Multi-field coupling behavior of frozen soil under impact loading based on phase-field model

Fulai Zhang, Zhiwu Zhu*, Taiyu Zhang, Jianguo Ning, Tao Li, Zhengqiang Cheng

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

The complex multiphase composition of frozen soil induces significant coupling interactions between the thermal, hydrological, mechanical, and damage fields during deformation, particularly under dynamic loading conditions. This study presents a hybrid decomposition phase-field model to investigate the multi-field coupling behavior and damage mechanisms of frozen soil. Unlike the spectral decomposition model, the proposed framework integrates isotropic degradation and the spectral decomposition methods, thereby enabling the simulation of damage evolution under compressive-dominated loading conditions. The model incorporates the viscous effects and strain rate sensitivity to accurately capture the dynamic response of frozen soil and establishes governing equations for coupled displacement, temperature, and fluid pressure fields. The applicability of the model was validated through confined compression experiments on frozen soil, demonstrating its capability to predict distinctive damage features, such as compaction bands oriented perpendicular to the loading direction, which represent the competitive interaction between the softening mechanism of pore collapse and the hardening mechanism of microstructural densification. This study provides significant advancements in the theoretical understanding and numerical simulation of the dynamic mechanical behavior of frozen soil.

Original languageEnglish
Article number111049
JournalEngineering Fracture Mechanics
Volume320
DOIs
Publication statusPublished - 13 May 2025

Keywords

  • Frozen soil
  • Geomaterial
  • Impact loading
  • Multi-field coupling
  • Phase-field model

Fingerprint

Dive into the research topics of 'Multi-field coupling behavior of frozen soil under impact loading based on phase-field model'. Together they form a unique fingerprint.

Cite this