Journal: Landslides

Authors: Yingbin Zhanga, Yao Xiaoa, Xiaofu Huangb, Pengcheng Yua, Dejian Lia, Qingdong Wangc, Dongliang Huangd & Xianqing Chenga.

Affiliations: a: Southwest Jiaotong University

              b: China Railway Tunnel Survey and Design Institute

              c: Lanzhou Jiaotong University

              d: Central South University

A large body of evidence indicates that earthquake-induced rockfalls exhibit greater mobility than gravity-induced rockfalls. Field data consistently show that rockfalls are often accompanied by slope shaking during movement. To investigate the influence of slope shaking on rockfall mobility, a series of shaking table tests exceeding 3000 runs were conducted. The tests considered multiple factors, including ground motion characteristics (amplitude and frequency), rockfall properties (shape and initial velocity), and slope conditions (straight and convex slopes). Using analysis of variance and trajectory analysis methods, the movement patterns of rockfalls under different conditions and their influencing factors were systematically summarized. Among these, runout distance and lateral displacement were the main indicators analyzed. The test results show that ground motion plays a crucial role in determining the kinematic characteristics of rockfalls. In particular, the amplitude of seismic loading has a significant effect on the kinematic behavior of rockfalls. Compared with the case without seismic loading, larger amplitudes lead to greater rockfall mobility, with runout distance and lateral displacement increasing significantly. Rockfall movement exhibits strong inherent randomness, and its runout distance and lateral displacement distributions approximately follow a Gaussian distribution. In addition, rockfall shape is another important factor affecting its kinematic characteristics. Rockfalls with more lateral edges have larger runout distances and lateral displacements. These findings help clarify the complex relationship between slope shaking and rockfall movement, thereby providing a deeper understanding of the dynamics and potential hazards of seismic events.

Fig. 1 Damage to houses and infrastructure caused by landslides triggered by the Luding earthquake and the Nepal earthquake (photographed by the authors)

Fig. 2 Photographs and schematic diagrams of the shaking table, transmission device, and slope model

Fig. 3 Rockfall samples used in the tests and simplified schematic diagrams: (a) TP, (b) QP, (c) PP, (d) HP

Fig. 4 Relationship between maximum error and number of tests. (a) Fitting curves and confidence intervals for the two displacements at the 5% confidence level, with the equations and goodness of fit R2 for each fitting curve. (b) First derivative of the fitting curves

Fig. 5 Distribution of results from all parallel tests. (a) Displacement and position distribution curves for all rockfalls. (b) Box plots of runout distance and lateral displacement for rockfalls of different shapes

Fig. 6 Three-dimensional surface plots of displacement under different frequency, amplitude, and shape effects. The left panels show runout distance, and the right panels show lateral displacement

Fig. 7 Effect of aspect ratio of different rockfall shapes on rockfall movement under ground motion with an amplitude of 600 cm/s2 and a frequency of 6 Hz. (a) Runout distance. (b) Lateral displacement

Fig. 8 Histograms of runout distance and lateral displacement distributions for rockfalls of different shapes in 60 parallel tests under different test conditions, including maximum, minimum, and mean values. (a) TP, amplitude = 600 cm/s2, frequency = 10 Hz. (b) QP, amplitude = 900 cm/s2, frequency = 6 Hz. (c) PP, amplitude = 300 cm/s2, frequency = 6 Hz. (d) HP, amplitude = 600 cm/s2, frequency = 4 Hz

This study used shaking table tests to analyze earthquake-induced rockfalls, focusing on their kinematic behavior and implications for hazard management. Through statistical analysis of more than 3000 tests, it was found that rockfall movement exhibits significant randomness, with runout distance and lateral displacement approximately following a Gaussian distribution. This randomness highlights the importance of probabilistic approaches in rockfall risk modeling rather than purely deterministic methods.

The results show that seismic ground motion provides additional kinetic energy through shaking, amplifying the bouncing and rolling behavior of rockfalls and thereby significantly enhancing their mobility. The study found a clear positive correlation between ground motion amplitude and runout distance. This finding emphasizes the need to consider seismic amplitude in hazard assessment, especially in earthquake-prone regions. However, the effect of frequency shows a more complex pattern, indicating that further research is needed to refine predictive models.

Rockfall characteristics, such as shape and aspect ratio, also play a key role in runout distance and lateral displacement. Spherical rockfalls have more lateral edges and dissipate less energy during movement. In contrast, elongated rockfalls are more prone to sliding rather than bouncing effectively. These insights are crucial for assessing the potential impact areas of different types of rockfalls, enabling more precise hazard mapping.

In addition, initial velocity and slope type significantly affect rockfall behavior. Owing to reduced friction losses, convex slopes produce larger runout distances and lateral displacements than straight slopes. These factors are particularly important for assessing risks along transportation corridors or near infrastructure, where slope geometry and detachment conditions vary greatly.

However, this study also has some limitations that warrant attention. First, the experiments used harmonic waves with a limited frequency range (4, 6, 8, and 10 Hz), which cannot fully capture the complex frequency components of real seismic events and may limit the applicability of the results to natural earthquake scenarios. Second, to ensure experimental consistency, the release direction of the rockfall samples was fixed, whereas in nature the detachment direction of rockfalls may vary widely, which may affect their trajectories and final positions. Third, the slope model was made of smooth steel plates, unlike the uneven and rough surfaces of natural slopes, which may have different effects on rockfall kinematics. In addition, although the simple prismatic rockfall model is effective for studying runout distance, it cannot fully explain the lateral dispersion observed in naturally occurring irregularly shaped rockfalls. Therefore, this limitation restricts the broader application of the lateral dispersion findings.

To overcome these limitations, future research should focus on the following directions: (1) incorporating actual slope materials and topography to enhance the realism of the experimental setup, simplifying the steel plate model, and improving its applicability under natural conditions; (2) exploring larger rockfalls and a greater variety of shapes to determine whether the observed movement behavior applies to large rockfalls in real natural environments, thereby overcoming the limitations of small-sized and single-shape samples and better studying the three-dimensional movement characteristics of rockfalls; (3) investigating ground motions with broader frequency spectra, which would help achieve a more comprehensive understanding of the effects of earthquakes on rockfall dynamics and overcome the limitation of the narrow frequency range used in this paper. The experimental results provide valuable insights into the kinematic characteristics of rockfalls under seismic conditions, help improve risk assessment models, and provide a reference for formulating disaster mitigation strategies in earthquake-prone regions. Although this study does not target a specific region or directly implement these applications, it lays a foundation for future research and practical work in rockfall hazard management.

Zhang, Y., Xiao, Y., Huang, X. et al. Rockfall movement and propagation during earthquakes: Insights from shaking table tests. Landslides (2025).