Introduction

Concrete is the most frequently used construction material in infrastructure development because its constituent components are locally available, easy to obtain, and cost-effective. It is estimated that 33 billion tons of concrete are used annually [1] and this is predicted to increase by 6% every year [2]. The main components of concrete consist of water, cement, as well as fine and coarse aggregates which make up more than half of the volume with approximately 60–85% [3]. River sand (NS) is the natural resource widely used as fine aggregate with approximately 40–50 billion tons reported to be used for construction projects every year [4]. The continuous use of NS in concrete has an impact on the environment in the form of easy erosion, flooding, and landslides. This, therefore, led to the use of industrial waste material which is another environmental problem as NS replacement to ensure a sustainable construction process.

Waste slag materials generated from the metallurgical industry have been investigated for their use as alternative construction materials in the past decades [5,6,7]. An example is a ferronickel slag (FNS) produced from the processing of nickel ore. The ore was smelted at a high temperature of 1500–1600 ℃ and approximately 1214 tons of FNS was generated for each one-ton production of nickel [8]. In Indonesia, about 1.67 million FNS are produced annually and mostly stacked in the open area near the site [9]. These waste materials have the ability to pollute the environment when exposed to acid rain due to the presence of several harmful substances. They are, therefore, required to be treated to avoid serious environmental problems. Several attempts have been made to deal with their environmental impacts and an example of these is their use as a substitute for fine aggregate in manufacturing mortar and concrete [2, 10, 11]. This approach helps to prevent the depletion of natural aggregate due to its excessive use in the concrete industry.

Corrosion is a phenomenon that occurs naturally over a long period of time. Chloride ingress and carbonation are closely related to the corrosion process in steel bars embedded in concrete. That corrosion degrades structures’ performance [12]. A proper corrosion monitoring system is required to select an appropriate repair strategy for reinforced concrete structures. Several corrosion monitoring systems have been developed, including pulse velocity [13], ground penetrating radar [14], and X-ray computer tomography [15]. However, these methods depend on highly developed corrosion in the steel bars having caused significant changes in the concrete’s volumetric properties and microstructure [16]. Accelerated corrosion testing is another technique used to evaluate the corrosion resistance of steel bars embedded in concrete. This technique works by accelerating the corrosion process through the application of a constant voltage or constant current. The corrosion process is monitored by measuring the current and the time until the specimen first cracks. The mass loss due to corrosion is controlled using a direct electrical current, and Faraday’s equation is used to predict the mass loss [17]. In this study, the corrosion resistance performance of steel bars embedded in FNS and fly ash (FA) concrete is evaluated using the accelerated corrosion test method.

Existing literature

Several studies have addressed the suitability of FNS as fine aggregate in mortar and concrete due to their physical characteristics is narrower to fine aggregates such as particles sizes, density, specific gravity, and sorption capacity [2, 3, 18,19,20]. Moreover, the heavy metals released in mortar and concrete using FNS are far below the environmental safety threshold required by USEPA and UK Environment Agencies and this denotes they are safe to use [21,22,23,24].

The durability performance is the main concern of adapting the FNS in concrete mixture due to the high content of magnesium oxide (MgO) which can lead to a high risk of expansion due to alkali-silica-reaction (ASR) [25,26,27]. Therefore, supplementary cementitious materials (SCM) have been widely adopted to replace cement in the production of low-carbon concrete such as the fly ash (FA) obtained from burning coal in thermal power plants [28, 29]. The use of FA in the concrete mixture has improved the durability in several cases as indicated by the resistance to ASR, chloride intrusion, and sulfate attack [30].

