山崩
地质学
碎片
岩土工程
腐蚀
管道
水文学(农业)
压力(语言学)
地貌学
环境科学
语言学
环境工程
海洋学
哲学
作者
Danyi Shen,Zhenming Shi,Hongchao Zheng,Jiangtao Yang,Kevin J. Hanley
出处
期刊:Geomorphology
[Elsevier]
日期:2022-06-30
卷期号:413: 108362-108362
被引量:11
标识
DOI:10.1016/j.geomorph.2022.108362
摘要
The sudden breach of a landslide dam seriously threatens the safety of people and property downstream. The failure mechanism of landslide dams is significantly affected by the grain size distribution of the dam material. In this study, a series of experiments are conducted to investigate the effects of grain composition on the stability, breach process, and breach parameters of landslide dams. The results show that dams remain stable with seepage for a coarse-dominated dam, fail by overtopping along with piping for a fines-dominated dam, and fail by overtopping alone for a balanced-composition dam. The failure mode is regulated by the composition of dam material, stress condition and hydraulic gradient. Backward erosion occurs for dams with fines contents <5 % due to the movement of debris, during which the width:depth ratio of the cross section first decreases and then increases. By contrast, for dams with fines contents higher than 15 %, headcutting occurs and a step-pool structure develops because the small grains are more easily washed away than the coarse grains. In this scenario, the width:depth ratio first increases and then decreases. Dam longevity nonlinearly increases with increasing fines content. The breach hydrography changes from a single peak to multiple peaks with increasing fines content, and the peak discharge is sensitive to the median diameter of dam material. In addition, the breach degree, defined as the ratio of the releasing potential energy (total potential energy minus the residual potential energy) to the total potential energy of the lake upstream of the landslide dam, is controlled by this median diameter. This study improves our understanding of failure mechanisms and aids the hazard assessment of landslide dams.
科研通智能强力驱动
Strongly Powered by AbleSci AI