1-s2.0-S0950061812008860-main

Construction and Building Materials 40 (2013) 710–718

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Influence of cement type in reinforcement corrosion of mortars under actionof chlorides

Victor Correia de Oliveira Pereira a,⇑, Eliana Cristina Barreto Monteiro a,b, Kalline da Silva Almeida a

a Department of Civil Engineering, University of Pernambuco, Pernambuco, Brazilb Department of Civil Engineering, Catholic University of Pernambuco, Pernambuco, Brazil

h i g h l i g h t s

" Increasing the curing period has improved the cement performance." Water–cement ratio presented itself as the most important factor to increase the durability, followed by the cement type." The CPIII-40 cement had the best performance with respect to corrosion by chloride ions to both water–cement ratios studied." To increase the durability in structures the cement to be used must be specified according to environmental conditions.

a r t i c l e i n f o

Article history:Received 18 April 2012Received in revised form 17 October 2012Accepted 22 November 2012Available online 25 December 2012

Keywords:CorrosionChloride ionsDurabilityCement

0950-0618/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.conbuildmat.2012.11.050

⇑ Corresponding author. Address: R. Des. CapistranoMartin, Recife, PE, CEP 50761-090. Tel.: +55 81 32295933.

E-mail address: [email protected] (V.C.O.

a b s t r a c t

In addition to the technological and environmental factors, the corrosion caused by chloride ions isstrongly influenced by the type of cement used in concrete; however, currently, cement is manufacturedand specified without taking into account its resistance to the action of aggressive agents. Given this con-text, a study on the protective capacity of some types of cement (CPII-Z-32, CPIII-40 and CPIV-32) wasconducted regarding the reinforcement structure under the action of chloride ions. The specimen moldedwith CPIII-40 cement clearly showed high resistance to corrosion caused by chloride ions, high compres-sive strength, and low capillary absorption.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Concrete represents the most suited building material to struc-tures, surpassing alternatives, also viable, such as steel and wood[1]. For a long time, it was believed that the durability of this mate-rial was limitless; however, during the decades of 1980 and 1990,the initial perception on durability was changed with the advent ofpathological manifestations that caused significant and frequentdamage to structures [2].

Reinforcement corrosion, one of the main causes of deteriora-tion of reinforced concrete structures, can be defined as an electro-chemical process that causes the degradation (oxidation) ofconcrete steel [3]. In advanced stages, it can compromise safetyof the structure and may lead to the collapse of the affected con-crete structures [4].

ll rights reserved.

de Moraes e Silva, n. 376, San9877, mobile: +55 81 9126

Pereira).

The literature on durability of reinforced concrete structureshas considered corrosion initiated by chloride ions as the most se-vere attack and the leading cause of premature corrosion in rein-forced concrete structures [5–11].

Various technological aspects of concrete (water/cement ratio;particle size distribution; chemical composition) contribute toreducing voids and increasing the compactness of concrete, thusreducing the transport of aggressive agents to the structure interior.

It is widely recognized that the corrosion of steel in concrete in-duced by chloride ions can lead to a rapid deterioration of reinforcedconcrete structures. This type of corrosion is influenced by severalfactors, such as pH, concentration of tricalcium aluminate (C3A) incement, water/cement ratio, cement content, and concrete cover.The presence of a critical concentration of chloride ions in contactwith the reinforcement will cause its depassivation, paving theway to the corrosion process, which, after corrosion initiation, willcontribute to the loss of the structure mechanical performance.

The current Brazilian standard NBR ABNT 6118:2007 (Project ofconcrete structures) has followed the international tendency, sinceit has specified levels of environmental aggressiveness; suchdetailing seeks, however, to ensure durability only through defini-

http://dx.doi.org/10.1016/j.conbuildmat.2012.11.050

mailto:[email protected]

http://dx.doi.org/10.1016/j.conbuildmat.2012.11.050

http://www.sciencedirect.com/science/journal/09500618

http://www.elsevier.com/locate/conbuildmat

Table 2Dependent variables and the corresponding levels to each factor.

Dependent variables

Corrosion evaluation Corrosion potential (ecorr)Mortar properties evaluation Capillary absorption

Compressive strength

V.C.O. Pereira et al. / Construction and Building Materials 40 (2013) 710–718 711

tion of maximum water/cement ratio (w/c), minimum coatings,and minimum compressive strength, not taking into considerationthe type of cement used and the minimum service life the struc-ture should reach [12].

The partial replacement of cement for products such as blastfurnace slag and pozzolanic materials leads to the beneficial effectsof reinforced concrete structures [13,14], especially when it comesto protection against chloride-induced corrosion of steel reinforce-ment [15,16]. The reduction in the diffusivity/permeability is a ma-jor purpose of using these products, particularly for chloride iontransportation [13,17,18]; besides increasing the resistivity of theconcrete [19]. In previous work, it was noted that the reinforce-ment corrosion initiated by chloride ions depends fundamentallyon the chemical composition of the cement used for the manufac-ture of mortars and concretes [11,20–23].

Due to this it is necessary to study the influence of type of ce-ment used for making mortar and concrete taking into consider-ation the particularities of the cements produced.

In view of the aforementioned, the present work aims to com-pare the protective capacity of some types of cement in Northeast-ern Brazil, CPII-Z-32 (Portland Composite Cement with Pozzolans),CPIII-40 (Portland Blast furnace Cement), and CPIV-32 (PortlandPozzolanic Cement), as the reinforcement corrosion under the ac-tion of chloride ions, which primarily depends on the chemicalcomposition of this material. The results presented here are partof a dissertation by Pereira [24] and will be useful for the concretetechnologists to specify correctly the type of cement to be used onstructures placed in potentially aggressive environments due tothe presence of chloride ions.

