The Mesoscopic Damage Mechanism of Jointed Sandstone Subjected to the Action of Dry–Wet Alternating Cycles

Abstract

The effect of the dry–wet cycle, characterized by periodic water level changes in the Three Gorges Reservoir, will severely degrade the bearing performance of rock formations. In order to explore the effect of the dry–wet cycle on the mesoscopic damage mechanism of jointed sandstone, a list of meso-experiments was carried out on sandstone subjected to dry–wet cycles. The pore structure, throat features and mesoscopic damage evolution of jointed sandstone with the action of the dry–wet cycle were analyzed using a-low-field nuclear magnetic resonance (NMR) technique. Subsequently, the impact on the mineral content of dry–wet cycles was studied by small angle X-ray scattering (SAXS). Based on this, the mesoscopic damage mechanism of sandstone subjected to dry–wet cycles was revealed. The results show that the effects of the drying–wetting cycle can promote the development of porous channels within sandstone, resulting in cumulative damage. Besides, with an increase in dry–wet cycles, the proportion of small pores and pore throats decreased, while the proportion of medium and large pores and pore throats increased. The combined effects of extrusion crush, tensile fracture, chemical reaction and dissolution of minerals inside the jointed sandstone contributed to the development of mesoscopic pores, resulting in the increase of porosity and permeability of rock samples under the dry–wet cycles. The results provide an important reference value for the stability evaluation of rock mass engineering under long-term dry–wet alternation.

1. Introduction

Because of the change of groundwater levels and rainfall, rocks and soils undergo alternating wetting and drying cycles. The wet–dry alternation action leads to obvious damage and deterioration in the rock mass, which will cause various geotechnical disasters such as landslides, collapse and rockfalls [1,2,3]. The fundamental reason for the performance degradation of rock masses is the accumulation of meso-damage to the rock [4,5]. Therefore, it is very necessary to study the meso-damage deterioration mechanism of jointed rock masses under the alternating dry–wet environment.

To date, a great number of studies and experiments have been conducted on the mechanical properties and damage properties of rock under alternating dry–wet action [6,7,8,9,10]. For example, Vasarhelyi [11] explored the strength and deformability of sandstones in dry and wet conditions. Hua et al. [12] investigated the influence of dry–wet cycles on the composite fracture toughness of sandstone. With an increasing number of dry–wet cycles, the fracture toughness of mode I and mode II decreased. Erguler et al. [13] and Wild et al. [14] concluded that cyclical dry–wet treatments lead to variations of mineralogical compositions and failure characteristics. Song et al. [15] explored the performance degradation and damage mechanism of sandstone under alternating wet and dry conditions. The results showed that the loss rate of wave velocity was positively correlated with the number of dry–wet alternating cycles. Shi et al. [16] investigated the evolution mechanisms of sandstone damage under the action of alternating dry–wet cycles. It was found that the strength of the rock gradually decreased under this action. Meng et al. [17] discussed the influence of dry–wet alternation action on the physico–mechanical parameters of sandstone, and found that the elastic modulus and density decreased with the increase of the number of dry–wet alternating cycles. Zhao et al. [18] revealed the relationship between the degradation mechanism and dry–wet alternation. Liu et al. [19] studied the mechanical properties of sandstone and declared that the tensile strength and compressive strength decrease with the increase of the wet–dry alternation number, while Colas et al. [20] found that the degradation of surrounding rocks was accelerated by long-term water circulation. Yuan et al. [21] researched the mechanical stability of sandstone under acidic dry–wet cycles and dynamic loading, and noted that peak stress and elasticity decreased as the acidity and cycle frequency increased. In the research mentioned above, the macroscopic damage features and evolution laws of rock masses subjected to dry–wet alternation is revealed. However, research on the internal damage of jointed rock masses during dry–wet alternation action is relatively scarce.

