- Email: [email protected]
Long-term chloride-induced corrosion monitoring of reinforced concrete coated with commercial polymer-modified mortar and polymeric coatings
Construction and Building Materials 48 (2013) 734–744
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Long-term chloride-induced corrosion monitoring of reinforced concrete coated with commercial polymer-modified mortar and polymeric coatings A. Brenna, F. Bolzoni, S. Beretta, M. Ormellese ⇑ Politecnico di Milano, Dipartimento di Chimica, Materiali e Ingegneria Chimica ‘‘Giulio Natta’’, Via Mancinelli 7, 20131 Milano, Italy
h i g h l i g h t s Efficiency of concrete coating was tested on concrete samples under accelerated chloride penetration. Organic coatings delay initiation of corrosion since they reduce chloride penetration. The coatings greatly increase the chloride apparent diffusion coefficient. Coatings reduce concrete water absorption thus reducing rebar corrosion rate.
a r t i c l e
i n f o
Article history: Received 4 December 2012 Received in revised form 23 July 2013 Accepted 25 July 2013 Available online 24 August 2013 Keywords: Coating Concrete Corrosion Chloride Diffusion
a b s t r a c t The efficiency of four commercial concrete coatings (a polymer modified cementitious mortar and three elastomeric coatings) against chloride-induced corrosion is discussed by means of steel corrosion longterm monitoring and by chlorides penetration profiles in concrete. The cement-based coating shows the best effect on delay chlorides penetration in concrete by acting as a physical barrier in addition to concrete cover. Despite its lower polymer content, the higher thickness guarantees a longer time-to-corrosion with respect to organic coatings. Once corrosion has started, corrosion rate is lower in the presence of coatings, due to their ability to reduce water ingress in concrete. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction On ordinary steel reinforcements embedded in alkaline concrete, a thin protective film (the so-called passive film) is thermodynamically stable and is formed spontaneously, as Pourbaix reported in his Atlas [1]. Corrosion of steel reinforcements represents the most widespread form of deterioration of concrete structures and its accurate knowledge is compulsory in order to predict the service life of reinforced concrete structures, which can be divided in two phases [2]. The first is corrosion initiation, during which steel is in passive condition and processes which can lead to steel depassivation (concrete carbonation or chloride penetration in the concrete cov-
⇑ Corresponding author. Tel.: +39 02 2399 3118; fax: +39 02 2399 3180. E-mail addresses: [email protected] (A. Brenna), [email protected] (F. Bolzoni), [email protected] (S. Beretta), [email protected] (M. Ormellese). 0950-0618/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2013.07.099
er) are taking place. In carbonation, carbon dioxide, diffusing into concrete, it neutralizes its alkalinity by the reaction with calcium hydroxide, so that the pH of the concrete pore solution decreases to a value lower than 9. In this condition, the passive film is not thermodynamically stable. Chloride ions penetrate into concrete in water solution causing a local breakdown of the passive film if their concentration at the metal surface reaches a critical threshold (in atmosphere between 0.4% and 1% by cement weight) [2]. The duration of this phase depends on concrete cover thickness and penetration rate of aggressive species. The second phase is corrosion propagation, which starts after corrosion initiation and ends when a limit state is achieved, over which corrosion cannot be further accepted. Once the passive layer is destroyed, corrosion occurs only in the presence of water and oxygen on the metal surface. Prevention of corrosion is firstly assured by casting an high quality concrete (i.e., proper concrete mixture proportion, W/C ratio and cover). If proper concrete cover cannot be assured, when a very long service life is required or in very severe environmental
735
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
exposure, the durability of the structure could be increased by adopting a specific preventative measure that modifies the characteristics of concrete, reinforcement, external environment or the structure itself. These techniques act by preventing aggressive species penetration in concrete or by controlling the corrosion process through inhibition of the anodic process or of the galvanic flow in the electrolyte. Nowadays, different preventative measures are claimed to prevent, or at least to reduce, steel corrosion in concrete [2]. In this wide context, concrete surface treatments and corrosion inhibitors [3] offer a possible way to improve concrete structures durability. By the way, the efficiency of organic substances as corrosion inhibitors and concrete coatings on chloride-induced corrosion was investigated during a research project started in 1997 and initially granted by an Italian company leader in the production of adhesives and chemical products for building. While the effectiveness of organic substances in preventing chloride-induced corrosion in synthetic alkaline pore solution and in concrete was recently discussed elsewhere [4,5], in this paper results about the effect of polymer-modified mortar and elastomeric coatings on time-to-corrosion and corrosion rate of reinforced concrete are reported. Commonly, it is possible to distinguish four principal classes of surface treatments for concrete [2]: organic coatings that form a continuous film, hydrophobic treatments that line the surface of the pores, treatments that fill the capillary pores and cementitious layers. Their effect is twofold: to reduce the permeability of aggressive agents in concrete and to decrease the water content of concrete with the resulting decrease of concrete electrical conductivity and corrosion rate [6–13]. In this work, the effect of three organic coatings and a polymer modified cementitious mortar on chloride induced corrosion of reinforced concrete was investigated. Organic coatings are commonly used to block the penetration of carbon dioxide or chloride ions by forming a continuous polymeric film on concrete surface of thickness from 0.1 to 1 mm. They are based on various types of polymers (e.g. acrylate, polyurethane, and epoxy), pigments and additives and their effectiveness is related to the absence of pores or defects. Cementitious coatings form a layer of low permeability and thickness of a few millimetres, typically lower than 10 mm. The mortar is generally fine grained and modified with polymers to decrease its permeability and to increase its bond to concrete. European standardizations [14,15] report several methods to test coating characteristics, as for instance water absorption, water vapour, chlorides and carbon dioxide permeability, adhesion, crack-bridging and self-repairing properties, mainly to short-term tests. There is a lack of data about long-term behaviour of organic and cement-based coatings. This work reports almost seven years corrosion monitoring of reinforced concrete specimens coated with cementitious and organic coatings. Coating efficiency was investigated by means of corrosion potential and corrosion rate monitoring and by chlorides penetration profiles in concrete.
2. Materials and methods 2.1. Concrete coatings Four commercial concrete coatings were tested: Coating A: two-components mortar based on cementitious binders, finegrained selected aggregates, special additives and synthetic acrylic polymers dispersed in water with polymer-to-cement ratio 0.33. Coating B: hydro-dispersed fibrous coating, based on elastomeric acrylic emulsions cement-free. Coating C: cement-free and elastomeric acrylic-based fibrous coating mixed with graded sand. Coating D: single-component acrylic resin-based paint in water dispersion which forms a flexible film on the concrete surface due to the action of natural light.
Polymeric coatings were applied on concrete after a previous surface treatment with a specific primer which aims to improve coating-to-concrete bonding. While the thickness of organic coatings (Coating B, C and D) is in the range between 0.2 and 1 mm, the mortar (Coating A) has a thickness between 1 and 3 mm. The tested coatings were commercial products available on the market on 2005; technical data are in conformity with the requirements of international standard tests. 2.2. Concrete mix design Concrete coatings (Section 2.1) were tested on laboratory concrete specimens prepared with a Portland-limestone cement, type CEM II A/L 42.5R that, according to EN 197-1 [16], indicates a cement containing from 6% to 20% by mass of limestone and with a minimum compressive strength of 20 and 42.5 MPa after 2 and 28 days, respectively. Concrete specimens were prepared with two water-to-cement ratios (W/C), 0.55 and 0.65, in order to study coating efficiency in concretes with different porosity. Limestone aggregates were used with maximum diameter 16 mm. An acrylic plasticizer was mixed to fresh concrete in order to guarantee S5 workability, according to EN 206-1 [17] (slump P220 mm). Concrete mix design is reported in Table 1. Two prismatic reinforced concrete specimens (Sample A and B, 340 250 50 mm, Fig. 1) were prepared for each experimental condition. Five carbon steel reinforcements (diameter 10 mm and length 290 mm) with chemical composition and mechanical properties according to EN 10080 [18] were placed in each concrete specimen; the ends (40 mm) of each steel bar were coated with heat shrinkage sleeve, so that only a length of 210 mm was exposed to concrete. Concrete cover was 20 mm. A Ti-MMO reference electrode and two stainless steel wires (diameter 2 mm) were placed next to each rebar for corrosion rate measurements. Specimens were cured for 28 days at 20 °C and 95% relative humidity, before the application of the coating on the top surface. The organic coatings (A, B and C) were applied only on concrete specimens prepared with the highest water-to-cement ratio (0.65). Plain concrete cubic samples (side 150 mm) were also prepared in order to allow concrete cores extraction for chloride concentration profiles determination. Concrete mix design and exposure condition are the same of reinforced specimens. 2.3. Exposure condition Both plain cubic samples and reinforced specimens were exposed to accelerated chlorides penetration wet–dry cycles, i.e. ponding cycles. A ponding cycle consist of one week wetting with a 5% sodium chloride solution (almost 30,000 mg/L chloride ions), and two weeks drying. The test solution was placed in contact with the upper surface of the specimen by putting the solution in a plastic box fixed on the top of the specimen. According to EN 206-1 [17], cyclic ponding to chloride solution can be classified as XS3 and XD3 exposure classes, which refer to corrosion induced by chlorides from seawater and other than from seawater, respectively. It should be pointed out that W/C ratio and concrete cover thickness (Section 2.2) were not adopted in agreement with standards requirements recommendations [17,19], in order to reduce chlorides penetration time into concrete, compatibly with the laboratory research time schedule. 2.4. Corrosion monitoring Almost seven years long tests were performed. Steel reinforcements corrosion was monitored by open circuit potential (in the following called corrosion potential, ECORR) measurement with respect to a saturated calomel reference electrode (SCE, +0.244 V SHE) placed on the upper surface of the concrete specimen by means of a wet sponge, and by linear polarization resistance (LPR) measurement, which is considered an accurate and rapid way to determine the instantaneous corrosion rate of steel reinforcements [20]. In LPR technique a small potential scan defined with respect to corrosion potential (DE = E ECORR) is applied to the metal and the polarization current (which varies approximately linearly with potential within a few millivolts from ECORR) is recorded. Corrosion current density (iCORR, mA/m2) is related to the specific LPR (X m2) by Stern–Geary equation [21]:
iCORR ¼ B=LRP
ð1Þ
where B is the Stern–Geary coefficient (related to anodic and cathodic Tafel slopes) which assumes approximately a value of 26 mV or 52 mV for steel in active or passive condition, respectively [2,21]. LPR measurement was carried out by means of an Table 1 Concrete mix design.
CEM II A/L 42.5 R Water Limestone aggregates Plasticizer
kg/m3 L/m3 kg/m3 % by cement weight
W/C = 0.55
W/C = 0.65
320 180 1876 0.64%
310 200 1812 0.14%
736
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
Fig. 1. Reinforced concrete specimens geometry.
