Effect of water on the damage and energy dissipation feature of coal under uniaxial cyclic loading–unloading condition

To gain insight into the reduction mechanism of water on the dynamic destabilization failure character of coal under mining disturbance conditions, the effect of water on the damage and energy dissipation feature of coal was investigated via cyclic uniaxial compressive loading–unloading test, acoustic emission (AE) monitoring technology, and theoretical analysis. It manifests that the softening effect of water on the mechanical properties of coal is still remarkable, lower uniaxial compressive strength and elastic modulus, while greater axial strain and higher Poisson's ratio are observed under the cyclic loading–unloading condition. Water‐saturated specimens dissipate more energy during the cyclic loading–unloading process, which reduces the proportion of input energy accumulated in the coal mass, thus lowering the burst proneness of coal. A greater amount of damage generates in air‐dried specimens than that in water‐saturated specimens at the first cyclic loading–unloading cycle, while lower damage increments are observed in the following cycles. This may be the reason for the minimal energy dissipation of air‐dried specimens in the following cycles. The elastic modulus behaves logarithmically versus the increasing cyclic loading–unloading cycle counts. The energy dissipated due to the damage evolution and plastic deformation of coal specimens positively correlates with the energy released by AE activities, the AE energy in air‐dried specimens has a greater concentration around the peak axial stress and a stronger correlation with the dissipated energy. The cumulative AE energy and energy dissipation density are linearly correlated during the cyclic loading–unloading process, which indicates a possibility in estimating the damage evolution and the dynamic failure risk of coal via cumulative AE energy, under the cyclic loading–unloading condition.


| INTRODUCTION
6][7] Owing to the increases in geostress and mining-induced stress caused by the ascending of the mining depth, 8,9 the threat of the coal burst on the safe and efficient production of coal mines is becoming more prominent in recent years.Therefore, it is crucial to find approaches to reduce the occurrence of the coal burst.
2][23] The existence of water is generally considered to weaken the mechanical properties of coal.Water-bearing coal usually has lower uniaxial compressive strength (UCS), 24,25 shear strength, 26 and elastic modulus, 27,28 while greater the peak strain, 29 expanded pore sizes, 24 declined acoustic emission (AE) activities. 302][33][34][35][36] Due to the roadway excavation and the rupture of the overlying rock layer occurring during the real coal mining process, 37,38 coal mass around mining areas is usually subject to various loading conditions, for instance, the stress impact, 39,40 stress concentration, 41,42 or stress relief. 43,44This complicates the various instability and failure mechanisms of coal mass.To reveal the mechanical response of coal during various loading conditions, many investigations were thus implemented, such as the impact load experiment conducted by Hopkinson compression bar, 45,46 uniaxial or triaxial compressive cyclic loading and unloading experiments, 47,48 and rock mechanics experiments with different loading paths. 49ompared with the loading approaches mentioned above, the cyclic loading-unloading testament is an effective approach in revealing the mechanical response of coal and rock under repeated disturbance of excavation, 47,50,51 through parameters, such as peak strength, elastic modulus, and energy dissipation features. 33,47,48,52hough many relevant reports were implemented, the result has manifested that cyclic loading-unloading has a more complicated influence on the mechanical response of coal and rock, 47,[51][52][53][54][55][56][57][58] in both uniaxial and triaxial conditions, since the influence of it on some parameters varies with rock types, loading rates, and other loading conditions.
Investigating the influence of water under the cyclic loading-unloading condition is significant in understanding the coal burst prevention and mitigation mechanism resulting from water infusion under the real mining environment. 15,46,59,60To obtain the failure mechanisms of coal mass during various excavation processes affected by water, numerous attempts were implemented, and corresponding mechanical properties variations were reported, such as elastic modulus, peak strength, and AE feature. 24,26,27,61This provides us with a beneficial reference for investigating the coal and rock instability and failure mechanism under cyclic loading-unloading conditions.However, due to the investigation limitations, the effect of water on the burst proneness, the damage evolution, and the energy dissipation features under the cyclic loading-unloading conditions has not been fundamentally investigated, nor fully understood.
Thus, in this investigation, the roles of water on the coal bump mitigation under mining-induced cyclic loading-unloading conditions were fundamentally investigated.A series of cylinder coal specimens were processed and, respectively, loaded under the air-dry and watersaturated state.The mechanical properties of UCS, elastic modulus, and Poisson's ratio were calculated, and energy dissipation and damage evolution features were revealed, based on the uniaxial cyclic loading-unloading testament, AE measurement, and theoretical analysis.

