ABSTRACT
Predictions of soil movements during construction of many geotechnical systems rely on accurate description of the response of soil at small-strain levels. In the laboratory the measurement of stiffness of soils in the small-strain domain can be achieved from static or dynamic laboratory tests. Static tests require precise control and measurement of very small stress and strain increments. Dynamic tests and the measurements of the body wave velocities within the soil element have been used since Shirley & Hampton (1978). Real challenges are present in data interpretation (Arroyo et al. (2006), among others) which are based on the theory of wave propagation and require either prior knowledge of a constitutive model for soil or manipulation of continuum medium theory assumptions. The measurement of body wave velocities is based on the generation of an elastic wave normally by piezoceramic bender/extender (B/E) elements, Lings & Greening (2001). A transmitting B/E element is placed on one end of a confined soil sample while a receiving B/E element is placed at the other end of
the sample Both shear and compressional wave velocities (and corresponding stiffnesses) can be determined non-destructively and relatively fast. Recent studies Lee & Santamarina (2005)), including numerical analysis of wave transmission (Arroyo (2001); Arroyo et al. (2006)), leave questions regarding the optimum properties of the input signal, identification of travel time, effect of boundary reflections. Accuracy of the interpreted bender stiffnesses is tied to the adequacy of the inversion model chosen, Arroyo et al. (2003). DEM is able to simulate both static and dynamic tests on the same sample and to resolve some of these issues.
