Les échantillons obtenus des expériences aluminium-alumina ont été analysés à l'aide du microscope optique, du microscope électronique à balayage -la spectroscopie aux rayons X à dispersion d'énergie (MEB-EDX), du microprobe MEB, et de la diffraction des rayons X (DRX). Les résultats montrent que les réactions chimiques entre l'alpha-alumine pure à haute densité et les alliages Mg-Al liquides ne sont pas rapide; mais, la présence des impuretés (telle que Na2O en tant que phase de bêta-alumine) et la structure poreuse de l'alumine augmentent l'étendue des réactions significativement. La phase riche en Na2<3(bêta-alumine) qui se trouve dans toutes les alumines commerciales semble être l'un des facteurs les plus importants pour les réactions spontanées avec la vapeur de magnésium, même au temps de résidence le plus court. Mg-spinelle a été trouvé comme le produit de réaction le plus important. L'analyse thermodynamique indique la même tendance. SUMMARYThis project was undertaken to study the interactions between alumina particles, aluminum alloys, and its inclusions under liquid aluminum flow conditions. The objective was to develop a test method which can simulate the conditions similar to those in the aluminum filtration process and to evaluate the interactions taking place between various types of alumina samples, aluminum alloys, and its inclusions. With this test method, it was aimed to determine how various alumina types behave under flow conditions during the filtration process.Chemical interactions between alumina, aluminum alloys, and its inclusions were investigated under both static and dynamic flow conditions. In order to study these interactions under dynamic flow conditions, a knowledge of the velocity field in the vicinity of the alumina particles is necessary. In this project, two unique experimental systems which can simulate the flow condition of the industrial bed were designed and built. A mathematical model was also developed to predict the flow field around the particles in the experimental system. The mathematical model was validated by comparing the predictions with the results from a physical model in which water was used as the fluid.The mathematical model was then used to conduct parametric studies to determine the design and operational parameters for the actual experimental system in which the tests were carried out. This allowed the generation of a flow field similar to that of the industrial filter.The experiments with various liquid Mg-Al alloys (0, 2, 5, and 7 wt% Mg) were conducted for different residence times (from 6 hours to 168 hours) using the above VI experimental systems. The effects of the liquid aluminum alloy velocity, the temperature of the melt, the physical (apparent porosity, surface roughness, etc.) and chemical (impurity content such as Na2Û, SiÛ2 5 etc.) properties of alumina samples on the extent of aluminum

The purpose of this work is to evaluate the performance of several wear models, either with different mathematical formulation or different definition of the unknown wear coefficients, on the prediction of the work-roll wear amplitude in Hot Strip Mills (HSM). To achieve this goal, a classical model calibration approach based on inverse optimization has been developed to calibrate these several wear models. A large industrial hot rolling database composed by roll wear amplitude measurements for both later finishing mill stands (F6 and F7) from ArcelorMittal Dofasco HSM was considered and a least-square cost function was applied to minimize the differences between both numerical and experimental results during the optimization process. The averaged roll wear gap between measurements and optimized numerical predictions was then used as a quantitative indicator to compare the performance between the wear models and identify the most suitable one for roll wear prediction. In addition, an Artificial Neural Network (ANN) approach was developed based on the most suitable wear model. Thus, roll wear predictions obtained using the ANN were compared with the ones obtained using Classical calibration to evaluate the performance of both approaches.