| Specimen preparation
The coal blocks were excavated from burst-prone coal seam.In this research, nine cylindrical specimens were drilled and polished with a diameter of 38.0 mm and a height-to-diameter ratio of 2.0, based on the recommendation of the International Society for Rock Mechanics, 62 as shown in Figure 1.
After being processed, specimens were randomly separated into three groups (namely, groups 1-3), specimens in group 1 were used for conventional uniaxial compressive tests, specimens in group 2 were tested under cyclic loading-unloading condition, and specimens in group 3 were chosen for water saturation and then been tested under the same cyclic loading-unloading condition.

| Water saturation test
Specimens for each group were air-dried, and then specimens in group 3 were chosen for the water saturation test.During the measurement, specimens were completely immersed in water and taken out as being weighted.The absorbed water content curves of specimens are shown in Figure 2. Obviously, the water content increases sharply at the first 400 min, and then gains gradually to a constant, after ~4000 min (2.78 days), with an average saturation content of 5.62%, respectively, ~6.22%, ~5.66%, and ~4.98% in three specimens.

| Experimental instruments
The uniaxial compressive and the cyclic loadingunloading tests were implemented by a servo control uniaxial loading frame (MTS).It has a maximum axial loading capability of 300 kN, two loading approaches are available, namely, displacement control and stress control.In this study, the uniaxial compressive test was first carried out on the specimens of group 1, with a stress control approach, considering the stress loading rate in coal mine less than 0.1 MPa/s is considered as the static load, 63 thus, the loading and unloading velocity here were chosen as 0.025 MPa/s.The UCS of group 1 was obtained among 14.08-16.55MPa.
To investigate the damage and energy evolution feature of coal, ensuring sufficient experimental data and cyclic loading-unloading cycles were necessary for being tested specimens.Thus, the beginning cyclic loading-unloading threshold value was chosen as 5.5 MPa, with an increasing interval of 0.5 MPa for each loading-unloading cycle, this ensured that each sample undergoes at least 10 cycles of cyclic loading and unloading (the maximum cyclic loading-unloading counts is 13).For specimens with cyclic loading-unloading counts greater than 13, the specimen was kept loaded until to destruction, with the loading velocity of 0.025 MPa/s.The cyclic loading-unloading process is shown in Figure 3.Meanwhile, the air-dried (group 2) and water-saturated (group 3) specimens shared the same cyclic loading-unloading approach.
The AE activities were monitored and recorded by an AE monitoring system (Physical Acoustics Corporation) during the testament.It consisted of a PIC-2 device, six micro-30 s sensors, and a six-type preamplifier.The PIC-2 device had a bandwidth frequency of 1 kHz-3 MHz with a maximum signal amplitude of 100 dB and a dynamic range greater than 85 dB.In this test, four microsensors were utilized, the loading frame system was shown in Figure 4, the amplification gain was set as 40 dB.

| Uniaxial compressive strength
The UCS and axial strain under different loading forms obtained in this experiment are summarized in Table 1.The UCS of natural coal specimens are 14.08, 16.55, and 14.95 MPa, with a mean value of 15.19 MPa.For air-dried specimens under the cyclic loading-unloading condition, it with values of 13.51, 9.57, and 13.96 MPa, however, considering the value of 9.57 MPa is far below the mean UCS value, which may be caused by significant cracks existing in this specimen.Thus, it is reasonable to discard this data and reduce the impact induced by accidental factors.Accordingly, the average UCS value of air-dried specimens is 13.74 MPa.Meanwhile, water-saturated specimens separately have UCS values of 11.76, 11.03, and 13.97 MPa, with an average value of 12.25 MPa.
Compared with natural specimens, the UCS of coal reduces under the cyclic loading-unloading condition, and the average reduction is 1.45 MPa, approximately 9.55%, as the loading approach changes from conventional loading to cyclic loading-unloading condition.Water weakens the UCS under the loading-unloading conditions, the decrement is 1.49 MPa, approximately 10.84%.The coupling effect of water and cyclic loading-unloading has a greater impact on the UCS, with a reduction of 2.94 MPa, approximately 19.3%.
Meanwhile, the axial strain is also affected by the cyclic loading-unloading condition and the water.The natural air-dried specimen under the conventional loading condition has greater axial deformation capability, with a value of 1.63%, than that of 1.38% and 1.57%.Evidently, water softens the deformation capability of coal, greater average axial strain value is observed in water-saturated specimens (1.57%) than that in the airdried specimens (1.38%).

