h), a long term test of 105h was also performed. In addition, Supplementary tests were done with a commercial catalyst in order of comparison plus one regenerability test. The catalyst characterization was done by XRD, FEG-SEM, BET and TPR. Test results of bio-oil MemU SR at 750-850 °C and WHSV ~ 1.8 g/gcat/h show a good production of biosyngas with a concentration range of 90-95% and a H2 selectivity of 80- 95%. The SR of bio-oil WOU in the same conditions resulted in less biosyngas and H2 produced because of high water content in the bio-oil. The catalyst was active for the whole duration of tests, XRD and SEM indicate that no carbon deposit was formed. However at WHSV > 6 g/gcat/h filamentous carbon was observed on the catalyst by SEM after the SR of bio-oil MemU but not after the SR of bio-oil WOU. The XRD showed also that after biooils SR Fe and Ni oxides that constitute the catalyst are reduced to metallic Ni and Ni-Fe alloys. BET test indicate that after activation the catalyst has a specific area of 10 m2 /g. TPR shows that more oxidized species are present in Ni-UGSO after bio-oil WOU SR than after bio-oil MemU SR which explains low H2/biosyngas yield. The tests of SR performed with the commercial catalyst show similar results as those performed with Ni-UGSO at low WHSV. However, at high WHSV the commercial catalyst was more resilient and better due to its high specific area. Regenerability test shows that NiUGSO is partially regenerated but its initial structure is not recovered. The positive results confirm that the production of biosyngas/biohydrogen from SR of biooils is technically viable for a biorefinery. The process is economically better than the HDO. The pyrolysis oil is a good feedstock for the reforming, we obtain an appreciable yield of biosyngas (or H2). The catalyst Ni-UGSO developed by the GRTP-C exhibits similar performances than commercial catalysts actually available in the market but needs more optimisation.