A detailed field investigation, including the unmanned aerial vehicle (UAV) survey, was conducted for a rock avalanche that occurred in southwest China in 2017. The dynamic fragmentation process of this rock avalanche has been reproduced via the 3D discrete element method (DEM). The numerical analysis and UAV image recognition were employed to perform a field-scale holistic analysis of the fragment distribution characteristics. The statistical results show that there is a decreasing trend of fragment size with transport distance, but the influence of runout on roundness and sphericity is weak. The holistic analysis indicates that fragment roundness and sphericity have an overall negative correlation with fragment size, and it is a non-linear feedback. Further, variability analysis suggests that along the motion direction, the fragment size distribution in the deposit tends to be uniform, while the roundness and sphericity distribution tend to be variable. Benefiting from the presence of a hillside in the accumulation area of this case, we found that the topography has a significant control on the distribution of geometric characteristics of the rock-avalanche fragments. Finally, according to the data from the vertical profiles in the DEM model, we found that the sphericity of the rock fragments tends to decrease with increasing vertical depth. Our analytical results indicate that channel sieving, frictional abrasion, and basal crushing processes within the avalanching mass lead to fragment sphericity to show better directivity in the vertical direction, compared to the planar distribution. These findings contribute to a better understanding of the distribution of geometric characteristics of blocks within the rock-avalanche deposit from a dynamic fragmentation perspective.Highlights<UnorderedList Mark="Bullet"> <ItemContent> UAV statistical results show that rock fragment size decreases with rock-avalanche propagation distance, but the influence of runout on roundness is weak. </ItemContent> <ItemContent> Both UAV and DEM model reveal that along motion direction, fragment size distribution tends to be uniform, while roundness distribution tends to be variable. </ItemContent> <ItemContent> DEM analysis suggests that fragment sphericity tends to show a better directivity in the vertical direction, compared with the planar distribution results. </ItemContent> </UnorderedList>

Spatial Distribution Characteristics of Rock Avalanche Fragments From Numerical Analysis and UAV Image Recognition

期刊论文
发表日期：

2025年3月31日

查看全文

Wenbin Chang,
Aiguo Xing,
Kaiping Jin

摘要

A detailed field investigation, including the unmanned aerial vehicle (UAV) survey, was conducted for a rock avalanche that occurred in southwest China in 2017. The dynamic fragmentation process of this rock avalanche has been reproduced via the 3D discrete element method (DEM). The numerical analysis and UAV image recognition were employed to perform a field-scale holistic analysis of the fragment distribution characteristics. The statistical results show that there is a decreasing trend of fragment size with transport distance, but the influence of runout on roundness and sphericity is weak. The holistic analysis indicates that fragment roundness and sphericity have an overall negative correlation with fragment size, and it is a non-linear feedback. Further, variability analysis suggests that along the motion direction, the fragment size distribution in the deposit tends to be uniform, while the roundness and sphericity distribution tend to be variable. Benefiting from the presence of a hillside in the accumulation area of this case, we found that the topography has a significant control on the distribution of geometric characteristics of the rock-avalanche fragments. Finally, according to the data from the vertical profiles in the DEM model, we found that the sphericity of the rock fragments tends to decrease with increasing vertical depth. Our analytical results indicate that channel sieving, frictional abrasion, and basal crushing processes within the avalanching mass lead to fragment sphericity to show better directivity in the vertical direction, compared to the planar distribution. These findings contribute to a better understanding of the distribution of geometric characteristics of blocks within the rock-avalanche deposit from a dynamic fragmentation perspective.

Highlights

UAV statistical results show that rock fragment size decreases with rock-avalanche propagation distance, but the influence of runout on roundness is weak.

Both UAV and DEM model reveal that along motion direction, fragment size distribution tends to be uniform, while roundness distribution tends to be variable.

DEM analysis suggests that fragment sphericity tends to show a better directivity in the vertical direction, compared with the planar distribution results.