Experimental study on carbonation behavior of seawater sea sand recycled aggregate concrete

Abstract

The carbonation behavior of seawater sea sand recycled aggregate concrete (SSRAC) was investigated in this study. Considering different added water, fine aggregates, and coarse aggregates, the specimens were divided into 12 groups for the accelerated carbonation test by 7 days, 14 days, and 28 days. Among them, river sand, sea sand, and mixed sand (the proportion of sea sand and shell sand was 4:1) were used as fine aggregates. The results show that the carbonation depths of concretes with different mixtures all increase over time. When the replacement ratio of recycled coarse aggregate (RCA) is not less than 50%, sea sand is the most suitable fine aggregate to acquire best carbonation resistance, while river sand is the worst. On the contrary, when the replacement ratio of RCA is not more than 30%, river sand is the most suitable fine aggregate, while sea sand is the worst. It can also be speculated that the most appropriate replacement ratio of RCA is 50% when sea sand is applied as fine aggregates in SSRAC. At the same time, 30% replacement ratio of RCA is appropriate when mixed sand is applied as fine aggregates in SSRAC.

Keywords

seawater sea sand recycled aggregate concrete recycled coarse aggregate shell particles chloride content accelerated carbonation test carbonation depth

Introduction

With the rapid development of economic growth and urbanization, the cement industry in China developed fast in recent 30 years. In the year of 2017, Chinese cement production was 2.38 billion metric tons (Bildirici, 2019). Today, concrete is the most common construction material in the world due to its availability and easy preparation, which is composed of four basic ingredients: water, cement, coarse aggregates, and fine aggregates (Castillo et al., 2020). The tremendous consumption of natural resources like river sand, gravel, crushed stone, and fresh water in concrete industry has raised severe environmental concerns. Among the components of concrete, aggregates (coarse and fine) take up about 70–80% of concrete’s volume (Verian et al., 2018). The global demand for concrete aggregates was 45 billion metric tons in 2017, and this number will reach an estimated 66 billion metric tons by the end of 2025 (De Brito et al., 2019). It is believed that mining of sand and gravel is the most disastrous activity which can destroy natural environments and thus affect native wildlife (Padmalal and Maya, 2014). In particular, river sand has been exploited in a great deal by mainland China and other developing countries to meet the demand for a large number of concrete structures (Guo et al., 2020). River sand or gravel mining is perhaps the most detrimental factor leading to river degradation (Zhai et al., 2020). Meanwhile, as one of the most precious resources in earth, fresh water is in great short. A World Meteorological Organization (WMO) report (WMO, 1997) indicates that more than half of the world’s population will encounter water stress by 2025.

To address such problems above, the trend of utilizing recycled coarse aggregate (RCA) as a replacement for natural coarse aggregate (NCA), sea sand as a replacement for river sand, and seawater as a replacement for fresh water in concrete (i.e., seawater sea sand recycled aggregate concrete (SSRAC)) industry has been intensively investigated and prompted in recent years.

Seawater sea sand recycled aggregate concrete is a way to form a new kind of environmental construction material. Zhang et al. (2019) concluded that although RCA may decrease the long-term mechanical behavior of concrete, these aggregates could still be used with seawater and sea sand if proper measures were applied. Due to the chloride content in SSRAC, the use of glass-fiber reinforced plastic (GFRP) rather than normal steel is alternative in reinforced concrete to avoid corrosion (D’Antino and Pisani, 2019; Robert and Benmokrane, 2013; Zhou et al., 2018) However, GFRP bars are vulnerable in alkaline environments (Jia et al., 2020). Chen et al. (2006) found the exposure to alkaline environments reduced the tensile strength of the GFRP bar to its 50% approximately after a half year exposure in alkaline solutions with a pH value of about 13.6. At the same time, carbonation is a common phenomenon in concrete which leads to the reduction of alkalinity. According to the investigation by Kim et al. (2008), alkaline environmental condition had more influence on the degradation of GFRP rods than other environmental conditions. Since carbonation can reduce the content of free hydroxyl ions in concrete, it is feasible to use the carbonation process to delay the deterioration of GFRP bars in concrete environments and that is why it is meaningful to study carbonization of SSRAC.

Most existing investigations have been concerned with the carbonation of either recycled aggregate concrete (RAC) or seawater and sea sand concrete (SSC), while a few studies have examined the effect of using seawater, sea sand, and RCA at the same time on the carbonation behavior of SSRAC. It is generally accepted that the carbonation behavior of RAC made with recycled fine and/or coarse aggregates is slightly inferior to that of ordinary concrete with the same water-to-cement ratio due to the higher water absorption and air permeability (Otsuki et al., 2003; Silva et al., 2015). With the increasing replacement ratio of RCA and the same mix proportion to that of natural aggregate concrete (NAC), there is an increasing trend of the carbonation depth in RAC (Amorim et al., 2012; Sagoe-Crentsil et al., 2001). Silva et al. (2015) drew a conclusion that the carbonation depth of RAC with 100% replacement ratio of RCA was nearly 2.15 times greater than that of ordinary concrete. Zhang and Xiao (2018b) collected the existing test results of carbonation depth of RAC and reported that the carbonation coefficient of RAC follows a gamma distribution. The carbonation behavior of RAC was sensitive to the water-to-cement ratio and the replacement ratio of RCA. Zhang and Xiao (2018a) evaluated the existing prediction models of carbonation depth for RAC and proposed a reasonable model to assess the carbonation behavior of RAC.

