Study on the Durability of Road Concrete with Blast Furnace Slag Affected by the Corrosion Initiated by Chloride

In order to assess the resistance of the road concrete to the penetration of the chloride ions, we used the rapid test RCPT in compliance with ASTM C1202. But, in order to increase the reliability of the rapid test of assessing the resistance of chloride ions’ permeability through measurements of the electrical conductivity, after ﬁnishing the RCPTtest, we have also measured the depth of the chloride front migrating in the concrete. Road concrete mixtures were prepared in which blast furnace slag (GGBS) was incorporated as a binder and blast furnace slag (ABS) as crushed slag aggregate to the size of 0/4mm. The eﬀect of the blast furnace slag on the RCPT results was investigated and correlated with porosity trends, carbonatation layer depth, and with the compression strength values, by comparison with the level of results obtained on concrete mixtures made with conventional materials. The measurement of the migrated chlorine front in the concrete together with the determined physical and mechanical characteristics conﬁrms the eﬃciency of the blast furnace slag incorporated in the concrete and supports the reliability of the rapid penetration test of chlorine ions RCPT using the procedure from ASTM C 1202.


Introduction
An important feature that affects the durability of road concrete is the corrosion of the reinforcements embedded in the concrete. Corrosion degradation attributed to the reinforcement in the concrete structure was estimated to represent 25% of the causes that influence the durability of concrete [1,2]. In the case of road pavements, the steel reinforcements are incorporated either at the longitudinal contact joints or at the expansion joints or on the entire surface of the slabs in the case of continuously reinforced road pavements. Usually, the concrete reinforcement is protected by an inert and dense layer, of nanometric thickness, of iron oxide, with a passive role, which surrounds the steel reinforcement, being supported by the alkalinity of the cement paste and by the thickness of the covering layer. But chloride ions can destroy the passive layer surrounding the steel reinforcement by increasing the chloride concentration, activating chemical corrosion [1,3].
When the passivating protective layer of the reinforcement is pierced by chloride ions, spot corrosion (pitting) is initiated. Local craters are formed, whose depth advances and which are partially filled with an unstable corrosion product of iron, ferrous oxide, friable, which makes it difficult to assess the degree of corrosion. At the same time, the corrosion products of iron with a volume much higher than the volume of iron (expansion factor of 2-10 times) will induce local stresses in the surrounding cement matrix, thus initiating cracks that will allow the penetration of chloride ions more easily and the advancing of the corrosion [4]. In addition to the presence of a significant amount of chloride ions at the reinforcement level, the destruction of the passive layer also occurs due to carbonation which reduces the pH of the water in the concrete pores to the unprotective level [5][6][7][8][9][10][11][12]. e carbonatation is a chemical reaction of portlandite Ca(OH) 2 and hydrated calcium silicate (CSH) with carbon dioxide (CO 2 ) present in air or water and forms calcium carbonate (CaCO 3 ) and water (H 2 O). At a certain moisture content, the carbon dioxide (CO 2 ) enters the pores of the concrete through gaseous diffusion. Normally, the high natural alkalinity conditions present inside the concrete protect the reinforcement against corrosion reactions. However, carbonatation lowers the pH value from typical values (12 ÷ 13) to less than 9 of the pore solution and destroys the passivity of the built-in reinforcement bars, triggering the corrosion process [13][14][15][16]. For permeable concretes, the increase of CO 2 concentrations in the external environment increases the carbonatation rate [17], but a compact cement matrix prevents the diffusion of carbon dioxide [18] and does not endanger the decrease of the compressive strength of the concrete [19]. Tuutti has developed a model for predicting the life of steel reinforcement, according to which the corrosion process distinguishes two stages: the first stage of corrosion initiation and the second stage of corrosion propagation. From the onset of the corrosion the duration until the degradation is influenced by relative humidity, ambient temperature, and oxygen level in the air [20,21]. e period of onset of corrosion will not immediately affect the performance of the reinforcement, but the period of corrosion propagation must be taken into account because the performance of the entire reinforced concrete element can be significantly affected [21,22]. e propagation of corrosion leads to accelerated corrosion, which is manifested by rust staining, cracking of the reinforcement coating, longitudinal cracking in the structure, chipping, and delamination of the reinforcement coating [20,23]. e research studies show that there are different types of calcium nitrite-based additives for concretes that generate chemical and physical inhibition mechanisms to keep the layer around the reinforcement inert [1,24,25]. Using the half-cell-potential method (HCP) which quantifies the progression of corrosion activity by the measured potential difference between the tested sample and a reference electrode, Pei et al. showed that an increase in the w/c ratio in the mortar from 0.5 to 0.7 led to a major decrease in the time until the reinforcement corrosion started.
However, the literature also indicates that this method (HCP) does not assess the actual corrosion rate, but a comparative evolution with a control, which suggests other more advanced techniques such as determining the resistance by linear polarization (LPR) [26]. Another way of reducing the risk of corrosion of concrete reinforcement can also be obtained by embedding optimal dosages of additional cementitious materials (SCM), such as fly ash, blast furnace slag, and ultrafine silica [27][28][29][30][31][32]. It is known that ultrafine silica provides the highest protection of steel reinforcement due to the small particle size (approximately 0.1 µm), which produces a pozzolanic reaction with portlandite (CH) forming a good bond of the interfacial transition zone (ITZ) between the cement paste and the aggregate [1,33].
In 1970, the Federal Highway Administration (FHWA) was concerned with discovering a rapid method of assessing the penetration of chloride ions into the reinforcement coating due to the need to use thawing salts in road coverings. e results from the methods existing at that time were specific to the evaluation of the permeability of concrete to water and air, without presenting a correlation with the methods for assessing the diffusion of chloride ions [1,34]. In 1981, a final version was developed used in the laboratory to determine the chloride penetrability class of concrete, and, in 1991, the RCPT test was documented in ASTM C1202 [35]. e concept of the RCPT method, in compliance with ASTM C1202, is based on the mobility of chloride ions forced to enter the concrete by inducing an external stress applied directly to the concrete surface. e migration of chloride through concrete is a slow process, but the researchers found that when an electric current is applied to a concrete test tube, the speed at which chlorides migrated into concrete was accelerated, which is why the test is currently commonly used [1]. However, the RCPT test is not recommended for use in concrete mixtures that include dispersed reinforcing fibers or corrosion inhibitors such as calcium nitrites, as they influence the electrical conductivity and artificially increased results are obtained. e reliability