Effect of Chemical Corrosion on Rock Fracture Behavior in Coastal Deep Mines: Insights from Mechanical and Acoustic Characteristics

Abstract

The demand for critical minerals has increased extraction activities in coastal deep mines where challenges such as high stresses, chemical corrosion, and mining disturbance impacts are present. This study investigated the effects of chemical corrosion and confining pressure on the mechanical and fracture behaviors of granite specimens, which are crucial for ensuring the stability of surrounding rock in coastal deep mines. Triaxial compression tests were conducted on uncorroded specimens and corroded specimens immersed in acid and alkali solutions under varying confining pressures, with real-time acoustic emission (AE) monitoring. Based on the test results, the strength and deformation properties, progressive fracture, and failure processes, as well as the AE response characteristics of the specimens under chemical corrosion and confining pressure were analyzed. Additionally, the influence of confining pressure on the chemical damage and brittle–ductile transition behavior of specimens was discussed, and the mechanism of chemical corrosion on the physical and mechanical behavior of specimens was revealed based on mineralogical analysis. These findings underscore the importance of understanding the interplay between chemical corrosion, confining pressure, and rock fracture behavior in coastal deep mining and contribute to evaluating the stability of underground surrounding rocks under corrosion environments.

1. Introduction

Extracting deep-sea or coastal mineral resources is essential to meeting the increasing demand for critical minerals in various industrial applications [1,2]. Moreover, developing these seabed resources can reduce dependence on limited terrestrial mineral resources and reduce the occupation of urban space [3]. Currently, due to the many technological, economic, and environmental obstacles of deep-sea mining, full-scale mining activities are carried out mainly in coastal areas [4,5]. However, coastal deep mining also faces many challenges, such as high stresses, high pore-water pressure, unusual temperatures, and the environmental impacts of artificial disturbances [6,7,8]. In addition, seawater intrusion and the geological complexity of seabed rock formations complicate extraction efforts, affecting the safety of mining operations and increasing the risk of mining equipment damage [9,10].

Granite is a typical geological formation in underground mining operations, ranging from host rock in mineral deposits to surrounding rock mass in tunneling and excavation activities. Although granite is a low-porosity crystalline rock with excellent durability, it is still susceptible to chemical corrosion when exposed to specific environmental conditions (e.g., coastal environments) [11,12]. Granite rock masses in coastal environments are often exposed to weathering processes that determine heavy chemical, physical, and mechanical alterations [13,14,15]. Chemical corrosion in underground mines arises from the interaction between granite surfaces and a myriad of corrosive agents in the surrounding environment. These agents may include acidic mine waters, sulfur-bearing compounds, metal-rich solutions, and microbial activity [16,17,18]. Over time, prolonged exposure to such corrosive substances can initiate chemical reactions that compromise the integrity of the granite matrix, leading to alterations in its physical and mechanical properties [19,20]. Such deterioration may increase the susceptibility of surrounding rock mass to fracturing, spalling, and disintegration, exacerbating safety risks and operational challenges in mine development and extraction activities [21,22]. For example, the Sanshandao gold mine, located at the northwestern edge of the Jiaodong Peninsula, North China Craton, is a typical coastal deep mine (Figure 1a). The mining depth has exceeded 1000 m, and the gold deposit in the north of the mining area is covered by seawater [23]. Three control faults (F1, F2, and F3) and some typical secondary joints are in the mining area [24]. Among them, F1 is the ore-controlling fault in a waterproof structure. F2 is a small-scale tensile fault that extends to the Bohai Sea in the northern direction and has a strong hydraulic conductivity. F3 is a tensile-shear fault that extends to the Bohai Sea in the NW direction and has experienced seawater intrusion. Frequent water inflow and significant chemical corrosion of the surrounding rock in some underground roadways pose considerable challenges to mine safety (Figure 1b) [25,26]. Therefore, it is crucial to research the fracture behaviors of the roadway surrounding rock under the influence of chemical corrosion and triaxial pressure to safeguard the stability of coastal deep mines.

Given the adverse effects of chemical corrosion on the physical and mechanical properties of granite, many scholars have researched this in laboratory tests. Shang et al. [27] studied the physicochemical interactions between different chemical solutions and granite fractures, finding that acidic solutions significantly influenced fracture shear properties, followed by alkaline solutions. Liu et al. [28] believe that different chemical solutions penetrate the granite interior through various micropores and fissures, enhancing chemical interactions through the dissolution and precipitation of minerals and destroying the mineral grains and the microstructures. Li et al. [29] investigated the impact of chemical corrosion, flaw distribution, corrosion time, and confining pressure on triaxial compression tests of four types of granite specimens. Miao et al. [30] conducted uniaxial and triaxial compression tests, splitting tests, and microscopic analyses on granite specimens subjected to acidic solution corrosion, revealing insights into strength loss, deformation behavior, mineral composition changes, and damage mechanisms. Moreover, Chen et al. [31] studied the degradation of the tunnel surrounding rock under flowing groundwater scouring with different pH and flow rates. Mei et al. [32] reveal how chemical corrosion in granite fractures leads to transitional behaviors characterized by fast ruptures followed by slow ruptures. The above research has played a positive role in analyzing the mechanical behavior and failure mode of rocks under chemical corrosion environments.

