Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2026 Feb 28;16:11339. doi: 10.1038/s41598-026-40868-x

Experimental study on layered cemented tailings backfill damage and failure mechanisms under blast loading

Hongjie Qiu 1, Xianyang Qiu 1,✉, Rihong Cao 1, Xin Chen 1, Xiuzhi Shi 1, Zhigang Tian 2, Xiaoyuan Li 1,3
PMCID: PMC13049052  PMID: 41764205

Abstract

The stability of cemented tailings backfill (CTB) is critical for the safe operation of underground mines. Inevitably, operational constraints introduce two types of layered interfaces within CTB: transition heterogeneous structural interface (THSI) and continuous homogeneous structural interface (CHSI), thereby transforming CTB into layered cemented tailings backfill (LCTB). In this study, three-dimensional physical models were developed to simulate rock–backfill systems in underground mines. Five blasting tests were conducted to investigate the effects of charge position and LCTB strength configuration. The analyses focused on dynamic volumetric strain responses (Δk, defined as the attenuation ratio of the first peak volumetric strain εv(max) at identical distances), pre- and post-blast sonic velocity change rates (η, used to quantify damage severity), as well as damage morphology and failure evolution.The results indicate that the dynamic failure of the rock–backfill system proceeds through three sequential stages: crack initiation and backfill extrusion, crack propagation and blasting gas invasion, and rock–backfill system destruction. For LCTB containing a THSI, placing the charge within the high-strength LCTB layer accelerates the attenuation of εv(max), reflected by an increase in Δk (from 0.71 to 2.32), while the damage index η (from < 13% to < 8%) decreases progressively. Conversely, for LCTB containing a CHSI, aligning the charge with the CHSI elevation results in more convergent εv(max) attenuation behavior, with Δk decreasing from 2.54 to 1.32, accompanied by reduced damage levels (η < 10%). These results suggest that, under the investigated model conditions, positioning the charge within the high-strength layer in the presence of a THSI, or aligning the charge with the CHSI, is favorable for mitigating LCTB degradation. The findings provide a mechanistic basis for understanding charge–layered interface interactions in two-step stope blasting and offer engineering-relevant insight into charge placement in layered cemented backfill systems.

Keywords: Layered cemented tailings backfill, Blast loading, Physical modeling, Layered interface

Subject terms: Engineering, Natural hazards, Solid Earth sciences

Introduction

High-stage open stoping with subsequent backfilling is widely recognized as an effective method for extracting thick, inclined ore bodies. This approach not only enhances resource recovery but also employs cemented tailings backfill (CTB) to provide structural support to the surrounding rock mass, thereby reducing the risk of roof collapse and dynamic hazards1–3. Moreover, the utilization of tailings as backfill material is consistent with principles of environmental protection and sustainable development4,5. However, in two-step stope mining, blasting operations in the subsequent stope inevitably impose vibration loading on the CTB placed in the previous stope. If blasting parameters for side holes are improperly designed, such vibration loading may induce crack initiation and propagation within the CTB, potentially leading to structural degradation or even collapse and threatening the overall stability of the backfill system6.

Although numerous studies have investigated the dynamic response and instability of cemented tailings backfill (CTB) under blast loading, in situ monitoring remains challenging because mined-out stopes are typically sealed and inaccessible. Consequently, laboratory-scale experiments using the split Hopkinson pressure bar (SHPB) system have been widely adopted to simulate blast-induced disturbances. For instance, Cao et al.7 and Chen et al.8 examined the dynamic peak compressive strength of CTB specimens with different material compositions and reported a pronounced strain-rate dependence. Tan et al.9 further showed that CTB subjected to cyclic impact loading exhibited significantly higher dynamic peak stress and enhanced resistance to failure compared with single-impact loading. In addition, Zheng et al.10 and Jiang et al.11 investigated energy dissipation characteristics of CTB under dynamic disturbances. However, owing to inherent limitations in specimen size, boundary conditions, and loading modes, SHPB-based tests cannot fully reproduce the systemic failure processes of CTB under practical blasting conditions. As a result, physical model experiments have been increasingly employed to reconstruct more representative blast scenarios. For example, Zhao et al.12 developed a planar physical model to investigate crack morphology under varying blast intensities, identifying localized high-intensity strain fields followed by rapid attenuation near the source. Their subsequent synchronous testing further demonstrated that delayed detonation is more favorable for CTB stability than synchronous blasting13.

In parallel, numerical simulation has been widely applied to examine the effects of blast loading on CTB behavior. Emad et al.14–16 developed a FLAC3D model to simulate blasting in adjacent stopes and identified vibration-induced loading as the primary cause of wedge-type failure. Suazo et al.17 employed LS-DYNA to evaluate the influence of blasting sequence and borehole geometry, determining critical conditions for internal pressurization within CTB. Li et al.18 and Xia et al.19 further elucidated failure propagation paths and the damaging effects of specific detonation sequences. Moreover, Hu et al.20 and Li et al.21 utilized LS-DYNA and SHPB systems, respectively, to investigate fracture propagation in composite rock–backfill models, highlighting the role of the rock–backfill interface in stress-wave transmission. Despite these advances, most studies have treated CTB as a continuous and homogeneous structural unit, thereby largely neglecting the presence of internal layered interfaces.

However, in practical filling operations, internal layered interfaces inevitably develop within CTB. These interfaces can be classified into transition heterogeneous structural interfaces (THSI) and continuous homogeneous structural interfaces (CHSI), depending on variations in backfill slurry mix ratios, as illustrated in Fig. 1a. THSI typically forms when slurries with different mix proportions are sequentially used to fill mined-out areas as a cost-effective strategy, resulting in pronounced strength stratification within CTB (Fig. 1b)22,23. Such filling schemes lead to a load-bearing structure characterized by higher strength in the upper and lower zones and relatively lower strength in the intermediate zone, ultimately giving rise to THSI. Wang et al.24 investigated CTB specimens containing THSI under uniaxial compression and acoustic emission monitoring, and reported that with decreasing CTB strength, the failure mode transitions from tensile-dominated failure to mixed tensile–shear failure, accompanied by increased fracture density. These findings are consistent with those reported by Wang et al.25, further confirming the influence of THSI on CTB failure behavior. To address the insufficient strength of the intermediate layer in layered cemented tailings backfill (LCTB), Zhang et al.26 proposed incorporating reinforcement layers to improve structural performance. In contrast, CHSI may develop even under identical mix ratios due to limitations in filling continuity, stope geometry, or particle settlement during slurry placement27,28. Chen et al.29 investigated CTB specimens containing CHSI through uniaxial compression tests and PFC simulations, showing that the presence of CHSI leads to a reduction in overall CTB strength and a shift in failure mode from shear-dominated to tensile-dominated failure. Under the combined influence of THSI and CHSI, CTB no longer behaves as a continuous medium but instead forms a LCTB. Existing studies on LCTB have predominantly focused on static mechanical properties, using laboratory experiments and numerical simulations to evaluate failure modes and stability.

