Dongqiao Liu ,Jie Sun ,Pengfei He ,Manchao He ,Binghao Cao ,Yuanyuan Yang
a State Key Laboratory for Geomechanics and Deep Underground Engineering,China University of Mining and Technology (Beijing),Beijing,100083,China
b School of Mechanics and Civil Engineering,China University of Mining and Technology (Beijing),Beijing,100083,China
c Institute for Deep Underground Science and Engineering,China University of Mining and Technology (Beijing),Beijing,100083,China
Keywords: Rockburst Water Prevention effect Crack evolution
ABSTRACT To investigate the mechanism of rockburst prevention by spraying water onto the surrounding rocks,15 experiments are performed considering different water absorption levels on a single face.High-speed photography and acoustic emission (AE) system are used to monitor the rockburst process.The effect of water on sandstone rockburst and the prevention mechanism of water on sandstone rockburst are analyzed from the perspective of energy and failure mode.The results show that the higher the absorption degree,the lower the intensity of the rockburst after absorbing water on single side of sandstone.This is reflected in the fact that with the increase in the water absorption level,the ejection velocity of rockburst fragments is smaller,the depth of the rockburst pit is shallower,and the AE energy is smaller.Under the water absorption level of 100%,the magnitude of rockburst intensity changes from medium to slight.The prevention mechanism of water on sandstone rockburst is that water reduces the capacity of sandstone to store strain energy and accelerates the expansion of shear cracks,which is not conducive to the occurrence of plate cracking before rockburst,and destroys the conditions for rockburst incubation.
With the increase in number of underground excavations,rockbursts are becoming a significant challenge in rock engineering practice(Hoek et al.,1995;He et al.,2012;Li et al.,2019;Feng et al.,2019).Rockburst is a nonlinear dynamic phenomenon characterized by instantaneous energy release from a rock mass with accumulated strain energy surrounding an excavation face (He et al.,2012;Mazaira and Konicek,2015),and thus poses a great threat to the construction safety of deep tunnels under high geostress conditions.Rockbursts have become an important topic of research(Bradley et al.,2020;Kaiser and Moss,2022;Lukasz et al.,2022;Li et al.,2023;Mao et al.,2023),because they pose high risk,involve complex mechanism,and are influenced by multiple factors (Sun et al.,2016;Selahattin et al.,2018;He et al.,2021).In deep tunnels,rockbursts seldom occur when the surrounding rock is waterrich.However,relatively dry rock masses are prone to rockburst(Ortlepp,1997;Fowkes,2011).For example,Ortlepp (1997) found that rock masses with a high water content encountered fewer rockbursts.Additionally,Fowkes (2011) reported that rockburst activity is much lower in tunnels with water and mud.Therefore,water absorption level will reduce the severity of rockburst.
The predominant influence of water on rocks is to weaken the mechanical properties of rocks (Cherblanc,et al.,2016;Liu et al.,2018a;Bao et al.,2022;Caselle et al.,2022;French et al.,2022).To understand the effect of water on rockburst,scholars have conducted true triaxial(Sun et al.,2016;Luo et al.,2019)and biaxial(Liu et al.,2018b;Luo,2020) rockburst tests under different water contents.In terms of acoustic emission (AE),the increase in water content reduces the accumulated AE count and energy (Sun et al.,2016;Liu et al.,2018b;Luo et al.,2019;Zhang et al.,2020),as well as the AE main frequency signal(Sun et al.,2016).In terms of rockburst failure mode,rock samples have shown a change from tension-shear mixed failure to single shear failure(Sun et al.,2016;Luo et al.,2019),and the kinetic energy of rockburst has been found to reduce with an increase in water content (Liu et al.,2018b;Luo et al.,2019).
Further,studies have been performed on how water affects rockburst considering different factors such as strength (Luo et al.,2019;Cai et al.,2021)and energy(Luo,2020).Liu et al.(2021)stated that water diminishes the rockburst conditions and Luo (2020)recognized that water causes reduction in the residual elastic strain energy of sandstone,thus avoiding energy concentration and preventing rockburst.Cai et al.(2021) reported that water absorption can prevent rockburst because water weakens the energy storage capacity of the surrounding rocks and reduces the stress near the working face.
There has been progress in understanding the relationship between water and rockburst.The water content of a specimen was considered in the corresponding laboratory rockburst experiments.These research efforts can only explain why rockbursts do not occur in surrounding rocks with higher water content or why the intensity of rockbursts is reduced,but cannot fully explain how spraying water on the surrounding rocks prevents rockburst(Song et al.,2014).This is because after spraying water on the surrounding rocks,the water on the surface of the surrounding rocks is nonlinearly distributed from saturation to relative dryness (Simona et al.,2016;Cai et al.,2022).However,previous laboratory studies only examined the relationship between water content and rockburst.As a result,the water-bearing state of the specimen did not reflect the occurrence state after spraying water on the surrounding rocks.According to some on-site rockburst cases,it is likely for rockburst to occur on the side wall (Feng et al.,2019;Yang et al.,2022).Therefore,it is necessary to conduct experimental research on the water-bearing state of side wall of the surrounding rocks after water spraying,to further evaluate the influence of different water absorption levels on the intensity of rockburst,and reveal the prevention effect of water absorption level on rockburst.
This study focused on the mechanism of preventing rockburst with spraying water on the surrounding rocks.Accordingly,single side of sandstone was absorbed with water to simulate the process of water infiltrating the interior from the surface after spraying to the surrounding rocks.The rockburst experiment was performed using a self-developed rockburst experimental system.High-speed photography and AE system were used to record the entire process of the experiment.The influence of water on rockburst and the mechanism of preventing rockburst by spraying water were studied from the perspective of energy and failure mode.
The rock samples were red sandstone with excellent integrity and uniformity collected from Shandong Province,China.They were prepared as cuboids with dimensions of 150 mm × 60 mm × 30 mm (Fig.1).The average density of the sandstone sample was 2.56 g/cm3and the average P-wave velocity was 3160 m/s.The average uniaxial compressive strength of these sandstone samples is 113.24 MPa,the average elastic modulus is 23.14 GPa,and the average Poisson’s ratio is 0.33.