of the RCPT test method has been criticized for various shortcomings, also in the use of additional materials with cementitious characteristics such as ultrafine silica [1,36]. Currently, there is still controversy regarding the influence of heat generated by the electric current applied on the results, respectively, on the migration of chloride ions, due to the fact that it cannot identify only them [37]. erefore, to minimize the effect of temperature on electrical conductivity, previous studies have suggested additional RCPT measurements after short periods of 1 minute and 30 minutes, as it has been observed that no large temperature differences are recorded during this interval. Multiplying the measurements at 30-minute intervals allows extrapolated calculation of passing charges and resistivity. e resistivity of the concrete calculated from the RCPT test can estimate the duration of the corrosion propagation stage until the concrete yields [20,[38][39][40]. Feliu et al. [41,42] made a correlation between the measured resistivity of the concrete and the corrosion rate at the level of the embedded reinforcement, these evolving inversely proportionally. is was also confirmed by Glass et al. [41,43], which showed that the resistivity of mortar is dependent on the relative humidity of the environment, while López et al. [41,44] showed that the volume of the concrete pores influences its resistivity and, therefore, the corrosion rate of the reinforcement. Another parameter for assessing the penetration of chloride ions is the diffusion coefficient of chloride ions, which can be evaluated by the NT Build 492 test [45], a test method used in Finland to determine the migration of chloride ions in an unstable state. is method is an alternative to ASTM C1202, differing in exposure time (24 to 96 hours) and applied voltage (30 V). e diffusion coefficient results from the values of chloride penetration, sample thickness, exposure time, voltage, and average sample temperature [46]. e diffusion of chloride ions from the concrete determines the stage of corrosion initiation while the resistivity of the concrete shows the speed of corrosion propagation [1,47]. e study by Bassuoni et al. showed that the determination of the migration coefficient of chloride ions can eliminate the uncertainties in the rapid RCPT test of electrical conductivity. Chloride migration coefficients are sensitive to the physical characteristics of the pore structure, and the passing charges are mainly influenced by the ionic concentration of the pore solution. e materials incorporated in concrete, such as ultrafine silica and CNI calcium nitrite corrosion inhibitors, can influence the chemical elements of the pore solution, so the RCPT rapid test method proved to be insufficient and was supplemented with the calorimetric method [1]. McGrath and Hooton [48] studied the influence of the potential difference on the diffusion coefficients of concrete chloride. e diffusion coefficients of chlorine ions in the unstable state were calculated by applying the Nernst-Planck equation, using the data from the accelerated chloride penetration tests under the conditions of applying a potential difference between 6 and 30 volts. e conclusions showed that the error of the results increased as the nominal potential decreased. Soleymani and Mohamed [49] conducted a laboratory study to estimate the corrosion of steel reinforcement for two types of concrete, ordinary and high performance. e results indicated that Tafel interpretation (TP), linear polarization resistance (LPR), semicellular potential measurement (HCP), and determination of chloride content by other methods obtained the same level of corrosion activity in only 24% of the samples. e study shows the need to combine these methods in order to increase the accuracy of the assessment of the degree of the concrete reinforcement corrosion. Coşa et al. tested by two methods the penetration depth of chlorine ions in the self-compacting concrete made with the addition of fly ash or limestone filler. e study shows that using the colorimetric method with silver nitrate solution (AgNO 3 ) results were obtained for the depth of chloride penetration, but additionally using the scanning electron microscopy (SEM) method, the concrete chloride concentration was determined, and the results for the penetration depth of chlorine ions were more precise. e advantages of using electron microscopy have proven to be remarkable, but not in the sense of undermining the efficiency and practicability of the colorimetric method [50]. Yeau and collaborators showed that the concrete with substitution of 40% blast furnace slag (GGBS) relative to the cement mass and even above this level showed higher corrosion resistance compared to cement concrete without (GGBS), after 56 days of immersion in water [27]. Sancharoen and Uomoto [51] obtained higher corrosion resistance of concrete reinforcement with (GGBS) compared to Portland cement concrete. However, with the cyclic and simultaneous exposure of concrete with (GGBS) to chlorine ion corrosion and carbonation, the corrosion resistance decreased compared with conventional concrete, so it was suggested that their durability be carefully analysed. Noushini et al. [52] studied the diffusion resistance of chloride on lowcalcium fly-ash-based geopolymer concrete at different temperatures and hardening times. For this concrete, the highest chloride diffusion coefficient was obtained by heat treatment at 60°C for 24 hours. e chloride diffusion coefficient decreased with the increase of the tartar temperature from 75°C to 90°C, but the values recorded were higher compared to concrete based on (OPC). Another observation of research shows that conventional concrete has a higher chloride binding capacity compared to geopolymer concrete. Most physically bound chlorides did not contribute to the corrosion of steel reinforcements, but were attacked by free chloride ions from concrete. e purpose of this work was to analyse the corrosion resistance of the reinforcement embedded in the road concrete prepared with blast furnace slag, started as a result of the penetration of chloride ions. e reliability of the rapid chloride ion penetration test, RCPT with ASTM C 1202, was also evaluated by supplementing the current method with the colorimetric method, aiming to evaluate the depth of the front and the migration coefficients of chloride ions. At the same time, the content of permeable pores was evaluated, which was a useful index for correlating with the migration coefficients of chlorine ions in concrete. In order to strengthen the results obtained by the rapid RCPT test, the concordance between the compressive strengths and the resistivity of the concrete to the penetration of chlorine ions were also analysed. e results of this study will complement the mechanical characteristics of concrete obtained by the authors of previous investigations. Based on sustainability criteria, the present study aims at selecting the optimal proportions of locally sourced blast furnace slag, incorporated in the composition of road concrete. is paper highlights both granulated and ground blast furnace slag (GGBS) and crushed aggregates (ABS) obtained from ungranulated blast furnace slag. e controlled application of blast furnace slag in the composition of road concrete would be a start at the local level, helping to reduce the need for cement in concrete production, respectively, reducing CO 2 emissions globally.
is reduction is necessary taking into account statistics showing that cement production is responsible for around 7.4% of global carbon dioxide emissions [53]. Also, the concrete in this study would contribute to the reduction of the consumption of aggregates from natural reserves, estimated Advances in Civil Engineering worldwide at a consumption of over 20.000 million tons/ year [54,55].