Fig. 1.

Fig. 1

Open in a new tab

Schematic illustration of the LCTB concept and its failure characteristics: (a) conceptual models of LCTB containing a THSI and a CHSI; (b) typical filling configuration of a two-step stope in underground mining; (c) characteristic failure and fracturing patterns of LCTB under blast loading.

The characteristics of rock mass structural planes exert a significant influence on the propagation path and energy attenuation of blasting-induced stress waves30–32. Compared with intact CTB, LCTB is more prone to stress accumulation along internal interfaces under blast loading, which may trigger crack initiation and subsequent propagation. Such interface-controlled damage processes can lead to severe degradation or even failure of LCTB in underground metal mines. Once compromised, the load-bearing and support capacity of LCTB is markedly reduced. In addition, during two-step stope mining, the mixing of fragmented LCTB with ore can reduce ore grade and increase safety risks during extraction (Fig. 1c).

Against this background, the present study investigates the failure mechanisms and dynamic response characteristics of LCTB containing different interface types under varying charge positions. Based on similarity principles, two three-dimensional physical models were constructed to simulate underground rock–backfill systems subjected to blast loading. Data acquisition and analysis were carried out using dynamic strain monitoring, sonic velocity measurements, high-speed photography, and three-dimensional surface scanning. The dynamic response of LCTB was systematically examined from multiple perspectives, including strain variation and attenuation behavior, damage severity, staged failure evolution, and destruction characteristics of the rock–backfill system. The results clarify the influence of charge position on damage patterns and dynamic response characteristics of LCTB under different strength configurations. These findings provide mechanistic insight and qualitative reference for side-hole charge placement in two-step stope mining, contributing to improved blasting design and the overall stability of LCTB.

Experimental materials and procedures

Model construction

Physical model blasting experiments were conducted using cement mortar and backfill slurry to fabricate the rock–backfill models, as illustrated in Fig. 2a. Due to constraints in specimen fabrication and transportation, cement mortar was selected as the ore body analogue, a practice widely accepted in rock mechanics research13,33,34. Furthermore, the wave impedance of cement mortar is significantly higher than that of LCTB, effectively simulating the wave reflection and transmission characteristics at the rock–backfill interface. Specifically, the ore body analogue was cast using a mixture of ordinary Portland cement, uniformly graded river sand, and water at a mass ratio of 1:1:0.35. The LCTB was simulated using a tailings-based slurry consistent with in-situ mine practices. The high-strength LCTB mix utilized a mass ratio of 1:4:1.65 (cement: tailings: water), whereas the low-strength LCTB utilized a ratio of 1:8:3. In most mines, cement-to-tailings ratios of 1:4 and 1:8 are commonly adopted in combination to reduce backfilling costs. Therefore, the mix proportions selected in this study are representative of those used in practical mine backfilling operations.

Fig. 2.

Fig. 2

Open in a new tab

Physical model preparation and dimensional specifications: (a) model fabrication process; (b) model geometry and layout.

To eliminate the influence of material variability on model strength, the same batch of raw materials was used throughout the preparation process. The model was cast in three phased. First, the cement mortar was poured. After an initial setting period, the bottom layer of the LCTB (LCTBb) was placed. Subsequently, a casting interval of 8–10 h was strictly maintained before pouring the top layer of the LCTB (LCTBt), simulating the actual operational cycle of underground filling. High-frequency vibration was applied during pouring to eliminate entrapped air and ensure compaction. During the casting process, four triaxial strain bricks (S1–S4) were embedded at specific heights of 5 cm and 15 cm. The model dimensions, along with the configuration and orientation of the strain bricks, are detailed in Fig. 2b.

A total of six models were fabricated (five for testing and one reserve) based on different interface and strength configurations. Models 1–3 were constructed as variable-strength systems containing a THSI, where the LCTBb possessed higher strength and the LCTBt possessed lower strength. Conversely, Models 4 and 5 were configured as uniform-strength systems containing a CHSI, utilising the low-strength slurry for both layers to simulate a continuous LCTB.

Parallel to model construction, standard cylindrical specimens were cast to determine the physical-mechanical properties of each component, as shown in Fig. 3. Tests including density, sonic velocity, and uniaxial compressive strength were performed (Table 1). The physical modeling adhered to the principle of similarity to ensure experimental fidelity13,35. Field investigations determined that the actual densities of the ore, high-strength LCTB, and low-strength LCTB were 2810 kg/m³, 1802 kg/m³, and 1760 kg/m³, respectively, with corresponding uniaxial compressive strengths of 73.05 MPa, 10.36 MPa, and 4.56 MPa. Based on these in-situ values and the model parameters, the calculated density similarity constants for ore body, high-strength LCTB, and low-strength LCTB were 1.33, 1.05, and 1.06, respectively, while the strength similarity constants were 2.67, 1.04, and 1.02, respectively. These values fall within the acceptable range for geomechanical modeling. With respect to geometric similarity, the model side-hole distance of 7 cm and single-layer height of 10 cm were scaled relative to actual field dimensions ranging from 1 to 3 m and 0.5 to 3 m, respectively, thereby satisfying the geometric requirements for blast loading simulation. Overall, the physical model of this experiment satisfies the similarity principle.

Fig. 3.

Fig. 3

Open in a new tab

Measurement of physical parameters of the cylindrical specimens.

Table 1.

Physical and mechanical properties of the cylindrical specimens.

Block type Density
/ρ
(kg·m− 3)		 Longitudinal wave velocity
/vp (m·s− 1)		 Elastic modulus
/E (GPa)		 Poisson ratio
/ν		 Compressive
strength
/σc (MPa)		 Tensile strength
/σt (MPa)
Cement mortar 2150 3704 23.7 0.3 27.31 2.4
High-strength LCTB 1902 3142 0.34 0.25 10.82 0.93
Low-strength LCTB 1867 2328 0.17 0.19 4.69 0.58
Open in a new tab

Charge structure and experimental scheme

PETN powder was selected as the explosive for the blasting experiments due to its good stability and encapsulation properties36. It was loaded into transparent plastic tubes with a diameter of 0.3 cm and a length of 8 cm. The borehole diameter in the physical model was 1.0 cm. With air as the coupling medium, a decoupled charge structure with a decoupling ratio of 3.3 was adopted. Considering the high detonation performance of PETN and the geometric constraints of the physical model, this decoupling ratio was selected to prevent excessive peak shock pressures that could lead to non-representative pulverization damage. The average charge mass and density were 0.67 g and 1193 kg/m³, respectively. Accordingly, the adopted decoupling ratio was determined to ensure that the peak stress intensity remains within a representative range for the model material, prioritizing failure mode similarity over strict geometric scaling of the charge structure. An external detonation method was adopted to prevent digital electronic detonators from influencing the experimental results. As shown in Fig. 4a, the detonators were connected to the detonating tubes using electrical tape, and the borehole openings were sealed with soil. The booster pipe was fabricated using the same procedure and enclosed in 0.15 cm-thick copper tubing to minimize its influence during blasting. Hot-melt glue was used to connect the charge and the booster pipe, with retaining rings formed at both ends to ensure that the explosive remained centered within the borehole during detonation.