As for concrete made with seawater and/or sea sand, Liu et al. (2016) reported that the carbonation of sea-sand concrete could be reduced 20–50% and claimed that chloride ions brought by sea sand could reduce the porosity and optimize the pore distribution of cement paste, indicating that the presence of chloride ions improved the carbonation resistance of concrete. However, studies by Cao et al. (2010) and Jiang et al. (2009) claimed that the carbonation depth development of sea-sand concrete is comparable to that of ordinary concrete. The carbonation depth is increasing as the carbonation time increases and the concrete grade decreases. Shen et al. (2019) explored the influence of carbonation on chloride binding. Ramezanianpour et al. (2014) investigated the combined effect of carbonation and chloride ion ingress on microscopic and mechanical properties of concrete. In addition, chloride diffusion is much faster than the speed of carbonation ingress (Wang et al., 2017). The content of shell sand is also a key difference between sea sand and river sand. Newman (1968) concluded that shell particles were strong and durable and had positive influence on the decrease of porosity. Existing study by Yang et al. (2010) showed that shell particles in fine aggregates improved the permeability resistance and had no apparent effect on carbonation behavior.

The purpose of this study is to investigate the carbonation behavior of SSRAC. Since the carbonation is likely to delay the deterioration of GFRP bars reinforced SSRAC, it is of much significance to learn the carbonation behavior of SSRAC. This research aims to explore the factors which can truly influence the carbonation behavior of SSRAC through the accelerated carbonation test.

Experimental program

Materials

The materials to be prepared included inorganic salt (used to prepare seawater), cement, river sand, sea sand, shell sand, NCA, RCA, and superplasticizer. Table 1 presents the chemical and mineralogical compositions of cement. Fine aggregates used in the test including river sand, sea sand, and mixed sand (the proportion of sea sand and shell sand was 4:1). River sand used in concrete was medium sand while shell sand was with grain size from 1 mm to 3 mm. Table 2 presents the properties of them, and Figure 1 shows the particle size distribution of the fine aggregates, which reveals that sands fell within the size limits of the Chinese code (JGJ 206-2010, 2010), except when the sea sand was below 0.6 mm and the shell sand was below 1.0 mm, and the limits in the ISO code (ISO 7033:1987, 2014) are much stricter than those in the Chinese code (JGJ 206-2010, 2010).

Table 1.

Chemical composition of cementitious material (%).

SiO₂	Al₂O₃	Fe₂O₃	CaO	TiO₂	MgO	SO₃	Na₂O	K₂O	Others
23.18	9.31	3.56	58.02	0.40	1.36	2.62	0.41	1.02	0.12

Table 2.

Physical properties of sea sand and shell sand.

Sand	Apparent density (kg/m³)	Fineness modulus	Content of Cl⁻ (%)	Content of SO₃ (%)	Shell particle content (%)
Sea sand	2660	2.6	0.057	0.123	1.66
Shell sand	2620	3.3	0.001	0.180	99.43

Figure 1.

Particle size distribution of the fine aggregates.

Natural coarse aggregate was made of natural stones with grain size from 5 mm to 25 mm while RCA used in concrete was with grain size from 4.75 mm to 31.5 mm. Based on the test method in GB/T 14685-2011 (2011), the physical properties of NCA and RCA were tested and presented in Table 3. The surface of RCA is usually attached with old mortar, which can greatly increase the water absorption of RCA. This influencing factor must be considered when RCA is used to prepare test specimens. Usually, additional water is added to ensure a certain water-to-cement ratio (Geng et al., 2019; Zhang et al., 2020). Therefore, the water absorption of RCA must be measured before mix proportion design. The determination of water absorption was carried out according to item 7.14 of national standard pebble and crushed stone for building (GB/T 14685-2011, 2011). As shown in Table 3, the mean water absorption of RCA was 6.6%. According to the water content and the water absorption of RCA, the additional water can be determined in the mix proportion.

Table 3.

Physical properties of coarse aggregates.

Aggregates	Bulk density (kg/m³)	Apparent density (kg/m³)	Clay content (%)	Crushing index (%)	Water content (%)	Water absorption (%)
NCA	1440	2660	0.80	5.1	0.40	1.03
RCA	1280	2590	0.71	11.0	3.87	6.60

RCA: recycled coarse aggregate; NCA: natural coarse aggregate.

The mixtures were prepared with ordinary Portland cement 42.5 R of Chinese standard [37]. Besides, the superplasticizer was used in the mixtures. The seawater was prepared according to Chinese standard (JGJ 55-2011, 2011), the chemical composition given in the seawater preparation regulation provided by American Society for Testing and Materials (ASTM D1141-98, 2013), in which the content of chemical substance below 0.1 g/L could be neglected. The specific details of prepared seawater can be found in the study of Xiao et al. (2017).

Mix proportion and specimens

In this carbonation experiment, two main variables were considered when designing mix proportion, namely, shell particle content and the replacement ratio of RCA. The type of water was not considered as the variable since the sea sand used in the experiment is non-desalted (i.e., fresh water was not used to prepare the concrete with sea sand). For the purpose to prepare recycled concrete with both seawater and sea sand, that is, SSRAC, the seawater was used in concrete once the sea sand was involved in this investigation. It is necessary to control the water-to-cement ratio, sand ratio, the mass of coarse aggregates, the mass of fine aggregates, and other unrelated factors to control variables in the experiment. In all, 12 concrete mixtures were designed as shown in Table 4.