This paper aims to investigate the effects of different chemical solutions on the mechanical properties and fracture behavior of granite to reveal the stability of underground surrounding rocks in coastal deep mines under chemical corrosion environments. Firstly, the granite specimens were subjected to prolonged immersion in chemical solutions, including acid and alkali solutions. Then, with the uncorroded granite specimens as the control group, triaxial compression tests were conducted on the corroded and uncorroded specimens under five different levels of confining pressure, while real-time acoustic emission (AE) signals were monitored. Based on the test results, the strength and deformation properties, progressive fracture, and failure processes as well as the AE response characteristics of the specimens under chemical corrosion and confining pressure were analyzed. Finally, the influence of confining pressure on the chemical damage and brittle–ductile transition behavior of granite specimens was discussed.

2. Materials and Methods

2.1. Specimen Preparation

The granite specimens extracted from an intact rock block were processed into cylinders with a height of 100 mm and a diameter of 50 mm. Subsequently, specimens with similar densities and longitudinal wave velocities were selected to minimize potential errors. The XRD spectrum and main mineral components of the granite specimens are shown in Figure 2. The granite specimen is primarily composed of quartz (17.0%), anorthite (45.0%), albite (25.7%), microcline (10.0%), and biotite (2.3%). A common way to shorten the test period is by elevating the ion concentration of chemical solutions [33,34]. Therefore, the granite specimens were immersed in the prepared chemical solutions, which included a 10% HCl solution and a 10% NaOH solution. By monitoring the changes in the pH value of the solutions, it was determined that the granite had completely reacted with the solutions, and the immersion period was set to 48 days.

2.2. Test Methods

The soaked granite specimens were dried for 24 h to mitigate the influence of moisture content on the mechanical test results. Triaxial compression tests were conducted using the MTS 815 rock mechanics testing system, supplemented by the AE technique to monitor the signals generated by microfractures inside the specimen during loading, as shown in Figure 3. The loading process for the triaxial compression test was as follows:

(1)

The specimen underwent preloading at a rate of 0.05 kN/s, with pre-pressure set at 0.5 MPa to ensure complete contact between the testing machine’s loading end and the specimen.

(2)

Following preloading, stress control was used to apply confining pressure at a loading rate of 0.1 MPa/s. Axial pressure was imposed at a loading rate of 0.25 MPa/s upon reaching the preset confining pressure.

(3)

Once the yield point was reached, the control mode transitioned from stress control to radial displacement control, with a control rate of 0.015 mm/min.

(4)

The loading process was halted upon specimen damage to ensure compliance with quasi-static loading conditions.

The AE instrument employed a PCI-2 AE monitoring system produced by American Physical Acoustics Corporation (PAC), which is composed of a cable, an amplifier, AE sensors, and a data-acquisition control system. Four AE sensors are arranged in the upper and lower parts of the granite specimen to receive the signals, which are then pre-amplified, discharged, and denoised by an amplifier to form the AE parameters (e.g., ringing counts, amplitude, frequency, etc.). The sampling interval of the AE signal during loading was set to 0.1 s, and the sampling threshold was set to 40 dB, meeting the requirements for filtering out most of the background noise and monitoring microcracking activity.

3. Mechanical Parameter Analysis

3.1. Triaxial Compressive Strength and Elastic Modulus

The stress-strain curves of granite specimens under different confining pressures obtained from the experiment are shown in Figure 4, and the plotted axial stress-volumetric strain curves are shown in Figure 5. The axial stress and lateral stress are represented by σ₁ and σ₃, respectively. The volumetric strain ε_V is calculated from axial strain ε_ax and lateral strain ε_lat based on the following equation:

$${\epsilon }_{\mathrm{V}}\approx {\epsilon }_{\mathrm{ax}}+2{\epsilon }_{\mathrm{lat}}$$

(1)

The variations in triaxial compressive strength (TCS) and elastic modulus E of granite specimens with confining pressure are shown in Figure 6. Under unconfined conditions, chemical corrosion significantly weakens the compressive strength of granite specimens, but this weakening effect is suppressed under confining pressure. Confining pressure shows the most pronounced strength enhancement in acid-corroded specimens followed by alkali-corroded specimens. In addition, significant strain-softening behavior is observed after reaching the TCS for uncorroded and alkali-corroded specimens. The stress-strain curves of the acid-corroded specimens show a sudden stress drop at the TCS, exhibiting sudden brittle failure characteristics.