Fig.1.The tested specimens: (a) Photo of sandstone,and (b) Geometry.
The single-face water absorption method was adopted to simulate the water spraying on the surrounding rocks.The basic concept is to wrap the side of the sandstone with an impervious film and fill the contact surface between the film and the sample with filler (Fig.2a),so that the sandstone can only absorb water from a single face.Fig.2b presents a schematic diagram of the single-face water absorption of sandstone.Fig.2c shows a typical single-face water-absorption curve for sandstone.As can be seen,the water absorption rate of the sample increased slowly after 30 h and then stopped increasing after 48 h.The water absorption rate after 48 h was regarded as single-face saturated water absorption,and the water absorption level of the sample was considered as 100%.The experiment was designed for different water absorption levels of 0%,25%,50%,75% and 100%.According to the water absorption curve,the times required for water absorption levels of 25%,50%,and 75% were 2 h,12 h and 24 h,respectively.The sandstones with five different water absorption levels were used in the experiment,with three specimens per group.

Fig.2.(a) Specific details,(b) Water absorption diagram,and (c) Measured water absorption curve.
Fig.3a shows the stress change state of the rock units before and after excavation.It was in a three-dimensional (3D) stress state before excavation,and a free face was generated after excavation.The stress perpendicular to the free face becomes zero and the rock mass unit changes from a six-face stress state to a five-face loading state(He et al.,2012).Fig.3b shows the typical stress loading path of a rockburst (He et al.,2012;Liu et al.,2021).Based on the different loading levels,the entire loading process can be divided into five stages:(1)stage I corresponds to the process of loading to the preset stress level;(2) stage II is the holding process after loading to the preset stress level,simulating the stress state of the sandstone before excavation;(3) stage III is the process of unloading σhto 0 MPa,which is used to simulate the process of underground rock mass subjected to excavation and unloading;(4)stage IV is the load retention process after unloading;and(5)stage V involves loading σVto rockburst,which is used to simulate rockburst caused by stress concentration.

