Search

The water retention behavior—a critical factor of unsaturated flow in porous media—can be strongly affected by deformation in the solid matrix. However, it remains challenging to model the water retention behavior with explicit consideration of its dependence on deformation. Here, we propose a data-driven approach that can automatically discover an interpretable model describing the water retention behavior of a deformable porous material, which can be as accurate as non-interpretable models obtained by other data-driven approaches. Specifically, we present a divide-and-conquer approach for discovering a mathematical expression that best fits a neural network trained with the data collected from a series of image-based drainage simulations at the pore-scale. We validate the predictive capability of the symbolically regressed counterpart of the trained neural network against unseen pore-scale simulations. Further, through incorporating the discovered symbolic function into a continuum-scale simulation, we showcase the inherent portability of the proposed approach: The discovered water retention model can provide results comparable to those from a hierarchical multi-scale model, while bypassing the need for sub-scale simulations at individual material points. [ABSTRACT FROM AUTHOR]

Water retention curve (WRC) is an important tool to study the hydraulic and mechanical properties of unsaturated soils, such as permeability, strength and deformation properties. Most of the existing WRC models fail to reflect the water retention mechanisms of unsaturated soils or they are complex in form, and these models are hard to give good performance on modelling the bimodal and multimodal WRCs. In this study, based on analyzing the water retention mechanisms of unsaturated soils, the WRC was divided into two domains that are governed by adsorption and capillary mechanisms, respectively. An adsorption water retention curve model (WRCM) was developed based on micro-pore filling theory and Kelvin’s law. A capillary WRCM was established based on the capillary condensation theory and Young-Laplace equation. Then, a new water retention curve model over the full suction range was built by superposing the adsorption and capillary WRCMs. Finally, the new model was validated through modelling the experimentally measured WRCs of six representative unsaturated soils, including Shanghai soft clay, Xi’an loess, Nanyang expansive soil, Guilin lateritic clay, Western Liaoning aeolian soil and Inner Mongolia Gaomiaozi (GMZ) bentonite. Results showed that the proposed model, which was simple in form with definite physical meaning parameters and successfully reflected the adsorption and capillary mechanisms of water retention, was able to simulate WRCs with different shapes for different types of soils under different conditions.

Due to climate change, extreme rainfalls happen more frequently with different patterns. Biochar and plant roots can affect soil water retention curve (SWRC) and hence slope stability. Until now, no available SWRC model considers soil–plant-biochar interaction, which can also be used for slope stability analysis under extreme rainfalls with different patterns (i.e., advanced, bimodal, and delayed). This study proposes a new SWRC model for vegetated biochar-amended soil by considering the influence of soil–plant-biochar interaction on the void ratio and pore structure. A series of numerical analyses were conducted to investigate the significance of new SWRC model on slope stability under extreme rainfall patterns. Good agreements between laboratory measurements and the new SWRC model predictions were obtained. With an increase of biochar content from 0 to 10%, the number of soil micropores increased by 22%, which improved root volume ratio (Rv) by 130% and hence air entry value (AEV) of vegetated soil by 50%. Based on numerical analysis, the factor of safety (FOS) for vegetated slope under advanced and bimodal rainfall patterns was up to 53% lower than that under the delayed one during rainfall. The rainfall influence depth and FOS of vegetated slope without biochar was nearly 100% deeper and 60% lower than that with biochar, respectively. The lowest FOS was caused by advanced rainfall pattern, which is therefore suggested for slope design. The application of 10% biochar can better stabilize shallow slopes due to its substantial effects on plant root, soil pore structure, and SWRC. [ABSTRACT FROM AUTHOR]

This paper presents a novel strain-dependent water retention model for improved predictive modelling of localized gas flow in bentonite. The proposed model uses fundamental material properties such as dry density and montmorillonite content to generate improved predictions of water retention under different strain conditions. The model was validated with the use of laboratory measurements of capillary pressure in MX-80 bentonite at different dry densities. An additional phenomenological test simulated microfracture induced gas flow in FEBEX bentonite, which showed strong local desaturation due to developing microfractures and a resulting decreasing gas entry pressure. The application of the approach provided first good results that can be relevant for modelling radioactive waste repositories. • Novel strain-dependent water retention model that utilizes fundamental properties of bentonite. • Calculation of the gas entry pressure based on dry density and montmorillonite content. • Validation through empirical measurements and phenomenological modelling to ensure reliability. • Simulating preferential gas pathways induced by high gas pressure within bentonite barriers. [ABSTRACT FROM AUTHOR]