Confining pressure has a strengthening effect on the capability of anti-deformation of granite specimens. As the confining pressure increases, the TCS increases approximately linearly. Linear fitting was performed on the experimental data to obtain the fitting relationship between the TCS and confining pressure as follows:

$$TC{S}_{\mathrm{N}}=244.30+10.74{\sigma }_3, R^2=0.990$$

(2)

$$TC{S}_{\mathrm{CH}}=221.85+12.02{\sigma }_3, R^2=0.981$$

(3)

$$TC{S}_{\mathrm{CN}}=229.84+10.94{\sigma }_3, R^2=0.988$$

(4)

where $TC{S}_{\mathrm{N}}$, $TC{S}_{\mathrm{CH}}$, and $TC{S}_{\mathrm{CN}}$ are the TCSs of the uncorroded, acid-corroded, and alkali-corroded specimens, respectively, MPa.

Under unconfined conditions, the elastic modulus of acid-corroded and alkali-corroded specimens decreases to varying degrees. The order of elastic modulus from highest to lowest is uncorroded > alkali-corroded > acid-corroded specimens. This indicates that chemical corrosion has a weakening effect on the elasticity of granite specimens. Under the application of confining pressure, the elastic moduli of the acid-corroded and alkali-corroded specimens approximate that of the uncorroded specimens. This suggests that confining pressure can suppress the weakening effect of chemical corrosion on elasticity, enhancing its ability to undergo elastic deformation. The elastic modulus of the uncorroded specimens shows an approximately linear increase with confining pressure, while the acid-corroded and alkali-corroded specimens exhibit a rapid increase within the range of 0–5 MPa confining pressure, followed by an approximately linear growth with increasing confining pressure. Nonlinear fitting was performed on the experimental data to obtain the fitting relationship between the elastic modulus and confining pressure as follows:

$$E_{\mathrm{N}}=58.86+0.45{{\sigma }_3}^{0.87}, R^2=0.997$$

(5)

$$E_{\mathrm{CH}}=54.03+4.31{{\sigma }_3}^{0.28}, R^2=0.969$$

(6)

$$E_{\mathrm{CN}}=56.84+2.10{{\sigma }_3}^{0.44}, R^2=0.895$$

(7)

where $E_{\mathrm{N}}$, $E_{\mathrm{CH}}$, and $E_{\mathrm{CN}}$ are the elastic modulus of the uncorroded, acid-corroded, and alkali-corroded specimens, respectively, GPa.

3.2. Cohesion and Internal Friction Angle

Cohesion c and internal friction angle φ were determined based on the linear fitting equations of the strength envelope of the Mohr–Coulomb criterion; the calculated results are presented in

Table 1. Triaxial compression test results for granite specimens.

Type	σ3 (MPa)	TCS (MPa)	σcd (MPa)	σci (MPa)	E (GPa)	c (MPa)	φ (°)
Uncorroded	0	235.48	181.43	104.41	58.86	37.27	56.06
	5	301.19	207.40	120.73	60.80
	10	363.10	233.95	133.55	63.62
	15	408.29	257.71	147.14	62.08
	20	450.44	282.64	156.96	64.99
Acid-corroded	0	205.85	157.18	90.82	54.03	32.00	57.82
	5	293.62	191.08	102.33	60.94
	10	354.48	225.13	120.52	61.86
	15	406.28	256.86	138.44	63.34
	20	450.02	279.09	152.66	63.98
Alkali-corroded	0	218.42	168.03	96.74	56.84	34.75	56.35
	5	292.35	197.61	109.19	61.54
	10	349.07	224.56	123.18	62.13
	15	396.51	256.63	140.75	63.33
	20	439.81	267.33	149.80	65.08

. The correlation coefficients of the linear fitting equations are all above 0.98, indicating that within the range of 0–20 MPa confining pressure, the Mohr–Coulomb criterion can effectively characterize the strength characteristics of granite specimens before and after corrosion.