Table 3Determination of the series.

2. Materials and methods

2.1. Experimental plan

A full factorial design was carried out in order to investigate concomitantly theeffects of multiple variables and their interactions in a response variable.

Thus, the independent or explanatory variables (type of cement, water/cementratio and curing period) are called factors, while the three types of cement used(CPII-Z-32, CPIII-40 and CPIV-32), the two water/cement ratios (0.4 and 0.7), andtwo curing periods (7–28 days) are the corresponding levels to each factor (Table 1).

In order to achieve the proposed objectives, capillary water absorption tests,compressive strength and accelerated corrosion, which are treated as dependentor response variables, were performed (Table 2). The water absorption test was car-ried out, since the penetration of water influences directly on the durability of rein-forced concrete structure exposed to chloride contaminated environment. Thecompressive strength test is of significant importance, since other properties of con-crete and mortar are directly related to this parameter. To corrosive process evalu-ation, an accelerated corrosion test was conducted, and the technique of corrosionelectrochemical potential was used.

Test specimens were molded using the three types of cement chosen to be ana-lyzed (CPII-Z-32, CPIII-40 and CPIV-32); two water/cement ratios with significantvariation (0.4 and 0.7) to evaluate the performance of cements in different micro-structural conditions, and two curing periods usually applied (7–28 days) [25,26].The features were defined by the determination index of mortar normal consistencytest in accordance with the procedures presented by the Brazilian standard NBRABNT 7215:1997 [25]. Table 3 shows the definition of the series used in the study.

Table 1Factors (explanatory variables) and the cor-responding levels to each factor.

Factors Levels

Type of Portland cement CPII-Z-32CPIII-40CPIV-32

Water/cement ratio 0.70.4

Curing period 7 days28 days

2.2. Materials

The physical and chemical properties of the three cements used are shown inTable 4. Table 4 also shows the limits specified by the Brazilian standards of the ce-ment used. CPII Z-32 and CPIV 32 had pozzolanic material in their composition, alsocalled natural pozzolan from volcanic rock, in levels of 12–43%, respectively. Therewas a percentage of 67% of blast furnace slag in the composition of CPIII-40.

The steel reinforcement used in prismatic specimens for electrochemical mea-surements of corrosion potential was the CA-60 class (reinforcing steel with flowresistance characteristic of 600 MPa) with 5 mm diameter.

2.3. Test procedures

The additional tests (water absorption, compressive strength) were performedaiming the understanding and interpretation of the accelerated corrosion test.

2.3.1. Capillary water absorptionThe molding of specimens for water absorption test was carried out according

to the Brazilian Standard NBR ABNT 7215:1997 [25], molding, therefore, cylindricalmortar specimens of 50 mm diameter and 100 mm tall. After molding, all speci-mens had the superior surface protected with a glass plate and remained in awet chamber (thermally insulated, climate-controlled environment, temperature23 ± 2 �C, fitted with shelves for storing specimens and water sprinklers to keepthe relative humidity P95%) for a period of 24 h, then they were de-molded. After-wards, the samples were kept in a wet chamber until the desired age (7 or 28 days).Before starting the testing, the specimens were dried in an oven at a temperature of(105 ± 5)�C until constant mass was obtained. After reaching constant mass, theywere cooled down at a laboratory environment (relative humidity P65% and tem-perature 23 ± 2 �C) for 24 h.

The methodology adopted in the experimental program was based on the Bra-zilian Standard ABNT NBR 9779:1995 [27]; to do so, three mortar cylindrical spec-imens were tested for each combination of independent variables. During thetesting, the water level was kept constant, 5 ± 1 mm above the lower surface ofthe specimens.

The absorption was monitored for 72 h in accordance with the Brazilian Stan-dard ABNT NBR 9779:1995 [27], from the weighing of specimens. It is noteworthythat, in addition to the measurements fixed by the standard after 3 h, 6 h, 24 h, 48 h,and 72 h; the weight evolution was also monitored after 30 min, 1 h, 2 h, 4 h, and5 h, for it is during the early hours when the highest capillary absorption speedoccurs.

2.3.2. Compressive strengthThe principle of the procedure of molding and performance of testing in itself is

described on the ABNT NBR 7215:1997 [25]. To determine the strength, four mortartest specimens were molded and conveniently capped with sulfur in order to rep-resent each series in the research; the equipment used has a upload speed of0,45 ± 0,15 MPa/s. All the test specimens were subject to the curing in a moist

Series Type ofcement

w/cratio

Curingperiod

Feature Cement content(kg/m3)

A1 CPII-Z-32 0.7 7 1:3.0 463A2 CPII-Z-32 0.7 28 1:3.0 463B1 CPII-Z-32 0.4 7 1:1.3 821B2 CPII-Z-32 0.4 28 1:1.3 821C1 CPIII-40 0.7 7 1:3.0 463C2 CPIII-40 0.7 28 1:3.0 463D1 CPIII-40 0.4 7 1:1.1 875D2 CPIII-40 0.4 28 1:1.1 875E1 CPIV-32 0.7 7 1:3.0 463E2 CPIV-32 0.7 28 1:3.0 463F1 CPIV-32 0.4 7 1:1.0 905F2 CPIV-32 0.4 28 1:1.0 905

Table 4Properties of Portland cement.