EG&G 273 Potentiostat/Galvanostat applying a potential scan in the range ±10 mV with respect to ECORR with a scan rate of 0.16 mV/s [21]. The polarization current was supplied by means of two stainless steel counter-electrodes placed on both sides of the reinforcement (Fig. 1). During LPR test, potential was measured with respect to the internal Ti-MMO reference electrode, in order to minimize the ohmic drop contribution. LPR was calculated by a linear regression of the potential–current curve with the PowerCORRÓ software. 2.5. Chlorides penetration Concrete cores (30 mm in diameter) were extracted from cubic concrete specimens, cut into 10 mm slices and then crushed and dissolved in nitric acid. Chlorides content was measured by potentiometric titration with AgNO3 0.01 N [22]. Chlorides penetration in concrete is due to the presence of different mechanisms, mainly diffusion and capillary sorption. Nevertheless, for comparison purposes, experimental profiles were interpolated using the analytical solution of the second Fick’s law of diffusion, valid in non-stationary condition. Supposing that chloride concentration at the concrete surface (Cs) is constant with time and considering an effective chlorides diffusion coefficient (Deff) that does not vary with time and space, i.e.concrete is homogeneous, the analytical solution is:
Cðx; tÞ x ¼ 1 erf pffiffiffiffiffiffiffiffiffiffiffiffiffi Cs 2 Deff t
ð2Þ
where C (x, t) is the chloride concentration at the depth x after time t.
3. Results The effect of the coatings on chloride-induced corrosion has been investigated by means of corrosion potential and corrosion rate (calculated by LPR measurements) monitoring and by the determination of chlorides penetration profiles in concrete. Coating A (Section 2.1) was tested on concrete specimens with both water-to-cement ratios (0.55 and 0.65), while elastomeric coatings were applied on concrete specimens prepared only with W/C ratio 0.65. Overall, the corrosion behaviour of ten steel reinforcements for each condition was monitored.
3.1. Corrosion monitoring At the end of the wetting week of the ponding cycle, corrosion potential was measured with respect to a SCE reference electrode by a high impedance voltmeter. Corrosion potential measurement allows to assess the corrosion state of the reinforcement: generally, corroded and passive steels in concrete show a difference in corrosion potential up to 500 mV, due to the different electrochemical behaviour of active and passive areas. In practical applications on extended structures, potential measurement allows to locate corroding reinforcements, corresponding to the most negative values (half-cell potential mapping) [23]. As reported by the standard ASTM C876-09 [24], if potential is more positive than 200 mV CSE (Cu/CuSO4 saturated reference electrode, +318 mV SHE), the probability that no reinforcing steel corrosion is occurring at the time of the measurement is greater than 90%. Otherwise, if potential is in the range from 200 to 350 mV CSE corrosion activity is uncertain and if potential is more negative than 350 mV CSE there is a probability greater than 90% that steel is in corrosion condition. This criterion (derived empirically and so not universally applicable), provides an indication of the corrosion behaviour of steel in atmospherically exposed concrete but does not indicate steel corrosion rate. Steel corrosion rate (CR, mm/y) was calculated by means of LPR measurements [25]:
CR ¼ K iCORR EW=q
ð3Þ
where EW is carbon steel equivalent weight (27.92), iCORR is the corrosion current density (lA/cm2) calculated assuming 26 mV as Stern–Geary coefficient (B in Eq. (1)), q is the metal density (7.86 g/cm3) and K is a constant (3.27 103 mm g/lA cm y). For mild steel, a corrosion current of 1 mA/m2 (0.1 lA/cm2) corresponds to a corrosion rate of 1.17 lm/y. Generally, corrosion rate can be considered low if lower than 5 lm/y and negligible if lower
737
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
than 1 lm/y [26], considering the usual minimal acceptable service life of new designed structures (50 years). It should be specified that corrosion rate by means of Eq. (3) is a mean value calculated on the entire surface area of the rebar, and its use as an ‘‘exact’’ value can be misleading in the case of localized corrosion, as pitting corrosion, depending on macrocell formation and on the cathodic-to-anodic surface ratio, in particular if the concrete resistivity is low [2]. Nevertheless, the initiation of corrosion can be detected by the decrease in the values of LPR of at least one order of magnitude with respect to passive condition. 3.1.1. Concrete with W/C ratio 0.65 3.1.1.1. Polymer modified cementitious mortar (Coating A). Figs. 2 and 3 show corrosion potential (mV SCE) and corrosion rate (lm/ y) monitoring of reinforced concrete specimens (Sample A and B, respectively) with water-to-cement ratio 0.65, each one containing five steel reinforcements. Data are compared with steel corrosion potentials and corrosion rate in the reference condition, without concrete coating. Almost seven years monitoring was performed. Initially, concrete is chlorides-free and carbon steel reinforcements are in passive condition. Corrosion potential is approximately in the range between 0 and 150 mV SCE, which corresponds to a corrosion probability lower than 10% [24] and mean corrosion rate is between 0.1 and 0.3 lm/y. Generally, corrosion potential is determined by the intersection of anodic and cathodic potential– current density curves. While the anodic process (described by the anodic passive curve) could be considered the same for all the reinforcements (in the absence of chloride ions), the variation
(a) 100
W/C = 0.65
of oxygen concentration corresponding to steel surface can lead to corrosion potential variation. Indeed, corrosion potential is determined by the availability of oxygen to the metal surface. Generally, in concrete exposed to atmosphere, steel reinforcements show corrosion potential between +100 mV and 200 mV SCE [2]. Nevertheless, no remarkable differences of corrosion potential data in passive condition between coated and uncoated specimens are observed. A corrosion potential drop is measured after five ponding cycles (about 100 days) for the reinforced uncoated specimens. Steel potential decreases to values between 500 mV and 600 mV SCE, with a net drop of about 500 mV. The potential decrease can be related to passive film breakdown with the formation of localized anodic area (pits) on the metal surface surrounded by not corroded regions. Accordingly, corresponding to the corrosion potential drop, corrosion rate increases from the initial values up to 30 lm/y. Corrosion initiation is so confirmed by the simultaneous decrease of corrosion potential and the increase of corrosion rate. The presence of Coating A allows to increase time-to-corrosion with respect to the uncoated condition. Nevertheless, an odd behaviour can be observed comparing the two specimens (Sample A and Sample B) prepared with the same water-to-cement ratio. For Sample A (Fig. 2) corrosion potential never goes down 276 mV SCE (350 mV CSE, dotted line), which corresponds to a corrosion probability greater than 90% [24]. A corrosion potential drop is observed for two reinforcements, which potential decreases to about 250 mV SCE, corresponding to an uncertain corrosion condition. Conversely, corrosion rate never exceeds 1 lm/y, i.e. corrosion is negligible. On the other hand, Sample B shows the
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
1800
2100
0.1
2400
0
300
600
900
Time (days)
1200
1500
1800
2100
2400
Time (days)
Fig. 2. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating A (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
Time (days)
1800
2100
2400
0.1
0
300
600
900
1200
1500
1800
2100
2400
Time (days)
Fig. 3. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating A (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
738
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
worst corrosion behaviour: after 49 ponding cycles (about 1000 days testing), corrosion potential decreases to values lower than 276 mV SCE for all the reinforcements, indicating corrosion initiation. Then, after about one year during which corrosion potential remains below this value (reaching a minimum value of 465 mV SCE), it increases and shifts to more noble values, up to 130 mV SCE. The increase of corrosion potential could be associated to the variation of the anodic and/or cathodic curves of steel in concrete, as well as to the ohmic drop contribution. The growth of corrosion products on steel surface and the variation of oxygen availability can cause a change of corrosion potential, even if, under active corrosion in concrete exposed to atmosphere, a relevant increase of corrosion potential should principally be related to the increase of overpotential of anodic reaction. Nevertheless, passivity conditions are not restored as well as it can be obtained with other protection methods, mainly cathodic protection and electrochemical chloride removal, or with conventional repair methods based on replacing of all contaminated concrete, less effective due to the pitting auto-catalytic mechanism. Corrosion rate measurements are in agreement with corrosion potential data (Fig. 3).
3.1.1.2. Polymeric coatings (Coating B, Coating C, Coating D). Figs. 4 and 5 show corrosion potential and corrosion rate data of reinforced concrete specimens (Sample A and B, respectively) coated with Coating B (Section 2.1) compared with steel corrosion potentials and corrosion rate in the reference condition, without con-
(a) 100
W/C = 0.65
crete coating. Initially, concrete is chlorides-free and carbon steel reinforcements are in passive condition. In the presence of Coating B, time-to-corrosion increases significantly: corrosion starts after 750 days, and after 800 days (about two years of exposure) only one reinforcement is still in passive condition. Corresponding to the corrosion potential drop, corrosion rate increases of more than one order of magnitude from the initial values up to 10 lm/y. After corrosion initiation, similarly to concrete coated with the cement-based coating, corrosion potential increases gradually, even though it never reaches the initial values of passive condition. Corrosion tests were interrupted after four years of exposition: at the end of the test corrosion potential is in the range from 130 to 200 mV SCE and corrosion rate never exceeds 10 lm/y, remaining always lower than corrosion rate measured in the reference condition. Figs. 6 and 7 show corrosion potential and corrosion rate monitoring of reinforced concrete specimens (Sample A and B, respectively) coated with Coating C (Section 2.1). The earliest corrosion potential drop occurs after 450 days (15 months). A strong dispersion in time-to-corrosion is observed. After 30 months two steel reinforcements are still in passive condition while the remaining eight rebars are in corrosion condition showing potentials lower than 400 mV SCE. Corrosion tests were interrupted after four years of exposition: at the end of the test corrosion potential varies from 200 to 500 mV SCE. Corresponding to the corrosion potential drop, corrosion rate increases of more than one order of magnitude from the initial values (lower than 1 lm/y) up to 30 lm/y
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
Time (days)
600
750
900 1050 1200 1350 1500
Time (days)
Fig. 4. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating B (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a)100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
0.1
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
Fig. 5. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating B (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
739
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
600
Time (days)
750
900 1050 1200 1350 1500
Time (days)
Fig. 6. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating C (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
600
Time (days)
750
900 1050 1200 1350 1500
Time (days)
Fig. 7. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating C (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
for Sample A, which shows the worst corrosion behaviour (Fig. 6). Conversely, for Sample B (Fig. 7), corrosion rate is lower than the one measured in the reference condition and only for one steel reinforcement goes over 10 lm/y. Figs. 8 and 9 show corrosion potential and corrosion rate data referred to concrete specimens (Sample A and B, respectively) coated with Coating D. In the presence of the coating, passive condition is maintained up to about 600 days of exposition (20 months) showing corrosion potential always more positive than 100 mV SCE. Then, corrosion rate increases from the initial passive values (lower than 1 lm/y) up to values in the range from
(a)100
W/C = 0.65
1 to 30 lm/y. After 830 days (40 ponding cycles) nine of ten steel reinforcements are in corrosion condition. Whereupon, corrosion potential increases slowly up to values in the range from 200 to 280 mV SCE. After four years exposition only one steel reinforcement (Sample A, Fig. 8) is in passive condition with a potential higher than 100 mV SCE. 3.1.2. Concrete with W/C ratio 0.55 Figs. 10 and 11 show corrosion potential (mV SCE) and corrosion rate monitoring of reinforced concrete specimens (Sample A and B, respectively) with water-to-cement ratio 0.55 coated with
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
0.1
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
Fig. 8. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating D (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
740
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
600
Time (days)
750
900 1050 1200 1350 1500
Time (days)
Fig. 9. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating D (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
Coating A. Initially, concrete is chlorides-free and carbon steel reinforcements are in passive condition, i.e. corrosion rate is negligible. In the uncoated specimens, the earliest corrosion potential drop occurs after 7 ponding cycles: steel potential decreases to values between 500 mV and 600 mV, with a net drop of about 500 mV and corrosion rate increases of about two orders of magnitude. Otherwise, for both the coated specimens no remarkable difference in corrosion potential data can be observed with time. Only for three steel reinforcements (Figs. 10a and 11a), a slight corrosion potential decrease is observed to about 200 mV SCE. Nevertheless, corrosion rate never exceeds 1 lm/y for both the specimens
(a) 100
W/C = 0.55
and the decrease of corrosion potential does not find a significant confirmation by corrosion rate measurements. 3.2. Chlorides penetration Chlorides penetration profiles were determined on concrete cores: At the 52nd ponding cycle on reinforced concrete specimen coated with the polymer modified mortar (Coating A). At the 40th ponding cycle on cubic samples.