Under the influence of chemical corrosion, the cohesion c of the granite specimens decreases and the internal friction angle φ increases. Notably, the cohesion c exhibits a higher sensitivity to chemical corrosion. Compared to alkali corrosion, acid corrosion induces more significant changes in the shear strength parameters of the granite specimens. Analysis of the mechanisms reveals that chemical immersion dissolves clay minerals between rock particles, weakening the interparticle bonding and, consequently, reducing the cohesion c [35]. Under compressive loading, the pores formed by the dissolution of the cementitious material are compacted and closed, causing direct contact between mineral particles and strengthening the internal friction angle φ. Due to chemical corrosion, which reduces the lubricating effect between mineral particles and increases their frictional resistance, the acid-corroded specimens are more prone to sudden failure after reaching compressive strength.

3.3. Stress Threshold for Cracking and Damage

As shown in Figure 7, the fracture of rocks usually has progressivity, which can be characterized by the initiation, propagation, and coalescence of cracks, corresponding to different fracture zones in the surrounding rocks of the roadway [36,37]. In rock engineering, the most important concerns are crack initiation stress σ_ci, crack damage stress σ_cd, and peak stress σ_c. Among them, σ_ci and σ_cd are considered as the lower and upper limits of the long-term strength of the rock mass, respectively. This is of great significance for the stability evaluation of engineering rock masses. In general, the method for determining σ_cd is relatively objective. The widely accepted view is to consider the axial stress corresponding to the maximum volumetric strain in the axial stress-volumetric strain curve as σ_cd [38,39]. However, the determination of σ_ci is challenging and highly subjective, mainly including the volumetric strain (VS) method, lateral strain (LS) method, lateral strain response (LSR) method, and AE method [40,41,42,43]. Among them, the LSR method is relatively more objective because the maximum value of the lateral strain difference in the axial stress-lateral strain difference curve is unique. Although the physical significance of this method is yet to be determined, it provides a convenient and objective method of obtaining σ_ci. Therefore, this article obtained the σ_ci and σ_cd of granite specimens based on the LSR method and axial stress-volumetric strain curve, respectively. The calculation results are shown in Table 1.

The σ_ci and σ_cd of granite specimens are approximately linearly positively correlated with confining pressure, as shown in Figure 8. Acid and alkali corrosion can reduce the σ_ci and σ_cd, making the granite specimens more prone to crack growth or propagation. For example, the σ_ci of the acid-corroded specimens and alkali-corroded specimens decreased by about 13% and 7%, respectively, compared to the uncorroded specimens under the unconfined condition. The σ_cd decreases in about the same proportion as the σ_ci. However, this variability was progressively minimized as the confining pressure increased, suggesting that the confining pressure effect can mitigate the deterioration of the stress threshold by chemical corrosion to some extent.

By analyzing the relationship between stress thresholds and peak stresses, it is easier to use compressive strength to evaluate the spalling and destabilization strength of rocks. Figure 9 shows the relationship between the σ_ci, σ_cd, and TCS. Overall, the σ_ci and σ_cd occur at approximately 0.36 and 0.65 of the TCS, respectively. More precisely, the σ_ci and σ_cd are about 0.33~0.45 and 0.60~0.78 of the TCS, respectively, as shown in Figure 10. It is worth noting that σ_ci/TCS and σ_cd/TCS first decrease rapidly with increasing confining pressure and then remain approximately at a constant value. The acid and alkali corrosion lead to specimens with lower σ_ci/TCS under the same confining pressure but have less impact on σ_cd/TCS.

3.4. Fracture Mode Analysis

The macro fracture morphology of granite specimens under confining pressure is shown in

Table 2. Macroscopic fracture morphology of granite specimens.

Type	Confining Pressure σ3 (MPa)
	0	5	10	15	20
Uncorroded
Acid-corroded
Alkali-corroded

. Under unconfined compression conditions, the failure mode of the specimen is splitting parallel to the direction of maximum stress, that is, tensile failure. Under low confining pressure, the specimen mainly exhibits shear failure mode, with local tensile failure and easy formation of a fracture surface perpendicular to the direction of maximum stress. With the increase in confining pressure, shear failure gradually dominates and obvious macro fracture surfaces appear, exhibiting strong brittle fracture characteristics. In addition, there is no significant difference in the failure mode between uncorroded and corroded specimens, indicating that under triaxial compression, the macroscopic failure mode of granite is more dependent on confining pressure and less affected by chemical corrosion.

4. Analysis of Acoustic Characteristics

4.1. Time Domain Characteristics of AE Counts

The AE signals released during the compression of granite specimens are closely associated with the formation and propagation of microcracks [44,45,46,47,48]. The characteristics of temporal AE sequences for granite specimens under different confining pressures are depicted in Figure 11, Figure 12 and Figure 13.

