Research on the Strength Damage and Permeability Characteristics of Cemented Paste Backfill under Chlorine Salt Erosion and Dry–Wet Cycles

Abstract

Cement paste backfill (CPB) suffers serious damage and deterioration under the dual erosion conditions of the dry–wet cycle caused by the high chloride salt concentration in mine water and the fluctuation of mine water level. In order to discuss the mechanical properties and permeability characteristics of CPB under erosion, this study designs an immersion experiment for CPB under chloride salt and dry–wet cycle conditions. Through a uniaxial compressive strength (UCS) test, the change law for the mechanical parameters of the CPB was investigated, the strength constitutive equation of the CPB was constructed and the deterioration process of the CPB was analyzed. The penetration test was used to investigate the diffusion characteristics of the packing under chloride salt and dry–wet cycle conditions. The results showed that the strength and Young’s modulus of the CPB initially increased and then rapidly decreased, with maximum decrease rates of 32.2% and 38.2%, respectively. The CPB structure exhibits an initial undamaged stage, an initial damaged stage, a damaged development stage, a damaged destruction stage and a residual damaged stage. The chloride ion penetration depth gradually increased with the number of dry–wet cycles, with a maximum diffusion depth of 20.5 mm. The maximum apparent diffusion coefficient of chloride ion was 18.99 × 10⁻¹⁰ m²/s, and the maximum concentration was 0.303 mol/L. Under the double erosion conditions of chloride salt and dry–wet cycle, the CPB structure was severely damaged.

1. Introduction

The CPB process allows mining wastes, such as fly ash and coal gangue, to be effectively converted into useful materials for backfilling underground voids in mines [1,2]. This method can not only reduce the environmental pollution caused by the accumulation of solid waste, but also helps to reduce backfilling costs by allowing the partial replacement of cement with waste materials during backfilling. Therefore, the use of CPB, which is a green mining technology, has been increasing in coal mining and goaf treatment [3,4,5]. Therefore, this study mainly investigates the strength reduction pattern and permeability characteristics of the backfill under chloride salt erosion and dry–wet cycles.

However, the high temperature caused by underground coal oxidation and heat release, blasting, deep well mining, stress in overlying strata and erosion of acids, chlorides, sulphates and other chemicals in groundwater lead to the destruction of the CPB structure, thereby affecting the stability of the CPB [6,7]. Among these factors, chloride salt erosion is an important factor; therefore, many researchers have conducted research on chloride erosion in CPB. Florea et al. [8] believed that chloride salts diffuse into the filler and chemically combine with hydration products to form expansive Friedel and Kuzel salts, resulting in a 20% reduction in strength. Mathias et al. [9] carried out an experimental study of the corrosion of cement-based materials caused by chloride salts. The results showed that the chloride salt erosion rate was high in the early stages and decreased in the later stages. Gao Meng et al. [10] soaked CPB in a 10% magnesium chloride and sodium chloride solution for 180 d and found that chloride crystals appeared inside the CPB. They also found that the UCS of the CPB decreased by 45.5% and 24.3% for the magnesium chloride and sodium chloride solutions, respectively. Du et al. [11] conducted an experiment on CPB erosion due to chloride salt solutions with mass fractions of 0.5%, 1.0% and 1.5%. They found that the mass change rate of the CPB showed an overall increasing trend with increasing mass fraction of chloride salts. The UCS first increased and then decreased significantly with a maximum decrease rate of 27.08%. He et al. [12] conducted a chloride salt erosion experiment on CPB mixed with limestone; the results showed that the CPB deteriorated significantly after 150 d and the UCS decreased by 43%.

