The Effect of Coastal Environment on the Degradation of Reinforced Concrete StructuresView this Special Issue
Effect of Sulphate and Chloride Ingress on Selected Cements Mortar Prisms Immersed in Seawater and Leather Industry Effluent
Cement structures are major capital investments globally. However, exposure of cement-based materials to aggressive media such as chloride- and sulphate-laden environments such as coastal areas affects their performance. Ordinary Portland cement (OPC) is the main cement used in buildings and civil structures such as dams and bridges. This paper reports the findings of an experimental investigation on the effect of ingress of Cl− and SO42− on compressive strength development and the ions’ diffusivity in selected OPC brands in Kenya. The aggressive media used included seawater (SW) and wastewater from leather industry (WLI). Three brands of commonly used cements of OPC in Kenya were used. Mortar prisms were prepared from each brand of cement at different water-to-cement ratios (w/c) of 0.5, 0.6, 0.65, and 0.7 and allowed to cure for 28 days in a highly humid environment. The aggressive ions’ ingress in the mortar prisms was accelerated using a potential difference of 12 V ± 0.1 V. Analysis of diffusivity and diffusion coefficient of Cl− and SO42− was finally done. Compressive strength analysis was done before (at the 2nd, 7th, 14th, and 28th day) and after exposure to the aggressive ions. The results showed that the diffusivity of chlorides was more pronounced than that of sulphates. Diffusivity was observed to be higher at higher w/c ratios for all cement categories. It was observed that compressive strength increased with curing age, with the highest observed at 28 days. Cement A was generally found to have the highest compressive strength for all w/c ratios. The compressive strength was observed to increase after the mortar prisms were exposed to SW as opposed to the ones exposed to WLI. Generally, it was also observed that the strength gain increased with increase in w/c. The loss in strength was also observed to increase with increase in w/c.
Long-term durability of concrete structures is the main concern for safety and economic reasons. Deterioration of concrete structures is generally caused by ingress of aggressive agents into the concrete . A large number of concrete structures such as harbors, decks, piers, floating offshore platforms, power plants, and waste disposal facilities are generally subjected to environments which contain aggressive agents . These aggressive agents may include sulphates, chlorides, moisture, and carbon (IV) oxide . The steel rebar corrosion of the reinforced concrete (RC) caused by the ingress of chloride ions is the most severe problem affecting the durability of concrete constructions, especially in saline environments. Once a sufficient quantity of chloride has been accumulated on the surface of the embedded steel rebar, pitting corrosion will occur . Corrosion of RC is a problem throughout the world, demanding significant amounts for repair and rehabilitation. Extensive research has been done to establish relationships between the chloride ions content and the onset of corrosion [4–8]. Haque and Kayyali  showed that not all chloride ions that ingress into concrete remain free in the pore solution. Some of the ions get bound by the hydration products in a chemical reaction to form chloroaluminate hydrate. The portion of chloride ions that remains free is responsible for causing damage to the RC structures. Elsewhere, Sagüés  observed that the concentration level of Cl− to OH− (Cl−/OH−) in the pore solution determines the depassivation of steel rebar in the concrete, while Mundra et al.  suggested that, for external chlorides, the Cl−/OH− ratio below 3 does not cause significant corrosion but with internal chlorides, corrosion will occur at a lower level.
Sulphate attack is considered a major deteriorative problem occurring when the cement-based materials, such as concrete, mortars, and buildings, are exposed to sulphate ions environment . The sources of aggressive ions to the mortar/concrete may include deicing salts, seawater, groundwater, bacterial action in sewers, wastewater from industries, use of sulphate or chloride contaminated mix water, and natural gypsum in the aggregate, among others [1, 13–16].
In the leather-making industry, a great deal of wastewater containing high concentrations of sulphate is discharged which comes from many processes such as liming, deliming, bating, pickling, and chrome tanning . This wastewater with high concentrations of sulphates is not treated especially in tannery, but discharged directly into the integrated wastewater [17, 18]. These ions may be carried into inner sections of concrete by ionic diffusion, capillary absorption, permeation, and convective flow through the pore system, among others .
Ordinary Portland cement (OPC) is the main cement used in buildings and civil structures such as dams and bridges in Kenya . It is mostly preferred to blended cements due to the fact that it exhibits shorter setting times and achieves high early strength (28 days) than blended cements [1, 20]. However, it is more prone to attack by aggressive media, for example, chlorides and sulphates, among others. This is attributed to the fact that the hydration of OPC results in production of about 20 percent by weight of Ca(OH)2, which makes it highly susceptible to the aggressive ions . This potentially results in degradation of the cement-based structures. This subsequently reduces the service life of these structures. The use of OPC has continued regardless. In the Kenyan market, there are many brands of OPC, yet not much study has been carried out on their resistance to ingress of aggressive of Cl− and SO42− ions. Therefore, the present study aimed at investigating the effect of ingress of Cl− and SO42− on compressive strength development and the ions diffusivity in selected OPC brands in Kenya. The aggressive media used included seawater (SW) and wastewater from leather industry (WLI).
2. Materials and Methods
Materials were sampled from their respective places in Kenya. OPC (42.5 N/mm) was obtained from the respective appointed distributors in Nairobi, Thika Town, and Githurai Township in Kenya, respectively. Letters A, B, and C are used to refer to OPCs from companies A, B, and C, respectively. For each company cement category, 20 kg was procured from the appointed distributor in the respective towns. The 20 kg of each cement category was mixed thoroughly to make a homogeneous 60 kg sample in each cement category.
