Calibration of the electrochemical methods for the corrosion rate measurement of steel in concrete. Nordtest project 1531-01.

Calibration of the Electrochemical

Methods for the Corrosion Rate

Measurement of Steel in Concrete

NORDTEST Project No. 1531-01

SP Swedish National Testing and Research Institute Building Technology

Calibration of the Electrochemical Methods for the

Corrosion Rate Measurement of Steel in Concrete

- Nordtest project No. 1531-01

Abstract

This report presents the results from the Nordtest project No. 1531-01. In this study, steel bars were embedded in concrete slabs with different chloride introductions (0%, 1.5%, 3% and 6 % by mass of cement) in order to induce corrosion at different rates. The slabs were cast in one laboratory and distributed to three different laboratories for monitoring of corrosion rate over a period of one year. Four instruments representing two types of electrochemical techniques, that is, linear polarisation technique (GECOR instruments) and galvanostatic pulse technique (GalvaPulse and SP’s instrument), were evaluated in the project through the comparative measurements. Finally, the standard gravimetric method ASTM G1 was used for calibration of the results measured by the non-destructive methods based on the electrochemical techniques.

The results show that the three different laboratories measured fairly comparable corrosion rates based on the gravimetric method, indicating that the method used in the project for production of specimens is basically suitable for use in the calibration of any type of non-destructive electrochemical techniques for measuring the corrosion rate of steel in concrete. The corrosion rates measured by GECOR are fairly close to the true mean corrosion rate (mass loss divided by the whole exposed area of steel), while the corrosion rates obtained by the 5 seconds short time galvanostatic pulse measurements from the chloride introduced specimens are close to the true actual corrosion rate (mass loss divided by the corroded area of steel). When multiplying the values measured by GECOR by a factor of 6 (as a pitting factor) for the chloride introduced specimens, the re-calculated corrosion rates become very comparable with the true actual corrosion rate. The condition of chloride content in concrete is necessary information for proper

judgement of passive or de-passive status of steel when using pulse techniques and proper use of pitting factor when using GECOR instruments.

Key words: concrete, corrosion, corrosion, electrochemical measurement, steel.

SP Sveriges Provnings- och Forskningsinstitut

SP Rapport 2002:25 ISBN 91-7848-916-4 ISSN 0284-5172 Borås 2002

SP Swedish National Testing and Research Institute

SP Report 2002:25

Postal address:

Box 857, SE-501 15 BORÅS Sweden

Telephone +46 33 16 50 00 Telex 36252 Testing S Telefax +46 33 13 55 02

Contents

Page Abstract ii

Preface iv

1 Introduction 1

2 Electrochemical techniques for non-destructive measurement of

corrosion rate in concrete 2

2.1 Expression of corrosion rate 2

2.2 Stern-Geary equation 2

2.3 Measurement techniques of polarisation resistance 3

3 Calibration 11

3.1 Concrete specimens 11

3.2 Electrochemical measurements 13

3.3 Destructive measurement of corrosion 13

4 Test results and discussions 15

4.1 Results from the first comparative measurement 15

4.2 Results from studies of effect of polarisation current and duration 16

4.3 Results from the second comparative measurements 21

4.4 Results from the third comparative measurement 24

4.5 Summary of the results from one year monitoring 27

4.6 Results from the destructive measurement 29

5 Comparison of the Results between Destructive and

Non-Destructive Methods 32

6 Concluding Remarks 35

7 References 37

Appendix 1 – Results from the first comparative measurement 39

Appendix 2 - Results from the second comparative measurement 40

Appendix 3 - Results from the third comparative measurement 41

Appendix 4 - Results from the tests of connected steel bars 42

Appendix 5 – Summarised results from one year monitoring 43

Appendix 6 - Results from the destructive measurements (Gravimetric

Preface

In the Nordic countries, there exist a limited number of portable instruments for measuring corrosion rate of steel in concrete. The engineers often arise a question how to interpret the measured results, because different instruments give different corrosion rates and the differences can sometimes be larger than one or two magnitude orders! It is difficult to answer the question because there is lack of information about the calibration of those

instruments based on non-destructive electrochemical techniques. This project is an attempt to the answer of the above question.

The author of this report would acknowledge Thomas Frølund (who currently works at Germann Instruments A/S in Denmark) and Henrik Sørensen at FORCE Technology, Denmark, Karin Pettersson (who currently works at Skanska Anläggning AB, Sweden) and Pål Skoglund at CBI, Swedish Cement and Concrete Institute, Per Fidjestøl and Anne-Marit Tonnesland at ELKEM Materials, Norway, for their participating in the project with valuable contributions. The author would extend special thanks to Carmen Andrade at Instituto de Ciencias de la Construcción Eduardo Torroja, Spain, for her constructive comments and suggestions.

The financial support from Nordtest is greatly appreciated.

Tang Luping

1 Introduction

Chloride-introduced corrosion of steel in concrete is a major problem with regard to the durability of reinforced concrete structures. Measurement of corrosion rate is an important measure for maintenance and repair planning of these structures. In the past decades a number of non-destructive methods have been developed for the measurement of corrosion rate especially for the field applications. Of these non-destructive methods, electrochemical methods such as linear polarisation technique and pulse technique seem more promising and have found relatively more applications worldwide. Both these techniques have also been used in the Nordic countries, for instance, the linear polarisation technique used in Sweden, Norway and Finland, and the galvanostatic pulse technique used in Denmark and Sweden. However, the measurement results from one method to another are often incomparable and the differences can sometimes be larger than one or two magnitude orders! With so large differences it is difficult to use the measurement results for practical maintenance and repair planning of concrete structures. In addition, there is lack of calibration method for these non-destructive techniques.

In order to minimise the confusion and to make use of the test results in the practical evaluation of concrete structures, Nordtest financed this project. The main objectives of the project are to calibrate different electrochemical (non-destructive and indirect) methods with a classic (destructive and direct) measurement method e.g. ASTM G1-90, and to find the practical measurement uncertainty of each method and the relations between these different methods, so as to be able to make use of the test results in the practical evaluation of concrete structures. Four institutions from three Nordic countries (see Table 1.1) participated in the project. This report presents the results from the project.

Table 1.1. Institutions participating in the project.

Institution Country/City Technique for corrosion

rate measurement

FORCE Technology Denmark/Copenhagen Galvanostatic pulse

technique SP Swedish National Testing

and Research Institute Sweden/Borås Galvanostatic technique pulse

CBI, Swedish Cement and Concrete Institute

Sweden/Stockholm Linear polarisation

2

Electrochemical techniques for non-destructive

measurement of corrosion rate in concrete

2.1

Expression of corrosion rate

Corrosion rate is often expressed in different terms: corrosion current, section loss or

corrosion depth, and sometimes, diameter reduction. Faraday’s law describes the relationship between section loss and corrosion current:

corr i zF M t x _⋅ ρ = ∆ ∆ (2.1) where x is the section loss of steel, t is the corrosion duration, M is the molecular weight of

metal (M = 56 g/mol for Fe), z is the number of ionic charges (z = 2 for Fe), F is the Faraday constant (F = 96480 C/mol or A⋅s/mol for Fe), ρ is the specific density of metal (ρ = 7.85 g/cm3 for Fe) and icorr is the density of corrosion current (often in µA/cm2). Thus 1 µA/cm2 of

corrosion current can easily be converted to 11.5 µm/yr of section loss.