Materials and Methods
2.1. Materials. CEM I 42,5R cement, according to EN 197-1 : 2011 [56], and granular blast furnace slag from local sources were used as binder, having characteristics in compliance with the limits of standard EN 15167-1 : 2007 [57]. Before being placed in concrete, the granulated blast furnace slag was mechanically ground to a granulation of less than 63 µm, a size that supports its activity index [58][59][60]. e value of the specific surface of the cement was 4385 cm 2 /g, with a specific weight of 3.00 g/cm 3 , and that of the ground slag of 3775 cm 2 /g, with a specific weight of 2.77 g/cm 3 ; these characteristics were determined in the laboratory of the Technical University of Cluj-Napoca. e ground and crushed slag had the sum of calcium, magnesium, and silicon oxides of 84.05% and a 28-day activity index of 74.97%, above the level of 70% in compliance with EN 15167-1 : 2007. e content of chloride within the granulated blast furnace slag (GGBS) was 0.004%, below the limit of 0.1% established in compliance with EN 15167-1 : 2007. e artificial aggregate from ungranulated blast furnace slag (ABS) is the result of a byproduct of the steel industry in Galati and has replaced in various proportions the natural aggregate with a grain size of 0/4 mm. e mineral phases of granulated (GGBS) and nongranulated (ABS) blast furnace slag were evaluated using X-ray diffraction (XRD) and polarized light optical microscopy methods. e X-ray diffraction (XRD) investigations were performed with the equipment from the laboratory of the Department of Geology of Babeş-Bolyai University of Cluj-Napoca, using a Bruker D8 Advance diffractometer with Cu Kα radiation (λ � 1.541874Å), Fe 0.01 mm filter, and a one-dimensional LynxEye detector. e microscopic images were taken using a Nikon Optiphot T2-Pol microscope with polarized light.
In granulated blast furnace slag (GGBS), the main mineral phase identified is merwinite, followed by magnesian gehlenite and carbon. Additionally, the glass appears in the amorphous phase (Figures 1 and 2). In the crushed aggregates from ungranulated blast furnace slag (ABS), the X-ray diffraction spectrum indicated as the main phase magnesian gehlenite and subordinate clinoferrosilite and aluminum silicate (Figures 3 and 4). From the results of the XRD measurements, the amorphous phase was estimated at 80% for granulated slag (GGBS) and at 100% for nongranulated slag (ABS) in the crystalline phase. e proportions of aggregates used were 32% natural sand (size 0/4 mm) and 68% coarse aggregate, of which the 8/4 mm sort was crushed river gravel and the 8/16 mm and 16/25 mm sorts were crushed chippings from the quarry. For the selection of the aggregates used and the particle size curve of the total mixture of aggregates, the conditions of the national standard NE 014 : 2002 were taken into account [61], and the physical and mechanical properties were in compliance with EN 12620 : 2003 and EN 12620 + A1 : 2008 [62]. MasterGlenium SKY 527 superplasticizer additive and Master Air 9060 air entraining additive, produced by Badische Anilin und Soda Fabrik (BASF), were used in the preparation. e preparation water, in compliance with EN 1008 : 2003 [63], was taken from the supply system of the city of Cluj-Napoca.