Fig.3.Stress transformation process in rockburst test: (a) Schematic diagram of stress transformation process of the rock element when unloading,and (b) Loading path.
The in situ stress was calculated according to the regression formula for ground stress in Shandong Province,China (Li et al.,2017),as follows:
where σHdenotes the maximum horizontal principal stress,σhdenotes the minimum horizontal principal stress,σVdenotes the vertical stress,andHdenotes the underground depth.The in situ stress level of 1000 m was selected as the initial stress in this study.According to Eq.(1),σH=33.63 MPa,σV=26.36 MPa,and σh=21.83 MPa.
He et al.(2012) conceptualized and designed a rockburst experimental system,as depicted in Fig.4a.The testing system includes the loading system(Fig.4b),AE monitoring system,force monitoring system,and high-speed photograph system (Fig.4c).Sudden and rapid unloading on one side can be realized under the condition of true triaxial loading state to form the stress state and geometric boundary conditions of rockburst (He et al.,2012).The maximum working pressure of the loading system was 450 kN.The loading system was independent in three directions,and uniform loading of the samples was realized through three rigid indenters.Sudden unloading can be carried out in one direction so that the force transfer bar and loading indenter can drop rapidly to simulate the rapid unloading process caused by excavation in engineering.

Fig.4.Rockburst testing system: (a) Sketch map of rockburst testing system,(b) Loading system,(c) AE and high-speed photograph system,and (d) AE sensor layout.
The AE signals were monitored with the Micro-II AE system.The system sampling rate was set to 1 MSPS,the threshold value was set to 40 dB,and the AE sensors were uniformly set on the diagonal on the side of the specimen.After the ultrasonic coupling agent was applied on the surface of AE sensors,they are closely connected with the surface of the specimen under the spring pressure in the prefabricated hole in side platens(Fig.4d).
To observe the fragment ejection process during fast-unloading rockburst tests under the conditions of different water absorption levels,high-speed camera technology (frame rate of 1000 f/s) was adopted in this study.Fig.5 shows the failure ejection process of fast-unloading rockburst for different levels of water absorption.

Fig.5.Typical phenomena of rockburst under different water absorption levels.
Fig.5 shows that the fragment ejection process generally exhibits the same characteristics.First,cracks appear on the free surface of the sandstone,then small fragments are spalled and ejected,and finally,the rockburst occurs on the entire exposed sample surface.Owing to the presence of water,there are some minor differences in the rockburst process under different water absorption levels.From examining the rockburst phenomenon of sandstone with five water absorption levels,it is clear that the effect of water on rockburst can be mainly manifested in the following two aspects: (1) With the increase in the water absorption level,the change in the jetting direction of debris during sandstone rockburst shows a trend as: a large amount of debris flying downward and upward →a small amount of debris peeling from free-surface;(2) The number of fragments ejected during a rockburst decreases with increasing water absorption level.These phenomena indicate that the rockburst intensity decreases with increasing water absorption level.
An apparent V-shaped pit was observed on the side of the specimen after the rockburst.The shape features of the rockburst pit reflect the rockburst intensity.Fig.6 shows the extraction process for shape features.The characteristic shape parameters of the rockburst pit under different water absorption levels are summarized in Table 1.

Table 1Characteristics of rockburst pit.

Fig.6.Extraction process of rockburst pit.
It can be seen from Table 1 that after absorbing water from a single side,the rockburst pit characteristics exhibit different changing laws,i.e.with increasing water absorption level,the pit angle increases,while the pit depth decreases.Meanwhile,the pit area decreases with increasing water absorption level.Summarily,with the increase in water absorption level,the pit shape changes from a narrow and deep V-shaped feature to a wide and shallow Vshaped feature,which indicates the less violent rockburst (Wang et al.,2019).
3.3.1.AE count
According to the real-time collection of AE events,the relationship between the AE count and cumulative count of the entire rockburst process is plotted in Fig.7a-e.As can be seen,the AE count features in the rockburst experiments with the five water absorption levels exhibit similar time distribution characteristics,i.e.the AE counts all reach the maximum value at the moment of rockburst.

Fig.7.AE count characteristics of sandstone rockburst with different water absorption levels: (a-e) The samples with water absorption levels of 0%,25%,50%,75% and 100%,respectively;and (f) Cumulative AE count of different water absorption levels.
The cumulative values of the sandstone rockbursts under different water absorption levels were calculated and are plotted in Fig.7f.Accordingly,it can be seen that with increasing water absorption level,the cumulative count exhibits a downward trend.The average values are 15.36 × 105,7.65 × 105,6.73 × 105,5.85 × 105,and 1.43 × 105,respectively,as listed in Table 2.Compared with the water absorption level of 0%,the cumulative count of sandstone at the water absorption levels of 25%,50%,75% and 100% decreased by 50.27%,62.34%,66.42% and 90.71%,respectively.The results demonstrate that the higher the water absorption level of sandstone,the lower the AE activity and the less severe the damage in sandstone.

