Compressive behavior and microstructure of concrete mixed with natural seawater and sea sand

Abstract

Noncorrosive reinforcement materials facilitate producing structural concrete with seawater and sea sand. This study investigated the properties of seawater and sea sand concrete (SSC), considering the curing age (3, 7, 14, 21, 28, 60, and 150 d) and strength grade (C30, C40, and C60). The compressive behavior of SSC was obtained by compressive tests and digital image correction (DIC) technique. Scanning electron microscope (SEM) and X-ray powder diffraction (XRD) methods were applied to understand the microstructure and hydration products of cement in SSC. Results revealed a 30% decrease in compressive strength for C30 and C40 SSC from 60 to 150 d, and a less than 5% decrease for C60 from 28 to 150 d. DIC results revealed significant cracking and crushing from 80% to 100% of compressive strength. SEM images showed a more compact microstructure in higher strength SSC. XRD patterns identified Friedel’s salt phase due to the chlorides brought by seawater and sea sand. The findings in this study can provide more insights into the microstructure of SSC along with its short- and long-term compressive behavior.

Graphical abstract

Keywords

seawater and sea sand concrete / compressive strength / strain field / microstructure / hydration products

1 Introduction

For the sustainability of the concrete industry, scholars have been exploring various sources of raw materials for making concrete. Desalting sea sand has been commercially used in the concrete industry for decades to address the shortage of river sand [1]. It requires an enormous amount of freshwater to reduce the attached salt, mainly the chloride, to avoid the corrosion of steel. This aggravates the shortage of freshwater, since large quantities of freshwater are also needed for mixing and curing concrete. According to the prediction of the World Meteorological Organization [2], there will not be enough drinking water for over 50% of the world’s population by the year 2025. For saving freshwater, scholars are exploring the possibility of using seawater in concrete since seawater accounts for 96.5% of all water on the earth [3].

Direct use of seawater [4–6] and sea sand [7,8] is generally considered not applicable for steel-reinforced concrete structures, mainly because of the high concentration of chlorides that can accelerate steel corrosion. However, noncorrosive reinforcement materials, such as fiber-reinforced polymer (FRP) and stainless steel, have raised the possibility of mixing concrete with untreated sea sand and seawater. FRP bars are considered promising alternatives to steel bars in concrete structures [9,10]. Khatibmasjedi et al. [11] embedded a batch of glass FRP (GFRP) bars in freshwater-mixed concrete and another batch in seawater-mixed concrete, and conducted accelerated aging tests for 24 months. According to mechanical test results, they concluded that GFRP bars under the two conditions had comparable tensile and shear properties. Robert and Benmokrane [12] tested concrete-wrapped GFRP bars after aging in saline solution of 3% sodium chloride, and observed only a minor change in tensile strength. Feng et al. [13] examined the influence of saturated sodium chloride solution on GFRP plates at 60°C for 90 d, and found little effects on flexural strength and modulus. Gooranorimi and Nanni [10] tested GFRP bars extracted from a bridge after 15 years of service, and Benmokrane et al. [14] investigated GFRP bars after 11 years of service in bridge barrier walls. During the tests, only a slight change was observed in the microstructure of GFRP bars [10,14]. Aforementioned study results revealed that GFRP composites maintained favorable properties after accelerated or natural aging process and showed significant resistance to chemical attack.

These findings confirm the possibility of mixing concrete with untreated seawater and sea sand for GFRP composites reinforced structures, and have promoted the research on seawater and sea sand concrete (SSC) [1,15]. Most of the results indicate that SSC has a higher early strength than conventional concrete because of the accelerated hydration of cement induced by the chlorides [16,17]. Younis et al. [18] and Wang et al. [17] adopted a scanning electron microscope (SEM) to reveal the more densified microstructure induced by the accelerated hydration of cement in seawater-mixed concrete at early ages as compared with conventional concrete. Wang et al. [17] also identified the formation of Friedel’s salt because of the presence of chloride in seawater, by X-ray powder diffraction (XRD). However, measurement of long-term strength shows some contradictory results and in some experiments SSC experienced strength reduction. Teng et al. [19] developed ultra-high-performance SSC to improve its durability in marine environments, and they observed a continuous compressive strength growth up to 90 d. Younis et al. [18] identified the strength growth of seawater-mixed concrete up to 56 d. Xiao et al. [20] mixed SSC with recycled aggregate to address the shortage of both natural aggregates and freshwater, and noted strength growth up to 180 d. On the contrary, Wegian [21] observed a reduction of 7.5%−26.5% in compressive strength of SSC at 90 d as compared to 28 d strength, and Kaushik and Islam [22] found the compressive strength of seawater-mixed concrete tended to decrease after 10 months.

