Inferring Dynamic Fragmentation Through the Particle Size and Shape Distribution of a Rock Avalanche

Abstract

On 28 August 2017, a large catastrophic rock avalanche occurred in Nayong, Guizhou Province, China. It claimed 35 lives and caused large property losses. The Nayong rock avalanche provides a rare opportunity to infer the dynamic fragmentation of rock masses from deposits. Field surveys combined with an uncrewed aerial vehicle image analysis system were performed to investigate the particle size and shape distribution along the runout path of the Nayong rock avalanche. The Weibull particle size distribution (PSD) model was tested and possesses the highest accuracy and capability over the entire range of the Nayong rock PSD, and it shows that the median size (D50) values of the particles range from 0.16 to 16.31 m along the runout path. Moreover, the evolution of D50 along the runout path confirms that the degree of rock fragmentation varies greatly in different regions. Additionally, the relative breakage (BR) of particles in the deposition area is 0.191. Particle R values increase significantly from 0.55 to 0.78 along the runout path, which indicates that the rock grinding (shape change) process continuously occurs throughout the avalanche path. The image analysis results confirm that the dominance of rock fragmentation and grinding alternates during granular flow transport. Based on these results, the classic fragmentation‐spreading model for long‐runout rock avalanches is optimized by taking into account the method by which grinding rocks increases the speed of avalanches.

Block‐Grain Phase Transition in Rock Avalanches: Insights From Large‐Scale Experiments

“…(2006) and Jin et al. (2022). It can be seen that our results closely match the distribution in Bowman et al.’s experiments (2012) under augmented gravity scenarios, and some of our experiments produce a higher content of finer fragments.…”

“…(2012) and Jin et al. (2022) are challenging in terms of discerning clear trends, insights from Locat et al. (2006) convincingly demonstrated the lack of a consistent relationship between h s / L s and fragmentation degree (Figure 13b).…”

“…In Figure 13, the values of h s /L s and h c /L c from our experiments and previous studies are plotted against the fragmentation degree. In these studies, both Bowman et al (2012), Bowman and Take (2015), Lin et al (2020), Jin et al (2022), and our experiments employ the relative breakage ratio (B r ) to quantitatively describe the fragmentation degree. While Haug et al (2016) quantifies fragmentation degree through calculating the total mass of the block divided by the mass of the largest fragment, they incorporate image processing to identify the fragment size.…”

“…The degree of fragmentation can be visualized with Figure 7, while a quantitative analysis can be conducted with the fragment size distribution. In Figure 8a, the fragment size is normalized by the maximum size and denoted as D. As shown in Figure 8a, the normalized fragment size distributions are compared with the experimental data conducted by Bowman et al (2012) in a centrifuge apparatus and the field data collected by Locat et al (2006) and Jin et al (2022). It can be seen that our results closely match the distribution in Bowman et al's experiments (2012) under augmented gravity scenarios, and some of our experiments produce a higher content of finer fragments.…”

“…Many studies aim to investigate the fragmentation process and how this process affects the subsequent runout process. The fact that grains with different sizes are distributed in the vertical profile of a rock avalanche deposit implies various breakage processes and mechanisms (Jin et al., 2022; Marc et al., 2021; M. Zhang & McSaveney, 2017). Perinotto et al.…”