The water retention behaviour of compacted bentonites is strongly affected by multi-physical and multi-scale processes taking place in these materials. Experimental data have evidenced major effects of the material dry density, the imposed volume constraints, and the soil fabric. This paper presents a new water retention model accounting for proper retention mechanisms in each structural level of compacted bentonites, namely adsorption in the intra-aggregate pores and capillarity in the inter-aggregate ones. The model is calibrated and validated against experimental data on different bentonite-based materials, showing good capabilities in capturing the main features of the behaviour. The model is able to reproduce experimental data on compacted bentonites over a wide range of suction values, within a unified framework, and using a limited number of parameters. Some of the parameters introduced are shown to take approximately the same value for several bentonites, providing a significant basis for preliminary design when dedicated experiments are missing. [ABSTRACT FROM AUTHOR]

Water retention in compacted clays is dominated by multi-modal pore size distribution which evolves during hydro-mechanical paths depending on water content and stress history. A description of the evolutionary fabric has been recently introduced in models for water retention, but mostly on a heuristic base. Here, a possible systematic approach to account for evolving pore size distribution is presented, and its implications in models for water retention are discussed. The approach relies on quantitative information derived from mercury intrusion porosimetry data. The information is exploited to introduce physically based evolution laws for the parameters of water retention models. These laws allow tracking simultaneously the evolution of the aggregated fabric and the consequent hydraulic state of compacted clays. The influence of clay microstructure, mechanical constraints and water content changes on the water retention properties is highlighted and quantified from experimental data on different compacted soils with different activity of the clayey fraction. The framework is discussed with reference to a widespread water retention model and validated against experimental data on a Sicilian scaly clay compacted to different dry densities and subjected to a number of hydro-mechanical paths. Copyright © 2014 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Investigation of the temperature effects on the water retention properties of compacted bentonite is of great importance in the context of geological disposal of high-level radioactive waste (HLW) based on the multi-barrier concept. Gaomiaozi (GMZ) bentonite, collected from the Inner Mongolia Autonomous Region of China, has been selected as a potential buffer/backfill material for the construction of Chinese HLW deep geological repository. In this study, soil water retention curves (SWRCs) of both confined and unconfined compacted GMZ01 bentonite specimens with an initial dry density of 1.70 g/cm were investigated at temperatures 20, 40, 60 and 80 °C. Results indicate that the water content decreases as temperature rises, even though the temperature effects on the water retention capacity of compacted GMZ01 bentonite depend on constraint-conditions. The influence of constraint conditions is more significant at relatively low suctions. Under unconfined conditions, the temperature effect is insignificant. However, under confined conditions, the influence on the water retention capacity is found to be significantly suction dependent. The hysteresis behavior of unconfined compacted bentonite becomes less significant as temperature increases. Based on the corresponding results obtained, a modified model was developed for describing the SWRCs of confined compacted GMZ01 bentonite with consideration of temperature effects. It appears that the calculated SWRCs agree well with the experimental ones, indicating that the modified model can satisfactorily describe the effect of temperature on the water retention properties of confined compacted GMZ01 bentonite. [ABSTRACT FROM AUTHOR]

Abstract: Ammonia and pesticide volatilization in the field is a surface phenomenon involving physical and chemical processes that depend on the soil surface temperature and water content. The water transfer, heat transfer and energy budget sub models of volatilization models are adapted from the most commonly accepted formalisms and parameterizations. They are less detailed than the dedicated models describing water and heat transfers and surface status. The aim of this work was to assess the ability of one of the available mechanistic volatilization models, Volt''Air, to accurately describe the pedo-climatic conditions of a soil surface at the required time and space resolution. The assessment involves: (i) a sensitivity analysis, (ii) an evaluation of Volt''Air outputs in the light of outputs from a reference Soil–Vegetation–Atmosphere Transfer model (SiSPAT) and three experimental datasets, and (iii) the study of three tests based on modifications of SiSPAT to establish the potential impact of the simplifying assumptions used in Volt''Air. The analysis confirmed that a 5mm surface layer was well suited, and that Volt''Air surface temperature correlated well with the experimental measurements as well as with SiSPAT outputs. In terms of liquid water transfers, Volt''Air was overall consistent with SiSPAT, with discrepancies only during major rainfall events and dry weather conditions. The tests enabled us to identify the main source of the discrepancies between Volt''Air and SiSPAT: the lack of gaseous water transfer description in Volt''Air. They also helped to explain why neither Volt''Air nor SiSPAT was able to represent lower values of surface water content: current classical water retention and hydraulic conductivity models are not yet adapted to cases of very dry conditions. Given the outcomes of this study, we discuss to what extent the volatilization models can be improved and the questions they pose for current research in water transfer modeling and parameterization. [Copyright &y& Elsevier]