By controlling the particle size distribution of an Italian silt, the influence of grading on its behaviour was investigated. As the clay content was reduced, the behaviour changed from a typical clay mode to a transitional form between that of clays and sands that had previously been seen only for gap-graded soils, emphasising that this type of behaviour is possibly much more extensive than previously thought. The work has highlighted several features of transitional behaviour, and in particular that unique normal compression and critical state lines do not exist, that Rendulic's principle does not apply, and that drained and undrained tests do not reach unique ultimate states.En contrôlant la répartition des dimensions de particules dans un limon italien, nous avons étudié l'influence de la granulométrie sur son comportement. À mesure que la teneur en argile baissait, le comportement changeait, passant d'un mode argileux type à une forme transitionnelle entre celle d'argiles et celle de sables -forme vue auparavant uniquement dans les sols gradés à espaces -, ce qui montre que ce type de comportement est peut-être bien plus étendu que préalablement estimé. Ce travail a fait apparaître plusieurs caractéristiques de comportement transitionnel et en particulier le fait que la compression normale unique et les lignes d'état critique n'existent pas, que le principe de Rendulic ne s'applique pas et que les essais drainés et non drainés n'atteignent pas des états ultimes uniques.

An analysis was made of an extensive series of oedometer and triaxial tests on intact samples of silts and silty clays from the lagoon of Venice, as well as a more limited series of carefully controlled oedometer tests on reconstituted specimens. The soils appear to have transitional modes of behaviour with compression curves that converge only slowly in the v–log [Formula: see text] plane and critical state lines defined for triaxial compression that are non-unique in the v–ln p′ plane. Two means of quantifying the degree of transitional behaviour are introduced, one factor, m, being based on the oedometer data and the second, P, on the critical state data. The values of m were in general lower than those of P, because of the higher stress level considered. The values of m were also found to depend on unloading, which must be accounted for when comparing data from intact and reconstituted specimens.

No abstract

To investigate the hydraulic behaviour of the fine-grained embankments along the mid-course of the Po River, research was carried out on a full-scale physical model, built on the floodplain along the existing embankment and forming a pond. The pond was filled to reproduce historical floods. The prototype was built according to recommendations formulated by the Po River Management Authority. The engineering properties of the foundation soils were investigated by in situ tests and complemented by some suction-controlled laboratory tests. Pore-water pressure was measured in the embankment and in its foundation before, during, and after the experimental reproduction of two floods that occurred in 1976 and 2000. Atmospheric variables were monitored at the prototype site. Monitoring data refer to 6 months of experimental activities. Pore-water pressure measurements were first presented and subsequently interpreted through an isothermal two-dimensional flow approach where boundary flow rates express the evolution of atmospheric variables. The distribution of the pore water provided by this interpretation was used for stability analyses of the prototype. Results from the experimental activities and their interpretation indicate that the river embankment remains extensively unsaturated during floods. Typical suction trends within the embankment, suction-associated soil strength, and the resulting slope stability safety factors are also shown

No abstract