2.2 Stern-Geary

equation

Stern & Geary (1957) first presented the relationship between polarisation resistance Rp (often

in kΩ) and corrosion current icorr, which is expressed as:

p corr

AR B

i = (2.2)

where B is a constant (often in mV) and A is the polarised area (often in cm2). Theoretically, B is determined by cathodic and anodic Tafel slopes βc and βa (the slopes of cathodic and

anodic polarisation curves in Tafel linear regions):

c a c a β + β β ⋅ β = B (2.3)

Practically, it is not an easy thing to determine the actual value of B. Conventionally, the value of B for steel-concrete systems is regarded within the range between 25 to 52 mV and B = 26 mV is usually assumed (Andrade & González, 1978). Song (2000) analysed B values on four different situations and concluded that a conservatively estimated range for B should be from 8 mV to infinite.

Nevertheless, if parameters B and A become or are assumed as known, the corrosion current icorr can be estimated from the measurement of polarisation resistance Rp.

2.3

Measurement techniques of polarisation resistance

Different techniques can be used for measuring polarisation resistance: linear polarisation technique, pulse technique, and electrochemical impedance spectroscopy (EIS). The linearity between potential drift η and current density i, as e.g. shown in Fig. 2.1, is a essential

requirement of all the above mentioned techniques.

Fig. 2.1 Polarisation curve for a reversible electrochemical system (Cox et at, 1997).

Linear polarisation technique

Linear polarisation technique directly makes use of Ohm’s law to the linear relationship between polarisation potential and current:

corr corr E E E E I I E I E R → → → ∆ = ∆ ∆ = p p p p p , 0 p p p (2.4)

where Ep and Ip are the polarisation potential and current, respectively, and Ecorr is the

corrosion potential, which is measured using the half-cell technique. When Ip is close to zero,

Rp will be in the linear range.

There are two ways to obtain ∆Ep/Ip:

• Potentiostatic way – applying a constant external potential ∆Ea and measuring the

• Galvanostatic way – applying a constant external current Ip and measuring the

potential response ∆Ea.

It should be pointed out that the applied or recorded ∆Ea is the total potential drop, which is a

sum of ∆Ep and ∆EΩ, the latter is called “ohmic drop” and attributed to the ohmic resistance

RΩ between the steel reinforcement and the counter electrode.

∆Ep = ∆Ea - ∆EΩ = ∆Ea - Ip·RΩ (2.5)

Therefore, it is important to know the actual value of RΩ in order to correctly quantify the

polarisation resistance Rp.

There are two types of equipment for the field measurement on concrete structures: 3LP (3-electrodes Linear Polarisation) developed in North America (Clear, 1989 and John et al, 1992) and GECOR developed in Spain (Feliu et al, 1990; and Rodríguez et al, 1995). Two GECOR instruments available in the Nordic countries were evaluated in this project. The measurement principles are shown in Figs. 2.2 and 2.3.

The GECOR instrument (version 6) measures the corrosion potential Ecorr by a reference

electrode (RE) of type Cu/CuSO4 placed at the centre of the disc before applying a

galvanostatic current. At the same time the instrument records the potential difference between a pair of small separated sensors (S1 and S2). Afterwards a trial current is applied to

the system and the potential response is analysed. According to the potential response the instrument applies a suitable galvanostatic current by the central counter electrode (CE) to the rebar with a target potential response of about 10 to 30 mV. Another current is applied by the external counter electrode or called “guard ring” to the rebar to keep the potential difference between the pair sensors S1 and S2 as at the original level as possible. A successful

maintenance of this potential difference implies a successful confinement of the polarised area on the steel. After the specified polarisation period (default value 100 seconds) the instrument records the potential response ∆Ea. The ohmic drop is measured at the end of the

trial during the switch off of the current and mathematically subtracted from the polarised potential. The duration for each measurement including initialisation and 100-seconds polarisation is about 3 to 5 minutes.

Fig. 2.2 Schematic of the electrodes arrangement on the GECOR 06 disc.

Fig. 2.3 Schematic of linear polarisation with guard electrode for confinement.

180 mm

Pulse technique

Pulse technique is, in fact, based on the same principle as linear polarisation technique. The main difference between these two techniques is that the linear polarisation technique measures responses of potential/current under a stationary state, while the pulse technique measures the responses under a non-stationary (transient) state, as shown in Fig. 2.4. In both these two techniques, the supplied current or potential must be small enough to assure a response in the linear polarisation range. Therefore, the pulse technique could be called as a non-stationary state linear polarisation.

Fig. 2.4 Illustration of response curves under a galvanostatic pulse (left) and a potentiostatic pulse (right).

When applying a constant current (galvanostatic way) or potential (potentiostatic way) to the steel-concrete system, the response of potential Ea or current Ia will change with time t in the

non-stationary state due to the capacitance behaviour of the system. Assuming that the system is like a Randles circuit, the potential response or current can be expressed as

        − + = ∆ + = _{Ω 1} − p dl p p p p 0 a C R t e R I R I E E E (2.6) or dl p p p p p C R t e K R R E I I I − Ω ∞ ₊ + ⋅ ∆ = ∆ + = (2.7)

where Cdl is the double layer capacitance and K is the amplitude of the transient response

process. Curve-fitting the measured values to the above equations one can obtain not only polarisation resistance Rp, but also other two informative parameters RΩ and Cdl.

Another way to evaluating the measured data is to transform Equation (2.6) to a linear form: I_p ∆Ep I_pRΩ t Pulse current Potential signal RΩ Rp Cdl ∆E_p I_∞ ∆I_p t Pulse potential Current signal I_p ∆Ep I_pRΩ t Pulse current Potential signal I_p ∆Ep I_pRΩ t Pulse current Potential signal RΩ Rp Cdl RΩ Rp Cdl ∆E_p I_∞ ∆I_p t Pulse potential Current signal

(

)

( )

dl p p p a max ln ln C R t R I E E − = − (2.8)

where Emax is the maximum potential reached after a long time (theoretically, t should be

larger than 5×RpCdl). Extrapolation of this straight line to t = 0, using least square linear

regression, yields an intercept corresponding to ln(IpRp) with a slope of 1/( RpCdl). Thus Rp

can be obtained from the intercept and Cdl can then be obtained from the slope (Elsener et al,

1997).

Two portable instruments based on galvanostatic pulse technique were reported in the literature (Elsener et al, 1997). One of them was developed at IBWK, ETH in Zurich,

Switzerland and another at FORCE Technology in Copenhagen, Denmark. The latter is called GalvaPulse and was evaluated in the project. Recently, SP also built up an instrument based on the principle of galvanostatic pulse technique. This instrument was also used in the evaluation.

The FORCE instrument GalvaPulse consists of a disc probe and a hand computer (Psion Workabout) attached with a galvanostatic generator, see Fig. 2.5. The instrument measures the corrosion potential Ecorr by a reference electrode of type Ag/AgCl placed at the centre of

the disc before applying a galvanostatic current to the counter electrode. Another current is applied to the guard ring to keep the potential difference between the counter electrode and guard ring close to zero. The instrument records the psignal responses of potential ∆Ea at a

sampling rate of 10 Hz in the first second and afterwards 5 Hz. The magnitude of

galvanostatic current can be pre-set between 5 and 450 µA and the measurement duration is normally 5 or 10 seconds, but can be up to 75 seconds. The potential-time curve can be roughly displayed on the screen. The first point of measured potentials is counted as an ohmic drop and Equation (2.8) is used for calculation of polarisation resistance.

Fig. 2.5. Schematic of FORCE’s instrument GalvaPulse.

