Radar Monitoring and Numerical Simulation Reveal the Impact of Underground Blasting Disturbance on Slope Stability

Abstract

Underground blasting vibrations are a critical factor influencing the stability of mine slopes. However, existing studies have yet to establish a quantitative relationship or clarify the underlying mechanisms linking blasting-induced vibrations and slope deformation. Taking the Shilu Iron Mine as a case study, this research develops a dynamic mechanical response model of slope stability that accounts for blasting loads. By integrating slope radar remote sensing data and applying the Pearson correlation coefficient, this study quantitatively evaluates—for the first time—the correlation between underground blasting activity and slope surface deformation. The results reveal that blasting vibrations are characterized by typical short-duration, high-amplitude pulse patterns, with horizontal shear stress identified as the primary trigger for slope shear failure. Both elevation and lithological conditions significantly influence the intensity of vibration responses: high-elevation areas and structurally loose rock masses exhibit greater dynamic sensitivity. A pronounced lag effect in slope deformation was observed following blasting, with cumulative displacements increasing by 10.13% and 34.06% at one and six hours post-blasting, respectively, showing a progressive intensification over time. Mechanistically, the impact of blasting on slope stability operates through three interrelated processes: abrupt perturbations in the stress environment, stress redistribution due to rock mass deformation, and the long-term accumulation of fatigue-induced damage. This integrated approach provides new insights into slope behavior under blasting disturbances and offers valuable guidance for slope stability assessment and hazard mitigation.

1. Introduction

Blasting technology, as the primary method for mine excavation and production, is widely applied in various mining operations due to its advantages of low cost, high efficiency, and operational simplicity [1]. However, the intense vibrations generated during blasting can induce disturbances within the rock mass, compromising its structural integrity and adversely affecting the stability of geotechnical structures [2]. This is particularly critical in slope engineering, where blasting-induced vibrations can activate potential weak structural planes or concealed fractures within the in situ rock mass, triggering localized failures or even large-scale slope sliding, thereby increasing the risk of deformation and instability [3].

Blasting vibrations, mining-induced disturbances [4], rainfall infiltration [5], and rock mass structure [6] are recognized as major triggers for slope hazards such as landslides and rockfalls. Among these, the academic focus on blasting-induced impacts remains limited, primarily due to the short duration, transient nature, and dynamic complexity of blasting loads, which are often coupled with other influencing factors. These characteristics make field monitoring and independent analysis of blasting effects particularly challenging. For example, Lu et al. [7], using borehole TV techniques, revealed that open-pit blasting not only causes direct damage within the mining area but also induces cumulative damage in surrounding rock masses, particularly near the slope crest, significantly reducing overall slope stability. Gou et al. [8] pointed out that compared with open-pit blasting, underground blasting generates more complex stress wave propagation patterns, and the attenuation of surface vibration amplitudes with decreasing charge weight is relatively milder. Chand et al. [9] integrated UAV mapping with 3D numerical modeling to visualize blasting-induced damage zones. By reconstructing 3D slope surfaces and applying both the finite difference method and limit equilibrium analysis, they evaluated slope stability under static and dynamic loading conditions. Similarly, Widodo et al. [10] employed vertical electrical sounding and resistivity tomography experiments to regressively predict peak particle acceleration (PPA) induced by tunnel blasting and used a pseudo-static approach to assess the dynamic slope stability. Due to the short duration and small magnitude of slope deformation under blasting, it is difficult to capture the full mechanical response of the slope through field monitoring alone. As such, numerical simulation has become a primary method for investigating slope stability under dynamic loading. Although existing research predominantly focuses on earthquake-induced dynamic responses, these studies provide valuable insights for understanding slope behavior under blasting vibrations. For instance, Chen et al. [11] utilized GeoStudio to investigate the failure mechanisms and stability of slopes under seismic loading and unsaturated rainfall infiltration, highlighting the deterioration effects of seismic excitation on rock mass stability. Tao et al. [12] simulated the impacts of pre- and post-seismic rainfall and seismic loading on loess slopes using GeoStudio, revealing the combined effects of structural degradation and external factors on slope stability. Zhang et al. [13] selected actual seismic records with different characteristics to establish 2D finite element models of soil slopes, analyzing deformation and stress responses under various ground motion conditions. In the field of mining-induced blasting disturbances, most existing studies have primarily focused on the propagation characteristics of blast waves and the impact-induced damage mechanisms in rock masses. For example, Jiang et al. [14] employed finite element methods to investigate the spatial distribution of peak particle velocity (PPV) under underground excavation blasting conditions. Similarly, Song et al. [15] and Wu et al. [16] analyzed vibration attenuation patterns and the extent of influence on open-pit slope rock masses during excavation blasting. Li et al. [17], Hu et al. [18], and Yang et al. [19] examined the damage effects of blast-induced impacts on slope rock integrity from the perspective of blasting excavation techniques. In addition, Guo et al. [20], Hu et al. [21], and Nath et al. [22] explored the vibrational behavior and wave propagation mechanisms of blasting disturbances through theoretical modeling approaches. These studies have significantly contributed to the understanding of blast-induced wave propagation and its localized impact on rock masses, providing essential theoretical guidance for blast control and rock response prediction. However, the majority of research remains concentrated on the dynamic process of wave propagation, while comprehensive investigations into the effects of blasting disturbances on the overall slope stability are comparatively scarce. In particular, the quantitative relationship between blasting vibrations and slope deformation remains poorly understood. As noted by Bazzi et al. [23], although individual blasting events typically induce only minor displacements on mine slopes, repeated blasting can lead to cumulative displacements that progressively become more pronounced and may eventually result in slope failure. In practical engineering scenarios, subtle slope deformations are often obscured by broader displacement trends, making them difficult to detect using conventional monitoring techniques. This limitation significantly constrains in-depth research and effective evaluation of the mechanisms through which blasting disturbances impact slope stability.