Since seawater and sea sand are locally available materials, their concentration of soluble salt differs globally, which could induce differences in mechanical properties and microstructure of SSC [1,22–24]. It is necessary to investigate the influence of local and natural seawater and sea sand on strength growth before promoting their use in concrete.

Our previous studies have investigated the compressive behavior of SSC with artificial seawater and natural sea sand, and found its difference from conventional concrete [20] and the influence of seashell content [25]. This paper adopted natural seawater and sea sand to produce SSC. A more detailed strength growth was observed by compressive tests on 3, 7, 14, 21, 28, 60, and 150 d after casting for C30, C40, and C60 SSC. Meanwhile, the strain development under compressive loading was captured by the digital image correlation (DIC) technique. For the microstructure observation of SSC, typical images were taken with an SEM at high magnifications. The phase characterization in SSC was obtained by the XRD technique for better interpretation of its microstructure.

2 Materials and methods

2.1 Raw materials and mixes

Natural seawater and sea sand were used to mix concrete without any desalting process. They were collected from Haizhou Bay on the coast of the Yellow Sea, China. The properties of seawater and sea sand contributed to the understanding of the performance of SSC; they were therefore obtained before mixing SSC.

2.1.1 Natural seawater for SSC

The variation of salt concentrations in seawater can induce different strength values of cement paste [16]; the salt concentration in seawater was therefore measured before the concrete was mixed. The seawater was first tested by ion chromatography for its chloride ( \text{C}{\text{l}}^{-}) and sulfate ( {\text{SO}}_4^{2-}) concentrations. The results are shown in Fig. 1.

Figure 1 also presents similar experimental results for seawater by other researchers. These include measurements of samples collected and tested from Al-Khor in Qatar [18], Biscayne Bay in the USA [11], a coastal beach near Melbourne in Australia [26], and Repulse Bay in The South China Sea [19]. These values were also compared with the ones for artificial seawater in ASTM D1141 [27] and with the average values of ocean water [28]. The comparison showed the difference of \text{C}{\text{l}}^{-} and {\text{SO}}_4^{2-} concentrations among global sea areas, and that ASTM D1141 [27] adopted the average values. The \text{C}{\text{l}}^{-} and {\text{SO}}_4^{2-} concentrations of seawater in this study were below the averages.

2.1.2 Natural sea sand for SSC

The original sea sand used in this study contained a large number of seashell fragments, which can have a negative influence on SSC [29]. The sea sand was first sieved by a 5 mm × 5 mm sieve to reduce extra-large fragments. Afterward, sieve analyses were performed and the resultant cumulative passing curve is given in Fig. 2, where error bars are presented and the x-axis is in logarithm scale. As shown in Fig. 2, seashell fragments can be easily observed in sea sand grains.

Since sea sand is frequently submerged in seawater, it adsorbs salt from seawater which can significantly influence the properties of concrete [1,23,30,31]. The attached salt content also needs to be quantified. A sea sand sample of 500 g was first collected and washed using one-liter tap water. After the sand settled, the resulting clear liquid was collected and tested under the same procedure as the seawater. The attached chloride ( \text{C}{\text{l}}^{-}) and sulfate ( {\text{SO}}_4^{2-}) content values are given in Table 1, along with other physical properties of the sea sand. Table 1 also presents properties of sea sand from coastal areas of Shenzhen, China [24], and a coastal beach near Melbourne, Australia [26]. The comparison showed that the sea sand in this study had a smaller apparent density and fineness modulus than sea sand from those locations. Typically, it contained more seashell fragments, despite the sieving of extra-large shell pieces. It also absorbed much more chloride and less sulfate than the sea sand from Shenzhen, whereas it attached similar chloride and sulfate to the sea sand from the Melbourne coastal beach.