(b)100.0
W/C = 0.55
0
Corrosion rate (µm/y)
Potential (mV SCE)
Uncoated concrete
-100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
1800
2100
0.1
2400
0
300
600
900
Time (days)
1200
1500
1800
2100
2400
Time (days)
Fig. 10. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.55 in the presence of Coating A (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a) 100
W/C = 0.55
(b)100.0
W/C = 0.55
0
Corrosion rate (µm/y)
Potential (mV SCE)
Uncoated concrete
-100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
Time (days)
1800
2100
2400
0.1 0
300
600
900
1200
1500
1800
2100
2400
Time (days)
Fig. 11. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.55 in the presence of Coating A (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
741
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
Fig. 12 shows chlorides penetration profiles in reinforced concrete in the presence of Coating A varying concrete water-to-cement ratio (0.55 and 0.65) after 52 ponding cycles (about 3 years exposure). Data are compared with concentration profiles in the uncoated specimens obtained after 40 ponding cycles. Coating A reduces chlorides containing solution penetration into concrete: the best effect is observed in combination with the lowest W/C ratio (0.55). At the rebar level (20 mm) chlorides concentration is about 0.35% by cement weight, below the critical level usually considered for concrete in atmosphere (in the range from 0.4% to 1% by cement weight). Conversely, the uncoated specimen with W/C 0.55 shows chlorides concentration almost one order of magnitude greater, close to 2.5% by cement weight. Considering the highest water-to-cement ratio (0.65), a strong difference between the two coated specimens is observed: chloride concentrations in Sample B are greater than in Sample A, which shows a penetration profile slightly higher than those obtained with W/C equal to 0.55. Chlorides concentration in Sample B at the concrete cover thickness is about 2.3% by cement weight, largely over the critical level. In the uncoated specimens, chloride concentration is very high (close to 4% at the rebar level). As expected, water-to-cement ratio has a great effect on penetration rate of aggressive agents from the external environment, i.e. chlorides penetration rate increases by increasing water-to-cement ratio due to the higher cement paste porosity. Chloride profiles are in good agreement with corrosion potential and corrosion rate data discussed previously (Figs. 2 and 3): rebars under active corrosion are embedded in the concrete specimen B, in which the highest chloride penetration was measured. In order to understand this odd behaviour, coating thickness measurements (Table 2) were carried out on the concrete cores extracted for chlorides profile measurement by means of a Leica stereomicroscope. As expected, the lowest coating thickness (1.26 mm) was measured for Sample B with W/C equal to 0.65, probably due to the low accuracy in the manual application of the coating carried out by the furnisher. Coating thickness shows a great effect in establishing concrete coating efficiency, which does not depends only on its permeability and waterproofing properties but also on the barrier effect provided by the coating. In order to compare all the tested coatings, chlorides penetration profiles (Fig. 13) were determined after 40 ponding cycles (2 years of exposure) by extracting a concrete core from cubic samples with W/C 0.65. In all the specimens, chlorides content is lower with re-
5,5 5,0 4,5 4,0
3,0 2,5 2,0 1,5 1,0 0,5 0
10
20
30
40
Coating thickness (mm)
0.65 0.65 0.55 0.55
A B A B
2.59 1.26 2.91 3.02
4. Discussion Concerning service life of reinforced concrete structures subjected to corrosion, concrete coatings performance can be described in terms of their ability to influence time-to-corrosion initiation and corrosion propagation time (by reducing corrosion rate). These effects will be discussed separately. 4.1. Effect on time-to-corrosion Time-to-corrosion was measured as the ponding cycle corresponding to which the simultaneous drop of corrosion potential and the increase of corrosion rate was observed. Table 3 reports time-to-corrosion of reinforced concrete specimens in the tested conditions. All concrete coatings seem to be able to increase corrosion initiation time in concrete subjected to accelerated chlorides penetration. As discussed previously, time-to-corrosion depends on concrete W/C ratio: mean values vary from 5 to 9 considering the uncoated concretes prepared with W/C 0.65 and 0.55, respectively. W/C ratio seems to be a significant parameter also if the polymer modified mortar (Coating A) is applied on concrete surface. Nowadays, corrosion is not occurring in the presence of W/ C equal to 0.55. Conversely, in concrete prepared with W/C 0.65 steel corrosion started after 44 ponding cycles, with a mean value (calculated considering only rebars in corrosion condition) of 46, about nine times greater than that measured for the uncoated
Cycle # 40
5,5
3,5
0,0
Sample
6,0
W/C = 0.55 - Sample A W/C = 0.55 - Sample B W/C = 0.65 - Sample A W/C = 0.65 - Sample B W/C = 0.55 - Uncoated W/C = 0.65 - Uncoated
Coating A Cycle # 52
W/C
spect to the one measured in the uncoated concrete. The lowest chloride concentrations are measured in concrete with the cementitious mortar, while no strong differences are observed comparing the three polymeric coatings (Coating B, C and D). At the rebar level (20 mm) chlorides content is close to 0.25% by cement weight for concrete with the cementitious mortar, while it is in the range from 0.4% to 0.7% in concrete coated with polymeric coatings.
Chlorides content (% by cement weight)
Chlorides content (% by cement weight)
6,0
Table 2 Coating A thickness measurements.
50
Depth (mm) Fig. 12. Chlorides penetration profiles on concrete specimens coated with Coating A (W/C = 0.65 and W/C = 0.55). Profiles of uncoated samples were obtained after 40 cycles of ponding.
5,0
Coating A
Coating B
Coating C
Coating D
Uncoated 4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0,5 0,0
0
10
20
30
40
50
Depth (mm) Fig. 13. Chlorides penetration profiles on concrete specimens coated with all the tested coatings (W/C = 0.65) after 40 cycles of ponding. Profile of uncoated sample is also shown.
742
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
Table 3 Time-to-corrosion data. Coating
W/ C
# Passive rebars
# Corroded rebars
Reference
0.55 0.65
0/10 0/10
10/10 10/10
Coating A
0.55 0.65
10/10 5/10
Coating B
0.65
Coating C
0.65
Coating D
0.65
Time-to-corrosion (ponding cycle) Min
Max
Mean
Median
5 5
13 6
9 5
8 5
0/10 5/10
– 44
– 49
– 46
– 46
1/10
9/10
25
33
30
29
1/10
9/10
18
37
26
25
1/10
9/10
24
32
28
28
specimens. The cementitious mortar seems to provide the best efficiency to delay time-to-corrosion. In the presence of polymeric coatings, mean time-to-corrosion varies in the range from 26 (Coating C) to 30 (Coating B), about six times the initiation time measured in the reference condition. Generally, the increase of corrosion initiation can be related to two different effects: Reduction of chlorides transport rate in concrete (Section 4.1.1). Increase of critical chlorides content for the initiation of corrosion (Section 4.1.2).
4.1.1. Effect on chlorides penetration Chloride penetration rate depends on concrete quality (porosity, related primarily to W/C ratio, compaction and curing condition), on exposure condition and on the presence of additional protective measures, as concrete coatings. Indeed, the application of a concrete coating provides an additional physical barrier to the chloride-containing solution penetration in the concrete cover. In order to compare the efficiency of the tested coatings to reduce chlorides penetration, chloride concentration profiles (Fig. 13) were analysed by means of the analytical solution of the Fick’s second law (Eq. (2)). Nevertheless, some considerations are necessary. The analytical solution of Fick’s second law is valid in non-stationary condition, supposing that chlorides concentration at the concrete surface (Cs) is constant in time (C = Cs for x = 0 and any t) and that chlorides diffusion coefficient (D) does not vary with time and space, i.e. concrete is homogeneous. While Fick’s law describes only diffusion transport, chlorides penetration in concrete occurs in general by means of different mechanisms, as capillary sorption due to the capillary action of concrete pores, and depends also on chlorides binding by concrete. In particular, pure diffusion transport is verified only in concrete completely and permanently saturated with water. Accordingly, Fick’s second law is not universally applicable to describe chloride transport in concrete and an effective diffusion coefficient (Deff) is usually considered in order to take into account the existence of different transport mechanisms. Moreover, the boundary condition expressed by the invariance of chlorides concentration at the concrete surface (Cs) in time could be too restrictive in the tested conditions. Cs is the chloride concentration at the interface between coating and concrete and increases with time due to the presence of the coating that acts as a physical barrier to chlorides penetration. Although the boundary condition on surface concentration is not verified, the fitting operation of chlorides penetration profile was carried out only for comparison purpose. The calculated diffusion coefficient assumes exclusively the significance of regression parameter and its use as an intrinsic property of concrete could be misleading.
Chloride concentration profiles (Figs. 12 and 13) were analysed by fitting the analytical solution of Fick’s second law to the experimental data by using chlorides surface concentration (Cs) and effective diffusion coefficient (Deff) as fitting parameters using the least squares method (Tables 4 and 5). Diffusion coefficient decreases by decreasing the W/C ratio (Table 4) that, as expected, represents a crucial parameter influencing penetration rate and time-to-corrosion. Once chloride-containing solution penetrates through coating thickness, it proceeds in the concrete cover until steel reinforcements are reached. By using an electrical analogy, concrete coating and concrete cover represent two resistors which act in series as barrier to chlorides penetration and which resistance depends on chlorides permeability and thickness. In the uncoated specimens, Deff increases from 10.9 108 to 29.8 108 cm2/s varying the W/C ratio from 0.55 to 0.65. Surface chlorides concentration (Cs) provides information about the barrier effect of the coating on reducing chlorides penetration from the external environment. Without coating, Cs assumes values higher than 4% by cement weight, similar to that in the splash zone of a marine structure where evaporation of water leads to an increase in the chloride content at the concrete surface [2]. In the presence of Coating A, Cs is reduced of about one order of magnitude up to 0.6% by cement weight (Table 4). Data obtained at the 40th ponding cycle (Table 5) show that all the tested coatings reduce diffusion coefficient of about one order of magnitude with respect to the uncoated concrete. Deff varies in the range from 2.2 to 6.4 108 cm2/s for polymeric coatings and is 2.3 108 cm2/s for the polymeric modified cementitious mortar (Coating A). The lowest chlorides content at the concrete-coating interface (Cs) is measured in the presence of Coating A which provides the best barrier effect against chlorides penetration. Results allow to state that all the tested concrete coatings show a ‘‘physical-barrier’’ effect, by reducing the chlorides ingress in concrete from the external environment. The best efficiency is shown by Coating A, in agreement with corrosion potential and corrosion rate monitoring. Table 4 Surface chloride content (Cs) data obtained by interpolation of experimental concentration profiles of Fig. 12 by non-linear regression with Fick’s second law. Coating
a b
W/C
Ponding cycle
Chlorides profile measured in reinforced specimens Deff (108 cm2/ s)
Cs (% by cement weight)
Reference
0.55 0.65
40 40
10.9 29.8
4.2 4.4
A
0.55a 0.65b
52 52
8.8 17.0
0.6 0.6
Mean of Sample A and B considered. Only Sample A considered.