| Elastic modulus
As a fundamental mechanical parameter, elastic modulus manifests the stress and deformation capability of coal.During the cyclic loading-unloading process, the elastic modulus has also been considered a damagerelated parameter in coal and rock. 58,64hus, the tangent modulus of elasticity during the loading and unloading process of the whole cyclic loading-unloading process were, respectively, calculated in this experiment, the tangent elasticity modulus during  loading cycles of per cyclic loading-unloading cycles are summarized in Figure 5. Within the 13 cyclic loading-unloading cycles, the tangent elasticity modulus under the loading condition in air-dried and watersaturated specimens share a similar increasing character as the cyclic loading-unloading cycle gain, and the increment is significant at the first to second cyclic loading-unloading cycle, as shown in Figure 5A.The elastic modulus of air-dried specimens separately increases from 0.82 to 1.39 GPa and from 0.97 to 1.46 GPa, with an average increment of 0.53 GPa.Meanwhile, the counterpart in the water-saturated specimens, respectively, gains from 0.66 to 1.23 GPa, from 0.84 to 1.32 GPa, and from 0.86 to 1.38 GPa, the average increment is 0.52 GPa.
In general, the average elastic modulus value of air-dried specimens is greater than that of water-saturated specimens, in the loading cycle during the whole cyclic loadingunloading process, as shown in Figure 5B.It has a minimum average value at the first cyclic loading-unloading cycle, respectively, 0.89 and 0.79 GPa, while has the maximum average value in the 13th cyclic loading-unloading cycle, separately 1.42 and 1.31 GPa.This is consistent with that reported by Ding et al. 47 and Zhong et al., 65 which consider that cyclic loading-unloading exhibits a hardening effect on coal mass.
Compared with the loading cycles, the elastic modulus also has an increasing feature, under the unloading cycle condition during the cyclic loading-unloading process.The average elastic modulus of air-dried specimens increases from 1.14 to 1.44 GPa, and it gains from 1.10 to 1.33 GPa in the water-saturated specimens, respectively, with an increment of 0.3 and 0.2 GPa.Which are less than that of under the loading cycle during the cyclic loading-unloading process.However, it increases more smoothly than that observed under the loading cycle condition during the cyclic loading-unloading process, which can be described by the logarithmic function where E represents the elastic modulus under the unloading cycle condition during the determined cyclic loading-unloading cycle, GPa; L c is the unloading cycle with the elastic modulus of E during the determined cyclic loading-unloading cycle; m, n, and E 0 are parameters obtained from regression analysis, parameters of Equation ( 1) are summarized in Table 2.The parameter m exhibits the growth rate of elastic modulus, obviously, it has a greater value in air-dried specimens, respectively, with values of 0.44 and 0.29, than that of water-saturated specimens, 0.22, 0.25, and 0.28.The elastic modulus of the unloading cycle is greater than that of the loading cycle, this may be correlated with the plastic deformation.Plastic deformation accumulates during the loading cycle, and cannot restored during the unloading cycle.This prevents the stress-strain curve return along that record during the loading cycle and results in a greater slope (elastic modulus) of unloading stress-strain curve (Figure 6).