To further understand the chemical compositions of the sea sand, the XRD technique was used. The details of the XRD test are described in the following section “hydration product identification using XRD”. The sea sand was gently ground into grains that could pass a 300 μm sieve. The XRD patterns of sea sand were then obtained and are illustrated in Fig. 3. The phases of quartz, feldspar, and calcium carbonate were mainly identified. As shown in Fig. 3, the quartz appears to be the main phase of the sea sand, which is similar to river sand. The main difference is the calcium carbonate phase, which is mostly due to the seashell fragments. These observations were also reported by Hasdemir et al. [32], who adopted XRD to compare the compositions of sea sand and river sand. They found quartz and feldspar in both types of sand, and revealed that the calcite was more abundant in sea sand.

2.1.3 Mix proportions of SSC

Before making concrete, the properties of crushed stone were also tested. Its percent passing is shown in Fig. 1 and its aggregate crushing value was 16.0%.

After testing the raw materials, three groups of SSC were designed to gain different strength grades, namely C30, C40, and C60. The mix proportions of SSC are given in Table 2. Note that, neither the seawater nor the sea sand was desalinated. Portland cement, P.O. 42.5, was adopted for C30 and C40 SSC, and P.O. 52.5 was used for C60 high strength SSC, as per the GB 175 cement standard of China [33]. A Polycarboxylate superplasticizer was added to achieve favorable workability, the amount of which was 0.3% by cement weight for C30 and C40 SSC and 0.2% for C60 SSC.

2.2 Test and observation methods

2.2.1 Compression tests

The compressive strength of SSC was obtained by uniaxially compressive loading tests on 100 mm concrete cubes. Two variables, the strength grade and curing age, were considered. After obtaining the workability of fresh SSC by slump tests, cubes were cast in plastic molds before being moved to a standard curing room. To simulate the curing condition in the marine environment, where only seawater is available, the specimens were cured by spraying with the seawater that was used for the SSC mixing. After 28 d of curing the specimens were moved to the laboratory. Each batch of three cubes was tested at age of 3, 7, 14, 21, 28, 60, and 150 d to obtain the strength growth of SSC. The average and standard deviation of compressive strength were then calculated.

2.2.2 Strain analysis using DIC

During the compression tests, the strain development of SSC was also obtained by the DIC technique. As a non-contact measurement technique, the DIC method has been widely used to investigate strains and displacements of components. It can detect the whole deformation field of specimens under loading. The underlying theory and principles are elaborated elsewhere [34]. At 28 d age, the DIC equipment was utilized to capture the surface strain development of SSC cubes under compression. These cubes were sprayed with white paint and black speckles for strain capture. The DIC equipment was calibrated using a standard calibration board. Two high-resolution images were taken every second during the compressive loading process for later strain analysis.

Afterward, an area of interest was taken to represent the whole exposed surface of each cube. Two image extensometers were set on the surface of the concrete cube along the compressive loading direction to record the engineering strain values. The principle of image extensometers is the same as that of physical extensometers. They derive the engineering strain as the average elongation over a certain distance. In this study, the compression tests were terminated when the load reached its maximum value, therefore only partial strain values were obtained. For the whole stress and strain curves of SSC, one can refer to our previous work [25].

2.2.3 Microstructure investigation using SEM

SEM images of various magnifications were captured to understand the microstructure of SSC. At each target age, SSC fragments with an approximate size of 5 mm × 5 mm were carefully collected from crushed SSC cubes after compressive loading. Because concrete is non-conductive, these fragments were first evenly coated with gold in an auto fine coater under pressure below 8 Pa. After coating, they were attached to conductive tapes and installed in an analytical SEM under a high vacuum condition for microstructure observation. Images were captured under an accelerating voltage of 20 kV at various magnifications.