Chemicaldeterminations(%)

CPIIZ-32

LimitsABNT NBR11578:1991

CPIII40


CPIV32


Heat loss 5.44 66.5 3.83 64.5 4.48 64.5Silica (SiO2) 23.99 N.S. 20.98 N.S. 32.37 N.S.Aluminum

oxide(Al2O3)

4.70 N.S. 4.91 N.S. 5.09 N.S.

Iron oxide(Fe2O3)

2.46 N.S. 3.75 N.S. 2.29 N.S.

Calcium oxide(CaO)

53.41 N.S. 59.37 N.S. 46.55 N.S.

Free calciumoxide (CaO)

0.80 N.S. 0.61 N.S. 0.78 N.S.

Magnesiumoxide (MgO)

3.98 66.5 3.78 N.S. 3.54 66.5

Trioxide sulfate(SO3)

3.26 64.0 2.29 64.0 3.17 64.0

Sodium oxide(Na2O)

0.14 N.S. 0.03 N.S. 0.15 N.S.

Potassiumoxide (K2O)

1.65 N.S. 0.42 N.S. 1.46 N.S.

Insolubleresidue

14.34 616.0 0.88 61.5 25.77 N.S.

Carbon dioxide(CO2)

4.02 65.0 2.88 63.0 2.79 63.0

Specific mass(g/cm3)

2.84 N.S. 2.90 N.S. 2.82 N.S.

Specific area(m2/kg)

470 P260 606 N.S. 560 N.S.

N.S.- no specification.

712 V.C.O. Pereira et al. / Construction and Building Materials 40 (2013) 710–718

chamber till the age of disrupting. Test specimens from series ‘‘1’’ (A1, B1, C1, D1,E1, and F1) were broken within 7 days, whereas the series ‘‘2’’ (A2, B2, C2, D2, E2,and F2), within 28 days.

Fig. 1. Vertical and horizontal sections of the specimens.

2.3.3. Accelerated corrosionFor the accelerated corrosion test, prismatic mortar specimens were made

according to the methodology proposed by Andrade et al. [28] and adapted byMonteiro [29], with dimensions of 60 � 80 � 25 mm, two steel reinforcement barsof 5 mm in diameter, 100 mm long, and 10 mm coatings, as shown in Fig. 1. Beforemolding the specimens, the bars were submitted to the cleaning procedure de-scribed on ASTM G1-03 [30]. The mold used was made from plastics, which facili-tates the specimens’ de-molding and avoids the use of release agents, possibleintervening factor in the results.

Even before the molding, the bar area exposed to attack had been limited byplacing tape at its ends (Fig. 1), settling a well-defined area exposed to corrosionrepresenting the anodic region of the bar.

For each combination, four specimens were made by varying the water/cementratio (0.4 and 0.7) and the curing period (7–28 days). The specimens were cured in awet chamber (relative humidity P95% and temperature 23 ± 2 �C); thereafter, theywere kept in a laboratory environment until they presented constant mass. Theywere then subjected to accelerated corrosion test.

The specimens dimension was the reason for using the mortar instead of con-crete; the latter could hinder the use of coarse aggregate, and create difficulties dur-ing the molding process. Studies conducted by Winslow and Liu [31] have shownthat the mortar paste has a pore structure similar to the pore structure in concretepaste. Therefore, it can be assumed that the porous structure of mortar paste mayhave been a suitable model to the study of the concrete paste porosity, althoughthere may be some influence in the transition zone.

Flexible cables were connected to the free end of the reinforcement, to provideelectrical connection during the corrosion potential testing. This connection wasshielded with tape and on the tape, epoxy resin was applied (Fig. 1). Epoxy resinhas been used to hamper the penetration of aggressive agents through the uppersurface of the specimens and protect the top of the reinforcement.

To accelerate the corrosion process, the procedure proposed by Monteiro [29]consisting of semi-cycles of drying (5 days) and partial immersion in aqueous solu-tion with 5% NaCl (2 days) was used because it allows the aggressive agents trans-port by both capillary absorption, and the ingress of chlorides by diffusion. Thismethodology has been chosen for providing the performance of the main mecha-nisms of chlorides conveyance (capillarity absorption by ionic diffusion) in concrete[8,32] and by the fact that these two mechanisms have governed the effects on dry-ing and wetting cyclic tests [33]. In addition, the sodium chloride is the most impor-

tant chloride salt to steel corrosion in reinforced concrete structures, because it ispresent in major situations of aggression from external contamination, such as mar-ine environment [5].

Once properly cured and dried, the specimens were submitted to drying andwetting half-cycles. At drying condition, the specimens are arranged in a ventilatedoven, keeping the temperature at 50 �C. The partially submerged condition is tomaintain the level of solution of the container in a position which corresponds tohalf of the height of the bar exposure area. Before the onset of the cycling, the firstmeasuring of the corrosion potential was held, and then measurements after eachhalf-cycle were carried out.

In order to enable data analysis, after each half-cycle, calculation of the averagepotential measurements values obtained in each of the eight built-in bars in fourspecimens of each series was performed. Based on the results of the averages, theevaluation of the corrosion potential in the specimens was performed using theparameters set forth in ASTM C 876 [34] for the copper–copper sulfate referenceelectrode (CSE) presented in Table 5, which indicates the corrosion probability inthe reinforced concrete structures.