| Poisson's ratio
To evaluate the effect of water on the deformation feature of coal, Poisson's ratio was calculated during the Elastic modulus variation of specimens under the loading condition during the cyclic loading-unloading process: (A) value for different specimens and (B) average value.
loading-unloading process for air-dried and watersaturated specimens, as shown in Figure 7. Unlike the elastic modulus, the values of Poisson's ratio all have a wave-rising character.For air-dried specimens, Poisson's ratio increases from 0.11 to 0.16 and from 0.12 to 0.15, respectively, has a fluctuation range of 0.05 and 0.03.It separately gains from 0.19 to 0.23, from 0.09 to 0.10, and from 0.10 to 0.15, with increments of 0.04, 0.01, and 0.05, in water-saturated specimens.
In general, the average Poisson's ratio of the air-dried specimens is lower than that of water-saturated specimens during the whole cyclic loading-unloading process.This means water strengthens the lateral deformation capacity during the cyclic loading-unloading process, contributing to a greater volumetric strain in watersaturated specimens, as shown in Figure 8.The typical volume strain variation of air-dried (specimen number 2-3) and water-saturated specimens (specimen number 3-3) are summarized in Figure 8.Though the UCS difference of specimen 2-3 and 3-3 is minimal, the volume strain of the water-saturated specimen obviously greater than the air-dried specimens during the whole cyclic loading-unloading process.
The effect of water on the UCS, elastic modulus, and Poisson's ratio under cyclic loading-unloading conditions is consistent with the weak effect reported by previous research in rock-like materials. 66,67In general, water reduces the strength, increases the deformation capability, and results in a more ductile failure of coal in cyclic loading-unloading conditions.

| ENERGY EVOLUTION FEATURE 4.1 | Dissipated and elastic strain energy calculation
The energy accumulation and dissipation features are essential attributes in revealing the damage and failure evolution of coal and rock. 53,68On the basis of previous research, the entire energy of the coal system is input from the loading system, then transferred into the accumulated energy and the dissipated energy.The accumulated energy is specifically manifested as the elastic strain energy.While the dissipated energy generally generates damage, and plastic deformation, which is accompanied by energy radiation. 47,64n general, the energy calculation during the loading-unloading period, the loading system is assumed as closed, and no heat exchange occurred with the outside system.On this basis, the energy transformation accords with the following equation 47,55,69 : | 4097 where U 0 is the entire input energy obtained by coal specimens, U e represents the elastic energy accumulated in coal specimens, and U d is the energy dissipated caused by the specimen damage and plastic deformation.
For a single loading-unloading cycle, the schematic diagram of the stress-strain curve is shown in Figure 9.
The area under the loading curve is the strain energy exerted by the axial load during one cycle, S L .The elastic strain energy is represented by the area with a blue color under the unloading curve, S UL .The difference in area between the loading and unloading curves is the dissipation energy, S D .The correlation of them can be expounded by the following equation: where σ is the axial stress, ε represents the corresponding axial strain, u 0 is the strain energy, u e is the elastic strain energy, and u d is the dissipation energy.
(A) (B) Poisson's ratio variation with the loading-unloading circle: (A) value for different specimens and (B) average value.
F I G U R E 8 Volume strain variation with time in air-dried and water-saturated specimens.
F I G U R E 9 Schematic illustration of coal energy calculation under cyclic loading-unloading condition.

| Dissipated and elastic strain energy variation
On the basis of Equation ( 3), the calculated energy dissipation and elastic energy liberation were summarized and shown in Figure 10.The dissipated energy shows a U-shaped feature, during the cyclic loadingunloading process.The maximum value is observed at the first loading-unloading cycle, which is considered caused by the long term loading as the axial stress increases from 0 to 5.5 MPa, more energy input plastic deformation occurred during the first cyclic loading-unloading process than others.However, the minimum value is generally obtained at the second loading-unloading cycle, after that, the dissipated energy has an overall rising trend in the following cyclic loading-unloading process, though inconsistency occurs in some periods of these specimens, as shown in Figure 10A.The second maximum dissipation energy values are generally obtained at the last (13th) loading-unloading cycle.This feature is similar to other types of coal reported before, 69 which indicates greater damage occurred in the first loading-unloading cycle and a relatively gentle increase of the damage and plastic deformation in the following cyclic loadingunloading cycles.Meanwhile, the water-saturated specimen has a greater dissipation energy value than that air-dried specimens, as shown in Figure 10A, they, respectively, have variation ranges of 1107.80-4167.87,933.97-2528.23,and 667.58-2571.05J/m 3 .Conversely, in airdried specimens dissipation energy separately with values of 618.57-3669.99and 521.10-1932.62J/m 3 .This reveals the water-saturated specimens maybe have greater plastic deformation than that of air-dried specimens, 58 in combination with greater axial strain observed during the cyclic loading-unloading process.
The elastic strain energy increases with loadingunloading cycles.Water-saturated specimens also have greater elastic strain energy than that of air-dried specimens, respectively, having ranges of 10717.54-59320.37,9481.68-38690.82,and 9201.12-38636.61J/m 3 .Considering the elastic modulus and the UCS of air-dried specimens are greater than water-saturated specimens during the cyclic loading-unloading process, the reason that greater elastic strain energy in water-saturated specimens is mainly caused by the greater strain value under the same stress increment condition.This may be result in less elastic strain energy storage as per unit strain occurred in water-saturated specimens, which may beneficial in reducing the dynamic failure risk under the mining disturbance.