Studies have been conducted to investigate the performance of FNS mortar and concrete with and without FA. Shoya et al. [31] focused on the fresh and mechanical properties of concrete produced using FNS as a sand replacement and found that the workability of the concrete decreased as the FNS content increased due to excessive bleeding caused by poor gradation as well as high angularity of FNS which led to an increase in the internal friction of particles. However, a 10.92% improvement in compressive strength was recorded for 50% FNS while the specimen with 100% FNS had similar compressive strength with the normal concrete. Malek et al. [20] evaluated the use of post-war FNS 0–25% as sand replacement in concrete and discovered that the slump value decreased while the density and air content increased linearly with an increment in the FNS content. Moreover, the compressive strength increased by 1.09–1.31 times, flexural strength by 1.05–1.66 times, and split tensile strength by 1.04–1.43 times for FNS concrete compared to normal concrete. Sun et al. [3] also investigated the properties of concrete produced with two different types of FNS including blast furnace slag (BS) and electric furnace slag (ES) as fine aggregate. The results showed that the inclusion of BS and ES decreased the workability of concrete while the compressive strength was improved with BS but decreased with ES due to its high porosity. It was also discovered that the use of BS had a positive effect on the chloride, sulfate, and fire resistance of concrete and this led to the recommendation of 75% BS or 25% ES to replace sand for practical purposes. Furthermore, Liu et al. [22] evaluated the effect of FNS as fine aggregate at different storage times on the durability of concrete and the storage time of FNS was found not to affect the durability of concrete. Meanwhile, the sulfate resistance increased by 27% with 27% FNS without any significant effect on chloride resistance. The abrasion resistance also improved by 33% with 40% FNS compared to the control due to its high density, angular shape, and well-graded particles.

Saha and Sarker [21] also substituted sand with FNS and 30% of cement with FA in concrete production and discovered that 50% FNS and 30% FA slightly improved workability and mechanical properties compared to the control concrete. The trace elements of harmful substances from FNS and FA concrete were found to be far below the standards specified by USEPA and the UK Environmental Agencies. Another study by the same authors [32] indicated that the combination of 50% FNS as a sand substitute and 30% FA for cement in mortar exhibited low expansion and cracked less under sulphate exposure compared to the control specimen. Moreover, the pozzolanic reaction of FA densified the microstructure of mortar, thereby, preventing ettringite formation. Nuruzzaman et al. [33] also investigated the effect of FNS in high-strength self-compacting concrete (SCC) using 0, 20, 40, and 60% of FNS by volume as sand replacement and 30% FA as cement replacement in all SCC mixtures. It was found that the fresh properties of SCC with 40% FNS achieved the EFNARC Standard without segregation. The compressive strength of the SCC mixtures was increased by 1, 34, and 31% for 20, 40, and 60% FNS, respectively, while the split tensile strength also increased by 2, 15, and 8%, respectively, than the normal specimen. There was also volume permeable voids reduction, water absorption, and chloride resistance with the SCC mixture containing 40% FNS and this means it exhibits good durability. Furthermore, Nguyen et al. [34] investigated the performance of 50% FNS used to replace sand with 25% FA included as a substitute for cement against ASR and reported that the mixture produced low-carbon concrete with a good performance against ASR and chloride contamination. A similar observation was made by [35] that there is good interaction between FNS aggregate and cement-FA matrix due to the improvement recorded in the interfacial transition zone (ITZ) of concrete after replacing the sand with 50% FNS and cement with 25% FA. This consequently enhanced the resistance against chloride diffusion, carbonation, water absorption, sorptivity, volume permeable voids, as well as bulk and surface electrical resistivity compared to the control concrete.

Most of these past studies showed that the mechanical and durability properties of concrete produced using 50% FNS as a replacement for sand with 30% FA are better than natural sand concrete. However, further research is needed to establish confidence in the material for specific cases such as durability against corrosion which has not been explored. This is considered important because corrosion of steel bars in concrete, especially those with FNS, usually leads to cracking and loss of steel bars area with subsequent effect on the service-life of reinforced concrete structures. Therefore, there is a need to evaluate the corrosion resistance of concretes containing FNS as fine aggregate when they are exposed to a chloride environment.

The main objective of this study is to evaluate the corrosion resistance of concrete containing 50% FNS as sand replacement which is combined with 15 and 30% FA as cement replacement as well as to investigate the uniaxial compressive behavior of the mixtures for the first time.