100 mm 70 mm

SP’s instrument is illustrated in Fig. 2.6. Similar to GalvaPulse, the instrument measures the corrosion potential Ecorr by a reference electrode of type Ag/AgCl placed at the centre of the

electrode array. A galvanostatic current ICE is applied to the counter electrodes. Another

current IGE is applied to the guard electrodes. The ratio of current density on the guard electrodes

to that on the counter electrodes is defined as γ,

CE GE CE CE GE GE i i A I A I ₌ = γ (2.9)

where the subscripts GE and CE represent guard electrodes and counter electrodes, respectively. If not otherwise stated, γ = 1 was used in the experiment. A computerised 16 bits data acquisition system was employed to supply the currents ICE and IGEto the steel bar and at the same time collect the

psignal responses of potential ∆Ea. The measurement parameters, such as sampling rate,

magnitude of current and polarisation duration, can be freely pre-set by the user. The potential-time curve is directly displayed on the computer screen. Equation (2.6) is used for calculation of ohmic resistance RΩ, polarisation resistance Rp, and double layer capasitance

Cdl. An example of digitalised instrument panel is shown in Fig.2.7.

Fig. 2.6. Schematic of SP’s instrument. The effective polarised area of steel bar is assumed to be 33 cm2 (π×D×L = 3.14×1×10.5)

Steel Bar

GE CE CE GE

A/D potential input D/A current output

I₁ I₂ U₁ RE Concrete Slab Epoxy coating 10 cm 25 cm 15 cm 0.5 cm 0.5 cm ~2.5 cm ~2.5 cm Steel Bar GE CE CE GE

A/D potential input D/A current output

I₁ I₂ U₁ RE Concrete Slab Epoxy coating 10 cm 25 cm 15 cm 0.5 cm 0.5 cm ~2.5 cm ~2.5 cm

Fig. 2.7. Example of the digitalised panel of SP’s instrument. Electrochemical impedance spectroscopy (EIS)

EIS is one of the most powerful electrochemical techniques that can provide sufficient information on corrosion process of steel in concrete. The steel-concrete system, like other electrolytic systems, contains large capacitance. If a series of small AC potentials or currents with different frequencies are applied to such a system, the response at every frequency f will be sinusoidal signal with different amplitudes and a phase shift relative to the input signal. The ratio of ∆E/∆I = Z (impedance) is a sinusoidal function that can be decomposed in resistive term in phase with the input signal and in a capacitive term with a phase shift of 90º:

dl p p j 1 R C R R Z ϖ + + = _Ω (2.10)

where ϖ is the angular frequency, ϖ = 2πf, and j is the imaginary unit of − . At a very high 1 frequency, Z = RΩ, whereas at a very low frequency Z = RΩ + Rp. Thus the polarisation

resistance Rp could be estimated from the difference in Z obtained at a very low and a very

high frequency.

Again, the applied AC potentials or currents must be small enough to assure a response in the linear polarisation range.

The equipment for EIS measurement is very complicated and costly. So far only one

instrument was reported for the field measurement (Homma et al, 1992), but not available in the Nordic countries. Therefore, this technique was not included in the project for evaluation.

3 Calibration

3.1 Concrete

specimens

Concrete with w/c 0.5 and with different chloride introductions (0%, 1.5%, 3% and 6 % by mass of cement) was used in this study. The mixture proportions and physical properties of concrete are listed in Table 3.1.

Table 3.1. Mixture proportions and physical properties of concrete.

Test series Mix 0 Mix 15 Mix 30 Mix 60

Chloride introduction* 0% Cl 1.5% Cl 3.0% Cl 6.1% Cl

Cement type Swedish SRPC (corresp. to CEM I 42.5R)

Cement content, kg/m3 375 385 362 353 Water-cement ratio 0.49 0.48 0.48 0.48 Aggregate, 0∼8 mm, kg/m3 939 934 958 972 Aggregate, 8∼16 mm, kg/m3 867 862 884 897 Water reducer: Type

Dose, wt% of cement

None None None None AEA: Type

Dose, wt% of cement

None None None None

Air content, vol% 0.7 1.5 1.6 1.3

Slump, mm 85 90 95 100

Strength** at 28 d, MPa 56.5 ± 0.5 63.0 ± 1.0 58.7 ± 1.3 50.4 ± 1.0 * in the form of NaCl salt and calculated in Cl% by cement mass;

** according to Swedish standard SS 13 72 10.

Plain cool-drawn carbon steel of diameter 10 mm was used as reinforcement in concrete. The steel bars were cleaned with degreasing agent followed with acetone. The ends of each steel bar were coated with cement grout followed with epoxy to avoid unexpected crevice corrosion. A picture of the steel bars after the treatment is shown in Fig. 3.1.

All the concrete specimens were produced at SP in Sweden. Each concrete was mixed in one batch by using a 120 litres paddle mixer. Concrete slabs of size 250 × 250 × 70 mm were cast in poly-wood moulds. Two steel bars were in parallel embedded in the centre portion of each slab at the mid-height (about 35 mm), with a space of 100 mm between each other, as shown in Fig. 3.2. Three concrete cubes of size 150 mm per each mixture were cast for determination of compressive strength. The moulds with the fresh concrete were covered with thick plastic films to prevent evaporation from the concrete surface. One day after casting the specimens

were demoulded. The side-surfaces of each slab were coated with epoxy and then each two slabs from the same type of concrete were placed in a plastic box with the bottom surfaces upward to facilitate the electrochemical measurements. The bottom of each box was filled with water to produce a moist condition (see Fig. 3.3). The boxes were stored at the room temperature in the laboratory. From the age 21 days on, the water in the boxes containing the slabs of Mixes15, 30 and 60 was replaced with saturated KCl solution to produce a condition of about 85%RH, while the boxes containing the slabs of Mix 0 were always kept water filled to prevent possible contamination from use of salt. The concrete cubes were cured and later (at 28 days) tested for compressive strength in accordance with SS 13 72 10.

Fig. 3.1 Steel bars after the treatment with cement grout and epoxy on the ends. 250 mm 70 mm 250 mm Ø10 mm steel 100 mm

Fig. 3.3 Illustration of curing conditions for the concrete slabs.

3.2 Electrochemical

measurements

SP carried out the measurements using SP’s instrument on all the slabs from the age one day and the measurements using ELKEM’s GECOR6 instrument on some of the slabs from the age 14 days.

In this project three comparative measurements have been carried out. The first comparative measurement was arranged at the age 4 weeks. FORCE’s personnel came to SP with their GalvaPulse instrument. CBI did not participate in the first comparative measurement owing to the delayed reparation of their instrument.

The second comparative measurement was arranged at the age 3 months for CBI’s personnel coming to SP with their GECOR6 instrument. Afterwards eight slabs (two per mixture) were transported to FORCE and other eight slabs were transported to CBI for continuously

measurements at the individual laboratories. To facilitate transportation, the liquid in each plastic box was taken away and the spaces between the slabs and the box were filled with cellular plastic. After arrival at the respective laboratory, the boxes containing the slabs of Mixes 15, 30 and 60were re-filled with saturated KCl solution and the box containing the slabs of Mix 0 was re-filled with water. The measurements in the respective laboratories were continued over the period of one year.

The third comparative measurement was arranged at the age 8 months and both FORCE’s and CBI’s personnel came to SP with their own instruments.

3.3

Destructive measurement of corrosion

At the end of the electrochemical measurements, the steel bars were removed from each concrete slab. To facilitate the broken of concrete, a groove of about 2 cm deep, parallel to the steel bar was cut on the cast-as surface of concrete, as shown in Fig. 3.4. After removal of the steel bars, the corrosion condition of each bar was visually inspected and photographed. The exposed length (about 195 mm) of each steel bar was machine-cut on a turner or alike. The steel bars were stored under dry conditions, e.g. in a desiccator with blue gel (or yellow gel), to prevent from further development of corrosion.

