This study introduces an analytical approach to treat pile-to-pile interaction for axially loaded, end-bearing piles. Building upon an available Tajimi-type soil model and an application of Graf’s addition theorem, the proposed solution simultaneously considers both the source and the receiver piles. In particular, results are shown for interaction factors, and pile and soil displacements for the case of two piles, which can additionally be subjected simultaneously to different loads. The predictions are compared with results from available methods in the literature, and additional finite-element analyses. The proposed interaction factors are in excellent agreement with rigorous models. The displacement of a single pile without the receiver is straightforward. At the location of the receiver pile, the displacement is calculated as the average displacement around the perimeter of the pile at its head. This calculation accounts for the dependence of the displacement at the soil–pile interface on a relative angle, as determined by the application of Graf’s addition theorem. The factors for piles in vertically inhomogeneous soils are also shown, by using an equivalent stiffness of a fictitious homogeneous soil layer from the literature. This study lays the groundwork for extensions to irregular pile groups subjected to multi-directional inertial and kinematic loads in inhomogeneous soils.

Mine-tailing dams have failed abruptly and frequently in recent years. This is notwithstanding the development and application of site investigation and monitoring techniques. The aim of this study was to develop a robust electrical conductivity formula based on Archie’s equation for fine particles in partially saturated mine-tailing dams. Three specimens of uncoated silt, haematite-coated silt and samples collected from actual mine tailings (Minnesota silt) were prepared in a temperature-compensated cell installed with four electrodes. The test results showed that the saturation exponent increased with respect to the pore water conductivity up to 1 S/m. In addition, the saturation exponent showed higher values in the order of Minnesota silt, haematite-coated silt and uncoated silt owing to the different wettabilities. From the comparative analysis, uncoated and haematite-coated silt differed only in their specific surface areas. Meanwhile, the surface conduction term of Minnesota silt was influenced by both specific surface area and zeta potential. In addition, the newly defined Archie’s ‘ms’ considering the surface conduction term fits effectively with a high R2 value. Thus, the surface conduction term should be considered for estimating the porosity by measuring the electrical conductivity of iron mine tailings.

The mechanical behaviour of rooted soils is anisotropic. In this work, an anisotropic constitutive model for rooted soils is developed, incorporating two independent fabric tensors to represent the soil fabric and the root network. The effect of root tensile strength mobilisation on soil’s dilatancy and plastic hardening mechanism is addressed by introducing new fabric anisotropic variables (AB, AR), expressed as a joint invariant of fabric and loading direction tensors. A new root network evolution rule is proposed to capture the progressive root tensile strength mobilisation as the root orientation evolves towards the perpendicular direction of the major principal stress. The model is validated against test data. The model can predict the transition of sand’s strain-softening to hardening upon undrained triaxial extension due to the presence of roots. The model predicts lower strength and more contraction for rooted silty sands than the bare case during drained triaxial compression, as the former has a higher void ratio due to root inclusion and minimal root tensile strength mobilisation (indicated by a high AR and slow root network evolution rate). Conversely, during drained triaxial extension, the model exhibits higher shear strength and less contraction due to substantial root tensile strength mobilisation (indicated by a low AR).

Deep-sea mining operations now exceed water depths of 3000 m, necessitating a thorough understanding of the role of low temperature of the seabed (approximately 5°C) in sand behaviour under ultra-high hydrostatic pressure. This study presents the first systematic experimental investigation into the effects of temperature on the undrained behaviour of granular materials in an ultra-deep marine environment, employing a high-pressure triaxial apparatus designed for precise effective stress control. Strain-controlled monotonic and cyclic undrained triaxial tests were conducted on Fujian sand under a temperature (T) range of 5–55°C and a back-pressure (bp) of 29 MPa. Comparative experiments were conducted under conventional back-pressure (bp = 300 kPa) to isolate the temperature effects from those observed in terrestrial environments. Results revealed that specimens exposed to lower temperatures exhibit significantly larger and more rapid positive excess pore pressure (u) during both undrained monotonic and cyclic shearing. Notably, this temperature effect is nearly independent of bp, although specimens subjected to ultra-high bp demonstrate a higher u and reduced undrained strength. Furthermore, the critical state line in the p′–q plane remains unique, independent of temperature and back-pressure. In addition, a framework for analysing the state-dependent mechanical properties of sand was established considering varying T and bp conditions.

The migration of fine particles of gravel–sand mixtures subjected to suffusion plays a significant role in understanding the development of internal erosion-induced failure of hydraulic structures. This paper presents an experimental study on the progress of the suffusion of gravel–sand mixtures under three cyclic hydraulic gradient amplitudes (i.e. Δi = 0·25, 0·375, 0·5) and three average hydraulic gradients (i.e. imean = 0·75, 1·5, 2·25). Examination of the composition and origin of eroded particles is performed using particle staining and image recognition techniques. The results indicate that the higher amplitude of the cyclic hydraulic gradient leads to more pronounced particle loss channels, resulting in more loss of fine particles and more significant changes in hydraulic conductivity. The increased mean hydraulic gradient facilitates the development of new particle migration channels, leading to soil transition into the subsequent erosion stage. The loss of soil particles is primarily composed of fine particles ranging from 0·075 to 0·25 mm in size, occurring mainly during the initial stage of hydraulic gradient loading and at locations experiencing high hydraulic gradients. With increased cyclic gradient amplitude and the mean hydraulic gradient, suffusion gradually progresses from the top layer to the bottom layer of the soil. These findings can deepen the understanding of the characteristics and mechanisms of suffusion of gravel–sand mixtures.