To address the above shortcomings, this paper takes the transition from open-pit to underground mining in the Shilu Iron Mine as the engineering background. A mechanical response analysis model of slope rock mass under blasting loads is established using GeoStudio. By incorporating real-time slope deformation data from radar monitoring and applying Pearson correlation analysis, the study investigates the relationship between blasting vibrations and slope deformation. The results reveal the mechanisms by which blasting disturbances affect slope stability and provide new approaches and tools for evaluating slope stability under blasting conditions.

2. Materials and Methods

2.1. Research Area

The Shilu Iron Mine was initially developed as an open-pit operation and later transitioned to underground caving methods. The geological conditions within the mining area are classified as simple, with a geological exploration category of Type III (moderate complexity within stratified rock formations). Although several relatively large-scale faults are present in the ore-bearing structural zone, their distribution has limited influence on rock mass stability in the study area. Intrusive rocks are extensively developed along the southern, northern, and western boundaries of the mining area, forming a nearly encircling pattern. These are primarily composed of diorite and granitic rock types. The surface rocks in the southwestern part of Mining Area I (North) are mainly gneissic monzonitic granite or migmatized granodiorite, while the southern and northern sections are dominated by biotite monzonitic granite. The western part is also mainly composed of biotite monzonitic granite.

The final bench height of the open pit is 12 m, with a pit bottom elevation of 0 m and a cutoff elevation of 168 m. The eastern and southern slopes of the pit are topographically higher, with the maximum elevation reaching 310 m in the east and 372 m in the south. In contrast, the northern slope is relatively lower, with a maximum elevation of 170 m. The underground orebody is mainly distributed beneath the bottom of the open pit and extends southeastward. At present, the underground mining depth has reached –75 m. The area extending 75 m below the original pit floor, as well as parts of the orebody on the eastern side, has already been extracted, resulting in voids partially filled by caved wall rock and collapsed surface material. Consequently, the eastern slope (Xiaoying Mountain) and southern slope of the open pit have exhibited varying degrees of gradual surface subsidence. Among these, the eastern Xiaoying Mountain area shows more pronounced landsliding, though the displacement and sliding velocity remain low, consistent with slow-moving landslides. These movements typically do not pose an immediate threat to underground operations. Landslide activity in other slope areas is of a smaller magnitude and has not yet materially impacted the safety of underground mining operations [24].

Field investigations have shown that the eastern slope of the open-pit mine is located within a current surface subsidence zone and demonstrates poor overall stability, with localized landslide occurrences. Multiple tensile fractures have developed on the back slope, with maximum displacements reaching up to 1.3 m, as illustrated in Figure 1. Daily high-intensity blasting operations are regularly conducted using a top–down fan-shaped medium-deep hole blasting technique. The blast holes are 85 mm in diameter, with a burden of 3.3 m and spacing of 2.2 m. Bulk porous ammonium nitrate–fuel oil (ANFO) serves as the primary explosive, with bottom-hole initiation. The specific charge is approximately 0.25 kg/t, indicating a relatively high energy input per unit mass of rock blasted. This repeated high-energy blasting constitutes the main source of dynamic disturbance in the area and significantly impacts slope stability. During blasting events, the eastern slope experiences strong vibrations and frequent rockfalls, indicating pronounced dynamic responses and an elevated risk of instability. In contrast, the western slope remains relatively stable, though localized disturbances are still observed. To systematically evaluate the differential effects of blasting-induced ground motions on the stability of both slopes, this study utilizes a combination of numerical simulation and ground-based radar monitoring. These methods are employed to analyze the stress evolution and displacement response of the rock mass under actual measured blasting loads, thereby elucidating the mechanisms by which repeated blasting influences overall slope stability.

2.2. Dynamic Response Analysis Model of Slope

Based on the topographic map and geological data of the Shilu iron mine in Hainan, a three-dimensional geological model of the rock mass and orebody was developed using 3DMine to represent the transition from open-pit to underground mining. Profiles were extracted from the highest points at the eastern and western slope crests, extending downslope along the direction of maximum gradient to the slope toes. Regions with high elevation and steep slopes were prioritized to capture representative structural features and deformation responses. The profile locations and corresponding surface morphology are shown in Figure 2.

GeoStudio 2018 R2 is a widely used numerical analysis software suite in geotechnical engineering. Its SLOPE module is commonly applied for slope stability assessments, while the QUAKE/W module enables dynamic analysis of geotechnical structures under seismic or blast loading. By integrating SLOPE and QUAKE/W, the dynamic response of slopes subjected to blasting disturbances can be effectively simulated. In this study, a robust and practical simulation method was developed through extensive programming and parameter calibration, resulting in a mechanical model that accounts for blast-induced vibrations from underground mining (Figure 3).

