The reuse of waste foundry sand (WFS) as a sustainable alternative to natural sand in geotechnical backfill applications necessitates a thorough understanding of its behaviour under dynamic loading. This study presents an experimental evaluation of the dynamic and liquefaction behaviour of using cyclic triaxial testing. A comprehensive test programme was conducted on cylindrical specimens of 50 mm dia. and 100 mm height under stress-controlled cyclic loading. Tests were performed over a range of loading frequencies (0 & centerdot;1-5 Hz) and cyclic stress ratios (CSR= 0 & centerdot;15-0 & centerdot;25) to simulate seismic and traffic-induced loading conditions. For comparison, selected tests were also carried out on natural silica sand under identical conditions. The response of was assessed in terms of excess pore water pressure generation, cyclic deformation characteristics and the number of cycles to liquefaction. The results indicate a clear influence of loading frequency and on the liquefaction resistance of WFS. Compared with silica sand, exhibits comparable dynamic performance. This study can also be useful for assessing seismic response and stability analysis of landfills.
Stabilising inclined composite roofs in coal mine roadways has long been a critical challenge in mining engineering. Traditional support systems often fail to prevent interlayer shear deformation, leading to cable shear failure, large roof deformations, and even collapse. To address these issues, this study proposes a spatially optimised anchor-grouting combined support system integrating space-interlaced anchoring and open grouting method for inclined composite roofs. Steel sleeves are configured on cables to improve the shear resistance. And a spatially sequential layout of variable-length cables is adopted, which enables anchoring into different rock strata to improve stress distribution. In addition, grouting capsules are employed to achieve efficient grouting and simplified construction without shotcreting. Through mechanical analysis and field implementation in a deep experimental roadway, the proposed support system is validated. Monitoring results show that the maximum roof displacement is reduced from 315 mm (traditional support) to below 105 mm (optimised support), and no cable shear failure occurs.
The increasing frequency of extreme rainfall necessitates slope protection methods that directly regulate infiltration. This study proposes engineered hydrophobicity as a delay-based infiltration control mechanism. Centrifuge model tests were conducted to evaluate slopes containing hydrophobically treated soil layers under prototype-scale rainfall (90 mm/h). Compared to untreated slopes, the hydrophobic layers delayed wetting-front progression and pore-water pressure buildup. Although thicker hydrophobic layers increased infiltration delay, they also reduced post-rainfall drainage efficiency and promoted water accumulation near the lower slope. Comparative tests using silica sand and recycled fine aggregates demonstrated the applicability of the method to different soil types, with recycled materials providing comparable resistance. These results indicate that engineered hydrophobic soils can function as delay-based infiltration-control layers under extreme rainfall conditions.
This study presents a proof-of-concept surrogate-assisted finite-element (FE) framework for multi-objective design optimisation of geosynthetic-reinforced pile-supported embankments (GRPSE), focusing on internal stability. A multilayer perceptron neural network was trained using data generated from an automated 2.5D full-scale model developed in through Python scripting. The design variables included pile spacing, pile diameter, gravel-platform friction angle, and embankment height. The trained surrogate was integrated into a multi-objective optimisation framework to identify the cheapest, safest, and most balanced designs using Pareto fronts in the cost-safety space. Optimised solutions were compared with those from the established CUR226 analytical framework. Results show that, for the cheapest and safest designs, the surrogate model provides cost and safety performance comparable to CUR226 while maintaining computational efficiency during optimisation. By capturing complex soil-structure interactions, it also identifies alternative balanced combinations of design variables. For the optimal design, the surrogate model achieves a higher factor of safety at comparatively lower cost. Although CUR226 remains effective, particularly when fill material properties govern the design, the surrogate framework offers added value for pile-soil interaction-informed optimisation. Future work should integrate CUR226 as guiding physics within neural network models trained on modest datasets for practical design.
Semi-probabilistic design codes, such as Eurocode 7, account for spatial variability in pile resistance through correlation factors applied to single-pile capacities. When piles are connected by a rigid cap, load redistribution reduces the likelihood that failure is governed by a single weak pile; however, this effect is represented through a fixed reduction in the correlation factors, thereby neglecting the additional contribution to characteristic resistance from larger pile numbers. A probabilistic–deterministic framework is proposed to evaluate the characteristic collapse domain of pile groups under eccentric vertical loads. Single-pile axial resistance is modelled as lognormally distributed and variability is propagated to the group by way of Monte Carlo simulation. Group collapse is described deterministically through limit analysis, capturing failure mechanisms and load redistribution induced by rigid pile caps. For each realisation of pile capacities, the corresponding collapse states are determined, and the characteristic collapse domain is defined as the 5th percentile of the ensemble, obtained by homothetic contraction of the mean domain. The resulting reduction factor for group resistance is smaller than for individual piles and decreases with group size, reflecting the increasing role of load redistribution. This provides a rigorous probabilistic interpretation and less conservative yet safe estimates of group capacity, offering a promising basis for future geotechnical design.