Seismic geotechnical practice has embraced concepts and analyses inspired by pseudo-static thinking and force-based methodologies. By demonstrating that such thinking and analyses may grossly underestimate the true dynamic seismic behaviour, it is shown how they have led to the huge underestimation of seismic acceleration levels universally adopted for design in all past seismic codes. In foundation earthquake engineering the pseudo-static methodology is partly reflected in ‘capacity design’, whose objective is to prevent all possible foundation failure modes by imposing strict upper limits on forces and moments transmitted onto the foundation–soil system. Thus, failure mechanisms (‘plastic hinging’) can occur only in the above-ground structural members. This often results in over-conservative foundations while limiting the soil, footing, or pile deformations to quasi linear-elastic levels even under unexpectedly high accelerations – but, surprisingly, to the detriment of the structure. The paper demonstrates the benefits of drastically reversing this established seismic design philosophy. It is proposed not just to allow but to provoke the development of the ‘foundation uplifting and soil bearing capacity failure mechanisms’ of tall slender structures, by deliberately and substantially under-designing its foundations; thus they act as the ‘weak link’ in the support chain of the system. The resulting highly non-linear inelastic soil–foundation response prevents structural failure of the columns, and restrains the accelerations transmitted to the (super) structure. Using finite-element analyses and shaking table (1g and centrifuge) experiments, it is shown that, when subjected to seismic motions that exceed the code design, a typical bridge pier as well as a structural frame perform much better if supported on unconventional rocking than on conventional oversized foundations – in spite of (or in fact, because of) the latter’s two times larger area of contact. A potential price to pay, increased residual foundation settlement and sometimes rotation, is also addressed, with a number of suggested ways to ameliorate them. Analyses of two historic case histories support the proposed ‘rocking’ foundation design. First, the collapse of the 18 piers of Hanshin Expressway’s Route 3 in the 1995 Kobe earthquake, undoubtedly arising from structural deficiencies and unexpectedly huge accelerations, is shown to have been detrimentally affected from the fixity against rotation of the overdesigned 17-pile-capped foundation. Ironically, a surface spread footing could have saved the bridge through mobilisation of soil failure mechanisms. Second, in Adapazari, during the 1999 Kocaeli earthquake, the overturning of many five- to six-storey slender buildings, founded on the surface of very soft/loose saturated soil, is shown to be caused by bearing capacity failures (whether limited liquefaction did or did not occur). However, the same buildings alone, with no other adjoining building at their back, could have (and indeed have in some cases) stood almost vertically after shaking. It is concluded that with the foundation ‘allowed’ (even unintentionally) to work beyond conventional thresholds, the structure would not necessarily overturn and it would most likely be protected from uncontrollable damage during very strong excitation. Nevertheless, a caveat is appropriate: prudently, unconventional rocking foundations must be thoroughly scrutinised before being chosen for very tall and slender buildings on very soft soil; such structures have a propensity to excessive rotation that may be triggered by an accidental asymmetry in their foundation – asymmetry, due to several factors, not only to the presence of an adjoining building.

Under complex hydraulic conditions, such as extreme rainfall, water level fluctuations and wave surges, the hydraulic load exerted on the soil differs from constant or monotonically varying hydraulic gradient loads, resulting in more complex contact erosion responses between soil layers. In this study, a coupled computational fluid dynamics–discrete element method (CFD–DEM) was employed to investigate the effects of different mean hydraulic gradients and cyclic hydraulic gradient amplitudes. The research focused on the macroscopic deformation of contact erosion between soil layers under cyclic hydraulic gradients and the underlying microscopic mechanical mechanisms. The findings revealed that under cyclic hydraulic loading, the erosion mass of fine particles significantly increased, with fine particles closer to the contact surface being more susceptible to migration due to the cyclic hydraulic gradients. During the migration process, fine particles were more likely to pack and clog at the bottom of the coarse particle layer. This was primarily due to the increase in both the strength and number of contact forces perpendicular to the seepage direction. Seepage caused the contact forces and their distribution to develop more along the direction of seepage, exhibiting anisotropy. As both the mean hydraulic gradient and the cyclic hydraulic gradient amplitude increased, the particle erosion rate increased, while the shear strength and stability of the sample decreased.

Gravel soils are ubiquitous in nature, yet they pose significant challenges for laboratory testing and numerical modelling. Unlike pure soils, gravel soils comprise both fine-grained particles and gravel fragments, resulting in complicated behaviours affected by both the soil matrix and gravel. In this study, triaxial compression tests are conducted to investigate the mechanical behaviours of gravel soil with varying gravel shapes and contents. The failure mechanisms of gravel samples are further analysed through numerical simulation of plane strain tests using a hypoplastic model, considering various gravel contents, shapes, stress levels and soil densities. Both experimental and numerical results indicate that the presence of gravel enhances soil strength, as evidenced by increased stiffness and resistance. The angularity characteristics of gravel alter the failure patterns and propagation of the localised strains in the soil matrix. In addition, the enhancing effects caused by the initial dense state and high-stress condition contribute to the delayed progression of shear band formation.