Table 2AE characteristics of rockburst.
3.3.2.AE energy
The time distribution of the AE energy in the sandstone rockburst experiment was similar to that of the counting feature and reached its maximum value at the moment of rockburst,as shown in Fig.8a-e.The absolute AE energy reflects the AE signal intensity.Fig.8f presents the cumulative value of the AE energy during the entire rockburst process.As shown in Fig.8f,with increasing water absorption level,the AE cumulative energy exhibits a downward trend: 2.49 ×109aJ,1.27 ×109aJ,1.09 ×109aJ,0.82 ×109aJ and 0.05 ×109aJ,respectively,as listed in Table 2.Compared with the sandstone rockburst at the water absorption level of 0%,the AE cumulative energy at water absorption levels of 25%,50%,75%,and 100% decreased by 49%,56.22%,67.23% and 97.95%,respectively,indicating that with the increase in water absorption level,the energy released by the rockburst decreased and the rockburst became less intense.

Fig.8.AE energy characteristics of sandstone rockburst with different water absorption levels: (a-e) The samples with water absorption levels of 0%,25%,50%,75% and 100%,respectively;and (f) Cumulative AE energy under different water absorption levels.
3.4.1.Crack classification method
In rock mechanics,different types of cracks exhibit different AE characteristics.In general,AE signals with low AF (average frequency,unit: kHz) and high RA (rise time/amplitude,unit: ms/V)values commonly represent the generation or development of shear cracks.In contrast,high AF and low RA values represent the generation or development of tensile cracks (Fig.9a) (Ohtsu et al.,2007;Ohno and Ohtsu,2010;Wang et al.,2019).

Fig.9.(a) AE crack classification chart,and (b) Basic schematic diagram of SVM.
Binary classification method was adopted in the classification of tensile and shear cracks.Consequently,the support vector machine(SVM),which is an important method for classification in the fields of pattern recognition and machine learning (Zhou 2016;Abolfazl et al.,2022),was introduced to the classification of tensile and shear cracks.
For the linear classification problem,suppose the training dataAare given by the input data X={x1,x2,x3,…,xn},and the learning target Y={-1,+1},the hyperplaneLwill make the data separable:
where ω is the normal vector andbis the intercept of the hyperplane(Zhou 2016).
For the nonlinear separable problems,the data can be mapped from the original space to the feature space,and the data can often be linearly separable in the feature space,i.e.a hypersurfacePwill make it separable:
where φ is a mapping function.
Because it is difficult to calculate the inner product of the mapping function,the inner product of the mapping function is usually defined as a kernel function k to avoid the display calculation of the inner product.Among them,the commonly used kernel function is Gaussian radial basis kernel function,i.e.k(X1,X2)=exp(-(||X1-X2||2)/(2σ2)) (Zhou,2016).The basic schematic diagram of SVM is shown in Fig.9b.
In this paper,the flow chart of crack classification method based on SVM is shown in Fig.10a.Specifically,it can be divided into four steps:
(1) Step 1: Data acquisition.The Brazil splitting test and direct shear test are conducted to obtain the AE parameters for each experiment.
(2) Step 2: Data preprocessing.The RA-AF density nephograms of the Brazil splitting test and direct shear test are respectively drawn in this step.According to the understanding of the crack types in the Brazilian test and direct shear test,the dark area in the RA-AF density nephograms of the Brazilian test and direct shear test represents most AE parameters of tensile and shear cracks,respectively.
(3) Step 3: SVM modeling.Output setOand input setIare defined.Hence,the cracks division model described asf:I→Oin the mathematics could be established based on SVM.I={(RA,AF)|RA∈S1∪S2,AF∈S1∪S2},where S1and S2are the dark parts in RA-AF density nephograms of the Brazilian test and direct shear test,respectively.O={“1”,“2”},where “2”and “1” are the labels corresponding to tensile and shear cracks,respectively.
(4) Step 4: Evaluation of results.The model is applied to crack classification in the Brazil splitting test and direct shear test to evaluate the rationality of the crack classification model.Fig.10b and c shows the confusion matrix of the SVM model.In general,more than 86% of the AE data could be correctly classified.
The classification result of rockburst experiment is shown in Fig.10d.It can be seen that the boundary between tensile and shear cracks is a nonlinear curve which could be described as
3.4.2.Crack evolution characteristics
To further obtain the crack evolution characteristics in the entire rockburst process,Eqs.(5)and(6)are used to calculate the ratios of the number of tensile cracks to the number of shear cracks per minute.The results are plotted in Fig.11.

