Dynamic Compressive Behavior of CFRP-Confined High Water Material

Abstract

As mining operations extend deeper underground, support structures are increasingly subjected to severe impact loads. The dynamic mechanical performance of column-type support systems has, therefore, become a pressing concern. In the present research, a Split Hopkinson Pressure Bar (SHPB) apparatus, combined with Scanning Electron Microscopy (SEM), is used to systematically examine how the water-to-cement ratio, number of carbon-fiber reinforced polymer (CFRP) layers, and strain rate influence the dynamic compressive behavior and microstructural evolution of CFRP-confined high-water material. The results indicate that unconfined specimens are strongly strain rate-dependent, with peak strength following a rise–fall trend. A lower water–cement ratio results in a denser internal structure and improved strength. Additionally, CFRP confinement markedly enhances peak strength and impact resistance, refines failure modes, and promotes the formation of denser hydration products by limiting lateral deformation. This confinement effect effectively mitigates microstructural damage under high strain rates. These findings clarify the reinforcement mechanism of CFRP from both macroscopic and microscopic perspectives, offering theoretical insights and engineering references for the design of impact-resistant support systems in deep mining applications.

1. Introduction

As shallow coal resources become increasingly depleted, mining activities are rapidly shifting to greater depths. This transition has led to an increase in rockburst occurrences, which now pose a significant threat to the safe operation of coal mines [1,2,3]. Current prevention strategies primarily involve multi-parameter precursor identification systems based on acoustic emission monitoring and other sensing technologies [4,5,6,7]. These are often combined with pressure relief techniques such as large-diameter borehole drilling [8,9,10,11,12], alongside roadway support methods like bolt–mesh systems for surrounding rock control [13,14,15]. Together, these approaches form an integrated “early warning–pressure relief–support” system for rockburst mitigation.

To further reduce the damage caused by rockbursts, researchers have proposed various novel energy-absorbing support systems based on different reinforcement strategies. Xu et al. [16] innovatively introduced a steel honeycomb sandwich panel as an energy-absorbing device to mitigate the impact energy from rockbursts. Zhang et al. [17] investigated the influence of joint roughness coefficient and interface morphology at the rock–shotcrete interface on the dynamic splitting behavior of rock–concrete composites. Fan et al. [18] developed a novel expansion–friction composite energy-absorbing anchor cable to address the vulnerability of conventional anchors under rockburst conditions in coal mines, significantly improving their impact resistance.

In addition to these reinforcement strategies, vertical support systems are also essential components of underground mining support technology [19,20]. Batchler [21] demonstrated that the mechanical performance of pumpable column supports is jointly influenced by external confinement and the properties of internal filling materials. In underground mine support, high-water quick-setting materials have attracted considerable attention due to their strong water affinity, high fluidity, rapid setting, and ease of application. Meanwhile, fiber-reinforced composite materials have been widely adopted by researchers owing to their high specific strength and excellent corrosion resistance. To address the limitations of conventional FRP-confined columns in accommodating large deformations in mining environments, Yu et al. [22] proposed a novel composite column structure composed of an outer FRP tube and an inner filling of coal gangue–calcium sulfoaluminate-based high-water cementitious material. Building upon this design, Zhao et al. [23] further investigated the effects of fiber type, number of confinement layers, and water–cement ratio on the static mechanical properties of FRP-confined high-water material. Their findings showed that FRP provides lateral confinement during compression, with the degree of confinement determined by the type and thickness of the FRP. Stronger confinement enhances both compressive strength and deformability. The drainage behavior of high-water material also contributes to its large-deformation capacity. Liu et al. [24] conducted triaxial compression tests to simulate lateral confinement and systematically studied the compressive behavior and bleeding mechanism of high-water material under varying confining pressures. The results revealed that the seepage threshold is most significantly influenced by the water–cement ratio, followed by confining pressure and curing time. The material initially undergoes slow volumetric shrinkage, followed by rapid compression and progressive densification. A lower water–cement ratio results in reduced free water content, requiring higher confinement to prevent damage caused by water migration. However, these studies have primarily focused on static conditions, and the dynamic mechanical behavior of such materials under high strain rates remains largely unexplored.