Research significance

Previous research has shown the suitability of FNS as a replacement for sand due to its better or comparable mechanical and durability performance. However, there are no previous studies conducted to assess the performance of FNS against corrosion which is an important factor affecting the service life of structures. Therefore, the primary objective of this study is to evaluate the influence of using FNS as sand replacement in concrete against corrosion. It also focuses on evaluating the effect of adding FA into the FNS concrete to improve its durability. The compressive behaviors of the concrete products such as the stress–strain relationship, peak stress, peak strain, ultimate strain, and toughness were also assessed due to their importance in structural concrete application. The findings are expected to be used to promote the application of FNS and FA as sustainable construction materials and provide a valuable guide on the co-disposal of these waste materials.

Experimental programs

Materials

Commercially available ordinary Portland cement (OPC) was used as the binder material while the FA used for its partial replacement was obtained from a local thermal power plant in Makassar, South Sulawesi. The FA was dried in an oven at a temperature of 105 ℃ for 24 h and sieved through mesh sizes of 75 µm before it was used in the experiment. The chemicals properties of OPC and FA obtained through the X-ray fluorescence (XRF) are presented in Table 1 and it was discovered that the total SiO2 + Al2O3 + Fe2O3 of FA was 58.01% while the CaO content was 20.5% and the SO3 < 5%, thereby, indicating it is categorized as Class C based on ASTM C 618 [36]. Moreover, the specific gravity of OPC and FA was 3.16 and 2.19, respectively. It is important to note that the particles of FA are shown through scanning electron microscopy (SEM) to be mainly spherical as presented in Fig. 1.

Table 1 The chemical properties of OPC and FA
Full size table
Fig. 1
figure 1

The SEM image of FA

Full size image

The raw of the FNS used to partially replace NS in the concrete mix was obtained from the local stockpile of the nickel industry in Sorowako, South Sulawesi, Indonesia. FNS is obtained after the melted slag undergoes a natural air-cooling process. It is further crushed to a fine fraction by a jaw crusher and sieved through a mesh of size 4.75 mm in the laboratory to remove coarser particles. The physical appearance of NS and FNS presented in Fig. 2 showed that FNS is coarser and has higher angularity than NS. Moreover, the physical properties of the materials based on ASTM C128 [37] listed in Table 2 showed that FNS has higher specific gravity but lower water absorption due to its fewer pores and smooth particles compared to NS as indicated in the SEM images shown in Fig. 3. Furthermore, the grading curve was determined based on the ASTM C136 [38] in Fig. 4 and it was discovered the combination of 50% NS and 50% FNS produced a well-graded particle sizes distribution. In addition, crushed stone (CS) with a maximum size of 19 mm and specific gravity of 2.76 was used as the coarse aggregate while Sika-Viscocrete 3115N superplasticizer (SP) was applied to maintain the target slump value.

Fig. 2
figure 2

The physical appearance of NS (left) and FNS (right)

Full size image
Table 2 Physical properties of sand and FNS
Full size table
Fig. 3
figure 3

The SEM images of NS (left) and FNS (right)

Full size image
Fig. 4
figure 4

The grading curve of NS and 50%NS:50%FNS

Full size image

Concrete mixtures

A constant water-to-binder (w/b) ratio of 45% was used for all the mixtures with a slump value of 12.5 ± 2 cm. FA was added by replacing the OPC at 15 and 30% of weights in line with the recommendation of the SNI 15-3500-2004 [39] for the application of FA in blended cement. Meanwhile, the FNS aggregate was used to partially replace NS at 50% by volume. A superplasticizer of 0.5% by weight of the binder was also added to all the mixture to achieve the target slump value. It is important to note that all the aggregates were used in saturated surface dry conditions. The detail of the composition of the concrete mixture is presented in Table 3.

Table 3 The concrete mixture composition for 1 m3
Full size table

The concrete was produced using a concrete mixture pan capacity of 100 L. Firstly, the CS, NS/FNS, and binder were mixed for 1 min after which the water containing SP was added and blended continuously for 2 min. This was followed by the manual mixing of the materials for 1 min to scrap those trapped at the bottom and side part of the pan. The mixing was continued for 2 min before the fresh concrete was placed in the molds. All the specimens were removed from the molds after 24 h and cured in water up to the period for the tests.

