Search

International audience; Poster présenté lors du IUFRO World Congress 2014.

International audience; The use of cable yarding systems constitutes an interesting solution for steep-slope harvesting in mountain forests. However, it requires many specific skills for both forest managers and operators. The purpose of this study is to provide a numerical tool to help managers in planning cable yarding projects and to facilitate the set-up of a cable line by operators. A model was developed and coded in Python language which allows fast calculation of the load path and tensile forces along the cable for different 3D configurations. This model uses theoretical equations initially developed for the mechanical design of cable structures. Its applicability in forestry was validated by field experiments. An additional module using a high-resolution digital terrain model optimizes the cable layout and automatically places required intermediate supports. If geographic information about timber volume is available, it also allows combining technical possibility with harvesting potential. From the managerial point of view, this tool presents significant advantages for forest operations planning. Moreover, operators’ productivity can be improved with the technical outputs of this model, which help to shorten the time spent in setting up a cable line and thereby reduce the total cost of a cable project.

International audience; Présentation réalisée lors du 1st Yade Workshop.

International audience; Forest is a well known and efficient natural protection solution against rockfall. In forested areas, the maintenance of forests is required to ensure their protective function and health. During this process that consists in removing some trees, the protective capacity of the forest decreases. To compensate the temporary loss of protection, some of the felled trees can be left in an oblique position to the slope. It is a financially feasible solution to ensure the protection against rockfall during the regeneration of forests. Thus, felled trees can become a useful protection system if they are correctly placed. No studies have been done concerning the efficiency of these devices and particularly their resistance to rock impacts and their energy dissipation capacity. In order to estimate the capacity of these devices to dissipate energy, it is necessary to study the dynamic response of tree stems under impact as well as rock's trajectory changes due the interaction with such structures. Experimental and numerical studies are carried out to determine the efficacy of this devices. Laboratory experiments enabled studying the response of fresh wood stems under dynamic and quasi-static loadings. A Mouton-Charpy pendulum was used on the dynamic loading tests performed onto simply supported stems. The experimental device was instrumented in order to obtain the impact force data and the stem's displacements fields. The mechanical properties of fresh wood are analyzed from the experimental results which also allow carrying out a detailed study of the stems dynamic response. A numerical model based on the Discrete Element Method (DEM) enables to simulate the interaction of a rock and a felled tree device. To simulate the rock - tree interactions, rocks are represented by spherical solid bodies while cylindrical bodies represent the trees. The fresh wood constitutive law and the contact law are integrated on the model allowing realistic simulations. The numerical model is calibrated and validated by the experimental data. The estimation of the protective capacity of these devices is assessed by the integration of the DEM model on a rockfall trajectographic one.

National audience; Ce travail de recherche est dédié à l'amélioration des modèles de rebond d'un bloc rocheux sur un sol naturel classiquement utilisés en analyse trajectographique. L'objectif final est de développer un modèle probabiliste de rebond d'un bloc sur un sol. Un modèle numérique, basé sur la méthode des éléments discrets, de l'impact d’un bloc sur un sol grossier a été développé et utilisé pour réaliser une campagne extensive de simulations numériques. L'analyse statistique, par une approche bayésienne hiérarchique, des résultats obtenus suite à cette campagne a permis de définir un modèle probabiliste de rebond reliant les vitesses de l'impactant après impact aux vitesses avant impact. Ce modèle est capable de quantifier la variabilité des vitesses après impact dans des cas de figure très variés. Une procédure d'intégration du modèle probabiliste de rebond dans le logiciel d'analyse trajectographique Rockyfor3D a été développée. La pertinence de cette démarche est étayée par la bonne adéquation entre les résultats d'essais en grandeur réelle menés à Irstea Grenoble et les prédictions d'une analyse trajectographique du site d'essais utilisant cette procédure.

