Catálogo de publicaciones - libros

Physically based mathematical models and statistically based empirical equations each may provide useful means of forecasting runout of rock and debris avalanches. This paper compares the foundations, strengths, and limitations of a physically based model and a statistically based forecasting method, both of which were developed to predict runout across three-dimensional topography. The chief advantage of the physically based model results from its ties to physical conservation laws and well-tested axioms of soil and rock mechanics, such as the Coulomb friction rule and effective-stress principle. The output of this model provides detailed information about the dynamics of avalanche runout, at the expense of high demands for accurate input data, numerical computation, and experimental testing. In comparison, the statistical method requires relatively modest computation and no input data except identification of prospective avalanche source areas and a range of postulated avalanche volumes. Like the physically based model, the statistical method yields maps of predicted runout, but it provides no information on runout dynamics. Although the two methods differ significantly in their structure and objectives, insights gained from one method can aid refinement of the other.

Continuum modelling of flow-like landslides is a possible approach that can be adopted to simulate landslide instability, and the transition to catastrophic failure up to flow development. Models based on continuum mechanics and associated with different rheological models are usually preferred to predict landslide runout and relevant parameters. A finite element method approach is here presented and contrasts previous research where depth-averaged equivalent-fluid approaches were adopted. We developed a 2D/3D finite element code to analyse slope stability and to model runout of mass movements characterised by very large displacements. Different material laws already known, tested and verified for granular materials have been implemented. Materials laws include classical elasto-plasticity, with a linear elastic part and different applicable yield surfaces with associated and non-associated flow rules. A series of simulations has been performed. The effects of different morphological conditions have been simulated to understand processes close to obstacles, both deformable and perfectly rigid, or of trough shaped profiles where accumulation is favoured. Mohr-Coulomb, Drucker-Prager, von Mises models with or without strain softening have been adopted for the granular materials. Results for Mohr-Coulomb material are presented and demonstrate the capability of this approach and the relevance of internal deformation within the flowing material.

Landslides are known to travel further than expected from the coefficient of friction of their material. In some cases, this is just because the ratio of the height lost to the horizontal distance travelled (H/L), which is compared to the coefficient of friction, is not computed from the centre of mass of the deposit, as it should be, but from the distal end. Simple spreading of the landslide mass can then explain the excess runout. However, spreading alone is not able to explain the spectacular runout of most landslides, for which the centre of mass does travel further than predicted for a frictionally-controlled slide. The long travel distance of the centre of mass cannot be explained by dry granular models. As it is well known that water reduces solid friction in debris flows, and that significant amounts of water are present in many landslides, it is proposed here that water is the main cause for the unexpectedly high mobility of landslides. Water in the debris also introduces a viscous dissipative stress which can account for the relatively channelled behaviour of landslides over topography. The difference between landslides and debris flows is wholly gradational and related to the water content.

Rock avalanches are relatively infrequent, but highly destructive. From the point of view of societal costs, direct damage due to rock avalanches is of similar magnitude as loss due to defensive actions, esp. slope stabilization and alienation of land. Runout estimates are needed in order to minimize hazard areas that need to be set aside to protect population against perceived rock avalanche hazards. Empirical methods reviewed include the travel angle approach and area-volume correlation. Theories attempting to explain high mobility of rock avalanches are reviewed. In particular, the hypothesis of lubrication by liquefied saturated soil entrained from the landslide path is discussed. In a discussion of analytical modelling approaches, preference is given to an empirical approach, calibrating simple rheological models by means of back-analysis of existing cases.

Several large rockslides and rock avalanches ranging in volume from 10µ m³ up to 10¸ m³ were triggered by underground nuclear explosions at the Novaya Zemlia test site. Rapid filming of rock avalanche formation allowed direct measuring of the velocities of debris spreading. Dynamics of two case studies derived from the real time observations and from the analysis of debris morphology and grain size composition is discussed in details. Factors determining runout of artificial rock avalanches such as variability of debris grain size composition and topography of the transition and deposition zones are examined. Relationships of rock avalanche runout and their volume are determined and compared with those of the natural events of different origin. Critical conditions of slope failure occurrence depending on intensity of seismic effects of the explosions and slope angles are examined as well.