Fig.11.Ratios of cracks per minute in sandstone rockburst with different water absorption levels.(a-e)Represent the samples with water absorption levels of 0%,25%,50%,75% and 100%,respectively.
wherertis the proportion of tensile cracks,rsthe proportion of shear cracks,Ntis the number of tensile cracks per minute,Nsis the number of shear cracks per minute,andNis the total number of microcracks per minute.
It can be observed from Fig.11 that the change in the crack proportion is closely related to the force application:
(1) At stage I,numerous types of cracks are generated,and the proportion of tensile cracks and shear cracks exhibit a zigzag interlacing characteristic owing to the compression and extension of primary microcracks in the sandstone.
(2) At stage II,a silent emission signal is generated,and the crack does not expand due to the load-preserving at this stage.
(3) At stage III,with the rapid unloading of σhto 0 MPa,the internal stress state of the sandstone changes dramatically,and the proportion of tensile cracks decreases.
(4) At stage IV,because of the load-preserving,a silent emission signal is generated,and the crack does not expand.
(5) At stage V,the proportion of tensile cracks remains higher than that of shear cracks,indicating that the extension of tensile cracks is still dominant in the entire rockburst process.However,in a more detailed view,the crack evolution before rockburst experienced two processes at stage V.Generally speaking,the proportion of tensile cracks remains unchanged firstly (stage V′),and then,the proportion of tensile cracks decreases and that of shear cracks increases before rockburst (stage V′′).The above analysis results indicate that rockburst experiences the process of dominant tensile cracks and the process of interaction of tensile and shear cracks.In addition,they also reflect that the extension of tensile crack is the prerequisite of rockburst,and the shear crack is an important factor that causes the occurrence of rockburst.
3.4.3.Influence of water absorption level on crack evolution
To further obtain the influence law of water absorption level on the crack evolution during rockburst,relevant parameters of crack evolution before rockburst were counted,including the duration of stable evolution of tensile crack(the duration of stage V′)(Fig.12a),the decreasing amplitude(Fig.12b),and the decreasing rate of the proportion of tensile crack proportion (Fig.12c) before the rockburst.

Fig.12.The influence of water absorption level on crack evolution characteristics: (a) Duration of stable evolution of tensile crack,(b) Decrease amplitude of tensile crack proportion,and (c) Decrease rate of tensile crack proportion.
As shown in Fig.12a,the duration of stable evolution of tensile crack shows a decrease trend with the increase of water absorption level,indicating an inhibitory effect of water on the stable evolution of tensile cracks and a promotional effect of water on shear crack evolution.In addition,as observed from Fig.12b,with increasing water absorption level,the decrease in the range of tensile cracks before rockburst increases.Compared with the water absorption level of 0%,the decrease in the proportion of tensile cracks at the water absorption levels of 25%,50%,75% and 100% increases by 12.37%,16.5%,66.85% and 95.55%,respectively.Meanwhile,as can be seen from Fig.12c,the decreasing rate of the proportion of tensile cracks at stage V presents an upward trend with increasing water absorption level.Compared with the case of 0% water absorption level,the decrease rates of tensile cracks at the water absorption levels of 25%,50%,75% and 100% increase by 8.19%,22.61%,85.72% and 204.18%,respectively.These results indicate that the water promotes the development of shear cracks and weakens the degree of spalling failure caused by tensile cracks.
In summary,the effect of water on crack evolution before rockburst is reflected in the following aspects.With the increase of water absorption level of sandstone,the duration of stable evolution of tensile cracks is shortened,and then,more rapid propagation of shear crack will be induced,resulting in a greater decrease in the proportion of tensile crack before rockburst.
The particle image velocity(PIV)technique is a common method used to determine the initial ejection velocity of fragments in rockburst experiment (Su et al.,2017;He et al.,2020).In order to evaluate the effect of water on rockburst intensity,single camera was used to record the ejection process during rockburst.The system sampling rate was set to 1000 fps.
In the calculation results of PIV,the ejection velocity of fragments in the rockburst area was mainly considered.Fig.13a-e shows the distribution curve of the average velocity in rockburst area under different water absorption levels.It can be seen from Fig.13a-e that the mean ejection velocity of fragments shows a trend of increasing first and then decreasing slowly during rockburst under different water absorption levels.

