M.Sc.

Le petit bassin minier de Ronchamp Champagney, situé dans les Vosges saônoises à environ soixante-dix kilomètres au Sud de Mulhouse, a été exploité de 1750 à 1958. Ce mémoire retrace l’histoire de l’exploitation charbonnière de 1812 à 1867, période de basculement technique au moment de la première révolution industrielle. Le mode d’extraction change de technologie et d’échelle pour stabiliser à la fois ses structures techniques et son espace d’exploitation. L’analyse prend le parti d’un double regard celui des enjeux de fond et des contraintes de surface, appréhendés principalement à partir des procès-verbaux des délibérations des conseils d’administration et des assemblées générales. Cette histoire reste liée à la vision entrepreneuriale de quelques notables locaux et à la persistance de relations familiales et personnelles. Ignorée par une métallurgie comtoise agonisante, paradoxalement la concession houillère contribue fortement au processus d’industrialisation alsacien, fournissant du charbon aux chaudières des machines à vapeur. L’histoire du bassin minier de Ronchamp, bien qu’enclavé dans son territoire, atteste néanmoins la réalité d’une modernité technique, et la quête permanente d’une rationalité productive. / The small-sized coal field of Ronchamp Champagney, in the Saône Vosges, about 44 miles south of Mulhouse was exploited from 1750 until 1958. This dissertation recounts the history of coal mining there from 1812 until 1867, a period of deep technical changes during the first industrial revolution. The technology and scale of coal extraction changed, which stabilized both its technical structures and its exploitation area. The focus of this analysis is twofold: first the stakes down the pit, second the constraints at the pithead, as they can both be perceived through the minutes of board and general meetings mainly. This history is linked to the entrepreneurial visions of a few local notables as well as to enduring family or personal ties. Ignored by the decaying steel industry of Franche-Comté, the coal concession contributed to the industrialization process in Alsace by feeding the boilers of steam engines with coal. The history of the coal field of Ronchamp – a landlocked territory - testifies to a real technical modernity and a constant search for productive rationality.

This thesis presents an engineering geological and geotechnical investigation of the proposed Terrace Opencast Coalmine highwall in the Reddale Valley, Reefton. The proposed pit will target the 4-11m thick No. 4 Seam coal, which exists on the Valley floor beneath outwash gravels and Brunner Coal Measures (BCM) overburden dipping at 15-30° to the northwest. Rock coatings are providing friable sandstone units with protection from weathering in existing cut faces and may contribute to short term pit wall stability. The BCM core was divided into four geotechnical units for rock material testing purposes: unit 1 siltstone, unit 2 carbonaceous mudstone, unit 3 interbedded sandstone and carbonaceous mudstone and unit 4 loose sandstone. The average results for units 1-3 gave classifications within the medium to high porosity (9-13%) and dry density (2250-2470kg/m³) ranges, and medium to medium high slake-durability Id2 values (72-94% retained). Unit 4 (loose sandstone) recorded very low dry density (1694 kg/m³) and slake-durability Id2 (9%) average values. Strength testing confirmed that the units can be classed as weak rocks, with average UCS values of 12.8-13.7MPa for units 1-3, and for all four units Is(50) from point load testing of 0.26-0.62MPa with low cohesion values (0-6.2MPa) from triaxial testing. Friction angles from triaxial tests gave high values of 32-45°, while direct shear tests established 15° internal friction for bedding planes in carbonaceous mudstone and 37° for a high angle defect in interbedded sandstone/carbonaceous mudstone. The average Young's modulus values ranged from 0.82 to 10GPa, and Poisson's ratio between 0.39 and 0.50. Eight scanline defect surveys established that the major discontinuities in existing cut faces consist of high angle tension joints, shallow dipping bedding, and faults related to regional uplift. The defect orientations from the scanlines located in the southwest were significantly different from those in the northeast, and may be due to the faults that cross the Valley. In general the majority of defects displayed low persistence (less than 3m), were clean and tight, and had low joint roughness coefficients (JRC less than 6). Joint wall compressive strengths gave an average of 32MPa, but were affected by case-hardening on weathered faces. The results from the 8 drill holes analysed show that 37% of core was within the excellent rock quality designation class (RQD = 90-100%), while 29% was in the very poor quality rock (RQD = 0-25%). A semi-confined aquifer in the outwash gravels that will drain into the proposed pit was found to have a transmissivity of 58m²/day and hydraulic conductivity of 3.1 x 10⁻⁵ m/s. Kinematic feasibility assessment determined an optimum highwall orientation of 65° dip to 120° (dip direction), which is within at least 20° of the coal seam strike. The likelihood of planar, wedge or toppling failure depends on whether the structural conditions are similar to those encountered in the southwest or northeast scanlines, as well as the persistence of the defects present. The occurrence of small scale (less than 1m offset) 'step-up' normal faults, and the three larger faults that cross the valley, all of which are related to regional uplift, will also affect which failure mode will be kinematically feasible. Other crucial slope stability considerations include groundwater inflow from the saturated overburden and bedding parallel failures on the footwall dip slope of the pit. An investigation into case hardening on existing cut faces identified three interconnected rock coatings: iron films, lithobiontic (biological) and clay-dominated crusts. Jarosite was found at sites with abundant pyrite and the oxidation of iron may have been aided by microbial activity. A green algae inhabiting pore spaces approximately 1mm below the surface was noted beneath an iron film and it is suggested to be similar to that found in arid environments. Although lithobiontic and clay-dominated crusts did not provide the weathered surface with any additional strength, they were observed to form relatively quickly (from months to less than 5 years) and will aid short term stability by providing the batters with protection from weathering processes. This project concluded that the overburden material in the proposed highwall can be expected to behave like weak rock and in some cases (such as the loose sandstone) can be expected to have soil characteristics. Highwall stability is more likely to be affected by substantial inflows of groundwater than highly persistent joint sets. Establishment of the highwalls in their final position in the early mining stages will enable development of rock coatings that are expected to aid short and long term stability.

