Electrochimica Acta 49(2004)
4447–4453
Chloride induced corrosion of reinforcing steel evaluated
by concrete resistivity measurements
W.Morris,A.Vico,M.Vázquez ∗,1
División Corrosión,INTEMA,Facultad de Ingenier´ıa,Universidad Nacional de Mar del Plata,J.B.Justo 4302,B7608FDQ Mar del Plata,Argentina
Received 28October 2003;received in revid form 29April 2004;accepted 2May 2004
Available online 7June 2004
Abstract
The chloride threshold (Cl TH )concentration for rebar corrosion initiation has received extensive attention over the last years.The chlo-ride threshold concentration depends on veral factors involving concrete composition and quality,exposure conditions and rebar surface characteristics.As a conquence,many rearchers have propod Cl TH ranges that take into account the relative influence of each of the many factors.On the other hand,the electrical resistivity of concrete has pr
oven to be an effective parameter that can be ud to estimate the risk of reinforcing steel corrosion,particularly when corrosion is induced by chloride attack.The prent study is bad on a correlation of electrochemical parameters such as corrosion potential (E corr )and current density (i corr )together with concrete resistivity (ρ)and chloride concentration data.A relationship between chloride threshold values for rebar corrosion initiation and resistivity values (indicative of concrete quality)is propod.According to this correlation,when the electrical resistivity of concrete increas from 2to 100k cm,the value of Cl TH increas from 0.44to 2.32%relative to the weight of cement.©2004Elvier Ltd.All rights rerved.
Keywords:Steel;Reinforced concrete;Corrosion;Resistivity;Chloride threshold
1.Introduction
It is widely accepted that chloride ions are responsible for causing local passive layer breakdown and subquent corrosion of reinforcing steel bars (rebars).This can be the ca when concrete is either expod to marine environ-ment or deiceing salts and also when concrete is prepared with contaminated aggregates.The chloride-induced corro-sion mechanism has been,and continues to be,extensively investigated.There is a general agreement in the existence of a certain value repre
nting the chloride threshold (Cl TH )concentration that must be reached at the rebar surface in order to initiate the corrosion process [1–4].The chloride threshold value is of great importance when evaluating the rvice life of reinforced concrete structures,as it deter-mines how long it takes to initiate the corrosion process [5,6].Many investigations have focud in determining the
∗
Corresponding author.Tel.:+542234816600;fax:+542234810046.
E-mail address:mvazquez@fi.mdp.edu.ar (M.V´a zquez).1ISE member.
Cl TH value [7–10],although the results that can be found in the literature show great variability.
As discusd by Alonso et al.[10],there are many vari-ables that affect the Cl TH value.Some of the depend on the concrete mix properties,such as the cement type con-tent,the water to cement ratio [11–15]and the pore so-lution pH [16,17].Other variables are related to the rein-forcement:the rebar composition and surface condition [18]should be considered.The environment also plays an im-portant role,determining temperature and relative humidity [19–22],eventually also being the source of chloride pene-tration [23].The oxygen availability at the rebar surface as well as its polarisa
tion potential [1,2,16,24]also need to be taken into account.The great number of variables involved in the chloride-induced corrosion process explains the rea-son why the Cl IH values reported in the literature prent such wide range of variability.
Another important issue still under discussion is the most appropriate way to define the chloride threshold value in concrete.Glass and Buenfeld [9]carried out an extensive re-view of chloride threshold levels published in the literature.They concluded that chloride threshold levels are best pre-
0013-4686/$–e front matter ©2004Elvier Ltd.All rights rerved.doi:10.1016/j.electacta.2004.05.001
4448W.Morris et al./Electrochimica Acta49(2004)4447–4453
nted as a total(bond plus free chlorides)content expresd relative to the weight of cement.The values of Cl TH re-ported in that review varied within a wide range(0.17–2.5% relative to the weight of cement).Alonso et al.[10]argue that chloride to hydroxyl ion ratio would be the best way of defining the value of Cl TH,although it is also suggested that either free or total chlorides expresd relative to the cement content in concrete are appropriate ways of defining Cl TH. According to them,the chlo
ride threshold may vary from 1.24to3.08%when considering total chlorides which,in turn,would reprent a variation from0.39to1.16%when only free chlorides are taken into account.