Rock masses are a complex solid material, and it is hard to accurately measure or describe their meso-damage by means of the conventional methods. In recent years, nuclear magnetic resonance technology (NMR), CT scanning technology and other technical methods have been widely adopted to research the meso-degradation law of rock [22,23,24,25]. With the aid of CT scanning technology, Zhou et al. [26] explored the cracking process in coal and rock under compression and found that the cracking process is partitioned into an initial stage, a stable growth stage and an accelerated growth stage. Wang et al. [27] studied the mesoscopic damage of fractured sandstone under compression, and proposed a macro–meso damage constitutive model using logistic growth theory. Xu et al. [28] concluded that cohesion and internal friction angle decrease under the action of dry–wet alternation. Guo et al. [29] researched the microscopic structure and mineral composition of sandstones subjected to the action of dry–wet alternation and found that the weakening mechanisms of rock are frictional, chemical and corrosive deterioration. Ma et al. [30] employed scanning electron microscopy (SEM) and NMR to investigate microstructure features and pore changes of sandstone specimens, and their results demonstrated that porosity increased from 10.22% to 12.30% after 20 wet–dry cycles. Zhang et al. [31] determined the degree of microscopic structural damage of sandstone under dry–wet alternation action. By means of X-ray diffraction and scanning electron microscopy, Huang et al. [32] analyzed the damage features of sandstone subjected to dry–wet alternation action, and a constitutive damage model was established. Pu et al. [33] adopted electron observation tests to explore the microstructure of sandstone under the action of dry–wet alternation, and constructed a coupling model for reappearing microstructure damage evolution. The existing research has promoted the understanding of the meso-damage mechanism of rock. However, few studies focus on the meso-structural characteristics and meso-deterioration law of jointed rock under dry–wet alternation conditions. Therefore, it is urgently necessary to explore in depth the meso-deterioration mechanism of rock masses considering dry–wet alternation action.

In the present study, the microstructure damage characteristics of jointed sandstone under dry–wet alternation action were investigated using the low-field NMR and SAXS methods. NMR was used to study the mesopore structure (T₂ curve, porosity, permeability, pore diameter and pore throat distribution) of jointed sandstone subjected to a different number of dry–wet cycles. Additionally, the XRD technique was employed to analyze the influence of dry–wet alternation action on rock mineral content. The meso-deterioration mechanism of jointed rock subjected to the action of dry–wet alternation was revealed. These research results can provide technical support for geological disaster prevention in reservoir areas.

2. Experimental Principles and Method

2.1. Nuclear Magnetic Resonance Technology (NMR) and Small Angle X-Ray Scattering (SAXS)

NMR has been widely employed to explore the evolution and distribution laws of rock mesopores in recent years [34]. NMR studies the relaxation properties of H atoms in rock pores by means of the magnetic interaction between the external magnetic field and H atom nucleus, and then forms NMR imaging. The recovery time for atomic nuclei from a high-energy state to a low-energy state under the action of radio frequency pulse RF is called relaxation time, which consists of longitudinal relaxation time T₁ and transverse relaxation time T₂. T₂ usually represents the internal pore structure of rocks directly because of its short measurement time. Different relaxation mechanisms of T₂ involve volume relaxation, surface relaxation and diffusion relaxation. The explicit expression of transverse relaxation time T₂ is as follows [35]:

$${\displaystyle \frac{1}{T_2}}={\displaystyle \frac{1}{T_{2surface}}}+{\displaystyle \frac{1}{T_{2volume}}}+{\displaystyle \frac{1}{T_{2diffusion}}}$$

(1)

The transverse relaxation time is mainly controlled by the surface relaxation, and the last two terms in the above equation can be ignored: T₂ = T_2surface. As a result, the above equation can be simplified as [36]:

$${\displaystyle \frac{1}{T_2}}\approx {\displaystyle \frac{1}{T_{2surface}}}={\rho }_2{\left({\displaystyle \frac{S}{V}}\right)}_{pore}$$

(2)

In Equation (2), S and V are the pore surface area (cm²) and volume (cm³), respectively; ρ₂ is the rock surface relaxation rate (ms⁻¹).

SAXS is a powerful method to study the overall shape and structural transitions of material, which is different from large angle X-ray diffraction. In SAXS experiments, powdered samples of rocks are usually placed in a plastic disk, and the intensity of the scattered X-rays is recorded by an X-ray detector [37]. The principles of the SAXS test are shown in Figure 1. The resulting scattering mode is relative to the overall shape and size of the particles under exploration [38]. The X-ray scattering occurs when it is projected into the crystal as a kind of electromagnetic wave, and specific diffraction patterns can be obtained by recording the X-ray diffraction line. The different mineral compositions can be distinguished by comparing the position and intensity of the diffraction lines since each mineral has a specific X-ray diffraction pattern.