Fig. 4.

Fig. 4

Open in a new tab

Configuration of the explosive charge and detonator placement: (a) detailed design and composition of the charge assembly; (b) schematic of the three designated charge positions relative to the LCTB layers.

Variations in the detonation position of PETN resulted in different disturbance responses of the LCTB. According to the configuration shown in Fig. 1a, three charge position schemes—bottom (Pb), middle (Pm), and top (Pt)—were designed. As illustrated in Fig. 4b, these positions correspond to PETN charges located within the LCTBb, at the layered interface, and within the LCTBt, respectively.

The hole in this experiment was situated in the cement mortar’s centre. The models 1–3 had their charge placed in the Pb, Pm, and Pt positions, respectively. In the models, the interface type between the layers was THSI, and the top and bottom layers of LCTB had varying strengths. The PETN detonation demonstrated symmetry in both the layered LCTBt and the LCTBb when the LCTB strengths were equal. A single detonation within an individual LCTB could meet the experimental goals. Consequently, the charge positions of models 4 and 5 containing CHSI were Pb and Pm, respectively. Table 2 displays the particular experimental configuration for the models.

Table 2.

experiment schemes.

Model number Types of structural interface Charge position Corresponding position
1 THSI Pb LCTBb
2 THSI Pm THSI
3 THSI Pt LCTBt
4 CHSI Pb LCTBb
5 CHSI Pm CHSI
Open in a new tab

Experimental monitoring system

Figure 5 depicts the four components of the comprehensive monitoring system utilized in this experiment: the dynamic strain monitoring system, the sonic wave measurement system, the high-speed camera system, and the 3D scanning system. The triaxial strain bricks were utilized to acquire dynamic strain data within the dynamic strain monitoring system and to explore the dynamic response characteristics of the LCTB under blast loading. The strain gauges used were the BE120-5AA, featuring a sensitive grid area of 5 × 2.8 mm², a base size of 8.5 × 4 mm², a resistance of 120 ± 0.2 Ω, and a sensitivity coefficient of 2.11 ± 1% with temperature compensation. The strain gauges were connected to a dynamic data acquisition system named TDEC NUXI-1004 via quarter-Wheatstone bridges. Accurate signal capture in the microsecond range was achieved with a sampling trigger level of 0.2 V, a sampling length of 2000 K, and a sampling frequency of 1 MHz. The first peak of the explosion stress wave was successfully monitored, and the reflected stress waves from the model boundaries were recorded by the four strain bricks with a time delay relative to the first explosion stress wave. Furthermore, to satisfy the optical requirements of high-speed photography, no measures were taken to attenuate stress wave reflection at the model borders37,38. The X-series high-speed camera was used within the high-speed camera system to efficiently capture the model’s complete dynamic process of damage development and failure under blast loading. It was synchronously triggered with the dynamic data acquisition device and operated at a frame rate of 19,607 fps, a sampling period of 51 µs, and a pulse width of 10 µs.

Fig. 5.

Fig. 5

Open in a new tab

Schematic diagram of the comprehensive experimental monitoring system.

The spatial distribution of damage at different locations within the LCTB reflects the attenuation characteristics of blast waves39. Changes in bulk density, elastic modulus, and longitudinal wave velocity are commonly used to quantify blast-induced damage in rock and backfill materials. In this experiment, the relative change rate of longitudinal wave velocity (η) was adopted to characterize the degree of damage in the models. An RSM-SY5(T) non-metallic sonic detector was employed using the transmission method to measure wave velocity at two stages: once after model curing and once after the blasting tests. Within the LCTB region, two monitoring lines were arranged along the x-direction at heights of 5 cm and 15 cm, starting from the rock–backfill interface, with each line consisting of six pairs of monitoring points. In each pair, one probe served as a transmitter and the other as a receiver, with a spacing of 4 cm between adjacent measurement points. The measurement points were numbered from 1 to 6, starting at the rock–backfill interface. The longitudinal wave velocity was determined using the transmission method, and the damage index η was calculated according to the following Eqs40–42. :

graphic file with name d33e969.gif 1

In Eq. (1), vp0 is the longitudinal wave velocity before blasting at the measurement point, and vp1 is the longitudinal wave velocity after blasting.

The 3D scanning system, utilizing an EinScan Pro 2X scanner, was employed to capture and analyze the three-dimensional point cloud data of the damaged surfaces. The scanning process was conducted with a speed of 1,100,000 points/s, a volumetric accuracy of 0.3 mm/m, and a scanning precision of 0.05 mm. The spatial point spacing ranged from 0.2 to 3 mm, ensuring that the resolution satisfied the accuracy requirements for reconstructing the failure surface. Following the scan, point cloud generation and model encapsulation were performed to reconstruct the digital model.

Experimental results and analysis

Dynamic strain response and attenuation law

Strain analysis of LCTB with different strengths

Due to the severance of signal transmission wires caused by the intense near-field shock wave, valid dynamic strain data for Model 1 could not be acquired. Consequently, Model 1 is excluded from the quantitative strain attenuation analysis in this section. This study examines the time-dependent fluctuation of volume strain (εv) to characterize the dynamic response features of distinct strain bricks inside a single model43,44. The following formula can be used to convert triaxial strain to εv by elasticity theory45,46:

graphic file with name d33e1033.gif 2

In Eq. (2), εx, εy, and εz are the strain signals recorded in the dynamic strain monitoring system in the X, Y, and Z directions, respectively.

The volumetric strain time histories recorded by the four strain bricks in Model 2, corresponding to the charge position Pm, are shown in Fig. 6a. The negative first peak volumetric strain εv(max) indicates that the LCTB is initially subjected to compressive loading induced by the incident blast stress wave. The subsequent positive peak results from the superposition of transmitted and reflected stress waves at the rock–backfill interface and model boundaries. In the near-blast zone, the strain amplitude εv(S1) is significantly greater than εv(S3), reflecting stronger stress-wave transmission into LCTBb due to its closer wave impedance match with the cement mortar, which is consistent with previous studies47,48. In contrast, at locations farther from the blast source, εv(S2) rapidly approaches zero.

Fig. 6.

Fig. 6

Open in a new tab

Dynamic volumetric strain response of Model 2: (a) typical volumetric strain-time history curves; (b) attenuation characteristics of peak volumetric strain with distance.

Although blast-wave attenuation in rock-like media is intrinsically nonlinear due to geometric spreading and material damping, the limited monitoring range (5–15 cm) and discrete measurement points in the present physical model preclude reliable fitting of a unique nonlinear attenuation function. Therefore, a linear fitting approach is adopted as a local approximation in the near-field zone to characterize the attenuation trend of the first peak volumetric strain. The attenuation gradient (k) is used to describe the decay rate with distance, and the relative attenuation ratio (Δk = kb/kt) is introduced to compare attenuation efficiency between different LCTB layers.

