Experimental Investigation on Unconfined Compressive Strength of Cemented Soils Admixed with Saturated GAC

This research investigated the compressive strength of cemented soils admixed with saturated granular activated carbon (GAC). The saturated GAC was obtained from the water filtration system. A series of unconfined compressive strength was performed on both compacted soil-cement specimens and compacted soil-GAC-cement specimens with GAC content of 30 percent. All specimens were prepared by compaction with energy equivalent to the modified Proctor test. The results from modified Proctor tests showed that the maximum dry unit weight and the optimum moisture content of soil-GAC sample was less than those of soil sample. From the unconfined compression tests, there was tiny development of strength for both types of specimens with cement content of 1 percent throughout the curing period of 28 days. For both types of specimens with cement content of 2 and 3 percent, the significant development of strength occurred after curing for 3 days. The strength of specimens typically increased with increasing cement content. Generally, the strength of compacted soil-GAC-cement specimens was less than that of compacted soil- cement specimens. It was also observed that the relationships between normalized compressive strength ratio and curing period was unique for the specimens with the same cement content.

Extending the Shields criterion to erosion of weakly cemented granular soils

This contribution tackles the issue of incipient conditions for initiation of erosion by a fluid flow at the surface of cohesive materials. To this end, a typical assessment procedure consists of subjecting a soil sample to progressive hydrodynamic stresses induced by a submerged impinging jet flow whose injection velocity is gradually increased. This paper presents the results of an extensive use of this protocol both in experiments and numerical simulations, the latter being based on a coupled DEM and LBM approach. Here we consider the specific case of weakly cemented soils, either made experimentally of glass beads bonded by solid bridges or modelled numerically by a solid bond rheology with a parabolic yield condition involving the micromechanical traction, shearing and bending of the bonds. The results show that, as expected, the hydrodynamic stress for erosion onset substantially increases with solid cohesion as compared to cohesionless cases but can, however, be satisfactorily predicted by a simple extension of the usual Shields criterion that only applies for cohesion-less granular sediments. This extension includes a cohesion number, the granular Bond number, with a simple definition based on tensile yield values.