Experimental Investigation on the Impact Resistance of Carbon Fibers Reinforced Coral Concrete

In this study, the impact resistance of coral concrete with different carbon fiber (CF) dosages subjected to drop-weight impact test was investigated. For this purpose, three concrete strength grades (C20, C30, C40) and six CF dosages (0.0%, 0.3%, 0.6%, 1.0%, 1.5%, and 2.0% by weight of the binder) were considered, and a total of 18 groups of carbon fibers reinforced coral concrete (CFRCC) were cast. For each group, eight specimens were tested following the drop-weight impact test suggested by CECS 13. Then, the two-parameter Weibull distribution theory was adopted to statistically analyze the variations in experimental results. The results indicated that the addition of CFs could transform the failure pattern from obvious brittleness to relatively good ductility and improve the impact resistance of coral concrete. Moreover, the impact resistance of CFRCC increases with the CF dosage increasing. The statistical analysis showed that the probability distribution of the blow numbers at the initial crack and final failure of CFRCC approximately follows the two-parameter Weibull distribution.


Introduction
The ocean is an essential space for the sustainable development of whole humans due to its abundant resources [1]. Recently, with the rapid development of society, the development and utilization of marine resources and the development of marine industry have received extensive attention [2,3]. Therefore, there are more and more island construction projects, which have led to a significant increase in demand for marine concrete [4,5]. In addition, the utilization of locally available resources on islands as materials to mix concrete has essential practical significance because it can solve the shortage of construction materials problem, shorten the construction period, and reduce costs for distant island reef construction projects [1,6,7].

Mix Proportions and Specimen Preparation
The designed strength grades of CFRCC without CF addition were C20, C30, and C40, respectively. The basic mix proportions were designed according to JGJ 51 [40] and presented in Table 3. The CFs dosage were 0.0%, 0.3%, 0.6%, 1.0%, 1.5%, 2.0% by weight of the binder (cement). The usage of HPMC and P803 was 0.4% and 0.15% by weight of the binder, respectively. Some studies [13,41,42] pointed out that preparation of coral concrete with pre-wetted coral coarse aggregates is beneficial for improving the compressive strength, improving the workability, reducing the self-shrinkage and dry shrinkage of coral concrete. Thus, the coral coarse aggregates have been pre-wetted before mixing. The procedure of mixing CRFCC is illustrated in Figure 4. After the uniform mixture was obtained, the stirred mixture was cast in molds and vibrated for about 30 s on a vibration table. For each mixture, three 100 mm × 100 mm × 100 mm cubes and two 150 mm × 300 mm cylinders were cast. All cast specimens were cured at room temperature for 24 h, then demolded and cured in a marine environment curing cabinet for 28 d. The marine environment curing cabinet has a seawater spray device and several related sensors that can simulate the humidity of the real marine environment. Then, cube specimens conducted the cube compressive strength test following the code GB/T50081 [43] (similar to the code of ASTM C33, but the specimen used is cube specimen instead of cylindrical specimen) to obtain the cube compressive strength of each CFRCC mixtures (see Table 4), and each cylindrical specimen was cut into four discs of 150 mm × (63 ± 2) mm for the drop-weight impact test. usage of HPMC and P803 was 0.4% and 0.15% by weight of the binder, respectively. Some studies [13,41,42] pointed out that preparation of coral concrete with pre-wetted coral coarse aggregates is beneficial for improving the compressive strength, improving the workability, reducing the selfshrinkage and dry shrinkage of coral concrete. Thus, the coral coarse aggregates have been prewetted before mixing. The procedure of mixing CRFCC is illustrated in Figure 4. After the uniform  mixture was obtained, the stirred mixture was cast in molds and vibrated for about 30 s on a vibration  table. For each mixture, three 100 mm × 100 mm × 100 mm cubes and two 150 mm × 300 mm cylinders were cast. All cast specimens were cured at room temperature for 24 h, then demolded and cured in a marine environment curing cabinet for 28 d. The marine environment curing cabinet has a seawater spray device and several related sensors that can simulate the humidity of the real marine environment. Then, cube specimens conducted the cube compressive strength test following the code GB/T50081 [43] (similar to the code of ASTM C33, but the specimen used is cube specimen instead of cylindrical specimen) to obtain the cube compressive strength of each CFRCC mixtures (see Table 4), and each cylindrical specimen was cut into four discs of 150 mm × (63 ± 2) mm for the drop-weight impact test.