2.2. Preparation of Rock Samples

The sandstone specimens were collected from the hydro-fluctuation zone in the Three Gorges Reservoir. The specimens were processed into standard cylinders measuring Φ50 mm × 100 mm. Based on the X-ray diffraction pattern, the main mineral composition of the sandstone comprised quartz, feldspar, calcite, siderite and anhydrite and the total amount of clay minerals (TCCM), respectively, which accounted for 41.2%, 36.0%, 1.8%, 0.8%, 2.3% and 17.9%, respectively. The proportion of quartz minerals in the samples was less than 75%, while the proportion of feldspar was more than 25%. To ensure the consistency of the meso-experiment, the rock specimens were derived from the same rock block. Besides, the rock samples with large discreteness were removed by wave velocity testing technology. Subsequently, cracks with a dip angle of 45° and a length 2a of 10 mm were prefabricated using a high-pressure water jet directed at the center of the cylinders. The properties of gypsum are similar to those of filling materials in natural joints [39]. It was used to fill the precast jointed rock masses studied in this paper to simulate the jointed rock masses in practical engineering [40]. In the experiment, the gypsum was injected into the cracks when it was fluid, and solidified after several minutes. Experiment samples are shown in Figure 2.

2.3. Experimental Procedure and Device

To simulate the drying and wetting process in a more realistic way, the water samples were taken from the hydro-fluctuation zone of the Three Gorges Reservoir. According to the chemical elements tests, the water consisted of Ca²⁺, Mg²⁺, Na⁺, K⁺, ${\mathrm{HCO}}_3^{-}$,Cl^-, ${\mathrm{SO}}_4^{2-}$ and ${NO}_3^{-}$, with a PH = 7.2~8.3. Generally, the composition of water is stable perennially. All the samples for the experiments were taken from the same place. In nature, the water level of the Three Gorges Reservoir rises from June to August, while it drops from November to December. To accelerate the damage of sandstone under dry–wet alternation action, the vacuum saturation and oven drying methods were employed to simulate the effect of dry–wet alternation [41,42]. In detail, rock samples were firstly put into water under tank pressure of −80 kPa for 6 h. Then, samples were soaked for 18 h until the negative pressure increased to 0 kPa. After that, rock samples were dried at 105 °C for 24 h. The test flow is shown in Figure 3, where n represents the dry–wet alternation number. After the tests (n = 0, 1, 5, 10, 15 and 20), the jointed sandstones were taken out, and the meso-damage tests by means of NMR and SAXS technology were conducted. Throughout the testing process, we performed parallel tests (i.e., three sets of tests under identical conditions for each protocol) to minimize the uncertainty of the results.

In this research, a vacuum saturation instrument and electric drying oven were selected to carry out testing, as illustrated in Figure 3a. The advent and application of NMR technology provides a new and effective method to study rock pore structures. To quantitatively analyze the variation of mesoscopic pore structure of jointed sandstone subjected to the dry–wet alternation action, MacroMR12-150H-I low-field nuclear magnetic resonance spectrometer systems were employed to observe the mesostructure of the saturated rock samples. The non-destructive NMR equipment is shown in Figure 3b, and the main technical parameters of the low-field NMR system are compiled in

Table 1. CPMG sequence parameters of low-field NMR system.

System Parameters	Value	System Parameters	Value
Temperature of magnet T (°C)	32	RF 90° pulse width P1 (μs)	3.80
RF frequency SF (MHz)	12.68	RF 180° pulse width P2 (μs)	10.20
RF offset frequency O1 (Hz)	682,035.1	Repeat sampling interval TW (ms)	500
RFD (ms)	0.020	Sampling number TD	99,980
Analog gain (RG1)	20.0	Data radius DR	1
Cumulative collecting times (NS)	8	Pulse echo times NECH	5000

. The jointed sandstone samples with different dry–wet cycle numbers were processed into rock powder after drying and grinding. After that, the mineral composition and proportion of rock samples after the action of dry–wet cycles were identified using the small angle X-ray scattering instrument (Figure 3c).

3. Mesostructural Evolution of Sandstone Under Dry–Wet Cycles

In nature, dry–wet cycles generally cause damage to sandstone, which will decrease the strength of the rock, leading to instability. Essentially, damage to sandstone is mainly caused by the deterioration of the mineral structure of the rock. To reveal the damage mechanism of the sandstone under dry–wet alternation action, the variations of the porosity, permeability, pore size distribution, throat and mineral composition of jointed sandstone during the process of dry–wet alternation are investigated in this section.