Numerous studies have investigated the dynamic mechanical behavior of fiber-reinforced confined concrete. Yang et al. [25] conducted Split Hopkinson Pressure Bar (SHPB) tests to examine the effects of different numbers of AFRP wrapping layers and strain rates on the dynamic response of concrete. Their results indicated a positive correlation between dynamic strength and strain rate, with AFRP-confined concrete exhibiting significantly higher strength and toughness than unconfined concrete. Xiong et al. [26] evaluated the compressive behavior of CFRP-confined concrete under high-strain rates and compared it with unconfined specimens. While unconfined concrete was highly sensitive to strain rate, showing increased strength with increasing strain rate, the confined concrete exhibited improved overall strength but less strain rate sensitivity. Guo et al. [27] developed a unified model for the dynamic compressive strength of FRP-confined concrete based on an experimental database. Under certain FRP confinement conditions, both the compressive strength and corresponding strain of unconfined and FRP-confined concrete increased with strain rate. Moreover, as the FRP confinement ratio increased, the dynamic enhancement factor also rose. However, the confinement effectiveness of FRP is more prominent under quasi-static loading. Under dynamic loading, concrete may fail before the FRP confinement is fully mobilized, resulting in a non-linear relationship between confinement ratio and enhancement efficiency. Jiang et al. [28] identified a typical three-stage stress–strain behavior of FRP-confined concrete under dynamic compression: initial rise, post-peak decline, and secondary rise. Their experiments confirmed this pattern and revealed that the lateral–axial strain curve under dynamic loading is similar in shape to that under quasi-static conditions, though the confinement effect of BFRP is more pronounced in the dynamic regime. Strain rate history analysis showed asynchronous changes between axial and lateral strain rates near the peak, leading to a lower dynamic confinement ratio compared to the static case and contributing to the post-peak decline. To characterize this effect, the concept of dynamic confinement ratio (fl,d/fcd) was introduced, with a threshold value of 0.045 proposed to ensure effective confinement. While these studies provide a solid theoretical foundation for the present work, the dynamic behavior of high-water backfill materials confined by fiber-reinforced composites remains largely unexplored.

Building upon the above, this study focuses on the dynamic behavior of CFRP-confined high-water material. A series of Split Hopkinson Pressure Bar (SHPB) impact compression tests was conducted, accompanied by Scanning Electron Microscopy (SEM) analysis. Both unconfined and CFRP-confined groups were tested under varying conditions, including different water–cement ratios (1.25, 1.5, 1.75), numbers of CFRP confinement layers (0, 1, 3), and strain rates. The experiments systematically examined the effects of water–cement ratio, confinement condition, and strain rate on the dynamic compressive strength and microstructural characteristics of CFRP-confined high-water material.

2. Experimental Program

2.1. Test Specimens

The specimens were primarily composed of high-water material, water, adhesive, and carbon fiber-reinforced polymer (CFRP) sheets. The high-water material, supplied by angzhou Zhongkuang Building New Materials Technology Co., Ltd., Yangzhou, China, consisted of two separate components: Component A, composed mainly of calcium aluminate and calcium sulfoaluminate, and Component B, primarily composed of gypsum and lime. Tap water was used for mixing. The adhesive and CFRP sheets, responsible for bonding and confinement, respectively, were provided by Shanghai Hanma Construction Technology Co., Ltd., Shanghai, China. The tensile properties of the CFRP are listed in

Table 1. Mechanical Properties of CFRP.

Material	Average Tensile Strength (MPa)	Average Strain (×10−2)	Elastic Modulus (GPa)
CFRP	3346.89	1.23	267.57

[29]. The adhesive was prepared by mixing carbon fiber impregnating resin (HM-180C3P, Part A) and structural epoxy resin (HM-120CP, Part B) at a ratio of 2:1.

As shown in

Table 2. Grouping and Number of Specimens.