The above research only makes a detailed analysis of the strength erosion mechanism of the CPB under the condition of a single chloride salt, and the obtained results provide certain guidance for improving the durability of the mine CPB. However, the mine environment of deep mining is more complex and the erosion factor faced by the CPB is not singular, but a combination of factors. In Wali Coal Mine, Beizao Coal Mine and Liangjia Coal Mine in Longkou Coalfield, Shandong, China, the Cl-contents in mine water are 930 mg/L, 3207 mg/L and 2827 mg/L, respectively [13]. The micro-cracks formed in the early stage penetrate into the fill material, and the CPB is eroded by chloride salts. Due to the effect of monsoon rainfall, the mine water level in the goaf will fluctuate up and down. This causes the CPB to experience dry–wet cycles in the goaf, and the CPB is subjected to such dry–wet cycles [14], which can aggravate its expansion, cracking, delamination and loss of strength and adhesion. Under the actions of chlorine erosion and dry–wet cycling, the strength of the backfill gradually decreases. At present, there are few reports on the strength degradation and permeability characteristics of CPB under the action of chloride salt erosion and dry–wet cycling.

Chloride erosion causes a reduction in the strength of the CPB, and the dry–wet cycle increases this degree of damage. At present, most of the research focuses only on chloride erosion or wet–dry cycle conditions, but the backfill material inside the actual mining site has always been under dual erosion conditions of chloride and wet–dry cycle. If the damage pattern and erosion mechanism under dual damage are not clear, it is difficult to ensure the stability of the CPB. In this paper, through the immersion test of the CPB subjected to chloride erosion and dry–wet cycle, the UCS mechanical test and chloride ion penetration test were carried out, the variation law of the mechanical parameters of the CPB was explored, the strength constitutive equation of the CPB was constructed and the damage process of the CPB was analyzed. The penetration test was used to investigate the diffusion characteristics of the packing under chloride salt and dry–wet cycle conditions. Carrying out the research on the CPB with chloride erosion and dry–wet cycle can provide scientific guidance for the implementation of backfill mining under chloride environment and effective maintenance of the stability of the goaf. Through this study, the variation law and erosion mechanism of the strength of CPB under the chloride salt and dry–wet cycle can be clarified. Based on this, a theoretical basis can be provided for the selection of additives, ratio of CPB, determination of concentration and other parameters, and scientific guidance can be provided for the implementation of backfilling mining in the chloride salt environment, effectively maintaining the stability of the goaf.

2. Test Material and Methods

2.1. Test Material

Fly ash and coal gangue are the most commonly used aggregates in coal mine backfilling. Therefore, the fill material was composed of fly ash, coal gangue and cement, the chemical composition of which is shown in

Table 1. Chemical contents and physical properties of backfill material (wt.%).

Element	SiO2	Al2O3	Fe2O3	CaO	MgO	K2O	Others
Cement	21.38	4.23	3.58	66.49	2.50	1.07	0.74
Fly ash	53.94	30.91	1.38	6.53	0.92	2.02	4.29
Coal gangue	59.10	18.90	4.30	2.36	1.41	1.89	12.03

. The coal gangue was obtained from Daizhuang Coal Mine, Shandong Province, China. The cement is 42.5 ordinary Portland cement and is supplied by Shandong Shanshui Cement Group Co., Ltd., (Jinan, China). The particle size diagram of cement and fly ash is shown in Figure 1c. The particle size ratio of the coal gangue is shown in

Table 2. Particle size distribution of coal gangue used.

Size (mm)	+10	10 to 6	6 to 3	3 to 1.5	1.5 to 0
Content (%)	5.18	13.68	22.55	14.97	43.62

.

Due to the differences in coal sources, coal types and extraction methods, the performance of fly ash in China is also very different. According to China’s national standard, fly ash is classified into three grades according to its loss on ignition. The fly ash in this paper, obtained from the Huangdao Power Plant in Qingdao City, was light grey in appearance, with a loss on ignition of less than 8%, Class II fly ash. The apparent density is 1950 kg/m³ and the bulk density is 780 kg/m³, which contains a large amount of SiO₂ and Al₂O₃. The schematic diagram of fly ash is shown in Figure 1, which shows that the shape of the fly ash particles is spherical. These chemical compositions and spherical particles are beneficial to exert the pozzolanic and microaggregate effects of fly ash [15]. The routineity and rationality of the fly ash have been verified.

