Correlation between macroscopic creep behavior and multi-scale pore characteristics of loess landslide soil: Experimental analysis and model prediction

Read full text
Authors: Pan Wang, Bin Zhi, Enlong Liu, Xiaochan Wang
Cited by: 1
References: 65

Abstract

The multi-scale pore evolution during the creep process of loess affects the shear strength and permeability characteristics, which is an important physical mechanism in landslides. This study analyzes the multi-scale mechanisms through creep tests and nuclear magnetic resonance (NMR) tests, and develops a macro-mesoscale creep model. The primary observations are as follows: (1) As the shear stress level increases, the creep mode varies from attenuating type to non-attenuating type. With the increases in shear stress, mean stress, and stress ratio, the final creep strain exhibits a linear increasing trend. With the increase in shear stress level, the macro-microscale pore area after creep exhibits an evolution pattern of first decreasing, then increasing, and ultimately exceeding the initial state under low confining pressure (100 and 200 kPa) and high shear stress level (0.95) in terms of macro-pore area. The relationship between macro-pore area and creep strain can be described by a quadratic function. (2) During the attenuating creep deformation process of loess, the macro-pore area decreases and the micro-pore area increases, thereby reducing pore channels and enhancing frictional effects. This, in turn, results in decreased permeability and increased strength, thereby improving the slope stability. In contrast, in the non-attenuating creep stage, the macro-pore area increases abruptly, resulting in increased permeability and decreased strength, thereby weakening the slope stability. That is, the stability of loess slopes is affected by the positive-negative feedback mechanism. (3) Based on this, a local bonding damage parameter that accounts for the multi-scale pore evolution mechanism is constructed, and a macro-mesoscale creep model is developed, which can more accurately predict test data compared with previous models.

Fields of Study & Subjects

Landslides and related hazards Geotechnical Engineering and Soil Mechanics Soil and Unsaturated Flow

Publication details

Journal / source
Journal of Rock Mechanics and Geotechnical Engineering
Publisher
Elsevier BV
Volume
18
Issue
8
Pages
6472–6488
Publication date
7 February 2026
Language
EN