Limestone samples from a reef on an island in Western Pacific Ocean waters were synthetically pulverized into particles, each not exceeding 16 mm in size. Particles measuring 4.75–16 mm in size were processed as coarse aggreagte, whereas those ranging from 0.15 to 4.75 mm were selected for fine aggreagte (Fan et al., 2023; Liu et al., 2018). Figure 1 elaborates on the ongoing categorization of both coarse and fine aggregates of coral reef limestone.

Contrasting with traditional land-based sand and gravel, fragments of reef limestone display features like suboptimal particle form, uneven particle shapes, coarse surfaces, significant porosity, extensive specific surface area, and reduced strength (Luo et al., 2023). Therefore, more cement paste is required to enhance workability when mixing concrete, making it less suitable for high-strength concrete applications. This research involved blending both coarse and fine reef limestone aggregates with standard Portland cement, seawater, fly ash, slag, and various additives, as detailed in Table 1, to create concrete with a C40 strength. Notably, the aggregates of reef limestone underwent water immersion before being mixed to eliminate detrimental ions and contaminants. After processes such as mixing, vibrating, and curing, cylindrical specimens with height-to-diameter ratios of 1:2 (φ50 mm × 25 mm) and 2:1 (φ50 mm × 100 mm) were prepared.

Figure 2 illustrates the entire process of the dry-wet carbonation cycle that the reef limestone concrete samples underwent. Before the experiments, the samples were subjected to a drying pretreatment (assessed by changes in mass to determine whether they reached a dry state). In order to ensure the rationality of the dry-wet carbonation cycle pretreatment scheme, the existing research results are referred to Wang et al. (2018); Jiang et al. (2017); Zhang and Shao (2016). First, the dried samples were placed in a vacuum saturation container filled with seawater (soaked in a vacuum state for 12 h). Subsequently, the samples, once saturated, underwent a drying phase in an oven maintained at a steady 60°C for 12 h. Ultimately, the specimens were positioned in a carbonation chamber to undergo carbonation treatment, characterized by a carbon dioxide level of 20% ± 3%, a humidity of 70% ± 5%, a temperature maintained at 20°C ± 5°C, and a 24-h period. The trio of stages forms a single cycle, with the count of dry-wet carbonation cycles adjusted to 0, 10, 20, 30, 40, and 50 cycles.

Micro-fissures in the reef limestone concrete samples will progressively widen and enlarge due to dry-wet carbonation cycles, resulting in the unavoidable seepage of internal elements. Initially, the specimens have relatively high density and stronger resistance to dry-wet cycles and carbonation. However, with the rise in cycle count, there will be a noticeable alteration in the ultimate weight of the samples. Consequently, the rate at which mass is lost serves as an indicator of the extent of damage and degradation. The method and procedure for calculating the mass loss rate are as follows:

1) Dry the specimens after completing the dry-wet carbonation cycles, and wrap them in plastic wrap to prevent moisture from causing errors, then weigh each specimen three times and take the average value as its initial mass; 2) Place the specimens into the corresponding equipment for the deterioration test according to the experimental protocol, and after undergoing the specified number of deterioration tests, remove the specimens, dry them, and weigh them; 3) Utilize the following formula to determine the rate at which the specimens lost mass following the nth deterioration test. Formula 1 is as follows:

In the equation, Δm_n represents the mass loss rate of the specimen (%); m₀ denotes the initial mass (g); and m_n represents the mass after undergoing n deterioration actions (g). Furthermore, the E_rd relative dynamic modulus of elasticity serves as a crucial measure for evaluating the specimens’ damage and degradation. This refers to the proportion between the specimen’s dynamic modulus of elasticity post the nth cycle and its initial dynamic modulus of elasticity. To prevent secondary damage during the testing process from causing experimental errors, the elastic modulus is calculated using parameters such as sound travel time and wave speed while ultrasonic waves propagate through the specimen. This method is completely non-destructive to the specimen, as shown in Equations 2, 3. E rd = E dn E d 0 = V n 2 V 0 2 (2) E d = 1 + v 1 − 2 v ρ V 2 1 − v (3)

In the equation, E_d represents the dynamic modulus of elasticity of the specimen (GPa); v represents the Poisson’s ratio of the specimen; ρ represents the density of the specimen (kg/m³); V is the ultrasonic wave speed (m/s); E_d0 is the initial dynamic modulus of elasticity of the specimen (GPa), and E_dn is the dynamic modulus of elasticity after undergoing n deterioration actions (GPa).

Using a divided Hopkinson pressure bar, dynamic compression experiments were performed on reef limestone concrete samples that had experienced cycles of dry and wet carbonation. The experimental setup is shown in Figure 3. The incident bar and the transmission bar are both 2,800 mm, the bar diameter is 50 mm, the elastic modulus is 210 GPa, and the density is 7,800 kg/m³. Introducing high-pressure gas into the emission chamber propels the projectile at a specific speed, hitting the incident bar’s end surface and creating a longitudinal compressive wave. Upon the arrival of this compressive wave at the specimen, a sequence of transmission and reflection events take place at the terminal surfaces of both the incident and transmission bars. Strain gauges affixed to both the incident and transmission bars record the incident strain ε_i, the reflected strain ε_r, and the transmitted strain ε_t. Ultimately, by presuming one-dimensional stress waves and uniformity, one can determine the specimen’s stress-strain and loading rate data.

