Most of the robust artificial intelligence (AI)-based constitutive models are developed with synthetic datasets generated from traditional constitutive models. Therefore, they fundamentally rely on the traditional constitutive models rather than laboratory test results. Also, their potential use within geotechnical engineering communities is limited due to the unavailability of datasets along with the model code files. In this study, the data-driven constitutive models are developed using only laboratory test databases and deep learning (DL) techniques. The laboratory database was prepared by conducting cyclic direct simple shear (CDSS) tests on reconstituted sand, that is, PDX sand. The stacked long short-term memory (LSTM) network and its variants are considered for developing the predictive models of the shear strain (gamma [%]) and excess pore pressure ratio (ru) time histories. The suitable input parameters (IPs) are selected based on the physics behind the generation of ru and gamma (%) of the liquefiable sands. The predicted responses of gamma (%) and ru agree well in most cases and are used to predict the dynamic soil properties of the PDX sand. The same modeling framework is extended for other sand and compared with existing AI-based constitutive models to verify its practical applicability. In summary, it is observed that though the trained models predicted the time histories of ru and gamma reasonably well; however, they struggled to predict the hysteresis loops at higher cycles. Therefore, more research is needed to verify and enhance the predictability of existing AI-based models in the future before using them in practice for simulating cyclic response.

This paper presents a novel analytical framework to predict short-term pile setup in natural structure clay, considering the influence of soil destructuration in installation and consolidation. Based on the cavity expansion method, a simulation of pile installation has been conducted, with an analytical solution formulated for cavity expansion under undrained conditions to capture soil destructuration effectively. The flow rate in the unit cell is determined by Darcy's law based on the soil mass volume change, leading to the consolidation equation, which is obtained in a fully analytical form for excess pore water pressure (EPWP) dissipation. The utilization of the average compression curve aimed to depict a partially disturbed state due to the effects of installation. Based on the rewritten effective stress method (beta method), which involves the time-dependent factor while properly incorporating the effects of relaxation and thixotropy by introducing the requisite parameters. Finally, the analytical framework for predicting short-term pile setup is established and validated through a comprehensive pile field test conducted at St-Alban. The close correspondence between the analytical results and the empirical data indicates the effectiveness of the proposed framework in forecasting short-term pile behaviour with reasonable accuracy.

Subsidence from coal mining is a major environmental issue, causing significant damage to soil structure. Soil microorganisms, highly sensitive to environmental changes, adapt accordingly. This study focused on four areas of the Burdai coal mine: a non-subsidence area (CK), half-yearly (HY), 1-year (OY), and 2-year (TY) subsidence areas. Using high-throughput sequencing and molecular ecological network analysis, we examined soil microbial community diversity and structure across these zones, exploring microbial community assembly and functional predictions. Results showed that compared to the control, subsidence areas experienced reduced soil water content, organic matter, available phosphorus, and alkaline nitrogen, with the lowest levels observed at 1 year. These values began to rise after 1 year, suggesting natural recovery after subsidence stabilized. Microbial communities were closely related to soil organic matter, water content, and alkaline nitrogen. At the 1-year mark, soil property changes significantly reduced microbial diversity, which then began to recover after 2 years. The microbial network during 1-year subsidence was simpler, with 102 nodes, 179 edges, and an average degree of 3.51, indicating that early subsidence was unstable, and the microbial community was still adapting. By 1 year, community structure and interactions had begun to stabilize. Stochastic processes played a key role in microbial variability during short-term subsidence.

Soil-atmospheric boundary interaction is vital for the geotechnical design as the soil behaviour is moisture dependent, especially for expansive soils. Understanding the soil-atmospheric boundary interaction and the effect of climate change can be important for ensuring the resilience of geotechnical infrastructure. Thornthwaite Moisture Index (TMI) has been adopted in many geotechnical designs to account for the climate-induced moisture variations within the soil. The average TMI deducted from long-term climate data (continuous 25+ years) is often correlated with the design parameters such as the suction change depth and hence the characteristic surface movement. The behaviour of many structures can be influenced by shorter-term weather events and a shorter-term TMI may present a better correlation in such scenarios. However, high variability and the non-stationary character of a short-term TMI can be hindrances for any real-life application. This study assesses 1, 3, 6 monthly, and yearly TMI values estimated from 30 years of climate data. The result showed a significant difference exists between these monthly and annual average TMI values. This highlights the significance of incorporating short-term climate events and integrating climate change into geotechnical structures for the betterment of the built environment through safer, more resilient, and sustainable design.