Seawater from the Indian Ocean and Pirates Beach, Mombasa County, Kenya, and wastewater from Leather Industries of Kenya Limited, Thika, Kiambu County were used. The seawater and wastewater from leather industry were labelled as SW and WLI, respectively.
River sand was obtained from sand transporters and distributors in Githurai Township, Kiambu County, Kenya. Sand as obtained from the sampling point was washed by spraying tap water and sun-dried for two days to a constant weight (ASTM C0117, 2004). The dried sand was sieved to meet the standard sand aggregates grade (ASTM C 593, 2005). To remove coarse materials, the dry sand was sieved through a 2360 μm sieve. The aggregates that passed through the 2360 μm sieve were then passed through a series of sieves which were 1180 μm, 600 μm, 300 μm, and 150 μm (ASTM C 593, 2005). Standard sand was finally prepared by mixing 384.00 g of sand retained on 1180 μm sieve, 427.00 g of sand retained on 600 μm sieve, 181.00 g of sand retained on 300 μm sieve, 223.00 g of sand retained on 150 μm, and 135.00 g of sand that passed through a 150 μm sieve to make 1350 g of standard sand (ASTM C 593, 2005). The sand retained in each sieve was stored in dry polythene bags until needed.
2.2.1. Preparation of Mortar Prisms
Mortar prisms measuring 40 mm × 40 mm × 160 mm were prepared in accordance with EAS 148-1 : 2000. In this regard, 450 ± 1 g of different brands of OP was separately weighed and placed in the mixing bowl of an automatic programmable mixer, model number 1305. For a water-to-cement ratio of 0.5, 225 ± 1 ml of distilled water was then added. The mixing bowl with its contents was clamped onto the mixer. 1350 ± 1 g of the graded sand was placed in an automatic pour trough and added automatically to the mixing bowl little by little while the machine was running at a speed of 30 revolutions per second. The machine was allowed to run for ten minutes. The mould and the hopper were firmly mounted on the vibrating table after the vibrator had been switched on. The first layer of the mortar into the compartments of the mould was placed from right to left of the mould within 15 ± 1 seconds so that the compartments were approximately half-full. After an interval of 15 ± 1 seconds, during which the vibrator remained running, the second layer was placed in the mould within the next 15 seconds, again working from right to left. The vibrator switched off automatically after 120 ± 1 seconds. The excess mortar was removed using a straightedge spreader. The surface of the mortar was smoothened using the same straightedge spreader held almost level. The vibration machine mould clamps were loosened to release the mould. The mould with the mortar was gently lifted off the vibrating table and stored in a cabinet maintained at 22 ± 1°C. The mortar prisms were left in the mould for 24 hours after which they were demoulded and marked with a crayon marker for identification. The mortars were finally cured in deionized water for additional 27 days in a curing room maintained at 22 ± 1°C. To prepare 0.6, 0.65, and 0.7 w/c ratio mortar, 270 ± 1 ml, 292.5 ± 1 ml, and 315 ± 1 ml water was used, respectively, following the same procedure. 27 mortar prisms were made for each category of cement.
Compressive strength measurements were taken using compressive strength machine model number ADR-Auto 250 at the age of 2, 7, 14, and 28 days and after separate curing in tap water and aggressive media. The 28-day water-cured mortars were also exposed to aggressive media for 36 hours, and the changes in the compressive strength after exposure were calculated using the following equation:where is the calculated percent gain or loss in compressive strength, is the strength after exposure, and is the strength at the 28th day.
For each category of cement, three prisms were removed from the curing tank; any deposits wiped out and covered with a damp cloth until tested. Their identities were noted down. A prism was placed in the test machine with one face on the supporting rollers and with its longitudinal axis normal to the supports. The load was then applied vertically by means of the loading roller at a rate of 50 N/s until failure to obtain prism halves. The prism halves were kept damp until tested. The halves were then crushed by smoothly applying load at a rate of 2400 ± 200 N/s. Triplicate measurements of compressive strength were taken.
2.2.2. Chemical Analysis of Test Cements
About 100 g of each cement sample was pulverized to pass through a 76 μm sieve and used for the analysis of cement oxides in accordance with KS EAS 18:2008.
