Low-volume roads (LVRs) represent an important part of road network in Burkina Faso and in many countries in west Africa (more than 80

Lateritic soils are the most common materials in tropical areas and widely used in pavement engineering. In fact, in these areas, pavement compaction control is generally performed through the measurements of the in-situ unit weight of the soils. However, these methods are time-consuming and expensive. So, in many countries, methodology based on the use of the dynamic penetrometer has become widespread for rapid and less costly control of pavement quality. This methodology is efficient but requires calibration between the cone resistance and the dry unit weight. Due to the specific nature of lateritic soils, to their assumed wide variability and probably to other reasons, no calibration has actually been performed in-situ for these soils. The calibrations performed until now, has been carried out using chambers in the laboratory. This process provides some bias for the prediction of in-situ dry unit weight, due to the chamber boundary effects which affect cone resistance. This paper proposes a methodology that consists of testing the soils in-situ and in the laboratory, studying cone resistance according to compaction parameters in order to estimate the boundary effects and to propose a model to take them into account when predicting in-situ dry unit weight.

Sand tailings dams have historically been the most commonly used technology for tailings storage in Chile. Although engineering advances have resulted in the construction of approximately 250-m-high facilities, some operational challenges still remain, including compaction control. Control is currently performed at a few control points in a dam embankment, without considering a series of factors that affect its mechanical behavior (e.g.,layer thickness and material variability). Within this context, geostatistics can be applied in combination with low-cost geotechnical tools as an alternative to improve compaction control in tailings storage facilities. In this study, an extensive field investigation was carried out. A total of 91 PANDA penetrometer tests were conducted to monitor the degree of compaction in an experimental classified sand tailings dam. The results were analyzed using stochastic interpolation for ordinary kriging and considering the spatial distribution of the cone resistance and the degree of compaction determined for the dam. The results showed that spatial variability was associated with the material variability of sand tailings and the compaction method used, and deviations from design requirements. The article shows the value of the use of geostatistics in decision-making in the case of classified sand tailings dams. This is mainly due to the fact that it allows optimization of the compaction process used in these tailings dams. Additionally, a useful database is generated to continue deepening studies of physical stability during the useful life of the tailings storage facilities.

Conventional geotechnical soil classifications aim to classify soils into families with geotechnical characteristics and therefore similar behaviour; however, they require core samples and laboratory identification testing. Several empirical systems for estimating the nature of the soils have been developed on the basis of several in situ geotechnical tests. However, at present these systems remain empirical and they are often only used on an indicative basis. The objective of this article is, based on the analysis of dynamic penetrometric signals, to develop a methodology able to provide an estimate of the nature of the soil crossed. The methodology developed provide an automatic classification based on artificial neuron networks (ANNs) tools. Two types of ANN architectures were considered: multi-layer feedforward perceptron (MFP) and probabilistic neural network (PNN). The learning of these two tools was achieved through a base carried out in the laboratory and in situ. Both classification models were then tested in blind conditions and showed a good efficiency for calibrated soils and promising results for in situ soils.

Pavement design with some soils, such as lateritic soils, is still a “real challenge” in some countries. In the tropical countries guide, recommended for Burkina Faso, the pavement design is performed according to the class of traffic and the soil bearing capacity, characterized by the soaked CBR index (Californian Bearing Ratio). In practice, the validation of pavement structures is possible using numerical models requiring the thickness, the Poisson’s ratio and the elastic modulus of the pavement’s layer materials, as input data for the stresses and the strain calculation. Until now, the guides provide none value of the modulus for the lateritic soils. So, the modulus is estimated by assimilating the lateritic soils to untreated coarse gravels (or GNT: Graves Non Traitées in the French guideline) with known modulus or by using empirical correlations based on the soaked CBR index. This approach lead to an erroneous estimating of modulus of lateritic soils, and to an under or overdesign of pavement structures, due to the fact that the specific correlation suitable for these soils in Burkina Faso is not known. So, this paper was interested in the characterization of lateritic soils from Burkina Faso, according to their CBR bearing capacity, their elastic modulus and to their compaction parameters, after having been assured that the selected soils were representative of country’s road soils. The objective is to perform a specific model for estimating the elastic modulus of these soils. Analysis of the studied parameters showed that CBR and elastic modulus, both depend on moisture content and dry unit weight of soils. A correlation was thus performed between elastic modulus and CBR index. The specific correlation obtained is in power form: E= a.CBR b and relatively close to whose obtained by Green and Hall, where a =66.50 and b =0.526.

Field determination of small-strain modulus (E0) and compressional wave velocity (cp) is a major issue. Even though many techniques allow determining these parameters, most of them are not able to provide precise and reliable data at low-cost. Moreover, most of these techniques do not allow to assess a soil strength parameter. Panda 3® is a dynamic penetrometer for shallow soil characterization. The device’s principle consists to measure wave propagation within the rod to each blow during driving. Force and acceleration measurements combined with signal analyses allow the obtain stress and velocity generate by the impacts. An approach based on the shock theory is proposed to determine cp and E0 from shock related equations. This paper presents the principle and the theoretical background applied. Additionally, shock relationships provide an estimating of soil axial stress-strain response during each shock. The results from a series of test carried out in calibration chamber on two sands at different densities, moisture contents and confining stresses are presented. These results are compared to those reported in the literature.