3.1. The Evolution of Porosity and Permeability of Jointed Sandstone

The spaces in rocks that are not filled with particles, impurities and cementing substances are called the pore structure. The spatial volume and connectivity of the spaces can be characterized by porosity and permeability. In this section, the change of the porosity and permeability of sandstone specimens under the action of dry–wet alternation were observed by NMR technology. The mesostructure of sandstone was analyzed.

The evolution of cumulative porosity of jointed sandstone with the dry–wet alternation number is depicted in Figure 4. This shows all the cumulative porosity curves consisted of three phases, namely the slow accumulation stage, the rapid accumulation stage and the steady development stage. To be specific, the cumulative porosity curves changed slightly when t = 0–0.1 μs. Subsequently, the growth rate increased rapidly (0.1–100 μs). After 100 μs, the growth rate of the cumulative porosity curve decreased, while cumulative porosity stayed constant when it reached the maximum value. To explore the influence of the wet–dry cycle on the porosity of sandstone, the changes of the maximum accumulated porosity and permeability with the wet–dry cycle number were analyzed and are shown in Figure 5. In this figure, the accumulated porosity under different dry–wet cycles represents the peak value in Figure 4, and the permeability was calculated using the empirical formula (Coates model) according to the volume of movable and bound fluids of T₂ spectrum [43]. The study found that both accumulated porosity and permeability increased with the increase of the dry–wet alternation number. To determine the relationship between the total porosity of jointed sandstone and the dry–wet alternation number, the explicit fitting expression was deduced:

$${\mathrm{\Phi }}_n=p+q\ln \left(n+1\right)$$

(3)

where p denotes the initial porosity, n is the number of dry–wet alternation cycles, Φ_n is the total porosity corresponding to the dry–wet alternation number and q is the logarithmic fitting coefficient of the total porosity. In the experiment, p = 7.005% and the fitting coefficient is q = 0.372, yielding the correlation coefficient R² is 0.993.

On the basis of the experimental results, it was found that dry–wet alternation action promotes the development of porosity channels, resulting in damage to sandstone. When water infiltrates the sandstone, hydrophilic minerals such as potassium feldspar, albite and illite dissolve as the hydration reaction proceeds (Equations (4) and (5)); thus, the number of micropores increases gradually. Besides, the expansion of hydrophilic clay minerals leads to uneven stress inside the rock. Therefore, microcracks occur in the sandstone. Under the action of dry–wet alternation, the micropores and microcracks expand from the rock surface to the interior. As a result, the pore connectivity of rock samples is improved [44]. The more new pores and pore throats generated by dry–wet alternation action, the higher the total porosity and permeability. Although the variation trends of micropores, mesopores and macropores in jointed sandstone are different under dry–wet alternation conditions, the total porosity and permeability increase with an increasing dry–wet alternation number. The irreversible cumulative damage of jointed rock masses caused by dry–wet alternation action ultimately causes the significant weakening of mechanical properties [45].

$${\mathrm{CaCO}}_3{+\mathrm{H}}_2{\mathrm{O}=\mathrm{Ca}}^{2+}{+\mathrm{HCO}}_3^{-}{+\mathrm{OH}}^{-}$$

(4)

$$\mathrm{CaMg}{\left({\mathrm{CO}}_3\right)}_2{+2\mathrm{H}}^{+}{=\mathrm{Ca}}^{2+}{+2\mathrm{HCO}}_3^{-}{+2\mathrm{Mg}}^{2+}$$

(5)

3.2. The Variation of Pore Size Distribution and Throats in Jointed Sandstone

The pore structure of rock can be divided into pores (spaces surrounded by particles) and throats (channels between pores) according to the size and interrelationships of pore space, which represent the reservoir performance and the permeability of rocks, respectively. The pore diameter and pore throat are important indexes, which influence the porosity and permeability of rock samples. To further study the mesostructure of the sandstone, the pore diameter and throats of jointed sandstone under different dry–wet alternation conditions were investigated; the evolution of the mesostructure of sandstone was studied in this part.