2.2. Test Plan

The ratio of cement: fly ash: gangue of the prepared test piece was 1:4:6, and the mass concentration was 72% [16,17]. The material was weighed and poured into a mixer. When the slurry was evenly mixed, it was put into a mold with dimensions of 70.7 mm × 70.7 mm × 70.7 mm. After curing for 1 d, the specimen was demolded and placed in a constant temperature and high humidity curing box (95% relative humidity and 20 °C temperature) for 28 day(d). The standard strength test and the chloride ion and dry–wet cycle tests were performed on the 28d cured test block.

In the dry–wet cycle test, a sodium chloride solution with a mass concentration of 10%, prepared with analytically pure sodium chloride, was used, and the test box containing the sodium chloride solution was placed at room temperature (20 ± 5 °C) and immersed for 7d. The indoor drying cycle system was for 7d each for the dry and wet cycles, i.e., two weeks per dry–wet cycle, for eight such cycles (16 weeks). Four samples were taken at the end of each cycle, providing a total of 64 samples. Of the 4 specimens removed each time, 3 were subjected to the uniaxial compression test and the other to the chloride ion penetration test.

2.3. Test Method

The uniaxial compressive strength of all specimens was measured using a Shimadzu AX-G250 (Shimadzu Enterprise Management Co., Ltd., Beijing, China) testing machine.

The chloride ion penetration depth d was measured using the silver nitrate chromogenic method [18,19], and the chloride ion penetration image was processed using digital image processing (DIP) technology. A digital image in a computer consists of a series of pixels arranged in a matrix. In a grayscale image or a binarized image, the grayscale values are 0-255 and 0-1, respectively. The greyscale image of the color rendition surface can eventually become a two-dimensional matrix, and the essence of image processing is the processing of the two-dimensional matrix.

The free chlorine ion concentration at the color change limit was then determined using a Lemag ZDJ-4B automatic potentiometric titrator and the water extraction method. Finally, the apparent diffusion coefficient of the chloride ions in the sample was calculated. The calculation process is in accordance with the following Formulas (1)–(3) [20]:

$$w=\frac{\mathrm{M}·\mathrm{R}·N·V_1}{1000·V_2}$$

(1)

$$c=\frac{1000·w}{\mathrm{M}}$$

(2)

$$D_{app}={\left[\frac{d}{2F_{erfc}\left(1-\frac{c}{c_i}\right)\sqrt{t}}\right]}^2$$

(3)

where w is the mass fraction of chloride ions, M is the amount of chloride ions (35.45), R is the ratio of water to solid during water extraction, N is the concentration of the silver nitrate standard titration solution (mol/L), V₁ is the amount of silver nitrate standard titrant (mL), V₂ is the volume of the sample taken (mL) and c is the chloride ion concentration (mol/L) at the color change boundary. D_app is the apparent diffusion coefficient of chloride ions (m²∙s⁻¹), d is the penetration depth of chloride ions measured by the silver nitrate coloration method (mm), F_erfc is the inverse function of the error function F_erf, c is the color change boundary chloride ion concentration (mol/L), c_i is the chloride ion concentration on the surface of the test piece (mol/L) and t is the chloride ion penetration time (s).

3. Analysis of Test Results

3.1. Mechanical Parameter Analysis of CPB

The strength value and elastic modulus value of each age are compared with the strength value and elastic modulus value of the specimen cured for 28d, and the strength change rate I_σ and elastic modulus change rate I_E are obtained. The calculation formula is shown in Formula (4):

$$\left.\begin{array}{c}{I}_{\sigma }=\frac{{\sigma }_n-{\sigma }_0}{{\sigma }_0}\times 100\%\\ I_E=\frac{E_n-E_0}{E_0}\times 100\%\end{array}\right\}$$

(4)

In the formula, I_σ and I_E are the strength change rate and elastic modulus change rate of the CPB under different cycle ages; σ_n and E_n are the strength value and elastic modulus value of the CPB under different cycle ages; σ₀ and E₀ is the strength value and elastic modulus value of the CPB after curing for 28 d.