Submitted by Mariana Dornelles Vargas (marianadv) on 2015-04-27T12:20:51Z No. of bitstreams: 1 redes_neurais.pdf: 847349 bytes, checksum: fb1a76a712ea4ee0f08c3e3f3e243fa7 (MD5) / Made available in DSpace on 2015-04-27T12:20:51Z (GMT). No. of bitstreams: 1 redes_neurais.pdf: 847349 bytes, checksum: fb1a76a712ea4ee0f08c3e3f3e243fa7 (MD5) Previous issue date: 2009 / PROSUP - Programa de Suporte à Pós-Gradução de Instituições de Ensino Particulares / As técnicas de perfilagem geofísica e de testemunhagem, utilizadas na identificação de litologias, representam alto custo financeiro e envolvem uma quantidade considerável de tempo por parte de um especialista. Nesse sentido, este trabalho propõe a modelagem e aplicação de um método alternativo de classificação litológica, através de Redes Neurais Artificiais (RNAs), para auxiliar no processo de interpretação de dados geofísicos. A área de estudo da aplicação é a Jazida do Leão, localizada em sua grande parte nos municípios de Rio Pardo, Minas do Leão e Butiá (RS). O conjunto de treinamento e de validação da RNA contém informações de oito furos de sonda provenientes das Formações Palermo e Rio Bonito. As variáveis de entrada incluem dados de profundidade e informações geofísicas de perfis raios gama, potencial espontâneo, resistência e resistividade. Para todos os experimentos, as litologias a serem classificadas foram: arenito, siltito e carvão. O modelo de rede neural utilizado foi o de múltiplas camadas (MLP) alimentadas adiante (feedforward). As redes foram treinadas com o algoritmo de retropropagação de Levenberg- Marquardt e Resilient Backpropagation. Obteve-se uma taxa de acertos de aproximadamente 80% na classificação. / The techniques of geophysical logging and core descriptions, used on lithology identification, represent a high financial cost and involve a considerable amount of time from a specialist. On that direction, the main objective of this research is to propose an alternative method of lithological classification, through Artificial Neural Networks (ANNs), to assist the process of geophysical data interpretation. The study area is located in Leão coal field, where a major part of its territory is inside the municipalities of Rio Pardo, Minas do Leão and Butiá (RS). The set of ANN training and validation contain information of eight boreholes coming from Palermo and Rio Bonito formations. The input variables include depth data and geophysical information of gamma-ray profiles, spontaneous potential, resistance and resistivity. For all experiments, the lithologies to be classified were: sandstone, silt and coal. The neural network model utilized was feedforward multilayer perceptron (MPL). Networks were trained by Levenberg-Marquardt and Resilient backpropagation algorithms. A success rate of approximately 80% was obtained on classification.