2.2.4 Hydration product identification using XRD

XRD is a convenient method of phase identification of samples and has been widely adopted in revealing hydration products of cement [35,36]. It was therefore used to identify the hydration products of cement in SSC in this investigation. The results could contribute to a better understanding of the results from SEM images, since most hydration products would be embedded in C-S-H gel and were hardly identified by SEM images alone. After the compression tests, SSC fragments were carefully selected to avoid attached sea sand and crushed stone. They were gently ground into powders with a diameter of less than 300 μm. The powders were then gathered and compacted on a glass holder before being installed in the scanning machine. Afterward, the XRD patterns were collected at 2θ from 5° to 20° with a scanning step of 0.02° and a scanning speed of 10°/min. The anode material was Cu of {\lambda }_{{\text{k}}_{\alpha }}\text{=}1.54187\text{{AA}}, under a 40 kV generator voltage and a 40 mA tube current. The intensities at corresponding 2θ were measured by counting the number of photons incident on the detector. Then the chemical phases were identified by comparing intensity peaks to intensity data from known minerals. More details regarding the XRD technique are described elsewhere [37].

3 Experimental results and discussions

3.1 Workability of fresh SSC

The workability indexes of fresh SSC by slump tests are illustrated in Fig. 4. They are compared with those of SSC mixed by Li et al. [26], Yang et al. [38] and Xiao et al. [20], and seawater-mixed concrete by Younis et al. [18].

Mixing concrete with seawater generally decreases the workability of concrete [18], mainly due to the soluble chloride that accelerates the cement hydration [39,40]. Li et al. [26] added hydrated lime slurry to improve the workability of SSC. Yang et al. [38] used a polycarboxylate-based water reducer to maintain the slump values between 120 and 150 mm. Xiao et al. [20] utilized a polycarboxylate superplasticizer that was 0.9% by cement weight to gain a 220 mm ± 20 mm slump for C40/C50 SSC. Younis et al. [18] adopted a polycarboxylic ether-based superplasticizer to gain a 480 mm slump flow value.

As observed from Fig. 4, both slump and slump flow values in this study were comparable to those from other experiments. Adding 0.3% polycarboxylate superplasticizer provided a slump of over 150 mm for C30 SSC and over 220 mm for C40 SSC, and 0.2% content provided a 200 mm slump for C60 SSC. The results showed that the amount of polycarboxylate superplasticizer in this investigation could provide enough workability for mixing SSC.

3.2 Compressive strength growth of SSC

Our previous experiments revealed that the compressive strength of SSC was 36%–76% higher at 7 d, and similar at 28 d, and 3.7%–10.2% lower at 180 d, as compared with conventional concrete [20]. This study added more detailed strength growth data for SSC at various ages. Figure 5(a) presents the strength growth curves with error bars for three groups of SSC. For each group, the deviation of strength after 28 d was generally larger than that before 28 d. The compressive strength tended to decline after 28 d for C60 SSC, while the value decreased after 60 d for C30 and C40 SSC. The strength decrease of SSC could be mainly attributed to the leaching of hydration products [22] and salt crystallization in the pores [21]. Younis et al. [18] adopted SEM along with EDX microanalysis to reveal the salt impurities at 56 d due to the reaction of calcium with sulfate ions, which could also contribute to the decrease in strength.

Currently, Eurocode 1992-1-1 [41] and ACI 209R [42] both provide equations to predict compressive strength growth for conventional concrete, depending on the cement type. They are illustrated in Fig. 5(b) with experimental data from this study, done by Xiao et al. [20], Dhondy et al. [43], and Wegian [21]. Generally, Eurocode 1992-1-1 overestimates the strength gain whereas ACI 209 underestimates the strength gain of SSC before 7 d in this study. After 28 d, Xiao et al. [20], and Dhondy et al. [43] reported a continuous strength growth that was larger than the standard prediction. However, Wegian [21] reported a smaller growth and even a decrease in strength at 90 d. Unlike the above results, the experimental data in the current study showed significant growth of over 30% at 60 d followed by an approximate 30% decrease in strength for both C30 and C40 SSC at 150 d. The different trends in strength growth revealed the complexity of predicting the strength of SSC.