Specimens and test methods

The slump value of each concrete mixture was measured according to ASTM C143 [40]. Moreover, the density and porosity (volume permeable voids) of the hardened concrete were tested based on the ASTM C642 [41] while the compressive strength was determined at 3, 7, and 28 days using cylindrical specimens which is 100 mm in diameter and 200 mm in height with a loading rate of 0.25 MPa/s in line with the SNI 1974:2011 [42]. The axial stress–strain behaviors were determined at 28 days according to the ASTM C469 [43] using a Ø100 mm × 200 mm cylindrical specimen. Furthermore, one set of compressometer consisting of two linear variable displacement transducers (LVDTs) with a capacity of 10 mm was used to measure the axial deformations as shown in Fig. 5. The load was applied to the surface of the specimens at a loading rate of 0.25 MPa/s until failure and the average value obtained from the three specimens for each mixture was reported for the hardened tests.

Fig. 5
figure 5

Setup for the axial stress–strain test

Full size image

The corrosion resistance test was conducted using an accelerated corrosion test after 28 days of curing the specimens in tap water. The process involved the production of Ø100 mm × 200 mm cylindrical specimens with centrally located deformed 16 mm-diameter steel bar embedded at a length of 150 mm as indicated in Fig. 6. The initial weight of the steel bars was recorded before they were placed in the molds. These specimens were partially immersed in a 3% NaCl solution and a constant DC voltage of 10 V was maintained to accelerate the corrosion process. The negative terminal of the DC power supply was connected to the stainless plate (counter electrode) and the positive terminal to the steel bars (working electrode). The corrosion current was monitored using an ammeter up to 240 h as indicated in the setup presented in Fig. 7. The procedure for measuring corrosion resistance performance using corrosion current was similar to that in previous studies by Refs [44, 45].

Fig. 6
figure 6

Specimens for corrosion test

Full size image
Fig. 7
figure 7

Setup for the corrosion test

Full size image

The corrosion resistance was determined based on the first crack appearance in the specimens as well as the weight loss of the steel bars after the accelerated corrosion tests. The weight loss was determined using calculation and gravimetric methods. The calculation method involves converting the corrosion current to weight loss using Faraday’s law in Eq. (1).

$$ \Delta M \, = \, (M \cdot I \cdot \Delta t)/\left( {Z \cdot F} \right) $$
(1)

where ΔM = weight loss of steel bar (g), M = atomic mass of Fe (56 g/mol), I = corrosion current (A), Δt = time duration (s), Z = valency of Fe (2), and Faraday constant is 96,500 C/mol.

The gravimetric method involves removing the steel bars from the concrete as well as adhering mortar on the steel surface after the test. The bars were immersed in a 10% diammonium hydrogen citrate (C6H14N2O7) solution at a temperature of 40 ℃ for 24 h, cleaned with a steel brush and distilled water, and weighed. The gravimetric weight loss was determined by comparing the weight of steel bars before and after the accelerated corrosion test. It is important to note that the average weight loss from the three specimens for each mixture was reported and the value was compared with those obtained from the calculation method.

Results and discussion

Slump and density

Figure 8 shows the slump and hardened density of the concrete mixture. It was discovered that the replacement of the fine aggregate with 50% FNS and the addition of fly ash increased the slump of the fresh concrete mixture. For instance, the slump value for the 50% FNS specimens was 8.33% higher than the control specimen with approximately 16.67% and 20.83% improvement recorded for the specimens with 15% and 30% fly ash, respectively. This is attributed to the decrease in aggregate surface area due to the addition of FNS [22]. Moreover, FNS has a higher fineness modulus than sand and this means less water was needed to wet the aggregate, thereby, leading to the increment in the workability of the concrete produced. Furthermore, the replacement of cement with FA caused a rise in the slump of the concrete produced due to the well-known ball-bearing effect of its spherical shape particles.

Fig. 8
figure 8

Slump and hardened density of concrete

Full size image

The density of the concrete with 50% FNS also increased by 7.22% compared to the control specimen as depicted in Fig. 8 and this is due to the higher density of FNS aggregate. A similar result has been previously reported by Ref. [33] that the density of the concrete produced with 20, 40, and 60% FNS increased by 0.58, 3.15, and 4.73%, respectively, compared to the normal concrete. Moreover, the replacement of cement with 15 and 30% fly ash reduced the density of FNS concrete by 1.05 and 1.46%, respectively, due to the lower density possessed by FA than cement.

