Statistical Characteristics of a Natural Marine Sand under Constant Volume Cyclic Direct Simple Shear

Dynamic loads induced by environmental forces and seismic events significantly influence the performance of foundation systems throughout their lifespan. However, the dynamic resistance of soil is usually estimated from a limited number of laboratory tests, introducing uncertainty into design parameters. For example, the uncertainty can hinder the definition of reliable safety margins and lead to errors inherent in numerical simulations and constitutive models. To quantify the uncertainty and to evaluate the potential risks in design, a series of cyclic direct simple shear (CDSS) tests was performed on a sandy soil. Eight combinations of consolidation stress, relative density, and cyclic stress ratio are examined, with each combination being tested 30 times under stress-controlled, constant volume (equivalent undrained) conditions. The resulting data set provides robust statistical insight into the uncertainty and variability characteristics of the cyclic resistance of the soil. Clear distinctions in strain accumulation behavior and hysteresis response are observed between loose and dense samples. Variability analyses of (apparent) excess pore pressure, shear modulus, and damping ratio are performed. The variability of excess pore pressure increased with its magnitude, while the trend of pore pressure development remains relatively stable. The variability in shear modulus is observed to be sensitive to consolidation stress, while the damping ratio remains stable in all testing conditions. The number of cycles to reach a 5% strain threshold follows an approximately normal distribution, with a standard deviation around 20% of the mean. Applications based on probabilistic cyclic resistance are presented according to the results. This statistical feature enables a probabilistic framework for cyclic design, including confidence-based correction of design parameters. The findings improve understanding of test repeatability, highlight the degree of uncertainty in CDSS data, and support more reliable application of CDSS results in geotechnical design.