3. Result and discussion

3.1. Capillary water absorption

Fig. 2 shows results for all series with w/c ratio equals to 0.7.The graph indicates that the specimens cured for 28 days absorbedless water per unit area, when compared with those cured for7 days and molded with the same type of cement. The results indi-cate a decrease in water absorption up to 13.9% in the series thatextended the curing period. This result is due to improvement ofmicro-structure of specimen provided by prolonging the curingperiod.

The results confirm the beneficial effect of additions to reducethe water absorption for both curing ages studied, since the useof compound cement with larger amounts of mineral additions(CPIII-40 and CPIV-32) have reduced about 11–34% of waterabsorption by capillarity in relation to CPII-Z-32. Moreover, theCPIII-40 cement may be seen as the most efficient in preventingthe penetration of water by capillarity in the mortar.

Fig. 3 shows results for all series that used w/c ratio equals to0.4. For the series with w/c ratio 0.4, it is also observed that by pro-longing curing results in less water absorption in the specimens,which shows clearly the beneficial effect provided to mortar bycuring. A decrease of 22.5% in absorption when the curing periodis increased from 7 to 28 days was observed. The data also indicatethe beneficial effect provided by the use of CPIII-40 cement in themortar presenting a 30% decrease in water absorption.

It is noteworthy that most of the specimens with w/c ratio 0.4absorbed less water, when compared with the specimens withw/c ratio 0.7.


It is also noticed the capacity of CPIII cement in increasing thewater penetration resistance in the mortar. This is because of thebenefit brought by the addition of blast-furnace slag to the cementcomposition and the higher fineness presented by the CPIII-40(Table 4).

Fig. 4 shows the capillary absorption values obtained at the endof the testing for each type of cement studied.

Through the graph of Fig. 4, the tendency of increased absorp-tion by capillarity with increased water/cement ratio is verified.This phenomenon can be explained by more open and intercon-nected porosity to samples with high water/cement ratio [35].The water/cement ratio, as the porosity controller parameter, willinfluence the properties linked to the transport mechanisms, sinceby reducing the water/cement ratio there will be a decrease in themortar voids [36].

Through the graph of Fig. 4, it is clearly seen the difference inbehavior of the cement used as the absorption by capillarity, beingthe CPIII-40 cement with the smallest capillary absorption shownto two w/c ratios and the two curing periods studied. The worst re-sult is assigned to CPII-Z-32 cement, since it had the highestabsorption values for the conditions studied, as expected.

3.2. Compressive strength

For this test, the recommendation of Brazilian Standard ABNTNBR 7215:1997 [25] was followed, using as satisfactory coefficientof variation the one equal to or less than 6%. When this coefficientpresented values higher than 6%, the furthest average value wasexcluded, calculating again the average, and the coefficient ofvariation.

Fig. 5 shows the compressive strength values of mortar at 7–28 days.

The results shown correspond to what was expected, indicatingthat, as the w/c ratio decreases, the compressive strength in-creases. These results are characteristics of Abrams Law, in whichthe compressive strength is presented as an exponential functionhaving the w/c ratio as the exponent [9].

The highest value of average compressive strength at 7 days, forthe series with w/c ratio equal to 0.7, was noted in the CPIII-40;being even above to the series of CPII-Z-32 and CPIV-32 cementsat 11.0–28.7%, respectively. For the age of 28 days, the averagecompressive strength value of the series of CPIII-40 was also highercompared to the others, being above the series of CPII-Z-32 andCPIV-32 cements at 40.0–56.2%, respectively.

The tendency was the same for the series with w/c ratio equal to0.4, with higher average values at 7–28 days, to the series of CPIII-40 cement. At 7 days of age, the average resistance value of the ser-ies of CPIII-40 cement was superior to the series of CPII-Z-32 andCPIV-32 cements at 37.7% and 28.12%, respectively. For the ageof 28 days, the average resistance value of CPIII-40 was superiorto the series of CPII-Z-32 and CPIV-32 cements at 8.2–14.7%,respectively.

It is also observed that, for the series of CPIII-40 with w/c ratioequal to 0.4, there is a zone of coincident values in the variation ofresults region (variation bar between maximum and minimum val-ues) which indicates a very low increase of resistance, about 5.1%,

Table 5Parameters to evaluate the corrosion potential values according to ASTM C 876 [21].

Corrosion potential related to copper-copper sulfatereference electrode (mV)

Corrosionprobability (%)

Ecorr < �350 >90Ecorr > �200 <10�350 < Ecorr < �200 Uncertain

due to the specimens’ disruption age. That shows a fast resistancegain of the specimens with blast-furnace slag in their composition.

Fig. 6 shows the effect of the type of cement in the compressivestrength.

According to the results, it is clear that there are differences be-tween the compressive strength of the different types of the ce-ment studied. This difference can be explained by the fact thateach type of cement has a specific curve of resistance evolutionover time due to significant differences in the physical and chem-ical of different types of cement in early stages of hydration [35].

Fig. 6 shows that the CPIII-40 cement had the best result, withregard to compressive strength. The lowest resistance achieved isrelated to the CPIV-32 cement, which contains a large amount ofpozzolan (43%) in its composition.

Besides the importance of the chemical composition of the ce-ment, Silva [37] states that the grain size and the granulometricdistribution of cement are factors that impact at the final and ini-tial resistance, for, the thinner the cement, the faster its hydrationreaction, and the higher its reactivity. This way, the good perfor-mance of the CPIII-40 cement is on account of its being thinnercompared to other cement studied (Table 4).