To determine the mass loss, the steel bars were placed on a pair of triangle strips as shown in Fig.3.5 and subjected to the acid bath in accordance with ASTM G 1, Designation C.3.1. After a number of (5~6 times) repeated acid treatments, the weight of each steel bar was recorded to 0.001 g and the length was measured to 0.01 mm. The mean diameter of steel bar

Plastic box with cover Concrete slab

Water or saturated KCl Bottom (test) surface upward

after the acid treatment was determined from the non-corroded bars. The mass loss of each steel bar was then estimated by comparing its actual weight with its theoretical weight (density×πr2_×length).

Fig. 3.4 Cutting a groove facilitates the removal of steel bar.

Fig. 3.5 Teflon triangle strips for supporting the steel bars under the acid bath.

Bottom (test) surface

15 mm

4

Test results and discussions

4.1

Results from the first comparative measurement

At the age of 4 weeks, three instruments, FORCE’s GalvaPulse, ELKEM’s GECOR6 and SP’s instrument, were used in the comparative measurement. The measured results are given in Appendix 1 and summarised in Figs. 4.1 to 4.3.

Fig. 4.1 Corrosion currents from the first comparative measurement.

Fig. 4.2 Corrosion potentials from the first comparative measurement.

0.01 0.1 1 10 0% Cl 1.5% Cl 3% Cl 6% Cl Ad d e d ch lo rid e [mass % o f ce me n t] C o rro s io n ra te Ico rr [ µ A/ c m 2 ] _{GalvaPuls e} 100uA/5s GalvaPus e 5 uA/10s SP 100uA/23s Gecor6 5-30uA/100s -700 -600 -500 -400 -300 -200 -100 0 0% Cl 1.5% Cl 3% Cl 6% Cl

Adde d chloride [mass % of ce me nt]

C o rro s ion po te nt ia l ECS E [m V ] GalvalPuls e SP GEC OR 6

Fig. 4.3 Ohmic resistance from the first comparative measurement.

It appears that the values of corrosion currents measured by GalvaPulse (5 sec. measurement duration) are significantly higher than those measured by GECOR6 (100 sec. Measurement duration), especially for the specimens with 0%Cl, in which the steel bars should be in a passive status. While the values measured by SP’s instrument (23 sec. Measurement duration) lie between GalvaPulse and GECOR6. It seems that the measurement duration plays an important role. Therefore, SP carried out more studies of the effects of polarisation current and duration, which will be presented later.

It is natural that the results of corrosion potential from different instruments are very close, because all of the instruments employ a standard reference electrode for potential

measurement.

The ohmic resistance measured by GalvaPulse are the highest one probably due to its small size of sensor.

4.2

Results from studies of effect of polarisation current and

duration

After the first comparative measurement, it has been found that there exists large difference in corrosion rate measured by different instruments. To find the possible reasons, SP carried out some investigations using different measurement parameters. The measured response curves are shown in Figs. 4.4and 4.5.

0 0.2 0.4 0.6 0.8 1 0% Cl 1.5% Cl 3% Cl 6% Cl

Adde d chloride [mass % of ce me nt]

O h m ic r e si stan ce R Ω [k Ω ]

G

alvaPulse

SP

G

EC

O

R

6

(a) (b)

Fig. 4.4 Potential response from concrete without chlorides. (a) full data; (b) response in the first second.

(a) (b)

Fig. 4.5 Potential response from concrete with 3% chloride by mass of cement. (a) full data; (b) response in the first second.

As expected, the potential responses from the concrete without chlorides are significantly higher than that from the concrete with added chlorides. The potential response increases with the intensity of an imposed galvanostatic current, but the increment is non-linear, as will be discussed later. When a potential difference is less than 100 mV, e.g. the curves of the polarisation current less than 10 µA in Fig. 4.4 and all the curves in Fig. 4.5, the stationary polarisation could not be achieved even after a polarisation duration of 300 seconds. This is in agreement with the findings reported by Videm & Myrdal (1997). When the imposed galvanostatic current is larger than 20 µA for passive steel, the potential response looks close to a stationary state after more than 100~200 seconds polarisation (see Fig. 4.4). However, the potential difference in all of these curves is over 200 mV, which is far beyond of any of the conditions of linearity. Therefore, the stationary state of these curves could be some artefacts. 0 0.02 0.04 0.06 0.08 0.1 0 0.5 1

Polaris ation duration t, s e c

R esp o n se p o te n ti a l Ea , V 0 0.1 0.2 0.3 0.4 0.5 0 50 100 150 200 250 300 350

Polaris ation duration t, s e c

R esp o n se p o te n ti a l Ea , V 10 µA 20 µA 50 µA 100 µA 2 µA 5 µA Concrete without Cl 0 0.02 0.04 0.06 0.08 0.1 0 0.5 1

Polaris ation duration t, s e c

R e s p o n s e p o te n tia l Ea , V 0 0.025 0.05 0.075 0.1 0.125 0.15 0 50 100 150 200 250 300 350

Polaris ation duration t, s e c

R e s p o n s e p o te n tia l Ea , V 10 µA 20 µA 50 µA 100 µA Concrete with 3% Cl

The initial potential response (at t = 0) reflects the ohmic resistance of concrete. It can be seen from Figs. 4.4 (b) and 4.5 (b) that, at the same current level, the initial potential response from the concrete with 3% Cl is not significantly lower than that from the concrete without chlorides, implying that the resistivity of concrete is not a decisive parameter to corrosion of steel. This should be true, because the resistivity of concrete is strongly dependent on the content of moisture and the concentration of ions, especially hydroxides, in the pore solution. In this study, the concrete without chlorides were cured under a moist condition (>95%RH), while the concrete with added chlorides under a condition of about 85%RH (saturated KCl), the former has a higher moisture content than the latter. Therefore, addition of chloride or corrosion of steel in concrete may not necessarily mean a significant increase in conductivity, or decrease in resistivity, of concrete.

Since a potential shift of 60 mV from the corrosion potential may be the maximal limitation to a condition of linearity, only those data of potential shift less than 60 mV should be taken as valid data. By curve-fitting equation (2.6) to these valid data from different polarisation durations, different values of RΩ, Rp and Cdl can be obtained, as shown in Fig. 4.6. It is noticed that each fitted curve is in a

good agreement with the corresponded data, implying that the polarisation behaviour of a steel-concrete system only “time-dependently” obeys the Randles circuit, and a better model is needed to describe this time-dependent behaviour. The increase in the curve-fitted RΩ, although not so

significant when compared with those in Rp and Cdl, is also due to the increased potential responses,

which statistically reduced the weight of the initial points of a ∆Ea-t curve in the curve-fit.

Nevertheless, in this study equation (2.6) was applied to the valid data of different durations of each ∆Ea-t curve to obtain the relationships between polarisation resistance and polarisation duration at

different intensities of polarisation current, as shown in Fig. 4.7. It is evident that the polarisation duration has remarkable effect on the curve-fitted polarisation resistance, while the effect of

polarisation current seems not significant when compared with the effect of polarisation duration. The fluctuations in the curves of 2 and 5 µA are probably due to the measurement uncertainty when the potential response was low.