Combined with the stress-strain curves, it can be observed that the AE count characteristics of granite specimens under different immersion conditions are generally similar during the deformation and failure processes. Specifically, the AE signals are relatively calm during the early stages of loading in the microcrack closure and elastic deformation. As the load increases, new cracks initiate and propagate, causing the AE signals to become more active, leading to a sharp rise in AE counts before failure. After reaching compressive strength, the frictional interaction between particles on the shear fracture surface and the tensile fracture surface continues to release AE signals, keeping the AE counts fluctuating within a relatively low range. Under confining pressure, uncorroded specimens exhibit a brittle–ductile transition characteristic after reaching compressive strength, without showing a significant stress drop. The cumulative AE counts of the acid-corroded and alkali-corroded specimens are fewer, indicating a higher susceptibility to sudden brittle failure.

4.2. AE Frequency Domain Characteristics

AE frequency domain characteristics can be used to analyze the fracture process and to identify fracture precursor information of rocks [44,45]. The evolution characteristics of the peak frequency distribution of granite specimens under different chemical corrosion conditions are shown in Figure 14, Figure 15 and Figure 16. According to the magnitude of the peak frequency, it is divided into three frequency ranges: high frequency (HF, >256 kHz), medium frequency (MF, 128~256 kHz), and low frequency (LF, <128 kHz). The peak frequency of the uncorroded specimen is mainly concentrated in the MF, and the HF signal gradually increases with the increase in confining pressure. Moreover, a large number of AE frequency signals are generated when a stress drop occurs, with the density of the frequency distribution increasing significantly. Acid- and alkali-corroded specimens have lower peak frequency densities than uncorroded specimens and relatively fewer AE frequency signals before a stress drop occurs. After the first stress drop, a large amount of AE frequency signals are still released, although a high-density frequency distribution is rarely observed. In summary, when the rock enters the unstable fracture stage, both uncorroded and corroded specimens show a large number of AE frequency signals, but corroded specimens may not necessarily exhibit high-density frequency distributions. Therefore, it is possible to determine whether the rock has reached the critical fracture state by monitoring the changes in the number of AE peak frequency signals.

4.3. AE b-Value

The AE events accompanying the failure of rocks can be approximated as a small-scale seismic activity, so the b-value in seismology can be introduced to study AE characteristic parameters. It is generally believed that the change in the b-value reflects the variations in microcrack propagation scales within rocks [46]. The b-value increases when the proportion of small AE events (small-scale fractures) increases and decreases when the proportion of large AE events (large-scale fractures) increases [47,48]. A significant change in fracture scales when the rock is likely to undergo damage results in a rapid increase or decrease in the b-value; therefore, the sudden change in AE b-values can serve as precursor information for rock failure.

In the b-value calculation process, the maximum likelihood method is generally considered to be simpler and less error-prone than the least squares method. The b-value based on the maximum likelihood estimation method is given by [46]

$$b=\frac{\lg e}{\overline{M}-M_{\min }}$$

(8)

where $\overline{M}$ is the average magnitude; $M_{\min }$ is the minimum magnitude; and e is a natural constant. The magnitude M can be calculated from the AE amplitude A_dB, and M = A_dB/20.

The AE b-values under different loading times were calculated using a set of 100 sample data, as shown in Figure 17 and Figure 18. Take the evolution of b-values under unconfined compression and 20 MPa confining pressure as examples for discussion. The b-value exhibits an evolutionary characteristic of stable variation followed by rapid fluctuation during loading, especially showing a significant sudden drop when a stress drop occurs. Compared with the unconfined compression condition, the b-value fluctuation under confining pressure is minor, and the rock fracture process is more concentrated.

To further discuss the effects of confining pressure and chemical corrosion on the fracture behavior of granite specimens, the variation patterns of the average value and coefficient of variation (CV) of b-values with confining pressure were analyzed, as shown in Figure 19. A higher CV indicates that the data set is more discrete, implying that the b-value fluctuates more and the rock is more prone to sudden failure. On the contrary, the smaller the CV, the more concentrated the data in this group, which means that the fluctuation of the b-value is smaller, and the rock is more prone to progressive failure. The CV is calculated as follows:

$$\mathrm{CV}=\frac{S}{\overline{A}}=\frac{\sqrt{\frac{1}{N}{\displaystyle {\sum }_1^N{(x_i-\overline{A})}^2}}}{\frac{1}{N}{\displaystyle {\sum }_1^N{x}_i}}$$

(9)

where $S$ is the standard deviation of data; $\overline{A}$ is the average value of data; N is the amount of data; $x_i$ is the i-th data.

