Simple shear testing is used to study a number of practical geotechnical problems including: soil conditions directly below a loaded surface, adjacent to a driven pile shaft, soils deposited on a slightly inclined slope, and most notably the response of soils subjected to earthquake-type loading. While each of these problems still have important questions to be answered, earthquakes and earthquake triggered geohazards are the most complex and also pose the highest risk. An important aspect of assessing the risk associated with earthquakes is the need to accurately predict soil behavior. True field loading conditions involve multidirectional shearing and the rotation of principal planes and are much more complex than the triaxial laboratory testing methods and models often used to describe them. Simple shear testing allows for the in situ conditions to be replicated; however, several limitations of the device make data interpretation difficult. The inability to apply complementary shear stresses and the inability to measure the horizontal normal stresses results in non-uniform stresses across the boundaries, as well as an undefined stress state during shearing. This, in turn, requires assumptions to be made about the failure conditions before any state parameters can be determined. Even when only monotonic testing is conducted, there are still many important questions to be answered about the actual severity of the non-uniform stresses on the boundaries, as well as the internal stresses and the microscopic response of granular soils. Discrete element method (DEM) modeling has the advantage of being able to examine particle-to-particle interactions. Once validated with the measured laboratory data, these models provide a vast quantity of information about the fundamental mechanisms underlying the observed complexity of the response of the soil mass as a whole. The goal of this research is to gain insight into the particle-to-particle interactions driving the overall response of granular samples subjected to multi-directional cyclic simple shear conditions. The main objectives of this proposed project are to (1) characterize the macroscopic response of metal ballotini representing idealized sand under simple shear loading conditions and (2) model the physical element tests using DEM simulations to gain insight into the microscopic response of the gran℗Ưular material. Findings from this study showed that the DEM simulations could be successfully validated by laboratory data and that the overall trends observed agreed reasonably well with the experimental data from this study, as well as previous studies by other researchers. Analyses showed that density not only influences shear strength of a sample, it also affects the angle of shearing resistance, the magnitude of principal stress rotation, the angle of non-co axiality, and the orientation of the principal fabrics for strains below those needed to reach critical state. Vertical effective stress was instead shown to have very little influence on these parameters. The initial fabric appears to play the largest role in the behavior of samples tested at different vertical stresses. The simulations also showed the non-coaxial behavior of the granular samples in terms of principal stress and strain rate orientations, as well as particle displacements. A number of other sensitivity studies were conducted to examine the influence of the model simplifications on the observed response. Several of these simplifications were shown to affect the shear strength obtained and should be included in future analyses. The electronic version of this dissertation is accessible from http://hdl.handle.net/1969.1/151646