Fig. 4.6 An example of the curve-fitted results using the data from different durations of polarisation. 0 0.02 0.04 0.06 0.08 0.1 0 20 40 60 80 100

Polarisation duration t , sec

R es pons e pot ent ial E a , V t = 30 s RΩ = 0.791 kΩ, Rp = 9.22 kΩ, Cdl =4150 µF t = 10 s: RΩ = 0.723 kΩ, Rp = 4.61 kΩ, Cdl =3430 µF t = 3 s: RΩ = 0.688 kΩ, Rp = 1.82 kΩ, Cdl =2740 µF t = 1 s: RΩ = 0.661 kΩ, Rp = 0.565 kΩ, Cdl =1850 µF Concrete without Cl, ICE = 10 µA t = 5 s: RΩ = 0.701 kΩ, Rp = 2.82 kΩ, Cdl =3050 µF

Fig. 4.7 Polarisation resistances calculated by curve-fitting equation (2.6) to different durations of each ∆Ea-t curve.

It can be seen from Fig. 4.7 that the effect of polarisation current appears not significant if the potential shift is small enough to assure a linear polarisation. While polarisation duration has tremendous effect on polarisation resistance, no matter if the steel is corroded or not. This phenomenon has also been observed by other researchers (e.g. González et al 1985; Gowers et al 1994; Videm & Myrdal 1997). The effect of sweep rate on polarisation resistance in the dynamic polarisation measurement is in fact the similar phenomenon. A high sweep rate for a certain level of response implies the short polarisation duration, often resulting in a low polarisation resistance (see e.g. González et al 1985; Gowers et al 1994). Figure 4.7 also shows that the relationships between Rp

and t are logarithmically linear, especially when the polarisation duration is in the first few seconds. Therefore, these relationships could be empirically expressed as

b at

R_p = (4.1)

where a and b are constants.

At the present the constants a and b are, however, purely empiric without any physical meanings. It should be born in mind that the above relationship is derived based on the simple Randles model and it may not necessarily mean a zero resistance at t = 0 and an infinite resistance at t → ∞. Gowers et al (1994) also reported the similar relationships between total resistance (Rp + RΩ) and sweep rate. For

passive steel, when sweep rate changed from 0.1 to 1000 mV/min, the total resistance changed from some 10000 to 10 kΩ⋅cm2_{! When using the interrupt (“turn-off” and “tunr-off”) technique we}

observed no significant changes in ohmic drop, as shown in Fig. 4.8, implying that the ohmic resistance is almost constant during the polarisation. Therefore, the change in total resistance should be mainly attributed to the change in polarisation resistance. At the present, the exact mechanisms behind this phenomenon are, however, not clear. Further modelling work is needed to find the convincing explanations to this phenomenon.

0.01 0.1 1 10 100 0.1 1 10 100 1000

Po lar is atio n d u r atio n t , s e c

P o la ri sa ti o n r esi st a n c e R p , k 10 µA 20 µA 50 µA 100 µA 2 µA 5 µA Concrete with 3% Cl Concrete without Cl

Fig. 4.8 Potential responses from “turn-on” and “turn-off” experiment.

The curve-fitted Cdl also reveals a similar behaviour, that is, the capacitance increases with

polarisation duration, as shown in Fig. 4.9. The concrete with 3% Cl gives a higher

capacitance than that without chlorides, probably due to its relatively high ionic concentration in the pore solution.

Fig. 4.9 Double layer capacitances calculated by curve-fitting equation (2.6) to different durations of each ∆Ea-t curve.

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0 100 200 300 400 Po larisatio n d u ratio n , se c P o te n tia l r e s p o n s e , V 0%Cl, 5 uA 0%Cl, 100 uA 1.5%Cl, 100 uA 3%Cl, 100 uA 6%Cl, 100 uA 100 1000 10000 100000 0.1 1 10 100 1000

Po lar is atio n d u r atio n t , s e c

C a p aci ta n c e C dl , F 10 µA 20 µA 50 µA 100 µA 2 µA 5 µA Concrete w ith 3% Cl Concrete w ithout Cl

Nevertheless, from the curve-fitted data in Fig. 4.7 (except for those with fluctuation due to low potential responses), we can calculate the corrosion current density using Equation (2.2). The calculated results are shown in Fig. 4.10, where the results measured from the same specimens at the first comparative measurement are also given. The circled two marks were those data measured using 100 µA polarisation current. The deviation of these two points from the others is probably due to their high potential shift, which may exceed the limit of linear polarisation. In general, it can be concluded from Fig. 4.10 that the differences in corrosion rate measured by different instruments are mainly attributed to different polarisation durations. Therefore, SP re-evaluated the measured results using the first 1, 2, 3, 4, and 5 seconds data for curve-fitting Equation (2.6) and then utilising the logarithmical relationship as shown in Equation (4.1) to extrapolate polarisation resistance to 100 seconds in order to compare the results with both GalvaPulse and GECOR6.

Fig. 4.10 Relationships between corrosion rates measured by different instruments and polarisation durations.

4.3

Results from the second comparative measurements

The second comparative measurement was arranged at the age of 3 months (13 weeks). It is an opportunity to test two similar instruments (ELKEM’s and CBI’s GECOR6) on 64

separate steel bars (16 steel bars per each type of concrete). SP’s instrument was also used in the comparative measurement. The experiment was arranged in such a way that the interval between two measurement on the same specimen was at least three times as long as the polarisation duration so as to leave sufficient time for depolarisation. The measured results are given in Appendix 2. The comparative results between two similar instruments are shown in Figs.4.11 to 4.13 and a summary of the average corrosion rates measured by different instruments is presented in Fig. 4.14. As expected, the half-cell potential measured from two GECOR6 instruments is very comparable. The results of corrosion rate show some deviation, but most of the values are in the range of a factor of 2. The results of ohmic resistance show significant deviation, probably due to the fact that the concrete was relatively young and wet in this study, which resulted in such a low ohmic resistance that the instrument could not accurately measures it.

0.01 0.1 1 10 100 0.1 1 10 100 1000

P ola risa tion dura tion t , se c

C o rr os ion c u rr e n t de si ty ico rr , A/ c m ² GalvaPulse GECOR SP 10 µA 20 µA 50 µA 100 µA 2 µA 5 µA Concrete with 3% Cl Concrete without Cl

* The circled value might be an outlier and not included in averaging.

Fig. 4.11 Corrosion rates [in µA/cm²] measured by two GECOR6 instruments.

Fig. 4.12 Corrosion potentials [in mV CSE] measured by two GECOR6 instruments.

-800 -600 -400 -200 0 -800 -600 -400 -200 0

EL KEM 's GEC OR6

CB I' s G E CO R 6 0% Cl 1.5% Cl 3% Cl 6% Cl 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10

EL KEM 's GEC OR6

CB I' s G E CO R 6 0% Cl 1.5% Cl 3% Cl 6% Cl

Fig. 4.13 Ohmic resistances [in kΩ] measured by two GECOR6 instruments.

Fig. 4.14 Comparison of average corrosion rates measured by different instruments from the second comparative measurement.

From Fig. 4.14 it can be seen that, when extrapolated the results from 5 seconds polarisation to 100 seconds polarisation, the corrosion rates measured by SP’s instrument are better comparable with GECOR6 instruments.

0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8

EL KEM 's GEC OR6

CB I' s G E CO R 6 0% Cl 1.5% Cl 3% Cl 6% Cl 0.01 0.1 1 10 0% Cl 1.5% Cl 3% Cl 6% Cl Adde d chloride [m a ss % of ce m e nt] C o rr os ion r a te ico rr [ A/ cm 2 ] ELKEM's GEC OR 6 C BI's GEC OR 6 SP-ex100s SP-5s

4.4

Results from the third comparative measurement

The third comparative measurement was arranged at the age of 8.5 months (36 weeks). It is an opportunity to test all three instruments (FORCE’s GalvaPulse and CBI’s GECOR6 and SP’s instrument) on 16 disconnected steel bars (4 steel bars per each type of concrete). The test results are given in Appendix 3 and summarised in Figs. 4.15 to 4.17. After having tested the disconnected bars, the two bars in one slab from each type of concrete were connected and tested again after waiting a sufficient time for electrochemical equilibrium in the system. The test results are given in Appendix 4 and the ratios of corrosion rates measured on the

connected to those on the disconnected bars are shown in Fig. 4.18.