Table 4.

Concrete mixtures (kg/m³).

Specimen	Free water		Additional water	Cement	Fine aggregate			Coarse aggregate
Specimen	Fresh water	Seawater	Additional water	Cement	River sand	Sea sand	Shell sand	NCA	RCA
RAC-0	150	0	0	319	829	0	0	1099	0
RAC-30	150	0	9	319	829	0	0	769	330
RAC-50	150	0	15	319	829	0	0	549	549
RAC-100	150	0	30	319	829	0	0	0	1099
M-SSRAC-0	0	150	0	319	0	829	0	1099	0
M-SSRAC-30	0	150	9	319	0	829	0	769	330
M-SSRAC-50	0	150	15	319	0	829	0	549	549
M-SSRAC-100	0	150	30	319	0	829	0	0	1099
H-SSRAC-0	0	150	0	319	0	663	166	1099	0
H-SSRAC-30	0	150	9	319	0	663	166	769	330
H-SSRAC-50	0	150	15	319	0	663	166	549	549
H-SSRAC-100	0	150	30	319	0	663	166	0	1099

RCA: recycled coarse aggregate; NCA: natural coarse aggregate; SSRAC: seawater sea sand recycled aggregate concrete.

The water-to-cement ratio was fixed as 0.47 (corresponding to concrete C30), and the sand ratio was fixed as 0.43. According to shell particle content and the replacement ratio of RCA, the concrete mix proportion was divided into 12 groups, among which shell particle contents in fine aggregates and replacement ratios of RCA were divided into three and four grades, respectively. The shell particle contents of river sand, sea sand, and mixed sand (the ratio of sea sand to shell sand is 4:1) are 0, 1.66%, and 21.33%, respectively. Meanwhile, the replacement ratios of RCA included 0, 30%, 50%, and 100%. For RAC, since river sand was used as fine aggregates without shell particles, only the replacement ratio of RCA should be considered for them. For example, RAC-30 represents 30% replacement ratio of RCA adopted in RAC. Considering the other shell particle contents (1.66% and 21.33%), the first letter in the specimen’s name label including M and H stands for the medium and high shell particle content, respectively. For specimen M-SSRAC-30, M means sea sand with 1.66% shell particle content; 30 means 30% replacement ratio of RCA adopted in the mix proportion. The concrete mixtures for specimens are shown in Table 4. tIn this experiment, 6 identical specimens with the size of 100 mm × 100 mm × 300 mm were prepared for each mix proportion to acquire the mean value of carbonation depth.

The fabrication of concrete test specimens is as follows. First, add the fine aggregates, coarse aggregates, and cement to the concrete mixer and keep stirring until they were mixed evenly. The simulated seawater was prepared in advance by adding chemicals to fresh water. Then add the superplasticizer (3.19 kg/m³) to the simulated seawater and add them together to the mixer in the process of stirring. Pour the concrete slurry into the mold and compact it by a vibrating rod at the same time. After face processing to the concrete in the mold, the demolding and maintenance work was carried out. Finally, the specimens were exposed under outdoor conditions until the carbonation test.

Accelerated carbonation test

The carbonation test method in the standard GB/T 50082-2009 (2009) for testing long-term performance and durability of ordinary concrete was adopted in this research. The specimens were fully dried before the carbonation, and some adjustments were made according to actual conditions.

To investigate the carbonation behavior of SSRAC with a long period of curing, after the pouring of specimens, they were exposed to outdoor conditions for 90 days. The reason for longer age of curing is that the presence of salts such as NaCl and K₂SO₄ in seawater and sea sand can accelerate the hydration process of concrete leading to the rapid growth in the early strength (Xiao et al., 2018). At the same time, Cl⁻ and SO₄²⁻ ions diffusing in concrete can result in the formation of soft hydration products. The lower long-term strength compared to ordinary concrete was due to the leaching out of soft hydration products (Islam et al., 2012). Salt crystallization is also responsible for the lower long-term strength and development of internal cracking in concrete (Wegian, 2010). Therefore, considering the effects of hydration products and salt crystallization, 90-day curing is more suitable for SSRAC samples. Before the carbonation test, except the two opposite experimental surfaces for carbonation, the remaining four surfaces were sealed with paraffin under dry condition. Parallel lines at 10 mm intervals along the length of side were drawn by pencil to mark the predetermined carbonation depths. The test specimens were placed in the HTX-12W concrete carbonation test box, leaving a space of not less than 50 mm between the carbonized surfaces of each. Then seal the carbonation box tightly and turn on the power and carefully unscrew the carbon dioxide gas tank valve to inject carbon dioxide into the carbon box slowly. During this process, the carbon dioxide concentration was observed by the digital display, and the flow of carbon dioxide into the box was controlled by fine-tuning the carbon dioxide pressure reducer in order to keep the carbon dioxide concentration in the box within (20 ± 3)% finally. The relative humidity was set within the range of (70 ± 5)%, and the temperature was controlled to (20 ± 2)°C in the carbonation test box. Different carbonation periods including 7 days, 14 days, and 28 days were considered in this test. The whole process of carbonation was carried out in three stages: 0–7 days, 7–14 days, and 14–28 days. After 7-day carbonation, the destructive specimens were put into the carbonation box again after sealed by paraffin to continue carbonation for the results of 14-day carbonation. Similarly, after 14-day carbonation, the destructive specimens were put into the carbonation box after sealed by paraffin to continue carbonation for the results of 28-day carbonation.