As per the principles of one-dimensional stress wave theory, the stress and strain of the specimen can be determined by the following Formula 4 (Dai et al., 2010): σ t = ε i t + ε r t + ε t t E 0 A 2 A s ε ˙ t = ε i t − ε r t − ε t t C 0 L s ε t = ∫ 0 t ε ˙ t d t (4)

Within the formula, E₀, C₀, and A symbolize the velocity of the elastic wave within the bar, the elastic modulus, and the bar’s cross-sectional area, in that order; A_S and L_S denote the specimen’s cross-sectional area and length, respectively; and t denotes the stress wave’s length.

In addition, the calculation Formula 5 of the incident energy W_I, reflected energy W_R and transmitted energy W_T on the pressure rod during the whole process from the beginning of loading to the failure of the sample is as follows (Huang et al., 2014): W I = C 0 A E 0 ∫ σ I 2 t d t , W R = C 0 A E 0 ∫ σ R 2 t d t , W T = C 0 A E 0 ∫ σ T 2 t d t (5)

Given that the energy dissipation of the stress wave in both the bar and the specimen is deemed minimal, the aggregate energy the specimen absorbs and dissipates, denoted as W_L, is derivable from the subsequent Equation 6. W L = W I − W R + W T (6)

The primary cause of carbonation in reef limestone concrete lies in CO₂ penetrating its core, leading to a decrease in its alkalinity. Samples subjected to varying degrees of deterioration were cut and sprayed with a 1% phenolphthalein ethanol solution. The color reaction of the phenolphthalein ethanol solution is shown in Figure 4. After different numbers of deterioration cycles, the degree of carbonation represented by the phenolphthalein reagent varied. The undeteriorated samples showed a strong magenta color after treatment with phenolphthalein, while the degree of color change gradually weakened with an increasing number of deterioration cycles.

It is noteworthy that the reef limestone aggregate was also stained by the phenolphthalein reagent. This is because, after the aggregate underwent pre-absorption of water, more cement paste entered its interior during the mixing process, which increased the alkalinity of the aggregate. As the number of deterioration cycles increased, the cement content in the pores of the aggregate gradually decreased, resulting in a gradual reduction in the staining intensity under the phenolphthalein reagent. Further testing showed that the carbonation depth of reef limestone concrete at different deterioration levels increased by 1.5%, 1.58%, 2.87%, and 5.6% compared to the standard condition. Among them, the sample of reef limestone concrete with the deepest deterioration had a carbonation depth increase of 5.2 mm compared to the standard condition. He et al. research shows that cyclic dry-wet carbonation can lead to changes in the pore structure of concrete and cause the reorganization of hydration products, affecting its structural strength. For example, some hydration products (such as C-S-H gel) may convert into lower-strength calcium carbonate, thereby weakening the overall strength and reducing its durability (He et al., 2022).

Figure 5 summarizes the mass loss rates and relative elastic moduli of reef limestone concrete specimens subjected to different numbers of deterioration cycles. When the mass loss rate reaches over 5%, the specimen can be deemed damaged. For cylindrical specimens (φ50 mm × 100 mm), the mass loss rate did not exceed 5% even after more than 40 deterioration cycles, while the disc specimens (φ50 mm × 25 mm) exhibited a high mass loss rate, indicating that the dry-wet carbonation cycles caused significant damage to the disc samples. Overall, although the shapes of the specimens differ, both types experienced an increase in mass loss rates with the growing number of deterioration cycles, although the trends in their growth were distinct. Furthermore, the relative dynamic elastic modulus displayed a significant size effect influenced by the number of dry-wet carbonation cycles. With the ongoing rise in deterioration cycles, there was an acceleration in the decrease rate of the cylindrical specimens’ relative dynamic elastic modulus (φ50 mm × 100 mm), as evidenced by the reduced acoustic wave velocity. This suggests a swift escalation and ongoing spread of cracks within the specimens, thereby intensifying their damage severity. Conversely, the disc specimens’ dynamic elastic modulus (φ50 mm × 25 mm) varied in tandem with the escalation of deterioration cycles. This difference reflects the varying internal structural responses of different specimens during the deterioration process; the cylindrical specimens may exhibit better resistance to deterioration due to their structural characteristics compared to other types of specimens.

Furthermore, quasi-static compression and tension tests were conducted on reef limestone concrete samples subjected to varying degrees of deterioration, and the statistical results are shown in Figure 6. Despite variations in the data from uniaxial compression and tension tests, it's noticeable that the specimens’ uniaxial compressive and tensile strengths exhibit a general downward trajectory after varying levels of degradation, with a direct relationship between the rise in peak strain and the frequency of these deterioration cycles.

Referring to Deng et al. (2020a); Deng et al. (2020b) research results, the ratio of strength is used as an increasing coefficient. The strength reduction coefficient is defined as the strength reduction coefficient is the ratio of uniaxial compressive strength of the sample after dry-wet carbonization cycle to that of the initial sample, to evaluate the degree of deterioration of reef limestone concrete specimens. When the number of deterioration cycles is less than 20, the strength attenuation is not significant, with the average compressive strength remaining above 40 MPa. However, once the count of degradation cycles surpasses 20, the reduction in strength becomes more noticeable, evidenced by the average compressive strength falling from 32.53 to 25.55 MPa. When samples undergo 40 cycles of decay, their uniaxial compressive strength drops to merely 60% of their original level, and upon reaching 50 cycles, the mean compressive strength drops to just 21.32 MPa. The maximum average tensile strength of the specimens is 6.31 MPa, but after 50 cycles of deterioration, the maximum tensile strength decreases to 3.31 MPa, resulting in an overall tensile strength reduction of 49%. The reduction in tensile strength is lower than that of compressive strength.