It can be seen from Fig. 4.15 that the results measured by the two instrument based on

galvanostatic pulse technique are in fairly good agreement, implying a good reproducibility of this technique. Again, when extrapolated the results from 5 seconds polarisation to 100

seconds polarisation, the corrosion rates measured by SP’s instrument are in general comparable with those measured by GECOR6, as shown in Fig. 4.16.

Figure 4.17 shows that the galvanostatic pulse technique (5 seconds short time measurement) always produces a higher corrosion rate. This high corrosion rate seems incredible for those passive steel bars embedded in the concrete with 0%Cl.

It is surprising that all the instruments measured higher corrosion rates when two steel bars were connected, as shown in Fig. 4.18. It should be noticed that the distance between two bars is 100 mm, while the instrument with the largest electrodes disc is GECOR6, whose guard ring’s radius is 90 mm and confined radius is 53 mm, both are less than the distance 100 mm. Therefore, only one steel bar could be under the covered area of the electrodes. The increased corrosion rate measured on the connected bars may imply that part of the imposed current has flowed through the steel bar which was not under the covered area of the electrodes, due to the low ohmic resistance of concrete used in this study. It also implies that, when the resistance of concrete is low, the confinement of current is questionable.

Fig. 4.15 Comparison of corrosion rates [in µA/cm²] measured by two instruments based on galvanostatic pulse technique.

Fig. 4.16 Comparison of corrosion rates [in µA/cm²] measured by GECOR6 and SP’s extrapolation technique. 0.01 0.1 1 10 0.01 0.1 1 10

C BI's GEC OR6

S P e x 100 s e c 0% Cl 1.5% Cl 3% Cl 6% Cl 0.1 1 10 100 0.1 1 10 100 GalvaPu ls e SP 5 s e c 0% Cl 1.5% Cl 3% Cl 6% Cl

Fig. 4.17 Comparison of corrosion rates measured by different instruments from the third comparative measurement.

Fig. 4.18 Ratios of corrosion rates measured on connected bars to those on disconnected bars. 0.01 0.1 1 10 0% Cl 1.5% Cl 3% Cl 6% Cl Adde d chloride [m a ss % of ce m e nt] C o rro si o n ra te ico rr [ A/ c m 2 ] GalvaPuls e SP-5s GEC OR 6 SP-ex100s 0 2 4 6 0.01 0.1 1 10

i_{co r r} [µA/cm ²] m e a sure d on disconne cte d ba r

R a ti o of ico rr m eas u red o n c onn e c te d t o di sc o nne c te d ba rs GalvaPulse SP-5s GECOR6 SP-ex100s

4.5

Summary of the results from one year monitoring

The summarised results measured at SP during the period of about one year are shown in Figs.4.19 to 4.21. Other summarised results can be found in Appendix 5. It seems from Fig. 4.19 that the corrosion rate in the concrete with 6%Cl has a tendency to increase with time, while the corrosion rate in the concrete with 1.5%Cl or 3%Cl has a tendency to decrease with time. This is in agreement with the development of corrosion potential (see Fig. 4.20). By integrating the monitored curve of corrosion rate and then dividing the integrated value by the whole period of test the mean corrosion rate of each steel bar could be calculated. The

calculated values of mean corrosion rate are included in Appendix 6 and will be discussed in Chapter 5.

Fig. 4.19 Development of corrosion rate measured at SP during the period of about one year.

Corros ion Ra te by SP's Ins trum e nt (e x tra pola te d to 1 0 0 s e c )

0.01 0.1 1 10 0 100 200 300 400 500 C o n cr e te A g e , d ays ico rr , A/ c m ² M ix 0% Cl M ix 1.5% Cl M ix 3% Cl M ix 6% Cl

Corrosion Ra te by ELKEM 's GECOR6

0.01 0.1 1 10 0 100 200 300 400 500 C o n cr e te A g e , d ays ico rr , A/ c m ² _{M ix 0% Cl} M ix 1.5% Cl M ix 3% Cl M ix 6% Cl

Fig. 4.20 Development of corrosion potential measured at SP during the period of about one year. The ohmic resistances in all the types of concrete tested increased with time, probably due to the development of hydration that improved the pore structures of concrete. It seems difficult to judge the corrosion status from the ohmic resistance of concrete, because the difference in ohmic resistance between different types of concrete is not large enough to differ the passive and corroded steel bars.

Corros ion Pote ntia l by SP's Ins trum e nt

-800 -600 -400 -200 0 0 100 200 300 400 500 C o n cr e te A g e , d ays Eco rr , m V [ C S E ] Mix 0% Cl Mix 1.5% Cl Mix 3% Cl Mix 6% Cl

Corros ion Pote ntia l by ELKEM 's GECOR6

-800 -600 -400 -200 0 0 100 200 300 400 500 C o n cr e te A g e , d ays Eco rr , m V [ C S E ] Mix 0% Cl Mix 1.5% Cl Mix 3% Cl Mix 6% Cl

Fig. 4.21 Development of ohmic resistance measured at SP during the period of about one year.

4.6

Results from the destructive measurement

After the removal of steel bars from the concrete, it was found that the corrosion occurred on the side toward the bottom of concrete (see the photo picture in Fig. 3.4, and more pictures are available in Appendix 6), probably due to the defects of concrete underneath the steel bar caused from segregation. In most cases, only about a half of the surface of a steel bar in the concrete with 1.5 and 3%Cl was corroded, and less than 30% surface of a steel bar in the concrete with 6%Cl was corroded. This implies that the risk of localised corrosion increases with the chloride content in concrete. Therefore, the corrosion rate from the mass loss results was calculated in two ways: 1) mean corrosion rate (mass loss divided by the whole exposed area of steel), and 2) actual corrosion rate (mass loss divided by the corroded area of steel). The measured results are given in Appendix 6 and summarised in Fig. 4.22. The diameters of

Ohm ic Re s is ta nc e by SP's Ins trum e nt

0 0.2 0.4 0.6 0.8 1 1.2 0 100 200 300 400 500 C o n cr e te A g e , d ays RΩ , k M ix 0% Cl M ix 1.5% Cl M ix 3% Cl M ix 6% Cl

Ohm ic Re s is ta nc e by ELKEM 's GECOR6

0 0.2 0.4 0.6 0 100 200 300 400 500 C o n cr e te A g e , d ays RΩ , k M ix 0% Cl M ix 1.5% Cl M ix 3% Cl M ix 6% Cl

non-corroded bars measured by the three laboratories after the acid treatments are

summarised in Table 4.1. All the three laboratories reported a very small standard deviation (< 0.001 mm) in the diameter of steel bar calculated from the weight and length measurement, indicating the method used in the project is adequately accurate for measuring the mass loss of steel.

Table 4.1 Diameter of steel bar calculated from the weight and length measurement.

Laboratory SP CBI FORCE

Bar-1 9.970 9.968 Excluded* Bar-2 9.970 9.969 9.969 Bar-3 9.971 9.968 9.968 Bar-4 9.970 Excluded* 9.968 Mean 9.970 9.968 9.968 Std dev 0.0005 0.0006 0.0006

* Due to corrosion for unclear reasons.

