A multiphase approach to the stability analysis of reinforced earth structures accounting for a soil–strip failure condition

A multiphase model, developed in the context of elastoplasticity, is applied to the simulation and design of reinforced earth retaining structures. The main feature of this model is to combine the advantage of a homogenization approach, as regards its computational efficiency, with the ability to account for a specific failure condition at the interface between the soil and the reinforcing strips, which may have a decisive influence on the behavior of the structure. A particular emphasis is put on the stability analysis of this kind of reinforced soil structures, formulated within the framework of the yield design theory. Making use of a generalized rigid block failure mechanism, the stability of a reinforced earth retaining wall is investigated by means of the kinematic approach, leading to upper bound estimates for the stability factor of the structure, which are then favorably compared with the results of an elastoplastic analysis. Special attention is paid to assessing in a quantitative way how a specific soil–strip failure condition affects the stability of the reinforced earth structure as a whole.     

Numerical study of the biaxial compressive strength of high strength concrete based on a yield design micromechanical approach

This paper presents a numerical study of high strength concrete microstructure effects on its uniaxial and biaxial compressive strengths. Concrete is first represented as a set of angular aggregates interacting within a cement paste matrix. Then, a yield design kinematic approach is conducted at the mesoscopic scale in order to determine the concrete compressive strength for a given loading path. The proposed model, having a low computational cost, is able to capture the main microstructure effects already observed in literature on concrete uniaxial compressive strength, in particular, the aggregates volume fraction and maximal size effects. Finally, the proposed model also predicts the biaxial failure envelope of high strength concrete and confirms some experimental trends observed in literature.     

Pétrographie et géochimie comparées des pellets phosphatés et de leur gangue dans le gisement phosphaté de Ras-Draâ (Tunisie). Implications sur la genèse des pellets phosphatés

Paléocène-Early Eocene phosphate outcrop in the Ras-Draâ deposit (Tunisia) comprises alternances of P-rich strata (P₂O₅ ≥ 18%) and P-poor ones (P₂O₅ ≤ 2%). In phosphate rich strata, P is concentrated in rounded grains—so-called pellets—(28% ≤ P₂O₅ ≤ 38%) embedded in a matrix—so-called exogangue—much poorer in P (P₂O₅ # 7%) than pellets. The study of pellets (whose size lies mainly between 100 and 500 μm), of their surrounding matrix and of poorly phosphatic sediments, interlayered between the phosphatic strata, has been performed by optical microscopy and various chemical analyses, ICP-AES (Inductively Coupled Plasma-Atomic Emission Spectroscopy) and punctual chemical analyses by EDS (Energy Dispersive Spectroscopy). Mineralogical and chemical studies of the Tunisian phosphatic pellets show that the basic mineralogical component of these grains is carbonate-fluor-apatite, or francolite. The structural formula of a representative specimen of francolite in the basin is as follows: (Ca_4.63Mg_0.13Na_0.22)(PO₄)_2.51(CO₃)_0.48(OH_0.77F_0.23). The mineralogy of the constituents of the sediment surrounding pellets comprises carbonates, clays, silica (opal, quartz). Microscopic observations of Ras-Draâ phosphorites, added to geochemical results, establish that pellets are small bodies, allochtonous to their matrix and whose composition is independent of that of this matrix. Fecal cylindrical bodies and subspherical pellets, the latest being supposed to proceed from the fecal bodies by fragmentation, are considered to have been formed by fishes, as P-rich and organic matter-rich faeces. These faeces turn out to be resistant to mechanical dispersion and to chemical diffusion of soluble elements from and in direction of sea water. This closeness is responsible for the strong reducing conditions having prevailed inside and during the pellets diagenesis.     

Ultimate bearing capacity of a foundation reinforced by columns or cross trenches under inclined loads: a homogenization approach

This paper addresses the geotechnical engineering problem of evaluating the ultimate bearing capacity of a strip foundation resting upon a reinforced soil, by means of the yield design homogenization approach. The analysis is notably focused on the determination of the macroscopic strength criterion of such reinforced soils, where both constituents are purely cohesive, which can be conveniently expressed through the notion of anisotropic cohesion. A comprehensive comparison is made between the classical configuration of reinforcing columns and the more original one of orthogonal reinforcing trenches. Among the most outstanding results of the analysis is the conclusion that the cross trench configuration is notably more efficient in terms of load bearing capacity than the reinforcement by columns, notably when significantly inclined loading is concerned. Copyright © 2014 John Wiley & Sons, Ltd.     

A homogenization approach for evaluating the longitudinal shear stiffness of reinforced soils: column versus cross trench configuration

The macroscopic linear elastic behaviour of inclusion‐reinforced soils, regarded as periodic composite media, is investigated by means of the homogenization theory. Special attention is given here to the determination of their longitudinal shear stiffness properties, which strongly govern the reinforced ground response under lateral loading. Combining the use of analytical, variational and numerical methods, we thoroughly examined three particular engineering‐relevant configurations: single trench, column and cross trench reinforcements. Fairly accurate closed‐form expressions are thus obtained, giving the value of the reinforced soil longitudinal shear stiffness as a function of the individual components shear moduli and reinforcement volume fraction. It is shown in particular that adopting a cross trench reinforcement layout instead of the classical column configuration results in a much higher improvement of the longitudinal shear stiffness. The results are then applied to assessing the reduction of soil liquefaction risk, which can be attributed to the presence of the reinforcing inclusions. Again, they clearly demonstrate the excellent performance of the cross trench configuration as compared with the complete inefficiency of the column reinforcement technique. Copyright © 2013 John Wiley & Sons, Ltd.     