Group Label	Bullet Velocity (m/s)	Water–Cement Ratio	Number of CFRP Layers	Number of Specimens
B-C0-v4	4~5	1.50	0	4
B-C0-v5	5~6			4
B-C0-v6	6~7			4
B-C3-v9	9~10		3	3
B-C3-v10	10~11			3
B-C3-v11	11~12			3
B-C3-v12	12~13			3
A-C0-v5	5~6	1.25	0	4
C-C0-v5		1.75		4
A-C3-v12	12~13	1.25	3	3
C-C3-v12		1.75		3
B-C1-v9	9~10	1.50	1	3

, a total of 41 cylindrical specimens were prepared for the experiment, including 20 unconfined specimens and 21 CFRP-confined specimens. The tests were conducted using a Split Hopkinson Pressure Bar (SHPB) apparatus under impact compression loading. Grouping was performed based on the single-variable method. The water-to-binder ratio (WB) was categorized into 1.25, 1.50, and 1.75, labeled as A, B, and C, respectively. The number of CFRP layers was denoted as C0, C1, and C3. Different bullet velocity intervals were coded as “v + the lower bound of the interval.” For example, the interval of 4–5 m/s was labeled as v4. Each specimen was identified by a combination of its water-to-binder ratio, CFRP layer count, and bullet velocity range. For example, specimen A-C0-v2 refers to a sample with a water-to-binder ratio of 1.25, no CFRP confinement, and a bullet velocity range of 5–6 m/s.

The specimen preparation procedure was as follows. Components A and B, along with two portions of water, were weighed according to the designed mix proportions. Component A was first mixed thoroughly with one portion of water, and Component B was mixed with the other portion. The two slurries were then rapidly combined and stirred immediately, given the rapid setting behavior after their mixing. Once mixed, the slurry was quickly poured into cylindrical molds with a diameter of 50 mm and a height of 100 mm to form the matrix.

After complete setting, the specimens were demolded and wrapped in plastic film for natural curing. After 3 days of curing, CFRP sheets were prepared for confinement. Sheets with a width of 10 cm and lengths of 20 cm and 52 cm were cut to provide one-layer and three-layer confinement, respectively. The CFRP sheets were bonded to the specimen surfaces using pre-mixed adhesive. To prevent peeling at the ends, the terminal sections were wrapped and secured with plastic film. After the adhesive was fully cured, initial confinement was complete.

To meet the height-to-diameter ratio requirements of the Split Hopkinson Pressure Bar (SHPB) test—recommended to be between 0.5 and 1.0 according to the GB/T 34108-2017 standard [30]—all specimens were machined to a standard size of 25 mm in height and 50 mm in diameter. This ensured consistency in testing conditions and comparability of the results. The detailed specimen preparation process is illustrated in Figure 1.

2.2. Experimental Equipment and Procedure

The dynamic loading tests were conducted using a Split Hopkinson Pressure Bar (SHPB) apparatus with a bar diameter of 50 mm. A schematic diagram of the SHPB system is shown in Figure 2.

In the SHPB tests, different strain rate conditions were achieved by adjusting the air pressure in the cylinder, which controlled the striker velocity. This, in turn, modified the shape of the one-dimensional stress wave propagating through the incident bar. When the stress wave reached the specimen, part of the wave energy was reflected into the incident bar, while the remaining portion was transmitted into the transmission bar. Strain gauges positioned between the incident and transmission bars were used to capture waveform signals, allowing the incident, reflected, and transmitted waves to be recorded separately. Based on stress wave theory and the assumption of stress equilibrium, the time-dependent strain, strain rate, and stress of the specimen were calculated using Equations (1)–(3). To ensure experimental accuracy, specimens were carefully aligned axially during testing. To minimize the influence of end-face friction on the specimen’s stress state, a thin layer of petroleum jelly was applied to both ends of each specimen to reduce the friction coefficient at the specimen–bar interfaces.

$${\dot{\epsilon }}_s(t)=\left({\displaystyle \frac{2C_0}{L_s}}\right){\epsilon }_r(t)$$

(1)

$${\epsilon }_s\left(t\right)=\pm {\left({\displaystyle \frac{2C_0}{L_s}}\right)}^t{\int }_0^t{\epsilon }_r\left(t\right)dt$$

(2)

$$\sigma (t)=\pm E{\displaystyle \frac{A_B}{A_S}}{\epsilon }_t(t)$$

(3)