It can be seen from Fig. 4.22 that the results of mass loss reported from three different laboratories are fairly comparable. All three laboratories observed the significant localised corrosion and small mass loss of steel in the concrete with 6%Cl. When using the whole exposed area in the calculation, the calculated corrosion rate of steel in the concrete with 6%Cl is lower than that in the concrete with 3% Cl, while the actual corrosion depth of steel in the concrete with 6%Cl is the deepest among all the types of concrete. The small

percentage of the corroded area of steel in the concrete with 6%Cl is probably due to the fact that the high chloride concentration in the concrete induced pitting corrosion at a relatively high rate, resulting in a large cathodic area that protected the steel from further corrosion. On the other hand, all the concrete slabs with introduced chloride were stored under the similar humid conditions (about 85%RH). Since the moisture saturation pressure in the capillary pores containing high chloride content is relatively low, the 85%RH may cause condensation in the capillary pores of the concrete with 6%Cl. Each laboratory in this project has in fact observed this condensation. Every time when the cover of the plastic box was opened, the surfaces of the concrete slabs with 6%Cl were always wet when compared with the other slabs. The saturated capillary pores may reduce the oxygen transport. The starvation of

oxygen might be one of the reasons for the small mass loss of steel in the concrete with 6%Cl. The rust of steel in the concrete with 6%Cl was often in dark blue colour, as could be

observed from the photos in Appendix 6. This is an evidence of corrosion under the condition of oxygen starvation.

Fig. 4.22 Summary of corrosion area and corrosion rate calculated from the mass loss results. 0 10 20 30 40 50 0% Cl 1.5% Cl 3% Cl 6% Cl A d d e d ch lo r id e [m as s % o f ce m e n t] % o f co rr o d e d ar e a SP FORCE CBI 0 0.2 0.4 0.6 0.8 1 0% Cl 1.5% Cl 3% Cl 6% Cl A d d e d ch lo r id e [m as s % o f ce m e n t] C o rr os io n r a te [ A /c m ²] fr o m m a s s l o s s /w h o le a re a SP FORCE CBI 0 1 2 3 4 0% Cl 1.5% Cl 3% Cl 6% Cl A d d e d ch lo r id e [m as s % o f ce m e n t] C o rr os io n r a te [ A /c m ²] fr om m a s s l o s s /c or ro de d a re a SP FORCE CBI

5

Comparison of the Results between Destructive

and Non-Destructive Methods

Since the corrosion rate measured from all the non-destructive methods studied in this project is based on the whole confined area of the electrodes disc or unit, we should first compare the mean corrosion rate, as shown in Fig. 5.1, where two lines besides the equality line indicate the upper and lower limit when considering a tolerance factor of 2, which is commonly used in the field of corrosion measurement using electrochemical techniques.

It can be seen from Fig. 5.1 that the values measured from GECOR6 are in general close to the true values of the mean corrosion rate, especially for the steel bars with localised corrosion (6%Cl). According to GECOR’s criteria, when corrosion rate is less than 0.1 µA/cm², the steel is in passive status. The results measured from the steel bars in the concrete with 0%Cl are in good agreement with the observations. It appears a little underestimation for the steel bars in the concrete with 1.5%Cl.

It seems that the direct 5 seconds galvanostatic pulse measurements overestimate the mean corrosion rate, especially for the steel bars in passive status (0%Cl) and with localised corrosion (6%Cl). When extrapolating the 5 seconds measurement to 100 seconds, the

measured values become more comparable with the true values, especially for the steel bars in passive status (0%Cl) and with moderate mean corrosion rate (3%Cl), but it still appears an overestimation for the steel bars with localised corrosion (6%Cl). If the overestimation from the galvanostatic pulse measurement is due to the improper confinement of the polarised area, the deviation should be less than a factor of 2 to 3, because in this study the total exposed area of a steel bar is about 60 cm², while the confinement of SP’s instrument and GalvaPulse is about 30 cm² and 20 cm², respectively. Therefore, there must be some other reasons causing the overestimation. When compared with the actual corrosion rate (mass loss divided by the corroded area), the values measured by the direct 5 seconds galvanostatic pulse measurements become much closer to the true values for all the corroded steel bars, as shown in Fig. 5.2. This implies that the direct 5 seconds galvanostatic pulse measurements may in fact measure the actual pitting corrosion. However, because the technique was supposed to measure the mean corrosion rate, it is not clear if this is a mere coincidence for the chosen set-up or if the mechanisms behind the technique still are not well understood. In the electrochemical measurements, the imposed galvanostatic current is assumed homogeneously distributed through the steel surface. This is not true for the locally corroded steel. The corroded part has much lower resistance than the non-corroded part. The current will mainly flow through the corroded part, resulting in a much lower signal response (potential shift) than it would be if the current were homogeneously distributed. The heterogeneity of current distribution in the locally corroded steel might be one of the reasons for the coincidence shown in Fig. 5.2. Further studies are, however, needed to clarify this coincidence. A challenged question for the galvanostatic pulse technique is how to effectively differ the passive and de-passive status of a steel bar. Extrapolation technique as used by SP’s instrument might be one of the ways, but this way may also underestimate the actual corrosion rate of a corroded steel bar, e.g. as shown in Fig. 5.1 for the steel bars in the concrete with 1.5%Cl. Probably a combination of the extrapolation technique with the condition of chloride content in concrete may be helpful for a proper judgement of the corrosion status of reinforcement steel.

Fig. 5.1 Comparison between corrosion rates [in µA/cm²] measured by non-destructive and destructive methods.

In practice to know the actual corrosion rate of steel is very important for the reinforced concrete structures because the broken of any portion of steel may imply the failure of the whole structure. When using GECOR instruments, it has been suggested to multiply the measured value by a pitting factor of 4∼8 to obtain the actual localised corrosion rate (González & Andrade et al, 1995). Multiplying the values measured by GECOR

0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 Gr avim e tr ic M e th o d (m as s lo s s /w h o le ar e a) GE C O R 6 ELKEM, 0% C l C BI, 0% C l ELKEM, 1.5% C l C BI, 1.5% C l ELKEM, 3% C l C BI, 3% C l ELKEM, 6% C l C BI, 6% C l 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 Gr avim e tr ic M e th o d (m as s lo s s /w h o le ar e a) G a lvaP u ls e FOR C E, 0% C l FOR C E, 1.5% C l FOR C E, 3% C l FOR C E, 6% C l 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 Gr avim e tr ic M e th o d (m as s lo s s /w h o le ar e a) SP 's i n s tr u m e n t SP-5s , 0% C l SP-100s , 0% C l SP-5s , 1.5% C l SP-100s , 1.5% C l SP-5s , 3% C l SP-100s , 3% C l SP-5s , 6% C l SP-100s , 6% C l

instruments used in this study by a factor of 6 for all the specimens introduced with chlorides and comparing them with the actual corrosion rate, as shown in Fig. 5.3, we can find that the agreement between the re-calculated values (×6) and the true values is even better than that shown in Fig. 5.1. This means that the actual pitting factor found from this study is in good agreement with that reported by González et al (1995). Also a challenged question for GECOR is how to judge the localised corrosion. Probably the condition of chloride content in concrete is also necessary information for a proper use of pitting factor.

Fig. 5.2 Comparison of the corrosion rate [in µA/cm²] measured by the 5 seconds galvanostatic pulse technique with the actual corrosion rate [in µA/cm²] measured by the gravimetric method.