As listed in Table 3 and visualized in Fig. 6b, for Model 2, the attenuation gradient in the high-strength LCTBb is markedly steeper than in the low-strength LCTBt (kb = 512.16 and kt = 221.06), yielding a high attenuation ratio (Δk = 2.32). This indicates that while the high-strength layer receives higher initial energy, it possesses a superior capacity to dissipate blast energy over a short distance.

Table 3.

Summary of volumetric strain attenuation parameters for all models.

Model number Attenuation gradient (k) Relative attenuation ratio (Δk)
2 kb = 512.16 2.32
kt = 221.06 (kb > kt)
3 kb = 272.09 0.71
kt = 383.50 (kb < kt)
4 kb = 281.11 2.54
kt = 110.60 (kb>kt)
5 kb = 306.87 1.32
kt = 231.62 (kb>kt)
Open in a new tab

Figure 7a displays the volume strain-time curve for Model 3, with the charge position at Pt. Model 3 differs from Model 2 in that LCTBt bears most of the initial transmitted blast stress wave. The corresponding linear attenuation trends are presented in Fig. 7b. Unlike Model 2, the data in Table 3; Fig. 7b reveals that when the charge is located within the LCTBt, the attenuation behaviors of the two layers become more comparable, with Δk dropping to 0.71. Although LCTBt attenuates the wave slightly faster locally (kt>kb), the difference is not significant.

Fig. 7.

Fig. 7

Open in a new tab

Dynamic volumetric strain response of Model 3: (a) typical volumetric strain-time history curves; (b) attenuation characteristics of peak volumetric strain with distance.

A comparison between Models 2 and 3 highlights the critical influence of charge position on interlayer attenuation behavior. Specifically, the attenuation contrast between layers intensifies substantially when the charge shifts from the low-strength LCTBt (Pt) to the interface (Pm), with Δk rising from 0.71 to 2.32. This confirms that although the high-strength LCTBb intercepts a higher initial shock load as the charge moves vertically closer, it functions as an energy dissipation buffer, rapidly attenuating the stress wave over a short propagation distance. Consequently, LCTBb leverages its superior blast resistance to preserve the overall stability of the THSI-containing LCTB.

Strain Analysis of LCTB with the same strengths

For Model 4, Fig. 8a shows that the strain responses in LCTBb precede and exhibit significantly larger amplitudes than those in LCTBt, indicating that the initial blast-induced stress wave primarily acts on the layer containing the charge. Quantitatively, εv(max) decreases by approximately 26% across the interface (from S1 to S3) in the near-field zone, whereas the corresponding reduction in the far-field is limited to about 7% (from S2 to S4). This discrepancy indicates that although the negligible wave-impedance contrast facilitates stress-wave transmission across the interface49, the attenuation of the first peak strain within LCTBb occurs more rapidly over a short propagation distance. This behavior is further reflected by the steeper attenuation gradient in LCTBb shown in Fig. 8b (Δk = 2.54), suggesting that the influence of the blast loading is concentrated in the near-field region and becomes increasingly balanced between layers with increasing distance.

Fig. 8.

Fig. 8

Open in a new tab

Dynamic volumetric strain response of Model 4: (a) typical volumetric strain-time history curves; (b) attenuation characteristics of peak volumetric strain with distance.

Figure 9a presents the volumetric strain time histories for Model 5, with the charge positioned at the interface (Pm). The responses of the symmetric monitoring points S1 and S3 are nearly identical, exhibiting synchronous arrival times and comparable peak magnitudes. In the far-field (S2 and S4), while the response times remain synchronized, minor discrepancies in peak strain are observed, which are attributed to the complex superposition of secondary transmission and reflection of the blast stress wave. Regarding attenuation characteristics, Fig. 9b reveals that the decay gradients of the two layers are relatively close, yielding a low relative ratio (Δk = 1.32). These results demonstrate that for LCTB of uniform strength, aligning the charge coplanar with the CHSI fosters a highly symmetric dynamic response, characterized by comparable peak strains and vibration durations across the layers.

Fig. 9.

Fig. 9

Open in a new tab

Dynamic volumetric strain response of Model 5: (a) typical volumetric strain-time history curves; (b) attenuation characteristics of peak volumetric strain with distance.

A comparison between Models 4 and 5 indicates that for LCTB of uniform strength, both the peak volumetric strain εv(max) and the vibration duration of LCTBb and LCTBt become comparable when the charge is aligned with the CHSI. This suggests that the attenuation of the blast-induced strain response in the two layers proceeds in a synchronous manner under this configuration. As a result, the overall dynamic response of the layered structure becomes more uniform, thereby reducing the likelihood of relative displacement or interlayer instability induced by asymmetric blast loading.

Assessment of relative degradation

Since the mortar had entirely separated and been destroyed after blasting, the sonic speed after blasting cannot be measured or computed. Therefore, Eq. (1) is applied to determine the relative change rate of longitudinal wave velocity (η) in the LCTB at different measurement points before and after blasting. In this study, η is adopted as a non-destructive, semi-quantitative indicator to reflect the relative degree of internal degradation and its spatial distribution, rather than as a direct quantitative measure of mechanical property loss.

Relative degradation with different strengths

Figure 10a displays the spatial distribution maps of the sonic velocity change rate (η) for Models 1–3 before and after blasting. A larger η corresponds to a greater relative reduction in sonic velocity, reflecting more pronounced internal degradation within the LCTB. For Model 1, where the charge position is located within the high-strength LCTBb, the η values at all measurement points in the bottom layer are greater than those in the top layer (ηb > ηt), while the maximum change rate η1(max) remains below 8%.For Model 2, where the charge is coplanar with the THSI, ηt becomes slightly greater than ηb, with η2(max) remaining below 9%.In contrast, for Model 3, where the charge is located within the low-strength LCTBt, η values in the top layer exceed 12%, whereas ηb remains below 5%, indicating a pronounced interlayer disparity in relative degradation.

Fig. 10.

Fig. 10

Open in a new tab

Spatial distribution of η in LCTB with different strengths under varying charge positions (Models 1–3): (a) spatial maps of η (damage distribution); (b) interlayer difference in η between the LCTBt and LCTBb (Δη).

To facilitate interlayer comparison, Δη is introduced as the difference in η between LCTBb and LCTBt, serving as a relative indicator of degradation contrast. As shown in Fig. 10b, Δη1 and Δη2 are less than 2.1 and 1.7, respectively, with Δη2 being slightly larger than Δη1. The most significant contrast is observed in Model 3, where Δη3 exceeds 2.9 and reaches a maximum value of 7.45.