Compressive strength

The compressive strength of concretes at different curing ages is presented in Fig. 9 and it was discovered that the specimen with 50% FNS increased the compressive strength by 67.77, 48.10, and 27.77% at 3, 7, and 28 days, respectively, than the control specimen. The FNS used in this study had higher angularity, density, fineness modulus, and better aggregate gradation than sand, and this can improve the particle packing and compressive strength of the concrete produced. A similar trend was reported by Ref. [2] where approximately 8% higher compressive strength was obtained by replacing 50% of sand with FNS compared to the normal concrete. Another study by Ref. [20] also showed that the substitution of sand with 25% FNS improved the compressive strength by 31.1% at 28 days compared to normal concrete. Moreover, Sun et al. [3] utilized two types of slag including blast furnace (BS) and electric furnace (ES) as sand replacements to manufacture concrete and concluded that there is an approximately 6, 9, and 11% increase in the compressive strength of the concrete produced with 25, 50, and 75% BS. The concrete containing ES aggregate showed a decrease in compressive strength as the content increased due to the spherical shape and smoothness of its particles which further affects the interaction between the components of the concrete. The increment recorded for the compressive strength of the concrete with 50% FNS in this study was estimated to be 35% higher than the values recorded in similar studies. This is possibly associated with the high angularity and sharp edges of the FNS used compared to these previous studies which increased the interlocking between the particles in the concrete and significantly enhanced the compressive strength.

Fig. 9
figure 9

Compressive strengths of the concrete mixture at different curing ages

Full size image

The results also showed that the replacement of cement with fly ash in the FNS concrete caused a lower compressive strength at an early age of 3 and 7 days than the FNS concrete without fly ash (50% FNS). However, after 28 days of curing, the 50%FNS-15%FA concrete provided a 5.13% higher compressive strength than 50%FNS and this is possibly due to the class C fly ash with higher calcium oxide (CaO) content used in this study which has the ability to accelerate the pozzolanic reaction of concrete. Similar results have been reported in the previous study by Ref. [46] in concrete containing high calcium FA. The 50%FNS-30%FA concrete produced lower compressive strength compared to 50%FNS-15%FA and 50%FNS at 28 days. This is due to the reduction in the cement content caused by the higher level of FA used as a replacement which led to the delay in the pozzolanic reaction of concrete.

Axial stress–strain properties

Table 4 summarizes the stress–strain behavior of the concrete mixtures while the average value of the axial stress–strain curve of each concrete mix for the three specimens is presented in Fig. 10. The control concrete (C) generally showed a lower stress–strain curve than FNS concrete as indicated by the peak stress which is mostly known as compressive strength, peak strain, modulus elasticity, and toughness. It was also observed that the FNS concrete had a steeper curve than the control concrete. However, it is important to note that all the specimens include those with FNS and the control showed a linear behavior at the initial stage till when they reached the peak stress.

Table 4 Axial stress–strain behavior of mixtures at 28 days
Full size table
Fig. 10
figure 10

Axial stress–strain curve of concrete mixtures at 28 days

Full size image

Peak stress, peak strain, and ultimate strain

Table 4 shows that the peaks stress of concrete increased by 27.77% when the sand was replaced by 50% FNS and this was associated with the increment in the particle packing due to the higher density and large particle sizes of the FNS used. Similar behavior has also been reported in past studies [21, 31, 47, 48]. Moreover, the addition of 15% FA to replace cement in the FNS concrete improved the compressive strength by 5.13% compared to the specimen without FA. This is attributed to the Class C FA with CaO content > 10% used which possibly accelerated the hydration process at an early age but the trend declined as the FA content increased to 30%. A similar observation was made in a previous study by Refs. [28, 46] in relation to the concrete and mortar containing high calcium FA at 28 days.