Figure 2a shows the changes in UCS and Young’s modulus of the filler; Figure 2b shows the rates of change in UCS and Young’s modulus of the CPB under different chloride salt and dry–wet cycle conditions. The effect on the strength of the CPB is that its UCS does not decrease significantly with the increase in the number of cycles, but shows a tendency to first increase slightly, then decrease and then decrease significantly. After the third cycle, its strength value increased by 15.5% compared to that of the unmachined 28d specimen. After the fourth and fifth cycles, the strength decreased slightly. At this time, the strength of the specimen was still higher than the strength value of the uncured 28d specimen. After the 6th, 7th and 8th cycles, the strength value decreased by 10.6%, 19.1% and 32.2%, respectively. This is much greater than the decrease in strength of the restorative material seen in previous studies [10,11].

The effect on the Young’s modulus of the CPB is that the Young’s modulus value first increases and then decreases as the number of cycles increases. At the early stage of erosion, i.e., after the 1st, 2nd, 3rd, 4th and 5th cycles, the Young’s modulus of the specimens as a whole showed a trend of first slightly increasing and then slowly decreasing, with a maximum increase of 9.5%, which was higher than that of the 28d cured specimens. There is no obvious change in the modulus of elasticity of the specimens, and the specimens still have good plasticity in the early stage of erosion, and the strengths tested in the same period are all greater than 4MPa, which is beneficial to the joint action of the CPB and the surrounding rock to support the roof. At the later stage of erosion, after the 6th, 7th and 8th cycles, the Young’s modulus values decreased by 11.1%, 18.6% and 38.2%, respectively, compared to that of the 28d cured specimen, which showed a significant decrease, and the CPB appeared to be “softened” as the CPB and the surrounding rock jointly supported the roof load. As the modulus of the CPB decreases, most of the roof load is transferred to the surrounding rock, causing stress concentration and potential safety hazards.

The reason for this can be explained in stages: in the first three cycles, on one hand, the hydration of the CPB was not completely stopped due to the cement and fly ash contained in the CPB. The formation of hydration products overcame the damage caused by the chloride salt to the CPB in the initial phase. On the other hand, the chloride ions penetrating into the interior of the material at the initial stage reacted with the composition of the CPB to produce expansion products such as Friedel’s salt, which filled the pores inside the CPB, causing its UCS and Young’s modulus to increase slightly. However, as the number of cycles increased, the hydration weakened. On the contrary, the erosion caused by chloride is more obvious and excessive expansion products lead to increased internal stress in the CPB, and the UCS and Young’s modulus of the CPB are greatly reduced.

3.2. Strength Model and Damage Process Analysis of CPB

3.2.1. Damage Constitutive Model of CPB

Material deterioration is a reduction in the effective bearing area due to flaws, which can be described in terms of continuity. Assuming that the bearing area of the filler material in the undamaged state is A, under uniaxial loading, the internal cracks will continue to develop and the effective bearing area after damage will be reduced to $\tilde{A}$ and the continuity can be defined by a scalar φ:

$$\phi =\frac{\tilde{A}}{A}$$

(5)

A complementary parametric damage variable D of the degree of continuity φ is introduced:

$$D=1-\phi$$

(6)

When no load is applied to the CPB, it is considered to be in a non-destructive state at this time and D = 0; when it reaches the ultimate damage state, D = 1.

The strain equivalence assumption proposed by Lemaitre is widely used to establish the damage constitutive equations [21]. It can be expressed as, for damaged elastic-brittle materials under the action of the true stress σ, the strain in the damaged state is equivalent to the strain of the virtual element under the action of the effective stress $\tilde{\sigma }$.

Using the strain equivalence assumption, the stress–strain relationship of the damaged material in the one-dimensional case can be expressed as:

$$\epsilon =\frac{\tilde{\sigma }}{E}=\frac{\sigma }{{\rm E}\left(1-D\right)}$$

(7)

Then the damage constitutive relation of the CPB under one-dimensional load is:

$$\tilde{\sigma }=\frac{\sigma }{\left(1-D\right)}$$

(8)

The damage constitutive model of the CPB is established by $\tilde{\sigma }=E\epsilon$:

$$\sigma =E\epsilon \left(1-D\right)$$

(9)

The damage evolution of the CPB is generated by the continuous initiation and expansion of micro-elements. Assuming that there is a relationship P(ε) between the failure probability of micro-elements and strain and assuming that the damage distribution is isotropic, the damage variable D can be expressed as [22]:

$$D={\int }_0^{\epsilon }P\left(x\right)dx$$

(10)

Assuming that the micro-element failure probability of the CPB conforms to the Weibull distribution, the micro-element failure probability density is:

$$P\left(\epsilon \right)=\frac{m}{n}{\left(\frac{\epsilon }{n}\right)}^{m-1}\mathit{exp}\left[-{\left(\frac{\epsilon }{n}\right)}^m\right]$$

(11)

where m and n are parameters related to material properties.