[Departement_IRSTEA]Territoires [TR1_IRSTEA]SEDYVIN; International audience; While significant progress has been achieved regarding the modeling of rockfall propagation, including the interaction with forest cover, the estimation of the spatial distribution of rockfall frequency in release areas remains a challenging task. Indeed, the very low number of observed events makes any statistical modeling difficult, so that estimations are generally based on expert knowledge regarding the rock structure of start zone and a few recorded events. Meanwhile, recent dendrochronological studies have shown that trees, as silent witnesses of numerous events, bring information about the spatial distribution of rockfall trajectories. This data can be compared to simulated trajectories in order to reconstruct the most probable start zones. This study presents a methodological framework for the resolution of this problem and a real-case example. Consider a forest patch f located below a cliff with Ns potential rockfall start points. The expected value of the number of impacts in this forest patch during lapse T can be calculated as (1) : Impacts(f)=T*SUM(Pstart(i)*Ppropagation(i,f)) with Pstart(i) the probability of rockfall for the start point I, Ppropagation(i,f) the probability that a rockfall from start point i impacts at least one tree in the forest patch f . The matrix formulation for the whole forest divided in Nf forest patches is (2): Impact=T*Pstart*Propagation. Ppropagation can be estimated with numerical simulations, by taking into account the current state of the forest stand if trees are measured and georeferenced, and if the volume of rockfall is known. Impact can be also be observed on the field. Under the assumption that the forest stand and the start frequencies do not change during lapse T, the equation system (2) can be solved for the coefficients of Pstart. This framework was tested on a rockfall area located in Valdrôme, France, where more than 1000 trees were measured and geolocated in a one hectare stand located below a cliff with rockfall activity. The number of impacts on each tree was observed and starting points were determined based on a slope criterion applied to a digital terrain model. The rockfall trajectories were simulated with the RockyFor3D software. The system (2) was solved with an optimization routine. Results show that the rockfall frequencies display high spatial variability and that 21% of starting points account for more than 50% of rockfall releases. The coefficients of Pstart can then be used to weight the simulated trajectories in order to map the spatial variability of rockfall trajectories and corresponding kinetic energies, which is an important step for a better identification of hazard areas. The proposed framework can thus lead to a numerical solution, but it relies on numerous assumptions. The spatial variations of rockfall volume are not taken into account and the forest stand is supposed time-invariant. This last point could be tackled by simulating trajectories during past stages of the stand evolution with growth simulation models.

International audience; This study concerns biological degradation of non-treated wood used in ecological engineering structures. The structural design requires knowing the mechanical properties of the wood parts composing the structures at the building time as well as their evolution due to wood degradation. A bibliographic review showed a lack of scientific knowledge on these aspects, especially for wood degraded by an entire fungi community as observed in natural conditions.

International audience; In this paper, experimental and numerical studies carried out to determine the efficacy of felled trees made structure are presented. In this research work, felled trees are simplified and represented by simply supported fresh wood stems. The aim of this study is to analyze the impact of a projectile on those stems. The energy transfers between the projectile and the stem during an impact and the dynamic response of the stems are the two main subjects of this study.

International audience; Using plant roots for soil mechanical reinforcement is a common practice in the field of ecological engineering to improve slopes and river banks stability. The study of the influence of roots on the stability of forested slopes requires identifying the effect of the different root-soil interaction processes depending on soil and roots properties. For that purpose, a numerical model of rooted soils based on the Discrete Element Method was developed. The model was used to study the interaction between the roots and the soil for different loading cases and, in particular, for direct shear tests. The soil is modelled as an assembly of locally interacting spheres and the roots are modelled as deformable cylinders in the soil matrix. The model allows accounting for the root tensile loading until breakage, the root bending loading, the root-soil cohesive interaction until cohesion breakage, the root slippage associated with a frictional resistance at the root-soil interface. The study focuses on identifying the different root-soil interaction mechanisms depending on the soil type. Both frictional and cohesive soil types were used in the simulations. The effects of the roots mechanical properties - tensile, bending modulus and root-soil interfacial friction angle - and of the root number were also analyzed for the different soil types. The analysis was held both at the macroscopic scale, that is on the shear stress-strain relationships for direct shear test, and at the particles scale by analysing the spatial distribution of the contact forces magnitude, sliding contacts, and adhesion breakage for cohesive soils. The effect of the roots strongly depends on the shear strain for any soil type. The roots do not provide reinforcement but entail decrease in the shear resistance until the shear strain is large enough to induce a significant loading of the roots. For frictional soils and increasing shear strain values, the processes inducing the increase in the soil shear resistance are successively the pure tensile loading of the roots and the tensile loading combined with slippage of the root-soil interface. For cohesive soils, the pure tensile loading of the roots is followed by a progressive breakage of the cohesive root-soil links and by a complete slippage of the roots. The results show that the influence of the root number is significant if the prevailing processes are root tensile loading combined with slippage whereas it is less important if root loading is combined with progressive breakage of the cohesive links for the root configurations explored. These results illustrate the interest of the model developed to analyze the stability of rooted soils.