Using NMR technology, the measurements of pores with different diameters were obtained (i.e., the pore size distribution of the jointed rocks). The results are depicted in Figure 6. We found three main peaks, corresponding to pore sizes of 0.001–0.1 μm (micropores), 0.1–10 μm (medium pores) and 10–100 μm (macropores), respectively. What is more, the first main peak value gradually decreased as the dry–wet process went on, while the second and third main peak presented the opposite trend, and the growth rate of the second main peak was more obvious. According to the classification criteria of the internal pores in concrete [25], the pores in jointed sandstone were further divided into five grades: micropores (G_I), small pores (G_II), mesopores (G_III), macropores (G_IV) and larger pores (G_V) (

Table 2. Classification of pores of rock specimens.

Grade	Symbol	Name	Pore Diameter rp	Remarks
I	GI	Micropore	<0.01 μm	Containing 0.002–0.01 μm transitional pores, partially cementing pores
II	GII	Small pore	0.01–0.1 μm	Containing 0.01–0.1 μm transitional pores, partially cementing pores
III	GIII	Mesopore	0.1–10 μm	Containing 0.1–100 μm capillary opening
IV	GIV	Macropore	10–100 μm	Containing 0.1–100 μm capillary opening
V	GV	Larger pore	>100 μm	Containing air pores, visible to the naked eye

). The proportion of pores with different pore size ranges under different dry–wet alternation action is presented in Figure 7. With the increase of the dry–wet alternation number, the proportion of G_I pores fluctuated slightly, staying at about 25%. The proportion of G_II pores decreased from 39.107% to 34.512% with an increasing dry–wet cycle. The proportion of G_V pores was the smallest, and the value was close 0. In addition, the proportions of G_III pore and G_IV pore increased with an increasing dry–wet alternation number. These phenomena also indicate that the G_II gradually changed to the range of a high-grade pore size.

The results above indicate that the pore diameter of G_II continued to expand to G_III, then the pores continued to expand and gradually turned to G_IV. According to the studies conducted by Fang et al. [46], during wetting, when the water infiltrates into the sandstone, the soluble minerals, including feldspar and calcite, are dissolved, and the space occupied by the soluble minerals before becomes a cavity. In addition, the dissolution of minerals will weaken the bond between the skeleton particles. After that, the uneven shrink of minerals will lead to microcracks between the skeleton particles in the drying process. Finally, the skeleton particles are disengaged from the rock in the dry–wet cycling. Especially, the water usually infiltrates into the sandstone through the pores and throats in the sand. Thus, the micropores transform to medium and macropores, and medium pores transform to macropores gradually. The more micropores exist in the sandstone, the more obvious the transformation from micropores to medium pores becomes. The variation of pore structure will inevitably weaken the strength of sandstone specimens, causing the decrease of compressive and tensile strength.

Similarly, the proportion of pore throats in the jointed rock samples with different dry–wet alternation numbers is illustrated in Figure 8. Based on size, the pore throats can be divided into three categories: small (GT_I), medium (GT_II) and large (GT_III). The classification standards of pore throats are presented in

Table 3. Pore throat classification of jointed rock specimens.

Grade	Symbol	Name	Pore Diameter Φ
I	GTI	Small pore throat	<0.1 μm
II	GTII	Medium pore throat	0.1–1 μm
III	GTIII	Large pore throat	1–25 μm

. According to the pore throat classification, the evolution of the grade of pore throats with the dry–wet cycles is plotted in Figure 9. As can be seen in the figure, with the increase of the dry–wet alternation number, the proportion of GT_I pore throats (small pore throats) of jointed sandstone decreased gradually, while the GT_II and GT_III pore throats (medium and large pore throats) increased gradually. As the dry–wet alternation went on, the original small pore throats inside the rock developed continuously into medium pore throats due to the hydration reaction and dissolution of soluble minerals. Meanwhile, the medium pore throats gradually transformed into large pore throats during the wet–dry cycle. Due to the large pore throat volumes, the number of small pore throats converted into medium pore throats was greater than the number of medium pore throats changed into large pore throats. Therefore, the proportion of small pore throats decreased, while the proportion of medium and large pore throats showed an increasing trend during the dry–wet alternation action, which was similar to the variation of the pores. The continuous development and connection of various pores and throats inside rock samples led to significant changes in the pore space structure, which in turn influenced the porosity and permeability of rock samples. Simultaneously, the sandstone microstructural changes during dry–wet alternation action also seriously influenced the macroscopic mechanical properties of the jointed rock.

