Dissipators are the components that are integrated into the flexible net barrier to absorb the impact energy during rockfall events. Upon impact, the dissipator experiences dynamic loads which makes its quasi-static characterization under low loading rate not totally adequate to describe their behavior observed on-site. This article aims to address this gap by conducting new high loading rate tests on snake dissipators under conditions that closely replicate real-case rockfall scenarios, enabling a more accurate characterization of their performance. A dynamic mechanical model of this dissipator was established and compared to the quasi-static one. This comparison highlighted the presence of a dynamic effect in the response of the snake dissipator indicating the inaccuracy of the use of quasi-static characterization. In addition, a loss of snake dissipator effectiveness was observed under dynamic loading compared to the expected effectiveness retained under low loading rate. This decrease indicates an overestimation in the performance of the snake dissipator in the flexible barrier when relying only on quasi-static characterization. Moreover, the numerical simulations also highlighted the sensitivity of the performance of the barrier by the dissipator mechanical model.

Impact tests on full-scale masonry panels are undertaken in order to explore the vulnerability of vernacular construction to rockfall hazards in mountainous areas. Seven 2.5 m high walls made of bricks subjected to a static gravity overload are subjected to a dynamic impact load on their centre, provided by ETAG concrete blocks launched at different energies (from 5 to 15 kJ). This experimental campaign is analysed through numerical simulations. Two different approaches have been explored: a micro-modelling based on the discrete element method (DEM) using the free software Siconos from Inria and a macro-modelling based on the finite element method (FEM) using Abaqus. Models have been adapted to the specificity of masonry structures under dynamic stress by hard shock. They show the strong influence of the overload on the resistance of the wall. The present work aims at assessing the damage generated by an impact on a masonry structure in real conditions. It also leads to question the failure characterisation of the structure after impact. On a long-term perspective, this work is intended to provide damage curves for masonry buildings and thus contribute to the development of normative prescriptions for natural hazard prevention.

Masonry tunnels are underground structures which can experience degradation due to ageing resulting in damage or cracking. In this work, the stability of these structures is studied using the upper-bound kinematic approach of yield design theory. This method allows estimating the ultimate load of any system knowing the geometry and the strength of its constitutive materials. However, the determination of the optimal failure mechanism, that is to say the one giving the upper-bound of the ultimate load, can be tricky as regards the interactions between the ground and the structure itself. To overcome this issue, it has been chosen to decompose the problem in two sub-problems. On the one hand, existing kinematic models dealing with the stability of excavations are extended to cohesive-frictional grounds. On the other hand, a mechanism representing the deformation observed on-field is explored to assess the masonry lining stability. Yield design theory is used to determine upper-bound estimations of the ultimate load for excavations in purely frictional grounds, cohesive-frictional grounds, and for a masonry vault subjected to a concentrated load.

Smart Rock (SR) sensors designed at the University of New Hampshire have been used to instrument laboratory and field rockfall experiments. This paper summarizes the results of instrumented small- and medium-scale tests using SRs embedded in the test blocks, performed with the aim to enhance input parameters in rockfall models and to evaluate the contribution of block rotation, often disregarded in energy computations. Fabricated concrete blocks weighing 1 and 13.5 kg were released under controlled conditions onto a poorly graded sand from a height of 1 to 10 m. The impacted surface was also varied at angles between 0 degrees and 42 degrees. High-speed video and SR data were used to determine block translational and rotational velocities of 45 tests and to calculate their energy restitution after each drop. The results demonstrate that different test conditions produce distinct responses in terms of energy restitution, highlighting the difficulty of using fixed coefficients of restitution to realistically model rockfall trajectories.

