In geotechnical engineering, the development of efficient and accurate constitutive models for granular soils is crucial. The micromechanical models have gained much attention for their capacity to account for particle-scale interactions and fabric anisotropy, while requiring far less computational resources compared to discrete element method. Various micromechanical models have been proposed in the literature, but none of them have been conclusively shown to agree with the critical state theory given theoretical proof, despite the authors described that their models approximately reach the critical state. This paper modifies the previous CHY micromechanical model that is compatible with the critical state theory based on the assumption that the microscopic force-dilatancy relationship should align with the macroscopic stress-dilatancy relationship. Moreover, under the framework of the CHY model, the fabric anisotropy can be easily considered and the anisotropic critical state can be achieved with the introduction of the fabric evolution law. The model is calibrated using drained and undrained triaxial experiments and the results show that the model reliably replicates the mechanical behaviors of granular materials under both drained and undrained conditions. The compatibility of the model with the critical state theory is verified at both macroscopic and microscopic scales.

The stress state and density of soil have been considered as the key factors to determine the liquefaction resistance. However, the results of seismic liquefaction case histories, laboratory tests and centrifuge model tests show that the fabric characteristics also influence liquefaction resistance, even more significantly than the contributions of stress state and density. In this study, anisotropic specimens with different consolidation histories were prepared using the 3D Discrete Element Method (DEM) to investigate the influence of fabric characteristics on the mechanical behavior of granular materials and the underlying mechanisms. The simulations revealed that under monotonic shear conditions, horizontally anisotropic specimens exhibited strain hardening and dilatancy characteristics, as well as higher peak strength. Under cyclic shear condition, the normalized liquefaction resistance of the specimens showed a strong linear relationship with the degree of anisotropy, independent of confining pressures and density. Microscopic results indicate that the fabric arrangement aligned with the loading direction leads to the evolution of the mechanical coordination number and average contact force in a manner favorable to resisting loads, which is the underlying mechanism influencing macroscopic mechanical properties. Additionally, the evolution patterns of contact normal magnitude and angle in anisotropic granular materials under cyclic loading conditions were also analyzed. The results of this study provided a new perspective on the macroscopic mechanical properties and the evolution of the microstructure of granular soils under anisotropic conditions.

The macroscopic mechanical properties of granular systems largely depend on the complex mechanical responses of force chains at the mesoscopic level. This study offers an alternative to rapidly identify and predict force chain distributions under different stress states. 100 sets of gradation curves that effectively represent four typical continuous gradation distributions are constructed. Numerical specimens corresponding to these gradation curves are generated using the discrete element method (DEM), and a dataset for deep neural network training is established via biaxial compression numerical simulations. The relationship between particle distribution characteristics and force chain structure is captured by the Pix2Pix conditional generative adversarial network (cGAN). The effectiveness of the generated force chain images in reproducing both particle gradation and spatial distribution characteristics is verified through the extraction and analysis of pixel probability distributions across different color channels, along with the computation of texture feature metrics. In addition, a GoogLeNet-based prediction model is constructed to demonstrate the accuracy with which the generated force chain images characterize the macroscopic mechanical properties of granular assemblies. The results indicate that the Pix2Pix network effectively predicts and identifies force chain distributions at peak stress for different gradation

The study of macroscopic discrete granular materials holds significance in hydraulic engineering, geotechnical engineering, as well as road and bridge engineering. Its foundational scientific exploration bears profound theoretical implications and is of pivotal practical value to engineering endeavors. Within the realm of engineering construction, issues such as dam breakages, earth-rock dam damage, and geological disasters involving loose particles pose substantial threats to the safety of both national livelihoods and property. Thus, delving into the examination of the structural stability of granular materials at the mesoscopic scale becomes an imperative pursuit. In this study, the topological structure of granular materials is identified and segmented based on image processing techniques, and the relationship between the compressive capacity of polygonal structures and the number of polygonal sides is studied. The redundancy function is defined to evaluate the structural stability of granular materials. In addition, the definition of structure tensor is introduced, and redundancy and structure tensor are applied to the study of biaxial compression of shale materials. The research results contribute to improving engineering safety and have guiding merits for the research and application of granular materials. Future work could focus on extending these methods to other types of granular materials and exploring their behavior under different loading conditions. (c) 2025 Published by Elsevier B.V. on behalf of Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy of Sciences.