Preparing the Road Concrete Mixtures.
In standardized European models, there are no predictions of the service life of reinforced concrete structures exposed to corrosion caused by chloride attack. Also, the type of cement is not considered as a parameter in the evaluation for chlorideinitiated corrosion. However, Jan Bijen proposes some limits for concrete cement with blast furnace slag: a maximum corrosion rate of 5 µm Fe/year for any point in the reinforcement, a chloride content of up to 2% of the mass of the slag cement, a w/c ratio of 0.45 which together with the reinforcement coating, will not endanger the performance of the structure [22]. e design parameters of the road concretes in this experiment followed the minimum requirements for exposure class XD3 (exposure to corrosion D due to chlorides of nonmarine origin, alternating moisturedrying, level 3, in compliance to [64,65]). us, the max w/c ratio (0.45), minimum strength class C35/45, and minimum cement dosage 320 kg/m 3 are provided. In the general study related to the durability of road concretes with blast furnace slag, two series of concretes with the same amount of (GGBS) and the same proportions of crushed aggregates from blast furnace slag (ABS) were poured. is study is part of the second batch series and comprises five compositions, coded S 360, S 414, S 54/20, S 54/40, and S 54/60, having the quantities of raw materials used according to Table 1. e first two control mixtures were made with Portland cement and natural aggregates, and in the next three mixtures, we used granulated and grounded blast furnace slag under 63 µm (GGBS) and aggregates from air-cooled crushed blast furnace slag (ABS) at the dimension of 0/4 mm in different proportions: S 360, control mixture with 360 kg/m 3 dosage Portland cement and natural aggregates S 414, control mixture with 414 kg/m 3 dosage Portland cement and natural aggregates S 54/20, 360 kg/m 3 (cement) + 54 kg/m3 (GGBS) and 20% (ABS)_0/4 mm + 80% natural sand S 54/40, 360 kg/m 3 (cement) + 54 kg/m 3 (GGBS) and 40% (ABS)_0/4 mm + 60% natural sand S 54/60, 360 kg/m 3 (cement) + 54 kg/m 3 (GGBS) and 60% (ABS)_0/4 mm + 40% natural sand e water content and additives were adjusted in the second series of casting so as to ensure a consistency of at least 30 mm, in compliance with NE 014:2002 [61]. Due to a water adjustment error in the preparation of the S 360 mixture, the w/b ratio limit of 0.45 was exceeded. e tests, trials, and analysis of the results were continued for the mixture of S 360 but for the compositions with built-in blast furnace slag only those obtained from the control mixture S 414 were considered as reference results. 4 Advances in Civil Engineering    Advances in Civil Engineering 5