International audience; The term rockfall is often used ambiguously to describe various mass movement processes. Here we propose more precise terminology based on the physical nature of the moving mass, differentiating between two distinct types of rockfall: fragmental rockfall and rock mass fall. For both rockfall types, the current knowledge of the mechanisms controlling propagation of the mass movement are described, showing how these mechanisms can be simulated with different modelling approaches. However, we point out that almost no development has been realized concerning dynamic behaviour of the transitional processes between these two end-member rockfall types. Some simplified means of dealing with these complications are suggested, but we emphasize that a considerable amount of fundamental methodological development remains necessary.

National audience; L’utilisation des arbres en travers comme systèmes de protection contre le chute de bloc et stabilisation du manteau neigeux est en train d’augmenter significativement. Ces systèmes temporaires et financièrement très intéressantes n’ont pas été étudiés jusqu’à l’actualité. L’objectif du projet de thèse « étude expérimentale et numérique des systèmes para bloc avec du bois mort (arbres en travers) » est d’analyser le rôle protecteur de ces ouvrages et la capacité de dissipation de l’énergie des pierres se propageant le long du versant.

International audience; In mountainous areas, forests may hold a protection function against natural hazards. To evaluate the influence of tree anchorage on these processes, a model of tree root system/soil interactions has been developed. The model, based on the Discrete Element Method (DEM), has been developed to identify the physical processes involved for shallow landslide and rockfall hazards. The soil is modelled as an assembly of rigid locally deformable spheres interacting at their contact points. The roots are either modelled as deformable cylindrical elements embedded into the soil for structural roots or by additional cohesive contact forces between the soil particles for fine roots. For shallow landslides, direct shear tests of rooted and non-rooted soils were modelled to identify the processes involved into roots reinforcement. The model presented can be potentially used to analyze the role played by root anchorage in the protection against other natural hazards (rockfall, in particular).

International audience; This research consists on an detailed study of the physical phenomena governing the impact between the rock and the tree stem, the mechanical resistance, the energy propagation into the felled tree and the evolution of these properties depending on wood degradation. The final aim of this research work is to develop an accurate numerical model that allows simulating the protective role of the felled trees in a wide range of situations.

International audience; To compensate for the temporary loss of protection due to the reduced forest in the felling area, a considerable part of the trees felled in rockfall protection forests are left in place. Although being a widely used technique, there is little objective information on the efficacy of these trees felled across the slope. To quantify the protective function of such structures, full scale rockfall tests on a protective structure made with trees felled across the slope were carried out. The results show significant energy losses when impacting the structure. However, the rock energy significantly increases again after the structure if the rock is not stopped. Consequently, a larger number of felled trees is required to stop all rocks and, to increase the number of stopped rocks, the structure height has to be determined in relation to the rock size which promotes, in most cases, structures made of stacked trees.

International audience; Attenuation, is the reduction in intensity of any flux through a medium. Rock fall attenuator systems are structures made of flexible wire netting designed reduce the intensity (kinetic energy) of rock fall and guide their trajectory. In order to evaluate the protective effect of an attenuating system and their capacity to absorb high energy impacts, high speed video recording methods have been applied to capture full scale rock rolling and attenuator barrier tests. The application of video methods requires that individual frames are scaled to the plane of the rock fall. Following this the rock fall trajectory and velocity over time can be extracted from the video. Herein, an assessment of the error in different approaches to video image scaling and position tracking for video analysis is made. Additionally the trajectories and velocity of rock falls from both un-attenuated rock rolling experiments and experiments testing the rock fall attenuator systems are analysed. A comparison of these data has allowed an assessment of the energy absorbed by the attenuator systems. Additionally, restitution coefficients for impacts on hard rock slopes have been drawn from the data, assisting in a simple model applied to estimation the attenuation effects of rock fall attenuator systems.

International audience; La gestion des forêts à fonction de protection est facilitée par la télédétection par scanner laser aéroporté. Les modèles de terrain et informations forestières qu'elle permet de cartographies sont en effet des variables cruciales pour la cartographie des forêts de protection et pour leur caractérisation par la simulation de chute de blocs.