Substitute Equation (11) into (10) to get:

$$D={\int }_0^{\epsilon }P\left(x\right)dx=1-\mathit{exp}\left[-{\left(\frac{\epsilon }{n}\right)}^m\right]$$

(12)

Substitute Equation (12) into Equation (9) to get:

$$\sigma =E\epsilon \mathit{exp}\left[-{\left(\frac{\epsilon }{n}\right)}^m\right]$$

(13)

Taking the logarithm of Equation (13), we get:

$$\mathrm{l}\mathrm{n}\frac{\sigma }{E\epsilon }=-{\left(\frac{\epsilon }{n}\right)}^m$$

(14)

By defining the stress and strain at the peak load as ${\epsilon }_c$ and ${\sigma }_c$, respectively, and $\frac{d\sigma }{d\epsilon }=0$ at the peak:

$$\frac{d_{\sigma }}{d_{\epsilon }}{|}_{\epsilon ={\epsilon }_c}=Eexp\left[-{\left(\frac{{\epsilon }_c}{n}\right)}^m\right]\left[1-m{\left(\frac{{\epsilon }_c}{n}\right)}^m\right]$$

(15)

in the formula, $Eexp\left[-{\left(\frac{{\epsilon }_c}{n}\right)}^m\right]\ne 0,$ so $\left[1-m{\left(\frac{{\epsilon }_c}{n}\right)}^m\right]=0$, we get:

$${\left(\frac{{\epsilon }_c}{n}\right)}^m=-\frac{1}{m}$$

(16)

Substitute Equation (16) into Equation (14) to get:

$$m=\frac{1}{\mathrm{l}\mathrm{n}\left(E{\epsilon }_c/{\sigma }_c\right)}$$

(17)

$$n=\frac{{\epsilon }_c}{{\mathrm{l}\mathrm{n}\left(E{\epsilon }_c/{\sigma }_c\right)}^{\mathrm{l}\mathrm{n}\left(E{\epsilon }_c/{\sigma }_c\right)}}$$

(18)

Substitute Equation (18) into Equation (9) to obtain the damage constitutive model of the CPB:

$$\sigma =E\epsilon \mathit{exp}\left[-\frac{1}{m}{\left(\frac{\epsilon }{{\epsilon }_c}\right)}^m\right]$$

(19)

According to the stress–strain curve obtained in the uniaxial compressive strength test, the mechanical performance parameters of the CPB such as peak stress, peak strain, elastic modulus and the property parameter m can be calculated. The mechanical parameters of the CPB under different cycles have obvious differences. Therefore, the damage constitutive models are different. The fitting coefficients of all models are greater than 0.98, and this damage constitutive model is reasonable and reliable. When compared with the stress–strain values obtained from uniaxial compression tests, the error values obtained are all less than 5%, indicating that this damage constitutive model is reasonable and reliable. This model can describe the variation law of the filling strength under chloride salt and dry–wet cycles.

Table 3. Damage constitutive equation for backfill under different erosion ages.