Measurement of carbonation depth

After each stage of carbonation, the test specimens were taken out and cut with a concrete cutting machine. The thickness of each cut was about 50 mm. The carbonation depth was measured with phenolphthalein solution of 1% concentration sprayed on the cutting face as an indicator. After the measurement, the damaged paraffin covers were repaired, and the cutting face was sealed with paraffin as well. Then, the repaired test specimens were put into the carbonation box again to continue the next carbonation stage. Phenolphthalein solution was chosen as an indicator due to the pH range getting phenolphthalein solution red is about 8.2–10, so it would get red when applied to the non-carbonated zone of concrete. On the other hand, the phenolphthalein solution would still remain colorless when applied to the carbonated zone. Therefore, it is feasible to detect the complete carbonation zone in this way. For each specimen, multiple points were chosen to acquire a mean carbonation depth and measured by a stick with 1 mm precision after 30 s (GB/T 50082-2009, 2009), as shown in Figure 2. In all, six specimens were required for each mix proportion to acquire the mean carbonation depth.

Figure 2.

Measurement points of carbonation depth.

Test results

Carbonation depth

The mean values of carbonation depths by different carbonation periods are drawn in Figure 3.

Figure 3.

Carbonation depths of seawater sea sand recycled aggregate concrete.

As shown in Figure 3, the carbonation depths of concrete test specimens with different mixtures all increased over time. The average increase of 14-day carbonation depths was 145.0% compared with the 7-day carbonation depths. Meanwhile, the average increase of 28-day carbonation depths was 48.7% compared with the 14-day carbonation depths. For instance, the increases of carbonation depths of M-SSRAC-30 by 14-day and 28-day carbonation test compared with the previous stage were 144.7% and 37.7%, respectively. The increases of carbonation depths of M-SSRAC-100 by 14-day and 28-day carbonation test compared with the previous stage were 222.2% and 41.4%, respectively. It can be concluded that although the carbonation depths increased over time, carbonation mostly occurred in early age. The possible reason is that the reaction product between calcium hydroxide and carbonate, namely, calcium carbonate, plays a role in fulfilling the pores in concrete (Puatatsananon and Saouma, 2005) and in turn impedes the CO₂ penetration.

Influence of replacement ratio of RCA

When the shell particle content in fine aggregates was fixed at 0, 1.66%, and 21.33%, respectively, the carbonation depth differed with the replacement ratio of RCA, as shown in Figure 4.

Figure 4.

Influence of the replacement ratio of recycled coarse aggregate on carbonation depth: (a) shell particle content of 0 in fine aggregates; (b) shell particle content of 1.66% in fine aggregates; and (c) shell particle content of 21.33% in fine aggregates.

When the fine aggregates used in concrete were river sand without any shell particles, as shown in Figure 4(a), RAC-100 with 100% replacement ratio of RCA had the highest carbonation depth in each carbonation stage. The 7-day carbonation depth of ordinary concrete RAC-0 was only 0.3 mm, while that of RAC-100 was 2.7 mm, which was 9 times greater than that of RAC-0. The 14-day carbonation depth of RAC-0 was 0.7 mm, while that of RAC-100 was 4.2 mm, which was 6 times of that of RAC-0. The 28-day carbonation depth of RAC-0 was 1.0 mm, while that of RAC-100 was 5.9 mm, which was about 6 times greater than that of RAC-0. The carbonation depth of RAC-50 which was with 50% replacement ratio of RCA was second-highest after RAC-100 in each carbonation stage. When the carbonation period was 7 days, the carbonation depths of RAC-100 and RAC-50 were 2.7 mm and 1.6 mm, indicating that the carbonation depth of RAC-100 was about 1.7 times greater than that of RAC-50. When the carbonation period was 14 days, the carbonation depths of RAC-100 and RAC-50 were 4.2 mm and 2.8 mm, indicating that the carbonation depth of RAC-100 was 1.5 times greater than that of RAC-50. When the carbonation period was 28 days, the carbonation depth of RAC-100 and RAC-50 was 5.9 mm and 5.4 mm, respectively, indicating that the carbonation depth of RAC-100 was about 1.1 times greater than that of RAC-50. The carbonation depth of RAC-30 had no obvious difference to that of RAC-0, which suggests that the replacement ratio of RCA not more than 30% has no apparent effect on the carbonation behavior of RAC. But the overall trend is that the carbonation depth of RAC increases with the increase in the replacement ratio of RCA, and the difference between the carbonation depth of RAC and that of ordinary concrete shows a tendency of gradually increasing over time. According to the investigation by Otsuki et al. (2003), the carbonation performance of RAC is influenced mainly by the presence of RCA and interface transition zone (ITZ). First, the porosity of RCA is greater than that of NCA, which increases the porosity of RAC compared with its NAC counterpart. Second, since ITZ is the weak zone of RAC, the initial location of crack tends to appear in ITZ which gives CO₂ access of ingress. Therefore, the carbonation resistance gets worse with the increasing replacement ratio of RCA, consistent with the experimental phenomenon.

