Assigned Session: WS 3.504: Numerical Modeling for Landslide Risk Assessment
Uncertainty of measurement and variability of direct shear parameters
Abstract ID: 3.13446 | Accepted as Talk | Talk | TBA | TBA
Gerald Innocent Otim (1)
Giulio Barbato (2), Irene Rocchi (1), Gianmario Sorrentino (1), Stefan Trapp (1), Alena Zhelezova (1)
(2) Politecnico di Torino University, Torino, 10129, Italy
The diverse and variable nature of ground conditions and soil properties presents a challenge when it comes to determining design parameters for a soil model. In these numerical analyses, one of the key properties to focus on is shear strength. It’s important to balance the number of data sets to achieve an accurate linear Mohr-Coulomb failure surface within budget and time constraints. Careful consideration of important parameters such as normal stress is necessary, and it’s recommended to use a set of three data sets in practice. However, this approach may still result in a notable amount of data variability. To evaluate this, the direct shear test was performed in accordance with ASTM D3080/D3080M-23 on loose sand of size 0.3-0.5 mm on a set of 6 normal loads (0 – 100 kPa) each with 10 repetitions. This sand was a simplified material that could be compared to alluvial fans in mountainous subsoils but equally cohesionless. By leveraging the Mohr-coulomb shear strength equation and applying an area correction to the data sets, the uncertainty in this measurement was determined using the guide to the expression of uncertainty in measurement approach. The relative expanded uncertainty obtained from normal stress ranged between 3.2 – 3.5 % while that from the shear stress was 3.9 %. When applying the shear strength equation, a combined relative expanded uncertainty of 5.3 % was observed. Furthermore, the uncertainty measurements for the angle of internal friction and cohesion were determined as 33 ± 10 and 7.5 ± 1.1 kPa respectively, with residuals exhibiting randomness. The three primary factors influencing the uncertainty in both the normal and shear stresses included reproducibility and bias on the measuring forces, and bias on the diameter of the specimen. Combinations within the data set were derived to investigate the variability in shear strength parameters. The variability in the angle of internal friction and cohesion were 34 ± 1.80 and 3.6 ± 2.0 kPa respectively. The exploration of the variability in shear strength parameters with sand demonstrated the impact on the long-term stability of a natural homogeneous slope when simulated in plaxis software.
N/A | ||||||||
|