The peak strain was determined from the strain produced at the peak stress and the results presented in Table 4 showed that the peak strain for the control concrete (C) was recorded at 2097.5 µ and the value increased by approximately 12.24% at 50%FNS concrete, thereby, indicating its higher strength. Meanwhile, the 30%FA-50%FNS specimen had a lower peak strain which is close to the control concrete (C) and this shows that the addition of FA to the mixture causes the concrete to be more brittle. It is pertinent to note that the peak strain recorded for all the specimens is between 2000 and 3000 µ which is considered suitable for concrete structural application [49]. The ultimate strain was also determined at 80% peak stress [50] and FNS concrete specimens were observed to have a lower value than the control concrete (C) with 50%FNS, 50%FNS-15%FA, and 50%FNS-30%FA being approximately 13.12, 17.59, and 12.27% lower, respectively. This indicates that the FNS-FA concrete has a lower deformation capacity than the control concrete due to its brittle behavior.

Modulus of elasticity and toughness

The modulus elasticity values presented in Table 4 were observed to have increased significantly from 21.86 GPa to 26.01 GPa and are 15.96% higher due to the use of 50% FNS to replace sand compared to the control concrete (C). However, the value s slightly decreased as FA was added to the FNS mixture but still higher than the control concrete. These results are similar to those recorded in related previous studies such as the 11% increase reported by Ref. [21] and 14.29% improvement by Ref. [51] with the usage of 50% FNS concrete compared to control. This is associated with the high density and angular shape of FNS which enhanced the concrete density and subsequently improved the stiffness and rigidity as indicated by the high modulus of elasticity.

The toughness of concrete is defined as the area under the stress–strain curve up to the ultimate strain and the results presented in Table 4 showed that the value for the normal concrete (C) was 0.0792 MPa which increased to 0.082 for the 50%FNS concrete. This is attributed to the higher peak stress of concrete containing FNS. It was also observed that the replacement of cement with FA in the FNS concrete has little effect on the toughness.

Porosity

The ability of aggressive ions to penetrate concrete depends on the porosity and this is the reason it is an important factor influencing the durability of concrete. This study used volume permeable voids (VPV) to determine the total porosity and the results presented in Fig. 11 showed that the porosity of the concrete at 28 days of curing in tap water decreased based on the quantity of FNS and FA used. This was indicated by the 3.15 reduction recorded for the 50%FNS specimen compared to the control concrete (C) and a further 4.69 and 7.98% reduction due to the replacement of the cement by FA at 50%FNS-15%FA and 50%FNS-30%FA, respectively. This means the application of SCMs materials such as FA can improve the durability of concrete through the enhancement of the microstructures by the pozzolanic reaction. It has also been previously reported that the VPV of 50% FNS concrete with 25–30% FA was reduced by approximately 7.81–13.38% at 28 days [32, 35].

Fig. 11
figure 11

Porosity of concrete at 28 days

Full size image

It is also interesting to note that the VPV of 50%FNS-30%FA is lower than for 50%FNS and 50%FNS-15%FA which have higher compressive strengths. This is possibly due to the lower absorption capacity of FNS compared to NS as indicated in Table 2 which ensures there is no potential for absorption or release of water in the presence of FNS during the hydration process of ITZ. This is expected to support the pozzolanic reaction between cement and FA at ITZ by providing sufficient water needed for the improvement to reduce the interconnectivity of the pores. Therefore, the resistance of the concrete to chemically aggressive ion diffusion is improved. The observation is supported by the findings of previous studies that the correlation between compressive strength and porosity of concrete with SCMs is not very high [35, 52].

Accelerated corrosion test

Corrosion current with time

Figure 12 presents the average corrosion current values for each mixture with time. It was discovered at the initial time that the control concrete (C) has a higher value than the FNS concrete with approximately 38.75 mA compared to 26.63, 22.67, and 18.3 mA recorded for 50%FNS, 50%FNS-15%FA, and 50%FNS-30%FA, respectively. This implies using 50% FNS as fine aggregate and 15–30% FA as the replacement for cement reduced the permeability of concrete. Moreover, the corrosion current of all specimens first showed a descending trend before gradually increasing over the time up to the moment cracking was observed on the surface. The decreasing trend at the early time is due to the fact that the pores were being filled with NaCl ions and other substances in the NaCl solution [44, 53].