It is further noted that such factors as water/cement ratio, theage and type of cement, significantly influence the compressivestrength. With the reduction in water/cement ratio and the in-crease in age, the features used had significant increases in com-pressive strength.

3.3. Accelerated corrosion

Throughout the test, considerably higher values of corrosion po-tential in the drying stage at 50 �C was observed, because of waterloss and consequent electrolyte volume reduction of the pores. Inpotential analysis, the tendency of measured values to presentthemselves more negative in the wetting stage, indicating an in-creased corrosion probability has also been observed.

In Fig. 7, results of series with w/c ratio equal to 0.7 subjected to7 days of curing are shown. According to the graph, the series ofCPII-Z-32 and CPIV-32 cements have revealed amplitude of similarpotential variation. However, the series of CPIII-40 cement showeda higher variation in the potential values when changing the partialimmersion stage to the drying one, showing more significantly po-sitive values on drying.

It is interesting to note that the three series entered the range ofvalues smaller than �350 mV (corrosion probability above 90%) atthe partial immersion stage at 21 days of test. It also appears thatthe potential values at the partial immersion stage were stabilizednear �600 mV.

In Fig. 8, results of series with w/c ratio equal to 0.7 submittedto the period of 28 days of curing have been presented. It is alsoverified that the series of CPII-Z-32 and CPIV-32 cements haveshown amplitude of similar potential variation but smaller thanthe CPIII-40 cement.

It turns out that the series of CPII-Z-32 and CPIV-32 cementshave joined the range of values more negative than �350 mV at21 days of testing at the wetting stage, just as it happened to theseries submitted to a shorter period of curing. However, theCPIII-40 cement series only reached values for corrosion probabil-ity higher than 90% after 28 days of testing, i.e. one more cycle thanthe other series.

For the series with w/c ratio equal to 0.7, specimens moldedwith CPIII-40 cement showed better results, since they were thelast ones to show values indicating high corrosion probability(>90%) when submitted to 28 days of curing. The series of CPIII-40 and CPIV-32 cements had similar results and did not show sig-nificant difference on the results due to the increased curingperiod.

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

0 0.5 1 2 3 4 5 6 24 48 72

Cap

illar

y ab

sorp

tion

(g/c

m²)

Time (h)

w/c=0.7Curing 7 and 28 days

CPII-Z -7 CPII-Z -28 CPIII -7 CPIII -28 CPIV -7 CPIV -28

Fig. 2. Capillary absorption as a function of time to specimens with w/c ratio = 0.7.

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

0 0.5 1 2 3 4 5 6 24 48 72

Cap

illar

y ab

sorp

tion

(g/c

m²)

Time (h)

w/c=0.4Curing 7 and 28 days

CPII-Z -7 CPII-Z -28 CPIII -7 CPIII -28 CPIV -7 CPIV -28

Fig. 3. Capillary absorption as a function of time to specimens with w/c ratio = 0.4.

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

CPII-Z CPIII CPIV

Cap

illar

y ab

sorp

tion

(g/c

m²)

Type of cement

w/c=0.7 - 7 days of curing




Fig. 4. Type of cement effect in the capillary absorption.


In Fig. 9, results of series with w/c ratio equal to 0.4 submittedto a curing time of 7 days have been presented. Analyzing thegraph, it is seen that these series show different amplitudes of po-tential variation.

It is verified that the series joined the range of more negativevalues than �350 mV at different times, being the CPIV-32 cementthe first to join at 28 days of testing, the CPII-Z-32 the second at

35 days (1 cycle after CPIV-32) and the CPIII-40 the last at 42 days(2 cycles after the CPIV-32).

Fig. 10 shows results of series with w/c ratio equal to 0.4 sub-mitted to 28 days of curing time. According to the graph, variableamplitudes of potential to all three series have been verified.

It is clearly noticed the beneficial effect caused by the extendedcuring period for the series with w/c ratio 0.4, since the series(other than the cement CPIV-32) delayed to provide corrosionprobability higher than 90%, which represents an increase of corro-sion resistance. The series of CPII-Z-32 cement entered the regionof more negative values than �350 mV after 42 days of testing,i.e. by the end of the 6th cycle. The series of CPIII-40 cementshowed the best results, considering that it only started to presentcorrosion probability above 90% at the end of the 8th cycle (56 daysof testing), two cycles after the series of CPIII-Z-32 cement.

For the series with w/c ratio equal to 0.4, the specimens made ofCPIII-40 cement also showed the best results, followed by the ser-ies of CPII-Z-32 and CPIV-32 cements, following this order; con-firming, in general, the results of water absorption test.

In general, there was an improvement of corrosion resistance onthe series with w/c ratio reduction; this is because of the fact thatthe water/binder ratio is an important factor for the concrete rein-forcement protection, as long as alteration in the variable causessignificant variations to the corrosion rate presented in the struc-ture. Moreover, the less the water/cement ratio, the greater the

16.8220.38

34.50

46.17

18.68

28.55

47.52 49.96

14.5118.28

37.0943.57

0

10

20

30

40

50

60

Com

pres

sive

str

engt

h (M

Pa)

SeriesA1 -CPII-Z -7 A2 -CPII-Z -28 B1 -CPII-Z -7 B2 -CPII-Z -28C1 -CPIII -7 C2 -CPIII -28 D1 -CPIII -7 D2 -CPIII -28E1 -CPIV -7 E2 -CPIV -28 F1 -CPIV -7 F2 -CPIV -28

Fig. 5. Compressive strength values for each series studied.