Fig.13.Influences of water on tensile crack: (a-e) The samples with water absorption levels of 0%,25%,50%,75% and 100%,respectively;and (f) Maximum ejection velocity of fragments under different water absorption levels.
Fig.13f shows the maximum average ejection velocity during rockburst under different water absorption levels.It can be seen from Fig.13f that the water has a great impact on the ejection velocity of fragments.Under the water absorption level of 0%,the ejection velocity of fragments is the maximum,with an average value of 2.85 m/s.With the increase of water absorption degree,the maximum ejection velocity of fragments gradually decreases.Compared with the water absorption level of 0%,the maximum ejection velocities of fragments at the water absorption levels of 25%,50%,75% and 100% decrease by 0.67%,3.04%,29.79% and 71.23%,respectively.
Relevant studies show that the magnitude of rockburst intensity is related to the initial ejection velocity of fragments (Feng et al.,2019).Table 3 lists the characteristics of rockburst with different intensities.Based on Table 3 and Fig.13,the magnitude of rockburst intensity under different water absorption levels can be summarized in Table 4.It can be seen from Table 4 that the average ejection velocity decreases with the increase in water absorption level,and the magnitude of rockburst intensity is medium under the water absorption level from 0% to 75%,and slight for the sandstone under the water absorption level of 100%.The results show that the sandstone rockburst can be prevented after absorbing water from single side.In addition,with the increase in water absorption level,the reduction in the depth of the rockburst pit reflects that less debris is produced during the rockburst,and the rockburst is less hazardous at a lower ejection velocity.

Table 3Characteristics of rockburst with different intensities (Feng et al.,2019).

Table 4Characteristics of rockburst with different intensities.
Fig.14 shows the stress evolution of sandstone rockburst in the form of a stress circle.Taking the dry state(Fig.14a)as an example,the stress state of a sandstone can be represented by the stress circle CIwhen the sandstone is a 3D six-face stress state in the initial in situ stress.The minimum principal stress σ3decreases to 0 MPa at the formation of a free surface,and then,the stress state can be represented by the stress circle CII.Since the stress state at this time has not yet been tangent to the envelope in dry state,i.e.it has not reached its limit state,hence the rockburst has not occurred.With the loading of σv,the rockburst will occur when the circle reaches its limit (σ1reaches σb).

Fig.14.Stress evolution of sandstone rockburst in (a) dry state and (b) water state.
Similarly,for sandstone in water state,the stress circle CIturns into CIIas a result of the unloading of σ3.Under the softening effect of water,the envelope of sandstone gradually decreases from a black line to a blue curve (Fig.14b).Due to this fact,σvreaches its limit state only when it increases to σ′b(σ′b<σb),and the rockburst occurs immediately.In other words,with the increase in water absorption level,the softening effect of water on sandstone is more obvious,which leads to the greater decline of the envelope,and the smaller peak stress of rockburst.
Fig.15 shows that the peak stress of the rockburst decreases gradually with increasing water absorption level,indicating that the water has a certain weakening effect on the mechanical properties of rocks.Compared with the 0% water absorption level,the average peak stresses of the rockburst at the water absorption levels of 25%,50%,75% and 100% decrease by 2.23%,4.86%,18.33% and 41.07%,respectively.Thus,with higher water absorption level,the weakening effect of water on rock is more significant.