As shown in Figure 4(b), the carbonation depths of specimens with 1.66% shell particle content had no obvious trend with changes in the replacement ratio of RCA. The carbonation depths of M-SSRAC-0 and M-SSRAC-30 rather than M-SSRAC-100 were the highest. Similarly, Figure 4(c) shows the carbonation depths of specimens with 21.33% shell particle content and different replacement ratios of RCA. Although the carbonation depths of H-SSRAC-100 were highest, the carbonation depths of H-SSRAC-30 and H-SSRAC-50 were lower than H-SSRAC-0. Thus, it can be speculated that the presence of sea sand and mixed sand leads to the difference between the carbonation behavior of SSRAC and RAC. The carbonation depth of SSRAC does not increase with the increase in the replacement ratio of RCA any more. Compared to the single influence of the replacement ratio of RCA, the coupled influence of RCA and fine aggregates seems more important and complicated on the carbonation behavior of SSRAC.

Influence of shell particle content

When the replacement ratio of RCA was fixed at 0, 30%, 50%, and 100%, the carbonation depth varied with different shell particle contents, as shown in Figure 5.

Figure 5.

Influence of the shell particle content on the carbonation depth: (a) 0 replacement ratio of RCA; (b) 30% replacement ratio of RCA; (c) 50% replacement ratio of RCA; and (d) 100% replacement ratio of RCA. RCA: recycled coarse aggregate.

As shown in Figure 5(a) and (b), the replacement ratios of RCA were not more than 30%, and thus, the main component of coarse aggregates was NCA. It has been confirmed in section 3.2 that such small replacement ratio of RCA has no apparent effect on the porosity of RAC and the carbonation behavior is almost the same as ordinary concrete. The 28-day carbonation depths of RAC-0, M-SSRAC-0, and H-SSRAC-0 were 1.0 mm, 6.2 mm, and 3.0 mm, indicating that the 28-day carbonation depth of M-SSRAC-0 was about 6 times and 2 times greater than those of RAC-0 and H-SSRAC-0, respectively. The 28-day carbonation depths of RAC-30, M-SSRAC-30, and H-SSRAC-30 were 1.0 mm, 7.9 mm, and 1.5 mm, indicating that the 28-day carbonation depth of M-SSRAC-30 was about 8 times and 5 times greater than those of RAC-30 and H-SSRAC-30, respectively. When the replacement ratio of RCA was not more than 30%, the carbonation depth of concrete which used sea sand with 1.66% shell particle content was the maximum. Conversely, the carbonation depth of concrete which used river sand without any shell particle content was the minimum. When fine aggregates were mixed sand with 21.33% shell particle content, the carbonation depth was between the maximum and the minimum.

As shown in Figure 5(c) and (d), the replacement ratios of RCA were not less than 50%, and the results were totally different. The 28-day carbonation depths of RAC-50, M-SSRAC-50, and H-SSRAC-50 were 5.4 mm, 1.1 mm, and 1.4 mm, indicating that the 28-day carbonation depth of RAC-50 was about 5 times and 4 times greater than those of M-SSRAC-50 and H-SSRAC-50, respectively. The 28-day carbonation depths of RAC-100, M-SSRAC-100, and H-SSRAC-100 were 5.9 mm, 3.4 mm, and 5.4 mm, indicating that the 28-day carbonation depth of RAC-100 was about 1.7 times and 1.1 times greater than those of M-SSRAC-100 and H-SSRAC-100, respectively. It could be concluded that when the replacement ratio of RCA was not less than 50%, the carbonation depth of concrete which used river sand without shell particles was the maximum. Conversely, the carbonation depth of concrete which used sea sand with 1.66% shell particle content was the minimum. When fine aggregates were mixed sand with 21.33% shell particle content, the carbonation depth was between the maximum and the minimum.

No regularity can be derived from the angle of shell particle content to analyze the carbonation behavior of SSRAC. It should be mentioned that although the mixed sand selected in this experiment had a higher content of shell particles than sea sand, as shown in Table 2, the content of chloride ions was lower than that of sea sand. Chloride ions can react with the cement hydration product of tricalcium aluminate (3CaO·Al₂O₃, C₃A) to generate refractory chlorine aluminate hydration. Friedel’s salt (FS, 3CaO·Al₂O·CaCl₂·10H₂O) forms by the dissolution and precipitation reactions: C₃A + 2CH + 2NaCl + 10H₂O → FS + 2NaOH. Besides, chloride ions react with Ca(OH)₂ produced by cement hydration to generate calcium chloride (CaCl₂) with a low solubility on the surface of C-S-H, which reduces the pore size and hence the permeability (Kayyali and Haque, 1988). The presence of calcium chloride further accelerates tricalcium silicate (C₃S) hydration to generate composite salt and increase the solid phase proportion in cement pastes (Shi et al., 2012). As a result, appropriate chloride content and RCA are beneficial for carbonation resistance. What’s more, the surface of RCA is usually attached with unhydrated cement, which can make it exhibit gel activity when encountering water. The reaction may happen on the surface of RCA, and the reaction products will fill the pores of RCA first rather than the cement matrix. However, excessive amounts of chloride relative to the amount of RCA will cause cracking and in turn accelerate the diffusion of CO₂ (Qiao et al., 2018). The cracking can be attributed to the too much leaching out of hydration products and the crystallization pressure if the salt content is in excess. In a sense, there is a most suitable ratio of RCA and chloride content for best carbonation resistance. Compared to the shell particle content, the carbonation behavior is more affected by the high concentration of chloride ions.