Figure 5(c) compares the C60 SSC data in this study with standard predictions and results from Younis et al. [18]. Younis et al. [18] mixed concrete with untreated seawater without sea sand to achieve a compressive strength of 62 MPa at 28 d, and the results were therefore taken for comparison. The two standard curves generally underestimate the early strength of C60 SSC at 3 d in this study. With some fluctuation before 28 d, the strength of C60 SSC showed a gradual decrease of less than 5% after 28 d. This decrease was also observed by Wegian [21]. The standard predictions do not match the compressive strength of C60 SSC after 28 d, whereas results from Younis et al. [18] generally comply with the curve from Eurocode.

3.3 Strain development of SSC under compression

Pictures and strain fields of SSC cubes are compared in Fig. 6. These pictures illustrate the broken surfaces of cubes at the peak compressive load, while the colored strain contours by DIC demonstrate the corresponding strain distribution. The strain here refers to the Lagrange strain in the horizontal direction which is perpendicular to the compression direction. The colored strain contours in Fig. 6 clearly show the cracking paths and crushing patterns of all SSC cubes, and they have good agreement with the physical pictures. Meanwhile, the same contours also could identify the large strain that would induce cracks that one cannot observe with the naked eye.

Figure 7 shows the strain field evolution under different stress levels from 20% to 80% of peak load for C60 SSC at the age of 28 d. The strain contour under 100% peak load is displayed in Fig. 6(c). Figures 6 and 7 share the same strain legend for clear observation. As can be seen in Fig. 7, cracking of SSC surface initiated from corners of the concrete cube under approximately 20% of the peak load. The cracking lengthened into a deeper area of the cube when the load increased. Before 60% of peak load, the cracking evolved gradually without clear cracking in other areas. From 60% to 80% of the peak load, multiple cracking paths appeared on the concrete surface. These paths propagated significantly and connected when the load increased from 80% to 100% of the maximum value, resulting in the eventual crushing of SSC. A similar evolution of strain fields was also observed for C30 and C40 SSC.

According to the average of two image extensometers, the compressive engineering strain values at peak loads were 2292 με for C30, 2498 με for C40, and 3487 με for C60 SSC. The strain value became larger when the SSC possessed a higher strength. This trend was also noted for SSC prisms from our previous work [25].

3.4 Microstructure development of SSC

Typical SEM images were captured for all groups of SSC at different ages, and are presented in Figs. 8–10. The figures show the change of microstructure of SSC with time and can help to understand the strength growth of SSC. For C30 SSC at 14 d in Fig. 8(a), a clear expansive formation of ettringite could be observed, forming a loose and unevenly distributed microstructure in SSC. These long and thin ettringite crystals grew on C-S-H gel in all directions. The ettringite crystals were the resultant products of the reaction of calcium aluminate with water and sulfate. The {\text{SO}}_4^{2-} from seawater enhanced the hydration of calcium aluminate, producing more ettringite crystals [17]. The increased formation of ettringite could induce expansion to reduce the autogenous shrinkage of concrete [17,44]. Afterward, C-S-H grew to fill in pores and made the microstructure of SSC more compact from 21 to 28 d, as shown in Figs. 8(b) and 8(c). These observations were similar to those of Wang et al. [16].

For C40 SSC at 14 d in Fig. 9(a), the ettringite crystals were not as abundant as those in C30 SSC shown in Fig. 8(a). This observation was predictable since the lower water-to-cement ratio of C40 SSC resulted in smaller pores that restricted the growth of ettringite crystals. Figure 9(b) shows the honeycomb constitution of C-S-H gel at 21 d in C40 SSC. This honeycomb structure was also described by Kurdowski [45]. At 28 d, the microstructure became much denser, resulting in a higher strength value.

For C60 SSC from 14 to 28 d in Fig. 10, fewer fibrous crystals could be observed as compared with C30 and C40 SSC. This observation of a more compact microstructure for C60 SSC can contribute to the explanation of the higher strength than C30 and C40 SSC at different ages.

