In order to estimate accumulated excess pore pressures in the soil around a cyclically loaded (offshore) foundation structure, cyclic laboratory tests are required. In practice, the cyclic direct simple shear (DSS) test is often used. From numerous undrained tests (or alternatively tests under constant-volume condition) under varying stress conditions, contour diagrams can be derived, which characterize the soil's behavior under arbitrary cyclic loading conditions. Such contour diagrams can then be used as input for finite element models predicting the load-bearing behavior of foundation structures under undrained or partially drained cyclic loading. The paper deals with the general behavior of a poorly graded medium sand in cyclic DSS tests under undrained loading conditions. The main objective of the research was to investigate and parametrize the soil's behavior and to identify possible effects of sample preparation. Numerous tests with varying cyclic stress ratios (CSR) and mean stress ratios (MSR) have been conducted. Also the relative density of the sand was varied. A new set of equations for a relatively easy handable mathematical description of the resulting contour plots was developed and parametrized. In the original tests, the sand was poured into the testing frame and carefully compacted to the desired relative density by tamping. In offshore practice, a preconditioning of a soil sample is usually realised by cyclic preshearing with a certain CSR-value or additionally by preconsolidation under drained conditions. By that, a more realistic initial state of the soil shall be achieved. In order to investigate the effect of such a preconditioning on the resulting contour diagrams, additional tests were conducted in which preshearing and preconsolidation was applied and the results were compared to the test results without any preconditioning. The results clearly show a significant effect of preshearing and an even more pronounced effect of preconsolidation for the considered poorly graded medium sand.

Despite several parameters having been identified as having an impact on the undrained monotonic response of granular soils, the impact of the overconsolidation ratio (OCR) is still a contentious issue. One of the significant reasons for the inconsistencies in the undrained behavior is the method by which the stresses are applied--specifically, the effective preconsolidation and confining pressures. To address this, two separate series of triaxial compression tests were realized in order to examine and compare the influence of the OCR (OCR = 1, 2, 4, and 8) on the mechanical response of Chlef River (Algeria) sand, considering the way the stress state was applied. During the first series, the OCR was accomplished by consolidating the specimens to an effective preconsolidation pressure (sigma p ' = 100, 200, 400, and 800 kPa) and subsequently unloading them to a constant desired effective confining pressure of 100 kPa. In the second series, all specimens were consolidated to a maximum effective preconsolidation pressure of sigma p ' = 800 kPa (constant effective preconsolidation pressure) and then unloaded to different effective confining pressures (sigma c ' = 800, 400, 200, and 100 kPa), using two different sample preparation techniques--dry funnel pluviation and moist tamping. The test results revealed a suitable increase in the shear strength with an increase in OCR in the first series, with the opposite trend observed in the pore water pressure. For the second series, an increase in the OCR parameter resulted in a minimized shear strength and pore water pressure (although the trend in pore water pressure evolution did not really reflect the behavior of the deviator stress for this series). In addition, certain parameters, such as normalized behaviors, the brittleness index, ratio of excess pore water pressure to deviator stress at the critical state, and flow potential, appear to be reliable predictors for clarifying and, consequently, explaining the studied behaviors.