National audience; Afin de mieux comprendre le rôle de protection des forêts de montagne contre les chutes de blocs, un modèle d'impact d'un bloc sur une tige de bois vert est proposé. Ce dernier est basé sur une modélisation par éléments finis (MEF) développée à partir du code libre CAST3M. Pour décrire l'interaction bloc/arbre, la cinématique et les phénomènes dissipatifs au cours de l'impact, les non-linéarités géométriques, matérielles et de contact sont prises en compte. En particulier, un algorithme de gestion du contact tige/bloc, basé sur la méthode des éléments discrets (MED), a été développé. La calibration est faite à partir d'essais d'impact à l'aide d'un pendule de Mouton-Charpy instrumenté. L'objectif de cette étude est de quantifier les échanges énergétiques et les modifications de la cinématique du bloc lors de l'impact, données très utiles pour prédire la trajectoire de blocs rocheux se propageant le long d'un talus boisé.

Rockfall is a major risk in mountainous zones. In the field of falling rocks trajectory simulations, many scientific problems related to modelling the bouncing phase remain to be solved. The large variability of slope properties and boulder removal conditions pleads for a stochastic modelling of these trajectories. In particular, the rebound of falling rocks has to be modelled as a random process (Bourrier et al., 2009a). The aim of this study is to define a stochastic impact model that can capture the variability of the bouncing phenomenon on coarse soils, in particular. For this purpose, a numerical approach based on the Discrete Element Method (DEM) is used to investigate the impact of a boulder on a coarse granular soil in 2D. This model is used to perform an intensive simulations campaign. The analysis of results from this campaign is carried out using advanced statistical methods leading to the definition of a stochastic impact model. The stochastic impact model is finally integrated in a falling rocks trajectories simulation code and the code developed is evaluated by comparing real-scale experimental results to rockfall trajectory simulations.

International audience; The use of rockfall protection forests is becoming a common practice in the European alps. Despite several recent research works, the estimation of the efficiency of standing trees and wooden protection structures remains very difficult. This efficiency is closely related to the resistance of - fresh, dry or decayed - tree stems to rock impacts. This resistance can be estimated from specific field experiments or from results from other research fields such as tree resistance to wind, for example. However, the impact of a standing or cut tree by a rock constitutes, from a mechanical point of view, a very complex and specific problem. Consequently, depending on the phytosanitary state of the stem, the type of structure impacted (standing trees or wooden structure), and the size and shape of the impacting rocks, specific studies should be done. In this context, the study presented aims at developing an innovative half-scale laboratory experiments to study the resistance of stems to rock impacts on different configurations in terms of tree properties (specie, dimensions ), projectile size and shape, and boundary conditions (fallen or standing trees). A methodology developed to use these experimental results is also presented for rock impacts on small fresh beech stems. This methodology consists in using the experimental results to calibrate a numerical model of impacts on fresh beech stems. The model calibrated can be used to design a rockfall wooden protection structure or to implement a tree interaction model into rockfall codes, for example.

Dans ce livre, le Cemagref, institut de recherche en sciences et technologies pour l'environnement, fait le point de ses travaux scientifiques menés sur la forêt. Pourquoi une recherche sur la forêt ? D'abord parce qu'elle occupe plus d'un quart de la surface de notre territoire métropolitain. Ensuite, parce qu'elle fournit tout un ensemble de services variés et, pour certains, essentiels à notre société : la production de bois, la protection contre les aléas naturels, la chasse et la promenade en sont de très bonnes illustrations. Enfin, parce qu'elle constitue un milieu naturel riche en biodiversité et occupe une place notable dans les débats sur les enjeux environnementaux. Forts de ce constat, les chercheurs et experts du Cemagref apportent leur contribution en abordant des problématiques majeures pour la gestion durable des forêts : état et dynamique de la forêt, de sa biodiversité, vulnérabilité des forêts aux changements globaux et stratégies d'adaptation, gestion multifonctionnelle des forêts, rôle de protection contre les aléas naturels, place de la forêt dans les territoires et enjeux sociaux. L'interaction entre forêt et gestion se situe ainsi au c½ur de nos travaux.

International audience; Classical rockfall simulations codes could be highly improved by defining models of the interactions between the falling rocks and the natural environment that integrate the variability associated with these interactions. Probabilistic models of the rebound of the rock on the soil and of the interaction of the rock with standing trees were developed. The probabilistic rebound model is based on the statistical analysis of numerical simulations of the impact of a rock on a soil using the Discrete Element Method. The procedure for the integration of the effect of forest is based on the probabilistic analysis of results from field experiments. The models allow the probabilistic calculation of the kinematics of the falling rock after interaction depending on the kinematics before interaction using a reduced number of easily identifiable field parameters. These models have been integrated into the trajectory simulation code Rockyfor3D. Comparisons of rockfall simulations using these models with field experiments have shown the efficiency of the approach.