(1) Analysis of Al2O3, SiO2, and Fe2O3. Analysis of Al2O3, SiO2, and Fe2O3 was carried out in accordance with KS EAS 18:2008. A buffer solution was prepared by dissolving 2.5 g of SnCl2·H2O in 10 ml concentrated HCl in a 100 ml volumetric flask. The solution was then topped to the mark with distilled water. 1000 ppm stock solution of aluminium was prepared by dissolving 1.0 g of aluminium in 20.0 ml HCl and 1 ml of HgNO3 in a 1000 ml volumetric flask. The resultant solution was then topped to the mark using distilled water. For the preparation of working standards, 10.0 ml of the stock solution was put in a 100 ml volumetric flask. 2.0 ml of the buffer solution and 2.0 ml of a potassium chloride solution containing 100 g/l were then added. The resultant solution was then topped to the mark using distilled water. From the working standards, solutions of 20 ppm, 40 ppm, 60 ppm, and 100 ppm were made through serial dilutions. SiO2 and Fe2O3 standards were made through serial dilution of titrosol solutions of the oxides provided with the AAS machine. 0.100 g of the pulverized cement sample was weighed and placed at the bottom of a 100 ml plastic bottle. 1.0 ml of aquarengia (HNO3 : HCl, 1 : 3) and 3.0 ml of hydrofluoric acid were added and the bottle was corked. The mixture was then stirred taking care so as to maintain the sample at the bottom of the bottle. The resultant mixture was allowed to stand for 12 hours. 50.0 ml of boric acid was added, and the solution was left to stand for one hour. The resultant solution was topped to the mark using distilled water and left to stand for two hours. AAS model AA.10 was used to determine the amounts of Al2O3, SiO2, and Fe2O3. Calibration of the machine was done using the working standard solution diluents and the blank solution. Analyses were done in triplicate. Each time analysis was done, the standard solutions were prepared afresh.
(2) Analysis of CaO and MgO. Analysis was done according to KS EAS 18:2008. A buffer solution was prepared by dissolving requisite amounts of La2O3 and CsCl in 20.0 ml distilled water−25.0 ml concentrated HCl mix in a 100 ml volumetric flask. The resultant solution was then topped to the mark using distilled water. 1000 ppm of CaO and 200 ppm of MgO stock solution was prepared by dissolving 2.5 g anhydrous CaCO3 and 0.7 g of anhydrous MgCO3 in 500.0 ml of 1 : 4 nitric acid : distilled water mixture in a 1000 ml volumetric flask. 20.0 ml of the buffer solution was added to the solution, and the resultant solution was topped to the mark using distilled water. From the stock solution, 50, 100, 300, and 500 ppm CaO and 10, 20, 60, and 100 ppm MgO were prepared through serial dilution. Blank solution was prepared using the same procedure but without the addition of CaCO3 or MgCO3. 0.5 g pulverized cement sample and 0.5 g of ammonium acetate were mixed in a 250 ml beaker. 10.0 ml of concentrated HCl was added into the beaker while stirring until no more bubbles were observed. The beaker’s content with a glass cover lid was heated over a water bath at 80–100°C for about 30 minutes. 50.0 ml hot water was added while stirring. The solution was then filtered through a Whatman filter paper number 542 into a 1000 ml volumetric flask. The residue was washed with 5.0 ml hot concentrated HCl twice followed by hot water until it was confirmed free of chloride ions using AgNO3 solution. 10 ml buffer solution was added to the solution while stirring. The solution was cooled and then topped to the mark using distilled water. Analysis was then done in triplicate using AAS model AA.10. Calibration curve was prepared from the stock solution diluents and the blank solution. Each time analysis was done, standard solutions were prepared afresh.
(3) Analysis of K2O and Na2O. A buffer solution was prepared by dissolving 30.0 g aluminium in 400 ml of 50% HNO3 acid in a 1000 ml volumetric flask, and the solution was topped to the mark using distilled water and mixed well by inversion. 500 ppm K2O and 250 ppm Na2O stock solution was prepared by dissolving 0.7915 g KCl and 0.4715 g NaCl in 500 ml distilled water in a 1000 ml volumetric flask. To the resultant solution, 20.0 ml buffer solution was added and the solution was topped to the mark using distilled water. From the stock solution, 10, 20, 30, 40, and 50 ppm K2O and 5, 10, 15, 20, and 25 ppm Na2O were prepared through serial dilution. In each case, an amount of concentrated HCl equivalent to half the amount of the stock solution was added. Blank solutions were made using the same procedure but without the addition of KCl and NaCl. 0.5 g of the pulverized cement sample was put in a 50 ml beaker and 25.0 ml distilled water and 5.0 ml concentrated HCl were added as stirring continued. The resultant solution was topped to the mark using distilled water and heated to boiling on a hot plate for fifteen minutes. The solution was filtered through a Whatman filter paper number 542 into a 500 ml volumetric flask. The filter paper was washed six times with hot water. 10.0 ml of the buffer solution was added to the solution while stirring. The solution was cooled and then topped to the mark using distilled water. Calibration curve was prepared using the stock solution diluents and the blank solution. Analysis in triplicate was finally done by flame photometry using flame photometer number WFX 210A.
2.2.3. Bogue’s Calculation of the Major Cement Phases
The approximate mineralogical compositions of the cements were presented in percentages of the main phases.
2.2.4. Migration Test
Migration test was carried out in accordance with ASTM C 1556 (2004). The mortar prisms cured for 28 days were cut to 100 mm length using a cutting machine model number MC 100 type EFNOUT KT with a 2.0 mm diamond blade. An electrochemical cell setup consisting of 500 ml of 0.3 M NaOH was placed in the anodic compartment and an equal volume of SW, WLI, or TW was placed in the cathodic compartment. Stainless steel electrodes were then placed in the two compartments as the electrodes. The electrodes were connected to a 12 V ± 0.1 V electric source. Current between the two electrodes was recorded using a milliampere ammeter after every 30 minutes. Temperature was also monitored throughout the exposure period. The exposure period of 36 hours started when the solutions were placed in the respective compartments, covered, and the electrodes were connected in the circuit. The solutions were stirred periodically using a glass rod.
