Characterization of Unsaturated Soils Using Acoustic Techniques

Recently there has been a great interest in the ability to relate the hydro-mechanical properties of soils to their acoustic response. This ability could enhance high resolution non-destructive evaluation of the shallow subsurface, and would have applications in a variety of fields including groundwater and contaminant hydrogeology, oil recovery, soil dynamics, and the detection of buried objects. Groundwater hydrologists and environmental engineers are challenged with the task of characterizing the material, mechanical and hydraulic properties of the subsurface with limited information generally collected from discrete points. Geophysical testing offers a suite of measurement techniques that allow for the non destructive evaluation of potentially large areas in a continuous manner. Acoustic testing is one geophysical method used by many professions to characterize the subsurface. Unsaturated and multiphase flow modeling relies on the relationship between the capillary pressure and the level of saturation of the porous media. It has been previously suggested that this relationship may be non-unique and rate dependent. A theory which relates this dynamic relationship to the acoustic properties of the soil was developed by others. This research attempts to experimentally verify this theory by meeting the following three objectives: (1) develop an apparatus and procedure to collect acoustic waveforms on laboratory sized unsaturated soil samples, (2) develop a forward modeling technique using a one-dimensional wave propagation model as an alternative analysis method for waves collected on relatively small laboratory specimens, and (3) apply the theory to the measured acoustic data in an attempt to predict the dynamic behavior of the capillary pressure relationship. The acoustic data collected showed variation in compressional wave velocity and attenuation with saturation, and the trends were consistent with data collected by others in partially saturated rocks. The forward modeling technique was shown to provide objective results with reasonable accuracy and low computational time. The dynamic effects predicted with these acoustic measurements did not sufficiently explain the dynamic behavior seen in the laboratory. This is attributed to other causes of significant attenuation not accounted for in the wave propagation theory that was evaluated.