3.3. Variation of Mineral Composition of the Sandstone

In addition to changing the pore structure, the dry–wet alternation conditions also led to the change of the mineral composition of the rock specimens. To further analyze the change of mineral composition during the dry–wet alternation process, an X-ray diffraction spectrum test on sandstone powder was conducted as the dry–wet alternation numbers changed. Furthermore, the precipitates in the water solution after the dry–wet cycles were dried. Sandstone powder samples with different dry–wet alternation numbers are shown in Figure 10a. The mineral contents of rock specimens with different dry–wet alternation numbers are plotted in Figure 11.

Figure 11 shows the contents of feldspar (potassium feldspar and albite) and calcite in the test rock samples gradually decreased with the increase of the dry–wet alternation number. In nature, the proportions of feldspar and calcite are 36% and 1.8%, respectively, while the proportions of feldspar and calcite decrease to 25.5% and 0.5% when n = 20. TCCM also decreases from 17.2% to 10.6% as the dry–wet alternation goes on. Due to the ion hydrolysis reaction, exchange reaction and protonation reaction of feldspar after wetting, the content of chlorite, montmorillonite and illite gradually decreases. In particular, the precipitation of chlorite is more obvious than that of illite and kaolinite, while the variations of kaolinite content is not obvious with the increase of dry–wet cycles [47,48,49]. As a result, the volume of pores and pore throats in the sandstone increases gradually [50]. Besides, the total content of clay minerals decreases with the increase of the dry–wet alternation number. The experimental results were similar to the results obtained by Zhang et al. [51]. However, quartz and anhydrite increased gradually due to the reduction of water-soluble mineral content in the rock samples. The decreases in the clay minerals weakened the strength of the grain skeletons in the rock specimens, leading to the damage and degradation of the rock.

After the tests, the mineral composition of the sediment in the solution was measured, which reflected a series of physicochemical reactions between the rock samples and the solution. Table 1 shows the solution comprised quartz (42.3%), feldspar (26.0%), calcite (7.9%), dolomite (4.5%), anhydrite (2.7%) and TCCM (16.6%), while the rank order of the mineral contents was the following: quartz > TCCM > plagioclase > calcite > potassium feldspar > anhydrite. The quartz in the sediment mainly came from quartz particles on the surface of the rock samples, which peeled off during the dry–wet cycle. The process was relatively intense during the testing, so the quartz content in the sediment was larger than that of other minerals. The content of feldspar, calcite and clay minerals in the sediment was relatively large because these minerals were dissolved in water or reacted with water during the dry–wet cycle, resulting in the continuous increase of pore volume in the rock samples. The decrease of feldspar, calcite, clay minerals and the continuous development of pores and pore throats eventually led to the rock strength weakening. The results were consistent with the variation of total porosity and pore throat values in Section 3, which explains the essential reason for the increase of porosity of sandstone specimens with an increasing dry–wet alternation number from the microscopic perspective. Meanwhile, in the experiments, it was found that the mass of the gypsum accounted for about 2.7% of the sediment after 20 dry–wet cycles. This shows that the filler gypsum of the samples also dissolved in water during the dry–wet alternation action. Therefore, the water’s effect on the filler gypsum was also a factor. The combined effects of the water susceptibility of the gypsum and physicochemical reactions of mineral components degraded the strength of the jointed rock specimens.

4. Damage Mechanism of Mesostructure in Sandstone

Based on the above results of NMR and SAXS testing on the meso-damage of jointed rock, the mesoscopic structural damage development law of sandstone during the dry–wet alternation action can be revealed. On the basis of the mesoscopic results, the meso-damage propagation law of jointed sandstone subjected to dry–wet alternation is illustrated in Figure 12, and the damage evolution process is as follows.

(i)

Pristine condition (Figure 12a). The sandstone was formed by mechanical, chemical and biological deposition of ore-forming materials. In nature, the mineral particles inside the sandstone are usually connected with cementitious materials in the form of porosity and contact; i.e., porous connection and contact connection. Besides, the gap between the mineral particles divided into pores and pore throats (Figure 6 and Figure 8), and pore structure and throat features are all inner factors that affect and determine the mesoscopic damage evolution of rock samples under the dry–wet cycle.