After the exposure period, the mortar prisms were removed from the setup and allowed to drain for about 20 minutes. Three mortar prisms were subjected to compressive strength analysis while the other three were subjected to chloride and sulphate profiling. The mortar prisms were sliced into 10 mm slices using a well-oiled block cutter model MC-100 number 88-07-521 with a 2.0 mm diamond blade. The slices were then dried in an oven Binder model at 50–70°C for 60 seconds ± 5 seconds. Each slice was pulverized using a standard pulverizing machine Siebtechnik model TS 250 so as to pass through a 76 μm mesh sieve. Each pulverized sample was stored in a dry well-labelled plastic container and shaken well to enhance mixing. The pulverizer machine basins were thoroughly cleaned and dried before another sample was ground. The pulverized samples were then analyzed for chlorides and sulphates. Triplicate analyses were done for each cement category.
2.2.5. Sulphate Analysis
Sulphate analysis was done using the turbidimetric method in accordance with ASTM C 1580 (2007). Buffer solution was prepared by mixing 30.0 g magnesium chloride, MgCl2·6H2O, 5.0 g sodium acetate, CH3COONa·3H2O, 10.0 g potassium nitrate, KNO3, and 20.0 ml acetic acid, and CH3COOH in 500 ml distilled water and made up to 1000 ml using distilled water. A 100 ppm sulphate stock solution was prepared by dissolving 0.1479 g anhydrous sodium sulphate, Na2SO4, in 500.0 ml distilled water in a 1000 ml volumetric flask, and the resultant solution was made to the mark using distilled water. From the stock solution, 1 ppm, 5 ppm, 10 ppm, 20 ppm, and 30 ppm were prepared through serial dilutions. The diluents were used to prepare the calibration curve.
Ten grams of the pulverized sample was placed in a 250 ml beaker, and 75 ml of distilled water was added slowly while stirring. The solution was placed on a hot plate and allowed to boil for two minutes then diluted to 100 ml using distilled water and heated for 15 minutes. The resultant solution was then filtered using Whatman filter paper number 542 into a beaker. The residue was washed with hot water until no white precipitate was formed when AgNO3 solution was added to the filtrate. The filtrate was diluted to 250 ml and boiled. 10 ml of the diluted sample solution was placed in a 250 ml conical flask. 20 ml buffer solution was added, and the resultant solution was stirred with a magnetic stirrer, while stirring about 0.2 g of BaCl2 was added and the solution was stirred at a constant speed for 60 ± 2 seconds. The resultant solution was left to sit for five minutes during which the absorption cell was filled with distilled water and used to zero the instrument at 420 nm wavelength. Analysis of the samples was run on a Beckman DU 520 spectrophotometer. The sample solution was then transferred into the cell of the photometer and turbidity measured at 5 ± 0.5 minutes. The same procedure was applied to the standards and their turbidity determined. The turbidity of the test solution was calculated by taking the difference between solution with BaCl2 and the one without. Calibration curves were used to determine sulphate concentration in ppm of cement. Triplicate analyses were done for each category of the test cement. The results were presented as sulphate concentration versus depth of ingress. From the graph, sulphate diffusion coefficients, Dapp, were derived from error fitting curves in equation (6) using the method of least squares . where C is the concentration of the aggressive ion at a depth x and at the moment t and D is the diffusion coefficient.
The diffusion of chloride or sulphate ions in concrete is approximated from solution to Fick’s second law equation  under non-steady-state conditions assuming boundary conditions C(x,t) = 0 at t = 0, 0 < x<∞, C(x,t) = Cs at x = 0, 0 < t < ∞, constant effects of coexisting ions, linear chloride binding, and one-dimensional diffusion into semi-infinite solid. The analytical solution to equation (7) is equation (8) [23, 24]:where C(x,t) is the concentration of the ion at any depth x in the mortar bulk at time t, Cs is the surface concentration while Dmig is the migration diffusion coefficient and erf is the error function. Apparent diffusion coefficients, Dapp, are calculated from the following equation: where is the effective applied voltage in V, t is the duration of exposure in seconds, Dapp is the apparent diffusion coefficient, and Dmig is the migration diffusion coefficient.
2.2.6. Chloride Analysis
Cl− analysis was carried out using the Mohr method, precipitation titration. A 5% K2CrO4 indicator was prepared by dissolving 1.0 g of K2CrO4 in 20.0 ml of distilled water. Standard (0.1 M) AgNO3 solution was prepared by dissolving 9.0 g of AgNO3 in about 200.0 ml distilled water in a 500 ml volumetric flask and the solution topped to the mark using distilled water. The resultant solution was standardized against NaCl. NaCl was dried for one hour at 140°C in an oven and cooled to room temperature in desiccators. 0.25 g portions of NaCl were weighed into a 250 ml Erlenmeyer flask and dissolved in about 100.0 ml of distilled water. Small quantities of NaHCO3 were added until effervescence ceased. About 2 ml of K2CrO4 was added and the solution was titrated to the first permanent appearance of red colour.