Rockfall hazard in mountainous areas requires the construction of protective structures for the buildings or transport facilities. Reinforced soil embankments can be effective protections against rockfalls due to the combination of the damping characteristics of the soil and the tensile resistance of the reinforcements. Following the advances in soil reinforcements, the research is ongoing for this type of structures aiming to optimize their shape and reinforcement design. In order to study the influence of the different geosynthetics design, two reinforced soil embankments are tested experimentally. The embankments had vertical facings and a slenderness ratio of two with the purpose to reduce the footprint of the conventional trapezoidal shape. Embankment 1 had two vertical layers of geogrids that divided its cross section in three equal parts, while Embankment 2 had multiple geogrid strips installed horizontally close to the front facing. Horizontal impact tests are performed on the two embankments using reinforced concrete blocks as impactors with the aid of a pendulum device. Several instruments and sensors are used to monitor the behavior of the embankments, namely accelerometers, pressure sensors, strain gauges, rapid cameras and a laser scanner. During the tests, the two embankments experienced local shearing at the impact position and overall backward leaning. Moreover, a waves' propagation effect is observed during the first moments of the impact at the sensors installed in the embankments. The speed of the propagation of these waves appears to be influenced by the reinforcement design of the embankments. After the tests, Embankment 2 was less deformed than Embankment 1, which is attributed to their different reinforcement design. According to the strain gauges measurements, the geogrids of Embankment 2 were more mobilized in the vicinity of the impact compared to the ones of Embankment 1. These tests showed that the geogrids installed in horizontal position close to the front facing are more efficient compared to the ones installed vertically and deeper in the embankment.

The shear strength of concrete-rock interface is a key factor to evaluate the stability of gravity dams. The shear strength assessment by achieving tests on small samples gives values different from those estimated by back-analysis on the existing dams. This work aims to study the shear behaviour of concrete-rock interface in the metric scale. Five direct shear tests were performed on bonded meter-scale concrete-granite interfaces in the range of normal stresses to which gravity dam foundation is subjected. Specific instrumentation were installed to monitor the failure mechanisms during the tests. The five concrete-rock interfaces have not broken by shearing of materials (concrete, rock) in the shear plane imposed by the test device, but by debonding of the contact between concrete and rock. Considering roughness of the contact surface in the decimeter scale and the results of shear tests carried out in the same scale, the decimeter scale is demonstrated to correspond to the elementary surface for the shear behaviour of the metric concrete-rock interface. According to the level of normal stress, the stiffness of both materials and the main asperities in the decimeter scale, different failure mechanisms occur locally to justify the overall failure in the metric scale.

This experimental campaign aimed to observe the response of reinforced soil bunds subjected to horizontal impacts. Three reinforced soil bunds are constructed next to each other separated by a double polystyrene layer to avoid mutual influence. The tested bunds were designed using a reduction factor of three, compared to the real situation. This allowed us to target energies close to failure using the available equipment (150 kJ at reduced scale corresponding to 4 MJ at real scale).

Brittle materials are very complex in their behavior, which is characterized by anisotropy in tension and in compression. Therefore, numerical modeling of brittle materials requires an insight view of material behavior and an advanced constitutive material model. This paper presents an approach to model brittle materials by using the concrete damage plasticity model (CDP). Firstly, the CDP model, including the procedure to determine its parameters is described. Secondly, size effects due to tensile cracking and compressive crushing are considered within calibration of parameters and via development and implementation of a subroutine in ABAQUS. Finally, this numerical approach is validated through numerical simulation of different mechanical tests: three-points bending test, uniaxial compression tests.

Rockfall phenomenon is a common event in mountainous areas and presents a serious risk to the nearby buildings or infrastructure. Different protection measures may be taken depending on the site conditions. A reinforced soil bund can be effective in stopping blocks with relatively high speed and kinetic energy. This article presents the response of three reinforced soil bunds subjected to horizontal impacts. The tests are dimensioned in a reduced scale with a factor of 1/3 by following specific similarity rules and assumptions. A rectangular shape with ratio height/width equal to 2 is chosen aiming to have a lower footprint than the dominating trapezoidal shape of existing soil bunds. In order to find the most efficient design, different reinforcement patterns are installed in each bund. The failure mechanisms and the role of the reinforcements are observed with several instrumentations.