(ii)

Dry state (Figure 12b). In the dry state, due to the thermal expansion caused by changes in temperature, the compression and fragmentation of the inter-particles occurs in contact cementation, resulting in the formation of mineral particle debris. Meanwhile, the expansion of particles causes compression and fragmentation of the cementitious material between the clastic particles, inducing the generation of tiny amounts of cement debris. The combined effect of the two kinds of fragmentation mechanisms leads to microscopic structure changes in sandstone.

(iii)

Water-saturated state (Figure 12c). On the one hand, when the water permeates the rock specimens, the particle debris, cement debris and unstable mineral particles inside the rock (such as feldspar, calcite, etc.) are dissolved. The cohesion between the mineral particles will be weakened. This was demonstrated by the mineral analysis experiment in Section 3.3. On the other hand, the mineral particles will shrink slightly when the rock is cooled by water. The compression between particles will become tension, leading to microcracks in the contact cementation structure. Finally, the micropores continue to expand and new throats are produced, which further enhances the permeability of the rock sample.

(iv)

The second dry state (Figure 12d). When the water-saturated rock sample is dried again, the contact and porous cementation are squeezed again due to thermal expansion. The force between clastic particles changes from tension to compression, resulting in the formation of mineral particle debris and cement debris. The mineral particles in the rock sample go through tension-compression transformation under alternating dry–wet cycles, leading to the continuous fragmentation and dissolution of soluble minerals and cements. Besides, some closed pores are transformed into open pores due to the formation of new cracks in the inter-particle structure [52]. Therefore, with an increasing dry–wet alternation number, the proportion of small pores and pore throats decreases, while the proportion of medium and large pores and pore throats increases (Figure 7 and Figure 8).

(v)

Final state (Figure 12e). During the dry–wet alternation action, the size of pores and pore throats continues to expand due to the weakening of the cements between mineral particles. The proportion of mesopores and medium pore throats is significantly increased. Besides, the content of soluble minerals and cements in the pores of rock samples decreases continuously due to the erosion of water, so new pores and pore throats are gradually generated under the action of dry–wet alternation [44]. Meanwhile, the cementation effect on the inter-particles gradually decreases during the dry–wet cycle. Ultimately, the micropore morphology and structure change significantly (Figure 4 and Figure 5), resulting in the irreversible cumulative damage of jointed sandstone during dry–wet alternation action.

In conclusion, under a dry–wet cycle, the extrusion crush, tensile fracture, chemical reaction and dissolution action, etc., of mineral particles inside jointed sandstone jointly led to the continual expansion of mesoscopic pores, and ultimately resulted in the increase of porosity and permeability of the sandstone samples. The damage of the meso-scale structure of the sandstone samples was the main reason for the rock strength weakening.

5. Conclusions

In the present study, dry–wet alternation tests were conducted on jointed sandstone specimens. The pore structure, throat features, mesoscopic damage evolution and mineral composition of jointed sandstone were measured by NMR and SAXS. Subsequently, the mesoscopic damage evolution mechanism of jointed sandstone during the dry–wet alternation process was investigated. The main conclusions are summarized as follows:

(1)

The dry–wet alternation action can cause the development of the porosity channel, leading to cumulative damage in sandstone. Therefore, the accumulated porosity and permeability of sandstone increase as the wet–dry cycle goes on, and the cumulative porosity goes through the slow accumulation stage, the rapid accumulation stage and the steady development stage.

(2)

During the dry–wet alternation process, the dissolution soluble minerals, hydration reaction and uneven shrink increase the size of pores and pore throats. In detail, micropores transform to medium and macropores, while medium pores transform to macropores gradually. The more micropores exist in the sandstone, the more obvious is the transformation from micropores to medium pores. Besides, the proportion of small pore throats decreases gradually, while the proportion of medium and large pore throats increases gradually.

(3)

The contents of feldspar (potassium feldspar and albite), calcite and clay minerals (TCCM) in the test rock samples decreased as the dry–wet cycle went on, which led to the continuous development of pores and pore throats.

(4)

In the alternating dry–wet environment, the combined effects of extrusion crush, tensile fracture, chemical reaction and dissolution action, etc., of clastic particles inside the jointed sandstone led to the continual development of mesoscopic pore structures, which ultimately caused the porosity and permeability to increase. The damage process was the essential reason for the weakening of rock strength.