Individual 10.0 g pulverized sample was placed in a 250 ml Erlenmeyer flask. 150 ml of 2 : 1 nitric acid : distilled water was added to the sample in the flask. A glass cover lid was placed, and the flask and its content were placed on a hot plate until the volume of the solution reduced to about 100 ml. The solution was allowed to cool for about 15 minutes. The solution was then neutralized by adding small quantities of calcium carbonate until effervescence ceased. The solution was then filtered through a Whatman filter paper number 541 into a 200 ml volumetric flask. The filter paper was rinsed thrice with distilled water, and the resultant solution was topped to the mark. 10.0 ml aliquot of the solution was pipetted into a conical flask. About 2.0 ml the indicator was added, and titration was carried out in the usual way with the 0.1 M AgNO3. Triplicate analyses were done for each cement category. A chloride-free CaCO3 blank was run through the same procedure. The amount of chloride was determined using the following equation: where V1 is the volume of AgNO3 solution used for sample titration (equivalent point), V2 is the volume of AgNO3 solution used for blank titration (equivalent point), M is the molarity of AgNO3 solution, and W is the weight of sample in grams.
The results were represented in chloride profiles. From the profiles, chloride diffusion coefficients, Dmig, were derived from error fitting curves in equation (7) using the method of least squares. Apparent diffusion coefficients, Dapp, were calculated using equation (8).
2.2.7. Chemical Analysis of SW and WLI
About 500 ml of each media was sampled for the analysis of Na+, K+, Ca2+, Mg2+, Cl−, and SO42−. The samples were filtered to remove solid matter. Sodium and potassium ion concentrations were analyzed using flame photometry, while Ca2+ and Mg2+ were analyzed using atomic absorption technique. Buffer solutions for the analysis of Ca2+ and Mg2+ were prepared as described in Analysis of CaO and MgO in Section 2.2.2 and for Na+ and K+ as described in Analysis of K2O and Na2O in Section 2.2.2. Sulphate concentration in sampled SW and WLI was determined using the procedure described in Section 2.2.5 while chloride concentration was determined using the procedure described in Section 2.2.6.
3. Results and Discussion
3.1. Chemical Analysis of the Test Cements
Table 1 gives the results for chemical analysis of oxides in percent by mass (except for Cl−) of the test cements.
Table 2 gives the results of the Bogue calculation.
From Table 1, it was observed that for all the test cements, the sum of the proportions of CaO and SiO2 was at least 50 percent by mass. This allows the formation of the most important cement phases during the formation of clinkers. The phases are C3S and C2S. Kenya Standards stipulates a minimum of 50 percent by mass of CaO and SiO2 for OPC (KS EAS 18-1:2001). The ratio (CaO/SiO2) was 2.99, 3.00, and 2.99 for cements A, B, and C, respectively. The Kenya Standards stipulates a minimum of 2.0 for OPC. From Bogue’s approximations, all the test cements met the two-thirds by mass of calcium silicates (C3S and C2S) stipulated by the Kenya Standards. The silicates upon mixing with water hydration make cement gain strength.
The Kenya Standards (KS EAS 18-1:2001) stipulates a maximum of 0.10 percent of chloride content and a maximum of 3.5 percent sulphate of SO3. From the results, all the test cements used in this study met this requirement. In the chemical analysis of cement, gypsum is indicated by SO3. The role of gypsum is to control the setting time of cement when mixed with water. Slower setting results in greater strength to the set mass.
The alkali and alkaline metal oxides are important because they provide alkalinity in hydrated cement and/or concrete pore systems. The alkalinity protects reinforcing bars against corrosion. A disruption of the pore system would result in corrosion of the rebar. The test cement alkali metal oxides were within the recommended range of 0.100–2.000%, and the alkaline metal oxides were within the stipulated range of 40–67% (KS EAS 18-1 : 2008). The oxides besides providing alkalinity (pH > 10) for the pore water system also provide a medium through which cement phases react .
The aluminium oxide and iron oxides of the test cements were within the recommended range of 1.5–8% for OPC . These oxides serve as a flux during clinkerization, thus promoting fusion. This aids in the formation of the calcium silicates at a lower temperature (<2000°C). This makes the process of cement manufacture economical . Given that the cement samples were obtained from the market, the results indicate that the Kenya Standards and East African Standards are observed by the various cement manufacturers in Kenya.
3.2. Compressive Strengths
Compressive strength is a performance measure used by engineers in designing buildings and other structures. The compressive strength results are used in quality control, acceptance of concrete/mortar, and evaluation of adequacy of curing. Compressive strength results for the test cements at various w/c ratios and at 2, 7, 14, and 28 days are represented in Figures 1–4.
From Figures 1–4, it was observed that the compressive strengths increased with curing period at all w/c ratios and for all the test cements. The increase in the compressive strength with curing days was significant. The increase in the compressive strength was significant for all w/c ratios. Hydration is mainly achieved through curing of cement-based materials. The older the concrete/mortar, the greater the hydration that has occurred and the higher the compressive strength [20, 28–31]. Cement gains strength upon hydration. The hydration reactions of C3S and C2S are given by
During hydration, CSH is formed which contributes to the strength of the mortars. The Kenya Standard (KEBS KS02-1262, 1993) recommends a minimum strength of 42.5 MPa for OPC at 28 days for w/c = 0.5 and 10 MPa at 2 days. All the test cements met the standard requirement.
