Speaker
Description
For coarse-grained granular materials, manual specimen preparation is inherently associated with variability, particularly with respect to the density. This effect is expected to be even more pronounced for large-scale specimens, where the larger volume and increased handling requirements can further amplify density variations and substantially increase labor intensity.
The Robotic Automated Specimen Preparation (RASP) system enables an automated specimen preparation with the air pluviation out of a single nozzle technique using a programmable robotic arm, allowing controlled adjustment of fall height, deposition rate and spatial trajectory. By prescribing layer-wise motion paths and monitoring the deposited mass, both global and local void ratios can be reproducibly achieved. Previous investigations have already demonstrated a high degree of homogeneity for small-scale specimens prepared using RASP.
In the present study, the RASP air-pluviation concept was extended to substantially larger specimen diameters using Karlsruhe Sand. Cylindrical specimens with a diameter of 50 cm were prepared by robotic deposition. Stacked rings were used exclusively as a geometric control system to determine vertical density uniformity and to quantify potential density gradients. The experimental results showed that homogeneous large-scale samples with predefined target density can be reliably generated using robotic deposition alone. The measured density profiles confirmed that the spatial variability remained within the same order of magnitude as previously attained for small-scale specimens.
A mechanical comparison between homogeneous small and large specimens prepared to the same target density using the RASP method was subsequently performed. One-dimensional compression tests on Karlsruhe Sand were conducted on conventional oedometer specimens (75 mm diameter) and compared with large-scale specimens (500 mm diameter). Compressibility, constrained modulus evolution and stress-dependent stiffness were evaluated to identify scale-dependent behavior.
The study demonstrated that the RASP method can be successfully implemented for large-scale specimens, reduce labor effort and provides a reproducible experimental framework for systematically investigating intrinsic specimen size effects in granular soils by decoupling preparation quality from mechanical testing.