e Rapid Chloride Penetrability Test
RCPT. For the RCPT test, cylindrical specimens with a diameter of 100 mm and a length of 200 mm were casted and covered with polyethylene foil, kept in the air for 24 hours at a temperature of (20 ± 2)°C and relative humidity of about 40%, after which they were stripped and immersed in water at a temperature of (20 ± 2)°C. e curing period of the test specimens in water was extended from 28 days to the test age of 150 days in order to observe the evolution of the durability characteristics of concrete (mechanical strength and resistance to chemical aggression) due to the continued hydration of slag particles over time.
ese characteristics are highlighted only after a longer curing period, such as at least 90 days. It is known that blast furnace slag is characterised by a slower hydration rate and a lower hydration heat compared to Portland cement [66][67][68][69][70]. Two days before the test, the 100 mm diameter cylindrical specimens were removed from the water, and a 50 ± 3 mm thick strip from the top of the specimen was cut from each specimen to simulate the reinforcement coating. A water-impermeable epoxide was applied to the side of the specimens, and the upper and lower circular faces were left free. After drying the epoxy product, the specimens were kept in polyethylene foil until the start of the test to prevent evaporation of water. e preconditioning of the samples for the RCPT test was performed in a dry vacuum environment for 3 h, after which the samples were covered with boiled water and cooled to room temperature. e container with the immersed samples was emptied, thus maintaining the samples for 20 hours for complete saturation. e preconditioning of the samples complied with the specifications of ASTM C1202 [35], except that the container was emptied with a standard apparatus in the absence of a vacuum saturator equipped with a desiccator, vacuum pump, and manometer. e electrical charges passed through the concrete sample for 6 hours, Q (expressed in coulombs), was measured with the Controls device, model 58-E0052 to determine the capacity of the concrete in case of penetration of chloride ions, applying a potential difference of 60 V between the ends of the specimen (Figures 5 and 6).
e ends of the sample were fixed in the two test cells after which the area between the sample and the test side in which electrolytic solutions were introduced was waterproofed, in one part sodium chloride solution, NaCl, 3%, and in the other solution of 0.3% sodium hydroxide, NaOH, in dilution with distilled water. e cell containing 3.0% NaCl was connected to the negative pole (cathode); then, the cell containing 0.3% NaOH was connected to the positive pole (anode). e air temperature around the samples was maintained in the range (20 ± 0.5)°C, and the temperature sensor that monitored the solution was placed in the cell with sodium chloride. e standard ASTM C1202 [35] provides a correlation between the electrical charge passed at 6 h through a sample with a diameter of 95 mm and a thickness of 50 mm and between the penetrability of chloride ions, presented in Table 2.
e evaluation of the resistance of concrete to the penetration of chloride ions determined by the ASTM C1202  Advances in Civil Engineering procedure can be sustained by determining the resistivity of the concrete. Julio-Betancourt and Hooton [71] recommended calculating the electrical resistivity of concrete 1 minute after the start of the test to avoid the influence of temperature on the mobility of chloride ions. Bassuoni et al. [1] noticed that at 30 minutes the maximum temperature gradient did not exceed 6°C, resulting in comparable values of electrical resistivity. In this study, the temperature difference recorded from 1 minute to 30 minutes was up to 3°C, which allowed the calculation of the resistivity of the concrete at 30 min, as follows: where ρ � resistivity (Ωm) at 30 minutes, V � applied voltage (V), A � transversal area of the specimen (m 2 ), I � current intensity (A) at 30 minutes, and L � thickness of the specimen (m). e electric heating of the solution results in increased ion mobility in the pore solution. To avoid this effect, the passage of the electrical charges at 30 min was recorded and extrapolated linearly to estimate the passing charges at 6 h. e electrical resistivity of the concrete was used to determine the corrosion rate of the reinforcement Cr (µm Fe/yr), calculated with equation (2) [1,47,72] and classified as in Table 3 [20,73].

Advances in Civil Engineering
Immediately after the completion of the RCPT test, similarly to the tests performed by Bassuoni et al. [1] on the same samples, the migration coefficient of chloride ions was determined by colorimetric method, in compliance with the NT Build 492 Test [45]. e 50 mm thick concrete strips were split axially and sprayed with 0.1 mol/dm 3 (0.1 M) silver nitrate AgNO 3 solution, forming a white precipitate of silver chloride on the surface of the specimen after 15 minutes. e white precipitation of silver nitrate was formed in the part of the face where chloride ions were present. e depth of chloride penetration (x d ) for each sample is the average value of seven measurements made at different positions along the width of the sample. Chloride ion migration coefficients (D) were calculated with the Nernst-Planck equation as follows: where D � non-steady-state migration coefficient (×10 −12 m 2 /s), V � applied voltage (V), T �average value of initial and final temperatures in the anolyte solution, NaCl (°C), and L � thickness of the specimen (mm).

