The hydraulic conductivity of sand-bentonite mixtures (SBMs) was measured to investigate the effects of mixing method, uniformity, and hydration of the mixtures. Triaxial tests were completed to determine the hydraulic conductivity of each specimen. Specimens using Ottawa sand and Wyoming bentonite, prepared with dry and suspension mixing conditions that altered the degree of hydration and swelling of bentonite, had varying bentonite content by percentage dry weight of sand. The conclusions of this experiment can be applied to the construction of cut off walls used in levees to mitigate groundwater seepage through underlying pervious layers. Eleven sand-bentonite specimens were tested in this study: nine were prepared using dry mixing and two were prepared using suspension mixing. The results do not show strong correlations between hydraulic conductivity and bentonite content, mixing method, clay void ratio, or time. Therefore, further investigation of the results was necessary. The bentonite void ratio (clay void ratio) assumes that bentonite is fully swelled for both blocked and partially blocked flow. Blocked flow occurs when the swelled bentonite blocks all the sand voids, forcing the water to flow within the bentonite voids. However, the results in this study shows that the concept of clay void ratio doesn’t capture the performance of SBMs when the bentonite is partially swelled; therefore, a new concept of effective clay void ratio was introduced to account for bentonite partial swelling. The effective clay void ratio determines the volume of swelled clay as a function of the volume of fully swelled bentonite. This is useful when comparing results with literature or predicting hydraulic conductivity in cases where only partial swelling of bentonite is expected. / text

The use of lime for soil stabilization has greatly increased since the second half of the 20th century. A lot of research has been conducted to understand the mechanisms of stabilization. These mechanisms are caused by pozzolanic reactions between lime and clay minerals. However, it has not yet been possible to quantify the factors affecting the evolution of these reactions. The variety of soils and the disruptive elements do not allow comparing these soils with each other and giving any quantitative and generalized conclusions in terms of mechanical improvement. The goal of this study is to build a progressive understanding of this phenomenon by avoiding any disruptive elements and controlling most of the parameters. Consequently, the choice is made to study a synthetic soil by controlling its particle size distribution and composition. This soil is a mixture of sand and bentonite taken at different compositions. An analysis of the evolution of lime treatment on such model soils is carried out on two scales: the macroscopic scale and the microscopic scale. The macroscopic scale studies the evolution through unconfined compressive strength, lime consumption, electrical resistivity as well as complementary studies such as sonic and triaxial tests. Results from macroscopic tests show that sand takes an important part into soil stabilization, meaning that a soil containing a lot of clay does not necessary give the best long term mechanical characteristics. Tests that allow a much more detailed comprehension of stabilization are also presented. Microscopic evolution is studied through X-Ray Computed Tomography and Mercury Intrusion Porosimetry. A study on tomographic image treatment has also been carried out to segment the images from its different constituents. / L'utilisation de la chaux pour stabiliser le sol s'est considérablement déve-lop-pée depuis la seconde moitié du 20e siècle. De nombreuses recherches ont vu le jour pour comprendre les mécanis-mes de cette stabilisation. Ces mé-ca-nis-mes sont causés par les réactions pouzzolaniques entre la chaux et les minéraux argileux. Cependant, il n'a pas encore été possible de quantifier totalement les facteurs influençant le déroulement de ces réactions. La diversité des sols et la présence d'éléments perturbateurs ne permettent pas de les comparer et d'en tirer des conclusions quantitatives et généralisables en termes d'amélioration des paramètres mécaniques. Cette étude envisage donc de construire une compréhension progressive du phénomène en ne prenant pas en compte les éléments perturbateurs et en contrôlant à priori un maximum de paramètres. Pour ce faire, il est proposé d'étudier un sol synthétique dont la granulométrie et la composition peuvent être contrôlées. Ce sol est un mélange de sable et de bentonite pris à différentes compositions. Une analyse sur l'évolution du traitement à la chaux est alors effectuée sur ces mélanges sur deux échelles :l'échelle macroscopique et l'échelle microscopique. L'échelle macroscopique envisage des essais de compression simple, de consommation de chaux, de résistivité électrique ainsi que des essais complémentaires tels que les essais soniques et triaxiaux. Les résultats macroscopiques montrent que le sable joue un rôle important dans la stabilisation, le sol présentant la fraction argileuse la plus importante n'ayant pas les meilleures caractéristiques mécaniques à long terme. Les tests permettant une compréhension plus fine de cette stabilisation sont également présentés. L'échelle microscopique est étudiée via la tomographie aux rayons-X et la porosimétrie au mercure. Une étude sur le traitement des images tomographiques est aussi mise en oeuvre pour segmenter de manière adéquate les images de ses différents constituants. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished