Preparation of consistent sand densities around scaled models in physical modelling is a challenge. Manual sand-pouring techniques require a significant amount of time, are difficult to control, and yield unreliable results. The new device utilises a moving hopper and custom-designed tip allowing control over the sand discharge rate, drop height, traversing speed, sieve mesh size, and pouring spacing. To assess the impact of each of these factors on the relative density, extensive calibration tests were conducted. The calibration results are presented in this paper. Tests were also conducted on a pile with sand poured around the pile to demonstrate sample repeatability and uniformity. The findings revealed that sand discharge rate, drop height, and traversing speed are primary controls on relative density, while pouring spacing influences layer thickness uniformity. Both dense and loose samples can be reliably prepared by appropriate parameter selection or by removing sieves for loose deposition. Repeatability checks confirm that the automated system can produce uniform, reproducible samples suited for soilstructure interaction studies. This work contributes to improving the effectiveness and reliability of geotechnical testing by providing valuable insights into optimising automated sample preparation for physical modelling applications.
11th International Conference on Physical Modelling in Geotechnics (ICPMG2026)
Session 1: New facilities, new equipment, and measuring techniques