|Zusammenfassung||Rock slides and avalanches, and gravitational mass movements in general, are not well understood geological phenomena, the simulation of which is difficult because of the complexity and multitude of acting processes. After mechanical failure of a rock slope, the rocks are transported by gravity and emplaced somewhere downslope. How that transport occurs, and how far the material is transported is determined by numerous processes, working simultaneously across a large range of scales, making prediction of the system behavior difficult. One of these processes is fragmentation, which by breaking the material apart can change the dynamic properties of the rock mass by transforming the intact rock mass into a flowing granular material, as well as potentially changing the effective basal friction. Despite this, fragmentation is rarely considered in studies of gravitational rock movement.
Previous studies on fragmentation in gravitational rock movements have mainly considered the final deposits. Thus, little is known of the conditions under which the fragmentation occurred, and how it affected the movement. Observations of these systems as they occur in nature, at sufficient temporal and spatial resolution are hard to come by, and modeling serves as an important tool to constrain the physics of their transport. Existing models rarely take the process of fragmentation into account, assuming instead that the movement can be approximated with constant material properties throughout the transport event.
Here, a new experimental approach is presented where the fragmentation of a rock analogue material during transport down a chute is considered at lab-scale.
The analogue material used in the models is produced by cementing sand with either gypsum or potato starch. The cement provides the material with a cohesion, which if loaded above a failure criterion is lost. Material testing using a ring shear tester and a triaxial tester reveal that the cohesion of the material is determined by the amount and type of cement added, such that the cohesion can be controlled from $3.5to 360 kPa.
Rock slides and avalanches are modeled by releasing a cuboid-shaped block of the analogue material down a 1 m slope, at an angle of 45 degrees. After accelerating down the slope, the blocks impact on a horizontal plate, on which they fragment, slide and come to rest. The models are scaled to nature by considering a set of characterizing dimensionless parameters derived from dimensional analysis. Of these parameters, the aspect ratio of the sliding material, the cohesion versus potential energy ratio and the basal coefficient of friction are studied in a parameter study.
The parameter study reveals that the degree of fragmentation increases the thicker the sample is compared to its length, or if its strength increases with respect to its initial potential energy. The relative contribution of these two parameters on the degree of fragmentation is considered through an analytical model of an elastic bending plate. Based on this model, a new parameter is suggested which can be used to predict the degree of fragmentation solely from initial conditions, taking into account the geometry, rock strength and potential energy.
The fragmentation is observed to consume energy, causing a reduced transport of the center of mass. An analysis based on the conservation of energy suggests that the energy consumed by fragmentation can be described by a logarithmic scaling law. Despite the loss of energy, fragmentation is observed to cause spreading of the material, leading to an increased travel distance of the front position of the deposits.
However, the front position is seen to depend on the degree of fragmentation in a complex manner, increasing for intermediate degrees of fragmentation but decreasing for higher ones. Careful observations of the analogue models suggest this behavior to be caused by a competition between spreading and increased internal friction. Because of this competition, an intermediately fragmented rock, i.e. a fragmenting strong rock, is expected to have a higher mobility than a highly fragmented one, i.e. a collapsing weak rock.
A comparison between the mobility as a function of the degree of fragmentation from the analogue model and a data set of natural rock avalanches reveal a remarkably good fit between model and nature and thereby the models applicability. This shows that fragmentation is a governing process in the transport of rock avalanches, and for gravitational rock movements in general.