0

10

20

30

40

50

60

CPII-Z-32 CPIII-40 CPIV-32Com

pres

sive

Str

engt

h (M

Pa)

Type of cement

w/c=0.7 - 7 days

w/c=0.7 - 28 days

w/c=0.4 - 7 days

w/c=0.4 - 28 days

Fig. 6. Effect of the type of cement in the compressive strength.


protection provided to the reinforcement, because of the pore sizereduction, and therefore, the higher the penetration resistance byaggressive agents and fluids [2].

The curing period has influenced more significantly the serieswith w/c ratio equal to 0.4, since the corrosion resistance had beenincreased with the prolonged curing period. The better the curingprocedure adopted, the higher the corrosion resistance of reinforcedconcrete structure, since the electrical resistivity of concrete hasbeen increased [38] and the chloride ions penetration reduced [39].

- 700

- 600

- 500

- 400

- 300

- 200

- 100

0

100

200

300

0 5 7 12 14 19 21 26 28 33 35 40 42 47

CSE

Cor

rosi

on P

oten

tial

(m

V)

Time (Days

CPII-Z -7 CPIII

Fig. 7. Evolution of the corrosion potentials of the s

Results show that for the two w/c ratio studied, the seriesmolded with CPIII-40 cement have presented higher corrosionresistance. Tumidajski and Chan [40] have confirmed this asser-tion, stating that concrete incorporating blast-furnace slag in par-tial replacement of Portland cement have been more efficient inpreventing the inflow of chloride ions than concrete with ordinarycement.

3.4. Statistical analysis

In order to intensify the analysis of work and the result inter-pretation of the developed experimental procedure, resultsachieved from the corrosion potential test have been submittedto statistical analysis using a model of experimental plan.

Aiming to verify whether the independent variables and theirinteractions influence on the dependent variable potential corro-sion, i.e., whether the factors are statistically significant for theexperimental procedure of this study; a full factorial design that al-lows to investigate simultaneously the effects of multiple variablesand their interactions in a variable response was carried out.

Through the analysis of variance, all hypotheses were tested fora confidence level equal to 95%, i.e. to a level of significance (errorprobability) equal to 5%.

49 54 56 61 63 68 70 75 77 82 84

)

w/c = 0.77 days of curing

-7 CPIV -7

Low corrosion probability

(10%)

Uncertain zone

Highcorrosion probability

(90%)

eries with w/c ratio = 0.7 and 7 days of curing.

-700

-600

-500

-400

-300

-200

-100

0

100

200

300

0 5 7 12 14 19 21 26 28 33 35 40 42 47 49 54 56 61 63 68 70 75 77 82 84

CSE

Cor

rosi

on P

oten

tial

(m

V)

Time (Days)


CPII-Z -28 CPIII -28 CPIV -28

Lowcorrosion probability

(10%)

Uncertain zone

Highcorrosion probability

(90%)

Fig. 8. Evolution of the corrosion potentials of the series with w/c ratio = 0.7 and 28 days of curing.

-700

-600

-500

-400

-300

-200

-100

0

100

200

300

0 5 7 12 14 19 21 26 28 33 35 40 42 47 49 54 56 61 63 68 70 75 77 82 84

CSE

Cor

r osi

o n P

oten

tial

(mV

)

Time (Days)



Lowcorrosionprobability

(10%)

Uncertainzone

Highcorrosionprobability

(90%)



In the corrosion test it was verified that the potential values(Ecorr) for the wetting half cycle (partial immersion) tended to sta-bilize close to �600 mV due to accelerated attack provided by thetesting. Thus, the performance of a single analysis on the corrosionfinal potential values could lead to distorted conclusions. There-fore, the analysis of potential results was chosen to be done inthree different stages, namely:

� at 21 days, when the series molded with water/cement ratioequal to 0.7 and submitted to 7 days of curing showed potentialvalues lower than �350 mV (corrosion probability above 90%),which represented 1=4 of the total test time;� at 42 days, representing half of the testing;� and at 84 days, at the end of the testing.

The potential values used in the analysis were always of thepartial immersion stage (wetting), since this stage has representedthe moment at which the corrosion process operates with higherintensity on the specimens due to the electrolyte presence.

Table 6 is a summary of variance analysis performed for thecorrosion potential test. It is seen in Table 6 that the type ofcement variables and curing period are statistically significant inthree analyzed moments, although a significance level less than

water/cement ratio in the two first moments have been shown.Despite the water/cement ratio to present higher importance inthe first two moments evaluated, this factor has not demonstratedto be meaningful to final potential values. This is probably the ten-dency, observed during the test, of the corrosion potential values tostabilize in �600 mV, reducing the variability of potential valuesand influencing the effect exerted by the water/cement ratio factor.Results indicating that the water/cement ratio is the most impor-tant factor about the potential results of corrosion have beenpresented by Tessari [41].

It may also be observed, that the interaction between the water/cement ratio and the curing period has not rendered any signifi-cance in any of the three analyzed moments, indicating that the ef-fect exerted by a variable on the corrosion potential is independentof the other.

