Effect of Sample Preparation on the Reliability of Large-Scale Physical Modeling in Geotechnical Systems: ACase Study

Preparing uniform soil samples is essential for obtaining reliable results in the physical modeling of geotechnical structures, such as retaining walls and piles. The air pluviation method is a common approach adopted by researchers and engineers to reconstitute large-scale sand beds with high degrees of homogeneity and uniformity. This paper investigates the uniformity of sand beds prepared using the conventional manual sand pouring method, its potential impact on test results, and how creating a more uniform sand bed can enhance the reliability of the results. To achieve this goal, a novel portable sand pluviator capable of reconstituting sand beds inside a frustum confining vessel (FCV) apparatus at a high level of uniformity is adopted. The pluviator is utilised to prepare sand beds inside the FCV, and the uniformity of the resulting sand bed is assessed and compared with those prepared using the manual pouring method. Next, in order to investigate the influence of uniformity of sand beds on the outcomes of a large-scale test, a large-scale experiment involving the determination of the bearing capacity of a open-ended single-helix helical pile within the FCV is considered. Numerical analysis investigates how sand bed uniformity affects bearing capacity without extensive testing. It uses data from initial research and employs Latin Hypercube Sampling to determine the required simulations for accurate measurement of the impact of sand bed uniformity on bearing capacity. The results of this set of numerical simulations are then thoroughly analysed to assess the impact of the sand pouring method on the reliability of the experimental outcomes. Analysis of the results show that preparation of uniform sand beds at the desired relative density is one of the key elements in obtaining accurate and reliable experimental results in geotechnical physical modelling problems.