Table 5 Chloride effective diffusion coefficient (Deff) and surface chloride content (Cs) data obtained by interpolation of experimental concentration profiles of Fig. 13 by nonlinear regression with second Fick’s law. Coating
Reference A B C D
W/ C
0.65 0.65 0.65 0.65 0.65
Ponding cycle
40 40 40 40 40
Chlorides profile measured in cubic specimens Deff (108 cm2/ s)
Cs (% by cement weight)
29.8 2.3 2.2 6.4 3.1
4.4 0.6 2.5 1.2 1.8
743
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
ð4Þ
where Cs,ext is the surface chloride concentration at the interface between the coating and the external environment, Cs is the chloride concentration at the concrete-coating interface (Table 5) and tcoat is the coating thickness. Cs,ext was considered equal to that measured for the uncoated specimen in the same exposure condition (4.4% by cement weigh). Due to the highest thickness, Coating A shows the lowest specific coating resistance with respect to organic coatings. The highest value is shown by the organic paint (Coating D): the thin layer of the paint (0.22 mm) and the high polymer content provide an effective barrier with a great chlorides concentration drop for unit thickness. Polymeric coatings exhibit the highest specific coating resistance to chlorides penetration. Nevertheless, the cementitious mortar, despite its higher porosity, provides the best effect against chlorides penetration due to the higher thickness (2.33 mm, Table 6). This confirms that for concrete coatings, a minimum thickness value must be assured on the entire concrete surface in order to guarantee the protective effect of the coating and then the design lifetime of the structure. 4.1.2. Effect on critical chlorides threshold Critical chlorides threshold was calculated taking into account the measured time-to-corrosion (Table 3) by using the parameters obtained by chloride profiles interpolation (Deff, Cs, Table 5). In respective to the type of the applied coating, the obtained value is close to 0.5% with respect the cement weight. Moreover, values are in the typical interval generally considered for concrete structures exposed to atmosphere (0.4–1% by cement weight): coatings have no effect on critical chloride threshold. 4.2. Effect on corrosion rate The effect of concrete coatings on corrosion propagation time can be related to their ability to reduce water penetration in concrete, increasing concrete electrical resistivity (ohmic control of the corrosion process). In any case, concerning the designed service life of a reinforced concrete structure subjected to chloride-induced corrosion, the conservative approach is to assume the time at which repair actions should be planned on the structure equal to the initiation time. Propagation time may be shorter than initiation time and is traditionally not taken into account because of the uncertainty with regard to the consequences of localized corrosion. Nevertheless, depending on the limit state and on the actual corrosion rate, propagation time could be sufficiently long to be
Table 6 Specific coating resistance to chlorides penetration (rcoat). Coating
A B C D
tcoat
Cs,ext
Cs
DC = Cs,ext Cs
mm
% by cem. weight
% by cem. weight
% by cem. weight
rcoat = DC/ tcoat (% Cl)/ mm
2.33 0.69 0.71 0.22
4.4 4.4 4.4 4.4
0.6 2.5 1.2 1.8
3.8 1.9 3.2 2.6
1.6 2.8 4.5 11.8
0.9 0.8
Negligible corrosion rate
r coat ¼ ðC s;ext C s Þ=t coat
1.0
Cumulative frequency
In order to estimate coating effectiveness on reduce chlorides ingress in concrete, an empirical parameter was introduced to calculate the resistance of unit coating thickness to chlorides penetration. This parameter (called ‘‘specific coating penetration resistence, rcoat’’) does not depend on coating thickness and is proportional to the concentration gradient in the coating, assumed linear because of the lower thickness of coatings with respect to the concrete cover. Specific coating penetration resistance (Table 6) is defined as:
0.7 0.6 0.5 0.4 0.3
Uncoated Coating A Coating B Coating C Coating D
0.2 0.1 0.0
0
2
4
6
8
10
12
14
16
18
20
Corrosion rate (µm/y) Fig. 14. Cumulative frequency of corrosion rate for rebars embedded in uncoated concrete and concrete with coatings.
considered, at least for economic reasons. Fig. 14 shows the cumulative frequency curve of corrosion rate of steel reinforcements in the coated concrete specimens. Generally, corrosion rate can be considered negligible if below 1 lm/y, low between 2 and 5 lm/ y, moderate between 5 and 10 lm/y and very high for values above 100 lm/y [26]. Corrosion rates of steel reinforcements in the uncoated concrete are more dispersed and the 70% of data are greater than 10 lm/y, corresponding to a quite high corrosion rate. In the presence of the cementitious mortar (Coating A), this value decreases to about the 15%. The best behaviour is shown by Coating B and D: the 90% of corrosion rate values are lower than 7 and 9 lm/y, respectively. Conversely, corrosion rate data distribution in the presence of Coating C presents a great dispersion and the 40% of data are higher than 10 lm/y. Concrete coatings seem to be able to reduce corrosion rate of steel reinforcements since they the increase concrete electrical resistivity as a consequence of the reduction of the water content into concrete. Although corrosion rate is not negligible, steel reinforcements corrosion is in ohmic control and corrosion rate is limited to lower values with respect to the uncoated condition. The adoption of concrete coatings as a preventative corrosion method can extend corrosion propagation time by a reduction of corrosion rate. Nevertheless, it should be pointed out that the calculated corrosion rate is the mean value on the entire surface of the reinforcement and does not take into account the localized mechanism of chloride-induced corrosion. Pitting penetration rate can be higher than corrosion rate calculated from LPR measurements depending on the ratio between anodic and cathodic area [27,28].
5. Conclusions The efficiency of four commercial concrete coatings (a polymer modified cementitious mortar and three elastomeric coatings) against chloride-induced corrosion was studied by means of steel corrosion long-term monitoring and by chlorides penetration profiles in concrete. The polymer containing mortar shows the best effect on delay chlorides penetration in concrete by acting as a physical barrier in addition to concrete cover. Despite its lower polymer content with respect polymeric coatings, the higher thickness guarantees a longer time-to-corrosion. In the presence of aggressive environments or for long service life, the use of coating
744
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
must be associated with a good quality concrete (low W/C ratio and an adequate concrete cover). Once corrosion has started, corrosion rate is lower in the presence of coatings, due to their ability to reduce water ingress in concrete. References [1] Pourbaix M. Atlas of electrochemical equilibria in aqueous solutions, NACE Cebelcor; 1974. [2] Bertolini L, Elsener B, Pedeferri P, Polder RB. Corrosion of steel in concrete. Prevention, diagnosis, repair. Weinheim: Wiley-VCH; 2004. [3] Elsener B. Corrosion inhibitors for reinforced concrete – an EFC state of the art report, in: Corrosion of reinforcement in concrete – mechanisms, monitoring, inhibitors and rehabilitation techniques. Published for the European Federation of Corrosion by Woodhead Publishing and Maney Publishing, No. 38; 2007. p. 170–84. [4] Ormellese M, Bolzoni F, Lazzari L, Pedeferri P. Effect of corrosion inhibitors on the initiation of chloride-induced corrosion on reinforced concrete structures. Mater Corros 2008;59(2):98–106. [5] Ormellese M, Bolzoni F, Lazzari L, Brenna A, Pedeferri MP. Organic substances as inhibitors for chloride-induced corrosion in reinforced concrete. Mater Corros 2011;62(2):170–7. [6] Swamy RN, Tanikawa S. An external surface coating to protect concrete and steel from aggressive environments. Mater Struct 1993;26:465–78. [7] de Vries J, Polder RB. Hydrophobic treatment of concrete. Constr Build Mater 1997;11(4):259–65. [8] Seneviratne AMG, Sergi G, Page CL. Performance characteristics of surface coatings applied to concrete for control of reinforcement corrosion. Constr Build Mater 2000;14:55–9. [9] Al-Zahrani MM, Al-Dulaijan SU, Ibrahim M, Saricimen H, Sharif FM. Effect of waterproofing coatings on steel reinforcement corrosion and physical properties of concrete. Cem Concr Compos 2002;24:127–37. [10] Al-Zahrani MM, Maslehuddin M, Al-Dulaijan SU, Ibrahim M. Mechanical properties and durability characteristics of polymer- and cement-based repair materials. Cem Concr Compos 2003;25:527–37. [11] Almusallam AA, Khan FM, Dulaijan SU, Al-Amoudi OSB. Effectiveness of surface coatings in improving concrete durability. Cem Concr Compos 2003;25:473–81. [12] Raupach M, Wolff L. Investigations on durability of surface protection systems by accelerated laboratory and outdoor exposure tests. In: Proceedings of the European corrosion congress eurocorr. Lisbon, Portugal; 2005.
[13] Medeiros MHF, Helene P. Surface treatment of reinforced concrete in marine environment: influence on chloride diffusion coefficient and capillary water absorption. Constr Build Mater 2009;23(3):1476–84. [14] Cen, EN 1062, Paints and varnishes – coating materials and coating systems for exterior masonry and, concrete;2002. [15] Cen, EN 1504–2, Products and systems for the protection and repair of concrete structures - Definitions, requirements, quality control and evaluation of conformity - Part 2: Surface protection systems for, concrete, 2004. [16] CEN EN 197-1, Cement – Part 1: composition, specifications and conformity criteria for common cements; 2011. [17] CEN EN 206-1, Concrete—Part 1: specification, performance, production and conformity; 2000. [18] CEN EN 10080, Steel for the reinforcement of concrete – Weldable reinforcing steel – general; 2005. [19] Cen, EN 1992-1-1, Eurocode 2: design of concrete structures – Part 1-1: general rules and rules for, buildings; 2004. [20] Andrade C, Gonzalez JA. Quantitative measurements of corrosion rate of reinforcing steels embedded in concrete using polarization resistance measurements. Mater Corros 1978;29(8):515–9. [21] ASTM G59, Standard test method for conducting potentiodynamic polarization resistance measurements; 2009. [22] UNI 9944, Corrosion and protection of reinforcing steel in concrete. Determination of the carbonation depth and of the chlorides penetration profile in concrete; 1992. [23] Elsener B, Andrade C, Gulikers J, Polder R, Raupach M. Half-cell potential measurements – potential mapping on reinforced concrete structures. Mater Struct 2003;36:461–71. [24] ASTM C876, Standard test method for corrosion potentials of uncoated reinforcing steel in concrete; 2009. [25] ASTM G102, Standard practice for calculation of corrosion rates and related information from electrochemical measurements; 2010. [26] Andrade C. In: Corrosion of steel in reinforced structures, Final Report, COST action 521, Luxembourg, Office for Official Publications of the European Communities; 2003. [27] Gonzalez JA, Andrade C, Alonso C, Feliu S. Comparison of rates of general corrosion and maximum pitting penetration on concrete embedded steel reinforcement. Cem Concr Res 1995;25:257–64. [28] Ormellese M, Berra M, Bolzoni F, Pastore T. Corrosion inhibitors for chlorides induced corrosion in reinforced concrete structures. Cem Concr Res 2006;36:536–47.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Long-term chloride-induced corrosion monitoring of reinforced concrete coated with commercial polymer-modified mortar and polymeric coatings A. Brenna, F. Bolzoni, S. Beretta, M. Ormellese ⇑ Politecnico di Milano, Dipartimento di Chimica, Materiali e Ingegneria Chimica ‘‘Giulio Natta’’, Via Mancinelli 7, 20131 Milano, Italy
h i g h l i g h t s Efficiency of concrete coating was tested on concrete samples under accelerated chloride penetration. Organic coatings delay initiation of corrosion since they reduce chloride penetration. The coatings greatly increase the chloride apparent diffusion coefficient. Coatings reduce concrete water absorption thus reducing rebar corrosion rate.