Further than the aspects addresd above,and accepting that Cl TH depends on many factors,one more question aris that needs a practical answer:how to establish an appropri-ate Cl TH value for a given structure expod to a particular environment.
In a previous publication[25],the electrical resistivity of concrete was propod as an effective parameter to evalu-ate the risk of reinforcing steel corrosion,particularly when corrosion is induced by chloride attack.Likewi the Cl TH value,the resistivity of concrete is strongly dependent on the concrete quality and on the exposure conditions,such as the relative humidity.Also,temperature affects the degree of concrete pore saturation[19,26,27]and so the resistiv-ity values.Therefore,the idea offinding a relationship be-tween concrete resistivity and Cl TH values would em quite feasible.In order to determine the existence of such corre-lation,certain electrochemical parameters as the corrosion potential(E corr)and current density(i corr)are investigated together with concrete resistivity and chloride concentration data.
2.Experimental data
The results evaluated in this work were taken from pre-vious investigations conducted by the authors[25,28,29]. The performance of four different concrete proportioning was investigated over a period of1000days of exposure to ashore and immersion conditions.Chloride concentrations varied from0.16to1.6%relative to the cement content.A detailed description of the experimental t-up can be found in a previous publication[25].
2.1.Samples design
Cylindrical concrete test specimens(15cm diameter, 22cm height)were prepared.Each contained four rebar gments positioned in such a way that a1.5cm concrete cover was achieved(e Fig.1).The rebar gments prent an expod area of40cm2.
Table1prents the mix proportioning lected for the study.Mixes A and B were prepared using a sand of the same type of that ud by the local construction indus-try(siliceous,fineness modulus=2.7±0.3and
specific Fig.1.Schematic reprentation of the cylindrical concrete samples ud in the study.
gravity=2.7).Mixes C and D were prepared using river sand(fineness modulus=2.7and specific gravity=2.65) containing less than0.1%per weight of chloride ions.In the ca of mix design C,NaCl was incorporated in a known proportion to reproduce the ca of reinforced concrete struc-tures heavily contaminated.Prepartion D has no admixed chlorides.The total initial chloride concentration(ASTM C-1152),expresd in percentage by weight of cement is prented in Table2.
2.2.Exposure conditions
Two exposure conditions,referred as ashore and im-merd,were evaluated.Four specimens per each mix de-sign were prepared(totalling16).Two samples per each mix were placed at the terrace of a40-floor building located at less than100m from the acoast in Mar del Plata city (latitude S:3756;longitude W:5735).The samples are labelled as ashore specimens and were directly expod to rainfall,a spray and wind.The other eight specimens Table1
Concrete mix proportioning,slump test results and sample identification Mix proportioning
identification
A B C D
Water/cement ratio(w/c)0.600.400.600.60 Cement content(kg m−3)300400300300 Fine aggregate(FA)
River sand(kg)––851858 Sea sand(kg)858789––River rock
MAS=10mm(kg)
1003104210031003
Sodium chloride(kg)––7.4–Superplastisizer(%) 1.0 2.5––Slump test(cm) 3.0 3.0 6.58.0
W.Morris et al./Electrochimica Acta 49(2004)4447–4453
4449
Table 2
Compressive strength (f C ),porosity (P =%of air voids)and the total initial chloride concentration ([Cl −]0)in the concrete mixes lected for the study Concrete mix
f C (Mpa)P (%)
[Cl −]0(%)
7days
28days A 14.421.318.30.78B 21.531.412.10.43C 16.221.017.7 1.65D
14.0
22.5
18.5
0.16
The compressive strength was determined after 7and 28days of curing.Chloride concentration is expresd as percent by weight respect to cement content.The concrete porosity and the total initial chloride concentration (acid soluble)were determined after approximately 100days of casting the specimens.Values are average results of duplicate measurements.