Impact Tests
The impact test was conducted following the China CECS 13 [34] drop-weight impact test that was modified from the ACI 544 [44] suggested method. The details of the drop-weight impact test setup are illustrated in Figure 5. As shown in Figure 5, a steel hammer with a mass of 4.5 kg drops from a height of 500 mm on a steel ball with a diameter of 63 mm located on the central surface of the disc specimens. The number of blows causing the first visible crack was recorded as the initial crack resistance factor (N 1 ), and the number of blows until the pieces of specimen touching three of the four steel lugs was recorded as the final failure resistance factor (N 2 ). For each mixture, eight discs were tested, and the impact resistance was represented based on the average of eight specimens. The impact energy at initial crack and final failure were calculated by using the following equation: where W i is the impact energy (J); N i is the number of blows; m is the weight of steel hammer with a mass of 4.5 kg; v is the velocity of the steel hammer (m/s); g is the acceleration of gravity (9.81 m/s 2 ); h is the falling height of the steel hammer (500 mm); and i = 1, 2 is representing the initial crack and final failure, respectively.

Impact Tests
The impact test was conducted following the China CECS 13 [34] drop-weight impact test that was modified from the ACI 544 [44] suggested method. The details of the drop-weight impact test setup are illustrated in Figure 5. As shown in Figure 5, a steel hammer with a mass of 4.5 kg drops from a height of 500 mm on a steel ball with a diameter of 63 mm located on the central surface of the disc specimens. The number of blows causing the first visible crack was recorded as the initial crack resistance factor (N1), and the number of blows until the pieces of specimen touching three of the four steel lugs was recorded as the final failure resistance factor (N2). For each mixture, eight discs were tested, and the impact resistance was represented based on the average of eight specimens. The impact energy at initial crack and final failure were calculated by using the following equation: where is the impact energy (J); is the number of blows; m is the weight of steel hammer with a mass of 4.5 kg; is the velocity of the steel hammer (m/s); is the acceleration of gravity (9.81 m/s 2 ); ℎ is the falling height of the steel hammer (500 mm); and i = 1, 2 is representing the initial crack and final failure, respectively.

Failure Patterns under Impact
After drop-weight impact tests, the failure patterns of part of the specimens with and without CFs are shown in Figure 6. As expected, for all the specimens without CFs, when the first visible crack appears, the specimens suddenly broke down into two pieces and showed an obviously brittle failure behavior. For the specimens with CFs at a low level, its failure pattern is similar to the specimens without CFs, but some specimens broke down into three pieces (Figure 6b). For the specimens with CFs at a high level, after the first visible crack appears, the specimen can continue to bear the impact