The resilient modulus (MR) of different pavement materials is one of the most important input parameters for the mechanistic-empirical pavement design approach. The dynamic triaxial test is the most often used method for evaluating the MR, although it is expensive, time-consuming, and requires specialized lab facilities. The purpose of this study is to establish a new model based on Long Short-Term Memory (LSTM) networks for predicting the MR of stabilized base materials with various additives during wet-dry cycles (WDC). A laboratory dataset of 704 records has been used using input parameters, including WDC, ratio of calcium oxide to silica, alumina, and ferric oxide compound, Maximum dry density to the optimal moisture content ratio (DMR), deviator stress (sigma d), and confining stress (sigma 3). The results demonstrate that the LSTM technique is very accurate, with coefficients of determination of 0.995 and 0.980 for the training and testing datasets, respectively. The LSTM model outperforms other developed models, such as support vector regression and least squares approaches, in the literature. A sensitivity analysis study has determined that the DMR parameter is the most significant factor, while the sigma d parameter is the least significant factor in predicting the MR of the stabilized base material under WDC. Furthermore, the SHapley Additive exPlanations approach is employed to elucidate the optimal model and examine the impact of its features on the final result.

Biochar has been found to be an effective soil amendment in agriculture based upon its manifold functional groups as well as porous structure. However, the impacts of this material on soil mechanical properties are still poorly explored, especially under oscillatory shear conditions (as common due to traffic of agricultural machinery). Hence, our study investigates how short-term application of different rates and types of biochar in successive crops affects soil microstructural resistance, viscoelasticity, and resilience under oscillatory shear. In a completely randomized greenhouse pot experiment, wheat and soybean were grown successively in a sandy loam soil under single addition of two types of biochar (derived from either rice or soybean straw) at application rates (0 - control, 10 and 20 t ha-1). After crop harvesting, disturbed soil samples were collected in three layers to conduct amplitude sweep and thixotropy tests and analyze soil chemical properties. Biochar application resulted in extended elastic behavior, whereas soil strength decreased at low shear strain. Conversely, at high shear strain biochar had a destabilizing effect on soil microstructure, as indicated by the advancement of the flow point and lower overall viscoelasticity in biochar amended soils. Despite reduced microstructure stiffness exhibited in thixotropy tests, soil amended with biochar almost recovered completely its stiffness after high shear impact. However, significant effects were only noticed in topsoil layer independent of biochar type applied. Hence, accumulated biochar on soil surface layer had an overall negative impact on soil mechanical stability.

Hydromechanical behaviour of unsaturated expansive soils is complex, and current constitutive models failed to accurately reproduce it. Different from conventional modelling, this study proposes a novel physics-informed neural networks (PINN)-based model utilising long short-term memory as the baseline algorithm and incorporating a physical constraint (water retention) to modify the loss function. Firstly, a series of laboratory tests on Zaoyang expansive clay, including wetting and drying cycle tests and triaxial tests, was compiled into a dataset and subsequently fed into the PINN-based model. Subsequently, a specific equation representing the soil water retention curve (SWRC) for expansive clay was derived by accounting for the influence of the void ratio, which was integrated into the PINN-based model as a physical law. The ultimate predictions from the PINN-based model are compared with experimental data of unsaturated expansive clay with excellent agreement. This study demonstrates the capability of the proposed PINN in modelling the hydromechanical response of unsaturated soils and provides an innovative approach to establish constitutive models in the unsaturated soil mechanics field.

Cement-admixed clay (CAC) is a widely-used soil stabilization technique for enhancing the strength and stiffness of soft clay. However, the stress-strain behavior of CAC is complex and nonlinear, and also depends on various factors such as mixing proportion, confining pressure, stress path, and shearing condition. In this study, we propose a novel approach for modeling the stress-strain behavior of CAC using recurrent neural networks (RNNs), which are a type of deep learning (DL) technique that can well capture the temporal dependencies and nonlinearities in sequential data. We compare three types of RNNs: traditional RNN, long short-term memory (LSTM) neural network, and gated recurrent unit (GRU) neural network, and evaluate their performance in simulating the strain- and stress-controlled triaxial test results of 25 CAC specimens with different mixing proportions and confining pressures. The results demonstrate that the LSTM model, incorporating a 2-time step backward, exhibits superior prediction accuracy and generalization capability compared to other evaluated models, achieving a mean absolute percentage error (MAPE) of 4%. This LSTM model is capable of capturing the stress-strain behavior of CACs across various loading conditions and mixing proportions within a unified framework. Therefore, we suggest that the LSTM model is a promising tool for modeling and analyzing the mechanical behavior of CAC in geotechnical engineering applications.