Figure 19a shows that the average b-value first increases and then decreases with increasing confining pressure. This indicates that small-scale fractures are more predominant at low confining pressures, while large-scale fractures are more predominant at high confining pressures. Figure 19b shows that the CV of the b-value decreases rapidly under low confining pressure and remains almost flat under high confining pressure. Compared with the unconfined compression condition, the effect of confining pressure weakens the fluctuation of the b-value, and the granite specimen gradually changes from sudden failure to progressive fracture. These phenomena are related to the brittle–ductile transition behavior of rocks under confining pressure. Notably, the CV of the b-value of the acid-corroded specimens is much larger under the same confining pressure, suggesting that many small-scale fractures were generated during loading, followed by a rapid convergence to form large-scale fractures. The analysis suggests that the acid solution corrodes the bonds between the mineral particles, weakening cementation and making the rock more prone to sudden failure.

5. Discussion

5.1. Effect of Confining Pressure on Chemical Damage

Compressive strength can be used to reflect the ability of rock materials to resist external deformation and fracture under compressive loads and is one of the main mechanical indicators. Therefore, chemical damage variables can be defined by changes in compressive strength to reflect the level of strength degradation of chemically damaged rocks. The definition of chemical damage variables is as follows:

$$D=1-\frac{TC{S}_{\mathrm{cor}}}{TC{S}_{\mathrm{uncor}}}$$

(10)

where D is the chemical damage variable; and $TC{S}_{\mathrm{uncor}}$ and $TC{S}_{\mathrm{cor}}$ represent the compressive strengths of uncorroded and corroded specimens under the same confining pressure, respectively.

The variation in chemical damage variables of chemically corroded granite specimens with confining pressure is shown in Figure 20. It can be observed that the chemical damage variable decreases nonlinearly with the increase in confining pressure, and its quantitative relationship can be established using nonlinear fitting equations. The fitting equations are as follows:

$$D_{\mathrm{CH}}=0.126-0.039{\sigma }_3^{0.390}，R^2=0.913$$

(11)

$$D_{\mathrm{CN}}=0.072-0.008{\sigma }_3^{0.627}，R^2=0.703$$

(12)

where $D_{\mathrm{CH}}$ and $D_{\mathrm{CN}}$ are the chemical damage variables for acid and alkali corrosion specimens, respectively.

The lower the chemical damage variable, the better the corrosion resistance of the specimen. Therefore, confining pressure can enhance the compressive strength of granite, thereby reducing the deterioration of strength caused by chemical corrosion. Under unconfined compression conditions, acid corrosion has a greater degradation effect on the strength of the specimen than alkali corrosion, while under confining pressure, the degradation effect of alkali corrosion becomes more prominent. It should be noted that chemical corrosion has a slight impact on the strength deterioration effect of granite specimens, and even under unconfined compression conditions, the chemical damage variable is still not greater than 0.15. A lower confining pressure can restore the strength of the chemically corroded specimen to near that of the uncorroded specimen.

5.2. Effect of Confining Pressure on Brittle–Ductile Transition

Granite often exhibits brittle failure under low confining pressure, while it may gradually transform into ductile failure under high confining pressure. This brittle–ductile transition behavior can be represented by the ductility factor, the figure of dividing ultimate strain by yield strain, expressed as follows

$$F=\frac{{\epsilon }_{\mathrm{c}}}{{\epsilon }_{\mathrm{cd}}}$$

(13)

where F is the ductility factor; ε_c is the peak strain of the specimen, which is the axial strain corresponding to the TCS in the deviatoric stress-strain curve; and ε_cd is the yield strain of the specimen, which is the axial strain corresponding to the maximum volumetric strain ε_V in the deviatoric stress-volumetric strain curve.

The variation in ductility factors of granite specimens with confining pressure is shown in Figure 21. It can be observed that the ductility factor increases nonlinearly with the increase in confining pressure, and its quantitative relationship can be established through nonlinear fitting equations. The fitting equations are as follows:

$$F_{ \mathrm{N}}=1.294+0.136{\sigma }_3^{0.325}，R^2=0.985$$

(14)

$$F_{ \mathrm{CH}}=1.199+0.230{\sigma }_3^{0.257}，R^2=0.935$$

(15)

$$F_{ \mathrm{CN}}=1.305+0.135{\sigma }_3^{0.383}，R^2=0.986$$

(16)

where $F_{\mathrm{N}}$, $F_{\mathrm{CH}}$, and $F_{\mathrm{CN}}$ are the ductility factors for uncorroded, acid-corroded, and alkali-corroded specimens, respectively.