Failure Patterns under Impact
After drop-weight impact tests, the failure patterns of part of the specimens with and without CFs are shown in Figure 6. As expected, for all the specimens without CFs, when the first visible crack appears, the specimens suddenly broke down into two pieces and showed an obviously brittle failure behavior. For the specimens with CFs at a low level, its failure pattern is similar to the specimens without CFs, but some specimens broke down into three pieces (Figure 6b). For the specimens with CFs at a high level, after the first visible crack appears, the specimen can continue to bear the impact loads, and finally break into two or three or four pieces (Figure 6c,d). It is worth noting that no matter whether the dosage of CFs is high or low, the specimens will eventually be wholly separated into several parts, which is similar to the basalt fibers reinforced concrete [45] but different from the impact failure phenomenon-the specimen still remains intact-of steel fibers reinforced concrete, macro polypropylene fibers reinforced concrete, NiTi-SMA fibers reinforced concrete, and polypropylene fibers reinforced concrete [21,36,46]. The reason is that the diameter of CFs is only 7.3 µm, and the elongation at the break of CFs is no greater than 2.05%; when CFs are added into coral concrete, there are tens of millions of micro CFs that exist in the coral concrete matrix, and almost all the microcracks have micro CFs, which can restrain the microcracks propagation and hence enhance the impact performance of CFRCC at the microcrack stage, but many CFs have been broken or pulled out at macrocracks stage, so the CFs mainly act in the microcrack stage, and have less hindrance effect on large cracks. Moreover, with the strength grade and CF dosage increasing, a more profound impact pit and more debris were observed at the central surface of the specimen when the specimen fails. loads, and finally break into two or three or four pieces (Figure 6c,d). It is worth noting that no matter whether the dosage of CFs is high or low, the specimens will eventually be wholly separated into several parts, which is similar to the basalt fibers reinforced concrete [45] but different from the impact failure phenomenon-the specimen still remains intact-of steel fibers reinforced concrete, macro polypropylene fibers reinforced concrete, NiTi-SMA fibers reinforced concrete, and polypropylene fibers reinforced concrete [21,36,46]. The reason is that the diameter of CFs is only 7.3 μm, and the elongation at the break of CFs is no greater than 2.05%; when CFs are added into coral concrete, there are tens of millions of micro CFs that exist in the coral concrete matrix, and almost all the microcracks have micro CFs, which can restrain the microcracks propagation and hence enhance the impact performance of CFRCC at the microcrack stage, but many CFs have been broken or pulled out at macrocracks stage, so the CFs mainly act in the microcrack stage, and have less hindrance effect on large cracks. Moreover, with the strength grade and CF dosage increasing, a more profound impact pit and more debris were observed at the central surface of the specimen when the specimen fails.  Figure 7 shows the fracture surface of part of the specimens after repeated drop-weight impact tests. It can be seen that, when the concrete strength grade is C20, there is a small amount of coral coarse aggregates broken (see Figure 7a), but almost all coral coarse aggregates broke (see Figure 7c) when the concrete strength grade is raised to C40. It can be concluded that the fracture rate of coral aggregates on the fracture surface increases with the increase of concrete strength grade. This phenomenon can be attributed to the relatively low strength of the coral coarse aggregates and the excellent bonding properties between the coral coarse aggregates and the cement matrix due to the rough surface morphology of the coral coarse aggregates [47].  Table 5 summarizes the drop-weight impact test results for all the CFRCC mixtures (the detailed results of each specimen see Appendix Table A1) where an increase in the number of post-first crack blow (INPB) is introduced, and the INPB is calculated as follows:

Effect of CFs on the Impact Resistance
where N1 and N2 are representing the number of blows at initial crack and final failure, respectively.  Figure 7 shows the fracture surface of part of the specimens after repeated drop-weight impact tests. It can be seen that, when the concrete strength grade is C20, there is a small amount of coral coarse aggregates broken (see Figure 7a), but almost all coral coarse aggregates broke (see Figure 7c) when the concrete strength grade is raised to C40. It can be concluded that the fracture rate of coral aggregates on the fracture surface increases with the increase of concrete strength grade. This phenomenon can be attributed to the relatively low strength of the coral coarse aggregates and the excellent bonding properties between the coral coarse aggregates and the cement matrix due to the rough surface morphology of the coral coarse aggregates [47]. loads, and finally break into two or three or four pieces (Figure 6c,d). It is worth noting that no matter whether the dosage of CFs is high or low, the specimens will eventually be wholly separated into several parts, which is similar to the basalt fibers reinforced concrete [45] but different from the impact failure phenomenon-the specimen still remains intact-of steel fibers reinforced concrete, macro polypropylene fibers reinforced concrete, NiTi-SMA fibers reinforced concrete, and polypropylene fibers reinforced concrete [21,36,46]. The reason is that the diameter of CFs is only 7.3 μm, and the elongation at the break of CFs is no greater than 2.05%; when CFs are added into coral concrete, there are tens of millions of micro CFs that exist in the coral concrete matrix, and almost all the microcracks have micro CFs, which can restrain the microcracks propagation and hence enhance the impact performance of CFRCC at the microcrack stage, but many CFs have been broken or pulled out at macrocracks stage, so the CFs mainly act in the microcrack stage, and have less hindrance effect on large cracks. Moreover, with the strength grade and CF dosage increasing, a more profound impact pit and more debris were observed at the central surface of the specimen when the specimen fails.  Figure 7 shows the fracture surface of part of the specimens after repeated drop-weight impact tests. It can be seen that, when the concrete strength grade is C20, there is a small amount of coral coarse aggregates broken (see Figure 7a), but almost all coral coarse aggregates broke (see Figure 7c) when the concrete strength grade is raised to C40. It can be concluded that the fracture rate of coral aggregates on the fracture surface increases with the increase of concrete strength grade. This phenomenon can be attributed to the relatively low strength of the coral coarse aggregates and the excellent bonding properties between the coral coarse aggregates and the cement matrix due to the rough surface morphology of the coral coarse aggregates [47].   Table A1) where an increase in the number of post-first crack blow (INPB) is introduced, and the INPB is calculated as follows:

Effect of CFs on the Impact Resistance
where N1 and N2 are representing the number of blows at initial crack and final failure, respectively.

Effect of CFs on the Impact Resistance
where N 1 and N 2 are representing the number of blows at initial crack and final failure, respectively. For the specimens of CC20C00, the first crack impact energy (W 1 ) equals the failure impact energy (W 2 ). For the specimens of CC30C00 and CC40C00, the failure impact energy (W 2 ) is only 2 J and 6 J more than the first crack impact energy (W 1 ). That is to say, when the first visible crack appears, the final failure of the specimen will occur at the same time, and the specimens without CFs show distinct brittle behavior. Figure 8 shows the effect of CFs dosage on the impact energy at first crack (W 1 ) and final failure (W 2 ) of CFRCC of three strength grades. It is easily found from the Figure 8 that adding CFs in coral concrete can improve the first impact energy and the final failure impact energy, and further improvement was recorded for the final failure impact energy, as compared to the first impact energy. With the increasing of additional CFs in coral concrete, the increase percentage of W 1 and W 2 is also increasing. In other words, the addition of CFs in coral concrete can improve both the initial crack and ultimate failure impact resistances of CFRCC, and its improvement increases with the increase of CF dosage.  For the specimens of CC20C00, the first crack impact energy (W1) equals the failure impact energy (W2). For the specimens of CC30C00 and CC40C00, the failure impact energy (W2) is only 2 J and 6 J more than the first crack impact energy (W1). That is to say, when the first visible crack appears, the final failure of the specimen will occur at the same time, and the specimens without CFs show distinct brittle behavior. Figure 8 shows the effect of CFs dosage on the impact energy at first crack (W1) and final failure (W2) of CFRCC of three strength grades. It is easily found from the Figure 8 that adding CFs in coral concrete can improve the first impact energy and the final failure impact energy, and further improvement was recorded for the final failure impact energy, as compared to the first impact energy. With the increasing of additional CFs in coral concrete, the increase percentage of W1 and W2 is also increasing. In other words, the addition of CFs in coral concrete can improve both the initial crack and ultimate failure impact resistances of CFRCC, and its improvement increases with the increase of CF dosage.    and 6.1 for C20, C30, and C40, respectively, and the INPB/N 1 for all the mixture is no more than 14%. Mastali et al. [27] conducted the drop-weight impact test on CF reinforced self-compacting concrete and obtained similar results. That is to say, the improvement effect of CFs on the impact resistance of specimens after cracking is not apparent, which is obviously different from the test results of steel fibers reinforced concrete and macro PP fibers reinforced concrete obtained by Zhang, Rahmani, Ding, and Murali et al. [36,[48][49][50]. The explanation for this is that the steel fibers are macro fibers (the diameter is generally higher than 0.4 mm) and have a relatively large elongation at break (more than 3.5%), so the steel fibers can play an excellent bridging role in macrocracks after the first visible crack appeared of specimens. However, the CFs have a diameter of only 7.3 µm and an elongation at break of only 2.05%, so the CFs mainly play a positive role in microcracks and a less positive role in macrocracks under the drop-weight impact test. The previous research data [26,50,51] also clearly indicated that, in the drop-weight impact test, the larger the diameter of fiber and elongation at break is, the larger the INPB will be, when other conditions are the same.  [36,[48][49][50]. The explanation for this is that the steel fibers are macro fibers (the diameter is generally higher than 0.4 mm) and have a relatively large elongation at break (more than 3.5%), so the steel fibers can play an excellent bridging role in macrocracks after the first visible crack appeared of specimens. However, the CFs have a diameter of only 7.3 μm and an elongation at break of only 2.05%, so the CFs mainly play a positive role in microcracks and a less positive role in macrocracks under the drop-weight impact test. The previous research data [26,50,51] also clearly indicated that, in the drop-weight impact test, the larger the diameter of fiber and elongation at break is, the larger the INPB will be, when other conditions are the same.