Fig. 12
figure 12

Corrosion current of the specimens with time

Full size image

The time for the appearance of the first crack on the specimens is presented in Fig. 13 and the control concrete (C) specimens were observed to have experienced it at 143 h while the time for 50%FNS specimens was at 177 h which is 1.2 times longer. This implies that the replacement of sand with 50% FNS improved the corrosion resistance of the concrete due to its higher compressive strength. A similar finding has been reported by Ref. [54] that the corrosion resistance of steel bars embedded in concrete depends on the compressive strength. Moreover, the replacement of cement with 15% and 30% of FA in the FNS concrete extended the time to 189 h and 206 h for 50%FNS-15%FA and 50%FNS-30%FA, respectively, which is 1.3 and 1.4 times longer than the time required by the control concrete (C) specimen. This is associated with the less porosity and compacted microstructure caused by the pozzolanic reaction of FA which can inhibit the penetration rate of chloride ions in reaching the reinforcement surface. These results are in line with the observation by Refs. [55, 56] that the corrosion resistance of concrete was increased due to the pozzolanic reaction of the FA with calcium hydroxide (Ca(OH)2) from cement hydration to create denser microstructure by reducing the porosity. Figure 14 further shows the relationship between time to first cracking and porosity and the findings indicated a strong relationship with a correlation of coefficient (R2) of 0.96.

Fig. 13
figure 13

Time to first cracking of specimens under accelerated corrosion test

Full size image
Fig. 14
figure 14

The relationship between time to first cracking and porosity

Full size image

Visual observation after corrosion test

Figure 15 shows the physical appearances of specimens after the accelerated corrosion test and the control concrete (C) specimen was found to have suffered more deterioration than the FNS concrete specimens in line with the corrosion current. It is pertinent to state that a higher corrosion current leads to more deterioration. It was also discovered that there was a larger crack width and massive brown rust of corrosion products on the surface of the control concrete (C) specimens. Moreover, Fig. 15 also indicates that the replacement of 15 and 30% of cement with FA in the FNS concrete provided less deterioration than specimen without FA at the same corrosion time. This is probably due to the ability of FA to fill the pores of concrete through the pozzolanic reaction which usually makes the concrete less diffusive and permeable [56].

Fig. 15
figure 15

Physical appearances of specimens after accelerated corrosion test

Full size image

The physical appearances of the steel bars after they were removed from the concrete specimens are shown in Fig. 16 and the steel bar of the control concrete (C) specimen was observed to have deeply corroded as indicated by the flaky oxide product on its surface as well as the significant loss of diameters and ribs. Meanwhile, those embedded in FNS specimens showed lesser corrosion without loss of ribs, especially in the specimen with 30% FA. This indicates the positive effect of introducing 30% FA as a partial replacement for cement which acted as the pozzolanic material by contributing to the refined pores of the concrete and reducing the risk of steel corrosion due to chloride ingress. This is similar to the findings of Sandra et al. [57] that the concrete containing 20% FA and 30% cooper slag as the replacement for cement and fine aggregates, respectively, showed better corrosion resistance than the control specimen. Al-Yaqout et al. [58] also reported the positive effect of replacing cement with 25% GGBS against corrosion in 50% RAC when compared to the normal concrete mix.

Fig. 16
figure 16

Steel surface condition after accelerated corrosion test

Full size image

Weight loss of steel bar

The weight loss of the steel bars during the accelerated corrosion test was estimated using Faraday’s law and the results are presented in Fig. 17. The control concrete (C) specimen was observed to have lost higher weight compared to FNS specimens as indicated by the sharp increase in the value after 150 h which is attributed to the presence of crack on its surface. Meanwhile, the FNS specimens exhibited a lower rate with those containing FA discovered to have the lowest value. For instance, the control specimen showed 15.9 g weight loss after 240 h and this was approximately 38.29, 43.41, and 42.44% higher than the values recorded for 50%FNS, 50%FNS-15%FA, and 50%FNS-30% FA, respectively. A similar trend was also observed using the gravimetric method in Fig. 18 with the steel bar in 50%FNS-30%FA recorded to have experienced a weight loss which is approximately two times lower than the one in the control concrete (C). It is important to note that the weight loss of steel bars is closely related to the porosity and time to first cracking of the concrete. Moreover, the values obtained from the calculated and gravimetric methods have a small difference which is less than 10% and this indicates the accelerated corrosion test method adopted in this study is reliable and similar to the previous studies by Refs [44, 59].