By combining Figs. 6 and 7, the spatial variation of η with charge position is found to be qualitatively consistent with the volumetric strain attenuation patterns discussed in Sect. 3.1.1. Notably, for Model 1, the degradation distribution inferred from the sonic velocity measurements provides complementary information in the absence of valid volumetric strain data, thereby supporting the dynamic response analysis. Overall, these results indicate that when the explosive is positioned within the high-strength LCTBb, the relative degradation of the LCTB remains limited, and the blast-induced response attenuates rapidly over a short propagation distance.

Relative degradation with the same strengths

Figure 11a presents the spatial distribution maps of the sonic velocity change rate for Models 4 and 5. For Model 4, where the charge position is located within the LCTBb, ηb is greater than ηt, and η4(max) exceeds 12%, whereas the ηt values in the LCTBt remain below 8%, indicating a marked interlayer contrast in relative degradation. Conversely, for Model 5, where the charge is coplanar with the CHSI, the η values at corresponding measurement points in both layers are approximately equal, with η5(max) remaining below 10%. This reflects a relatively uniform degradation distribution across the layered structure.

Fig. 11.

Fig. 11

Open in a new tab

Spatial distribution of η in LCTB with same strengths under varying charge positions (Models 4–5): (a) spatial maps of η (damage distribution); (b) interlayer difference in η between the LCTBt and LCTBb (Δη).

As shown in Fig. 11b, Δη4 in Model 4 exceeds 1.9 and reaches a maximum value of 5.08. In contrast, Δη5 in Model 5 remains below 0.7, suggesting minimal interlayer contrast. By combining Figs. 8 and 9, the degradation patterns inferred from sonic velocity measurements show strong qualitative agreement with the volumetric strain attenuation behavior discussed in Sect. 3.1.2. These results indicate that for LCTB with uniform strength, positioning the charge coplanar with the CHSI promotes coordinated stress-wave attenuation and results in relatively mild and evenly distributed internal degradation.

Analysis of crack propagation mechanisms and failure characteristics

Analysis of crack propagation patterns

High-speed camera observations revealed that the failure processes followed a consistent pattern across different conditions, evolving through three main stages. (1) Crack initiation and backfill extrusion: A tensile crack initiates along the X-direction, accompanied by the expansion and extrusion of backfill material at the rock-backfill interface due to shock compression, marking the initial decline in system mechanical strength. (2) Crack propagation and gas invasion: The primary crack propagates to the interface, splitting the cement mortar. This creates a pathway for high-pressure explosive gas to penetrate and wedge apart the rock-backfill interface, further weakening the system. (3) Rock-backfill system destruction: The combined effects of stress waves and gas expansion lead to the complete decoupling of the cement mortar from the LCTB.Taking Model 5 as a representative example (Fig. 12), the failure progression is detailed as follows:

Fig. 12.

Fig. 12

Open in a new tab

Failure evolution of Model 5 captured by high-speed photography.

During Stage 1 (0–459 µs), the initial tensile crack appeared at 306 µs, reaching a length of approximately 5 cm. Simultaneously, backfill extrusion was observed along the interface (Fig. 12a), measuring approximately 18 cm. Minor spalling observed in the boundary region was identified as a secondary edge effect50–52. By 459 µs, the crack length extended to 13 cm, while the backfill extrusion intensified to 20.5 cm.

In Stage 2 (816–1632 µs), the primary crack bridged the rock-backfill interface (Fig. 12b), creating a channel for gas invasion. High-pressure gas penetration became visible at 1173 µs, driving the lateral separation of the cement mortar. This gas wedge effect accelerated the displacement of backfill material, leading to significant interfacial opening by 1632 µs. Secondary fractures propagated rapidly from the boundaries due to the intensified displacement.

By Stage 3 (> 1632 µs), the interface was completely compromised (Fig. 12c). For visualization clarity, the time intervals are extended in this phase. Secondary fractures continued to develop at 2295 µs, and by 12,750 µs, the cement mortar and LCTB were fully decoupled. Crucially, despite the systemic collapse of the rock-backfill system, the LCTB itself did not separate along the internal CHSI, retaining its structural coherence as a single entity.

Analysis of the destructive characteristics of LCTB

The degree of damage differed after the blast loading was applied, although the dynamic destruction processes were the same for all models. The cement mortars’ failure mechanism was essentially uniform, characterized predominantly by tensile failure. As shown in Fig. 13, the three-dimensional scanning results of the partially damaged surface of the cement mortar after the explosion of Model 3 are presented. The analysis mainly focuses on the magnitude and features of the damage to the LCTB parts of each model following the blast. Hence, damage to the cement mortar section is not the subject of this study.

Fig. 13.

Fig. 13

Open in a new tab

Reconstructed 3D failure morphology of the cement mortar interface in Model 3: (a) left-side view; (b) right-side view.

The damage morphology for all LCTB segments is illustrated in Figs. 14 and 15. A critical observation across all models is that while the rock–backfill interface experienced significant decoupling, the internal THSI and CHSI remained intact. This disparity is attributed to differences in stress-wave incidence relative to the structural planes. The incident stress wave impinges on the rock–backfill interface in a near-normal direction, resulting in pronounced tensile damage. In contrast, columnar stress waves propagate parallel or obliquely to the internal LCTB interfaces. Owing to the lower energy density along these paths and the relatively small impedance contrast between adjacent fill layers, stress concentration is insufficient to induce interlayer delamination.

Fig. 14.

Fig. 14

Open in a new tab

Damage and failure modes of Models 1–3.

Fig. 15.

Fig. 15

Open in a new tab

Damage and failure modes of Models 4 and 5.

The damage and destruction characteristics after the blast with different strength models are shown in Fig. 14. Models 1 and 2 exhibit comparatively stable behavior. When the charge is located within the high-strength LCTBb or at the THSI, damage is largely limited to minor spalling along the lateral boundaries. The high-strength layer acts as an effective barrier, limiting fracture network expansion. The crack observed in the LCTBt of Model 1 resulted from post-test handling and is therefore excluded from the analysis. In contrast, Model 3 shows severe lateral damage, with extensive fracturing developing within the low-strength LCTBt. The three-dimensional scanning results are consistent with the dynamic strain measurements, indicating a clear protective buffering effect of the high-strength layer.

Figure 15 illustrates the damage morphology for uniform-strength models. Unlike different strength models, edge spalling occurred in both models but with distinct distributions. Model 4 exhibited asymmetric failure concentrated within the charge-bearing LCTBb. In contrast, Model 5 displayed a symmetric damage profile across layers. This symmetry reflects the consistent wave propagation and attenuation properties of the identical strata. Consequently, aligning the charge with the CHSI promotes a coordinated dynamic response, effectively mitigating instability risks from differential interlayer displacement.

Discussion

This study shows that the rock-backfill interface is the first macroscopic failure zone in the system under blast loading. There are three main stages in the destruction of an ore or rock body in this system: crack initiation and backfill extrusion, crack propagation and blasting gas invasion, and rock-backfill system destruction. To guarantee the LCTB material’s stability while optimizing the ore’s lump rate, it is crucial to fully utilize the rock-backfill interface’s separation properties and failure patterns under explosive loads.