There was a decrease in the compressive strength with increase in w/c ratios for all the test cements. This can be attributed to increased porosity at high w/c ratio. The key constituents of all concrete mixes are the binder system and the amount of water present [32, 33]; w/c ratio is considered as the most important factor affecting mortars/concrete strength . This is because it affects the porosity of the hardened paste. The quantity of water used also affects the flow or rheology of the mixture as well as cohesion between paste and aggregate . As a result, it influences the overall strength of the mortar.
It was observed that cement A had generally the highest compressive strength compared to the other test cements. From Bogue’s calculation (Table 2), A, B, and C had 79.652%, 75.812%, and 74.299%, respectively, of combined C3S and C2S. These are the two main phases that contribute to the strength development of cements . Hydration of silicates contributes significantly to the strength development of cements. Cement A had the highest amount of C3S and C2S (79.652%) and hence the observed compressive strength. Hydration of silicates leads to substantial strength gain.
3.3. Chemical Analysis of SW and WLI
Table 3 gives the results for analysis of SW and WLI in ppm.
SW was found to have a higher chloride concentration than WLI. It was therefore expected that mortar prisms subjected to SW would have more gain in compressive strength than the ones subjected to WLI. This was observed in this study. Chlorides are known to be accelerators of compressive strength development in mortars or concrete . On the other hand, WLI was found to have a higher sulphate concentration than SW. Mortar prisms subjected to it were expected to suffer more from sulphate attack [36–38]; this was observed in this study.
3.4. Compressive Strength Development on Exposure to Aggressive Media and Tap Water
The change in the compressive strength of the cement mortars subjected to SW, WLI, and TW was compared to their respective strength at 28 days, and their percentage gain/loss was determined. Figures 5–7 show these results after exposure to SW, WLI, and TW, respectively.
There was an observed increase in the strength for all the test cements immersed in SW for all w/c ratios. It was also observed that the strength increased with w/c ratio. Chlorides are known to be accelerators of compressive strength in mortars/concrete. Seawater contains both chloride and sulphate ions. The presence of chloride ions in the hydrated cement pore water activates residual cement hydration, hence enhancing the strength. Chlorides are more active in enhancing the strength gain than sulphates due to their small charge and ionic sizes. The small size allowed more chlorides into the mortar. This in turn may have allowed more residual hydration products to be formed majorly, Friedel’s salt that might have crystallized in the pores. This results in pore refinement and densification of the mortar matrix hence increased compressive strength. Both chlorides and sulphates are known to initiate residual cement hydration .
It was also observed that gain in compressive strength increased with increased w/c ratio. High w/c ratio results in increased voids and spaces through which the residue cement hydration reactions occur. Increased voids and spaces due to high w/c ratios give more room for more residual hydration. Generally, cement A had the highest percent gain across all the w/c ratios. This could be attributed to the high content of C2S and C3S phase which is observed in Table 2. The phase may have been activated by ingressed Cl−.
The percent loss in compressive strength after exposure of the mortar to WLI is given in Figure 6 at different w/c ratios.
It was observed that there was a decrease in compressive strength when the mortars were exposed to WLI for all w/c ratios. The loss of the compressive strength may be attributed to the formation of many different compounds such as Na2SO4 (thenardite), Na2SO4·10H2O (mirabilite), Na3H(CO3)2·2H2O (trona), Na2CO3·H2O (thermonatrite), and NaHCO3 (nahcolite), and CaSO4·2H2O (gypsum), CaSiO3·CaCO3·CaSO4·15H2O (thaumasite), and 8.5CaSiO3·9.5CaCO3·CaSO4·15H2O (birunite). Sodium sulphates can result in damage to the concrete due to cyclic formation of the anhydrous and hydrated forms of these salts . Expansion of concrete may be attributed to the formation of ettringite . Ettringite forms as a result of the chemical reactions between sulphate ions with aluminate phase of cement and Ca(OH)2 .
Mortars subjected to a 3.5% sodium chloride and 5% sodium sulphate solution were found to have a percent reduction of about 20.7% and those subjected purely to a 5% sodium sulphate solution 68.3%. This trend elucidated the predominant role of the Cl− in mitigating sodium sulphate attack. Presence of chloride ions leads to a reduction in the deterioration effects of sulphate ions . The percent expansion of the specimens subjected to combined action of sulphate and chloride ions is less than would be if the specimens were subjected to individual action of sulphate ions. The rate of diffusion of chloride ions is much higher than that of sulphate ions and this allows the chlorides to react with C3A and C4AF . The reaction leaves little C3A available for sulphate ions to react to form ettringite. In this study, cement A had the highest amount of C3A (4.781%). This may mean that little amount of C3A was available to react with sulphate ions to form ettringite; hence, cement A experienced least expansion for all w/c ratios. The aggressive media used had both chloride and sulphate ions, and the loss in the compressive strength was significantly different for all w/c ratios. Magnesium attack may also have occurred resulting in decalcification. In the case of magnesium sulphate attack, brucite, Mg(OH)2, which has low solubility, and its relatively great amount of gypsum released lead to degradation . This is so because magnesium also takes part in the reactions, replacing calcium in the solid phases. The displaced calcium precipitates as gypsum. OPC with low C3A content is more easily attacked by sulphates . In this study, cement C had the lowest C3A content and was therefore more attacked. Figure 7 gives the percent gain in compressive strength after exposure of mortars to TW.