Effect of Concrete Strength Grade on the Impact Resistance
As shown in Figure 10, the impact energy at first crack (W1) and final failure (W2) and strength grade is approximately in a linear relationship, which indicates that, for CFRCC, the higher the concrete strength grade is, the higher the impact resistance will be. For polypropylene fibers reinforced coral concrete, Wang et al. [21] also reached a similar conclusion.

Correlation between Cube Compressive Strength, CFs Dosage, and Impact Energy
After regression analysis, it is found that the effect of CFs dosage and cube compressive strength on the impact resistance of CFRCC can be illustrated by Equation (3): where W1 and W2 are the impact energy at the first visible crack and final failure, respectively (J); fcu is the cube compressive strength (MPa); ρc is the CFs dosage (%); a, b, c, and d are fitting parameters.

Effect of Concrete Strength Grade on the Impact Resistance
As shown in Figure 10, the impact energy at first crack (W 1 ) and final failure (W 2 ) and strength grade is approximately in a linear relationship, which indicates that, for CFRCC, the higher the concrete strength grade is, the higher the impact resistance will be. For polypropylene fibers reinforced coral concrete, Wang et al. [21] also reached a similar conclusion.  [36,[48][49][50]. The explanation for this is that the steel fibers are macro fibers (the diameter is generally higher than 0.4 mm) and have a relatively large elongation at break (more than 3.5%), so the steel fibers can play an excellent bridging role in macrocracks after the first visible crack appeared of specimens. However, the CFs have a diameter of only 7.3 μm and an elongation at break of only 2.05%, so the CFs mainly play a positive role in microcracks and a less positive role in macrocracks under the drop-weight impact test. The previous research data [26,50,51] also clearly indicated that, in the drop-weight impact test, the larger the diameter of fiber and elongation at break is, the larger the INPB will be, when other conditions are the same.

Effect of Concrete Strength Grade on the Impact Resistance
As shown in Figure 10, the impact energy at first crack (W1) and final failure (W2) and strength grade is approximately in a linear relationship, which indicates that, for CFRCC, the higher the concrete strength grade is, the higher the impact resistance will be. For polypropylene fibers reinforced coral concrete, Wang et al. [21] also reached a similar conclusion.

Correlation between Cube Compressive Strength, CFs Dosage, and Impact Energy
After regression analysis, it is found that the effect of CFs dosage and cube compressive strength on the impact resistance of CFRCC can be illustrated by Equation (3): where W1 and W2 are the impact energy at the first visible crack and final failure, respectively (J); fcu is the cube compressive strength (MPa); ρc is the CFs dosage (%); a, b, c, and d are fitting parameters.

Correlation between Cube Compressive Strength, CFs Dosage, and Impact Energy
After regression analysis, it is found that the effect of CFs dosage and cube compressive strength on the impact resistance of CFRCC can be illustrated by Equation (3): where W 1 and W 2 are the impact energy at the first visible crack and final failure, respectively (J); f cu is the cube compressive strength (MPa); ρ c is the CFs dosage (%); a, b, c, and d are fitting parameters. The fitting results of Equation (3) to test data are presented in Table 6, Figure 11, and Figure 12. It can be seen that the standardized residuals of most of the points are in the range of −2 to 2, and the Adjusted R 2 are 0.995 and 0.996 for W 1 and W 2 , respectively, which indicates that Equation (3) fits the experimental data well. Moreover, Figures 11c and 12c also indicate that the fitting values are very close to the experimental values. The fitting results of Equation (3) to test data are presented in Table 6, Figure 11, and Figure 12. It can be seen that the standardized residuals of most of the points are in the range of −2 to 2, and the Adjusted R 2 are 0.995 and 0.996 for W1 and W2, respectively, which indicates that Equation (3) fits the experimental data well. Moreover, Figure 11c and Figure 12c also indicate that the fitting values are very close to the experimental values.