In these equations, ${\dot{\epsilon }}_s\left(t\right)$ denotes the strain rate of the specimen at time $t$ (s⁻¹), ${\epsilon }_s\left(t\right)$ is the strain of the specimen, and $\sigma \left(t\right)$ is the stress (Pa). ${\epsilon }_r\left(t\right)$ and ${\epsilon }_t\left(t\right)$ represent the strain signals of the reflected and transmitted waves, respectively. $C_0$ is the wave velocity in the pressure bar (m/s), $L_s$ is the specimen length (m), and $E$ is the elastic modulus of the pressure bar material (Pa). $A_B$ and $A_S$ are the cross-sectional areas of the pressure bar and the specimen, respectively (m²).

Figure 3 shows the incident, reflected, and transmitted waves recorded during the tests. The dynamic compression experiments were conducted at strain rates ranging from 100 to 300 s⁻¹. High-water materials often exhibit relatively high strain rates in SHPB tests, primarily due to their low strength, low elastic modulus, and high deformability. Compared to conventional concrete, high-water materials contain more free water and have a looser internal structure, making them more prone to rapid deformation under impact loading. This results in greater strain accumulation within a short period.

Moreover, the acoustic impedance of high-water material is significantly lower than that of the steel pressure bars, leading to strong wave reflections at the specimen–bar interfaces. This reduces the effective loading duration and accelerates strain accumulation. The material tends to fracture rapidly under impact, often before stress redistribution is completed, which further increases the strain rate. Therefore, even under the same loading velocity, the inherent characteristics of high-water material—such as their high water content and soft structure—cause them to respond more quickly, reaching strain rate levels significantly higher than those typically observed in concrete.

2.3. Microstructural Observation via SEM

To investigate the internal microstructure of high-water material after impact loading, scanning electron microscopy (SEM) observations were performed on the residual fragments of specimens subjected to SHPB dynamic compression tests. Representative fragments were selected from the fractured specimens, and each was split or cut along the fracture surface to expose a fresh internal cross-section suitable for observation.

The selected surfaces were then coated with a gold film approximately 20 nm thick using a vacuum sputter coater to ensure adequate electrical conductivity. Sputtering was conducted for at least 20 min to form a uniform conductive layer. After coating, the samples were placed in a high-vacuum drying chamber for more than 30 min of evacuation. Following this standard SEM preparation procedure, the fragments were mounted on specimen holders for microstructural observation.

3. Results and Discussion

3.1. Dynamic Compressive Behavior of Unconfined High-Water Material

3.1.1. Effect of Strain Rate on Peak Stress

As shown in Figure 4, at a fixed water–cement ratio of 1.50, the peak stress generally follows an initial increase followed by a decrease as the bullet velocity increases. When the velocity rises from v1tov2, the peak stress increases significantly, reaching a maximum of 9.75 MPa in the v2 range, with absorbed energy of about 9.01 J. However, further increasing the velocity to v3 results in a decline in peak stress. This reduction is attributed to rapid failure before the internal stress redistribution is completed. Under such high loading rates, microcracks initiate and propagate almost instantaneously, leading to premature macro-scale failure and a consequent drop in strength. Additionally, strain rate is positively correlated with bullet velocity: higher velocities result in higher strain rates.

3.1.2. Effect of Water–Cement Ratio on Peak Stress

As illustrated in Figure 5, under a constant bullet velocity, the peak stress decreases significantly with increasing water–cement ratio. When the ratio increases from 1.25 to 1.75, the peak stress drops from 16.75 MPa to 5.70 MPa, and the absorbed energy decreases from 14.05 J to 5.30 J. These results demonstrate that the strength of high-water material is highly sensitive to the water–cement ratio. At lower ratios, hydration products more fully fill the internal pores, resulting in a denser microstructure, higher strength, and inhibited crack propagation. In contrast, higher ratios increase the free water content and porosity after setting, making the material more porous and reducing its strength. Under unconfined conditions, the dynamic compressive strength of high-water material is highly strain-rate sensitive. It tends to increase with rising strain rate within a certain range, but when the loading rate becomes excessively high, the strength declines instead. This is because at very high strain rates, internal pores cannot close in time to bear increased stress, and rapid pulverization under impact leads to a sudden loss of strength.

