Preloading-based DEM Modeling on Dynamic Processes of Loess-granular Mass Flow Landslides

Zhi-jun WANG , Zuo-xiang XIE

Science Technology and Engineering ›› 2026, Vol. 26 ›› Issue (23) : 9796 -9807.

PDF
Science Technology and Engineering ›› 2026, Vol. 26 ›› Issue (23) : 9796-9807 DOI: 10.12404/j.issn.1671-1815.2505430
research-article
Preloading-based DEM Modeling on Dynamic Processes of Loess-granular Mass Flow Landslides
Author information +
History +
PDF

Abstract

The fourth-level river terraces on the south bank of the Yellow River in Lanzhou are frequently affected by loess-debris flow landslides that threaten nearby communities. To clarify their kinematics and dynamics, a preloading DEM(discrete element method) was established in PFC-2D that incorporates gravitational stress and the Yunsheng Printing Plant landslide (north Fulongping) was used as a test case.Compared with the conventional “sand-rain” method, the preloading approach increased computational efficiency by up to 7.5× and produced a more realistic gravitational stress field, as evidenced by denser particle bonding and a more homogeneous force-chain network.The simulation faithfully reproduced the retrogressive failure process and the final deposit morphology observed in the field. Analysis of the micro-mechanical evolution shows that basin-scale disintegration of the particle skeleton and the consequent loss of inter-particle cohesion govern the catastrophic transition.The results validate the preloading DEM framework for loess-debris flow landslides and offer a reliable tool for future hazard assessment and mitigation design.

Graphical abstract

Keywords

loess-granular mass flow / landslides / PFC-2D / dynamic processes / preloading method

Cite this article

Download citation ▾
Zhi-jun WANG, Zuo-xiang XIE. Preloading-based DEM Modeling on Dynamic Processes of Loess-granular Mass Flow Landslides. Science Technology and Engineering, 2026, 26(23): 9796-9807 DOI:10.12404/j.issn.1671-1815.2505430

登录浏览全文

4963

注册一个新账户 忘记密码

References

Funding

PDF (36285KB)

26

Accesses

0

Citation

Detail

Sections
Recommended

/

〈 〉