Experimental Studies on The Mechanical Behaviour of Cohesive Frictional Granular Materials

Abstract

Thss thesis presents the results of an experimental programme on the static mono-tonic response of cohesive-frictional granular materials. The purpose of this experimental programme was to gain insight into the mechanical behaviour of uncemented sands, and sands with small percentages of cementation. With this objective in sight, the research involved understanding and delineating the e ects of four variables: the intermediate principal stress, stress inclination, cohesion (or cementation), and particle morphology. The hollow cylinder torsion (HCT) apparatus, which allows control over both the magnitude and direction of principal stresses, was used in this study to carry out a series of elemental tests on the model materials. The test results were analysed in a plasticity theory based framework of critical state soil mechanics. Drained and undrained HCT tests were conducted on a model angular sand to understand the combined influence of intermediate principal stress ratio (b) and principal stress inclination ( ). Sand specimens were reconstituted to a given density and confining pressure, and were sheared to large strains towards a critical state. The stresses at the critical state with varying `b' were mapped on an octahedral plane to obtain a critical state locus. The shape of this locus closely resembles a curved triangle. Also these specimens showed increased non-coaxiality between the stress and strain increment directions at lower strains. This non-coaxiality decreased significantly, and the response at the critical state was by and large coaxial. The effect of `b' and ` ' on the flow potential, phase transformation, and critical state was also investigated. At phase transformation, ` ' plays a more dominant role in determining the flow potential than `b'. The shape and size of the critical state locus remained the same immaterial of the drainage conditions. Next, small amounts of cohesion (using ordinary Portland cement) was added to this sand ensemble to study the mechanical behaviour of weakly cemented sands. The peak in the stress strain curve was used to signal the breakdown of cohesion further leading to a complete destructuring of the sand at the critical state. The response of the cemented sand changes from brittle to ductile with increase in confining pressure, while reverses with increase in density and `b'. Stress-dilatancy response for the weakly cemented materials shows the non coincidence of peak stress ratio and maximum value of dilation unlike purely frictional materials. This mismatch in peak stress ratio and maximum dilation diminishes with increase in confining pressure. The peak stress (cemented structured sand) locus and the critical state (destructured) locus were constructed on the octahedral plane from these HCT tests. The critical state locus of the cemented sand when it is completely destructured almost coincides with the critical state locus of the clean sand. Using this experimental data set, some important stress-dilatancy relationships (like Zhang and Salgado) and failure criteria (Lade's isotropic single hardening failure criteria and SMP failure criteria) were benchmarked and their prediction capabilities of such models were discussed in detail. The effect of particle morphology was also investigated in this testing programme. Rounded glass ballotini and angular quartzitic sand which occupy two extreme shapes were selected, and a series of HCT tests at different `b' values were con-ducted. A larger sized CS locus was obtained for angular particles and it encompassed the critical state locus of the spherical glass ballotini. Spherical particles exhibit a predominantly dilative behaviour, however present a lower strength at the critical state. The mobilization of strength as a result of rearrangement of angular particles and the consequent interlocking is higher. Even with contractive behaviour which is reflected in the higher values of critical state friction angle and the larger size of the yield locus for sand. Finally, a series of unconfined compression tests were performed to understand if there exists a scale separation in cohesive frictional materials. Specimens were reconstituted to a range of sizes while maintaining a constant aspect ratio and density. As the specimen size increased, the peak strength also increases, counter to an idea of a generalized continuum for all model systems. The observed secondary length scale (in addition to the continuum length scale) is obverse to the one observed in quasi-brittle materials such as concrete, rock. In order to ascertain the reason behind this phenomenon, a series of tomography studies were carried out on these contact-bound ensembles. The presence of cohesion between the grains brings about an \entanglement" between the grains, which contributes to increase in strength, with increase in the size of the sample. This in e ect bringing forth a second length scale that controls the behaviour of these cohesive frictional granular materials. This experimental data set provides quantification of various aspects of the me-chanical response of both cemented and uncemented granular materials under myriad stress conditions. This data set is also extremely useful in developing and bench-marking constitutive models and simulations.