The lower the ductility factor, the stronger the brittleness of the specimen. As the confining pressure increases, the brittleness of the granite specimen weakens and the ductility increases. Under the same confining pressure, the ductility of alkali-corroded specimens is highest, followed by those of acid-corroded and uncorroded specimens. In addition, the influence of confining pressure on the brittle–ductile transition of granite has a characteristic of being strong first and then weak, reflected in the gradual decrease in the tangent slope of the fitted curve. For example, when the confining pressure increases from 0 MPa to 5 MPa, the ductility factor of the specimen increases significantly, while the growth rate of the ductility factor gradually slows down when it increases from 5 MPa to 20 MPa. As shown in Figure 22, there is a nonlinear positive correlation between the stress threshold and the ductility factor. The strength enhancement and brittle–ductile transition caused by confining pressure are the basic reasons for the elevated stress threshold of the granite specimen.

5.3. Mineralogical Analysis

The XRD spectra of granite specimens under different corrosion conditions are shown in Figure 23. It can be found that the feldspar, biotite, and oxidation products dissolve to varying degrees under prolonged immersion in different solutions. However, quartz is chemically stable and difficult to react with ordinary acidic or alkaline solutions at room temperature. In the acidic solution, the dissolution of cations, such as Na⁺, K⁺, Ca²⁺, and Al³⁺, will destroy the original microstructures inside the granite specimens, leading to changes in physical and mechanical properties. In the alkaline solution, partially dissolved cations will combine with OH^- to form hydroxide precipitates that are slightly soluble in water, such as Al(OH)₃, Fe(OH)₂, Mg(OH)₂, and Ca(OH)₂. These precipitates will fill the microcracks and micropores inside the specimen to some extent, thus mitigating the effects of chemical corrosion.

Table 3. Potential chemical reactions between the main minerals and solutions in granite specimens.

Type	Reactant	Chemical Reaction Formula
Acid-corroded	Quartz	${\mathrm{SiO}}_2{+4\mathrm{H}}^{+}\to {\mathrm{Si}}^{4+}+2{\mathrm{H}}_2\mathrm{O}$
	Albite	${2\mathrm{NaAlSi}}_3{\mathrm{O}}_8{+2\mathrm{H}}^{+}{+9\mathrm{H}}_2\mathrm{O}\to {2\mathrm{Na}}^{+}+{\mathrm{Al}}_2{\mathrm{Si}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4{+4\mathrm{H}}_4{\mathrm{SiO}}_4$
	Microcline	${2\mathrm{KAlSi}}_3{\mathrm{O}}_8{+2\mathrm{H}}^{+}{+9\mathrm{H}}_2\mathrm{O}\to {2\mathrm{K}}^{+}{+\mathrm{Al}}_2{\mathrm{Si}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4{+4\mathrm{H}}_4{\mathrm{SiO}}_4$
	Anorthite	${\mathrm{CaAl}}_2{\mathrm{Si}}_2{\mathrm{O}}_8{+2\mathrm{H}}^{+}{+\mathrm{H}}_2\mathrm{O}\to {\mathrm{Ca}}^{2+}{+\mathrm{Al}}_2{\mathrm{Si}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4$
	Biotite	${\mathrm{KFeMg}}_2{(\mathrm{AlSi}}_3{\mathrm{O}}_{10}{\left)\right(\mathrm{OH})}_2{+10\mathrm{H}}^{+}\to {\mathrm{Al}}^{3+}{+\mathrm{K}}^{+}{+\mathrm{Fe}}^{2+}{+2\mathrm{Mg}}^{2+}{+3\mathrm{H}}_4{\mathrm{SiO}}_4(\mathrm{aq}.)$
	Muscovite	${\mathrm{KAl}}_3{\mathrm{Si}}_3{\mathrm{O}}_{10}{\left(\mathrm{OH}\right)}_2{+10\mathrm{H}}^{+}\to {3\mathrm{Al}}^{3+}{+\mathrm{K}}^{+}{+3\mathrm{H}}_4{\mathrm{SiO}}_4$
	Kaolinite	${\mathrm{Al}}_2{\mathrm{Si}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4{+6\mathrm{H}}^{+}\to 2{\mathrm{Al}}^{3+}{+2\mathrm{H}}_4{\mathrm{SiO}}_4+{\mathrm{H}}_2\mathrm{O}$
Alkali-corroded	Quartz	${\mathrm{SiO}}_2+2{\mathrm{OH}}^{-}\to {\mathrm{SiO}}_3^{2-}+{\mathrm{H}}_2\mathrm{O}$
	Albite	${\mathrm{NaAlSi}}_3{\mathrm{O}}_8{+6\mathrm{OH}}^{-}+2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Na}}^{+}{+3\mathrm{H}}_2{\mathrm{SiO}}_4^{2-}+{\mathrm{Al}\left(\mathrm{OH}\right)}_4^{-}$
	Microcline	${\mathrm{KAlSi}}_3{\mathrm{O}}_8{+6\mathrm{OH}}^{-}+2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{K}}^{+}{+3\mathrm{H}}_2{\mathrm{SiO}}_4^{2-}+{\mathrm{Al}\left(\mathrm{OH}\right)}_4^{-}$
	Anorthite	${\mathrm{CaAl}}_2{\mathrm{Si}}_2{\mathrm{O}}_8{+4\mathrm{OH}}^{-}{+4\mathrm{H}}_2\mathrm{O}\to {\mathrm{Ca}}^{2+}{+2\mathrm{H}}_2{\mathrm{SiO}}_4^{2-}{+2\mathrm{Al}\left(\mathrm{OH}\right)}_4^{-}$
	Biotite	${\mathrm{KFeMg}}_2{(\mathrm{AlSi}}_3{\mathrm{O}}_{10}{\left)\right(\mathrm{OH})}_2{+5\mathrm{OH}}^{-}+4{\mathrm{H}}_2\mathrm{O}\to {\mathrm{K}}^{+}{+3\mathrm{H}}_2{\mathrm{SiO}}_4^{2-}{+\mathrm{Al}\left(\mathrm{OH}\right)}_3\downarrow {+\mathrm{Fe}\left(\mathrm{OH}\right)}_2\downarrow {+2\mathrm{Mg}\left(\mathrm{OH}\right)}_2\downarrow$
	Muscovite	${\mathrm{KAl}}_3{\mathrm{Si}}_3{\mathrm{O}}_{10}{\left(\mathrm{OH}\right)}_2{+8\mathrm{OH}}^{-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{K}}^{+}{+3\mathrm{SiO}}_3^{2-}+{3\mathrm{Al}\left(\mathrm{OH}\right)}_4^{-}$
	Kaolinite	${\mathrm{Al}}_2{\mathrm{Si}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4{+6\mathrm{OH}}^{-}+{\mathrm{H}}_2\mathrm{O}\to 2{\mathrm{Al}\left(\mathrm{OH}\right)}_4^{-}{+2\mathrm{H}}_2{\mathrm{SiO}}_4^{2-}$