In engineering practice, the interface height is preliminarily estimated by correlating void scanning data with daily backfill volumes. However, to mitigate discrepancies arising from complex site conditions, complementary verification methods such as Measurement While Drilling (MWD) or borehole cameras are recommended for critical zones prior to charging. These techniques enable the identification of THSI or CHSI depths with engineering-acceptable accuracy via mechanical response or visual confirmation. Importantly, such localization allows the charge positions to be calibrated relative to the interface, thereby supporting optimized blast design based on actual in-situ conditions, which is a key aspect of precision blasting in two-step stope mining.

As illustrated in Fig. 16, the experimental results clarify the interaction between charge elevation and layered interfaces. This understanding offers mechanistic guidance for charge placement in two-step stopes. Specifically, for LCTB containing a THSI, positioning the charge within the high-strength layer was found to be favorable in the present tests. In this configuration, the high-strength layer acts as a protective buffer, rapidly dissipating incident blast energy and thereby reducing the stress intensity transmitted to the adjacent low-strength layer.

Fig. 16.

Fig. 16

Open in a new tab

Proposed optimization strategy for side-hole charge placement in two-step stope blasting adjacent to LCTB.

Conversely, for LCTB containing a CHSI, or when there is minimal material property contrast between layers, aligning the charge elevation with the interface resulted in a more uniform dynamic response in the physical models. This configuration ensures similar stress-wave propagation and attenuation across the layers, reduces secondary transmission and reflection effects at the interface, and helps minimize the risk of displacement and interlayer instability.

It should be noted that, in the present physical models, the interface type (THSI versus CHSI) is intrinsically linked to the strength stratification of the LCTB. Under otherwise consistent material compositions and geometric conditions, the comparative results therefore reflect the combined mechanical effect of interface configuration and strength contrast, rather than the influence of strength contrast alone.

In physical model tests, boundary reflections were distinguished from genuine failure based on spatiotemporal characteristics. Primary failure modes, particularly backfill extrusion, initiated near the blast source during the early loading phase, distinctly preceding the late-stage boundary spalling. In addition, strain bricks were embedded in the central region, and the analysis focused exclusively on the first signal peak corresponding to the direct wave, thereby effectively excluding interference from boundary reflections. Furthermore, since the comparative evaluation between THSI and CHSI was conducted under identical boundary conditions, boundary effects are considered systematic influences that do not compromise the validity of the main conclusions.

The main subjects of this investigation are the damage and failure of the LCTB section. Due to restrictions in the experimental equipment’s data-collecting capabilities, this experiment did not analyze dynamic strain data for the cement mortar component. The same portions in each model were cast from the same batch, and the variations in PETN dosage were within a manageable range. Consequently, the dynamic strain data of the cement mortar section was excluded from the analysis as it was not critical to the primary objectives of this study. Future work may further decouple interface geometry from strength contrast by independently varying interface morphology and material strength to quantify their respective roles. Additionally, future studies should investigate the energy evolution of the blast wave across the rock-backfill interface by monitoring dynamic strain changes in the cement mortar section. Ultimately, these mechanistic insights are of great importance for optimising the fragmentation and ore block size in the perimeter zones of stope blasting.

Conclusion

In this study, three-dimensional physical models obeying similarity laws were constructed to investigate the dynamic response and failure mechanisms of rock–backfill systems containing THSI and CHSI under blast loading. Based on comprehensive analyses utilizing dynamic strain monitoring, sonic velocity measurement, high-speed photography, and 3D scanning, the following main conclusions are drawn:

  1. For LCTB containing THSI, positioning the charge within the high-strength layer significantly accelerates the attenuation of the first peak volumetric strain εv(max) (Δk increases from 0.71 to 2.32), despite the higher initial wave energy in this layer. Conversely, for LCTB containing CHSI, aligning the charge position with the interface promotes synchronization in strain attenuation and duration between layers (Δk decreases from 2.54 to 1.32).

  2. Sonic wave analysis showed that for LCTB with different strength configurations, blast-induced damage remained limited and comparable when the charge was located within the LCTB height range (η3(max) < 13%, η2(max) < 9%, η1(max) < 8%). In contrast, when the charge elevation coincided with the CHSI, LCTB with identical strength exhibited similar damage levels and wave velocity evolution (η4(max) < 13%, η5(max) < 10%).

  3. The rock–backfill interface was identified as the initial macroscopic failure surface under blast loading. The failure of the rock–backfill system progressed through three stages: crack initiation and backfill extrusion, crack propagation and blasting gas invasion, and rock-backfill system destruction.

  4. The observations indicate that charge elevation significantly influences interlayer stress-wave attenuation and degradation. For LCTB with pronounced strength stratification (THSI), placing the charge within the high-strength layer reduces degradation in the low-strength layer by taking advantage of the high-strength layer’s superior blast resistance and rapid energy dissipation. For LCTB with uniform strength (CHSI), aligning the charge elevation with the interface promotes more balanced attenuation across layers, reducing instability from asymmetric blast loading. These findings provide refined guidance for charge placement in two-step stope blasting operations adjacent to LCTB.

Acknowledgements

The authors acknowledge the financial support from the National Natural Science Foundation Project of China (Grant No. 52374152), and the Guangxi Key R&D Plan (Grant No.2022AB31023), and the Postgraduate Scientific Research Innovation Project of Hunan Province (Grant No. CX20250283) for carrying out this research work.

List of symbols

CTB

Cemented tailings backfill

LCTB/LCTBb/LCTBt

Layered cemented tailings backfill (bottom and top layers)

THSI/CHSI

Transition heterogeneous / continuous homogeneous structural interface

k/kb/kt

Attenuation gradient of the first peak volumetric strain (including bottom / top)

Δk

Relative attenuation ratio of the first peak volumetric strain between layers

η/ηb/ηt

Relative change rate of sonic velocity (bottom and top layers)

η1(max) ~ η5(max)

Maximum relative change rate of wave velocity in Models 1 to 5

Δη/Δη1 ~ Δη5

Difference in the relative change rate of wave velocity between layers

εv/εv(max)

Volumetric strain / First peak volumetric strain

εv(S1) ~ εv(S4)

Peak volumetric strain recorded at strain bricks S1 to S4

Pb/Pm/Pt

Charge positions at the bottom, middle, and top of the model

Author contributions

Hongjie Qiu - Conceptualisation, Data Curation, Operational Experiment, Methodology, Software, Visualisation, Writing-Original Writing-Review & Editing; Xianyang Qiu - Conceptualisation, Funding Acquisition Resources, Supervision, Validation, Writing-Original Draft, Writing-Review &Editing; Rihong Cao - Supervision, Validation; Xin Chen - Data Curation, Investigation; Xiuzhi Shi - Supervision, Guidance; Zhigang Tian - Experimental Site, Safety Management; Xiaoyuan Li - Experimental Site.