Discussions

The process of carbonation in concrete can be divided into three steps: The diffusion of CO₂ in concrete; the dissolution of CO₂ in the pore solution; and the formation of calcium carbonate. In fact, the third step, namely, the reaction of Ca²⁺ and CO₃²⁻ to generate calcium carbonate, is fairly quick. As a result, the most important factor controlling carbonation behavior is the diffusion rate of CO₂, which is determined by pore size distribution and the porosity of concrete. Thus, pore size distribution and porosity are key factors which influence the carbonation rate. The computed tomography (CT) technique (NIKON XTH 320/225, Nikon Corporation, Tokyo, speed of 0.5 s/rotation, exposure of 300 kV and 250 mA, recon matrix of 2000 × 2000, and slice thickness of 1.0 mm) is used to analyze the porosity of concrete to understand its relation with carbonation resistance. The results are shown in Figure 6.

Figure 6.

Porosities tested by CT: (a) RAC specimen; (b) M-SSRAC specimen; (c) H-SSRAC specimen. CT: computed tomography; SSRAC: seawater sea sand recycled aggregate concrete.

For RAC, the porosity of RAC-100 was larger than that of RAC-0 which is in line with the conclusion drawn in section 3.2 that the carbonation depth of RAC increases with the increase in the replacement ratio of RCA. Combined with the experimental phenomenon shown in Figure 4(a), it can be deduced that when the replacement ratio of RCA is not less than 50%, the porosity and ITZ caused by RCA is apparent and notable.

For M-SSRAC, the porosity of M-SSRAC-0 was larger than that of M-SSRAC-100, indicating that chloride ions in sea sand were so excessive that the leaching out of hydration products and crystallization pressure resulted in large porosity. The presence of RCA in M-SSRAC-100 abated such adverse influence because the initial pores in RCA could admit a large number of crystals, reducing the possibility of cracks and pores in concrete. Combined with Figure 4(b), it can be speculated that 50% replacement ratio of RCA is most suitable for the chloride content in M-SSRAC. The pores brought by 100% replacement ratio of RCA itself are larger than the reduced pores filled by reaction products. As for M-SSRAC-0 and M-SSRAC-30, they lacked adequate RCA, so the chloride content was relatively excessive for them and hence, porosities caused by crystallization pressure were large.

For H-SSRAC, the porosity of H-SSRAC-0 was larger than that of H-SSRAC-100. From Figure 4(c), it seems that 30% replacement ratio of RCA was appropriate for the chloride content in H-SSRAC, consistent with the less chloride content compared with M-SSRAC. Consequently, 50% and 100% replacement ratio of RCA was relatively excessive and hence large porosities brought by RCA. For H-SSRAC-0, the coarse aggregates were all NCA, and the chloride in fine aggregates was detrimental to carbonation resistance.

For specimens with 0 replacement ratio of RCA, the porosities of RAC-0, M-SSRAC-0, and H-SSRAC-0 are consistent with the experimental phenomenon. As shown in Figure 5(a) and (b), in the case of 0 or 30% replacement ratio of RCA, not enough RCA could afford spare room on its surface to accommodate reaction products. Therefore, the more chloride, the more pores were caused by the leaching out of hydration products and crystallization pressure. The carbonation resistance of M-SSRAC was the worst due to its most chloride ions followed by H-SSRAC. However, for specimens with 100% replacement ratio of RCA, the porosity of RAC-100 is more than the others in accordance with its highest carbonation depth. The reason why the porosity of M-SSRAC-100 is more than H-SSRAC-100 is probably due to stochasticity in the CT test. As shown in Figure 5(c) and (d), in the case of 50% or 100% replacement ratio of RCA, the amount of RCA was so large that the porosity brought by it was considerable. Pores of RCA could be filled with the reaction products of chloride and unhydrated cement particles attached on the surface of RCA. Hence, the carbonation resistance of M-SSRAC was the best due to its most chloride ions followed by H-SSRAC. Summarily, when the replacement ratio of RCA is not more than 30%, the carbonation resistance is vulnerable due to the presence of chloride ions. On the other hand, when the replacement ratio of RCA is not less than 50%, the carbonation resistance improves instead under the influence of chloride ions.

In a word, for RAC, the replacement ratio of RCA which is not more than 30% has negligible impact on the carbonation behavior of RAC. Thereafter, the porosity increases with the increase in RCA and hence worse carbonation resistance. As for SSRAC, the most important influencing factor is not shell particle content but the ratio of RCA and chloride content. In this investigation, although sea sand and shell sand had different shell particle contents, no apparent effect for this reason could be observed, indicating that the content of shell particles had negligible influence on carbonation behavior of SSRAC. Conversely, due to the content of chloride ions in the sea sand and mixed sand was different, a regular phenomenon could be found. Considering the surface of RCA is usually attached with unhydrated cement, which react with chloride ions and form products filling the pores of RCA, moderate chloride ions can make up for the defect owing to the usage of RCA, beneficial for carbonation resistance. However, excessive chloride ions are negative in that the leaching out of hydration products and crystallization pressure tend to cause cracking in concrete.