| Damage evolution character
Energy dissipation is considered as a damage evolutionrelated parameter, 55,68,69 since the dissipated energy was consumed by internal damage and plastic deformation of coal specimens.On the basis of the previous research, 47 the cumulative dissipated energy can be normalized to define and characterize the damage coefficient via formulas as follows: loading-unloading cycle of i, and D d is the damage coefficient in the ith cycle.The damage coefficient was calculated and assumed that the damage is 1.0 as the loading-unloading process finished.It is found that airdried specimens have a greater average damage value than that of water-saturated specimens during the whole loading process, as shown in Figure 11.Meanwhile, the water-saturated specimen has a wider damage coefficient range during the whole cyclic loading-unloading process, with a minimum value of 0.13 after the first cyclic loading-unloading, while it is 0.24 in the air-dried specimens.
Mathematically, the damage coefficient in watersaturated specimens has greater increments than that of air-dried specimens during the cyclic loadingunloading process.This indicates a lower percentage of damage occurred after the first cycle of loading-unloading in the air-dried specimens, while a greater proportion of damage is generated in the water-saturated specimens.This indicates that water injection before the mining activities may be beneficial in reducing the damage of coal mass at the beginning of the cyclic loading-unloading disturbance, as shown in Figure 11, the damage coefficient at the first cyclic loading-unloading cycle in watersaturated specimens is lower than that in air-dried specimens.
Meanwhile, the damage coefficient is found positively correlates with the loading-unloading cycles, and a new exponential function is proposed in describing the correlation, namely, where α, β, and C are the specimen-related variables, determined by the internal structure of a specific specimen, which can be acquired via regression analysis.The regression result is shown in Figure 12, and the parameters of α, β, and C are summarized in Table 3. Apparently, Equation ( 5) behaves a good correlation with the experimental data in Figure 12, and the correlation coefficients were greater than 0.995 in experimental specimens.

| Elastic energy index variation
Water injection is considered an effective approach in reducing the coal burst proneness.Thus, more details should be investigated to reveal the weak mechanism in the burst proneness of coal under the cyclic loading-unloading condition.Herein, the elastic energy index was chosen as the evaluating parameter, it was generally calculated by the following formula 5,70 : where W ET is the elastic energy index, U e and U d share the same meaning as been introduced before.Equation ( 6) is available for a single loading-unloading condition.Meanwhile, the elastic energy index obtained by Equation ( 6) is normally calculated as the coal specimen being loaded to the 0.70-0.90times of the UCS, then unloaded to 0.01-0.05times of UCS, under the uniaxial compressive condition.However, the elastic energy index calculation under multicyclic loading-unloading conditions has still not been determined.

Considering the energy dissipation exists in per cyclic
F G U R E 11 Damage evolution with the cyclic loading-unloading cycle.
Comparison of damage coefficient with cyclic loading-unloading cycles.
loading-unloading cycles before the 0.70-0.90times of the UCS, thus, a cumulative elastic energy index-based calculated approach was promoted and verified applicable, with the equation form as 70 where n represent the amount of cyclic loading-unloading counts, W ′ Ei is the elastic energy index in a single loading-unloading cycle, and W ′ ET represents the elastic energy index for a specimen.Thus, for a series of specimens, the average elastic energy index can be calculated as where N is the number of specimens used for one test.The calculated elastic energy index is summarized in Table 4.
Burst proneness of coal specimens experienced an evident reduction as been affected by water under the cyclic loading-unloading.The average elastic energy index in airdried specimens is 17.93, approximately 21.64% greater than that of water-saturated specimens, 14.05.
On the basis of the character of dissipated and elastic strain energy revealed in Section 4.2, though the density of elastic energy and dissipated energy in water-saturated specimens are greater than that in the air-dried specimens, the obvious increment of dissipated energy increases the proportion of total energy input from the outside and reduces the burst proneness of water-saturated specimens.

