Assessing the undrained strength cross-anisotropy of three tailings types

The cross-anisotropic nature of soil strength has been studied and documented for decades, including the increased propensity for cross-anisotropy in layered materials. However, current engineering practice for tailings storage facilities (TSFs) does not appear to generally include cross-anisotropy considerations in the development of shear strengths. This being despite the very common layering profile seen in subaerially-deposited tailings. To provide additional data to highlight the strength cross-anisotropy of tailings, high quality block samples from three TSFs were obtained and trimmed to enable Hollow Cylinder Torsional Shear tests to be sheared at principal stress angles of 0 and 45 degrees during undrained shearing. Consolidation procedures were carried out such that the drained rotation of principal stress angle that would precede potential undrained shear events for below-slope tailings was reasonably simulated. The results indicated the significant effects of cross-anisotropy on the undrained strength, instability stress ratio, contractive tendency and brittleness of each of the three tailings types. The magnitude of cross-anisotropy effects seen was generally consistent with previous published data on sands.

Pure cross-anisotropy for geotechnical elastic potentials

The pure cross-anisotropy is understood as a special scaling of strain (or stress). The scaled tensor is used as an argument in the elastic stiffness (or compliance). Such anisotropy can be overlaid on the top of any elastic stiffness, in particular on one obtained from an elastic potential with its own stress-induced anisotropy. This superposition does not violate the Second Law. The method can be also applied to other functions like plastic potentials or yield surfaces, wherever some cross-anisotropy is desired. The pure cross-anisotropy is described by the sedimentation vector and at most two constants. Scaling with more than two purely anisotropic constants is shown impossible. The formulation was compared with experiments and alternative approaches. Static and dynamic calibration of the pure anisotropy is also discussed. Graphic representation of stiffness with the popular response envelopes requires some enhancement for anisotropy. Several examples are presented. All derivations and examples were accomplished using the algebra program Mathematica.

Subsurface settlements of shield tunneling predicted by 2D and 3D constitutive models considering non-coaxiality and soil anisotropy: a case study

Appropriate constitutive models and reliable excavation and support sequences are believed to be the major concern in using Finite element (FE) analysis to simulate shield tunnel excavation. This paper presents systematic 2D and 3D FE analyses employing a number of constitutive models accounting for initial soil anisotropy and non-coaxial plasticity, as evidenced within site investigations from the Tsinghuayuan Tunnel of the Jing-Zhang High-Speed Railway in China. The aim is to assess the effects of both the initial soil anisotropy and non-coaxiality on longitudinal and transverse tunneling-induced surface settlements. It is shown that the excavation procedures combined with the degree of cross-anisotropy are key towards the accurate prediction of maximum vertical displacements from tunneling, matching field data. Knowledge of the initial soil strength anisotropy can further improve the shape prediction of the transverse tunneling-induced surface settlement troughs. When considering n = 0.6 and "β =" 〖" 0" 〗^"°" in simulations, the transverse surface settlement trough obtained is almost coincided with monitored field data. Initial stiffness anisotropy used in the prediction of shield tunnel-induced surface settlements in sandy pebble soils, does improve realism of results significantly. The maximum longitudinal settlement predicted by considering cross-anisotropy is larger than that predicted by its isotropic counterpart.

Consideration of the Cross-Anisotropy of Different Materials of the Asphalt Surface Layer and Temperature to Determine Pavement Responses

The effects of the cross-anisotropy of different materials of the asphalt surface layer and the depth-temperature relationship on pavement responses and damage are investigated. A three-dimensional Finite-Element Model (FEM) of the pavement, which considers the depth-temperature relationship of the surface layer under moving tire load, is developed. Pavement damage models are established to evaluate the damage ratio for primary rutting and fatigue cracking. The results show that the compressive strain at the bottom of the surface layer increases as the temperature increases, and the cross-anisotropy (n-value) decreases, indicating that a decrease in the horizontal modulus of different materials of the surface layer increases the damage ratio for primary rutting at high temperatures. The tensile strain at the bottom of the surface layer declines as the n-value increases to 1. For the same change in the n-value, the rate of change of the damage ratio for fatigue cracking is greater at low temperatures than at high temperatures, demonstrating that the number of allowable load repetitions is more sensitive at low temperatures. In addition, the effect of cross-anisotropy and temperature on the vertical stress are larger on the top of the base than in the subbase and subgrade.

Field Characterization of Pavement Materials using Falling Weight Deflectometer and Sensor Data from an Instrumented Pavement Section

This study deals with the backcalculation of the mechanical properties of pavement layers using not only the falling weight deflectometer (FWD) sensor data but also pavement response under that FWD load from the embedded sensors in an instrumentation section. To perform the backcalculation, a layered viscoelastic pavement model incorporating asphalt concrete (AC) cross-anisotropy is developed as the forward model. Field degree of cross-anisotropy in AC is determined at the maximum magnitude frequency obtained through continuous wavelet transform of the material response signal. The material response signal is obtained from the deconvolution between the loading signal and the signal registered at the embedded sensors. An inverse analysis methodology is also developed to calculate the gross vehicle weight (GVW) of any vehicle passing through the instrumentation section using only the backcalculated material properties and pavement responses. From the results, it is observed that inclusion of the AC cross-anisotropy reduces the error norm, and a good agreement is observed with the laboratory dynamic modulus in both horizontal and vertical directions. It is also observed that the maximum magnitude frequency of material response and degree of cross-anisotropy in AC both decrease with an increase in average AC temperature. Furthermore, using the backcalculated material properties and pavement responses, it is possible to determine the GVW with 95% accuracy.