a r t i c l e
i n f o
Article history: Received 4 December 2012 Received in revised form 23 July 2013 Accepted 25 July 2013 Available online 24 August 2013 Keywords: Coating Concrete Corrosion Chloride Diffusion
a b s t r a c t The efficiency of four commercial concrete coatings (a polymer modified cementitious mortar and three elastomeric coatings) against chloride-induced corrosion is discussed by means of steel corrosion longterm monitoring and by chlorides penetration profiles in concrete. The cement-based coating shows the best effect on delay chlorides penetration in concrete by acting as a physical barrier in addition to concrete cover. Despite its lower polymer content, the higher thickness guarantees a longer time-to-corrosion with respect to organic coatings. Once corrosion has started, corrosion rate is lower in the presence of coatings, due to their ability to reduce water ingress in concrete. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction On ordinary steel reinforcements embedded in alkaline concrete, a thin protective film (the so-called passive film) is thermodynamically stable and is formed spontaneously, as Pourbaix reported in his Atlas [1]. Corrosion of steel reinforcements represents the most widespread form of deterioration of concrete structures and its accurate knowledge is compulsory in order to predict the service life of reinforced concrete structures, which can be divided in two phases [2]. The first is corrosion initiation, during which steel is in passive condition and processes which can lead to steel depassivation (concrete carbonation or chloride penetration in the concrete cov-
⇑ Corresponding author. Tel.: +39 02 2399 3118; fax: +39 02 2399 3180. E-mail addresses: [email protected] (A. Brenna), [email protected] (F. Bolzoni), [email protected] (S. Beretta), [email protected] (M. Ormellese). 0950-0618/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2013.07.099
er) are taking place. In carbonation, carbon dioxide, diffusing into concrete, it neutralizes its alkalinity by the reaction with calcium hydroxide, so that the pH of the concrete pore solution decreases to a value lower than 9. In this condition, the passive film is not thermodynamically stable. Chloride ions penetrate into concrete in water solution causing a local breakdown of the passive film if their concentration at the metal surface reaches a critical threshold (in atmosphere between 0.4% and 1% by cement weight) [2]. The duration of this phase depends on concrete cover thickness and penetration rate of aggressive species. The second phase is corrosion propagation, which starts after corrosion initiation and ends when a limit state is achieved, over which corrosion cannot be further accepted. Once the passive layer is destroyed, corrosion occurs only in the presence of water and oxygen on the metal surface. Prevention of corrosion is firstly assured by casting an high quality concrete (i.e., proper concrete mixture proportion, W/C ratio and cover). If proper concrete cover cannot be assured, when a very long service life is required or in very severe environmental
735
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
exposure, the durability of the structure could be increased by adopting a specific preventative measure that modifies the characteristics of concrete, reinforcement, external environment or the structure itself. These techniques act by preventing aggressive species penetration in concrete or by controlling the corrosion process through inhibition of the anodic process or of the galvanic flow in the electrolyte. Nowadays, different preventative measures are claimed to prevent, or at least to reduce, steel corrosion in concrete [2]. In this wide context, concrete surface treatments and corrosion inhibitors [3] offer a possible way to improve concrete structures durability. By the way, the efficiency of organic substances as corrosion inhibitors and concrete coatings on chloride-induced corrosion was investigated during a research project started in 1997 and initially granted by an Italian company leader in the production of adhesives and chemical products for building. While the effectiveness of organic substances in preventing chloride-induced corrosion in synthetic alkaline pore solution and in concrete was recently discussed elsewhere [4,5], in this paper results about the effect of polymer-modified mortar and elastomeric coatings on time-to-corrosion and corrosion rate of reinforced concrete are reported. Commonly, it is possible to distinguish four principal classes of surface treatments for concrete [2]: organic coatings that form a continuous film, hydrophobic treatments that line the surface of the pores, treatments that fill the capillary pores and cementitious layers. Their effect is twofold: to reduce the permeability of aggressive agents in concrete and to decrease the water content of concrete with the resulting decrease of concrete electrical conductivity and corrosion rate [6–13]. In this work, the effect of three organic coatings and a polymer modified cementitious mortar on chloride induced corrosion of reinforced concrete was investigated. Organic coatings are commonly used to block the penetration of carbon dioxide or chloride ions by forming a continuous polymeric film on concrete surface of thickness from 0.1 to 1 mm. They are based on various types of polymers (e.g. acrylate, polyurethane, and epoxy), pigments and additives and their effectiveness is related to the absence of pores or defects. Cementitious coatings form a layer of low permeability and thickness of a few millimetres, typically lower than 10 mm. The mortar is generally fine grained and modified with polymers to decrease its permeability and to increase its bond to concrete. European standardizations [14,15] report several methods to test coating characteristics, as for instance water absorption, water vapour, chlorides and carbon dioxide permeability, adhesion, crack-bridging and self-repairing properties, mainly to short-term tests. There is a lack of data about long-term behaviour of organic and cement-based coatings. This work reports almost seven years corrosion monitoring of reinforced concrete specimens coated with cementitious and organic coatings. Coating efficiency was investigated by means of corrosion potential and corrosion rate monitoring and by chlorides penetration profiles in concrete.
2. Materials and methods 2.1. Concrete coatings Four commercial concrete coatings were tested: Coating A: two-components mortar based on cementitious binders, finegrained selected aggregates, special additives and synthetic acrylic polymers dispersed in water with polymer-to-cement ratio 0.33. Coating B: hydro-dispersed fibrous coating, based on elastomeric acrylic emulsions cement-free. Coating C: cement-free and elastomeric acrylic-based fibrous coating mixed with graded sand. Coating D: single-component acrylic resin-based paint in water dispersion which forms a flexible film on the concrete surface due to the action of natural light.
Polymeric coatings were applied on concrete after a previous surface treatment with a specific primer which aims to improve coating-to-concrete bonding. While the thickness of organic coatings (Coating B, C and D) is in the range between 0.2 and 1 mm, the mortar (Coating A) has a thickness between 1 and 3 mm. The tested coatings were commercial products available on the market on 2005; technical data are in conformity with the requirements of international standard tests. 2.2. Concrete mix design Concrete coatings (Section 2.1) were tested on laboratory concrete specimens prepared with a Portland-limestone cement, type CEM II A/L 42.5R that, according to EN 197-1 [16], indicates a cement containing from 6% to 20% by mass of limestone and with a minimum compressive strength of 20 and 42.5 MPa after 2 and 28 days, respectively. Concrete specimens were prepared with two water-to-cement ratios (W/C), 0.55 and 0.65, in order to study coating efficiency in concretes with different porosity. Limestone aggregates were used with maximum diameter 16 mm. An acrylic plasticizer was mixed to fresh concrete in order to guarantee S5 workability, according to EN 206-1 [17] (slump P220 mm). Concrete mix design is reported in Table 1. Two prismatic reinforced concrete specimens (Sample A and B, 340 250 50 mm, Fig. 1) were prepared for each experimental condition. Five carbon steel reinforcements (diameter 10 mm and length 290 mm) with chemical composition and mechanical properties according to EN 10080 [18] were placed in each concrete specimen; the ends (40 mm) of each steel bar were coated with heat shrinkage sleeve, so that only a length of 210 mm was exposed to concrete. Concrete cover was 20 mm. A Ti-MMO reference electrode and two stainless steel wires (diameter 2 mm) were placed next to each rebar for corrosion rate measurements. Specimens were cured for 28 days at 20 °C and 95% relative humidity, before the application of the coating on the top surface. The organic coatings (A, B and C) were applied only on concrete specimens prepared with the highest water-to-cement ratio (0.65). Plain concrete cubic samples (side 150 mm) were also prepared in order to allow concrete cores extraction for chloride concentration profiles determination. Concrete mix design and exposure condition are the same of reinforced specimens. 2.3. Exposure condition Both plain cubic samples and reinforced specimens were exposed to accelerated chlorides penetration wet–dry cycles, i.e. ponding cycles. A ponding cycle consist of one week wetting with a 5% sodium chloride solution (almost 30,000 mg/L chloride ions), and two weeks drying. The test solution was placed in contact with the upper surface of the specimen by putting the solution in a plastic box fixed on the top of the specimen. According to EN 206-1 [17], cyclic ponding to chloride solution can be classified as XS3 and XD3 exposure classes, which refer to corrosion induced by chlorides from seawater and other than from seawater, respectively. It should be pointed out that W/C ratio and concrete cover thickness (Section 2.2) were not adopted in agreement with standards requirements recommendations [17,19], in order to reduce chlorides penetration time into concrete, compatibly with the laboratory research time schedule. 2.4. Corrosion monitoring Almost seven years long tests were performed. Steel reinforcements corrosion was monitored by open circuit potential (in the following called corrosion potential, ECORR) measurement with respect to a saturated calomel reference electrode (SCE, +0.244 V SHE) placed on the upper surface of the concrete specimen by means of a wet sponge, and by linear polarization resistance (LPR) measurement, which is considered an accurate and rapid way to determine the instantaneous corrosion rate of steel reinforcements [20]. In LPR technique a small potential scan defined with respect to corrosion potential (DE = E ECORR) is applied to the metal and the polarization current (which varies approximately linearly with potential within a few millivolts from ECORR) is recorded. Corrosion current density (iCORR, mA/m2) is related to the specific LPR (X m2) by Stern–Geary equation [21]:
iCORR ¼ B=LRP
ð1Þ
where B is the Stern–Geary coefficient (related to anodic and cathodic Tafel slopes) which assumes approximately a value of 26 mV or 52 mV for steel in active or passive condition, respectively [2,21]. LPR measurement was carried out by means of an Table 1 Concrete mix design.
CEM II A/L 42.5 R Water Limestone aggregates Plasticizer
kg/m3 L/m3 kg/m3 % by cement weight
W/C = 0.55
W/C = 0.65
320 180 1876 0.64%
310 200 1812 0.14%
736
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
Fig. 1. Reinforced concrete specimens geometry.
EG&G 273 Potentiostat/Galvanostat applying a potential scan in the range ±10 mV with respect to ECORR with a scan rate of 0.16 mV/s [21]. The polarization current was supplied by means of two stainless steel counter-electrodes placed on both sides of the reinforcement (Fig. 1). During LPR test, potential was measured with respect to the internal Ti-MMO reference electrode, in order to minimize the ohmic drop contribution. LPR was calculated by a linear regression of the potential–current curve with the PowerCORRÓ software. 2.5. Chlorides penetration Concrete cores (30 mm in diameter) were extracted from cubic concrete specimens, cut into 10 mm slices and then crushed and dissolved in nitric acid. Chlorides content was measured by potentiometric titration with AgNO3 0.01 N [22]. Chlorides penetration in concrete is due to the presence of different mechanisms, mainly diffusion and capillary sorption. Nevertheless, for comparison purposes, experimental profiles were interpolated using the analytical solution of the second Fick’s law of diffusion, valid in non-stationary condition. Supposing that chloride concentration at the concrete surface (Cs) is constant with time and considering an effective chlorides diffusion coefficient (Deff) that does not vary with time and space, i.e.concrete is homogeneous, the analytical solution is:
Cðx; tÞ x ¼ 1 erf pffiffiffiffiffiffiffiffiffiffiffiffiffi Cs 2 Deff t
ð2Þ
where C (x, t) is the chloride concentration at the depth x after time t.