Improper disposal of substantial scrap tires can result in significant environmental issues, such as air pollution, water resources, and soil contamination. The scrap tires can be turned into valuable materials by preparing tire rubber powder into Crumb rubber modified asphalt (CRMA). This method can effectively reduce environmental pollution and achieve the concept of energy conservation, environmental protection, and low-carbon. However, during the production process of CRMA mixtures, the elevated construction temperature may induce more severe short-term aging of the mixtures. The standard laboratory short-term aging scheme of asphalt binders, the Rolling Thin Film Oven Test (RTFOT), cannot simulate the short-term aging of CRMA due to the increase in construction temperature. Besides, the determination methods and indices of RTFOT aging temperature are still unclear and inconsistent. In this study, three CRMA binders were treated by RTFOT with five temperatures, and the short-term oven aging (STOA) protocol was conducted on CRMA mixtures with three types of gradations. Firstly, the chemical and rheological performance of CRMA binders and mixtures were investigated by conducting Gel permeation chromatography (GPC), Fourier transformation infrared spectroscopy (FTIR), Dynamic shear rheology (DSR) tests, and Bending beam rheometer (BBR) test. Then, the equivalent aging temperatures were determined to optimize RTFOT temperatures by comparing the chemical indicators of binders and mixtures. Finally, the correlations of chemical and rheological indices were established, and the rheological properties of aged asphalt in mixtures were predicted. The analysis results indicate that suitable RTFOT temperature is not only related to the technical routes and viscosity of CRMA binders, but also relevant to the mixture gradations. Binders with higher viscosity need elevated RTFOT temperatures from 173 degrees C to 193 degrees C, especially simulating short-term aging of mixtures with higher air voids. The swelling and degradation of crumb rubber contribute to the enhancement of the anti-aging performance. The aging index (AI) can reflect the anti-aging performance of CRMA binders, and the variation of AI can validate the rationality of the optimized RTFOT temperatures.

The temporal variations (diurnal and seasonal) of the optical properties and direct aerosol radiative forcing (DARF) of different aerosol components (water-soluble, insoluble, black carbon (BC), and sea-salt) were analyzed using the hourly resolution data (PM2.5\) measured at an urban site in Seoul, Korea during 2010, based on a modeling approach. In general, the water-soluble component was predominant over all other components (with a higher concentration) in terms of its impact on the optical properties (except for absorbing BC) and DARF. The annual mean aerosol optical depth (AOD, tau) at 500 nm for the water-soluble component was 0.38 +/- 0.07 (0.06 +/- 0.01 for BC). The forcing at the surface (DARF(SFC)) and top of the atmosphere (DARF(TOA)), and in the atmosphere (DARF(ATM)) for most aerosol components (except for BC) during the daytime were highest in spring and lowest in late fall or early winter. The maximum DARF(SFC) occurred in the morning during most seasons (except for the water-soluble components showing peaks in the afternoon or noon in summer, fall, or winter), while the maximum DARF(TOA) occurred in the morning during spring and/or winter and in the afternoon during summer and/or fall. The estimated DARF(SFC) and DARF(ATM) of the water-soluble component were in the range of -49 to -84 W m(-2) and +10 to +22 W m(-2), respectively. The DARF(SFC) and DARF(ATM) of BC were -26 to -39 W m(-2) and +32 to +51 W m(-2), respectively, showing highest in summer and lowest in spring, with morning peaks regardless of the season. This positive DARF(ATM) of BC in this study area accounted for approximately 64% of the total atmospheric aerosol forcing due to strong radiative absorption, thus increasing atmospheric heating by 2.9 +/- 12 K day(-1) (heating rate efficiency of 39 K day(-1) tau(-1)) and then causing further atmospheric warming. (C) 2017 Elsevier B.V. All rights reserved.