It was observed that there was a slight increase in compressive strength subjecting the mortar to TW. TW contains ions such as chlorides which were expected to accelerate compressive strength development. However, the concentration of these ions was too low to cause a substantial increase in strength development. The slight increase in strength was therefore more due to continued hydration process than the presence of the stated ions. Most concrete structures are expected to come into contact with TW as a normal media. From the results, it was observed that WLI is an aggressive media since all the mortars experienced a reduction in compressive strength when compared to those exposed to TW. On the contrary, mortars exposed to SW had gained compressive strength even more than those exposed to TW. This was expected due to the high concentration of chlorides in SW than in TW. TW has a chloride concentration of less than 250 ppm and sulphate concentration of less than 400 ppm (KS 05-459, 1996). However, this did not mean that SW was not an aggressive media. Chlorides, as stated earlier, are used as compressive strength development accelerators in concrete structures [44, 45]. Ingressed chlorides and sulphates cause a gain in compressive strength. Increased strength of the hydrated cement mortars results in a densification of the microstructure of the hydrated cement after exposure to aggressive ions.
3.5. Chloride and Sulphate Profiling
3.5.1. Chloride Profiles
Figures 8–15 show the chloride profiles against the depth of cover of each category of mortar prisms after exposure to SW and WLI, respectively, at various w/c ratios. The profiles involved the determination of the concentration of the chloride ions (×10−1) at different depths of cement mortar bulk of the test cements.
From Figures 8–15, it is observed that, as w/c increased, there was an increase in chloride ingress across all the test cements. This can be attributed to continuous and interlinked voids that provided pathways for ion ingress in hydrated cement mortars at increased w/c ratios. The penetrability of concrete is related to the pore structure of the cement paste matrix. As w/c ratio increases, the porosity of the resultant mortar increases . Increased porosity could have resulted in a higher diffusivity of chlorides and sulphates into the mortar. From the figures again, it was also observed that the ion ingress was highest in the first few millimeters (20 mm) from where their concentration dropped significantly. This was attributed to the proximity to the exposed surface to the aggressive media. Diffusivity of chloride ions is affected by the amount of C3A and C4AF in the cement . A higher amount of C3A in cement results in a lower diffusivity of chloride ions. C3A and C4AF are known to bind chlorides, thus decreasing their ingress. The reaction between C3A and free chlorides in the hydrated cement results in its reduction from the pore solution. In this study, Figures 14 and 15 show clearly that cement C had a higher total chloride ingress than B and A. The order of the amount of C3A was C < B < A. It can be seen from Table 2 that C had the least amount of C3A. This showed that it had the lowest proportion of the phase that binds chlorides. C was therefore observed to have the highest chloride diffusivity compared to B and A. This can be attributed to chloride binding capacity of the cement involved. Cements with a high C3A content are likely to bind chlorides more than those with low C3A content [47, 48]. When more chlorides are bound, their diffusivity and thus RC corrosion risk are lowered.
Generally, higher w/c ratios were expected to show increased chloride ingress into the bulk compared lower w/c ratio. This was observed in this study. As the w/c ratio increased, there was a marked rise in the total chloride ingress in all the profile depths of all the test cements. Increase in w/c ratio leads to an increase in porosity. This was dependent upon the depth of penetration which was again dependent upon the brand of OPC.
3.5.2. Sulphate Profiles
It was observed that diffusivity of sulphates increased with increase in w/c ratio. This was attributed to porosity and permeability of the mortar/concrete [47, 48]. Expansive products are expected to be formed when mortars are subjected to sulphates and fill the voids at high w/c ratio . Possible cracks may have provided a pathway for the ingress of sulphate ions. The mortars may also have suffered sulphate attack which may have resulted in the formation of gypsum, brucite, and ettringite. These products are expansive and thus create voids in the mortar making it more permeable. From the profiles, the concentration was highest in the first 20 mm and then a drop was experienced. This may be attributed to their ionic size. Sulphate ions being bulky ions were not expected to ingress to deep depths. In the presence of chloride ions, sulphate ions are first bound to the hydration products of cements . However, the chloride and sulphate diffusivity can be used as a general quality parameter for evaluation of resistance of concrete against chloride and sulphate intrusion.
3.5.3. Apparent Diffusion Coefficient
Figure 24 shows one of the error fitting curves for apparent chloride diffusion coefficient (Dapp) of cement A mortar. Similar curves were used for the determination of the apparent diffusion coefficients for other test cements. In this study, the equation to solution of Fick’s law under non-steady-state conditions for diffusion in a semi-infinite solid was used. Tables 4 and 5 give the Dapp and r2 values for different cement mortars and w/c values after exposure to SW and WLI, respectively, for Cl−, and Tables 6 and 7 give the Dapp and r2 values for different cement mortars and w/c values after exposure to SW and WLI, respectively, for SO42−.
The Dapp generally increased with increasing w/c ratio for both chlorides and sulphates. As would have been expected generally, the diffusivity and permeability are higher for the 0.7 w/c systems than 0.5, 0.6, and 0.65. Lowest diffusivities of chloride and sulphate ions for 0.5 w/c were observed for all the test cements. This shows that low w/c ratio gave better resistance to ingress of ions. Increasing w/c ratio increased diffusivity of chlorides and sulphates. High w/c ratio affects porosity. Although increased porosity gives more room for increased cement hydration, extremely high w/c ratios are harmful as the increased porosity acts as pathways for harmful ions and products .