Determination of the Depth of the Carbonatation Layer of the Hardened Concrete.
e depth of the carbonatation layer on the surface of the reinforced concrete was determined at the age of 28, 56, and 150 days, in compliance with the standard SR CR 12793 : 2002 [74], on fragments of prismatic specimens of 150 × 300 mm (obtained from prisms with a length of 600 mm after the bending tensile test performed at the age of 28 days). Each of the two remaining fragments was used one for the carbonatation test and one for the 150-day compressive strength test. After the 28 days of hardening in water, the prism halves used for the carbonatation test were placed up to 150 days in the climatic chamber at a humidity of (65 ± 5)%, a temperature of (20 ± 2)°C, and a normal carbon dioxide level. To check the ventilation of the air in the climate chamber, the average evaporation rate of water at 7 days from the free surface of a 30 mm glass was evaluated; the result was 44.59 g. By the age of 56 days, the specimens were immersed in water for 6 hours after every 7 days, in accordance with the climatic conditions related to the exposure class (CX3) in an environment with alternating humidity-drying. e immersion in water for 6 hours was repeated before the test performed at the age of 150 days. At each test term, a strip of approximately 50 mm was made for each mixture and the solution of 1% phenolphthalein in 70% ethyl alcohol was sprayed on the freshly split face. e carbonatation depth d k (mm) was determined by measuring the distance from the outer surface of the concrete to the edge of the red-purple coloured region.

Determining the Density, Water Absorption, and Absorption of the Pore Content of the Hardened Concrete (1) Apparent Density on Specimens in Natural Humidity
State. e cylindrical fragments remaining after cutting the cylindrical strips used in the RCPT test determined the density and content of permeable pores. e apparent density, ρ a , at the age of 150 days was determined by measuring the volume of water displaced in cylindrical fragments with a diameter of 100 mm according to the reference method of EN 12390-7 : 2000 [75], respectively, as follows: where m � the mass of the specimen in natural humidity state (g); V a � apparent volume of the specimen (cm 3 ); m a � mass of the specimen saturated measured in air (g); m st � mass of the immersed caliper (g) (was 604 g), m w � mass of the saturated specimen, immersed in water (g); and ρ w � water density, at 20°C, with a calculus value of 1 g/ cm 3 .
(2) Density after Immersion and Boiling, Water Absorption, and Absorption of Permeable Pore Content. For density after immersion and boiling, water absorption and permeable pore content were obtained according to ASTM C 642-2006 [76], a method also used in the study by Mardani-Aghabaglou et al. [77]. e ASTM C 642-2006 method allows the use of samples having a minimum volume of 350 cm 3 or a specific weight of approximately 800 g. e water-saturated specimens with the dry saturated surface were weighed in air (b) and then dried in an oven at 100-110°C until a constant mass (a) was reached. e specimens were placed in a container, covered with tap water, and boiled for 5 hours. ey were then allowed to cool to room temperature (22 ± 2°C). e moisture from the surface was removed by swabbing with a towel and the dry mass on the surface of the immersed and boiled samples was determined (c); the mass was measured in air. e apparent mass (d) after immersion and boiling was measured in water with the hydrostatic balance. Absorption after immersion (m 1 ) and absorption after immersion and boiling (m 2 ) were calculated with the following equations: e density in dry state ρ 1 and the density after saturation and boiling ρ 2 were calculated with the following equations: and for water density ρ w we used the value of 1 g/cm 3 . e content of the permeable pores (P 0 ) was calculated using the following equation: in compliance with ASTM C 642-2006 [76]. e determination of the electric charge (Q), the depth of penetration of chloride ions (X d ), the migration coefficient of chloride ions (D), and the content of permeable pores (P 0 ) were determined from the average results of 3 cylindrical specimens with a diameter of 100 mm at the age of 150 days.

Determination of Compressive Strength at the Age of 150 days.
e determination of the compressive strength, according to EN 12390-3 : 2002 [78], was performed on a prism fragment with a size of 150 × 150 × 300 mm resulting from the bending stress test at the age of 28 days. Preservation of the specimens was performed in the climatic chamber at temperatures of (20 ± 2)°C and relative humidity of (65 ± 5)%, until the age of 150 days.