International audience; In mountainous regions forested slopes play an important protective role against rockfall. Up to now, most of the researches on rockfall protection forest have been and are focused on the dissipative effect of standing and living trees. There are hardly any studies on the protective capacity of tree stumps and lying stems against snow avalanches and rockfalls. Although, these techniques are more and more used throughout the Alps. In Austria, the felling technique Alpi has been developed, which allows a specialised lumberjack to create small rockfall barriers using one or two tree stems anchored on high tree stumps. Lying tree stems can be then used to increase efficiently the roughness of the soil and so to limit or avoid triggering and propagation falling rocks. But, due to the wood decay, the efficiency of such bio-engineering type protective works is decreasing with time. One of the questions that the forest and natural hazard managers have to answer is: what is the lifetime of such protective structures? In order to answer to this question we have developed a specific research on this thematic. The main objectives of this research program are to quantify the efficacy of these bio-engineering type rockfall fences depending on their characteristics (stump and stems density, position on the slope, tree species, etc.), and to evaluate their resistance over time. To achieve these objectives we have developed two types of experiments. The first one, performed during the summer 2009 on our experimental site test of Vaujany (France), are full scale rockfall experiments on four felled trees, which have been anchored on their stumps and are lying in an oblique direction to the slope. In total, fifty rocks (from 522 at 2242 kg) have been released one by one. For each rock, the trajectory has been filmed with high speed digital cameras. The second type of experiment is the uprooting of tree stumps (spruce, fir and beech) of different ages and diameter. To uproot the stumps we are using the same technique as for winching tests. For each experiment, stumps samples are taken in order to determine the wood decay ratio. Finally, a statistical analysis will be made in order to study the correlation between the force required to uproot a stump and its wood decay ratio. The first uprooting test campaign has been held during the summer 2009. First results of both experiment types will be presented. This study is part of the European Interreg IV Alpine Space project "MANFRED".

International audience; This paper proposes numerical investigations for the stochastic modeling of a boulder impacting a coarse granular soil. The soil is considered as a noncohesive granular medium using a discrete element method and the soil model is calibrated compared to results from half-scale experiments. Based on this numerical model, an extensive numerical simulation campaign is carried out. The statistical analysis of the numerical results allows the definition of a stochastic bouncing model that quantifies most of the variability of the numerical results. Comparisons with classical bouncing models in the field of trajectory analysis highlight that this model expands classical approaches. The comparison of results from real-scale experiments to trajectory simulations based on the stochastic model show the relevance of the proposed approach to modeling rockfall trajectories.

National audience; Dans ce chapitre, différentes approches de l'évaluation de l'aléa sont exposées. Certaines sont utilisées dans la pratique, d'autres, issues de la recherche, sont proposées dans l'objectif de rendre plus quantitative cette évaluation.

International audience; The impact of a block on a coarse granular soil corresponding to rockfall events is investigated using the Discrete Element Method. Different impacting particle and medium characteristics are considered. Based on these numerical results, the impacting particle bouncing occurrence diagram is defined for various impacting particle sizes, incident kinematic parameters and sample heights. The bouncing occurrence diagram brings out three impact regimes. For a small impacting particle, the impact is mainly determined by the first interaction between the impacting particle and the soil, whereas for an intermediate-sized impacting particle, the compression wave propagation through the sample is the leading phenomenon. For large impacting particles, bouncing is associated with the formation of a compact layer below the impacting particle.

