Investigations on secondary compression behaviours of artificial soft sand-clay mixtures

International audienceSpecial soft soils with notable coarse and fine fractions are not only naturally deposited in ravines or interactive marine-terrestrial sedimentation areas but also artificially formed in land reclamation projects via the hydraulic filling method and excavation cuts of different soil layers in open pit mines. The behaviour of these soils is not like that of similar normal sands or clays but rather exhibits the characteristics of transitional mixtures. A sand-clay mixture was prepared in the laboratory to simulate these special soil types, and an incremental loading oedometer test was performed. Thereafter, a secondary compression coefficient (Cα) was determined to investigate the long-term rheological properties of the mixture. The results in the Cα-log σv′, Cα-e, Cα-e/eL and Cα-Cc planes revealed that a threshold e/eL value (i.e. approximately 0.45) in the case of sands. Note that in a traditional Cα-Cc analysis system, the soils with sand exhibited the same secondary compression behaviour (SCB) as that without sand before reaching the transformation threshold, whereas after this threshold, they appeared to behave much like sandy soils. Hence, a four-phase frame of sand-clay mixture was further introduced to clarify the mechanism of the SCB of soils with and without sand. An updated Cαc/Ccc reflecting the creep mechanism of the clay phase can be effectively adopted to exclude the influence of sand particles. Because the quantitative relationship between Cαc and Ccc for soils with sand was consistent with that without sand, it can be assumed that the SCB of clay-sand mixtures was dominated by the creep of fine fractions. It is porposed that the existence pattern of the sand particles in the mixtures undergoes three stages with an increasing vertical stress, i.e. non-skeleton, partial skeleton and complete skeleton. A further discussion on the mutual relationship between phases on the behaviour of mixtures with sand revealed that isolated sand had no effect on the compressibility and SCB during the initial stage. When a partial skeleton of the sand was established in the second stage, characterized by less stress on the clay matrix, the compressibility and SCB were suppressed. When the vertical stress continuously increased or additional sand content was supplied, a complete skeleton formed. Hence, the behaviour of mixtures (regardless of compressibility and secondary compression) was completely governed by the sand fractions. Generally, the same relationship was observed between Cαc and Ccc during the first and second stages with sand, indicating that the influence of sand fractions on the compressibility and SCB can be counterweighted