3. Results The effect of the coatings on chloride-induced corrosion has been investigated by means of corrosion potential and corrosion rate (calculated by LPR measurements) monitoring and by the determination of chlorides penetration profiles in concrete. Coating A (Section 2.1) was tested on concrete specimens with both water-to-cement ratios (0.55 and 0.65), while elastomeric coatings were applied on concrete specimens prepared only with W/C ratio 0.65. Overall, the corrosion behaviour of ten steel reinforcements for each condition was monitored.
3.1. Corrosion monitoring At the end of the wetting week of the ponding cycle, corrosion potential was measured with respect to a SCE reference electrode by a high impedance voltmeter. Corrosion potential measurement allows to assess the corrosion state of the reinforcement: generally, corroded and passive steels in concrete show a difference in corrosion potential up to 500 mV, due to the different electrochemical behaviour of active and passive areas. In practical applications on extended structures, potential measurement allows to locate corroding reinforcements, corresponding to the most negative values (half-cell potential mapping) [23]. As reported by the standard ASTM C876-09 [24], if potential is more positive than 200 mV CSE (Cu/CuSO4 saturated reference electrode, +318 mV SHE), the probability that no reinforcing steel corrosion is occurring at the time of the measurement is greater than 90%. Otherwise, if potential is in the range from 200 to 350 mV CSE corrosion activity is uncertain and if potential is more negative than 350 mV CSE there is a probability greater than 90% that steel is in corrosion condition. This criterion (derived empirically and so not universally applicable), provides an indication of the corrosion behaviour of steel in atmospherically exposed concrete but does not indicate steel corrosion rate. Steel corrosion rate (CR, mm/y) was calculated by means of LPR measurements [25]:
CR ¼ K iCORR EW=q
ð3Þ
where EW is carbon steel equivalent weight (27.92), iCORR is the corrosion current density (lA/cm2) calculated assuming 26 mV as Stern–Geary coefficient (B in Eq. (1)), q is the metal density (7.86 g/cm3) and K is a constant (3.27 103 mm g/lA cm y). For mild steel, a corrosion current of 1 mA/m2 (0.1 lA/cm2) corresponds to a corrosion rate of 1.17 lm/y. Generally, corrosion rate can be considered low if lower than 5 lm/y and negligible if lower
737
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
than 1 lm/y [26], considering the usual minimal acceptable service life of new designed structures (50 years). It should be specified that corrosion rate by means of Eq. (3) is a mean value calculated on the entire surface area of the rebar, and its use as an ‘‘exact’’ value can be misleading in the case of localized corrosion, as pitting corrosion, depending on macrocell formation and on the cathodic-to-anodic surface ratio, in particular if the concrete resistivity is low [2]. Nevertheless, the initiation of corrosion can be detected by the decrease in the values of LPR of at least one order of magnitude with respect to passive condition. 3.1.1. Concrete with W/C ratio 0.65 3.1.1.1. Polymer modified cementitious mortar (Coating A). Figs. 2 and 3 show corrosion potential (mV SCE) and corrosion rate (lm/ y) monitoring of reinforced concrete specimens (Sample A and B, respectively) with water-to-cement ratio 0.65, each one containing five steel reinforcements. Data are compared with steel corrosion potentials and corrosion rate in the reference condition, without concrete coating. Almost seven years monitoring was performed. Initially, concrete is chlorides-free and carbon steel reinforcements are in passive condition. Corrosion potential is approximately in the range between 0 and 150 mV SCE, which corresponds to a corrosion probability lower than 10% [24] and mean corrosion rate is between 0.1 and 0.3 lm/y. Generally, corrosion potential is determined by the intersection of anodic and cathodic potential– current density curves. While the anodic process (described by the anodic passive curve) could be considered the same for all the reinforcements (in the absence of chloride ions), the variation
(a) 100
W/C = 0.65
of oxygen concentration corresponding to steel surface can lead to corrosion potential variation. Indeed, corrosion potential is determined by the availability of oxygen to the metal surface. Generally, in concrete exposed to atmosphere, steel reinforcements show corrosion potential between +100 mV and 200 mV SCE [2]. Nevertheless, no remarkable differences of corrosion potential data in passive condition between coated and uncoated specimens are observed. A corrosion potential drop is measured after five ponding cycles (about 100 days) for the reinforced uncoated specimens. Steel potential decreases to values between 500 mV and 600 mV SCE, with a net drop of about 500 mV. The potential decrease can be related to passive film breakdown with the formation of localized anodic area (pits) on the metal surface surrounded by not corroded regions. Accordingly, corresponding to the corrosion potential drop, corrosion rate increases from the initial values up to 30 lm/y. Corrosion initiation is so confirmed by the simultaneous decrease of corrosion potential and the increase of corrosion rate. The presence of Coating A allows to increase time-to-corrosion with respect to the uncoated condition. Nevertheless, an odd behaviour can be observed comparing the two specimens (Sample A and Sample B) prepared with the same water-to-cement ratio. For Sample A (Fig. 2) corrosion potential never goes down 276 mV SCE (350 mV CSE, dotted line), which corresponds to a corrosion probability greater than 90% [24]. A corrosion potential drop is observed for two reinforcements, which potential decreases to about 250 mV SCE, corresponding to an uncertain corrosion condition. Conversely, corrosion rate never exceeds 1 lm/y, i.e. corrosion is negligible. On the other hand, Sample B shows the
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
1800
2100
0.1
2400
0
300
600
900
Time (days)
1200
1500
1800
2100
2400
Time (days)
Fig. 2. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating A (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
Time (days)
1800
2100
2400
0.1
0
300
600
900
1200
1500
1800
2100
2400
Time (days)
Fig. 3. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating A (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
738
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
worst corrosion behaviour: after 49 ponding cycles (about 1000 days testing), corrosion potential decreases to values lower than 276 mV SCE for all the reinforcements, indicating corrosion initiation. Then, after about one year during which corrosion potential remains below this value (reaching a minimum value of 465 mV SCE), it increases and shifts to more noble values, up to 130 mV SCE. The increase of corrosion potential could be associated to the variation of the anodic and/or cathodic curves of steel in concrete, as well as to the ohmic drop contribution. The growth of corrosion products on steel surface and the variation of oxygen availability can cause a change of corrosion potential, even if, under active corrosion in concrete exposed to atmosphere, a relevant increase of corrosion potential should principally be related to the increase of overpotential of anodic reaction. Nevertheless, passivity conditions are not restored as well as it can be obtained with other protection methods, mainly cathodic protection and electrochemical chloride removal, or with conventional repair methods based on replacing of all contaminated concrete, less effective due to the pitting auto-catalytic mechanism. Corrosion rate measurements are in agreement with corrosion potential data (Fig. 3).
3.1.1.2. Polymeric coatings (Coating B, Coating C, Coating D). Figs. 4 and 5 show corrosion potential and corrosion rate data of reinforced concrete specimens (Sample A and B, respectively) coated with Coating B (Section 2.1) compared with steel corrosion potentials and corrosion rate in the reference condition, without con-
(a) 100
W/C = 0.65
crete coating. Initially, concrete is chlorides-free and carbon steel reinforcements are in passive condition. In the presence of Coating B, time-to-corrosion increases significantly: corrosion starts after 750 days, and after 800 days (about two years of exposure) only one reinforcement is still in passive condition. Corresponding to the corrosion potential drop, corrosion rate increases of more than one order of magnitude from the initial values up to 10 lm/y. After corrosion initiation, similarly to concrete coated with the cement-based coating, corrosion potential increases gradually, even though it never reaches the initial values of passive condition. Corrosion tests were interrupted after four years of exposition: at the end of the test corrosion potential is in the range from 130 to 200 mV SCE and corrosion rate never exceeds 10 lm/y, remaining always lower than corrosion rate measured in the reference condition. Figs. 6 and 7 show corrosion potential and corrosion rate monitoring of reinforced concrete specimens (Sample A and B, respectively) coated with Coating C (Section 2.1). The earliest corrosion potential drop occurs after 450 days (15 months). A strong dispersion in time-to-corrosion is observed. After 30 months two steel reinforcements are still in passive condition while the remaining eight rebars are in corrosion condition showing potentials lower than 400 mV SCE. Corrosion tests were interrupted after four years of exposition: at the end of the test corrosion potential varies from 200 to 500 mV SCE. Corresponding to the corrosion potential drop, corrosion rate increases of more than one order of magnitude from the initial values (lower than 1 lm/y) up to 30 lm/y
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
Time (days)
600
750
900 1050 1200 1350 1500
Time (days)
Fig. 4. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating B (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a)100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
0.1
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
Fig. 5. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating B (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
739
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
600
Time (days)
750
900 1050 1200 1350 1500
Time (days)
Fig. 6. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating C (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
600
Time (days)
750
900 1050 1200 1350 1500
Time (days)
Fig. 7. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating C (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
for Sample A, which shows the worst corrosion behaviour (Fig. 6). Conversely, for Sample B (Fig. 7), corrosion rate is lower than the one measured in the reference condition and only for one steel reinforcement goes over 10 lm/y. Figs. 8 and 9 show corrosion potential and corrosion rate data referred to concrete specimens (Sample A and B, respectively) coated with Coating D. In the presence of the coating, passive condition is maintained up to about 600 days of exposition (20 months) showing corrosion potential always more positive than 100 mV SCE. Then, corrosion rate increases from the initial passive values (lower than 1 lm/y) up to values in the range from
(a)100
W/C = 0.65
1 to 30 lm/y. After 830 days (40 ponding cycles) nine of ten steel reinforcements are in corrosion condition. Whereupon, corrosion potential increases slowly up to values in the range from 200 to 280 mV SCE. After four years exposition only one steel reinforcement (Sample A, Fig. 8) is in passive condition with a potential higher than 100 mV SCE. 3.1.2. Concrete with W/C ratio 0.55 Figs. 10 and 11 show corrosion potential (mV SCE) and corrosion rate monitoring of reinforced concrete specimens (Sample A and B, respectively) with water-to-cement ratio 0.55 coated with
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
0.1
0
150
300
450
600
750
900 1050 1200 1350 1500
Time (days)
Fig. 8. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating D (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
740
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
(a) 100
W/C = 0.65
(b)100.0
W/C = 0.65 Uncoated concrete
Corrosion rate (µm/y)
Potential (mV SCE)
0 -100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
150
300
450
600
750
0.1
900 1050 1200 1350 1500
0
150
300
450
600
Time (days)
750
900 1050 1200 1350 1500
Time (days)
Fig. 9. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.65 in the presence of Coating D (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
Coating A. Initially, concrete is chlorides-free and carbon steel reinforcements are in passive condition, i.e. corrosion rate is negligible. In the uncoated specimens, the earliest corrosion potential drop occurs after 7 ponding cycles: steel potential decreases to values between 500 mV and 600 mV, with a net drop of about 500 mV and corrosion rate increases of about two orders of magnitude. Otherwise, for both the coated specimens no remarkable difference in corrosion potential data can be observed with time. Only for three steel reinforcements (Figs. 10a and 11a), a slight corrosion potential decrease is observed to about 200 mV SCE. Nevertheless, corrosion rate never exceeds 1 lm/y for both the specimens
(a) 100
W/C = 0.55
and the decrease of corrosion potential does not find a significant confirmation by corrosion rate measurements. 3.2. Chlorides penetration Chlorides penetration profiles were determined on concrete cores: At the 52nd ponding cycle on reinforced concrete specimen coated with the polymer modified mortar (Coating A). At the 40th ponding cycle on cubic samples.