Finite element simulations of the behavior of piled‐raft foundations using a multiphase model

Finite element simulations of the behavior of a piled raft foundation have been carried out using a multiphase model conceived as an improved homogenization approach. According to this model, the ground reinforced by a group of piles is treated as a homogeneous continuous medium. In this approach, no specific interface elements are necessary to account for the mechanical interaction between the piles and the ground: this interaction is described by means of two scalar parameters, one stiffness parameter and one which can easily be derived from the maximum ground‐pile friction. The implementation of the model into a finite element code provides an efficient tool for the analysis of the influence of the pile number or length on the settlement and bearing capacity of a square piled raft foundation and of the way the total applied load is shared between the raft and the piles. Results are compared with a standard 3D finite element analysis. The comparison highlights the fact that the proposed approach remains to be improved to account for tip resistance. Copyright © 2012 John Wiley & Sons, Ltd.     

Etude de la sensibilité d’estimation de la période de retour d’une crue connue seulement par la cote atteinte

Résumé

Cet article propose une méthodologie d’étude de sensibilité pour cerner l’incertitude d’estimation de la période de retour d’une crue maximale annuelle exceptionnelle connue seulement par la cote des plus hautes eaux. Deux autres contraintes ont présidé au choix méthodologique. Tout d’abord, nous ne disposions pas des débits maximaux de crue à la station (hors ceux de la crue annuelle) et nous disposions d’une courte série des débits maximaux annuels. L’étude a ainsi été focalisée sur les débits moyens journaliers dont la série est complète avec une seule lacune qui est celle de la crue qui nous intéresse. La taille de la série des débits moyens journaliers la plus complète étant assez modérée, un modèle statistique à dépassement de seuil POT (Peaks Over Threshold) a été adopté. A partir de la laisse de crue, le débit maximal de la crue non jaugée a été estimé par cinq approches différentes (méthodes d’extrapolation par régression statistique à partir des paramètres hydrauliques de la station, deux modèles de Manning-Strickler en lit homogène, et deux modèles de Manning-Strickler en lit composé) en adoptant des valeurs de rugosité du lit couvrant l’incertitude reflétée d’après les jaugeages existants. Nous avons proposé d’étudier la distribution statistique du rapport α du débit maximum instantané de la crue annuelle à son débit moyen journalier. Par ce biais, l’observation de la laisse de crue a été mise à profit pour compléter la série des débits moyens journaliers moyennant une hypothèse sur la valeur de α. L’article traite l’incertitude sur α pour diverses hypothèses sur la rugosité du lit. Une distribution lognormale a été ajustée pour α et des simulations de Monte Carlo ont été entreprises. Une valeur du débit moyen journalier a alors été attachée à chaque simulation de Monte Carlo de α. Cinquante (50) simulations ont été effectuées dans chaque cas. Le modèle à dépassement de seuil a été appliqué à chaque série de débits moyens journaliers ainsi complétée par simulation. Les quantiles des débits moyens journaliers pour différentes périodes de retour ont été estimés pour chaque série complétée. La dispersion de ces derniers a été représentée par une boite à moustaches et a été quantifiée par l’intervalle interquartile. L’application a porté sur une station hydrométrique contrôlant un bassin de 9000 km² en région semi-aride nord africaine. Les séries d’observation ont été étudiées pour la période 1984–2006 pour laquelle 114 jaugeages étaient disponibles. La médiane du quantile quinquagennal du débit moyen journalier varie de 1410 à 1840 m³/s en fonction du seuil de dépassement retenu et de la valeur du coefficient de rugosité du lit. Les différentes méthodes, en considérant les différents seuils de troncature, conduisent à un intervalle interquartile de la période de retour du débit moyen journalier de cette crue variant entre 30 et 70 ans.

Editeur Z.W. Kundzewicz; Editeur associé C. Leduc

Citation Aridhi, H., Bargaoui, Z., et Assia Chebchoub, F., 2014. Etude de la sensibilité d’estimation de la période de retour d’une crue connue seulement par la cote atteinte. Hydrological Sciences Journal, 59 (5), 978–992.     

Summer phytoplankton pigments and community composition related to water mass properties in the Gulf of Gabes

Variations in phytoplankton pigments and community composition were examined in the Gulf of Gabes in relationship to water mass properties, characterised by the influence of the Modified Atlantic Water and by the thermal stratification. Data were collected on board the R/V Hannibal during July 2005.Distinct water masses were identified using cluster analysis of temperature–salinity (T–S) characteristics. Three major clusters appeared based on the combined effects of temperature and salinity. The first cluster was identified as the cool and less salty bottom Modified Atlantic Water (MAW). The warmer and saltier Mediterranean Mixed Water (MMW) represented the second cluster. The third cluster was the Transition Water (TW) separating the two previous clusters.The pigment and taxonomic composition of these water masses were examined. Chlorophyll a was rather low (<200 ng l⁻¹). Chlorophyll b was generally the most abundant accessory pigment and fucoxanthin dominated the accessory pigments in the MAW. Proportions of chlorophyll a associated with different phytoplankton classes were estimated using CHEMTAX software, and did not present significant variations among water groups. The results pointed out variations in the relative contribution of each phytoplankton taxa in each station group. Chlorophytes and prasinophytes accounted for 65% of chlorophyll a in the MMW. Diatoms and chlorophytes were relatively abundant in the MAW contributing to almost 63% of chlorophyll a. An unstructured community, slightly dominated by prasinophytes, chlorophytes and cryptophytes, characterised the TW. Different trophic statuses were observed in these water masses, the MMW and the MAW being characterised by mesotrophy, while an oligotrophy was observed in the TW. Nutrient availability, particularly the P-limitation supported by the summer stratification, as revealed by the high N:P ratio (greater than 20), seems to enhance the development of small-sized phytoplankton, thereby supporting the regenerated production.