3.5 Hydration products of cement in SSC

The high concentration of salt in seawater can change the hydration process of cement in SSC, which differs from that of cement in concrete mixed with freshwater. The formation of Friedel’s salt is one of the most remarkable changes mainly due to the high chloride content [16,46]. Friedel’s salt is formed when aluminum phases react with the chloride ions in seawater [46,47]. In this study, XRD patterns were recorded for all SSC at different ages from 7 to 60 d. After the elimination of the background, the patterns are presented in Fig. 11.

Figure 11 also illustrates the standard peak positions for ettringite (AFt), Friedel’s salt, and calcium hydroxide (CH) phases before 20°. Typical peaks of Friedel’s salt phase were identified for all groups of SSC, as a result of using natural seawater and sea sand. Clear ettringite phases were also observed for all SSC. This means, despite no clear observation with SEM, XRD could identify the existence of ettringite crystals. Meanwhile, it was difficult to directly distinguish CH crystals through SEM images, since most of them were embedded in C-S-H gels. They could be identified by XRD patterns and the corresponding peaks for CH are shown in Fig. 11 for all SSC. The stable existence of ettringite, Friedel’s salt, and CH phases was also found by Li et al. [48], who also adopted XRD to observe more intensified ettringite and Friedel’s salt at 28 and 63 d in seawater-mixed cement paste.

4 Conclusions

With noncorrosive reinforcement materials, seawater and SCC can be used in structures without considering steel corrosion. This study mixed concrete of different compressive strength grades with natural seawater and non-desalted sea sand, and investigated its compressive behavior, microstructure, and hydration products of cement. Based on the results from this investigation, some conclusions can be made as follows.

1) The deviation of compressive strength of SSC after 28 d was larger than that before 28 d. C30 and C40 SSC experienced a significant increase of more than 30% in compressive strength at 60 d followed by a 30% decrease at 150 d. However, C60 SSC demonstrated a continuous decrease in compressive strength from 60 to 150 d, though the extent was within 5%.

2) The DIC method captured the whole evolution of the strain field of SSC under various compression levels up to 100% of compressive strength, and illustrated the cracking and crushing patterns of SSC cubes perfectly. From 80% to 100% of peak load, the cracking of SSC grew much faster than the case for any previous compression level. This method also gave compressive strain values of SSC that were comparable to the ones obtained by physical extensometers.

3) SEM images illustrated the change of microstructure during the period from 14 to 28 d due to the cement hydration development for SSC. A clear difference was identified in microstructure between normal strength and high-strength SSC. The much more compact microstructure of C60 SSC could contribute to explaining its higher strength than that of other grades of concrete.

4) XRD patterns demonstrated the development of hydration products of cement mainly under the influence of seawater from 7 to 60 d. The typical phase of Friedel’s salt was identified for all groups of SSC. XRD also identified the CH phase that could not be easily observed by SEM images.

References

[1] Xiao J, Qiang C, Nanni A, Zhang K. Use of sea-sand and seawater in concrete construction: Current status and future opportunities. Construction & Building Materials, 2017, 155 : 1101– 1111

[2] Nishida T, Otsuki N, Ohara H, Garba-Say Z M, Nagata T. Some considerations for applicability of seawater as mixing water in concrete. Journal of Materials in Civil Engineering, 2015, 27( 7): B4014004–

[3] Dhondy T, Remennikov A, Shiekh M N. Benefits of using sea sand and seawater in concrete: A comprehensive review. Australian Journal of Structural Engineering, 2019, 20( 4): 280– 289

[4] JGJ 63–2006. Standard of Water for Concrete. Beijing: China Architecture and Building Press, 2006

[5] BS EN 1008: 2002. Mixing Water for Concrete—Specification for Sampling, Testing and Assessing the Suitability of Water, Including Water Recovered from Processes in the Concrete Industry, As Mixing Water for Concrete. Brussels: CEN, 2002

[6] ASTM C1602/C1602M–18. Standard Specification for Mixing Water Used in the Production of Hydraulic Cement Concrete. West Conshohocken, PA: ASTM International, 2018

[7] JGJ 206–2010. Technical Code for Application of Sea Sand Concrete. Beijing: China Architecture and Building Press, 2010