This paper presents a comprehensive study on the evolution of the small-strain shear modulus (G) of granular materials during hydrostatic compression, conventional triaxial, reduced triaxial, and p-constant triaxial tests using 3D discrete element method. Results from the hydrostatic compression tests indicate that G can be precisely estimated using Hardin's equation and that a linear correlation exists between a stress-normalized G and a function of mechanical coordination number and void ratio. During the triaxial tests, the specimen fabric, which refers to the contact network within the particle assembly, remains almost unchanged within a threshold range of stress ratio (SR). The disparity between measured G and predicted G, as per empirical equations, is less than 10% within this range. However, once this threshold range is exceeded, G experiences a significant SR effect, primarily due to considerable adjustments in the specimen's fabric. The study concludes that fabric information becomes crucial for accurate G prediction when SR threshold is exceeded. A stiffness-stress-fabric relationship spanning a wide range of SR is put forward by incorporating the influences of redistribution of contact forces, effective connectivity of fabric, and fabric anisotropy into the empirical equation.

Moisture accumulation within road pavements, particularly in unbound granular materials with or without thin sprayed seals, presents significant challenges in high-rainfall regions such as Queensland. This infiltration often leads to various forms of pavement distress, eventually causing irreversible damage to the pavement structure. The moisture content within pavements exhibits considerable dynamism and directly influenced by environmental factors such as precipitation, air temperature, and relative humidity. This variability underscores the importance of monitoring moisture changes using real-time climatic data to assess pavement conditions for operational management or incorporating these effects during pavement design based on historical climate data. Consequently, there is an increasing demand for advanced, technology-driven methodologies to predict moisture variations based on climatic inputs. Addressing this gap, the present study employs five traditional machine learning (ML) algorithms, K-nearest neighbors (KNN), regression trees, random forest, support vector machines (SVMs), and gaussian process regression (GPR), to forecast moisture levels within pavement layers over time, with varying algorithm complexities. Using data collected from an instrumented road in Brisbane, Australia, which includes pavement moisture and climatic factors, the study develops predictive models to forecast moisture content at future time steps. The approach incorporates current moisture content, rather than averaged values, along with seasonality (both daily and annual), and key climatic factors to predict next step moisture. Model performance is evaluated using R2, MSE, RMSE, and MAPE metrics. Results show that ML algorithms can reliably predict long-term moisture variations in pavements, provided optimal hyperparameters are selected for each algorithm. The best-performing algorithms include KNN (the number of neighbours equals to 15), medium regression tree, medium random forest, coarse SVM, and simple GPR, with medium random forest outperforming the others. The study also identifies the optimal hyperparameter combinations for each algorithm, offering significant advancements in moisture prediction tools for pavement technology.

Particle segregation is a widespread phenomenon in nature. Vertical vibration systems have been a focal point in studying particle segregation, providing valuable insights into the mechanisms and patterns that influence this process. However, despite extensive research on the mechanisms and patterns of particle separation, the consequences, particularly the mechanical properties of samples resulting from particle segregation, remain less understood. This study aims to investigate the segregation process of a binary mixture under vertical vibration and examine the consequences through monotonic and cyclic triaxial drained tests. The results reveal that large and small particles segregate nearly simultaneously, with more thorough separation observed for large particles. The segregation index, Ds, effectively describes this evolution process, offering a quantitative metric for both mixing and segregation. Granular temperature analysis unveils three distinct states during segregation: solid-like, fluid-like, and solid-liquid transitional phase, corresponding to varying activity levels of particle segregation. Drained triaxial shear tests demonstrate the sensitivity of stress-strain relationships to the degree of segregation. Interestingly, ultimate strength is found to be essentially unrelated to the degree of segregation. When the segregation index approaches zero, signifying particles approaching a uniform distribution, the granular system reaches a harmonic state. This state exhibits optimal mechanical performance characterised by maximum peak stress, friction angle, and the highest elastic modulus. These findings underscore the potential impact of segregation on the mechanical response of granular mixtures and emphasise the necessity of a comprehensive understanding of particle segregation in soil mechanics.