3.1.3. Analysis of Fractal Dimension Results

To quantitatively characterize the particle size distribution of fragments after impact failure, the cumulative mass method in fractal theory was used to calculate the fractal dimension. The calculated values were statistically analyzed and are presented in Figure 5. The fractal dimension was determined using the following equation:

$$\mathrm{D}=3-\mathrm{b}$$

(4)

$$b={\displaystyle \frac{\mathit{log}(\frac{M_R}{M})}{\mathit{log}R}}$$

(5)

In this equation, $\mathrm{D}$ is the fractal dimension; $\mathrm{b}$ is the slope of the linear fit of $\mathit{log}\left({\displaystyle \frac{M_R}{M}}\right)$ versus $\mathit{log}R$; $M_R$ is the cumulative mass (g) of fragments passing through a sieve with an aperture size $R$; and $M$ is the total mass of the fragments (g).

Figure 6 presents the fractal dimensions of fragmented specimens under different strain rates and water–cement ratios, along with representative images of the fracture morphologies. As shown in the figure, under the same water–cement ratio, an increase in strain rate leads to more severe fragmentation. Large blocks are significantly reduced, while the proportion of fine debris increases. The corresponding fractal dimension increases from approximately 3.0 to over 3.6, indicating a shift in fracture mode from block-like failure to pulverized failure. This demonstrates that higher strain rates result in more intense specimen fragmentation, with smaller fragment sizes and a clear upward trend in fractal characteristics. Under the same impact velocity, specimens with higher water–cement ratios show a higher proportion of fine particles after failure, resulting in higher fractal dimensions. In contrast, specimens with lower water–cement ratios tend to retain larger fragments, corresponding to lower fractal dimensions. The comparison of fracture morphologies further supports this trend: specimens with a water–cement ratio of 1.25 retained relatively large and intact fragments after failure, whereas those with a ratio of 1.75 were almost completely reduced to fine particles.

3.2. Dynamic Compressive Behavior of CFRP-Confined High-Water Material

3.2.1. Stress–Strain Curves Under Different Confinement Layers

Figure 7 compares the stress–strain curves of high-water specimens with one-layer and three-layer CFRP confinement under a bullet velocity of 9–10 m/s at a water–cement ratio of 1.50. For reference, the corresponding quasi-static stress–strain curves and typical post-impact failure cross-sections are also included. The results demonstrate that CFRP confinement significantly enhances the impact resistance of high-water material. As the number of CFRP layers increases from one to three, the peak stress rises from approximately 11 MPa to nearly 19 MPa—an increase of over 72%. The peak strain also increases, indicating that multilayer confinement improves the material’s deformation capacity.

This enhancement is consistent with previous findings on FRP-confined concrete, where external confinement induces a triaxial compression state in the core material, thereby improving both compressive strength and ductility. The observed failure morphologies further support this conclusion. The specimen with one-layer CFRP confinement exhibited edge spalling and localized fiber rupture, with scattered debris after impact. In contrast, the specimen with three-layer confinement maintained an almost intact appearance, with only minor cracks near the ends and no major structural disintegration, as shown in Figure 7. In addition, partial debonding was observed at the CFRP–high-water material interface. This localized debonding indicates that under high strain rates, interfacial stresses can exceed the bonding strength, resulting in interfacial slip or separation in specific regions. Nevertheless, the CFRP layer as a whole remained intact and continued to provide effective lateral confinement, consistent with the overall behavior of CFRP-confined high-water materials, where confinement primarily resists lateral expansion but may still experience localized weakening under extreme loading conditions.

The dynamic stress–strain response of CFRP-confined high-water material exhibits a distinct three-stage pattern, which differs from that observed in unconfined specimens. As shown in Figure 7, the curve initially features a nearly linear rise, corresponding to the elastic deformation of the material—this is the initial rising stage. It is followed by a plateau or slight decline, indicating the onset of matrix yielding and the progressive development of internal microcracks—this is referred to as the descending stage. As the strain approaches the peak strength of the matrix, some curves exhibit a secondary increase in stress, forming a secondary rising stage.

