No abstract

The technique for constructing geosynthetic-reinforced soil retaining structures is now in widespread use. Since its first use in the late 1960s, construction processes have become more technologically advanced, and the number of alternatives has reduced, with most of the construction techniques being standardised. This type of reinforced structure has been in use for nearly 30 years, during which time its reliability has been proved. At the same time, work is in progress to standardise the design of these structures using the new international format of Ultimate Limit States and Serviceability Limit States calculations and associated partial factors. This study presents the procedure currently being prepared for the future standard in France. First, a preliminary parametric calibration based on typical structure profiles is described, comparing different stability calculation methods and different partial factor combinations. Second, design validation of instrumented case histories is performed.

There can be no doubt that there are now real economic benefits to be gained from using retaining walls as bearing structures. To justify this technical option, research programmes have been undertaken in a variety of European countries, involving full-scale experiments on instrumented structures subjected to localised surcharge loading. In France two full-scale instrumented structures were built and tested until failure. At the same time, a finite-element numerical modelling system was established and used to optimise structural dimensions after calibration with respect to experimental results. The results of the model–experiment comparison are presented and analysed along with the results of the parametric study which sets forth a number of recommendations concerning the design of this type of structure. L'utilisation des murs de soutènement comme structure porteuse représente à l'heure actuelle un intérêt économique certain. Pour justifier ce choix technique, des programmes de recherche, mettant en œuvre des expérimentations en vraie grandeur sur des ouvrages instrumentés sollicités par des charges localisées en tête, ont été réalisés dans différents pays européens. En France deux ouvrages en vraie grandeur instrumentés ont été réalisés et testés jusqu' à la rupture. Conjointement une modélisation numérique par éléments finis a été mise en œuvre et une fois validée sur les résultats expérimentaux, utilisée pour optimiser le dimensionnement des ouvrages. Les résultats de la confrontation modélisation–expérimentation sont présentés et analysés tout comme les résultats de l'étude paramétrique qui ont permis de dégager des recommandations techniques quant à la conception de ce type d'ouvrage.