Constructions located in mountainous areas are highly prone to rockfall impact risks. ln order to ensure people's safety, a robust design is necessary, then the understanding of concrete wall damages and failures caused by dynamic phenomena are highly needed. However, the scientific literature barely studied the behavior of concrete slabs, whereas beams under impact have been hugely investigated. This study focuses on walls, at the scale representative of common construction. The purpose is to provide a better understanding of the phenomena induced by rock-falls impacts. lndeed, slabs or walls subjected to outside plan loading are characterized by four types of failures: bending, shear, combination of both and punching. For a constant impacting energy, the failing mode can radical/y change; this is why the study dissociates the mass of the impactor from its velocity. Various configurations of rock-fall were experimentally processed; most of them either characterized low velocity and strong mass impactor or high velocity and low mass impactor. The different impact energies were set before the experiments. For each iso-energy cuNe, the state of the wall, its damage, and ils residual potential (characterized by a second impact) had been identified. The parametric study has enabled to define three criteria: a failure one, another describing a non-penalizing damage limit (which does not require repairs) and the last one representing a significant damage limit, involving necessary and possible repairs.

In mountainous areas, freezing is a prominent phenomenon for weathering processes in rock walls. A freezing front penetrates rock crack networks and causes its propagation. To study the evolution of rock mass stability, a suitable model of stress generated by freezing in open rock cracks is needed. This stress evaluated by the simple volume expansion model in a closed crack is too high to be realistic. In this paper, we present an assessment method for this stress and some results. Different experiments on notched limestone specimens submitted to freeze–thaw cycles were performed. Three different tight limestones (Larrys, Chamesson, Pierre de Lens) were tested. Actually, the stress generated by freezing begins to grow at the top of the notch where an ice plug is created and makes it possible for higher stresses to develop in deeper parts of the notch. Consequently, the stress induced by freezing depends on the geometry of the open crack represented by the notch. This value is, however, limited by the permeability of the surrounding rock matrix. A model of the stress evolution generated by freezing along an open crack was established and its envelope curve, named maximum stress, was parameterized. This maximum stress generated by freezing along the crack is completely defined by knowledge of the pore network of the limestone matrix and the geometry of the crack.

The shear design of concrete slabs is still an unsolved problem. The following study presents experimental and numerical investigations on the shear behaviour of reinforced concrete slabs (without shear reinforcement) under concentrated loads. The small thick slabs of 10 cm were tested in this study. Experimental tests were conducted to quantify the shear strength and the associated failure modes. The influence of several variables was addressed such as the influence of boundary conditions, four supported side slabs instead of two supported side slabs, the influence of loading plate length. A series of eight tests on six slabs were presented. The experiments are firstly used to evaluate the pertinence of Eurocode 2 and Model Code 2010 formulations using the levels of approximation LoA I and LoA II for the shear design of reinforced concrete slabs without shear reinforcement in comparison with the French approach, and secondly validated numerical modelling using the non-linear finite element method. The proposed numerical model showed good agreement with the experimental results in terms of slab behaviour.

The shear strength of the concrete–rock interface is a key factor to justify the stability of a hydraulic structure foundation. The Mohr–Coulomb failure criterion is usually used as shear strength and evaluated by extrapolating shear tests results carried out in a laboratory on small-sized samples. This paper presents an experimental study on the concrete–rock interface shear behavior. The effect of rock surface morphology on shear behavior was studied by performing laboratory direct shear tests on prepared square samples with a previously characterized rock surface. The scale effect and the test conditions were also studied by comparing the results to those obtained by performing usual laboratory shear tests on cored samples at lower scale. The tested interfaces were composed of the same concrete and granite and have a natural rock surface. The results displayed that the peak shear strength is strongly dependent on the concrete–rock bonding, the rock surface morphology and the applied normal load. A new surface morphology description tool was developed in order to characterize the main waviness. Moreover, the concrete–rock shear behavior at medium scale was reproduced by a 2D finite elements model to study the stress distribution along the sheared interface. Under low normal load, the concrete–rock adhesion is thus progressively mobilized according to the waviness on the rock surface and the local shear failure mechanisms depend on the type of this main waviness. Consequently the shear strength of a concrete–rock interface must be analyzed with respect to the various morphology aspects on its rock surface.

Conclusions pratiques du projet de recherche EDF-Ifsttar sur les fondations des barrages-poids et en particulier, la resistance au cisaillement a l'interface beton-roche.

Bilan de la campagne d'essais d'arrachements sur 36 ancrages scelles au rocher sur toute leur longueur dans le cadre du projet de recherche SNCF-GTS-Ifsttar 'ancrages au rocher'.