Rapid Chloride Penetrability Test (RCPT).
e electrical passing charges (Q) after 6 h, according to the RCPT rapid test, were in the range (1255 ÷ 3089)°C and the extrapolated passing charges from the recordings at 30 min were between (1262 ÷ 2798)°C, which in in compliance with ASTM C1202. Table 2 corresponds to a low, respectively, moderate chloride ion penetrability class. e report between the recorded data sets and the extrapolated data sets, Figure 7, is in the range (0.80 ÷ 0.91); they are generally lower for the extrapolated values. e exception is the mixture S 54/20 in which the ratio between extrapolated and recorded passages was 1.01.
is could indicate that for the mixture S 54/20 the effect of temperature on the passing charges had no influence. Table 4 shows the electrical resistivity at 30 minutes from the beginning of the RCPT test (ρ), average temperature of the sodium chloride solution (T med ), depth of penetration of chloride ions (X d ), nonequilibrium migration coefficient (D), and corrosion rate (C r ).
As seen in Figure 7, the electrical passing charges (Q) for compositions with blast furnace slag are lower than those of the control composition S 414 made with conventional materials. is is attributed to the increased resistivity as shown in Table 4, respectively, to the reduction in electrical conductivity of the pore solution that could be induced by blast furnace slag powder, consistent with the results obtained in the study [1]. But of the blast furnace slag-free compositions, the lowest passing charges were recorded for S 414 which had a w/c ratio of (0.42), lower than those in the S 360 mixture at which the w/c ratio was 0.46. is shows that an increased w/c ratio can lead to a decrease in the resistivity of the concrete to the penetration of chloride ions. Figure 8 shows the evolution of the results for the corrosion rate. e lowest values of this indicator are observed in the mixtures S 54/20 and S 54/60.

Determining the Chloride Ions' Migration Coefficients (D) and the Depth of Penetration of Chloride Ions.
e average penetration depth of chloride ions measured and the migration coefficients calculated according to equation (3) have values shown in Table 4. Figures 9 and 10 illustrate the chloride front for compositions S 360 and S 54/20, thus observing, visually, the depth of penetration of chloride ions into the concrete. Although these physical measurements are less dependent on the electrical conductivity of the concrete, the results have a common tendency with the results of the RCPT test. us, it was found that the average penetration depth of chloride ions of 5.20 mm registered in the composition S 54/20 and 9.43 mm in the composition S 54/60 resulted in migration coefficients (D) with values of, respectively, 4.41 and 8.42 × 10 −12 m 2 /s. ese values were the lowest, similar to the results obtained after the rapid RCPT test.

Determination of the Depth of the Reinforced Concrete Carbonatation Layer.
e carbonatation profile was determined using phenolphthalein as an indicator solution. One hour after spraying the surface, the noncarbonatated area will remain purple, with a pH higher than 9, while the carbonatated area will remain colourless, in compliance with SR CR 12793 [74]. From the visual analysis of the tested samples at the ages of 28, 56, and 150 days, it was found that all the tested samples did not lose their basicity, proved by maintaining the purple color on the entire sprayed surface. Figure 11 shows images with the appearance of the samples one hour after the application of the phenolphthalein solution compared to the freshly sprayed samples. Only a decrease in the intensity of the purple color is observed due to the drying of the solution on the sprayed surface. In this experiment, the depth of the carbonatation layer, d k , was zero.

Determining Density, Water Absorption, and Impermeable Pores Content Absorption for Hardened Concrete.
Following the determinations performed on fragments of cylinders with a side of 100 mm, the content of permeable pores (P 0 ), density in the dry state (ρ 1 ), density after saturation and boiling (ρ 2 ), apparent density in the state of natural humidity (ρ a ), absorption after immersion (m 1 ), and absorption after immersion and boiling (m 2 ) were obtained, with results shown in Figures 12 and 13. e content of permeable pores, or the apparent porosity, was in this experiment a useful index of correlation of the pore structure with the migration coefficients of chloride ions in concrete, in accordance with the specifications in the specialty literature [1]. e values obtained for the apparent porosity are in the range (5.63 ÷ 6.79)%, respectively, being lower than the limit of 10% above which the concrete is considered porous [79]. Gel pores (CSH) were not monitored in this experiment because they do not have a major contribution in the transport mechanism of chloride ions in the concrete matrix, their role being the discontinuity of capillary pores [1]. ese findings can be brought in line with the specifications in the specialty literature [22,80].
us, it is specified that the percentage of gel micropores (<30 nm) in slag concrete is