It was generally observed that increasing w/c ratio reduced the resistance of the mortars to both Cl− and SO42− diffusion. Diffusivity and permeability of mortars and pastes decrease with increased degree of hydration as a result of curing. Mortars made from cement A were found to have lower Cl− and SO42− diffusivity compared to both B and C for all w/c ratios. This may be due to higher degree of hydration in mortars made from A compared to B and C due to its highest C3S content which is responsible for the high degree of hydration.
From Table 2, A had the highest C3A, and this may mean that C3A may have reacted with chloride ions so that free chloride ions decreased. This then resulted in reduced chloride diffusivity. It was also observed that diffusivity of chlorides in mortars exposed to WLI was slightly higher than those exposed to SW. This was attributed to the higher sulphate ion concentration in WLI compared to SW. From Table 3, it was observed that sulphate concentration was higher in WLI than in SW. The amounts of free chloride ions in the pore water remain high when sulphate ions are added because chloride binding was reduced in the presence of sulphates. This is so because sulphates are bound first occupying sites otherwise available for chlorides [48, 50].
From Tables 4 and 5, Dapp values for chlorides were found to be in the range of 1.738 × 10−10 m2/s to 4.589 × 10−10 m2/s. With low ion ingress, low Dapp was expected. These values were in agreement with Dapp values for chlorides which have been found to be in the range of 10−9 to 10−10 m2/s depending on the concrete. In trying to fit the error function in all the test cement mortars, chloride profile results to obtain the Dapp, it was observed that the fitting was generally good. This was arrived at from the r2 values that were above 0.98.
From Tables 4 and 5, r2 values did not show an increase as w/c ratio increased or a decrease as w/c ratio decreased. This could probably show that there were other processes other than pure diffusion involved in the ingress. This change in Dapp could be attributed to other factors such as sulphate precipitation with the components of the pore solution, adsorption rather than w/c effect only.
In trying to fit the error function in all the test cement mortars, sulphate profile results to obtain the Dapp, it was observed that the fitting was generally not good. This was arrived at from the r2 values that were below 0.96. Dapp values have been reported by many researchers to be in the range of 10−10 to 10−13 m2/s depending on the concrete . From Tables 6 and 7, r2 values did not show an increase as w/c ratio increased or a decrease as w/c ratio decreased. This could probably show that there were other processes other than pure diffusion involved in the ingress .
Generally as expected, it was observed that the higher the w/c, the higher the Dapp. From Tables 4–7, there was more ion ingress in the mortars with higher w/c ratio. This in turn resulted in higher Dapp which can be attributed to increased permeability due to high porosity.
Based on the results, analysis, and discussions, the following conclusions were made:(i)Diffusivity of both Cl− and SO42− ions increased with w/c ratio for all the test cements(ii)Diffusivity of Cl− ions was higher than that of SO42− ions(iii)The use of high w/c > 0.5 should be discouraged since it resulted in more ingress of aggressive ions(iv)Mortars subjected to SW exhibited a gain in compressive strength as opposed to those subjected to WLI(v)All the test cements, A, B, and C, met the stipulated minimum requirements by KS EAS 18-1:2008
The data used in this paper will be provided upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
The authors sincerely thank the Department of Materials Testing under the Ministry of Roads in Kenya for providing the laboratory facilities.
K. K. Aligizaki, “Mechanisms of ion transport in cement-based materials during the application of corrosion mitigating electrical techniques,” in Proceedings of the NACE-2012-1542, pp. 1–15, NACE International, Salt Lake City, UT, USA, March 2012.View at: Google Scholar
J. Crank, The Mathematics of Diffusion, Clarendon Press, Oxford, UK, 2nd edition, 1975.
U. T. Bezerra, A. E. Martinelli, D. M. A. Melo, M. A. F. Melo, and F. M. Lima, “A correlation between Bogue’s equations and Taylor’s procedure for the evaluation of crystalline phases in special class Portland oilwell cement clinker,” Cerâmica, vol. 57, no. 341, pp. 122–128, 2011.View at: Publisher Site | Google Scholar
M. J. Mwiti, T. J. Karanja, and W. J. Muthengia, “Thermal resistivity of chemically activated calcined clays-based cements,” in Calcined Clays for Sustainable Concrete, F. Martirena, A. Favier, and K. Scrivener, Eds., pp. 327–333, Springer, Dordrecht, Netherlands, 2018.View at: Google Scholar
A. M. Neville, Properties of Concrete, Longman, London, UK, 4th edition, 1997.
E. Dan and I. Janotka, “Chemical resistance of Portland cement, blast-furnace slag Portland cement and sulphoaluminate-belite cement in acid, chloride and sulphate solution: some preliminary results,” Ceramics-Silikaty, vol. 47, pp. 141–148, 2003.View at: Google Scholar
A. E. Abalaka and A. D. Babalaga, “Effects of sodium chloride solutions on compressive strength development of concrete containing rice husk ash,” ATBU Journal of Environmental Technology, vol. 4, no. 1, pp. 33–40, 2011.View at: Google Scholar
N. R. Wangui, “Sulphate and chloride ingress and the effect in selected cements mortar prisms immersed in sea water and leather industry effluent,” Kenyatta University, Nairobi, Kenya, 2014, MSc thesis.View at: Google Scholar