Distribution of Impact Resistance Factors
Over the past few decades, several statistical models have been employed for analysis of the variations in impact test results of concrete [36,46,49,50,[52][53][54][55][56]. Among them, the normal distribution model is widely used. However, many researchers [54,56] pointed out that the impact test results exhibited poor fitness with normal distribution at a 95% confidence level. By contrast, the twoparameter Weibull distribution has been proved by some researchers [36,46,50] that it is appropriate to evaluate the impact performance of concrete under impact. Therefore, for analyzing the variations in the impact resistance of CFRCC under drop-weight impact test, the two-parameter Weibull distribution is employed in this study.
According to [46], the expression of the cumulative distribution function F(x) of two-parameter Weibull probability law is as follows: The fitting results of Equation (3) to test data are presented in Table 6, Figure 11, and Figure 12. It can be seen that the standardized residuals of most of the points are in the range of −2 to 2, and the Adjusted R 2 are 0.995 and 0.996 for W1 and W2, respectively, which indicates that Equation (3) fits the experimental data well. Moreover, Figure 11c and Figure 12c also indicate that the fitting values are very close to the experimental values.

Distribution of Impact Resistance Factors
Over the past few decades, several statistical models have been employed for analysis of the variations in impact test results of concrete [36,46,49,50,[52][53][54][55][56]. Among them, the normal distribution model is widely used. However, many researchers [54,56] pointed out that the impact test results exhibited poor fitness with normal distribution at a 95% confidence level. By contrast, the twoparameter Weibull distribution has been proved by some researchers [36,46,50] that it is appropriate to evaluate the impact performance of concrete under impact. Therefore, for analyzing the variations in the impact resistance of CFRCC under drop-weight impact test, the two-parameter Weibull distribution is employed in this study.
According to [46], the expression of the cumulative distribution function F(x) of two-parameter Weibull probability law is as follows:

Distribution of Impact Resistance Factors
Over the past few decades, several statistical models have been employed for analysis of the variations in impact test results of concrete [36,46,49,50,[52][53][54][55][56]. Among them, the normal distribution model is widely used. However, many researchers [54,56] pointed out that the impact test results exhibited poor fitness with normal distribution at a 95% confidence level. By contrast, the two-parameter Weibull distribution has been proved by some researchers [36,46,50] that it is appropriate to evaluate the impact performance of concrete under impact. Therefore, for analyzing the variations in the impact resistance of CFRCC under drop-weight impact test, the two-parameter Weibull distribution is employed in this study.
According to [46], the expression of the cumulative distribution function F(x) of two-parameter Weibull probability law is as follows: where x is the impact life of the concrete; k is the shape parameter; λ is the scale parameter; x0 is the minimum impact life of concrete and assumed to be 0 in this study. The function F(x) denotes the failure probability of concrete under impact loading. Thus, the probability estimator L(x) may be defined as: Take x 0 = 0 and the natural logarithm twice on both sides of Equation (5) to get: Thus, Equation (6) can be used to verify whether the impact resistance factors (N 1 , and N 2 ) of CFRCC follow the two-parameter Weibull distribution. Since Equation (6) represents a linear relationship between ln ln (1/L(x)) and ln x, if an appropriately linear relationship between ln ln (1/L(x)) and ln x is observed from the test results, the conclusion that using two-parameter Weibull distribution to characterize the statistical distribution of impact test results of CFRCC is feasible can be conducted. In order to verify whether there is an appropriately linear relationship between ln (1/L(x)) and ln x, first, the impact results (N 1 , and N 2 ) are arranged in an descending order, and then the probability estimator is assumed and the linear regression analysis is performed.
Many probability estimators have been used in previous studies and Murali et al. [50] summarized twenty probability estimators used in previous papers. It can be seen from the summaries of Murali [50] that there are two expression forms of the probability estimator: where j is the sequence number of the impact failure specimen; n is the total number of the impact specimens for each mixture; α and β are constants.
After trial calculating the test results with Equations (7) and (8), Equation (7) is chosen as the recommended probability estimator in this study, and the values of α and β are −0.6, and 0.9, respectively. Figure 13 shows the distribution of the impact resistance factor (N 1 , and N 2 ) of each CFRCC mixture and the corresponding fitted curves, and Table 7 gives the detailed linear regression results. Rahmani et al. [49] pointed out that a R 2 of 0.7 or higher is sufficient for establishing a reasonable reliability model. Since the appropriately linear relationship plot in Figure 13 and all the impact test results have Adjusted R 2 equal to or higher than 0.837, the two-parameter Weibull distribution is considered suitable for establishing the statistical distribution of impact test data of coral concrete incorporating CFs. These developed reliability curves are highly suitable as a useful tool to quickly investigate the impact resistance of CFRCC, thereby eliminating the necessity of time-consuming impact testing process. Some previous studies [36,46,49,50] have drawn similar conclusions for other types of fibers reinforced concrete. reasonable reliability model. Since the appropriately linear relationship plot in Figure 13 and all the impact test results have Adjusted R 2 equal to or higher than 0.837, the two-parameter Weibull distribution is considered suitable for establishing the statistical distribution of impact test data of coral concrete incorporating CFs. These developed reliability curves are highly suitable as a useful tool to quickly investigate the impact resistance of CFRCC, thereby eliminating the necessity of timeconsuming impact testing process. Some previous studies [36,46,49,50] have drawn similar conclusions for other types of fibers reinforced concrete.  According to Equation (5) and (6), the number of blows (N1, N2) of CFRCC at the corresponding failure probability P can be derived as follows: where P is the failure probability. Figure 14 shows the N2 of CFRCC acquired by reliability analysis at different failure probability. It is easy to note that the impact resistance performance of CFRCC increases approximately linearly with the CF dosage increasing at the same failure probability.  According to Equations (5) and (6), the number of blows (N 1 , N 2 ) of CFRCC at the corresponding failure probability P can be derived as follows: where P is the failure probability. Figure 14 shows the N 2 of CFRCC acquired by reliability analysis at different failure probability. It is easy to note that the impact resistance performance of CFRCC increases approximately linearly with the CF dosage increasing at the same failure probability. As an example of verifying whether the two-parameter Weibull distribution recommended in this study is also suitable to evaluate the impact performance of other fibers reinforced concrete, the test results of Ding et al. [36] for macro polypropylene fibers and steel fibers reinforced concrete are also analyzed by using Equation (6) and Equation (7), and the regression analysis results are given in Table 8. From Table 8, it can be seen that the Adjusted R 2 of each mixture is no less than 0.833, which indicates that the two-parameter Weibull distribution recommended in this study is also suitable to evaluate the impact performance of other types of fibers reinforced concrete.

Conclusions
In this study, the impact resistance of CFRCC under impact loading was investigated by conducting the drop-weight impact test. Based on the experimental results and regression analysis, the main conclusions can be drawn as follows: (1) The addition of CFs into coral concrete changed the failure pattern of coral concrete specimens under impact loading from obvious brittleness to relatively good ductility. (2) CF addition can improve the impact resistance at initial crack and final failure of coral concrete.
Still, the improvement of the impact resistance after initial cracking due to the addition of CFs is not as significant as steel fibers.  As an example of verifying whether the two-parameter Weibull distribution recommended in this study is also suitable to evaluate the impact performance of other fibers reinforced concrete, the test results of Ding et al. [36] for macro polypropylene fibers and steel fibers reinforced concrete are also analyzed by using Equations (6) and (7), and the regression analysis results are given in Table 8. From Table 8, it can be seen that the Adjusted R 2 of each mixture is no less than 0.833, which indicates that the two-parameter Weibull distribution recommended in this study is also suitable to evaluate the impact performance of other types of fibers reinforced concrete.

Conclusions
In this study, the impact resistance of CFRCC under impact loading was investigated by conducting the drop-weight impact test. Based on the experimental results and regression analysis, the main conclusions can be drawn as follows: (1) The addition of CFs into coral concrete changed the failure pattern of coral concrete specimens under impact loading from obvious brittleness to relatively good ductility. (2) CF addition can improve the impact resistance at initial crack and final failure of coral concrete.
Still, the improvement of the impact resistance after initial cracking due to the addition of CFs is not as significant as steel fibers.

Conflicts of Interest:
The authors declare no conflict of interest.
Appendix A