This thesis is dedicated to the study of falling rocks bouncing on natural soils in order to improve the models used for rockfall hazard assessment. The impact of a rock on a scree slope is modelled using the Discrete Element Method. The comparison between numerical results and half-scale experimental results of impacts on a coarse soil shows that the numerical model allows accurately predicting the bouncing of the rock for a reduced number of model parameters to be calibrated. The numerical results also emphasize that the interaction between the impacting particle and the granular soil is composed of three stages : the partial energy exchange from the impacting particle to the soil, the propagation of a shockwave from the impact point and the wave reflection on the substratum. The study of energy exchanges during these three stages allows defining the impacting particle bouncing occurrence diagram as well as three impact regimes. The statistical analysis of numerical results using Bayesian inference leads to the definition of a stochastic impact model. This model is relevant for rock velocity after impact depending on both rock velocity before impact and soil particles layout near the impact point. Finally, the stochastic impact model is integrated into trajectory analysis models which allows defining a global probabilistic approach of rockfall hazard assessment including protective structures design.; Ce travail de thèse porte sur la caractérisation du rebond d'un bloc sur un terrain naturel dans la perspective d'améliorer les modèles de détermination de l'aléa de chute de blocs. L'impact d'un bloc rocheux sur un sol composé d'éboulis est modélisé par la Méthode des Elements Discrets. La comparaison entre les résultats de simulation et les résultats d'essais à échelle réduite d'impact sur un sol granulaire grossier met en évidence que le modèle numérique développé assure une prédiction pertinente du rebond pour un nombre réduit de paramètres de simulation à calibrer. L'analyse de l'impact à l'aide du modèle numérique montre que l'interaction entre l'impactant et le sol peut être décomposée en trois phases : le transfert énergétique initial du bloc vers le sol, la propagation d'une onde de compression du point d'impact vers l'intérieur du sol et la réflexion de l'onde de compression sur le substratum. L'étude des échanges énergétiques lors de ces trois phases conduit à la définition d'un diagramme d'existence du rebond délimitant les domaines d'arrêt et de rebond de l'impactant et à l'identification de trois régimes d'impact. Le traitement statistique des résultats de simulation par des méthodes statistiques basées sur l'inférence Bayésienne permet également de définir une loi d'impact stochastique. Cette loi est représentative de la variabilité des vitesses du bloc après impact en fonction des paramètres cinématiques incidents et de l'arrangement géométrique des particules du sol au voisinage du point d'impact. Enfin, suite à l'intégration de la loi stochastique d'impact dans le contexte de l'analyse trajectographique, une approche probabiliste globale permettant la caractérisation détaillée de l'aléa de chute de bloc ainsi que l'implantation et le dimensionnement d'ouvrages de protection est proposée.

International audience; Rockfall hazard assessments are increasingly done using trajectory analysis. A key parameter in such approaches is the modelling of the interaction between the falling rock and the soil. The presented study focuses on this phase in the case of a boulder impacting a scree slope. In the context of trajectory analysis, the available information on the local configuration of soil particles near the impact point and on the incident kinematic parameters of the boulder are not sufficient to perform a deterministic prediction of boulder bouncing. Stochastic approaches are then necessary to properly model the impact phase. Numerical investigtions using a Discrete Element Method are performed in order to investigate the impact process. The statistical treatment of the results leads to the definition of a stochastic impact model linking the block kinematics parameters before and after its interaction with the granular soil. The comparison between the predictions of the model and the results from half-scale experiments emphasizes the relevance of a predictive use of the stochastic impact model and the interest of using stochastic approaches in the field of trajectory analysis.

International audience; A challenge in rockfall modelling is to develop a robust rebound algorithm for which parame-ter values can be estimated objectively in the field. In this paper, we tackle this challenge by presenting and testing a stochastic impact law that requires limited parameterisation by the model user. The approach provides promising results. We will present the simulation results and their validation as well as remaining unknowns and future developments.

International audience; Rockfall hazard assessments are increasingly done with rockfall numerical simulations. A key parameter in such models is the modelling of the interaction between the falling rock and the soil. The presented approach focuses on the impact phase modelling in the case of the interaction of a boulder with a coarse granular soil. The soil is modeled as a non cohesive granular medium using the Discrete Element Method. The statistical analysis of numerical results is carried out using Bayesian methods which leads to the definition of a stochastic impact model. The relevance of a predictive use of the model is evaluated by comparing the reflected velocities predicted by the model to results from half-scale experiments.

Colloque avec actes et comité de lecture. Internationale.; International audience; Les logiciels de trajectographie constituent un outil indispensable pour la détermination des zones soumises à un risque de chute de bloc en zones montagneuses. Une amélioration fondamentale à apporter à ces codes de calcul est la caractérisation fine du rebond du bloc sur le sol. Dans cette optique, une modélisation numérique de l'interaction entre le bloc et le sol par la Méthode des Elements Discrets est développée de façon à étudier l'impact d'un bloc rocheux sur un éboulis. Le traitement statistique des résultats issus des simulations numériques permet de définir une loi d'impact stochastique valable quels que soient le point d'impact et les conditions cinématiques initiales du bloc.