Experimental observation of the behaviour of a 2D granular material with inclusions

The objective of this paper is to present the results of a series of laboratory tests carried out in a large (56 cmx 56 cm) shear apparatus on dense assemblies of 2D rods. The tested material consists of a matrix of PVC rods of three different diameters (1.5, 3 and 3.5 mm), in which a varying number of larger wooden rods (28 mmin diameter) were included. Three gap-graded materials were used, depending on the number of wooden inclusions adopted (in particular, inclusions were 0%, 30% and 40% of the volume of all rods). An average of more than 30000 such rods constitutes the specimen, and thus it is considered that the scale of continuum mechanics is reached. Along with conventional boundary stress and strain macroscopic measurements, stereophotogrammetry was used in some tests to directly measure the evolution ofthe displacement field at the scale of the single rod. In all tests, shear strain was observed to localise, and a certain number of shear bands formed in the tested specimen. By comparing the analyses carried out with a continuum and with a discrete approach, it is observed that the displacement field can be considered homogeneous in the specimen either at a very early stage of the test, prior to strain localisation, or at very large strains, when a net of multiple shear bands has fully developed. The results indicate that for the three adopted gradations the critical state iriction angle is unique, being insensitive to the percentage of inclusions. An increasing percentage of inclusions has the only effect of decreasing the initial total void ratio, which hides an increase in matrix void ratio. The friction angles at peak and at maximum contraction point depend in all the investigated cases on the matrix void ratio.