(b)100.0
W/C = 0.55
0
Corrosion rate (µm/y)
Potential (mV SCE)
Uncoated concrete
-100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
1800
2100
0.1
2400
0
300
600
900
Time (days)
1200
1500
1800
2100
2400
Time (days)
Fig. 10. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.55 in the presence of Coating A (Sample A) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
(a) 100
W/C = 0.55
(b)100.0
W/C = 0.55
0
Corrosion rate (µm/y)
Potential (mV SCE)
Uncoated concrete
-100 -200 -300 -400 -500 -600 -700
10.0
1.0
Uncoated concrete
0
300
600
900
1200
1500
Time (days)
1800
2100
2400
0.1 0
300
600
900
1200
1500
1800
2100
2400
Time (days)
Fig. 11. Corrosion potential (a) and corrosion rate (b) monitoring on concrete specimens cast with W/C ratio 0.55 in the presence of Coating A (Sample B) during ponding cycles tests. Monitoring on uncoated concrete specimen is also shown.
741
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
Fig. 12 shows chlorides penetration profiles in reinforced concrete in the presence of Coating A varying concrete water-to-cement ratio (0.55 and 0.65) after 52 ponding cycles (about 3 years exposure). Data are compared with concentration profiles in the uncoated specimens obtained after 40 ponding cycles. Coating A reduces chlorides containing solution penetration into concrete: the best effect is observed in combination with the lowest W/C ratio (0.55). At the rebar level (20 mm) chlorides concentration is about 0.35% by cement weight, below the critical level usually considered for concrete in atmosphere (in the range from 0.4% to 1% by cement weight). Conversely, the uncoated specimen with W/C 0.55 shows chlorides concentration almost one order of magnitude greater, close to 2.5% by cement weight. Considering the highest water-to-cement ratio (0.65), a strong difference between the two coated specimens is observed: chloride concentrations in Sample B are greater than in Sample A, which shows a penetration profile slightly higher than those obtained with W/C equal to 0.55. Chlorides concentration in Sample B at the concrete cover thickness is about 2.3% by cement weight, largely over the critical level. In the uncoated specimens, chloride concentration is very high (close to 4% at the rebar level). As expected, water-to-cement ratio has a great effect on penetration rate of aggressive agents from the external environment, i.e. chlorides penetration rate increases by increasing water-to-cement ratio due to the higher cement paste porosity. Chloride profiles are in good agreement with corrosion potential and corrosion rate data discussed previously (Figs. 2 and 3): rebars under active corrosion are embedded in the concrete specimen B, in which the highest chloride penetration was measured. In order to understand this odd behaviour, coating thickness measurements (Table 2) were carried out on the concrete cores extracted for chlorides profile measurement by means of a Leica stereomicroscope. As expected, the lowest coating thickness (1.26 mm) was measured for Sample B with W/C equal to 0.65, probably due to the low accuracy in the manual application of the coating carried out by the furnisher. Coating thickness shows a great effect in establishing concrete coating efficiency, which does not depends only on its permeability and waterproofing properties but also on the barrier effect provided by the coating. In order to compare all the tested coatings, chlorides penetration profiles (Fig. 13) were determined after 40 ponding cycles (2 years of exposure) by extracting a concrete core from cubic samples with W/C 0.65. In all the specimens, chlorides content is lower with re-
5,5 5,0 4,5 4,0
3,0 2,5 2,0 1,5 1,0 0,5 0
10
20
30
40
Coating thickness (mm)
0.65 0.65 0.55 0.55
A B A B
2.59 1.26 2.91 3.02
4. Discussion Concerning service life of reinforced concrete structures subjected to corrosion, concrete coatings performance can be described in terms of their ability to influence time-to-corrosion initiation and corrosion propagation time (by reducing corrosion rate). These effects will be discussed separately. 4.1. Effect on time-to-corrosion Time-to-corrosion was measured as the ponding cycle corresponding to which the simultaneous drop of corrosion potential and the increase of corrosion rate was observed. Table 3 reports time-to-corrosion of reinforced concrete specimens in the tested conditions. All concrete coatings seem to be able to increase corrosion initiation time in concrete subjected to accelerated chlorides penetration. As discussed previously, time-to-corrosion depends on concrete W/C ratio: mean values vary from 5 to 9 considering the uncoated concretes prepared with W/C 0.65 and 0.55, respectively. W/C ratio seems to be a significant parameter also if the polymer modified mortar (Coating A) is applied on concrete surface. Nowadays, corrosion is not occurring in the presence of W/ C equal to 0.55. Conversely, in concrete prepared with W/C 0.65 steel corrosion started after 44 ponding cycles, with a mean value (calculated considering only rebars in corrosion condition) of 46, about nine times greater than that measured for the uncoated
Cycle # 40
5,5
3,5
0,0
Sample
6,0
W/C = 0.55 - Sample A W/C = 0.55 - Sample B W/C = 0.65 - Sample A W/C = 0.65 - Sample B W/C = 0.55 - Uncoated W/C = 0.65 - Uncoated
Coating A Cycle # 52
W/C
spect to the one measured in the uncoated concrete. The lowest chloride concentrations are measured in concrete with the cementitious mortar, while no strong differences are observed comparing the three polymeric coatings (Coating B, C and D). At the rebar level (20 mm) chlorides content is close to 0.25% by cement weight for concrete with the cementitious mortar, while it is in the range from 0.4% to 0.7% in concrete coated with polymeric coatings.
Chlorides content (% by cement weight)
Chlorides content (% by cement weight)
6,0
Table 2 Coating A thickness measurements.
50
Depth (mm) Fig. 12. Chlorides penetration profiles on concrete specimens coated with Coating A (W/C = 0.65 and W/C = 0.55). Profiles of uncoated samples were obtained after 40 cycles of ponding.
5,0
Coating A
Coating B
Coating C
Coating D
Uncoated 4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0,5 0,0
0
10
20
30
40
50
Depth (mm) Fig. 13. Chlorides penetration profiles on concrete specimens coated with all the tested coatings (W/C = 0.65) after 40 cycles of ponding. Profile of uncoated sample is also shown.
742
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
Table 3 Time-to-corrosion data. Coating
W/ C
# Passive rebars
# Corroded rebars
Reference
0.55 0.65
0/10 0/10
10/10 10/10
Coating A
0.55 0.65
10/10 5/10
Coating B
0.65
Coating C
0.65
Coating D
0.65
Time-to-corrosion (ponding cycle) Min
Max
Mean
Median
5 5
13 6
9 5
8 5
0/10 5/10
– 44
– 49
– 46
– 46
1/10
9/10
25
33
30
29
1/10
9/10
18
37
26
25
1/10
9/10
24
32
28
28
specimens. The cementitious mortar seems to provide the best efficiency to delay time-to-corrosion. In the presence of polymeric coatings, mean time-to-corrosion varies in the range from 26 (Coating C) to 30 (Coating B), about six times the initiation time measured in the reference condition. Generally, the increase of corrosion initiation can be related to two different effects: Reduction of chlorides transport rate in concrete (Section 4.1.1). Increase of critical chlorides content for the initiation of corrosion (Section 4.1.2).
4.1.1. Effect on chlorides penetration Chloride penetration rate depends on concrete quality (porosity, related primarily to W/C ratio, compaction and curing condition), on exposure condition and on the presence of additional protective measures, as concrete coatings. Indeed, the application of a concrete coating provides an additional physical barrier to the chloride-containing solution penetration in the concrete cover. In order to compare the efficiency of the tested coatings to reduce chlorides penetration, chloride concentration profiles (Fig. 13) were analysed by means of the analytical solution of the Fick’s second law (Eq. (2)). Nevertheless, some considerations are necessary. The analytical solution of Fick’s second law is valid in non-stationary condition, supposing that chlorides concentration at the concrete surface (Cs) is constant in time (C = Cs for x = 0 and any t) and that chlorides diffusion coefficient (D) does not vary with time and space, i.e. concrete is homogeneous. While Fick’s law describes only diffusion transport, chlorides penetration in concrete occurs in general by means of different mechanisms, as capillary sorption due to the capillary action of concrete pores, and depends also on chlorides binding by concrete. In particular, pure diffusion transport is verified only in concrete completely and permanently saturated with water. Accordingly, Fick’s second law is not universally applicable to describe chloride transport in concrete and an effective diffusion coefficient (Deff) is usually considered in order to take into account the existence of different transport mechanisms. Moreover, the boundary condition expressed by the invariance of chlorides concentration at the concrete surface (Cs) in time could be too restrictive in the tested conditions. Cs is the chloride concentration at the interface between coating and concrete and increases with time due to the presence of the coating that acts as a physical barrier to chlorides penetration. Although the boundary condition on surface concentration is not verified, the fitting operation of chlorides penetration profile was carried out only for comparison purpose. The calculated diffusion coefficient assumes exclusively the significance of regression parameter and its use as an intrinsic property of concrete could be misleading.
Chloride concentration profiles (Figs. 12 and 13) were analysed by fitting the analytical solution of Fick’s second law to the experimental data by using chlorides surface concentration (Cs) and effective diffusion coefficient (Deff) as fitting parameters using the least squares method (Tables 4 and 5). Diffusion coefficient decreases by decreasing the W/C ratio (Table 4) that, as expected, represents a crucial parameter influencing penetration rate and time-to-corrosion. Once chloride-containing solution penetrates through coating thickness, it proceeds in the concrete cover until steel reinforcements are reached. By using an electrical analogy, concrete coating and concrete cover represent two resistors which act in series as barrier to chlorides penetration and which resistance depends on chlorides permeability and thickness. In the uncoated specimens, Deff increases from 10.9 108 to 29.8 108 cm2/s varying the W/C ratio from 0.55 to 0.65. Surface chlorides concentration (Cs) provides information about the barrier effect of the coating on reducing chlorides penetration from the external environment. Without coating, Cs assumes values higher than 4% by cement weight, similar to that in the splash zone of a marine structure where evaporation of water leads to an increase in the chloride content at the concrete surface [2]. In the presence of Coating A, Cs is reduced of about one order of magnitude up to 0.6% by cement weight (Table 4). Data obtained at the 40th ponding cycle (Table 5) show that all the tested coatings reduce diffusion coefficient of about one order of magnitude with respect to the uncoated concrete. Deff varies in the range from 2.2 to 6.4 108 cm2/s for polymeric coatings and is 2.3 108 cm2/s for the polymeric modified cementitious mortar (Coating A). The lowest chlorides content at the concrete-coating interface (Cs) is measured in the presence of Coating A which provides the best barrier effect against chlorides penetration. Results allow to state that all the tested concrete coatings show a ‘‘physical-barrier’’ effect, by reducing the chlorides ingress in concrete from the external environment. The best efficiency is shown by Coating A, in agreement with corrosion potential and corrosion rate monitoring. Table 4 Surface chloride content (Cs) data obtained by interpolation of experimental concentration profiles of Fig. 12 by non-linear regression with Fick’s second law. Coating
a b
W/C
Ponding cycle
Chlorides profile measured in reinforced specimens Deff (108 cm2/ s)
Cs (% by cement weight)
Reference
0.55 0.65
40 40
10.9 29.8
4.2 4.4
A
0.55a 0.65b
52 52
8.8 17.0
0.6 0.6
Mean of Sample A and B considered. Only Sample A considered.
