Resumen de Modelling fragmentation in rockfalls
The fragmentation process in rockfalls is a complex phenomenon that is not well understood and only a few rockfall simulation models consider it explicitly. Fragmentation significantly affects the evaluation of the hazard and therefore of the risk. This thesis aims to develop a rockfall propagation model that is capable of reproducing the fragmentation phenomenon in rockfalls and to assess its consequences in the risk analysis. Four real-scale tests in a quarry and one laboratory test were performed for a better understanding of the fragmentation process. During these tests, several remote sensing techniques were used to capture the motion of the blocks and the fragment size distributions of the resulting deposit. The analysis of the empirical data acquired confirmed that the mass distribution produced by the fragmentation of a single block can be adequately described using fractal theory. Moreover, it was observed that the envelope of the trajectories of the newly generated fragments adopted the shape of a cone.
The knowledge gathered with these experiments led to the development of RockGIS, a stochastic program based on a lumped mass approach for the numerical simulation of rockfalls and their fragmentation using a fractal model. The model simulates the trajectories of the blocks using state-of-the-art methodologies and implements an innovative fragmentation module to consider block breakage using fractal theory. The code was developed within the framework of the Rockmodels project (https://rockmodels.upc.edu). In the simulation, the parameters that define the sizes of the fragments generated are computed at each impact according to the kinematic conditions. This approach allows different fragmentation patterns to be reproduced depending on the energy conditions of the impacts.
The performance of the RockGIS code was verified and validated by the real-scale rockfall tests carried out and by reconstructing three inventoried natural rockfall events that took place in Spain: a 10,000 m³ rockfall near Vilanova de Banat (Eastern Pyrenees) in 2011, a 800 m³ rockfall in Monasterio de Piedra in 2017 (Zaragoza) and a 10 m³ rockfall on the Ma-10 road (Mallorca). For the calibration of the model different goodness-of-fit indicators were considered depending on the information available in each case study. Two main calibration criteria were used: the runout distance and the size distributions of all the fragments generated. Moreover, the fragment scattering along the slope, the number of blocks crossing a reference line, the position of the center of gravity of the whole deposit and other criteria were used in some scenarios to validate the simulation results. The parametric analysis showed that the model is highly sensitive to the parameters that control the fragmentation process.
The performance of the fragmentation model developed is satisfactory and accomplishes the goal of representing the fragmentation process, as it is able to reproduce the field observations accurately. To use this approach for risk analysis and the design of protective measures, precise calibration is required to ensure the parameters are appropriate for each case study considered. Regarding the risk analysis, fragmentation has both a significant and a contrasting effect on the risk value and should not be ignored. The most significant effect is on the rockfall runout distance. Fragmentation may significantly reduce rockfall propagation if the slope is sufficiently gentle and long. In this case, the new fragments generated mobilize less energy and can be trapped by the topographic irregularities, obstacles and protection works. Conversely, a wide range of block sizes are able to reach corridors running below steep slopes. In such a situation, fragmentation facilitates the divergence of the blocks’ trajectories, which increases the probability of impact on people and vehicles and consequently the risk.