Fig.15.Influence of water on peak stress.
Because the weakening degree of water on the strength of sandstone is greater than that of water on the elastic modulus of sandstone (Luo,2020),according to the calculation formula of elastic strain energyU=σ2/(2E),the greater the water absorption level,the lower the accumulated elastic strain energy of sandstone.Studies have shown that the kinetic energy of rockburst fragments originates from the elastic strain energy accumulated in the rock before rockburst (He et al.,2012).Because the water reduces the elastic strain energy stored in the process of sandstone rockburst,when rockburst occurs,the kinetic energy converted into fragment ejection decreases,and the rockburst intensity is lower.Furthermore,the AE cumulative energy is positively correlated with the true released energy due to cracking activities within the rock material.Table 2 and Fig.8f show that the AE cumulative energy exhibits a downward trend with increasing water absorption level.Therefore,this demonstrates,from energy perspective,that the higher the water absorption level is,the less the released energy is,and the less intense the rockburst is.
Several studies have been conducted on the incubation conditions of rockburst.He et al.(2012) pointed out that the evolution process of rockburst can be divided into three stages:vertical plate cracking,vertical plate buckling deformation,and rockburst failure,and then proposed a plate structure evolution model of rockburst(Fig.16).Gong et al.(2018)indicated that the plate fracture failure can form unstable buckling rock plates,creating conditions for the sudden release of energy from rockbursts.Liu et al.(2021) summarized that the four stages of rockburst are extension crack propagation,splitting into plates,shearing into blocks,and block ejection.Based on the above research on the rockburst incubation process,plate cracking caused by the tensile fracture of hard rock under the condition of excavation and unloading(Diederichs et al.,2004)is the key stage of the rockburst incubation process(He et al.,2012),and the propagation of shear cracks contributes significantly to rockburst.

Fig.16.Evolution processes of rockburst: (a) Vertical plate cracking,(b) Vertical plate buckling deformation,and (c) Rockburst failure (He et al.,2012).
Fig.17 presents a typical picture of the failure mode on the side of the sample after the rockburst.Considering the sandstone sample with 0% water absorption level,several parallel tensile cracks (Fig.17a) can be seen near the rockburst pit (yellow area in Fig.17) after the rockburst,indicating that the sandstone with 0% water absorption level has a high degree of plate cracking characteristics.Meanwhile,the shear cracks can be observed in the upper left corner of the sandstone specimen,and there are multiple tensile cracks communicating and penetrating the side of the specimen,indicating that the rockburst results from the joint action of the tensile and shear cracks.When the water absorption level is 25%,a tensile crack is observed on the lower side of the sandstone explosion pit,which is connected to the rockburst pit.However,at the water absorption levels of 50%,75% and 100%,there is no evident tensile crack distribution around the rockburst pit.Meanwhile,the observation of shear cracks on the side of the sandstone specimens shows that the distribution of shear cracks on the side of the sandstone becomes more abundant with increasing water absorption level.

Fig.17.Distributions of cracks near the pit: (a) 0%,(b) 25%,(c) 50%,(d) 75%,and (e) 100%.
According to the above phenomena,with the increase in water absorption level,the decreasing rate and decreasing range of the proportion of tensile cracks increase before rockburst.After the sandstone absorbs water from a single side,the friction between particles is reduced by water,thus the shear cracks propagate more easily.Subsequently,the shear crack expands rapidly and causes rockburst in the absence of sufficient plate cracking in the rock interior.As a result,the incubation conditions of rockburst are destroyed,and then the rockburst intensity is reduced.
In this study,the sandstone was contrived to absorb water from a single side to simulate the process of water infiltrating the interior from the surface after spraying on the surrounding rock.The AE system and high-speed photography were used to monitor the entire rockburst process in real time.The results reveal the influence of water on sandstone rockbursts.The main conclusions of this study are drawn as follows:
(1) The rockburst processes of sandstone under different water absorption levels are similar,including particle ejection,continuous ejection,and comprehensive rockburst,and the rockburst intensity decreases with increasing water absorption level,as evidenced by the decreases in AE cumulative energy and shallower depth of the rockburst pit.
(2) The crack evolution characteristics before rockburst indicate that the extension of tensile crack is the dominant crack type of rockburst,and the shear crack is an important factor that causes the occurrence of rockburst.The effect of water absorption level on crack evolution during rockburst is reflected in that the water will shorten the duration of stable evolution stage of tensile crack,and induce a more rapid propagation of shear crack before rockburst.
(3) The effect of water absorption level on sandstone is reflected in the softening effect of water on sandstone.The water reduces the ability of sandstone to store elastic strain energy,resulting in reduced elastic strain energy that can be released during rockburst failure,which manifests as a change from violent ejection under low water absorption to detrital spalling under high water absorption.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The financial support from the National Natural Science Foundation of China (Grant Nos.52074299 and 41941018) and the Fundamental Research Funds for the Central Universities of China(Grant No.2023JCCXSB02) are gratefully acknowledged.
Journal of Rock Mechanics and Geotechnical Engineering2024年1期