No abstract

We propose a simple elastic-plastic contact model by considering the interaction of two spheres in the normal direction, for use in discrete element method (DEM) simulations of geomaterials. This model has been developed by using the finite element method (FEM) and nonlinear fitting methods, in the form of power-law relation of the dimensionless normal force and displacement. Only four parameters are needed for each loading-unloading contact process between two spheres, which are relevant to material properties evaluated by FEM simulations. Within the given range of material properties, those four parameters can be quickly accessed by interpolating the data appended or by regression functions supplied. Instead of the Von Mises (V-M) yield criterion, the Drucker–Prager (D-P) criterion is used to describe the yield behavior of contacting spheres in this model. The D-P criterion takes the effects of confining pressure, the intermediate principal stress, and strain rate into consideration; thus, this model can be used for DEM simulation of geomaterials as well as other granular materials with pressure sensitivity.

No abstract

Abstract. The Ms 8.0 Wenchuan earthquake in 2008 and Ms 7.0 Lushan earthquake in 2013 produced thousands of landslides in the southern region of the Longmen Mountains in China. We conducted field investigations and analyzed remote sensing data to determine the distribution law of earthquake-triggered landslides. The results show a strong negative power law relationship between the size and frequency of landslides in VII, VIII, and IX seismic intensity zones, a weak power law in the X seismic intensity zone, and a lognormal distribution in the XI seismic intensity zone. Landslide density increases with increasing seismic intensity. A sandpile cellular automata model was built under the conceptual framework of self-organized criticality theory to simulate earthquake-induced landslides. Data from the simulations demonstrate that, with increasing disturbance intensity, the dynamical mechanism of the sandpile model changes from a strong power law to a weak power law and then to a lognormal distribution. Results from shaking table experiments of a one-sided slope sandpile show that, for peak ground acceleration (PGA) in the range of 0.075–0.125 g, the relation between the amount and frequency of sand follows a negative power law. For PGA between 0.15 and 0.25 g, the relation obeys a lognormal distribution. This verifies that the abovementioned distribution of earthquake-induced landslides should be a universal law from a physical viewpoint and may apply to other areas. This new perspective may be used to guide the development of an inventory of earthquake-triggered landslides and provide a scientific basis for their prediction.