is a new hypothesis for the mechanism of rock-avalanche long runout. Low-strain-rate fragmentation is dominated by growth of a few flaws. It is the regime leading to initial failure of many landslides. Static rock strength is largely independent of loading rate. The dynamic regime is entered when growth of a few flaws does not relieve elastic strain fast enough, and stresses rise adjacent to the flaws, forcing many new ones to nucleate and grow. Strengths of dynamically fragmenting materials increase at about the 4th root of strain rate. Elastic strain energy, W, per unit volume, released at failure is given by W=Q²/(2E), where Q is strength and E is elastic modulus. Its explosive release as kinetic energy provides a large, isotropic, clast-dispersing stress, every time any clast is stressed to failure. Fragmentation-induced dilation is a positive granular “pressure”, but also causes low pore-fluid pressure, and is incompatible with saturation of voids by liquids and therefore is incompatible with high pore pressure and undrained loading. Driven entirely by internal deformation within the avalanching mass, dynamic fragmentation propels the distal margins of large avalanches of brittle rock further than they could travel had they just collapsed to joint-bounded clasts.

Massive rock slope failures and subsequent rapid motion of huge masses of debris is a complex process. Data required for its understanding and numerical modelling can be derived from detailed study of the morphology and internal structures of rockslide source zones and deposits. Proposed model to explain their peculiarities must not contradict any of the observable phenomena, which should be regarded as constraints with which to check a model's reliability. To develop reliable models for the formation and motion of rockslides and rock avalanches we must take into account the whole assemblage of morphological, structural and depositional features typical of them. Proceeding from case-by-case analysis to a comprehensive synthesis of the whole phenomenon requires systematisation and classification of all of the observable features typical of numerous rockslides and rock avalanches. Classification criteria and principles of data analysis and classification that allow selection of different types and genetic sequences of the phenomena in question are presented and discussed. They are illustrated by case studies from Central Asia, the Caucasus and some other regions.

A chronological review illustrates the evolution of the geological interpretation of the Flims rockslide event and leads to new field evidences and dating results. Fundamental is the new dating of the rockslide event to the Boreal time of about 8,200 - 8,300 yBP (uncalibr.). During that time, the climate excluded any direct glacial influence on the rockslide mechanisms and on the surface morphology. All the supposedly "clear" evidences for glacial influences during or after the Flims event therefore must and can be explained by other processes. The main rockslide body has moved by a slab by slab mode with extreme deformation along the bedding planes, responsible for a high mobility. Nevertheless the sedimentary rock structure was well preserved. Despite the good stability conditions of the material the rockslide dam has broken. The extreme long runout distance of the most distal parts of the rockslide material can be attributed to secondary effects. The impact of the rockslide on the valley bottom, occupied presumably by a pre-existing lake, displaced the alluvial sediments and transported large parts of the Flims rockslide and of the nearby and older Tamins rockslide material far into the Hinterrhein valley. The displaced alluvium resedimented as "Bonaduz gravels", a characteristic graded but unstratified sediment.

Maiella massif, in the easternmost sector of the central Apenninic chain. The massive rock slope failure involved about 30 x 10m volume of well-bedded Eocene and Miocene limestone and marly-limestone, which slid onto the Aventino river valley. A comprehensive study has allowed the identification of morphological, structural and lithological constraints on slope failure style, which is featured by a break-out geometry rather than a toe rupture one. The break-out developed by cutting bedding planes and utilizing low angle shear surfaces. According to a geological model of rupture, the stress-strain analysis performed through a finite difference code, revealed the fundamental role played by the rock mass anisotropies due to shear surfaces (inherited tectonic derived surfaces). Topographic stress itself could not have been sufficient for determining significant slope deformation. Shape and dimensions of the Lettopalena rock avalanche deposit suggest a low mobility and high energy dissipative event due to the geomorphic control of the pre-existing topography and the high deformability of the outcropping Mio-Pliocene formations, despite a significant run-up on the opposite side of the Aventino valley. The Lettopalena rock avalanche occurred about 4.8 kyr BP along the SE slope of the

Mountain slope movements have been recognized at numerous sites in Glacier National Park (GNP), British Columbia, especially along the transportation corridor through the Columbia Mountains, one of the most important in western Canada. The corridor links Golden and Revelstoke via Rogers Pass, and is utilized by the Canadian Pacific railways and the Trans-Canada Highway. Characterized by glacially overdeepened valleys and steep mountain slopes underlain by complex metamorphic rocks, the GPN landscape is especially prone to major slope movements. These slope movements are characterized by massive bedrock landslides and sagging slopes in which movement may result from sliding, toppling and flow, either singly or in combination. The influence of geology, rock mass fabric, and slope morphology on the mode of failure of slopes in GNP is illustrated by an ongoing multiple-mode slope movement in the Beaver River Valley, the East Gate landslide. The hazard assessment framework, as well as the influence of climate in triggering the failure, is discussed. Comparison with other similar examples of complex landslides in the French Alps is briefly outlined.