Since soils are natural and non-homogeneous materials, spatial variability of their properties has been recognised as one of the main sources of uncertainties affecting geotechnical analyses. However, soil testing is limited and the description and quantification of the resulting soil variability still remain largely subjective in practice. In this paper, anapproach is proposed to rationally evaluate the soil spatial variability using field data provided by a site investigation programme based on a lightweight dynamic cone penetrometer test. The first part deals with the boundary identification of statistically mechanical homogeneous soil units. The second part focuses on modelling spatial variability through 3D conditional random fields in the homogeneous soil units previously identified. This approach has been applied to and studied in a real site investigation carried out in an alluvial Mediterranean deltaic environment.

Knowledge of pore water pressure in an earth dam is crucial for analyzing its mechanical stability. In classical calculations of these pressures, great uncertainty exists regarding the permeability of the materials and the representation of their spatial variability. In this article, a probabilistic analysis of pore water pressures based on field data is performed to represent the permeability with a 2D random field established from statistical and geostatistical analyzes. This random field is introduced in a model based on the Finite Element Method (FEM) and the influence of the spatial variability of permeability on pore water pressure is then studied using Monte-Carlo simulations (MCS).

This paper performs a probabilistic stability analysis for an existing earthfill dam using a Stochastic Finite Element Method (SFEM) and considering the spatial variability of soil properties based on field data. Previous works on probabilistic slope stability analysis are generally based on hypothetical data while using data from existing earth structures is not widespread. A probabilistic procedure based on field data is here implemented to analyze the stability of an existing embankment dam. The spatial variability of several soil properties is modeled from the geostatistical analysis of the available dataset of the dam studied. Random variables and random fields representing the variability of dam materials are integrated into an FE model by performing Monte Carlo simulations (MCS). This probabilistic analysis based on field data allowed to characterize the variability of the sliding safety factor for the case study of an existing dam.

Les sols lateritiques sont les materiaux les plus repandus dans les regions tropicales et utilises en technique routiere pour la realisation des assisses de chaussees. Cependant, dans de nombreux pays, la qualite des routes et pistes lateritiques laisse a desirer. Ce niveau de qualite peut s’expliquer par des problemes de mise en œuvre mais egalement par des insuffisances des methodes de conception. En effet, la demarche actuelle de dimensionnement s’opere par le choix de structures de chaussees en fonction de la classe de trafic et de portance du sol caracterisee par l’indice CBR. La validation de ces structures est effectuee a l’aide de modeles numeriques necessitant l’utilisation de module elastique en donnee d’entree. Cette donnee est obtenue le plus souvent par des correlations empiriques avec l’indice de portance CBR propose par le guide CEBTP. L’objectif de cette etude est de comprendre les insuffisances de cette demarche et de proposer une approche permettant d’ameliorer la correlation entre les mesures des caracteristiques in situ de ces materiaux et leurs caracteristiques mecaniques (module et indice CBR).

The uplift load test is the most widely used technique for assessment of the performance of micropiles. Although this static load test is easy to perform and interpret, it is expensive, slow to implement, and cannot be used systematically because it destroys the micropile. A new methodology based on a simple low-strain dynamic load test (LSDT) is proposed to develop a systematic and inexpensive method for assessment of performance of a large number of micropiles. The advantage of this methodology is that it is easy to carry out, requires low-impact energy compared with a classical dynamic load test, and allows immediate interpretation. After describing the proposed methodology, the calibration, and the subsequent validation of the test method, results from full-scale micropiles at an experimental site and three real sites are presented. The results show that the methodology proposed allows confirming micropile performance under service loads in real time without disturbing the micropile. (C) 2016 American Society of Civil Engineers.

The shear strength of discontinuities plays a key role in the stability of rock masses, particularly in the case of analyzing the sliding stability of the rock foundations of gravity dams. This paper proposes a methodology for analyzing the spatial variability of shear strength along the joints of rock mass, based on the input parameters of the Barton and Choubey's model. The aim of this approach is to evaluate the reduction of the variance of the parameters involved at full-scale by identifying a deterministic trend varying in depth and a spatial correlation calculated from a variographic analysis. An advantage of this methodology is to use a simple experimental protocol (a laser profilometer, a portable shear test apparatus and a concrete sclerometer), which generates large sets of shear strength properties for assessing their spatial variability. The methodology is illustrated in the case of the rock foundation of a concrete gravity dam. Analysis of the spatial variability conducted for this case study led to a significant reduction of the variance of the variable analyzed. The advantage of this method is demonstrated through the evaluation of the probability of failure performed in a study of this structure's stability. Taking variance reduction into account in the case study led to significantly reducing the probability of failure assessed through a reliability analysis.