The primary steps for constructing the numerical model are as follows:

① A three-dimensional geological model of the transition from open-pit to underground mining was built using geological data. Representative cross-sections were extracted and imported into GeoStudio for analysis. The numerical model established in this study balances analytical accuracy with computational feasibility by adopting several simplifying assumptions. The main assumptions are as follows: The rock mass is considered a homogeneous and isotropic continuous medium, neglecting the fine-scale heterogeneity and the complexity of structural planes within the rock mass. This may lead to deviations in local stress or deformation characteristics. Material parameters are determined based on laboratory test results, assuming stable material properties throughout the loading process, without considering fatigue damage or time-dependent deformation. Simplified boundary conditions: Idealized boundary conditions are applied, which do not fully account for the influence of complex geological settings and groundwater flow on slope stability. Simplified blasting vibration treatment: Blasting loads are modeled as dynamic inputs with simplified vibration waveforms and durations, without capturing the full complexity of dynamic characteristics during actual blasting processes. The above assumptions are standard in conventional numerical simulation approaches. While they may somewhat restrict the model’s applicability, they do not compromise the validity of the blasting simulation analysis.

② In GeoStudio, mechanical parameters were assigned to each lithological unit based on engineering geological investigations. The slope strata in the Shilu Iron Mine mainly consist of dual-permeability rocks, hematite, and unconsolidated formations. To ensure the reliability of the input parameters, rock samples were collected on-site from representative locations across different strata, and a large number of laboratory tests—including uniaxial compression, triaxial compression, and indirect tensile strength tests—were conducted to determine fundamental mechanical properties such as the uniaxial compressive strength, Poisson’s ratio, Young’s modulus, cohesion, and internal friction angle. Subsequently, considering the influence of in situ structural defects, weathering, and scale effects, the obtained intact rock parameters were subjected to empirical reduction techniques to derive the equivalent mechanical parameters of the rock mass, which were then used in the numerical model. The final input parameters for each lithological unit are summarized in

Table 1. Fundamental physical and mechanical properties of the rock mass.

Rock Type	Density (kg/m3)	Friction Angle (°)	Cohesion (kPa)	Poisson’s Ratio	Elastic Modulus (GPa)
Dual-permeability rock	2814.82	11.89	0.241	0.21	81.08
Hematite	3675.68	38.38	1.396	0.23	112.905
Loose rock layer	2814.82	4.52	0.057	0.2	52.582

.

③ The QUAKE dynamic analysis module consists of two submodules: an initial static analysis and an equivalent linear dynamic analysis. Prior to evaluating the dynamic response induced by blasting disturbances, the initial static analysis under gravitational loading must be completed. The model’s boundary conditions are defined in two parts: initial stress state boundaries and dynamic analysis boundaries. In the initial stress configuration, the left and right boundaries of the 2D model are fixed in the X-direction and free in the Y-direction, while the bottom boundary is fixed in both directions. The top boundary is left unconstrained. Under these conditions, static analysis under self-weight is conducted to determine the stress–strain distribution within the slope, thereby providing the necessary initial input for subsequent dynamic simulations.

④ Following the static analysis, the calculated initial pore water pressure distribution is imported into the dynamic analysis module. The boundary conditions for dynamic analysis are as follows: the top boundary is defined as a free surface, the left and right boundaries are fixed in the Y-direction, and the bottom boundary is fixed in both X and Y directions.

⑤ The acquisition, processing, and implementation of blasting-induced ground motion data are critical in ensuring the accuracy of dynamic simulation inputs. In this study, field-measured vibration waveforms were directly used as boundary input conditions for the numerical model, thereby enhancing the physical realism and reliability of the simulation results. Vibration data were obtained from sensors installed at the −75 m level drift within the mine. These sensors recorded blasting-induced ground motions with a sampling delay of −100 ms, a frequency of 8000 Hz, and a total duration of 2 s, yielding 16,000 data points per event. According to the analysis results, the dynamic stress response and shear strength variation in the slope rock mass under blasting disturbance are predominantly governed by horizontal ground motions [3]. Therefore, the X-direction component of the field-recorded waveform—representing horizontal vibration—was selected as the primary dynamic input. Given the inherent time-step constraints in numerical simulations, the raw vibration signal could not be used directly. To address this, the waveform underwent a filtering and down-sampling process aimed at retaining the peak acceleration and key waveform characteristics while reducing the data volume to ensure computational stability and efficiency. The processed real waveform, as used in the model, is shown in Figure 4 and forms the basis for assessing the slope’s dynamic response under realistic blasting conditions.

⑥ Monitoring point configuration and dynamic response analysis of the slope were conducted to evaluate the slope’s behavior under blasting disturbances. Multiple stress monitoring points were placed at the toe, crest, and mid-slope regions to capture the spatial distribution of dynamic stress and displacement. Considering the complex lithological conditions on the eastern slope, three monitoring points were additionally deployed at varying elevations along the same horizontal section to enable a more comprehensive characterization of the dynamic response in this zone.