This phenomenon arises because, once the high-water matrix begins to crack and expand under impact loading, the outer CFRP layer becomes progressively activated and provides lateral confinement, which enhances the load-bearing capacity. This delayed confinement effect leads to a secondary stress rise beyond the matrix peak, continuing until the CFRP jacket either ruptures or the impact loading concludes.

A similar three-stage dynamic response was also reported by Yang et al. [24] in SHPB tests on AFRP-confined concrete. They attributed the secondary rise to the confinement effect of the fiber jacket that becomes active after yielding of the core material.

In the present study, since the specimens were not fully destroyed by a single impact, the stress began to recover after the peak and then gradually declined—this reflects the unloading process after loading termination, as seen at the end of the curves in Figure 6.

It is worth noting that the magnitude of the secondary rising stage depends on the matrix strength and the number of CFRP layers. Stronger matrices or thicker confinement layers tend to delay severe damage at the peak, and the CFRP confinement may not be fully activated, resulting in a less prominent secondary rise. In contrast, for weaker matrices or lower confinement levels, noticeable matrix damage occurs near the peak, which triggers a more significant confinement response from the CFRP, leading to a more pronounced secondary rise. For instance, in this study, the specimen with one layer of CFRP showed a clearer stress rebound after the peak, while the specimen with three layers exhibited a less noticeable secondary rise.

3.2.2. Stress–Strain Curves at Different Impact Velocities

As shown in Figure 8, under a water–cement ratio of 1.50 and three-layer CFRP confinement, the stress–strain curves at different bullet velocities exhibit a trend similar to that of the unconfined specimens: the peak stress initially increases and then decreases with increasing impact velocity. At lower velocities, the specimens show relatively low peak stresses, indicating that the impact resistance potential of the material has not been fully activated. As the impact velocity increases to a moderate level, both the peak stress and strain capacity increase significantly, suggesting that the material strength is better mobilized under the combined effects of inertia and lateral confinement.

However, when the bullet velocity increases further, the peak stress shows a slight decline, despite continued increases in strain rate and total strain. This may be attributed to the rapid compaction of internal pores within the high-water material during instantaneous loading, which weakens the load-bearing capacity under a single high-speed impact.

3.2.3. Stress–Strain Curves Under Different Water–Cement Ratios

As shown in Figure 9, under constant impact velocity and confinement conditions, increasing the water–cement ratio from 1.25 to 1.75 results in a decrease in peak stress from over 30 MPa to approximately 24 MPa, while the peak strain increases from around 2.5% to nearly 4%. A higher water–cement ratio leads to weaker dynamic load-bearing capacity and more pronounced deformation. In contrast, specimens with lower water–cement ratios have higher solid content and more complete hydration, forming a denser internal structure. These specimens maintain higher stress levels throughout the loading process, with significantly higher peak strength than those with higher ratios.

CFRP confinement mitigates the adverse effect of increasing water–cement ratio on peak strength. Compared to unconfined specimens, the reduction in strength caused by increasing the water–cement ratio from 1.25 to 1.75 is less significant under confinement. This indicates that external confinement can partially offset the strength loss caused by increased porosity in high water–cement ratio materials, thereby enhancing the effective utilization of the internal structure.

3.2.4. Influence of Inertial Effect Under High Strain Rates

Under high-strain-rate conditions, the inertial effect plays a significant role in the dynamic behavior of high-water material. It was observed in the experiments that at high strain rates, the inertial effect becomes more pronounced, limiting the material’s ability to redistribute stress, which leads to localized stress concentrations. These concentrations promote rapid crack initiation and propagation, particularly in the unconfined high-water material specimens, causing the failure mode to shift from ductile to brittle, with cracks propagating almost instantaneously. The CFRP-confined specimens, on the other hand, show delayed activation of lateral confinement, providing additional resistance during crack propagation, thereby altering the failure mode. The inertial effect also exacerbates stress concentration, especially under impact loading, where the stress in localized areas is significantly higher than in other regions, making microcracks more likely to form in the stress-concentrated areas and merge into larger cracks, ultimately accelerating material failure. Meanwhile, as the strain rate increases, the amplitude of the reflected wave also increases, leading to higher localized stress within the material, further accelerating crack formation and propagation. The inertial effect modulates the interaction between the incident and reflected waves, particularly in areas where stress has already concentrated, thereby accelerating the failure process.