| AE statistics
AE in coal and rock materials usually occurs due to crack generation, propagation, and coalescence. 33,53,68As part of energy dissipation, the AE signal is also deemed as a piece of significant precursor information in coal failure detection and prevention.To obtain the correlation between energy dissipation and failure of coal specimens, the AE feature during the cyclic loadingunloading process is summarized and analyzed.The cumulative AE counts and energy of air-dried and watersaturated specimens monitored during different loading conditions are summarized in Table 5.
Obviously, the cumulative AE counts and energy decrease as the specimen is exposed to the cyclic loading-unloading.Meanwhile, the water-saturated specimen has a greater reduction than air-dried specimens, under the cyclic loading-unloading condition, respectively, with the average cumulative AE energy, respectively, with values of 2.94 × 10 5 , 1.51 × 10 5 , and 0.62 × 10 5 , and the corresponding cumulative AE energy separately are 5.99 × 10 5 , 2.89 × 10 5 , 1.23 × 10 5 aJ.The reduction of AE activities as specimens are exposed to the loading-unloading conditions may be caused by the internal structure adjustment and readjustment of specimens, since it reduces the number of cracks generated during the loading-loading process.The reduced amount of AE activity in the water-saturated specimen under cyclic loading-unloading conditions also indicates that a greater amount of dissipated energy is used to overcome the crack generation and expansion in air-dried specimens than in the water-saturated specimen.

| Variation of AE feature and energy dissipated
To investigate the correlation of dissipated energy and AE feature, the dissipated energy and the AE feature during the loading-unloading process are analyzed, and a typical comparison of air-dried and water-saturated specimens is shown in Figure 13.
Accords with that reported by previous research. 33,55The Kaiser effect exists during the cyclic loading-unloading process in both air-dried and water-saturated specimens.The detected AE count has a significant increment as the axial loading strength is greater than that in previous loading-unloading cycles.Meanwhile, the AE activities are more pronounced in the beginning of several cycles, and this phenomenon is more significant in air-dried specimens, as shown in Figure 13.This also manifests the conclusion that more damage and plastic deformation occurred in air-dried specimens at the first cyclic loading-unloading process obtained by the damage coefficient.The AE activity is observed to have a greater evident variation correlation with the dissipated energy in air-dried specimens, the detected AE counts and dissipated energy increment behave in synchronicity.However, this phenomenon is not significantly observed in water-saturated specimens, since no greater sudden AE change is monitored, accordingly, the energy dissipation density curve increases more smoothly during the cyclic loading-unloading process.
2][73] To distinguish the energy dissipation feature of AE among airdried specimens and water-saturated specimens, time-series correlation fractal dimension was calculated, based on the Grassberger-Procaccia theory. 72,73On this basis, the time interval and phase space were, respectively, chosen as 3 s and 7, the variation of fractal dimension with the phase space in specimen 1-1 is summarized in Figure 14, with the k value of 0.1, 0.2, 0.3, …, 1.2, after multiple attempts.The calculated fractal dimensions are summarized in Table 6.
Here, to ensure the consistency of calculation parameters and the comparability of the calculated results, namely, the fractal dimension, the same phase space was chosen for specimens with different loading conditions.
The convention uniaxial compressive loading specimen has a greater fractal dimension value, with an average of 0.56.For specimens under the cyclic loading-unloading condition, the fractal dimension value reduces sharply to 0.12 in air-dried specimens and 0.13 in water-saturated specimens.
Generally, the lower fractal dimension value indicates a greater AE concentration during the loading process. 72reater fractal dimension value represents a more concentration of dissipated AE energy near the peak axial stress.This accords with the AE features exhibited in Figure 13.Meanwhile, the conventional uniaxial compressive loaded air-dried specimens with greater fractal dimension, it maybe correlate with the short loading time since most of them are loaded to failure  within 700 s, which is far lower than the specimens under the cyclic loading-unloading condition, approximately 9000-10,000 s.Which contributes to a relatively less AE energy concentration during the loading process.

