A study of the soil-water reliability model in the whole matric suction range

Soil–water characteristic curves (SWCCs) are considered as a basis for analyzing fluid–solid coupling of unsaturated soil. How to estimate the SWCC over the whole region with matric suction based on the limited local datasets is challenging. Laboratory test is separately carried out on the ranges wi...

Full description

Bibliographic Details
Main Authors: Qingling LIU, Wenbin JIAN, Xutang XU, Wen NIE
Format: Article
Language:zho
Published: Editorial Office of Hydrogeology & Engineering Geology 2022-01-01
Series:Shuiwen dizhi gongcheng dizhi
Subjects:
Online Access:https://www.swdzgcdz.com/en/article/doi/10.16030/j.cnki.issn.1000-3665.202104033
Description
Summary:Soil–water characteristic curves (SWCCs) are considered as a basis for analyzing fluid–solid coupling of unsaturated soil. How to estimate the SWCC over the whole region with matric suction based on the limited local datasets is challenging. Laboratory test is separately carried out on the ranges with low and high matric suctions of the undisturbed residual soil widely distributed in southeast China. With the test data, a model for SWCCs of soil in the whole range with matric suction is established by applying the reliability analysis method. Through analysis and verification based on the test data and data of 52 samples of three types of soil (sand, silt-loam and clay) obtained from UNSODA, it is found that the model parameters involved in the model show significant physics meaning. The model can be used to estimate the SWCC of soil within the whole range with matric suction according to part of the data (in the low or high matric suction part). Compared with the other models, the model is robust and is applicable for acquiring SWCCs of different types of soil within the whole range matric suction with R2>0.98. In addition, the model provides an important method to obtain SWCCs of soils with different texture, and is of the reference significance to the geotechnical AI analysis.
ISSN:1000-3665