Fig. 17
figure 17

Estimation of the weight loss of steel bar using Faraday’s law with the corrosion time

Full size image
Fig. 18
figure 18

Comparison of the weight loss between gravimetric and calculated methods

Full size image

SEM–EDS analysis

The SEM–EDS analysis of control (C) and 50%FNS-30%FA specimen at 28 days is illustrated in Fig. 19. The main hydration products of the control specimen consist of calcium hydroxide (C–H), calcium silicate hydrate (C–S–H), and ettringite, which is commonly found in the Portland cement–based material. The distribution of hydration product crystals is loose, and there are several large pores. Cracking was also observed in the control specimen. That cracking is caused by the excessive hydration heat of Portland cement, and those cracks in turn make the matrix less compact [29, 60, 61].

Fig. 19
figure 19

The SEM–EDS of control (C) and 50%FNS-30%FA specimens at 28 days

Full size image

The FA particles in the 50%FNS-30%FA specimen are evenly distributed and impeccably embedded in the matrix where the surface is covered with C–S–H. That surface covering is possibly the result of a pozzolanic reaction between FA and CH. Furthermore, long, linked crystals of ettringite can be seen in the matrix and contribute to reduced pores. The formation of C–S–H and ettringite hydration products at an early age depends on the content of CaO and SO3 as well as on the size of FA particles. FA particles contribute to the development of concrete’s compressive strength [61,62,63]. These results are supported by the EDS analysis, where the ratio between Ca/Si in the 50%FNS-30%FA specimen was lower than that in the control specimen due to the pozzolanic reaction. During the hydration process, additional silica from FA reacts with CH to form additional C–S–H, thus reducing the porosity. Moreover, the peaks of Al dan S elements in the 50%FNS-30%FA being higher than those in the control specimen indicates the growth of ettringite formation. The results of the SEM–EDS analysis reveal that FNS-FA concrete’s compressive strength and corrosion resistance are higher than those of the control concrete at 28 days due to the pozzolanic reaction of FA resulting in the former having less porosity.

Conclusions

The corrosion resistance and compressive behaviors of FNS aggregate with FA were studied experimentally and the following conclusions were drawn:

  1. 1.

    The replacement of sand with 50% FNS and the inclusion of 15 and 30% FA as a cement replacement increased the workability of concrete due to the combined low absorption capacity of FNS which produces free water and the ball-bearing effect of FA. Similarly, the density of FNS concretes increased due to the higher specific gravity of the material compared to NS.

  2. 2.

    The compressive strength of FNS concrete was found to be higher than control concrete for all curing ages. Moreover, the addition of 15% FA to the FNS concrete mixture also improved the compressive strength at 28 days by 5.13 and 34.32% compared to the 50%FNS and control concrete (C), respectively, due to the pozzolanic reaction of FA.

  3. 3.

    Regarding the axial stress–strain behaviors, 50%FNS-15%FA exhibited higher peak stress, peak strain, modulus of elasticity, and toughness than control concrete (C) but had a lower ultimate strain. This indicates the concrete became more brittle due to the presence of FNS and FA.

  4. 4.

    The inclusion of 30% FA in the concrete mixture containing FNS reduced porosity by 7.98% compared to the control due to the pozzolanic reaction of FA which contributed to the refinement of the pores and enhanced the microstructure of the concrete.

  5. 5.

    The inclusion of 30% FA also improved the durability of FNS concrete against corrosion as indicated by the time to first crack and weight loss of 50% FNS-30% FA concrete which was improved by 1.4 and 2.2 times, respectively, than the control concrete.

These findings showed that concrete with 50% FNS as sand replacement and 30% FA as cement replacement is a viable option to develop a low-carbon concrete with a better performance against corrosion in structures exposed to a chloride environment.