Understanding Strain Rate Response of Geomaterials through Orthogonal Cutting Experiments

Abstract

Cutting is a complex, large deformation process used for the removal of material during various engineering operations. In geomaterials, such as soils and rocks, the mechanism of cutting is encountered during operations such as ploughing, trenching, and excavation. This thesis presents the results of an experimental programme designed to understand the effect of strain rate on the mechanical response of three model geomaterials. A laboratory-scale physical model of in-plane orthogonal cutting was used in this suite of cutting experiments. In order to systematically investigate the geomaterial-tool interaction, three model geomaterial systems were chosen: clean quartz sands, cemented sands, and sands mixed with laponite, as they broadly span the characteristic geomaterial types commonly encountered in engineering soil cutting operations. These model materials allowed the controlled simulation of loose granular soils, natural or artificially cemented ground, and cohesive-viscous mixtures, respectively, and hence enabled the study of the influence of mechanical properties, such as cohesion, cementation, and viscous effects, on cutting behaviour in more practical scenarios. The detailed characterization of these model materials and protocols for preparing homogeneous specimens suitable for imaging were created in this research. The orthogonal cutting configuration allowed exploring a range of strains and strain rates by appropriately changing the boundary conditions. Complimentarily, images at high speed and high resolution were captured during these suites of experiments. Particle image velocimetry (image correlation) algorithms hitherto developed were improved to capture crucial features during deformation at various strain rates. A series of orthogonal cutting experiments was performed on clean quartz sand to understand the mechanics of severe plastic deformation during orthogonal cutting. The cutting experiments were performed on sands over 3 orders of strain rates. The velocity field maps of the region around the tool tip reveal a sharp change in the motion of sand particles, accompanied by the formation of a dead zone. The corresponding effective strain rate maps reveal regions of intense localized plastic deformation, termed “shear bands”. The inclination angle of these bands evolved periodically with time and showed a decreasing trend due to an increase in the surcharge and effective depth of cut. The morphology and overall characteristics of these shear bands do not change significantly with strain rate. The cutting force signatures exhibit fluctuations indicative of softening, alternating regions of contraction and dilation ahead of the tool, which is reflected in the periodic repositioning of shear bands. A limit equilibrium-based model was adequate to predict the tool-cutting forces, even with large variations in strain rates. A second set of orthogonal cutting experiments was performed to understand the response of weakly cemented sands to severe plastic deformation during orthogonal cutting. Orthogonal cutting tests were conducted on 4% and 8% artificially reconstituted model soft rocks (i.e., weakly cemented sand specimens). The model soft rocks were also extensively characterized through both compression and tensile tests over a wide range of loading rates. The experiments allowed a critical examination of the initiation and propagation of regions of localization or fractures. During the cutting process, the material ahead of the tool undergoes localized compaction, followed by the formation of a major fracture and progressive fragmentation of the deformed material, which occurs periodically. The force signatures showed distinct peaks, indicating the occurrence of major fractures. The damage to the material, defined as a percentage de-structuring, was quantified utilizing the grain size distribution of the cut material accumulated in front of the tool. Damage and specific energy reduced by ∼25% and ∼13%, respectively, with increasing cutting depth by twice, while both only marginally increased when the cutting speeds increased by 4 orders. By doubling the amount of cementation from 4% to 8%, although the damage was reduced marginally, the specific energy increased substantially by 2 to 3 times. The force estimates from three theoretical models were compared to the experimental force values, and it was found that the brittle tensile failure-based model, with an arc-shaped rock failure surface, provided the best force estimations, among others with planar failure surfaces. In the third and final suite, cutting tests were carried out on saturated laponite mixed sands. The material system was prepared by adding sand to a laponite suspension, followed by 72 hours of ageing under a moist environment. The effect of adding a viscous binder (laponite) to a sand matrix is studied using in-plane cutting. The cutting tests are performed at different depths and cutting speeds with concurrent imaging and recording of force components. PIV analysis revealed multiple regions of flow bifurcation. The effective strain rate and volumetric strain rate fields were calculated, revealing diffuse bifurcation regions that separate the soil into wedges/clods from the bulk. The formation of shear bands followed by the separation of the soil wedge was observed to occur periodically during cutting. Interestingly, intense shearing is also observed at the fracture plane, originating from the free surface and curving upward to the tool rake face. At both these regions of shearing, dilation is observed in the volumetric strain rate fields. The periodic occurrence of peaks and drops in force evolution during cutting is observed, where significant compaction/hardening of the soil near the tool and above the shear band is evident before the force peak. Conversely, the mobilization of new shear bands results in a drop in the magnitude of the force. The increase in depth has more pronounced effects on the force responses than the rate effects. The test results from the three material systems are compared in order to understand the influence of the binder on the mechanics of orthogonal cutting. The velocity field, strain rate fields, and force results were compared under the same boundary conditions. The deformation in clean sands during orthogonal cutting is dominated by the formation of shear bands, by fracturing and accumulated damage in soft rocks, and by almost a ductile chip in the case of laponite sands. In brittle material of cemented sands, the force peak -drop behaviour corresponds to the formation of major fracture planes and subsequent loss of partial tool engagement, while in relatively ductile materials of sands and laponite sands, the force peaks at the hardening of the soil wedge and drops at the subsequent formation of new shear bands. Clean sands and cemented sands are rate-insensitive, unlike laponite sands.