Pascal Villard

Design of geosynthetic reinforcements for platforms subjected to localized sinkholes

Construction of road and railway platforms in areas subject to localized sinkholes requires the use of specific reinforcements, for example, geosynthetics. The current design method for these structures is based on the assumption that there is no displacement of the geosynthetic in the anchorage areas on either side of the cavity. A new analytical method is proposed that takes into account the displacements and deformation of the geosynthetic reinforcement in the anchorage areas and the increase in stress at the edge of the cavity. To validate this new analytical method, a full-scale experiment was carried out; the use of optical fibre sensors integrated into the geosynthetic sheet made it possible to accurately measure the strain of the geosynthetic reinforcement. Comparison of the results obtained by this new analytical method with measurements of a full-scale experiment and the results of a finite element model confirmed the relevance of these new developments.

Investigation of Load-Transfer Mechanisms in Geotechnical Earth Structures with Thin Fill Platforms Reinforced by Rigid Inclusions

The reinforcement of soft soils by rigid inclusions is a practical and economical technique for wide-span buildings and the foundations of embankments. This method consists of placing a granular layer at the top of the network of piles to reduce vertical load on the supporting soil and vertical settlement of the upper structure. The study focuses on the modeling of load-transfer mechanisms occurring in the reinforced structure located over the network of piles with a coupling between the finite-element method (geosynthetic sheets) and discrete element method (granular layer; concrete slab in some cases). The importance of granular layer thickness to increase load-transfer intensity and to reduce vertical settlement was observed. However, without a basal geosynthetic sheet, the compressibility of soft soil has a great influence on the mechanisms. A method predicting the intensity of load transfers was proposed, based on Carlsson’s solution. The main parameters concerned are the geometry of the work and the peak and residual friction angles of the granular layer.

Effects of Satellite Transmitters Fitted to Breeding Cory's Shearwaters

ABSTRACT We studied the effects of a new technique to fix satellite transmitters on the feeding behavior and chick body growth of the Corys shearwater (Calonectris diomeded). Transmitters did not significantly affect the breeding birds body mass or diving performance. However, on average, pairs with ≥1 adult equipped with a transmitter brought a lighter meal (54.8 ± 18.9 g) to nestlings than did control birds or the same pair with the transmitter removed (77.4 ± 8.3 g) leading to a decreased chick body growth. However at fledging, chicks in control and treatment groups had similar body mass whether parents were equipped with a PTT or not. Our method of transmitter attachment could therefore be an appropriate alternative for studying the ecology of procellariiforms weighing < 1 kg given that birds are fitted with transmitters for <2 weeks and restricted to 1 parent only.

Prediction of Load Transfers in Granular Layers Used in Rigid Inclusions Technique—Experimental and Discrete Element Method Analysis

The rigid inclusions technique is based on the use of a network of piles covered by a load transfer granular layer which allows the construction of roads, railways and buildings on very poor subsoil. Today, there are no relevant design methods able to predict the behavior of such structures due to the lack of knowledge about the load transfer mechanisms in the granular layer. Consequently, experimental works and numerical studies based on the Discrete Element Method were carried out. Both experimental and numerical results show that the granular layer can reduce greatly the surface settlement and the load applied to the soft soil. The numerical analysis shows that the load transfer mechanisms take place in a specific area located over the piles and is highly influenced by the rigidity of the soft soil. Thanks to the good agreement obtained with the numerical analysis, a prediction of the load transfer amplitude close to the description of Carlsson was proposed.

Discrete Element Modeling of a Rockfall in the South of the “Massif Central”, France

This work aims to model the mechanical behavior of a natural rockfall of approximately 1,000 m3 using the discrete element method (DEM). The cliff geometry, fracture network, deposit geometry and slope topography, resulted from field measurements. The numerical model used for this study accounts for the energy dissipated by collisions and friction/abutment in the bulk of the rock mass as well as at the rock-slope interface. The contact parameters needed for the numerical simulation are issued from the literature. A sensitivity study was performed in order to analyze the role played by the contact parameters, the blocks shape and the slope geometry (roughness) in some important indicators (mass distance propagation, energy dissipation and rockfall deposit geometry). From the results obtained, it appears that the numerical model is able to provide, only from basic inputs collected in situ, a valuable description of natural rockfall propagation and stop. Assessment of different physical quantities such as energy dissipation modes (by collisions or friction/abutment), allowed us to also improve the understanding of rockfall events.