| Correlation of AE energy and dissipated energy density
AE is a significant precursor information for the dynamic failure of coal and rock, understanding the correlation between AE energy and dissipated energy is vital for the damage evolution monitoring and prediction under cyclic loading-unloading conditions.Thus, the correlation between AE energy and dissipated energy was investigated in this study.The percentage variation of dissipated energy during various loading-unloading cycles is summarized in Figure 15.
Obviously, the percentage of AE energy released during various loading-unloading cycles behaves in a similar changing character with that of dissipated energy, as shown in Figure 10A.This indicates a definite relationship exists between the AE energy and the plastic property.On the basis of the experiment data calculation the cumulative AE energy and dissipated energy were obtained and analyzed.It is found the cumulative AE energy and dissipated energy density are linearly correlated, this correlation can be described by a linear equation where U d is the cumulative dissipated energy during the cyclic loading-unloading process; U AE represents the cumulative AE energy released during the cyclic loading-unloading process; a, b are constant obtained by regression analysis.The fitting curves of air-dried and water-saturated specimens are shown in Figure 16.Obviously, the regression analysis slope of air-dried specimens, respectively, is 6.28 and 3.88, which are greater than that of water-saturated specimens, 0.78, 2.57, and 1.10.This indicates a greater cumulative AE energy increment with the increasing dissipated energy in air-dried specimens.It is also consistent with the cumulative AE counts and dissipated energy variation feature observed in this study.
This correlation indicates a possibility of revealing damage evolution and the energy dissipation of coal via cumulative AE energy, under the cyclic loadingunloading condition, this also indicates the applicability in detecting the dynamic failure of coal using the microseismic energy.In this study, combined with the conventional uniaxial compressive loading and uniaxial cyclic loading-unloading, AE measurement, and theoretical analysis, the of water on the damage and dissipation feature was experimentally investigated.On this basis, the main conclusion is summarized below.
a.The softened effect of water on the mechanical properties of coal is still remarkable during the cyclic loading-unloading process, greater axial strain and Poisson's ratio, and lower axial strength are observed in water-saturated specimens.b.Water benefits in generating more energy dissipation, reducing the proportion of elastic energy that inputs into the coal specimen, and thus lowering the burst proneness of water-saturated coal specimens.c.Air-dried specimens generate greater damage than water-saturated specimens at the first cyclic loading-unloading cycles, while lower damage increments are observed in the following cycles.d.The elastic modulus of air-dried specimens is greater than that of water-saturated specimens.The elastic modulus of coal has a logarithmic correlation with the cyclic loading-unloading cycles.e.Energy dissipation positively correlates with the energy released by AE activities, the air-dried specimen has greater energy concentration around the peak axial stress, and the cumulative AE energy and cumulative energy dissipation density are linearly correlated, during the cyclic loading-unloading process.

F I G U R E 1
Cylindrical specimens used in this study.FI G U R E 2 Water content variation with immersion time.

F I G U R E 3
Cyclic loading-unloading process during this testament.UCS, uniaxial compressive strength.F I G U R E 4 Layout of loading frame and monitoring system used in this research.AE, acoustic emission.T A B L E 1 Uniaxial compressive strength and axial strain obtained in this experiment.

4 )
where u d is the amount of dissipation energy, u i d represents the energy dissipated in the cyclic(A) (B) I G U R E 10Energy variation feature of coal specimens during the cyclic loading-unloading process: (A) dissipated energy and (B) elastic strain energy.

F
I G U R E 13 AE feature and the energy dissipation of coal specimens: (A) air-dried specimens and (B) water-saturated specimens.AE, acoustic emission.

F 13 F
I G U R E 14 Fractal dimension variation with the phase dimension.T A B L E 6 Fractal dimension of acoustic emission in specimens under different loading conditions.I G U R E 15 Acoustic emission energy dissipated during different loading-unloading cycles.
F I G U R E 6 Elastic modulus variation of specimens under the unloading condition during the cyclic loading-unloading process: (A) value for different specimens and (B) average value.SONG ET AL.