lists the potential chemical reactions between the main minerals in granite specimens and different types of solutions [34,35].

6. Conclusions

The main goal of this research was to investigate and characterize the fracture process and its acoustic characteristics of chemically corroded granite specimens under different confining pressures. However, there are still some limitations in the employed methodology. Firstly, this study conducted mechanical tests on granite specimens after chemical corrosion, which cannot reflect the real multi-field coupling environment of underground mines. Secondly, this study artificially increased the concentration of the chemical solutions to shorten the test period, which is much greater than the actual ionic concentration of corrosive mine water. In addition, the effect of sodium chloride solution, one of the main components of seawater, on the mechanical properties of the surrounding rocks in coastal deep mines needs to be further explored in subsequent studies. Nevertheless, this paper points out the importance of understanding the interactions between chemical corrosion, confining pressure, and rock fracture behavior in coastal deep mining environments. The results of the study contributed to evaluating the stability of underground surrounding rocks in coastal deep mines under corrosion environments. The main conclusions are as follows:

(1)

Chemical corrosion significantly weakens the compressive strength of the granite specimens under unconfined conditions, although this weakening effect is mitigated under confining pressure. After reaching the TCS, the uncorroded and alkali-corroded specimens exhibit strain-softening behavior, while the acid-corroded specimens exhibit sudden brittle failure.

(2)

The elastic modulus of the corroded specimens undergoes a varying degree of degradation under unconfined compression, whereas this difference is reduced under confining pressure. Under the effect of chemical corrosion, the cohesion of the granite specimens decreases, the internal friction angle increases, and the cohesion exhibits a higher sensitivity.

(3)

Chemical corrosion degrades the cracking stress threshold to a greater extent and has a lesser effect on the damage stress threshold, making the granite specimens more prone to crack propagation; the confining pressure can mitigate this deterioration to some extent.

(4)

Compared to uncorroded and alkali-corroded specimens, acid-corroded specimens show fewer accumulative AE counts and frequency signals and lower peak frequency densities and are more susceptible to sudden brittle failure. The fluctuations in AE b-values under confining pressure are less pronounced than those under unconfined compression, and the rock fracture process is more concentrated.

(5)

Under unconfined compression, acid corrosion has a more significant degradation effect on the strength of the specimen than alkali corrosion, while the degradation of alkali corrosion becomes more prominent under confining pressure. The influence of confining pressure on the brittle–ductile transition has a characteristic of being strong first and then weak, and the ductility of alkali-corroded specimens is highest, followed by those of acid-corroded and uncorroded specimens.