3.2.5. Energy Absorption and Damage Mechanism

The energy absorbed by the material during dynamic loading can be calculated using the stress–strain integral:

$$\mathrm{A}\mathrm{b}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{b}\mathrm{e}\mathrm{d}\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}={\int }_0^{{\mathsf{ϵ}}_{\mathrm{m}\mathrm{a}\mathrm{x}}}\mathsf{\sigma }(\mathsf{ϵ})\mathrm{d}\mathsf{ϵ}$$

(6)

where $\mathsf{\sigma }\left(\mathsf{ϵ}\right)$ represents the stress as a function of strain, and ${\mathsf{ϵ}}_{\mathrm{m}\mathrm{a}\mathrm{x}}$ is the maximum strain at failure. This integral captures the total energy absorbed by the material as it deforms under loading.

The energy absorption is closely related to the material’s damage mechanism. As the material undergoes deformation, part of the energy is absorbed in the form of plastic deformation, and part is used to initiate and propagate cracks. In unconfined high-water material specimens, a high absorbed energy corresponds to significant crack propagation, shifting the failure mode from ductile to brittle. In contrast, CFRP-confined specimens exhibit a more controlled energy absorption process, with less crack propagation due to the lateral confinement provided by the CFRP. This results in a higher peak stress and a more gradual failure process. The observed energy absorption values provide insight into the extent of material damage, with higher energy absorption typically indicating more extensive damage and a higher likelihood of catastrophic failure.

3.3. Microstructural Analysis

Figure 10 displays representative SEM images of the fractured microstructures of high-water material under different strain rates, water–cement ratios, and confinement conditions. Based on the SEM observations, the internal structures of the specimens were compared in terms of hydration product morphology and pore characteristics.

The main hydration products of high-water material include calcium silicate hydrate (C–S–H) gel and ettringite crystals (AFt). Ettringite (i.e., calcium sulfoaluminate hydrate) exhibits distinct morphological forms under different loading conditions. According to the classification proposed by Liu et al. [24], four typical ettringite morphologies were identified in this study, as shown in Figure 11:

Type I: loose, fibrous networks;

Type II: dense fibrous structures;

Type III: short, thick rod-like crystals;

Type IV: compact granular agglomerates.

As shown in Figure 10, the unconfined specimens primarily exhibited loose fibrous networks (Type I), suggesting sparse hydration product distribution and high porosity. In contrast, CFRP-confined specimens contained a greater number of short rod-like crystals and compact granular clusters (Types III and IV), along with denser fibrous structures (Type II). This indicates that lateral confinement induces a triaxial stress state during impact, causing compaction of the hydration products. As a result, fibrous ettringite undergoes morphological transformation into denser rod-like or granular forms. The confined specimens clearly exhibited more compact microstructures with reduced porosity, whereas the unconfined specimens showed loose, highly porous structures. These observations are consistent with findings from static triaxial compression tests, in which needle-like hydration crystals tend to agglomerate and rearrange under high confining pressures, transitioning from fibrous bundles to short rods or blocky structures.

With increasing confinement levels, the morphological transformation of ettringite becomes more pronounced. In specimens with three layers of CFRP confinement, abundant rod-like crystals and dense particle clusters (Types III and IV) were observed, whereas specimens with only one layer primarily featured compacted fibrous networks (Type II). This suggests that higher confinement strength promotes compaction and transformation of hydration products during high-rate impact, significantly increasing microstructural density.

The influence of water–cement ratio on microstructure is also evident in Figure 10. Regardless of confinement condition, specimens with lower water–cement ratios exhibited more continuous and dense hydration product distributions with fewer pores. In contrast, specimens with a ratio of 1.75 showed fragmented, discontinuous microstructures and a substantial increase in visible porosity. These trends align with the macroscopic mechanical results: higher water–cement ratios imply more free water and less solid content, leading to diluted hydration and insufficient filling of voids, which results in a looser hardened structure and reduced strength. Conversely, lower ratios promote greater formation of C–S–H and ettringite, which fill the pores and create a dense matrix, thus improving strength. Therefore, reducing the water–cement ratio is an effective strategy to enhance both the microstructure and macroscopic performance of high-water material.