Conclusions

In this investigation, 12 mix proportions were applied considering different shell particle contents and replacement ratios of RCA. The development trend of carbonation depth was experimentally studied, and the most influencing factors of carbonation behavior were deduced. Conclusions can be drawn, and recommendations on future researches are given based on the current state-of-the-art.

The carbonation depths of specimens with different mix proportion all increased over time. The average increases of carbonation depths by 14 days and 28 days compared with 7 days and 14 days were 145% and 48.7%, respectively. It can be concluded that although the carbonation depth increases over time, carbonation mostly occurs in early age.

For RAC, the replacement ratio of RCA which is not more than 30% has negligible impact on the porosity of RAC. The 28-day carbonation depth of concrete with 100% replacement ratio of RCA, RAC-100, was about six times greater than that of ordinary concrete RAC-0. The overall trend is that the carbonation depth of RAC increases with the increase in the replacement ratio of RCA, and the difference between the carbonation depth of RAC and that of ordinary concrete shows a trend of gradually increasing over time.

As for SSRAC, due to the presence of sea sand and mixed sand, the carbonation depth of SSRAC does not increase with the increase in the replacement ratio of RCA any more. However, it is not because of the content of shell particles but the content of chloride ions in fine aggregates. Although three levels were set according to different shell particle contents, no clear relation can be concluded between carbonation resistance and shell particle content. Conversely, some regularities can be found from the viewpoints of chloride content.

When the replacement ratio of RCA is not less than 50%, sea sand is the most suitable aggregate to acquire best carbonation resistance while river sand is the worst. On the contrary, when the replacement ratio of RCA is not more than 30%, river sand is the most suitable aggregate to acquire best carbonation resistance and sea sand is the worst. The leaching out of hydration products and crystallization pressure caused by chloride ions in sea sand or mixed sand can be a disadvantage to carbonation resistance. However, further investigations are need to clarify the effect of replacement ratio of RCA on the carbonation behavior of SSRAC.

Due to the hydration process is responsible for the porosity and the pore size distribution in cement-based material, further research is needed to investigate the hydration process of SSRAC and the microstructures of hydration products. It should be mentioned that the conclusions above are obtained only by the analysis of the test results in this investigation.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors wish to acknowledge the financial support by the National Natural Science Foundation of China (NSFC, No. 52008304, 52078358), the China Postdoctoral Science Foundation (2019M661620), the Science and Technology Innovation Research, and the Shanghai Jianfeng Yichang Engineering Technology Co., Ltd.

ORCID iD

Jianzhuang Xiao

References

Amorim

Brito

Evangelists

(2012) Concrete made with coarse concrete aggregate: influence of curing on durability. ACI Materials Journal 109(2): 195–204.

ASTM D1141-98 (2013) Standard Practice for the Preparation of Substitute Ocean Water. Philadelphia, PA: Annual Book of ASTM.

Bildirici

(2019) Cement production, environmental pollution, and economic growth: evidence from china and USA. Clean Technologies and Environmental Policy 21(4): 783–793.

Cao

Zhao

, et al. (2010) Experiments research on the durability of concrete made of sea sand. In: Proceedings of the 19th National Conference of Structural Engineering, Jinan, Shandong Province, China, 5–8 November 2010 (in Chinese).

Castillo

EDR

Almesfer

Saggi

, et al. (2020) Light-weight concrete with artificial aggregate manufactured from plastic waste. Construction and Building Materials 265: 120199.

Chen

Davalos

Ray

(2006) Durability prediction for GFRP reinforcing bars using short-term data of accelerated aging tests. Journal of Composites for Construction 10(4): 279–286.

D’Antino

Pisani

(2019) Long-term behavior of GFRP reinforcing bars. Composite Structures 227: 111283.

de Brito

Agrela

Silva

(2019) Construction and demolition waste. In: New Trends in Eco-Efficient and Recycled Concrete. Series in Civil and Structural Engineering. Cambridge, UK: Woodhead Publishing, pp. 1–22.

GB/T 14685-2011 (2011) Pebble and Crushed Stone for Construction. Beijing: Standards Press of China (in Chinese).

10.

GB/T 50082-2009 (2009) Standard for Test Methods of Long-Term Performance and Durability of Ordinary Concrete. Beijing: China Architecture and Building Press (in Chinese).

11.

Geng

Zhao

Yang

, et al. (2019) Creep model of concrete with recycled coarse and fine aggregates that accounts for creep development trend difference between recycled and natural aggregate concrete. Cement and Concrete Composites 103: 303–317.

12.

Guo

Xing

, et al. (2020) Characterization of the mechanical properties of eco-friendly concrete made with untreated sea sand and seawater based on statistical analysis. Construction and Building Materials 234: 117339.

13.

Islam

Al-Amin

(2012) Suitability of sea water on curing and compressive strength of structural concrete. Journal of Civil Engineering 40(1): 37–45.

14.

ISO 7033:1987 (2014) Fine and coarse aggregates for concrete - determination of the particle mass-per-volume and water absorption-pycnometer method. Geneva, Switzerland: The International Organization for Standardization.

15.

JGJ 206-2010 (2010) Technical Code for Application of Sea Sand Concrete. Beijing, China: Architecture & Building Press (in Chinese).

16.

JGJ 55-2011 (2011) Specification for Mix Proportion Design of Ordinary Concrete. Beijing, China: Architecture and Building Press (in Chinese).