Cycle Time	Damage Constitutive Model	Error/%
1	$1120\epsilon \mathit{exp}\left[-0.211{(\epsilon /0.0049)}^{4.745}\right]$	3.85
2	$1194\epsilon \mathit{exp}\left[-0.251{(\epsilon /0.0051)}^{3.982}\right]$	3.45
3	$1100\epsilon \mathit{exp}\left[-0.316{(\epsilon /0.0061)}^{3.163}\right]$	2.78
4	$1121\epsilon \mathit{exp}\left[-0.437{(\epsilon /0.0062)}^{2.287}\right]$	4.55
5	$1025\epsilon \mathit{exp}\left[-0.212{(\epsilon /0.0058)}^{3.235}\right]$	3.45
6	$1010\epsilon \mathit{exp}\left[-0.208{(\epsilon /0.0057)}^{3.225}\right]$	4.77
7	$882\epsilon \mathit{exp}\left[-0.377{(\epsilon /0.0057)}^{2.652}\right]$	4.21
8	$670\epsilon \mathit{exp}\left[-0.260{(\epsilon /0.0056)}^{3.944}\right]$	3.23

shows the damage constitutive equations at different erosion ages with error values.

3.2.2. Analysis of the Damage Process of CPB

The D-axial strain curve of the damage variable can be obtained from Equation (12), as shown in Figure 3. Observe the development characteristics of the damage variables of the CPB with different erosion ages, which basically correspond to the initial slow increase, then a rapid increase and finally a flat trend. According to the stage division characteristic points on the stress–strain curve, the damage development process of the CPB is divided into the following five stages [23]:

(1)

The initial undamaged stage, the OA section, corresponds to the non-linear compaction stage of deformation and the failure of the CPB. The internal micro-cracks are tightly closed, the surface is undamaged and the damage variable is essentially zero.

(2)

The initial damage stage, the AB segment, corresponds to the linear elastic deformation stage, wherein the internal micro-cracks are initiated and expanded and new cracks are generated. The damage variable is small but continues to accumulate.

(3)

The damage development stage, BC section, corresponds to the stage of steady micro-crack expansion, wherein the internal micro-cracks develop and penetrate steadily and macro-cracks gradually form, and the damage variable increases sharply.

(4)

The damage destruction stage after failure, the CD segment, corresponds to the failure stage, wherein the CPB has reached the peak strength, and there are still a large number of cracks expanding and converging to continue to damage the internal structure, and the damage variable increases linearly with the increase in strain.

(5)

The residual damage stage, after point D, corresponds to the residual deformation stage, wherein the CPB is almost completely destroyed under the effect of the load and the damage variable increases slowly and is finally completely destroyed with increasing strain.

4. Permeability Analysis of CPB

4.1. Analysis of CPB Penetration Test Results

The results of the chloride ion penetration experiment are shown in Figure 4. A clear boundary between the white area and the brown area can be observed on the cross section. The bright white area is the chloride ion penetration area and the brown area is the non-penetration area.

Influenced by the color of the specimen section, the color boundary is not obvious, which is inconvenient for measurement and calculation. Digital image processing (DIP) is used to process the image. Figure 5 is an image processed using digital image technology. The yellow area of the image is the chloride ion penetration boundary area, so the distance from the surface to the boundary can be checked; the red color is the penetration area, and the yellow area to the edge of the specimen is the chloride ion penetration depth. According to Figure 5, in the first three dry–wet cycles, there are few red areas and only a clear yellow boundary appears around the CPB, and the erosion effect is small; that is, the penetration depth of chloride ions gradually increases. During the eighth dry–wet cycle, the red area penetrated to the center of the CPB and the erosion effect of the CPB reached the maximum at this time.

4.2. Analysis of Chloride Ion Penetration Characteristics

The chloride ion concentration at the discoloration boundary is calculated to be about 0.2~0.3 mol/L. In the early stage of erosion, the concentration value increases rapidly, it is the chloride ion that is continuously accumulated into the material by Ca-pillar adsorption and diffusion, but a part of the chloride ion is subject to chemical combination and physical adsorption of the material itself, so the measured chloride ion concentration value is low. When the adsorbed chloride ions reach a saturation state, the growth rate decreases and shows a steady upward trend.

Chloride ions migrate into the material by diffusion, and the penetration depth of chloride ions increases linearly with time. The penetration depth d and penetration time t can be fitted as a linear function, as shown in Figure 6, the expression is as follows:

$$d=13.4733+{1.3515}^{-6}·t ({\mathrm{R}}^2=0.99)$$

(20)

In the formula, d—chloride ion penetration depth and t—chloride ion penetration time.