were partially immerd in aerated saline solution contain-ing 3.5%of NaCl by volume.2.3.Mechanical and chemical analysis
Standard size concrete specimens were also prepared fol-lowing the ASTM C-39standard in order to characteri the mechanical properties of the concrete mixes.Compres-sive strength was evaluated (ASTM C-617)at 7and 28days after casting the specimens.In order to determine the total chloride ions concentration and the percentage of air voids (porosity of hardened concrete,ASTM C-642),an-other group of concrete specimens was also prepared.
The results of the characterisation analys are pre-nted in Table 2.Replicate specimens were tested for each analysis.
Chloride concentration profiles were obtained from con-crete cylindrical cores (2cm diameter,5cm long)extracted from the test specimens using a drilling machine.The extraction procedure was carried out after 850days of ex-posure to both,ashore and immerd conditions.Cores extracted from the ashore specimens were drilled from the side that was facing the a.Fig.2illustrates the cores extraction method and the chloride profile
determination
Fig.2.Cores extraction and chloride profile determination procedure.
Table 3
Chloride effective diffusion coefficient (D eff )and surface concentration (C s )obtained from the mathematical curve fitting of the chloride content in-depth profiles [25]Concrete mix
Immerd condition Seashore environment D eff (cm 2s −1)C s (%)D eff (cm 2s −1)C s (%)A 5.09×
10−8 3.2 4.83×10−8 1.9B 1.88×10−8 2.4 1.79×10−8 1.5C 5.95×10−8 3.8 4.53×10−8 2.8D
4.84
×10−8
3.2
3.88
×10−8
1.2
procedure.In order to analy the chloride transport mech-anism within the concrete specimens,the effective diffusion coefficient (D eff )was determined by solving Fick’s cond law [25].D eff values are summarid in Table 3.The chlo-ride profiles in the two exposure conditions under analysis are prented in Fig.3.The chloride concentrations are expresd in percentage of total chlorides by weight of cement.
2.4.Electrical and electrochemical measurements The electrochemical parameters normally ud to charac-teri the corrosion behaviour of reinforcing steel in concrete were monitored periodically.The parameters included the corrosion potential,the corrosion current density obtained from polarisation resistance measurements (R p )and the elec-trical resistivity of concrete (ρ)determined from resistance measurements between the two uncoated rebar gments.The electrical resistance (R ,in )between two rebars bars was measured using a Nilsson 400soil resistivity meter.This instrument us a square wave of 97Hz,preventing po-larisation of the electrodes.The values of ρwere calculated as:ρ=kR ,where k is a geometrical factor that depends on the shape of the sample.In this ca,k =7.5and 12cm for the dry and wet condition,respectively.
Experimental procedures and the equipment employed to evaluate the electrochemical parameters can be found in de-tail elwhere [25].
4450W.Morris et al./Electrochimica Acta 49(2004)
4447–4453
Fig.3.Chloride concentration profiles obtained from concrete specimens after 850days of exposure to the immerd condition (left)and ashore environment (right).Vertical lines show the location of the rebars in the concrete sample.
The data reported in the figures correspond to average values of four independent measurements (two samples con-taining two rebars each one).
3.Results and discussion
On the basis of the compressive strength and porosity re-sults,the concrete mixes can be classified in two groups according to their quality.On one hand,mixes A,C and D reprent examples of standard quality concrete.They were prepared with the higher water to cement ratio (w /c =0.6)and chloride contents ranging from 0.16to 1.65%.The compressive strengths values (fc)at 28days varied be-tween 21and 22.5MPa and porosity (P )values between 17.7and 18.5%.On the other hand,mix B,prepared with w /c =0.4and a-sand,prented a fc value of 31.4MPa and a porosity of 12.1%,values typical of a good quality concrete.