Table 5 Chloride effective diffusion coefficient (Deff) and surface chloride content (Cs) data obtained by interpolation of experimental concentration profiles of Fig. 13 by nonlinear regression with second Fick’s law. Coating
Reference A B C D
W/ C
0.65 0.65 0.65 0.65 0.65
Ponding cycle
40 40 40 40 40
Chlorides profile measured in cubic specimens Deff (108 cm2/ s)
Cs (% by cement weight)
29.8 2.3 2.2 6.4 3.1
4.4 0.6 2.5 1.2 1.8
743
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
ð4Þ
where Cs,ext is the surface chloride concentration at the interface between the coating and the external environment, Cs is the chloride concentration at the concrete-coating interface (Table 5) and tcoat is the coating thickness. Cs,ext was considered equal to that measured for the uncoated specimen in the same exposure condition (4.4% by cement weigh). Due to the highest thickness, Coating A shows the lowest specific coating resistance with respect to organic coatings. The highest value is shown by the organic paint (Coating D): the thin layer of the paint (0.22 mm) and the high polymer content provide an effective barrier with a great chlorides concentration drop for unit thickness. Polymeric coatings exhibit the highest specific coating resistance to chlorides penetration. Nevertheless, the cementitious mortar, despite its higher porosity, provides the best effect against chlorides penetration due to the higher thickness (2.33 mm, Table 6). This confirms that for concrete coatings, a minimum thickness value must be assured on the entire concrete surface in order to guarantee the protective effect of the coating and then the design lifetime of the structure. 4.1.2. Effect on critical chlorides threshold Critical chlorides threshold was calculated taking into account the measured time-to-corrosion (Table 3) by using the parameters obtained by chloride profiles interpolation (Deff, Cs, Table 5). In respective to the type of the applied coating, the obtained value is close to 0.5% with respect the cement weight. Moreover, values are in the typical interval generally considered for concrete structures exposed to atmosphere (0.4–1% by cement weight): coatings have no effect on critical chloride threshold. 4.2. Effect on corrosion rate The effect of concrete coatings on corrosion propagation time can be related to their ability to reduce water penetration in concrete, increasing concrete electrical resistivity (ohmic control of the corrosion process). In any case, concerning the designed service life of a reinforced concrete structure subjected to chloride-induced corrosion, the conservative approach is to assume the time at which repair actions should be planned on the structure equal to the initiation time. Propagation time may be shorter than initiation time and is traditionally not taken into account because of the uncertainty with regard to the consequences of localized corrosion. Nevertheless, depending on the limit state and on the actual corrosion rate, propagation time could be sufficiently long to be
Table 6 Specific coating resistance to chlorides penetration (rcoat). Coating
A B C D
tcoat
Cs,ext
Cs
DC = Cs,ext Cs
mm
% by cem. weight
% by cem. weight
% by cem. weight
rcoat = DC/ tcoat (% Cl)/ mm
2.33 0.69 0.71 0.22
4.4 4.4 4.4 4.4
0.6 2.5 1.2 1.8
3.8 1.9 3.2 2.6
1.6 2.8 4.5 11.8
0.9 0.8
Negligible corrosion rate
r coat ¼ ðC s;ext C s Þ=t coat
1.0
Cumulative frequency
In order to estimate coating effectiveness on reduce chlorides ingress in concrete, an empirical parameter was introduced to calculate the resistance of unit coating thickness to chlorides penetration. This parameter (called ‘‘specific coating penetration resistence, rcoat’’) does not depend on coating thickness and is proportional to the concentration gradient in the coating, assumed linear because of the lower thickness of coatings with respect to the concrete cover. Specific coating penetration resistance (Table 6) is defined as:
0.7 0.6 0.5 0.4 0.3
Uncoated Coating A Coating B Coating C Coating D
0.2 0.1 0.0
0
2
4
6
8
10
12
14
16
18
20
Corrosion rate (µm/y) Fig. 14. Cumulative frequency of corrosion rate for rebars embedded in uncoated concrete and concrete with coatings.
considered, at least for economic reasons. Fig. 14 shows the cumulative frequency curve of corrosion rate of steel reinforcements in the coated concrete specimens. Generally, corrosion rate can be considered negligible if below 1 lm/y, low between 2 and 5 lm/ y, moderate between 5 and 10 lm/y and very high for values above 100 lm/y [26]. Corrosion rates of steel reinforcements in the uncoated concrete are more dispersed and the 70% of data are greater than 10 lm/y, corresponding to a quite high corrosion rate. In the presence of the cementitious mortar (Coating A), this value decreases to about the 15%. The best behaviour is shown by Coating B and D: the 90% of corrosion rate values are lower than 7 and 9 lm/y, respectively. Conversely, corrosion rate data distribution in the presence of Coating C presents a great dispersion and the 40% of data are higher than 10 lm/y. Concrete coatings seem to be able to reduce corrosion rate of steel reinforcements since they the increase concrete electrical resistivity as a consequence of the reduction of the water content into concrete. Although corrosion rate is not negligible, steel reinforcements corrosion is in ohmic control and corrosion rate is limited to lower values with respect to the uncoated condition. The adoption of concrete coatings as a preventative corrosion method can extend corrosion propagation time by a reduction of corrosion rate. Nevertheless, it should be pointed out that the calculated corrosion rate is the mean value on the entire surface of the reinforcement and does not take into account the localized mechanism of chloride-induced corrosion. Pitting penetration rate can be higher than corrosion rate calculated from LPR measurements depending on the ratio between anodic and cathodic area [27,28].
5. Conclusions The efficiency of four commercial concrete coatings (a polymer modified cementitious mortar and three elastomeric coatings) against chloride-induced corrosion was studied by means of steel corrosion long-term monitoring and by chlorides penetration profiles in concrete. The polymer containing mortar shows the best effect on delay chlorides penetration in concrete by acting as a physical barrier in addition to concrete cover. Despite its lower polymer content with respect polymeric coatings, the higher thickness guarantees a longer time-to-corrosion. In the presence of aggressive environments or for long service life, the use of coating
744
A. Brenna et al. / Construction and Building Materials 48 (2013) 734–744
must be associated with a good quality concrete (low W/C ratio and an adequate concrete cover). Once corrosion has started, corrosion rate is lower in the presence of coatings, due to their ability to reduce water ingress in concrete. References [1] Pourbaix M. Atlas of electrochemical equilibria in aqueous solutions, NACE Cebelcor; 1974. [2] Bertolini L, Elsener B, Pedeferri P, Polder RB. Corrosion of steel in concrete. Prevention, diagnosis, repair. Weinheim: Wiley-VCH; 2004. [3] Elsener B. Corrosion inhibitors for reinforced concrete – an EFC state of the art report, in: Corrosion of reinforcement in concrete – mechanisms, monitoring, inhibitors and rehabilitation techniques. Published for the European Federation of Corrosion by Woodhead Publishing and Maney Publishing, No. 38; 2007. p. 170–84. [4] Ormellese M, Bolzoni F, Lazzari L, Pedeferri P. Effect of corrosion inhibitors on the initiation of chloride-induced corrosion on reinforced concrete structures. Mater Corros 2008;59(2):98–106. [5] Ormellese M, Bolzoni F, Lazzari L, Brenna A, Pedeferri MP. Organic substances as inhibitors for chloride-induced corrosion in reinforced concrete. Mater Corros 2011;62(2):170–7. [6] Swamy RN, Tanikawa S. An external surface coating to protect concrete and steel from aggressive environments. Mater Struct 1993;26:465–78. [7] de Vries J, Polder RB. Hydrophobic treatment of concrete. Constr Build Mater 1997;11(4):259–65. [8] Seneviratne AMG, Sergi G, Page CL. Performance characteristics of surface coatings applied to concrete for control of reinforcement corrosion. Constr Build Mater 2000;14:55–9. [9] Al-Zahrani MM, Al-Dulaijan SU, Ibrahim M, Saricimen H, Sharif FM. Effect of waterproofing coatings on steel reinforcement corrosion and physical properties of concrete. Cem Concr Compos 2002;24:127–37. [10] Al-Zahrani MM, Maslehuddin M, Al-Dulaijan SU, Ibrahim M. Mechanical properties and durability characteristics of polymer- and cement-based repair materials. Cem Concr Compos 2003;25:527–37. [11] Almusallam AA, Khan FM, Dulaijan SU, Al-Amoudi OSB. Effectiveness of surface coatings in improving concrete durability. Cem Concr Compos 2003;25:473–81. [12] Raupach M, Wolff L. Investigations on durability of surface protection systems by accelerated laboratory and outdoor exposure tests. In: Proceedings of the European corrosion congress eurocorr. Lisbon, Portugal; 2005.
[13] Medeiros MHF, Helene P. Surface treatment of reinforced concrete in marine environment: influence on chloride diffusion coefficient and capillary water absorption. Constr Build Mater 2009;23(3):1476–84. [14] Cen, EN 1062, Paints and varnishes – coating materials and coating systems for exterior masonry and, concrete;2002. [15] Cen, EN 1504–2, Products and systems for the protection and repair of concrete structures - Definitions, requirements, quality control and evaluation of conformity - Part 2: Surface protection systems for, concrete, 2004. [16] CEN EN 197-1, Cement – Part 1: composition, specifications and conformity criteria for common cements; 2011. [17] CEN EN 206-1, Concrete—Part 1: specification, performance, production and conformity; 2000. [18] CEN EN 10080, Steel for the reinforcement of concrete – Weldable reinforcing steel – general; 2005. [19] Cen, EN 1992-1-1, Eurocode 2: design of concrete structures – Part 1-1: general rules and rules for, buildings; 2004. [20] Andrade C, Gonzalez JA. Quantitative measurements of corrosion rate of reinforcing steels embedded in concrete using polarization resistance measurements. Mater Corros 1978;29(8):515–9. [21] ASTM G59, Standard test method for conducting potentiodynamic polarization resistance measurements; 2009. [22] UNI 9944, Corrosion and protection of reinforcing steel in concrete. Determination of the carbonation depth and of the chlorides penetration profile in concrete; 1992. [23] Elsener B, Andrade C, Gulikers J, Polder R, Raupach M. Half-cell potential measurements – potential mapping on reinforced concrete structures. Mater Struct 2003;36:461–71. [24] ASTM C876, Standard test method for corrosion potentials of uncoated reinforcing steel in concrete; 2009. [25] ASTM G102, Standard practice for calculation of corrosion rates and related information from electrochemical measurements; 2010. [26] Andrade C. In: Corrosion of steel in reinforced structures, Final Report, COST action 521, Luxembourg, Office for Official Publications of the European Communities; 2003. [27] Gonzalez JA, Andrade C, Alonso C, Feliu S. Comparison of rates of general corrosion and maximum pitting penetration on concrete embedded steel reinforcement. Cem Concr Res 1995;25:257–64. [28] Ormellese M, Berra M, Bolzoni F, Pastore T. Corrosion inhibitors for chlorides induced corrosion in reinforced concrete structures. Cem Concr Res 2006;36:536–47.