[8] BS EN 12620: 2013. Aggregates for Concrete. Brussels: CEN, 2013

[9] Elmessalami N, El Refai A, Abed F. Fiber-reinforced polymers bars for compression reinforcement: A promising alternative to steel bars. Construction & Building Materials, 2019, 209 : 725– 737

[10] Gooranorimi O, Nanni A. GFRP reinforcement in concrete after 15 years of service. Journal of Composites for Construction, 2017, 21( 5): 04017024–

[11] Khatibmasjedi M, Ramanathan S, Suraneni P, Nanni A. Durability of commercially available GFRP reinforcement in seawater-mixed concrete under accelerated aging conditions. Journal of Composites for Construction, 2020, 24( 4): 04020026–

[12] Robert M, Benmokrane B. Combined effects of saline solution and moist concrete on long-term durability of GFRP reinforcing bars. Construction & Building Materials, 2013, 38 : 274– 284

[13] Feng P, Wang J, Wang Y, Loughery D, Niu D. Effects of corrosive environments on properties of pultruded GFRP plates. Composites. Part B, Engineering, 2014, 67 : 427– 433

[14] Benmokrane B, Nazair C, Loranger M, Manalo A. Field durability study of vinyl-ester-based GFRP rebars in concrete bridge barriers. Journal of Bridge Engineering, 2018, 23( 12): 04018094–

[15] Teng J, Yu T, Dai J, Chen G. FRP composites in new construction: Current status and opportunities. In: The 7th National Conference on FRP Composites in Infrastructure. Hangzhou, 2011

[16] Wang J, Xie J, Wang Y, Liu Y, Ding Y. Rheological properties, compressive strength, hydration products and microstructure of seawater-mixed cement pastes. Cement and Concrete Composites, 2020, 114 : 103770–

[17] Wang J, Liu E, Li L. Multiscale investigations on hydration mechanisms in seawater OPC paste. Construction & Building Materials, 2018, 191 : 891– 903

[18] Younis A, Ebead U, Suraneni P, Nanni A. Fresh and hardened properties of seawater-mixed concrete. Construction & Building Materials, 2018, 190 : 276– 286

[19] Teng J, Xiang Y, Yu T, Fang Z. Development and mechanical behaviour of ultra-high-performance seawater sea-sand concrete. Advances in Structural Engineering, 2019, 22( 14): 3100– 3120

[20] Xiao J, Zhang Q, Zhang P, Shen L, Qiang C. Mechanical behavior of concrete using seawater and sea-sand with recycled coarse aggregates. Structural Concrete, 2019, 20( 5): 1631– 1643

[21] Wegian F M. Effect of seawater for mixing and curing on structural concrete. IES Journal Part A: Civil & Structural Engineering, 2010, 3( 4): 235– 243

[22] Kaushik S K, Islam S. Suitability of sea water for mixing structural concrete exposed to a marine environment. Cement and Concrete Composites, 1995, 17( 3): 177– 185

[23] Zhao Y, Hu X, Shi C, Zhang Z, Zhu D. A review on seawater sea-sand concrete: Mixture proportion, hydration, microstructure and properties. Construction & Building Materials, 2021, 295 : 123602–

[24] Guo M, Hu B, Xing F, Zhou X, Sun M, Sui L, Zhou Y. Characterization of the mechanical properties of eco-friendly concrete made with untreated sea sand and seawater based on statistical analysis. Construction & Building Materials, 2020, 234 : 117339–

[25] Zhang K, Xiao J, Zhang Q. Experimental study on stress-strain curves of seawater sea-sand concrete under uniaxial compression with different strain rates. Advances in Structural Engineering, 2021, 24( 6): 1124– 1137

[26] Li Y, Zhao X, Singh Raman R K, Al-Saadi S. Thermal and mechanical properties of alkali-activated slag paste, mortar and concrete utilising seawater and sea sand. Construction & Building Materials, 2018, 159 : 704– 724

[27] ASTM D1141–98. Standard Practice for the Preparation of Substitute Ocean Water. West Conshohocken, PA: ASTM International, 2013

[28] Sillen L G. The ocean as a chemical system. Science, 1967, 156( 3779): 1189– 1197