⑦ Slip surface identification and factor of safety estimation were carried out using the SLOPE module, based on the results obtained from the QUAKE dynamic analysis. This process enabled the identification of potential failure surfaces and quantification of the safety factor, thereby assessing the impact of blasting-induced disturbances on slope stability. The key computational outcomes were extracted and exported to provide a scientific basis for slope stability evaluation.

2.3. Slope Surface Deformation Monitoring Using Radar and Correlation Analysis with Blasting Activities

(1) Slope surface deformation monitoring using radar

While numerical simulation of blast-induced vibrations can elucidate the dynamic response process and spatial distribution characteristics of slopes under blasting disturbances, its results are often difficult to directly correlate with the long-term deformation behavior of the slope and the cumulative effects of repeated blasting. To validate the accuracy of the numerical simulations, a slope monitoring radar system was employed to continuously acquire real-time three-dimensional coordinates of designated monitoring points across the slope surface. This allowed for the characterization of slope deformation evolution under the coupled effects of long-term blasting disturbances and gravitational loading. The monitoring equipment used in this study was the S-SAR M III slope radar, manufactured by Zhong’an Guotai Technology Development Co., Ltd., Beijing, China. The radar monitoring equipment and its coverage area are illustrated in Figure 5.

(2) Method of correlation analysis between blasting vibrations and slope deformation based on the Pearson correlation coefficient

Under mining-induced disturbances, the deformation of slow-moving slopes typically follows a complex and continuously evolving response process, driven by the coupled influence of multiple factors. These include mining activities, blasting vibrations, rainfall, and seepage, in conjunction with geological and hydrogeological conditions. Previous studies have shown that, in surface-to-underground transition mines employing block caving methods, the primary driver of slope deformation is rock mass subsidence induced by underground voids, while secondary factors such as blasting disturbances exert an indirect influence on the rate and scale of slope failures. Based on this understanding, a key challenge in slope stability assessment and hazard early warning lies in determining the extent to which blasting operations—within the broader context of mining disturbances—contribute to observed slope deformations. In this study, the sequence of blasting operations was used to divide the overall slope deformation time series into multiple stages. For each stage, observational data were analyzed using the Pearson correlation coefficient to quantitatively assess the linear relationship between blasting disturbances and slope deformation. The Pearson correlation coefficient, denoted as r, is a widely used statistical measure that reflects both the strength and direction of a linear relationship between two variables. Values of r range from −1 to 1, with values near 1 indicating strong positive correlation, values near −1 indicating strong negative correlation, and values near 0 suggesting no significant linear relationship. Let the variable sequence be denoted as

$$\begin{array}{l}X=\left\{x_1,x_2,x_3,\cdots x_n\right\}\\ Y=\left\{y_1,y_2,y_3,\cdots y_n\right\}\end{array}$$

(1)

$$r={\displaystyle \frac{{\displaystyle {\sum }_{i=1}^n(x_i-\overline{x})(y_i-\overline{y})}}{\sqrt{{\displaystyle {\sum }_{i=1}^n{(x_i-\overline{x})}^2}}\cdot \sqrt{{\displaystyle {\sum }_{i=1}^n{(y_i-\overline{y})}^2}}}}$$

(2)

where

$$\overline{x}={\displaystyle \frac{1}{n}}{\displaystyle {\sum }_{i=1}^n{x}_i}$$

(3)

$$\overline{y}={\displaystyle \frac{1}{n}}{\displaystyle {\sum }_{i=1}^n{y}_i}$$

(4)

To assess the dynamic deformation response of the slope under blasting disturbances, a phased statistical analysis was conducted based on high-frequency displacement data obtained from the radar monitoring system. The monitoring system recorded displacement variations on an hourly basis, capturing the full 24 h deformation process each day. Considering the potential lag in slope displacement in response to blasting loads, the monitoring timeline was divided into analysis stages of 6 h intervals, resulting in a total of 115 stages over the entire monitoring period. For each stage, the average displacement was calculated, and the difference between adjacent stages was determined to quantify the slope’s deformation rate over time. To identify the correlation between blasting disturbances and slope deformation across different periods, the Pearson correlation coefficient was applied to assess the linear relationship between the stage-wise displacement variations and the timing of blasting events. Taking a single day of monitoring data as an example, blasting typically occurred between 17:00 and 18:00. Given the 6 h interval segmentation, time nodes were set at 06:00, 12:00, 18:00, and 00:00, dividing each day into four analysis stages. The time period from 18:00 to 00:00 was defined as the post-blasting response stage, which was used to evaluate the immediate deformation behavior of the slope following blasting activities.

3. Results

3.1. Analysis of Slope Dynamic Response Under Blasting Vibration

(1) Stress response analysis under blasting vibration conditions

Figure 6 presents the X-directional stress distribution contour of the slope under static equilibrium conditions.

Analysis of Figure 6 reveals that significant stress concentration occurs near the toes of some remnant bench slopes. Field investigations corroborate that localized slumping failures had previously occurred in these bench areas during the open-pit mining stage. As mining transitions from surface to underground operations, the blasting disturbances generated during the block caving process may further exacerbate stress concentration in these regions. This could initiate tensile or shear cracks near the bench slope angles. Under continued blasting disturbance, such cracks tend to propagate and eventually coalesce, potentially leading to varying degrees of slope instability or failure.

Figure 7 shows the time–history curve of maximum shear stress on the eastern slope under a blasting load.