Under varying strain rates, the microstructural evolution of high-water material differs significantly depending on whether lateral confinement is applied. In the unconfined state, higher bullet velocities result in more extensive and larger pores and more fragmented microstructures. This is primarily due to severe radial expansion and rapid expulsion of free water under high-speed impact, which damages the original hydration framework and generates new cracks and voids. As strain rate increases, residual internal porosity increases, and the fracture surfaces become rougher and more porous. Similar findings were reported by Cao et al. [31], who used quantitative image analysis to show that, as the average strain rate increased from 46 s⁻¹ to 96 s⁻¹, the average pore diameter of CTC materials increased from 6.535 μm to 31.725 μm, and the fractal dimension DDD of the pore area showed a strong positive correlation with strain rate. These results suggest that under high strain rates, the pore structure becomes more complex and widespread, corresponding to a higher degree of macroscopic fragmentation.

In contrast, CFRP-confined specimens exhibit entirely different microstructural features under impact. Due to restricted lateral expansion, a triaxial stress state forms during high-rate loading, compressing and closing internal pores. Hydration products are realigned and compacted under directed stress, preventing the formation of new voids and reducing the size of existing ones. This is clearly visible in Figure 10: unconfined specimens under high-speed impact are filled with microvoids, while their confined counterparts appear dense with no significant large pores. These observations confirm that under impact loading, lateral confinement plays a dominant role in regulating microstructural evolution. Confinement effectively suppresses internal damage propagation and helps maintain the integrity of the internal skeleton, even at high strain rates. Thus, the macroscopic strength and deformability improvements offered by CFRP confinement are largely attributed to its role in restricting lateral expansion and promoting denser hydration product accumulation.

4. Conclusions

This study investigated the dynamic mechanical behavior and microstructural evolution of high-water material confined with CFRP (Carbon Fiber Reinforced Polymer) fabrics under impact loading, using a Split Hopkinson Pressure Bar (SHPB) apparatus. The effects of water–binder ratio, confinement layers, and impact velocity (strain rate) were systematically examined. The main conclusions are as follows:

(1)

Under unconfined conditions, the peak stress of high-water material shows a strain rate sensitivity: it first increases and then decreases with rising strain rate. Excessively high strain rates lead to rapid pulverization before internal stress redistribution, causing strength loss.

(2)

The water–cement ratio has a significant influence on the dynamic compressive strength. A lower ratio (e.g., 1.25) results in denser hydration products, higher peak stress, and greater absorbed energy, whereas a higher ratio (e.g., 1.75) produces porous structures, reducing strength and energy dissipation capacity.

(3)

The fractal dimension analysis of fracture fragments shows that both higher strain rate and higher water–cement ratio increase fragmentation degree, shifting the failure mode from block-like fracture to pulverized failure.

(4)

CFRP confinement effectively enhances the impact resistance of high-water material. Increasing the number of CFRP layers raises peak stress and peak strain, restrains lateral expansion, and changes the failure mode from severe fragmentation to localized cracking.

(5)

The dynamic stress–strain response of CFRP-confined high-water specimens exhibits a characteristic three-stage pattern: an initial rising stage, a descending stage, and a secondary rising stage. The secondary rise is attributed to the delayed activation of CFRP confinement after matrix cracking.

(6)

At a constant confinement condition, specimens with lower water–cement ratios maintain higher peak stress and denser structures, while those with higher ratios show reduced strength but larger deformation. CFRP confinement partially offsets the strength loss caused by higher water–cement ratios.

(7)

SEM analysis reveals that CFRP confinement transforms ettringite morphologies from loose fibrous forms into denser rod-like and granular structures, reducing porosity and enhancing compactness. In contrast, unconfined specimens subjected to high strain rates exhibit severe pore development and disrupted hydration frameworks.