17.

Jia

Guo

Mao

, et al. (2020) Durability of glass fibre-reinforced polymer (GFRP) bars embedded in concrete under various environments. I: experiments and analysis. Composite Structures 234: 111687.

18.

Jiang

Zhao

Song

(2009) Study on carbonation performance of sea sand concrete. Engineering Construction 41(4): 11–14 (in Chinese).

19.

Kayyali

Haque

(1988) Chloride penetration and the ratio of Cl−/OH– in the pores of cement paste. Cement and Concrete Research 18(6): 895–900.

20.

Kim

H-Y

Park

Y-H

You

Y-J

, et al. (2008) Short-term durability test for GFRP rods under various environmental conditions. Composite Structures 83(1): 37–47.

21.

Liu

Cui

Dong

, et al. (2016) Carbonation of concrete made with dredged marine sand and its effect on chloride binding. Construction and Building Materials 120: 1–9.

22.

Newman

(1968) Sea-dredged Aggregates for Concrete: Proceedings of a Symposium. Buckinghamshire: Sand and Gravel Association Great Britain.

23.

Otsuki

Miyazato

S-I

Yodsudjai

(2003) Influence of recycled aggregate on interfacial transition zone, strength, chloride penetration and carbonation of concrete. Journal of Materials in Civil Engineering 15(5): 443–451.

24.

Padmalal

Maya

(2014) Sand Mining Environmental Impacts and Selected Case Studies. Berlin, Germany: Springer.

25.

Puatatsananon

Saouma

(2005) Nonlinear coupling of carbonation and chloride diffusion in concrete. Journal of Materials in Civil Engineering 17(3): 264–275.

26.

Qiao

Suraneni

Weiss

(2018) Damage in cement pastes exposed to NaCl solutions. Construction and Building Materials 171: 120–127.

27.

Ramezanianpour

Ghahari

Esmaeili

(2014) Effect of combined carbonation and chloride ion ingress by an accelerated test method on microscopic and mechanical properties of concrete. Construction and Building Materials 58: 138–146.

28.

Robert

Benmokrane

(2013) Combined effects of saline solution and moist concrete on long-term durability of GFRP reinforcing bars. Construction and Building Materials 38: 274–284.

29.

Sagoe-Crentsil

Brown

Taylor

(2001) Performance of concrete made with commercially produced coarse recycled concrete aggregate. Cement and Concrete Research 31(5): 707–712.

30.

Shen

X-H

Jiang

W-Q

Hou

, et al. (2019) Numerical study of carbonation and its effect on chloride binding in concrete. Cement and Concrete Composites 104: 103402.

31.

Shi

Xie

Fortune

, et al. (2012) Durability of steel reinforced concrete in chloride environments: an overview. Construction and Building Materials 30: 125–138.

32.

Silva

Neves

de Brito

, et al. (2015) Carbonation behaviour of recycled aggregate concrete. Cement and Concrete Composites 62: 22–32.

33.

Verian

Ashraf

Cao

(2018) Properties of recycled concrete aggregate and their influence in new concrete production. Resources, Conservation and Recycling 133: 30–49.

34.

Wang

Nanukuttan

Bai

, et al. (2017) Influence of combined carbonation and chloride ingress regimes on rate of ingress and redistribution of chlorides in concretes. Construction and Building Materials 140: 173–183.

35.

Wegian

(2010) Effect of seawater for mixing and curing on structural concrete. The IES Journal Part A: Civil & Structural Engineering 3(4): 235–243.

36.

World Meteorological Organization (1997) Comprehensive Assessment of the Freshwater Resources of the World. Stockholm, Sweden: World Meteorological Organization.

37.

Xiao

Qiang

Nanni

, et al. (2017) Use of sea-sand and seawater in concrete construction: current status and future opportunities. Construction and Building Materials 155: 1101–1111.

38.

Xiao

Zhang

, et al. (2018) Basic mechanical properties of seawater sea-sand recycled concrete. Journal of Architecture and Civil Engineering 35(2): 16–22.

39.

Yang

E-I

Kim

M-Y

Park

H-G

, et al. (2010) Effect of partial replacement of sand with dry oyster shell on the long-term performance of concrete. Construction and Building Materials 24(5): 758–765.

40.

Zhai

Ding

, et al. (2020) An optimization model of sand and gravel mining quantity considering healthy ecosystem in Yangtze river, China. Journal of Cleaner Production 242: 118385.

41.

Zhang

Wang

Y-Y

Lehman

, et al. (2020) Autogenous-shrinkage model for concrete with coarse and fine recycled aggregate. Cement and Concrete Composites 111: 103600.

42.

Zhang

Xiao

(2018a) Prediction model of carbonation depth for recycled aggregate concrete. Cement and Concrete Composites 88: 86–99.

43.

Zhang

Xiao

(2018b) Time-dependent reliability analysis on carbonation behavior of recycled aggregate concrete based on Gamma process. Construction and Building Materials 158: 378–388.

44.

Zhang

Xiao

Zhang

, et al. (2019) Mechanical behaviour of seawater sea-sand recycled coarse aggregate concrete columns under axial compressive loading. Construction and Building Materials 229: 117050.

45.

Zhou

Chow

Lau

(2018) Structural behavior of GFRP reinforced concrete columns under the influence of chloride at casting and service stages. Composites Part B: Engineering 136: 1–9.