Analysis of Figure 7 indicates that the maximum shear stress undergoes significant fluctuations during the 0–1 s interval, highlighting the pronounced impact of blasting vibrations in this region. The maximum shear stress ranges from 1286.03 kPa to 1298.68 kPa, with a peak variation of 13.65 kPa. After 1 s, the amplitude of shear stress fluctuations gradually diminished and tended to stabilize after the vibration ceased. The final value was slightly higher than the initial shear stress before blasting, indicating that blasting vibrations not only induced instantaneous shear stress responses but also affected the final distribution of shear stress.

Figure 8 presents the temporal variations in stress in the x and y directions at monitoring point 1 located at the slope toe.

Analysis of Figure 8 reveals that stress fluctuation patterns in both x and y directions are essentially consistent with shear stress variations, exhibiting similar oscillatory characteristics during blasting disturbances. Further comparison shows that the amplitude of stress variation in the x direction significantly exceeds that in the y direction, indicating that horizontal disturbances dominate the stress response in this region.

(2) Analysis of displacement, velocity, and acceleration responses

The dynamic evolution of the model mesh effectively captures the deformation characteristics of the local slope mass. Focusing on the rock mass at the slope toe, mesh configurations at various time intervals under the influence of blasting disturbances were extracted from the numerical model. Key time points—0 s, 0.25 s, 0.5 s, 0.75 s, 0.875 s, 1 s, 1.5 s, and 2 s—were selected to analyze the progressive displacement and deformation process at the slope toe. The corresponding displacement variations over time are illustrated in Figure 9.

Analysis of Figure 9 reveals that during the initial phase of blasting vibration (t ≤ 1 s), the rock mass at the slope toe experiences significant instantaneous displacement increments due to the high acceleration of blast shock waves. As the amplitude of blasting vibration acceleration gradually attenuates, rock mass deformation stabilizes and displacement increments decrease. Following the cessation of vibrations, the rock mass develops permanent displacement toward the free face. These results indicate that the instantaneous impact of a single underground mining blast on open-pit slope displacement is relatively limited and insufficient to trigger large-scale geological hazards. However, under prolonged and repeated blasting disturbances, cumulative effects of displacement responses occur in the slope rock mass, leading to creep-type deformation or fatigue damage from cyclic disturbances, ultimately inducing localized instability in the slope rock mass.

Figure 10a shows the temporal variation in displacement in the X-direction (horizontal) for three monitoring points on the eastern slope, while Figure 10b presents the temporal variation in displacement in the Y-direction (vertical) for the same monitoring points.

Figure 10 shows that during blasting vibration, the slope face exhibits overall horizontal displacement fluctuations toward the free face direction. The horizontal displacement trends at all monitoring points are essentially consistent. Overall, the displacement amplitude at monitoring point M1 remains consistently smaller than those at monitoring points M2 and M3. This phenomenon can be attributed to the extensive presence of loosely structured rock strata in the central section of the eastern slope within the simulation model. These strata are characterized by poor rock mass integrity, along with lower elastic modulus and density compared to the surrounding formations. During the attenuation phase of peak blasting-induced vibration acceleration, such rock masses substantially reduce the propagation speed and frequency of vibration waves, potentially leading to wave distortion. Consequently, the displacement response at monitoring point M1, located within this rock stratum, is lower than the values recorded at monitoring points M2 and M3, which are positioned within intact rock formations.

Unlike the horizontal displacement characteristics, vertical displacement exhibits oscillatory behavior centered around the zero displacement baseline during blasting vibration. To further analyze the influence of blasting vibration on vertical displacement, the variance and standard deviation of vertical displacement at the three monitoring points were calculated, yielding corresponding displacement statistical indicators, as presented in

Table 2. Statistical parameters of vertical displacement.

ID	Mean	Standard Deviation	Variance	Minimum	Maximum
M1	0.000287	0.00187	0.000003	−0.00719	0.00488
M2	−0.000255	0.00358	0.000013	−0.01022	0.01106
M3	−0.000471	0.00345	0.000012	−0.01277	0.01485

.

Table 2 shows that monitoring point M1 exhibits a significantly lower vertical displacement mean and dispersion compared to monitoring points M2 and M3. This can be attributed to the location of monitoring point M1 within loose rock strata with poor integrity, resulting in reduced sensitivity to blasting vibration waves. Additionally, the lower topographic elevation of this monitoring point further diminishes vibration energy transmission and response amplitude. Therefore, throughout the entire blasting vibration process, the peak vertical displacement at monitoring point M1 is markedly smaller than that at monitoring points M2 and M3, which are located within the more compact and intact formations.

Figure 11 presents the temporal variations in velocity and acceleration vectors at vertical monitoring points MA, MB, and MC, while

Table 3. Statistical data for velocity and acceleration at vertical monitoring points.