Advances in Civil Engineering
higher than that of Portland cement concrete, and the capillary porosity is lower. e cause of this difference, suggested by Bakker, is due to a different precipitation of the gel (CSH), which leads to a relative decrease in capillary porosity. e gel (CSH) in the slag cement does not precipitate directly on the cement particles, but in the space between the granulation of the clinker particles and the blast furnace slag. is precipitation results in a better gel distribution (CSH), which leads to a lower diffusion and permeability of chloride ions to furnace slag concrete compared to Portland cement concrete [22,81]. Comparing the diagrams in Figures 12 and 13, it is observed that the apparent density of the mixtures has an inversely proportional evolution with the content of permeable pores, respectively, with the water absorption. Respectively, for compositions S 54/20 and S 54/60 with a lower content of permeable pores, the density of concrete is higher. e compositions S 414 and S 54/40 showed, respectively, a higher porosity and a lower density compared to the other compositions. For all three furnace slag compositions, the determined permeable pore content is lower compared to the control mixture S 414. e highest permeable pore content was recorded for the S 360 mixture. is was due the increased water/binder ratio (0. 46).
It is observed that the evolution of the permeable pore content (Figure 13) has the same upward trend with the migration coefficient of chloride ions (D) ( Table 4). is indicates, similar to the specifications in the literature [1], a direct relationship between chloride migration coefficients and permeable pore content. e standard error of the estimate for D is 2.02 × 10 −2 m 2 /s, approximately 12.59% of the average value of the migration coefficients of 16.05 × 10 −2 m 2 /s, as seen in Figure 14.
As shown in the work of Jan [22], the high level of concrete quality is influenced by a low porosity, due to a lower content of capillary pores and a smaller network of interconnected pores. Such porosity was also recorded in mixtures S54/20 and S54/60, which led to a reduced transfer of chloride ions through the concrete matrix.

e Evolution of Mechanical Strength, Concrete Resistivity, and Depth of Penetration of Chlorine Ions in Concrete.
As seen in Figure 15, the results of compressive strengths at the age of 150 days evolved similarly to those at the age of 28 days in the previous work of the authors [82]. Analyzing Figures 12 and 15, it results that the mechanical strengths at the age of 150 days are proportional to the values of the apparent density for all the mixing performed. e mechanical strengths of blast furnace slag mixtures were higher with values in the range (0.63 ÷ 3.08)% compared to the control mixture S 414, attributed to a higher compactness and a higher density, respectively. However, analyzing the evolution of the resistivity of concrete to the penetration of chlorine ions in Figure 16, increases between (56.24 ÷ 88.69)% are observed, with visibly higher values for blast furnace slag vs. (versus) control mixture S 414. e concentration of 3% sodium chloride (according to the RCPT rapid test) produced a lower impact on blast furnace slag mixtures, i.e., a lower content of free chlorine ions, which can be due to the increased chemical binding capacity of chlorine ions, a kinetic mechanism presented in the paper [52]. Further, in the diagram in Figure 16, the results obtained by the colorimetric method for the penetration depth of chlorine ions were completed. As expected, the depth of chlorine penetration into the concrete was inversely proportional to the resistivity and lower with values between (39.74 ÷ 77.83)% compared to the control mixture S 414.
All these results led to eliminating uncertainties related to electrical loads registered in this study, using the rapid RCPT test.

Conclusions
e effect of blast furnace slag on the corrosion resistance of the reinforcement was studied by the RCPT rapid test according to the ASTM C1202 standard supplemented with the colorimetric method from the NT Build 492 test. ese methods allowed the determination of the concrete resistivity and the corrosion propagation speed with the rapid RCPT test and identification of the depth of penetration and migration coefficients of chlorine ions by colorimetric method. e results were analysed and correlated with the porosity trends, with the depth of the carbonatation layer, with the compressive strengths, by comparison with the level of results obtained on the control mixture made with conventional materials. Based on the results of this study, the following conclusions can be presented: (i) e electrical pass-through charges (Q) obtained by the RCPT rapid test on slag compositions led to their classification in the low-penetrability class and for the control mixture in the moderatepenetrability class. (ii) e corrosion rate calculated according to the resistivity of the concrete was below the value of 2 µm/year for all the compositions made, thus determining the classification in the "very low" corrosion class, according to the classification in Table 3 of this article. (iii) e temperature recorded in the sodium chloride solution used for the RCPT rapid test did not affect the electrical charges passed through the concrete samples, which confirms the validation of the test. (iv) e high values of resistivity for mixtures with blast furnace slag content compared to the control ones, when chlorine ions penetrate, are due to the low concentration of sodium chloride (3%) according to the RCPT rapid test and especially the presence of blast furnace slag in the mixture. (v) Due to the lower apparent porosity of the slagcontaining concretes, the migration coefficients and the depth of penetration of chlorine ions in these compositions recorded lower values compared to the control mixture. (vi) e penetrability of chlorine ions through concrete evaluated by the colorimetric method together with the results of the tested physical and mechanical characteristics supported the results obtained by the rapid RCPT test. (vii) e depth of the carbonate layer was nonexistent, which reduced the risk of corrosion of the reinforcement in the mixtures, the result being attributed to increased compactness of the concrete matrix. (viii) e results of this study indicate that the proportions of blast furnace slag in combination with the conventional materials used were the most effective for the S 54/20 and S 54/60 road concrete compositions.
Data Availability e data sustaining the conclusions of this study are available within the tables and figures of the present paper.