As a result of an extensive investigation [25,28,29],it was found that rebars in contact with a good quality
concrete
Fig.4.Variation of the rebar corrosion potential (E corr )with time on mix designs A ([Cl −]0=0.78%,w /c =0.6);B ([Cl −]0=0.43%,w /c =0.4);C ([Cl −]0=1.65%,w /c =0.6);and D ([Cl −]0=0.16%,w /c =0.6).Horizontal lines define a region of active corrosion where E corr <−0.35V and a passive corrosion region where E corr >−0.2V vs.CSE [30].Vertical lines show the day of exposure to each environment,after a conditioning period indoors.(a)Samples were expod to ashore environment;(b)samples were partially immerd in aerated saline solution.
expod for 1000days to the ashore environment and pre-pared with a high content of admixed chlorides (0.43%)re-mained in the passive state,even when its surface chloride concentration reached 1%with respect to cement content (specimen B).Also,a standard quality concrete prepared with a w /c =0.6and 0.78%of admixed chlorides (sample A)would provide a protective environment for steel bars as long as the specimens were kept in a dry environment [25].The evolution in time of the corrosion potential and the corrosion current density support the statements and are prented in Figs.4a and 5a .
When immerd in a saline solution,all rebar gments achieved an active corrosion state,although the good quality concrete (samples B)showed a less active rebar corrosion behaviour than the standard quality one [25].The results of corrosion potential and corrosion current density for this con
dition can be en in Figs.4b and 5b .
The different quality of the concrete mixes under study can be clearly distinguished by obrving the electrical re-sistivity data (e Fig.6a and b ).Mix B (good quality con-crete)prents resistivity values that are approximately three
W.Morris et al./Electrochimica Acta 49(2004)4447–4453
4451
Fig.5.Variation of the apparent rebar corrosion rate (CR)and the corrosion current density (i corr )as a function of time on mix designs A (w /c =0.6,[Cl −]0=0.78%);B (w /c =0.4,[Cl −]0=0.43%);C (w /c =0.6,[Cl −]0=1.65%);and D (w /c =0.6,[Cl −]0=0.16%).Horizontal lines define a region of active corrosion where i corr >0.2A cm −2and a passive corrosion region where i corr <0.1A cm −2[31].(a)Samples were expod to ashore environment;(b)samples were partially immerd in aerated saline solution.
times higher than the values obrved on mixes A,C and D (all of them of standard quality).The difference is evi-dent both in ashore and immerd environments.As could be expected,the resistivity of the specimens expod to the ashore environment are approximately three times greater than the corresponding values measured in the immerd condition.
The results of the rebar corrosion measurements (E corr and i corr )together with the calculated values of rebar surface chloride concentration and the concrete resistivity values were reorganid to calculate chloride threshold values.The corrosion potential,the corrosion current density and the resistivity were measured at the same date.The chloride concentration at the rebar surface was calculated at each particular date using the data in Table 3.
Figs.7and 8prent a correlation between corrosion po-tential and corrosion rate values with respect to the
concrete
Fig.6.Variation of the electrical resistivity of concrete (ρ)as a function of time on mix proportions A (w /c =0.6,[Cl −]0=0.77%);B (w /c =0.4,[Cl −]0=0.43%);C (w /c =0.6,[Cl −]0=1.65%);and D (w /c =0.6,[Cl −]0=0.16%).Horizontal lines define a region of active corrosion where ρ<10k cm and a passive corrosion region where ρ>30k cm [25].Vertical lines show the day of exposure to each environment,after a conditioning period indoors.(a)Samples were expod to ashore environment;(b)samples were partially immerd in aerated saline
solution.
Fig.7.Correlation between corrosion potentials (E corr )and concrete re-sistivity (ρ)values measured at the same time on specimens expod to ashore and immerd conditions.Data points reprented by hollow symbols (᭺)where taken from reference 28.