	Monitoring Point	Mean	Standard Deviation	Variance	Minimum	Maximum
Velocity (mm/s)	MA	0.18115	0.20358	0.04145	0.00223	0.98950
	MB	0.13064	0.13105	0.01717	0.00223	0.59927
	MC	0.11085	0.133	0.01769	0.00369	0.67087
Acceleration (g)	MA	0.00158	0.00192	0.000004	0.00001	0.01048
	MB	0.00142	0.0016	0.000003	0.00002	0.00615
	MC	0.00137	0.00142	0.000002	0.00001	0.00718

lists the statistical values including mean, standard deviation, and variance of velocity and acceleration at each monitoring point. Combining the monitoring point locations with the data in Table 3 clearly demonstrates that both the velocity and acceleration means at monitoring points gradually increase with elevation, exhibiting a distinct “elevation effect.” Even though the vibration energy generated by caving blasting is limited, particle vibration velocity and acceleration show significant amplification with increasing elevation along profiles at the same horizontal distance. This is particularly evident at monitoring point MA located at the slope crest, where both velocity and acceleration standard deviations exceed those at monitoring points MB and MC within the same vertical profile, indicating greater data dispersion and larger fluctuation amplitudes. This demonstrates that under blasting vibration disturbances, higher slope elevations result in more pronounced dynamic responses of rock mass particles and more intense vibration responses, thereby exacerbating slope instability and failure.

The aforementioned research findings indicate that under blasting vibration, particle vibration velocity and acceleration at the slope crest and adjacent areas exhibit significant amplification effects. Such amplification readily induces the formation of new fractures and substantially activates pre-existing fissures, thereby weakening the tensile strength of the slope rock mass. Under prolonged and repeated blasting vibration loading, the mechanical properties of the slope rock mass gradually deteriorate, accompanied by fatigue-induced rock mass failure, potentially culminating in geological hazards such as landslides.

(3) Slope blast disturbance response characteristics under different working conditions

Figure 12 shows the temporal variations in horizontal displacement at monitoring points on each slope under underground mining blasting disturbances. With the blasting vibration duration set at 2 s, the horizontal displacement response characteristics at monitoring points on the eastern and western slopes were comparatively analyzed.

Figure 12 shows that under blasting vibration, the temporal displacement variations at monitoring points on both eastern and western slopes exhibit similar trends. The specific characteristics are as follows:

(1)

During the initial phase of blasting vibration generation and propagation (0–0.5 s), displacement responses at all monitoring points fluctuate dramatically, displaying an overall trend of initial decrease followed by increase, with displacement values exhibiting large-amplitude oscillations alternating between positive and negative values. The cumulative displacement formed during this phase comprises two components: elastic deformation displacement, which gradually recovers as vibration acceleration attenuates, and irreversible residual displacement.

(2)

During the sustained propagation phase of blasting vibration (0.5–1 s), blasting vibration acceleration rapidly attenuates, with all displacement values during this phase being negative, reflecting that displacement responses are primarily elastic with relatively small permanent deformation.

(3)

During the late propagation phase of blasting vibration (1–2 s), displacement gradually transitions from negative to positive values, exhibiting an overall slow increasing trend, indicating gradual attenuation of elastic displacement and continuous accumulation of permanent displacement.

(4)

Regarding differences in slope displacement responses, displacement fluctuation amplitudes at eastern slope monitoring points are generally smaller than those at the western slope. Combined with geological survey results and analysis of model rock stratum distribution, the eastern slope contains extensive loose rock strata with poor rock mass integrity and local fault and fracture structures, significantly enhancing vibration wave energy absorption and attenuation effects, resulting in relatively weak vibration displacement responses in this region. In contrast, the western slope exhibits superior rock stratum integrity and structural stability, displaying more sensitive displacement responses under vibration. Even after significant vibration acceleration attenuation, it still demonstrates pronounced vibration lag and sustained fluctuations.

(5)

During the middle and later stages of blasting vibration propagation, monitoring points located at higher elevations consistently exhibit greater displacement magnitudes compared to those at lower elevations on the same slope. This observation reflects a characteristic “elevation effect,” wherein increasing elevation amplifies particle velocity and acceleration responses to blasting vibrations, thereby intensifying the resulting displacement.

These findings reveal that slopes with hard lithology and high integrity typically exhibit large-amplitude elastic oscillatory responses during blasting vibration but demonstrate relatively small permanent deformation and strong resistance to vibrational disturbances. Conversely, slopes with soft lithology and poor integrity, due to their absorption and attenuation of vibration energy, show relatively small displacement responses in the short term; however, under prolonged and repeated blasting vibration disturbances, they are prone to micro-fracture propagation and coalescence within the slope rock mass, ultimately generating larger-magnitude irreversible deformations that severely compromise overall slope stability and increase the risk of geological hazards such as landslides and collapses.

(4) Variation patterns of safety factors under vibration loading

The factor of safety is a critical index for evaluating slope stability. In current engineering practice, it is typically calculated using limit equilibrium methods, among which the Bishop method is the most widely adopted. The Bishop method was used to calculate and analyze the factor of safety for the slope rock mass model at different time, with the results presented in Figure 13.

Figure 13 shows that the factor of safety variations for the western and eastern slope profiles under blasting vibration exhibit essentially consistent trends. Overall, the factor of safety displays alternating rise-and-fall dynamic responses corresponding to vibration acceleration fluctuations, reflecting the nonlinear dynamic behavior of rock mass during blasting disturbances.