As the erosion age increases, the apparent diffusion coefficient of chloride ions first decreases rapidly and then tends to stabilize. This trend can be characterized by a power function as shown in the following formula [24]:

$$D_{app}=D_0{\left(\frac{t_0}{t}\right)}^m$$

(21)

In the formula, D_app—the apparent diffusion coefficient of chloride ions of the filling material at any erosion age; D₀—the apparent diffusion coefficient of chloride ions at the age t₀; t₀—the time relative to D₀ (usually 28 d); m—environmental attenuation factor.

Use Formula (21) to fit the obtained experimental data, and the fitting result is shown in Figure 7 The function expression of the curve in Figure 7 is:

$$D_{app}={3.8182}^{-9}{\left(\frac{2419200}{t}\right)}^{2.3874} ({\mathrm{R}}^2=0.93)$$

(22)

The diffusion coefficient D_app decreases with increasing erosion time t. In the early stage of erosion, chloride ions have a high diffusion rate, which is dominated by capillary adsorption and diffusion, and the diffusion coefficient has a distinct downward trend. This is caused by the phenomenon of chemical bonding and physical adsorption; in the later stage of erosion, the diffusion coefficient decreases slightly, and the decrease in the last five cycles is only 11.85%. The migration of chloride ions to the deep part of the material is dominated by diffusion, and the secondary hydration and chemical bonding generate a large volume of salt, which reduces the porosity of the material, and the diffusion coefficient is relatively small and gradually decreases.

The aforementioned details have explored the change rule of the CPB damage characteristics under chloride dry–wet cycle conditions from mechanical experiments and permeability tests, respectively. According to the above analysis of the strength damage and permeability test, the CPB is more severely damaged under chloride salt and dry–wet cycle conditions. In the designed experiment in this paper, the strength is reduced by a maximum of 38.2% after only eight cycles, and the maximum depth of chloride ion damage is 20.5 mm. According to the conclusion of this experiment, we should pay attention to the erosion of chloride dry–wet cycle on the CPB.

5. Conclusions

(1)

Under the effect of chloride salt erosion and dry–wet cycling, the strength and Young’s modulus of the CPB first increased and then decreased rapidly. After the third cycle, its strength and Young’s modulus increased by 15.5% and 9.5%, respectively, compared to the strength values of the uneroded 28d specimen. After eight cycles, the strength decreased by 32.2% and 38.2%, respectively.

(2)

Based on the strain equivalence assumption, construct the constitutive damage equation of the backfill under uniaxial conditions. According to the constitutive equation, the damage variable D-axial strain curve of the CPB under different erosion ages can be obtained. They also have a similar trend; the damage development process of the CPB can be divided into five stages: the initial non-damage stage, the initial damage stage, the damage development stage, damage destruction and the residual damage stage, which is similar to the five stages of deformation and the damage process of the CPB corresponding to the stage.

(3)

Based on the chloride ion penetration images processed by digital image technology, it is found that the penetration depth of chloride ions gradually increases with the increasing number of dry–wet cycles. At the eighth dry–wet cycle, the penetration depth of chloride ions is the greatest. At this point, the CPB is subjected to the double erosion of the dry–wet cycle and the chloride salt, which eventually leads to structural damage.

(4)

The calculation shows that the maximum penetration depth of chloride ions is 20.5 mm, the maximum diffusion coefficient is 18.99 × 10⁻¹⁰ m²/s and the maximum concentration can reach 0.305 mol/L. Through function fitting, the development of the penetration depth is characterized as a linear function, and the change trend of the apparent diffusion coefficient of the ions is characterized by a power function.

Expectation

Due to the limitation of research time, the number of dry and wet cycles in this paper cannot represent all cycles. Therefore, if other scholars are interested in this, it is recommended to extend the number of dry and wet cycles. By aiming at the research topic in this paper, we have grasped the main existing erosion conditions of a mine: chlorine salt and dry–wet cycle. If other scholars are interested in this, they can look for other erosion conditions, such as sulfate, temperature and other factors, or multi-factor coupling analysis. In combination with macro and micro tests, we can carry out more detailed mechanism analysis to provide a theoretical basis for more mines.