Yet, according to Table 6, it is noteworthy that during the testthere had been tendencies regarding the significance of interac-tions between the variables. Note that during the test, the interac-tion between the type of cement and water/cement ratio had atendency to become significant. This finding has indicated thatfor old structures, with already initiated corrosive process, thevariables that interfere in the corrosion process are the type of ce-ment and water/cement ratio. Analogously, it is clear the contrary

-700

-600

-500

-400

-300

-200

-100

0

100

200

300

0 5 7 12 14 19 21 26 28 33 35 40 42 47 49 54 56 61 63 68 70 75 77 82 84

CSE

Cor

r osi

o n P

oten

tial

(mV

)

Time (Days)



Lowcorrosionprobability

(10%)

uncertainzone

Highcorrosionprobability

(90%)


Table 6Summary of variance analysis for the corrosion potential test.

Significance to the variables and their interactions

Source 21 days 42 days 84 days(last)

Type of cement S S SWater/cement ratio S S NSCuring period S S SType of cement � water/cement ratio NS S SType of cement � curing time S NS NSWater/cement ratio � curing time NS NS NSType of cement � water/cement ratio �

curing timeS NS NS

S – significant; NS – no significant.


convergence for interaction between the type of cement and thecuring period, as well as for triple interaction.

4. Conclusions

Taking into account the conditions of the test in which thesearch was carried out and the objective of this work to studythe ability to protect of some types of cement (CPII-Z-32, CPIII-40and CPIV-32) with regard to corrosion of reinforcement underthe action of chloride ions, it can be concluded that:

1. The reduction of the water/cement ratio has improved theproperties (such as compressive strength, corrosion resistanceand water penetration resistance) of mortars made for the threetypes of cement studied, as well as the corrosion resistance dueto the action of chloride ions.

2. For absorption by capillarity and compressive strength tests themortars molded with CPIII-40 cement performed the best results.

3. The water/cement ratio presented itself as an important factorto increase the durability of structures, together with the typeof cement used. This is because water/cement ratio is the con-troller parameter of porosity and permeability of concrete andmortar.

4. The CPIII-40 cement had the best performance with respect tocorrosion by chloride ions for the two water/cement ratio stud-ied, especially to the water/cement ratio equal to 0.4.

5. To increase the structures durability in marine environment,the cement to be used according to environmental conditions,potentially aggressive, in which structures will be exposed mustbe specified, involving the use of a low water/cement ratio anda prolonged curing period.

6. According to the additional tests carried out and the electro-chemical technique to detect potential corrosion applied onthe accelerated corrosion test, it was possible to classify thethree types of cement used in increasing performance order,as follows: CPIV-32, CPII-Z-32 and CPIII-40.

Acknowledgements

The authors would like to thank the Politécnica School–Univer-sity of Pernambuco (POLY-UPE), the PROCAD/NF and CAPES (Coor-dination of Improvement of Higher Education Personnel) for thefinancial support to conduct this survey.

References

[1] Helene PRL. Introduction. In: Helene PRL, Pereira F, editors. Rehabilitation andmaintenance of concrete structures (Rehabilitación y mantenimiento deestructuras de concreto). São Paulo: CARGRAPHICS; 2007. p. 17–32.

[2] Pereira E, Meneghetti LC, Resende A, Fabro G. Assessment of the level ofcorrosion of reinforcement for different thicknesses of coatings. In: BrazilianCongress of Concrete. Fortaleza: IBRACON; 2010. p. 52.

[3] Husni R, Benitez A, Manzelli A, Macchi C, Charreau G, Luco LF, et al. Actions andmechanisms of deterioration of the structures. In: Helene PRL, Pereira F,editors. Rehabilitation and maintenance of concrete structures (Rehabilitacióny mantenimiento de estructuras de concreto). São Paulo: CARGRAPHICS; 2007.p. 35–90.

[4] Cascudo O. Inspection and diagnosis of concrete structure with corrosionproblems of the reinforcement. In: Isaia GC, editor. Concrete: testing, researchand achievements, vol. 2. São Paulo: IBRACON; 2005. p. 1071–108.

[5] Ayra C, Buenfeld NR, Newman JB. Factors influencing chloride-binding inconcrete. Cem Concr Res 1990;20:29–300.

[6] DELLINGHAUSEN LM, GASTALDINI ALG, ISAIA GC, VANZIN FJF, VEIGA KK,ENGEL MK. Chloride penetration in concrete with various binders. In:BRAZILIAN CONGRESS OF CONCRETE, Curitiba: IBRACON, 2009; 51.

[7] Helene PRL. Contribution to the study of corrosion in reinforced-concreteprestressing. Thesis. Brazil: Department of Civil Engineering, University of SãoPaulo; 1993.

[8] Medeiros MHF, Helene P. Surface treatment of reinforced concrete in marineenvironment: influence on chloride diffusion coefficient and capillary waterabsorption. Constr Build Mater 2009;23:1476–84.

[9] Mehta K, Monteiro PJM. Concrete: microstructure, properties and materials.3rd ed. São Paulo: IBRACON; 2008.

[10] Nguyen TS, Lorente S, Carcasses M. Effect of the environment on the chloridediffusion through CEM-I and CEM-V mortars: an experimental study. ConstrBuild Mater 2009;23:795–803.

[11] Rasheeduzzafar, Al-Saadoun SS, Al-Gahtani AS, Dakhil FH. Effect of aluminatecontent of cement on corrosion of reinforced steel in concrete. Cem Concr Res1990;20:723–38.

[12] Silva MG, Silva VG, Tanesi J. A critical analysis on the life and durability in NBR6118/2003. Concr Constr 2010;58:96–101.