During the most intense 1 s of blasting vibration, the large-amplitude variations in vibration acceleration values cause dramatic factor-of-safety fluctuations with significant peak differences. The western slope profile exhibits a minimum factor of safety of 1.559, with a difference of 0.058 between maximum and minimum values, while the eastern slope profile shows a minimum factor of safety of 1.096 with a difference of 0.049. Compared to pre-blasting factors of safety, blasting disturbances induce fluctuation variations of approximately 0.02 within short time periods. Additionally, the eastern slope profile has a relatively low initial factor of safety, with more pronounced variations in factor of safety differences caused by blasting vibration, indicating stronger sensitivity of its rock mass structure to blasting disturbances and relatively poor stability. Although the final factor of safety essentially recovers to pre-blasting levels, single blasting events still exert certain influences on slope stability, necessitating continued attention to cumulative responses under long-term multiple blasting disturbances.

3.2. Correlation Between Underground Blasting Activities and Slope Surface Deformation

The relationship between slope displacement and time obtained from field slope radar monitoring is shown in Figure 14. To quantitatively analyze displacement correlations across different phases, the absolute values of Pearson correlation coefficients r(L) between displacement means at each phase throughout the entire data cycle, as well as the absolute values of Pearson correlation coefficients r(ΔL) between displacement increments (differences), were calculated, with the results presented in Figure 15.

As shown in Figure 14, slope displacement means exhibit a nonlinear dynamic growth trend with advancing mining activities. Time-series calculations reveal that multiple significant increases in slope displacement rates are concentrated within specific time periods following the completion of daily blasting operations (17:30–18:00), with these growth rates rapidly attenuating within short durations, displaying a “sudden increase–attenuation” response pattern that reflects the gradual stabilization characteristics of slope response after blasting disturbances. Additionally, the fluctuation amplitude of displacement increments reflects the degree of slope deformation during different time periods. Greater fluctuations indicate more intense slope responses during corresponding phases, potentially related to an enhanced disturbance to local structural planes or internal loosening and rearrangement of rock mass. This fluctuation pattern is consistent with the previously identified trend.

A further analysis of the temporal distribution of Pearson correlation coefficients (as shown in Figure 15) reveals that the first (6:00–12:00) and second (12:00–18:00) time intervals—both occurring prior to blasting—exhibited very low correlations between average slope displacement and time, with coefficients of 0.0006 and 0.002, respectively. These values are significantly lower than those recorded during the third (18:00–0:00) and fourth (0:00–6:00) post-blasting intervals, with increases nearly 5 to 10 times higher. In terms of displacement increments, the Pearson correlation coefficients during the first and second intervals were approximately 0.1, while the values for the third and fourth intervals reached 0.176~0.178—substantially higher than pre-blasting levels. A larger Pearson coefficient indicates a stronger correlation between displacement and time. These findings suggest a strong short-term correlation between blasting vibrations and slope deformation. Notably, the correlation in the fourth interval (0:00–6:00) was slightly higher than that in the third (18:00–0:00), indicating that the slope’s response to blasting is not instantaneous but exhibits a certain time-lag effect. This delay may be attributed to the propagation lag of vibration waves within the rock mass, the progressive development of microcracks, and the redistribution of localized stresses, which collectively result in a delayed peak in slope deformation following the blasting event.

Blasting-induced vibrations exhibit a delayed response and cause only limited displacements in the short term. Therefore, extending the monitoring period enables a clearer and more comprehensive assessment of slope deformation characteristics before and after blasting. In this study, a 72 h monitoring window was adopted. Based on slope radar monitoring results, cumulative displacement cloud maps were extracted for the fourth day at key time intervals: 6 h and 1 h before blasting and 1 h and 6 h after blasting, as shown in Figure 16.

Since the beginning of monitoring, three actively deforming zones have been identified along the open-pit slope. Zone A is located in the central-to-lower section of the southern part of the east slope; Zone B lies in the central area of the southeastern slope, directly connected to the upper part of a pre-existing landslide zone, and shows a significantly higher deformation trend than other areas; and Zone C is situated in the middle-to-upper section of the southern slope, primarily within the bench area, with a more scattered deformation distribution. As shown in Figure 16a,b, prior to the blasting event, the three zones already exhibited noticeable deformation, with displacement values ranging from −13.8 mm to 36.62 mm and displaying certain fluctuations. According to Figure 16c, one hour after blasting, both the magnitude and extent of deformation increased across the three zones. The maximum displacement reached 40.33 mm, with a cumulative increase of 10.13%. Significant sliding was observed in the upper part of Zone A, while the deformation in Zone C expanded outward in multiple directions. By 6 h post-blasting, as shown in Figure 16d, the maximum displacement further increased to 49.08 mm, representing a 34.06% rise in cumulative deformation relative to pre-blasting values. The sliding ranges in Zones A and C expanded downward and westward, respectively, with the extent nearly doubling compared to the pre-blasting state.

These findings, further corroborated by geological radar-based displacement cloud maps, confirm that the slope exhibits a pronounced disturbance response shortly after blasting. This response not only accelerates the deformation process but also reflects a distinct time-lag effect.

4. Discussion

In summary, the mechanisms by which blasting disturbances affect open-pit slope stability can be categorized into three major aspects: instantaneous stress environment perturbations, rock mass deformation and stress redistribution effects, and long-term damage mechanisms induced by fatigue accumulation, as illustrated in Figure 17. These mechanisms are mutually coupled and synergistically evolve, potentially culminating in slope stability failure.