The degradation of roads network in tropical countries is a major concerned nowadays. Many analysis link this problem to the specific environment of roads in their areas, but also to the limited knowledge on the real behavior of lateritic soils, mainly used for these roads construction and maintenance. To understand more this problematic in order to develop an appropriate and simple methodology of road control, we led a study to characterize laterites from Burkina Faso in a laboratory by using the penetrometer PANDA. This article presents the process of calibration of the lateritic soils and the different results obtained through this investigation.

The sustainability of transportation systems infrastructures remains a major challenge for managers in order to ensure safety and quality service to passengers. For example, the French company RAT? transports up to 10 million passengers daily in Paris. Most of its underground structures are composed of masonry tunnels built several decades ago. However, at present, monitoring methodologies often remain incomplete or qualitative when evaluating the real condition of tunnels and proposing a diagnosis. This article focuses on a new monitoring methodology that complements visual inspections, based on a quantitative analysis of each tunnel component. Using a coupled analysis of the various data (qualitative, quantitative and expert) in our possession, we suggest establishing a score to qualify the actual condition of a tunnel and improve its diagnosis. The application of this methodology to carry out an evaluation of the masonry lining of the Paris Subway System shows that the proposed method and tools are particularly well adapted for characterizing and evaluating masonry lining.

The Panda 2®, developed by Roland Gourvès in 1991, is a lightweight dynamic cone penetrometer. It provides the dynamic cone resistance (qd) and depth in real time with a high sampling frequency. Nevertheless it cannot take soil samples so the penetration test is called ‘blind’. The aim of this paper is to propose an automatic methodology to predict the soil grading from the cone resistance using artificial neural networks. We have built a database based on the Panda® laboratory tests on soil samples and on in situ tests conducted next to boreholes during various geotechnical studies performed in France. Then the neural network was used to classify the cone resistance logs according to grain size distributions of the tested soils by means of feature extraction using different signal analysis. The results show that we are able to separate 4 soil classes with 98% accuracy. 1 THE PANDA 2®, VARIABLE ENERGY DYNAMIC CONE PENETROMETER (DCP) The dynamic penetrometer Panda 2® has been designed to geotechnical investigation at shallow depth up to about 5 meters (Benz 2009). It is a light-weight dynamic, highly portable cone penetrometer, which uses variable energy manually delivered by the blow of a normalized hammer. After each blow, the dynamic cone resistance qd is calculated at the current depth using the Dutch formula. One of the major interest is the high acquisition resolution. Therefore the plot of the cone resistance values against the depth, the Panda penetrogram, is a rich amount of information on the stratigraphy of site and soil properties (Shahour & Gourvès 2005) with a large number of data. Despite all the benefits provided by the Panda 2® soil samples cannot be taken during the test thus there is no information about the nature of soil. However we notice empirically that the form of the cone resistance curve might differ between different types of soil. An example shows (Fig.1), with 3 Panda® tests conducted on laboratory calibration chamber of 80 cm height. The tested soils have a different nature and granulometry but a similar density and moisture content. We can easily note the morphological differences between the 3 resistance logs. The underlying idea for the methodology described in this paper relies on this observation: the expert knowledge and perspective of an experienced engineer can detect indices in the signal form to estimate the nature of the tested soil. In this study, we have developed a model based on artificial neural network to accomplish this empiricism. Table 1. Soil parameters for the samples (Fig.1) Nature DGA Silt Laschamps

In Chile, sand tailings dams represent the most common deposits of mining residues. These structures present a potential risk in terms of mechanical instability due to their potential susceptibility to seismic liquefaction. In order to manage these risks, it is necessary to take a probabilistic approach, thus accounting for inherent variability of material properties. However, in practice, implementing such an approach is impeded by the difficulty of acquiring and managing the data to be used in the reliability calculations and is conditioned by the relevance of the probabilistic models chosen to represent this variability. This paper proposes a method for onsite determination of the tailings relative density (DR%), and its variability, using dynamic penetration tests. This method was applied to typical Chilean sand tailings dams, and proposes a single model for all such tailings dams by associating a probability model to the variation of DR%. Finally, the validity of this approach is demonstrated by performing a reliability calculation of liquefaction potential (which is the main cause for the failure of this type of structure in this country) for a particular sands tailing dam.

In Chile, sand tailings dams represent the most common deposits of mining residues. These structures present a potential risk in terms of mechanical instability due to their potential susceptibility to seismic liquefaction. In order to manage these risks, it is necessary to take a probabilistic approach, thus accounting for inherent variability of material properties. However, in practice, implementing such an approach is impeded by the difficulty of acquiring and managing the data to be used in the reliability calculations and is conditioned by the relevance of the probabilistic models chosen to represent this variability. This paper proposes a method for onsite determination of the tailings relative density (DR%), and its variability, using dynamic penetration tests. This method was applied to typical Chilean sand tailings dams, and proposes a single model for all such tailings dams by associating a probability model to the variation of DR%. Finally, the validity of this approach is demonstrated by performing a reliability calculation of liquefaction potential (which is the main cause for the failure of this type of structure in this country) for a particular sands tailing dam.