[29] Vafaei D, Hassanli R, Ma X, Duan J, Zhuge Y. Sorptivity and mechanical properties of fiber-reinforced concrete made with seawater and dredged sea-sand. Construction & Building Materials, 2021, 270 : 121436–

[30] Sun W. Study on the influence of chloride ions content on the sea sand concrete performance. American Journal of Civil Engineering, 2016, 4( 2): 50– 54

[31] Dang V Q, Ogawa Y, Bui P T, Kawai K. Effects of chloride ions on the durability and mechanical properties of sea sand concrete incorporating supplementary cementitious materials under an accelerated carbonation condition. Construction & Building Materials, 2021, 274 : 122016–

[32] Hasdemir S, Tuğrul A, Yılmaz M. The effect of natural sand composition on concrete strength. Construction & Building Materials, 2016, 112 : 940– 948

[33] GB175-2007 . Common Portland Cement. Beijing: Standards Press of China, 2007

[34] Huang Y, He X, Wang Q, Xiao J. Deformation field and crack analyses of concrete using digital image correlation method. Frontiers of Structural and Civil Engineering, 2019, 13( 5): 1183– 1199

[35] De Weerdt K, Justnes H. The effect of sea water on the phase assemblage of hydrated cement paste. Cement and Concrete Composites, 2015, 55 : 215– 222

[36] Zhang X, Yang J, Li K, Pu H, Meng X, Zhang H, Liu K. Effects of steam on the compressive strength and microstructure of cement paste cured under alternating ultrahigh temperature. Cement and Concrete Composites, 2020, 112 : 103681–

[37] Cullity B D, Stock S R. Elements of X-Ray Diffraction. 3rd ed. Pearson Education Limited, 2014

[38] Yang S, Zang C, Xu J, Fan G. Determination of fracture parameters of seawater sea sand concrete based on maximum fracture load. Journal of Materials in Civil Engineering, 2020, 32( 1): 04019315–

[39] Younis A, Ebead U, Suraneni P, Nanni A. Performance of seawater-mixed recycled-aggregate concrete. Journal of Materials in Civil Engineering, 2020, 32( 1): 04019331–

[40] Li H, Farzadnia N, Shi C. The role of seawater in interaction of slag and silica fume with cement in low water-to-binder ratio pastes at the early age of hydration. Construction & Building Materials, 2018, 185 : 508– 518

[41] BS EN 1992-1-1: 2004+A1: 2014. Eurocode 2: Design of Concrete Structures—Part 1–1: General Rules and Rules for Buildings. London: British Standards Institution, 2004

[42] ACI 209R–92. Prediction of Creep, Shrinkage, and Temperature Effects in Concrete Structures (Reapproved 2008). Farmington Hills: American Concrete Institute, 2008

[43] Dhondy T, Remennikov A, Neaz Sheikh M. Properties and application of sea sand in sea sand-seawater concrete. Journal of Materials in Civil Engineering, 2020, 32( 12): 04020392–

[44] Bentz D P, Jensen O M. Mitigation strategies for autogenous shrinkage cracking. Cement and Concrete Composites, 2004, 26( 6): 677– 685

[45] Kurdowski W. Cement and Concrete Chemistry. Springer, 2014

[46] Birnin-Yauri U A, Glasser F P. Friedel’s salt, Ca₂Al(OH)₆(Cl, OH)·2H₂O: Its solid solutions and their role in chloride binding. Cement and Concrete Research, 1998, 28( 12): 1713– 1723

[47] Talero R, Trusilewicz L, Delgado A, Pedrajas C, Lannegrand R, Rahhal V, Mejía R, Delvasto S, Ramírez F A. Comparative and semi-quantitative XRD analysis of Friedel’s salt originating from pozzolan and Portland cement. Construction & Building Materials, 2011, 25( 5): 2370– 2380

[48] Li P, Li W, Yu T, Qu F, Tam V W Y. Investigation on early-age hydration, mechanical properties and microstructure of seawater sea sand cement mortar. Construction & Building Materials, 2020, 249 : 118776–

RIGHTS & PERMISSIONS

Higher Education Press 2021.