The rheological properties and creep dynamical behavior of the granular materials are significantly influenced by the packing fraction. The granular materials with a low packing fraction tend to transit from a solid-like to liquid-like state. The strain evolution and deformation characteristics of granular materials under different packing fractions are investigated by triaxial creep tests. The result indicates that a critical packing fraction exists for the granular system under specific external loading conditions, below which the system will be broken in a short period of time. Conversely, for packing fraction that exceeds the critical value, the granular material system exhibits logarithmic creep dynamics and eventually reaches a steady state. To characterize the creep behaviors of granular materials under dynamic loading, a state evolution model is introduced. The model is verified by combining the theoretical predictions with the experimental observations. Furthermore, parametric analysis is also implemented based on the introduced model. The results demonstrate that the model can capture the fundamental spatiotemporal evolution characteristics of granular materials which are subjected to dynamic loading conditions.

The recycling and reuse of construction and demolition materials offer significant environmental and economic benefits. This study investigated the performance of aggregates with varying proportions of recycled concrete aggregate (RCA) and natural aggregate (NA) as base materials. Key parameters, such as compaction behaviour, California bearing ratio (CBR), and resilient modulus, were evaluated. The findings revealed that the mixture with 50% RCA and 50% NA exhibited the highest CBR and resilient modulus values. Small-scale cyclic loading tests were then conducted on the samples of NA, RCA, and a 50% RCA-50% NA mixture to assess the suitability of RCA as a base material. Additionally, RCA and NA samples were reinforced with biaxial geogrids for material optimisation. The results showed that the 50% RCA-50% NA mixture exhibited the smallest permanent deformation, and the geogrid, placed at the middle depth of the base, significantly reduced rut depth. Findings of this experimental study suggest that RCA can be used as an alternative base material to partially replace NA in road construction. The results can help conserve natural resources, promote sustainability through the reuse of waste materials, and reduce the environmental impact associated with the use of NA in road construction.

The use of ordinary Portland cement for the stabilisation of granular materials in road construction undermines the effort on sustainability made by using recycled aggregate in substitution of natural ones. This requires the use of low-impact binders so that the road construction industry complies with the prevailing environmental regulations. This study compares the mechanical and environmental properties of construction and demolition waste (CDW) aggregates stabilised with different binders: (i) a Portland-limestone cement as a reference, (ii) a pozzolanic cement, (iii) an experimental pozzolanic cement containing waste clay from the lightweight aggregate production, and (iv) a binder with alkali-activated CDW fines. In the laboratory experiments, both strength and resilient properties were considered, while the environmental impact was assessed in a cradle-to-gate scenario through a life cycle analysis (LCA). The stabilised mixture with pozzolanic cement achieved comparable strength and stiffness while exhibiting a lower environmental impact than the mixture containing Portland-limestone cement. The addition of waste clay to the pozzolanic cement significantly reduces its environmental impact albeit more binder is required to compensate for the lower mechanical properties. The alkaline activation of the fine particles in the CDW aggregate enabled the creation of a stabilised mixture with high strengths and resilient modulus. However, this alternative stabilisation technique requires further optimisation to mitigate the significant environmental impact. The engineering evaluations of the stabilised granular mixtures studied have considered both mechanical and environmental factors intending to contribute to the scientific debate on how to make roadworks sustainable and conserve natural resources.