Breeding Biology of the Bridled Tern (Sterna anaethetus) in New Caledonia

Abstract. Bridled Terns (Sterna anaethetus) have only recently been discovered breeding in New Caledonia, the previous nearest breeding colonies being in Eastern Australia. Bridled Tern breeding biology was investigated in a single, small (c. 100 pairs) colony during the 2003 and 2004 breeding seasons. Of 93 nests monitored, 82 were under cover (70 under rock and twelve under vegetation). Nesting location did not predict hatching success. Almost half (49%) of the 37 nests monitored in 2003 were reused in 2004. Egg-laying took place between late October and mid-January; egg volumes did not vary between the two breeding seasons. Chicks reached their asymptotic weight (132 g) at 37 d, faster than chicks in Australia (40–53 d). Wing length was the most reliable predictor of chick age. An equation based on density was generated to age eggs, and graphs to age chicks. An estimated 81% of clutches produced fledglings. Predation of tern nests by Silver Gull (Larus novaehollandiae) was not observed, but Peregrine Falcons (Falco peregrinus) were observed hunting and catching adult terns.

Using Nonconvex Discrete Elements to Predict Experimental Behaviour of Granular Materials

A numerical model based on the DEM is proposed to reproduce the macroscopic behavior of a granular material without having to take into consideration the complexity of the real grain shape. It is based on nonconvex simple elements made up of interlocked, jointed or linked spheres which allow a high level of interlocking within the granular assembly. The ability of the numerical model to reproduce the mechanical behavior of a granular assembly with different porosities and several confined pressure was established by comparison with experimental axisymmetric triaxial compression tests on real sand. Since the shapes of the numerical elements are distant from the real shapes, the intrinsic porosities of the numerical and experimental samples cannot be directly compared. That’s why the notion of relative density was introduced, which was established based on the minimum and maximum porosities. In these conditions, we show that the numerical model, once it had been calibrated, is able to predict the experimental results of the triaxial tests on sands of different relative densities and under different confining stresses.

2 – DEM Applied to Laboratory Experiments

: This chapter aims to test the ability of the DEM to simulate well-controlled experiments that involve particle flows. Two experiments were chosen from the literature for this purpose: channelized avalanches of plastic pellets or glass beads, which can be assumed to be two-dimensional (2D) experiments, and a three-dimensional (3D) experiment carried out at the EPFL (Ecole polytechnique federale de Lausanne) involving small baked-clay bricks. The main difficulties met when comparing some experimental and numerical results relate to the definition of adequate parameters or in the ability of the numerical model to account for all specific experimental characteristics (boundary conditions, right number and realistic shapes of particles for example). The experiments of Savage and Hutter, Hutter and Koch and Hutter et al. were chosen because of the fact that the authors have made several attempts to measure the main physical parameters such as normal dissipation or friction coefficients, which will necessarily be used as input parameters in the modeling, and because the experiments can be considered as 2D problems. Nevertheless, due to the large number of particles and the complex shape of the elements used in the experiments (for plastic pellets in particular), it was not possible to model all details in the performed experiments. In this case, only the global macroscopic behavior of the granular flow is of interest; analyzing it from a microscopic viewpoint would be misleading. In contrast, the experiments performed at the EPFL were chosen because of their simple geometric configuration (low number of particles, simple boundary conditions, regular sizes and well-controlled shapes of the particles). Also, the materials used for the experiments (bricks and plastic coating for the propagation area) have been made available to us because of a collaboration established as part of the European project Alcotra MASSA (Medium and Small Size Rock Fall Hazard Assessment, 2010–2013) in order to assess the dissipation parameters via elementary rebound tests with single bricks.

4 – Application to Actual Rockfalls

: This chapter aims to present real-life studies using the code DEMbox. Each of the three cases presented was chosen for their remarkable specificities. The first case is a naturally triggered event that involves relatively few blocks – rather big and elongated – in which the propagation path is quite short. The second case is an artificially triggered event by means of explosive charges. The size grading of the blocks is consequently narrow. The propagation path is long with a soft ground in certain regions. The third case is an attempt to use the model as a predictive tool.

3 – Parameters that May Affect the Flow

: It is clear that many parameters have a more or less significant influence on the flow. This chapter aims at describing the role of a selection of parameters related to (1) the amount and the shape of the bricks that constitute the flowing mass, (2) the contact/collision parameters and (3) the roughness of the inclined slope and the “abruptness” of the transition between two slopes.