A significant amount of research has been carried out to characterize expansive clay behavior from either microstructural or macrostructural perspectives; however, there exists a current gap in our knowledge about the basic mechanisms that relate one structural level to another. This research works to fill the gap between our understanding of expansive clay behavior at a microstructural (particle and fabric scale) level with our understanding at the bulk (engineering) scale.;The main objectives of the work are: 1) to qualitatively examine how clay particle fabric, stress paths, water potential, and mineralogy affect the translation of crystalline interlayer swelling to bulk volume change and swelling pressure; 2) to quantify these effects by defining curves describing constitutive surfaces for the behavior of smectite in the crystalline swelling regime under free swell and constant strain (zero volume change) boundary conditions; 3) to interpret the experimental observations in the context of existing conceptual models, including the porosity evolution model (Likos & Lu, 2006) and the Barcelona Basic Model (BBM) (Gens & Alonso, 1992); and (4) to develop and evaluate a new conceptual model for examining microstructural changes in fabric associated with crystalline interlayer swelling.;Three types of clay were selected for testing: Na-smectite, Ca-smectite, and a Ca-exchanged form of the Na-smectite. Results obtained include: SEM imaging of Na and Ca smectite, void ratio as a function of compaction pressure, water vapor sorption isotherms, axial deformation vs. relative humidity (RH) during wetting and drying for free swelling boundary conditions, and swelling pressure vs. RH for constant strain boundary conditions.;Three scale-level porosities were identified; this work focused on the interaction between the interlayer (eIL) and the inter-particle (eIP) void spaces. Sorption isotherms follow a wavy behavior, with the steeper portions of the curve corresponding to the one and two layer hydration states. Bulk volume change and swelling pressure response reflected this behavior, especially for RH corresponding to the 1-to-2 hydration layer transition. Initial density has a significant effect on bulk volume change and swelling pressure for RH equal or greater than that corresponding to the second hydration layer; interlayer swelling for denser specimens is more effectively upscaled to bulk response. During the first layer transition, most of the crystalline swelling was absorbed into the inter-particle void space. Clay mineralogy and exchange cation identity are also important variables. Ca-smectite exhibits larger bulk volume changes and swelling pressures than Na-smectite for tests under the same stress conditions. For the Ca-exchanged smectite, the cation affects sorption response but it does not affect the volume change response. Hysteresis in the volume change response is more evident for denser specimens, and is larger for Na-smectite. Stress history is also an important factor. The ratio of macrostructural strains to microstructural strains (epsilon PVM/epsilonvm) decreases with lower values of maximum past pressure (po) for all suction values during wetting; as suction decreases, the epsilonPVM/epsilon vm ratio is larger in magnitude, and it becomes more sensitive to p o.;With better knowledge of how microstructural swelling translates to macroscopic behavior and what mechanisms and variables are important, the geotechnical engineering community will be more equipped to approach and resolve the several problems involving expansive clays. Likewise, numerous applications in industry can use expansive clays more effectively if quantitative values provide a guideline of what combinations of normal stress, void ratio, and water content of a soil within the crystalline swelling regime have a significant effect in the soil performance.
展开▼