Funding sources

This work was supported by the National Natural Science Foundation of China (Grant No. 52374152), the Guangxi Key R&D Plan (Grant No. 2022AB31023), and the Postgraduate Scientific Research Innovation Project of Hunan Province (Grant No. CX20250283).

Data availability

The datasets generated during and analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Liang, W. et al. Experimental study on the interaction between backfill and surrounding rock in the overhand cut-and-fill method. Minerals12, 1017. 10.3390/min12081017 (2022). [Google Scholar]
  • 2.Yin, S. et al. Active roof-contact: The future development of cemented paste backfill. Constr. Build. Mater.370, 130657. 10.1016/j.conbuildmat.2023.130657 (2023). [Google Scholar]
  • 3.Zou, S., Cao, S. & Yilmaz, E. Enhancing flexural property and mesoscopic mechanism of cementitious tailings backfill fabricated with 3D-printed polymers. Constr. Build. Mater.414, 135009. 10.1016/j.conbuildmat.2024.135009 (2024). [Google Scholar]
  • 4.Li, S., Zou, P., Yu, H., Hu, B. & Wang, X. Advantages of backfill mining method for small and medium-sized mines in China: Safe, eco-friendly, and efficient mining. Appl. Sci.13, 7280. 10.3390/app13127280 (2023). [Google Scholar]
  • 5.Xue, G., Yilmaz, E. & Wang, Y. Progress and prospects of mining with backfill in metal mines in China. Int. J. Min. Metall. Mater.30, 1455–1473. 10.1007/s12613-023-2663-0 (2023). [Google Scholar]
  • 6.Zhu, P., Song, W., Cao, S., Wang, F. & Zheng, D. Tensile mechanical response mechanism of cemented backfills under blasting load. J. Min. Saf. Eng.35, 605–611. 10.13545/j.cnki.jmse.2018.03.022 (2018). [Google Scholar]
  • 7.Cao, S., Yilmaz, E. & Song, W. Dynamic response of cement-tailings matrix composites under SHPB compression load. Constr. Build. Mater.186, 892–903. 10.1016/j.conbuildmat.2018.08.009 (2018). [Google Scholar]
  • 8.Chen, X. et al. High strain rate compressive strength behavior of cemented paste backfill using split hopkinson pressure bar. Int. J. Min. Sci. Technol.31, 387–399. 10.1016/j.ijmst.2021.03.008 (2021). [Google Scholar]
  • 9.Tan, Y., Davide, E., Zhou, Y., Song, W. & Meng, X. Long-term mechanical behavior and characteristics of cemented tailings backfill through impact loading. Int. J. Miner. Metall. Mater.27, 140–151. 10.1007/s12613-019-1878-6 (2020). [Google Scholar]
  • 10.Zheng, D., Song, W., Cao, S., Li, J. & Sun, L. Investigation on dynamical mechanics, energy dissipation, and microstructural characteristics of cemented tailings backfill under SHPB tests. Minerals11, 542. 10.3390/min11050542 (2021). [Google Scholar]
  • 11.Jiang, M., Sun, W., Li, J., Fan, K. & Liu, Z. Analysis of fracture characteristics and energy consumption of full tailings cemented backfill under impact load. Rock. Soil. Mech.44, 186–196. 10.16285/j.rsm.2022.0926 (2023). [Google Scholar]
  • 12.Zhao, Y., Yang, R., Fang, S., Wang, Y. & Wang, W. Model experimental study on strain field evolution and damage distribution of filling body under blast loading. Constr. Build. Mater.373, 130798. 10.1016/j.conbuildmat.2023.130798 (2023). [Google Scholar]
  • 13.Zhao, Y., Yang, R., Zuo, J., Liu, Z. & Wang, W. Dynamic response characteristics of backfill under blasting disturbance simulated through a three-dimensional model. Constr. Build. Mater.443, 137685. 10.1016/j.conbuildmat.2024.137685 (2024). [Google Scholar]
  • 14.Emad, M. Z., Mitri, H. S. & Henning, J. G. Effect of blast vibrations on the stability of cemented rockfill. Int. J. Min. Reclam. Environ.26, 233–243. 10.1080/17480930.2012.707527 (2012). [Google Scholar]
  • 15.Emad, M. Z. Numerical modelling approach for mine backfill. Sadhana42, 1595–1604. 10.1007/s12046-017-0702-0 (2017). [Google Scholar]
  • 16.Emad, M. Z., Mitri, H. & Kelly, C. Dynamic model validation using blast vibration monitoring in mine backfill. Int. J. Rock Mech. Min. Sci.107, 48–54. 10.1016/j.ijrmms.2018.04.047 (2018). [Google Scholar]
  • 17.Suazo, G. & Villavicencio, G. Numerical simulation of the blast response of cemented paste backfilled stopes. Comput. Geotech.100, 1–14. 10.1016/j.compgeo.2018.04.007 (2018). [Google Scholar]
  • 18.Li, G., Deng, G. & Ma, J. Numerical modelling of the response of cemented paste backfill under the blasting of an adjacent ore stope. Constr. Build. Mater.343, 128051. 10.1016/j.conbuildmat.2022.128051 (2022). [Google Scholar]
  • 19.Xia, Z. et al. Energy transfer and damage evolution process research of ore rock-filling body under the blasting load. Minerals12, 1362. 10.3390/min12111362 (2022). [Google Scholar]
  • 20.Hu, J. et al. Dynamic response mechanism of a rock-filling interfacial coupling body to blasting in it, Explos. Shock Waves. 41, 164–178 (2021). [Google Scholar]
  • 21.Li, X. et al. Study on energy dissipation characteristics and damage law of backfill under cyclic impact. Shock Vib.2022, 7618440. 10.1155/2022/7618440 (2022). [Google Scholar]
  • 22.Wang, J. Research and application of damage and failure evolution mechanism and strength model of layered cemented tailings backfill. Fuzhou Univ.10.26945/d.cnki.gbjku.2021.000111 (2021). [Google Scholar]
  • 23.Jiang, L., Su, Y. & Dai, Q. Dynamic response mechanisms of layered cemented backfill pillars under horizontal stress wave disturbance of far-field blasting. Chin. J. Rock Mech. Eng.39, 34–44. 10.13722/j.cnki.jrme.2019.0463 (2020). [Google Scholar]
  • 24.Wang, J., Fu, J., Song, W., Zhang, Y. & Wang, Y. Mechanical behavior, acoustic emission properties and damage evolution of cemented paste backfill considering structural feature. Constr. Build. Mater.261, 119958. 10.1016/j.conbuildmat.2020.119958 (2020). [Google Scholar]
  • 25.Wang, Z. et al. Impact of weak interlayer characteristics on the mechanical behavior and failure modes of cemented tailings backfill: A study on thickness, strength, and dip angle. Eng. Fail. Anal.165, 108795. 10.1016/j.engfailanal.2024.108795 (2024). [Google Scholar]