First, instantaneous high-amplitude vibrational disturbances induced by blasting cause abrupt perturbations in the internal stress field of rock mass. Particularly in regions with pre-existing micro-fractures and structural planes, rock mass elements originally in stable or metastable states may experience instantaneous sliding or sudden displacement jumps under blasting disturbances. Dai et al. [25] argued that the rapid release of strain energy during blasting can cause joint opening and sudden displacements in jointed rock slopes. Although the duration of this process is brief, it may significantly disturb the pre-existing stress equilibrium within the rock mass, potentially initiating localized deformation or triggering instability.

Second, blasting disturbances trigger stress redistribution and local plastic deformation accumulation within the rock mass. Especially at slope toe locations or areas with dense structural weakness development, dynamic disturbances generated by blasting intensify stress concentrations, promoting irreversible deformation in rock mass and weakening its structural integrity. Under long-term cumulative effects, rock mass may exhibit characteristics such as strength deterioration and stiffness degradation, progressively evolving to form shear zones or potential slip surfaces, thereby significantly undermining overall slope stability.

Furthermore, the fatigue damage effect induced by blasting vibrations is recognized as one of the primary mechanisms driving the long-term progressive instability of slopes. The initiation sites and failure modes of cracks in rock materials are highly influenced by complex stress conditions [26]. Blasting vibrations act on the rock mass in the form of high-frequency, short-duration, and repetitive loading—akin to low-cycle fatigue in engineering structures. This repeated dynamic loading facilitates the gradual initiation, propagation, and eventual coalescence of microcracks within the rock mass, leading to a continuous reduction in its load-bearing capacity and shear strength. Previous studies have shown that repeated blasting can reduce rock mass strength; our study further verifies and quantifies this through field measurements and simulations. Ray et al. [27] demonstrated that the uniaxial compressive strength of sandstone decreases progressively with increasing cyclic loading. Singh [28] experimentally confirmed that the fatigue life of rock extends as the amplitude of cyclic loading decreases. Tien et al. [29] investigated the fatigue behavior of water-saturated sandstone under cyclic loads and established a correlation between fatigue life and cumulative axial strain. This accumulation of fatigue damage ultimately undermines the structural integrity of the rock mass. Yang et al. [30] further found that under the cumulative effects of multi-stage blasting, damaged zones tend to evolve into potential slip surfaces, increasing the likelihood of localized slope failures or landslides.

Finally, when slopes are in slow creep sliding states or have experienced local sliding, instantaneous disturbances generated by blasting significantly accelerate sliding rates of slope rock mass and enhance their kinematic tendencies. This phenomenon is particularly pronounced when slopes are in limit equilibrium or critically stable states, constituting an important triggering factor for landslides. Furthermore, the deformation response exhibits a pronounced time-lag effect. Theoretical modeling by Wang et al. [31] suggests that blasting-induced slope failures may occur during the mid-to-late stages of the vibration process (approximately 0.17–0.5 s). Building upon this, radar monitoring data in the present study reveal that the lag effect extends well beyond the duration of the blasting wave itself. Specifically, accelerated slope surface deformation and heightened landslide risk were observed within 1 to 6 h following the blasting event, indicating a sustained post-blast response phase that warrants close attention in hazard assessment and monitoring practices.

5. Conclusions

(1)

The simulation results combined with field investigations indicate that blasting vibrations exhibit typical short-duration, high-amplitude pulse characteristics, with stress variations in the horizontal direction significantly exceeding those in the vertical direction. This suggests that horizontal dynamic loading is the critical trigger for slope shear failure. Furthermore, the intensity of the rock mass vibration response induced by blasting disturbances increases with elevation, with monitoring points at the slope crest exhibiting significantly higher velocity and acceleration, marking it as the most dynamically sensitive hazard zone. The eastern slope, characterized by poor rock mass integrity and pronounced variations in the factor of safety, represents the primary current instability zone, whereas the western slope remains relatively stable overall but still presents localized disturbance responses requiring attention.

(2)

By introducing the Pearson correlation coefficient combined with slope radar remote sensing technology, this study, for the first time, reveals a strong correlation between blasting sequences and slope deformation. In slope areas undergoing slow creep, both the magnitude and spatial extent of deformation increased following blasting events. Specifically, cumulative deformation increased by 10.13% and 34.06% at one and six hours after blasting, respectively, indicating a significant lag effect. These findings suggest that blasting vibrations may trigger or exacerbate short-term slope deformations, highlighting the necessity for enhanced post-blasting monitoring and assessment. Future work will focus on refining dynamic models and integrating long-term monitoring data to enhance predictive capabilities.

(3)

Mechanistically, the impact of blasting disturbances on slope stability can be understood through three interconnected processes: abrupt changes in the instantaneous stress field, stress redistribution caused by deformation within the rock mass, and the cumulative accumulation of fatigue damage over time. Repeated blasting accelerates the initiation, propagation, and coalescence of microcracks, facilitating the formation of continuous shear zones or slip surfaces. This progressive damage evolution may transform localized instabilities into large-scale slope failures. Therefore, blasting disturbances not only affect the short-term mechanical equilibrium of the slope but also have potential long-term impacts on stability through cumulative effects, warranting significant attention.