[13] Bijen J. Benefits of slag and fly ash. Constr Build Mater 1996;10:309–14.


[14] Bilodeau A, Malhotra VM. High-volume fly ash system: concrete solution forsustainable development. ACI Mater J 2000;97:41–8.

[15] Chindaprasirt P, Chotithanorm C, Cao HT, Sirivivatnanon V. Influence of fly ashfineness on the chloride penetration of concrete. Constr Build Mater2007;21:356–61.

[16] Ramezanianpour AA, Malhotra VM. Effect of curing on the compressivestrength, resistance to chloride-ion penetration and porosity of concretesincorporating slag, fly ash or silica fume. Cem Concr Compos 1995;17:125–33.

[17] Li Z, Peng J, Ma B. Investigation of chloride diffusion for high-performanceconcrete containing fly ash, microssilica and chemical admixtures. ACI Mater J1999;96:391–6.

[18] Chindaprasirt P, Rukzon S, Sirivivatnanon V. Effect of carbon dioxide onchloride penetration and chloride ion diffusion coefficient of blended Portlandcement mortar. Constr Build Mater 2008;22:1701–7.

[19] Fajardo G, Valdez P, Pacheco J. Corrosion of steel rebar embedded in naturalpozzolan based mortars exposed to chlorides. Constr Build Mater2009;23:768–74.

[20] Hussain SE, Rasheeduzzafar, Al-Musallam A, Al-Gahtani AS. Factors affectingthreshold chloride for reinforcement corrosion in concrete. Cem Concr Res1995;25:1543–55.

[21] Barberon F, Baroghel-Bouny V, Zanni H, Bresson B, la Caillerie JBE, Malosse L,et al. Interactions between chloride and cement-paste materials. Magn ResonImag 2005;23:267–72.

[22] Glass GK, Reddy B, Buenfeld NR. The participation of bound chloride in passivefilm breakdown on steel in concrete. Corros Sci 2000;42:2013–21.

[23] Ann KY, Kim T-S, Kim JH, Kim S-H. The resistance of high alumina cementagainst corrosion of steel in concrete. Constr Build Mater 2010;24:1502–10.

[24] Pereira VCO. Study on the influence cement type in corrosion of reinforcementunder action of chloride ions. Dissertation. Brazil: Department of CivilEngineering, University of Pernambuco; 2011.

[25] Abnt NBR. 7215 Portland cement – determination of compressivestrength. Brazil: Brazilian Association of Technical Norms; 1997.

[26] Abnt NBR. 6118 Design of concrete structures – procedure. Brazil: BrazilianAssociation of Technical Norms; 2007.

[27] Abnt NBR. 9779 Hardened Mortars and concrete – Determination of waterabsorption by capillarity. Brazil: Brazilian Association of Technical Norms; 1995.

[28] Andrade C, Castelo V, Alonso C, González JA. Determination of the corrosionrate of steel embedded in concrete by the polarization resistance and A.C.

impedance methods. Special technical publication 906, American Society forTesting and Materials; 1986.

[29] Monteiro ECB. Study of protection capacity of some types of cements, inrelation to the corrosion of reinforcement, under the joint action of chlorideions and CO2. Dissertation. Brazil: Department of Civil Engineering, Universityof Brasília; 1996.

[30] ASTM G1-03. Standard practice for preparing, cleaning, and evaluatingcorrosion test specimens. USA: American Society for Testing and Materials;2011.

[31] Winslow D, Liu D. The pore structure of paste in concrete. Cem Concr Res1990;20:227–35.

[32] Jucá TRP. Evaluation of free chlorides in concrete and Portland cement mortarsby the method of sprinkling silver nitrate solution. Dissertation. Brazil: CivilEngineering School, Federal University of Goiás; 2002.

[33] Hong K, Hooton RD. Effects of cyclic chloride exposure on penetration ofconcrete cover. Cem Concr Res 1999;29:1379–86.

[34] Astm C. 876 Standard test method for half-cell potentials of uncoatedreinforced steel in concrete. USA: American Society for Testing andMaterials; 2009.

[35] Neville AM. Properties of concrete. São Paulo: Pini; 1997.[36] Mendes MVAS, Cascudo O, Carasek H, Ferreira RB. Evaluation of transport

properties of concrete mass containing mineral additions. In: Braziliancongress of concrete. Fortaleza: IBRACON; 2010. p. 52.

[37] Silva FG. Study of high performance concrete face the action of chlorides.Doctoring Thesis. Brazil: Department of Science and Material Engineering,University of São Paulo; 2006.

[38] Guneyisi E, Ozturan T, Gesoglu M. A study on reinforcement corrosion andrelated properties of plain and blended cement concretes under differentcuring conditions. Cem Concr Compos 2004;27:449–61.

[39] Gastaldini ALG, Isaia GC, Saciloto AP, Missau F, Hoppe TF. Influence of curingtime on the chloride penetration resistance of concrete containing rice huskash: a technical and economical feasibility study. Cem Concr Compos2010;32:783–93.

[40] Tumidajski PJ, Chan GW. Effect of sulfate and carbon dioxide on chloridediffusivity. Cem Concr Res 1996;26:551–6.

[41] Tessari R. Study of the ability to protect some types of cements, in relation tothe corrosion of reinforcement under the action of chloride ions.Dissertation. Brazil: Federal University of Rio Grande do Sul; 2001.

1-s2.0-S0950061812008860-main

Documents

Transcript of 1-s2.0-S0950061812008860-main