  • 26.Zhang, Y., Xu, W. & Chen, W. The investigation into the mechanical properties and failure mechanisms of stratified cemented tailings backfill with enhancement layer under triaxial compression. Eng. Fail. Anal.182, 110128. 10.1016/j.engfailanal.2025.110128 (2025). [Google Scholar]
  • 27.Li, J. et al. Influence of layered angle on dynamic characteristics of backfill under impact loading. Minerals12, 511. 10.3390/min12050511 (2022). [Google Scholar]
  • 28.Lin, H. et al. Mechanical properties and fracture evolution of layered gangue-cemented backfill. Mining Metall. Explor.41, 3119–3131. 10.1007/s42461-024-01097-w (2024). [Google Scholar]
  • 29.Chen, X. et al. Micro-mechanism of uniaxial compression damage of layered cemented backfill in underground mine. Materials15, 4846. 10.3390/ma15144846 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Meguid, M. A., Saada, O., Nunes, M. A. & Mattar, J. Physical modeling of tunnels in soft ground: A review. Tunn. Undergr. Space Technol.23, 185–198. 10.1016/j.tust.2007.02.003 (2008). [Google Scholar]
  • 31.Liang, X. et al. Visualization study on stress evolution and crack propagation of jointed rock mass under blasting load. Eng. Fract. Mech.296, 109833. 10.1016/j.engfracmech.2023.109833 (2024). [Google Scholar]
  • 32.Shen, Y. et al. Experiments and discrete element simulations on the influence of symmetrical forms of joints on the propagation of blasting cracks. Eng. Fract. Mech.320, 111072. 10.1016/j.engfracmech.2025.111072 (2025). [Google Scholar]
  • 33.Zhang, F. et al. Explosive stress wave propagation and fracture characteristics of rock-like materials with weak filling defects. Eng. Fract. Mech.308, 110372. 10.1016/j.engfracmech.2024.110372 (2024). [Google Scholar]
  • 34.Huo, X. et al. Experimental and numerical investigation on the peak value and loading rate of borehole wall pressure in decoupled charge blasting. Int. J. Rock Mech. Min. Sci.170, 105535. 10.1016/j.ijrmms.2023.105535 (2023). [Google Scholar]
  • 35.Li, X. et al. Study on the bench cast-blasting effects influenced by explosive specific charge. Trans. Beijing Inst. Technol.36, 1233–1236. 10.15918/j.tbit1001-0645.2016.12.005 (2023). [Google Scholar]
  • 36.Huo, X. et al. Attenuation characteristics of blasting stress under decoupled cylindrical charge. Rock Mech. Rock Eng.56, 4185–4209. 10.1007/s00603-023-03286-3 (2023). [Google Scholar]
  • 37.Cheney, J. A., Brown, R. K., Dhat, N. R. & Hor, O. Y. Z. Modeling free-field conditions in centrifuge models. J. Geotech. Geoenviron. Eng.116, 1347–1367. 10.1061/(ASCE)0733-9410(1990)116:9(1347) (1990). [Google Scholar]
  • 38.Turan, A., Hinchberger, S. D. & El Naggar, H. Design and commissioning of a laminar soil container for use on small shaking tables. Soil Dyn. Earthquake Eng.29, 404–414. 10.1016/j.soildyn.2008.04.003 (2009). [Google Scholar]
  • 39.Dong, Y., Xia, C. & Duan, Z. Numerical analysis of plane explosive wave propagation with its attenuation behavior in semi-infinite medium. Eng. Mech.23, 60–65. 10.3969/j.issn.1000-4750.2006.02.011 (2006). [Google Scholar]
  • 40.Xiao, J.-Q., Ding, D.-X., Jiang, F.-L. & Xu, G. Fatigue damage variable and evolution of rock subjected to cyclic loading. Int. J. Rock Mech. Min. Sci.47, 461–468. 10.1016/j.ijrmms.2009.11.003 (2010). [Google Scholar]
  • 41.Zhong, G., Ao, L. & Fu, Y. Model experimental studies of vibration effect and damage evolution of tunnel’s surrounding rock under cyclic blasting excavation. Explos Shock Waves. 36, 853–860. 10.11883/1001-1455(2016)06-0853-08 (2016). [Google Scholar]
  • 42.Lv, G. & Zhou, C. Damage characteristics of grouted tunnel rock mass in fault zones induced by blasting. Chin. J. Rock. Mech. Eng.40, 2038–2047. 10.13722/j.cnki.jrme.2021.0390 (2021). [Google Scholar]
  • 43.Palchik, V. Is there link between the type of the volumetric strain curve and elastic constants, porosity, stress and strain characteristics?. Rock Mech. Rock Eng.46, 315–326. 10.1007/s00603-012-0263-9 (2013). [Google Scholar]
  • 44.Sainsbury, B.-A. Consideration of the volumetric changes that accompany rock mass failure. Rock Mech. Rock Eng.52, 277–281. 10.1007/s00603-018-1560-8 (2019). [Google Scholar]
  • 45.Zhao, K., Huang, Z. & Yu, B. Damage characterization of red sandstones using uniaxial compression experiments. RSC Adv.8, 40267–40278. 10.1039/C8RA06972G (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Lin, Q., Song, T., Lu, D. & Du, X. A novel analytical method for describing ground motion behaviour caused by tunnel excavation. Tunn. Undergr. Space Technol.161, 106546. 10.1016/j.tust.2025.106546 (2025). [Google Scholar]
  • 47.Fan, L. F., Wang, L. J. & Wu, Z. J. Wave transmission across linearly jointed complex rock masses. Int. J. Rock Mech. Min. Sci.112, 193–200. 10.1016/j.ijrmms.2018.09.004 (2018). [Google Scholar]
  • 48.Li, W., Fang, S., Zhu, Y. & Li, G. Effect of rock wave impedance on dynamic mechanical response in SHPB experiments. Geotech. Geol. Eng.43, 118. 10.1007/s10706-025-03087-1 (2025). [Google Scholar]
  • 49.Seinov, N. P. & Chevkin, A. I. Effect of fissure on the fragmentation of a medium by blasting. Sov. Min. Sci.4, 254–259. 10.1007/BF02501547 (1968). [Google Scholar]
  • 50.Hu, X. & Duan, K. Size effect: Influence of proximity of fracture process zone to specimen boundary. Eng. Fract. Mech.74, 1093–1100. 10.1016/j.engfracmech.2006.12.009 (2007). [Google Scholar]
  • 51.Wang, Y. & Hu, X. Determination of tensile strength and fracture toughness of granite using notched three-point-bend samples. Rock Mech. Rock Eng.50, 17–28. 10.1007/s00603-016-1098-6 (2017). [Google Scholar]
  • 52.Hu, X., Guan, J., Wang, Y., Keating, A. & Yang, S. Comparison of boundary and size effect models based on new developments. Eng. Fract. Mech.175, 146–167. 10.1016/j.engfracmech.2017.02.005 (2017). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets generated during and analyzed during the current study are available from the corresponding author on reasonable request.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES