毛细管作用
地质学
毛细管压力
多孔介质
岩土工程
磁导率
多孔性
润湿
机械
材料科学
土壤科学
饱和(图论)
含水饱和度
作者
Teng-Fei Fu,Tao Xu,Michael J. Heap,Philip G. Meredith,Tianhong Yang,Thomas M. Mitchell,Yoshitaka Nara
标识
DOI:10.1016/j.enggeo.2021.106070
摘要
Abstract The physics of water imbibition into initially unsaturated sandstone is critical to the understanding of displacement processes and fluid transport in the vadose zone. The distribution of water within rock is important due to its significant influence on rock mechanical behavior. Here, therefore, we used nuclear magnetic resonance (NMR) technology to visualize and quantify the dynamics of water infiltration and distribution in initially unsaturated sandstone. The progression of water imbibition in sandstone specimens under the following two conditions were analyzed: (1) specimens soaked in water for different durations and (2) specimens soaked in water for the same duration and then held in this state for different durations. An analytic function was developed to estimate the sandstone moisture profile and to determine the unsaturated flow within the sandstone when the water distribution matched laboratory observations. Finally, a three-dimensional discrete element grain-based model was formulated that incorporates the local parallel-plate method, the unsaturated flow function, and the generalized effective stress principle. We used this model to effectively reproduce the process of water imbibition in the laboratory sandstone specimens. The effect of water imbibition on the mechanical properties of the studied sandstone was also evaluated. These results show that the strength of the core of the sample was reduced as water migrated from its surface to its center, resulting in a decrease in bulk sample strength as standing duration (i.e. water distribution uniformity) increased. The results of this study aid in our understanding of the influence of water imbibition on the mechanical behavior of sandstone, which is important for rock slope stability assessments following rainfall.
科研通智能强力驱动
Strongly Powered by AbleSci AI