Fig. 5.3 Comparison of the corrosion rate [in µA/cm²] measured by the GECOR instruments (× a pitting factor of 6) with the actual corrosion rate [in µA/cm²] measured by the gravimetric method. 0.001 0.01 0.1 1 10 100 0.001 0.01 0.1 1 10 100 Gr avim e tr ic M e th o d (m as s lo s s /co r r o d e d ar e a) G E C O R 6 x 6 (p itti n g fa c to r) fo r a ll C l in d u c e d s p e c im e n s ELKEM, 0% C l C BI, 0% C l ELKEM, 1.5% C l C BI, 1.5% C l ELKEM, 3% C l C BI, 3% C l ELKEM, 6% C l C BI, 6% C l 0.01 0.1 1 10 100 0.01 0.1 1 10 100 Gr avim e tr ic M e th o d (m as s lo s s /co r r o d e d ar e a) 5 s e c G a lvan o s ta ti c P u ls e SP-5s , 0% C l FOR C E, 0% C l SP-5s , 1.5% C l FOR C E, 1.5% C l SP-5s , 3% C l FOR C E, 3% C l SP-5s , 6% C l FOR C E, 6% C l

6 Concluding

Remarks

In this project, the reinforced concrete slabs with different chloride introductions were cast in one laboratory and distributed to different laboratories. After storage for over one year, three different laboratories measured fairly comparable corrosion rates based on the gravimetric method as shown in Fig. 4.22. This indicates that the method used in the project for production of specimens is basically suitable for use in the calibration of any type of non-destructive techniques for measuring the corrosion rate of steel in concrete.

From the results of the comparative measurements as presented in Chapter 4 it could be concluded that the corrosion rates obtained by the short time galvanostatic pulse

measurements (GalvaPulse and SP’s instrument) are always a several times as large as those obtained by GECOR.

From the results of the electrochemical study as described in Section 4.2, the following concluding remarks could be drawn:

• Polarisation duration has tremendous effect on the polarisation resistance calculated based on the simple Randles circuit, no matter if the steel is corroded or not. • The relationships between polarisation resistance and polarisation duration are

logarithmically linear and can be expressed as Rp = atb. At the present the constants a

and b are, however, purely empiric without any physical meanings.

• The effect of polarisation current appears not significant if the potential shift is small enough to assure a linear polarisation.

• The double layer capacitance calculated based on the simple Randles circuit reveals a similar behaviour as the polarisation resistance, that is, capacitance increases with polarisation duration.

• The differences in corrosion rate measured by two commercial instruments, GECOR and GalvaPulse, are mainly attributed to different polarisation durations.

From the results of the comparison with the true corrosion rates by the gravimetric method as reported in Chapter 5, the following concluding remarks could be drawn:

• The corrosion rates measured by GECOR6 are fairly close to the true mean corrosion rate (mass loss divided by the whole exposed area of steel), while the 5 seconds short time galvanostatic pulse measurements overestimate the true mean corrosion rate, especially for the passive steel bars. When extrapolating the 5 seconds data to 100 seconds utilising the logarithmical relationship (equation Rp = atb), the corrosion rates

measured by the galvanostatic pulse technique become closer to the true mean corrosion rate.

• The corrosion rates obtained by the 5 seconds short time galvanostatic pulse measurements from the chloride introduced specimens are close to the true actual corrosion rate (mass loss divided by the corroded area of steel), probably due to the heterogeneous distribution of the imposed galvanostatic current. The further studies are, however, needed to clarify this coincidence.

• When multiplying the values measured by GECOR6 by a factor of 6 (as a pitting factor) for the chloride introduced specimens, the re-calculated corrosion rates become very comparable with the true actual corrosion rate.

• The condition of chloride content in concrete is necessary information for proper judgement of passive or de-passive status of steel when using galvanostatic pulse techniques and proper use of pitting factor when using GECOR instruments.

Finally, it should be kept in mind that the steel bars in the real reinforced concrete structures are “infinitely” long and cross-connected each other. The measurement conditions in the field are far more complicated than those in the laboratory. The calibration in the laboratory on the small specimens under controlled conditions is the first important step toward the true world. Further investigations on large specimens or real structures are needed in order to make a comprehensive evaluation of the non-destructive electrochemical methods for their applications to various real reinforced concrete structures.

7 References

Andrade C. & Gonzáles, J.A. (1978), “Quantitative measurements of corrosion rate of reinforcing steels embedded in concrete using polarisation

resistance measurements”, Werkstoffe und Korrosion, Vol. 29, pp. 515-519.

ASTM G 1, (1990) “Standard practice for preparing, cleaning, and evaluating corrosion test specimens”, American Society for Testing and Materials, Philadelphia.

Clear K.C. (1989). “Measuring the rate of corrosion of steel in field concrete structures”, Transportation Research Record 1211, Transportation Research Board, National Research Council, Washington, DC.

Cox R.N., Cigna R., Vennesland Ø. & Valente T. (1997). “Corrosion and Protection of Metals in Contact with Concrete”, COST 509 Final Report, EUR 17608 EN, European Commission, DG XII Science, Research and Development, Brussels.

Elsener B., Klinghoffer O., Frolund T., Rislund E., Schiegg Y. & Böhni H. (1997),

“Assessment of reinforcement corrosion by means of galvanostatic pulse technique”, in “Proceedings of International Conference on Repair of Concrete Structures - From Theory to Practice in a Marine Environment”, ed. A. Blankvoll, pp. 391-400, Svolvær, Norway.

Feliu S., González J.A., Feliu S.Jr. & Andrade M.C. (1990), Confinement of the electrical signal for in situ measurement of polarisation resistance in reinforced concrete”, ACI Materials Journal, 87 (5), pp. 457-460.

Gecor 06—Corrosion rate meter for steel in concrete, Instruction manual, manufactured in Spain by Geocisa

González J.A., Andrade C., Alonso C. and Feliú S. (1995), “Comparison of rates of general corrosion and maximum pitting penetration on concrete embedded steel

reinforcement”, Cement and Concrete Research, 25 (2), 257-264.

González J.A., Molina A., Escudero M.L. & Andrade C. (1985), “Errors in the

electrochemical evaluation of very small corrosion rates. Part I. Polarisation resistance method applied to corrosion of steel in concrete”, Corrosion Science, Vol. 25, pp. 917-930.

Gowers K.R., Millard S.G., Gill J.S. & Gill R.P. (1994), British Corrosion Journal, Vol.29:pp.25

Homma K., Goto N., Matsuoka K. & Ito S. (1992), Utilization of electrochemical impedance techniques to estimate corrosion damage of steel infrastructures”, in “Corrosion Forms and Control for Infrastructure”, ed. V. Chaker, ASTM STP 1137, pp. 155-169, American Society for Testing and Materials, Philadelphia.

John D.G., Hladky K., Gaydecki, P.A. & Dawson J.L. (1992), “Recent developments in inspection techniques for corrosion damaged concrete structures”, in “Corrosion Forms and Control for Infrastructure”, ed. V. Chaker, ASTM STP 1137, pp. 246-257, American Society for Testing and Materials, Philadelphia.

Rodríguez J., Ortega L.M., Garcia A.M., Johansson L. & Petterson K. (1995), “On-site corrosion measurements in concrete structures”, Construction Repair, Nov./Dec. 1995.

reinforced concrete”, Cement & Concrete Composites, Vol. 22, pp.407-415.

Stern M & Geary A.L. (1957), “Electrochemical polarisation, I. A theoretical analysis of the shape of polarisation curves”, J. Electrochem. Soc., Vol. 104, pp. 56-63.

Videm K. & Myrdal R. (1997), “Electrochemical behavior of steel in concrete and evaluation of the corrosion rate”, Corrosion, 53 (9) 734-742.