corrosion of metals in concrete

CORROSION OF METALS 222R-1

This committee report has been prepared to reflect the stateof the art of the corrosion of metals, and especially steel, inconcrete. Separate chapters are devoted to the mechanismsof the corrosion of metals in concrete, protective measuresfor new concrete construction, procedures for identifyingcorrosive environments and active corrosion in concrete,and remedial measures. A selected list of references isincluded with each chapter.

Keywords: admixtures; aggregates; blended cements; bridge decks; calcium chlorides; carbonation; cathodic protection; cement pastes; chemicaanalysis; chlorides; coatings; concrete durability; corrosion; corrosion re-sistance; cracking (fracturing); deicers; deterioration; durability; marine at-mospheres; parking structures; plastics, polymers and resins; portlancements; prestressing steels; protective coatings; reinforced concrete; re-inforcing steels; repairs; resurfacing; spalling; waterproof coatings.

ACI 222R-96

Corrosion of Metals in Concrete

Reported by ACI Committee 222

ACI Committee Reports, Guides, Standard Practices, andCommentaries are intended for guidance in designing, plan-ning, executing, or inspecting construction and in preparingspecifications. Reference to these documents shall not bemade in the Project Documents. If items found in these doc-uments are desired to be part of the Project Documents, theyshould be phrased in mandatory language and incorporatedin the Project Documents.

222R

ACI COMMITTEE 222Corrosion of Metals in Concrete

Brian B. HopeChairman

Keith A. PashinaSecretary

John P. Broomfield Bret James Robert E. PriceKenneth C. Clear Thomas D. Joseph D. V. ReddyJames R. Clifton David G. Manning William T. ScannellIsrael Cornet Walter J. McCoy David C. StarkMarwan Daye Theodore L. Neff Wayne J. SwiatBernard Erlin Charles K. Nmai Thomas G. WeilJohn K. Grant William F. Perenchio Richard E. WeyersKenneth C. Hover Randall W. Poston David A. Whiting

Associate MembersStephen L. Amey Odd E. Gjorv Mohamad A. NagiSteven F. Dailey Clayford T. Grimm Morris SchupackStephen D. Disch Alan K. C. Ip Ephraim SenbettaHamad Farzam Andrew Kaminker Robert E. ShewmakerPer Fidjestol Mohammad S. Khan Bruce A. SuprenantRodney R. Gerard Philip J. Leclaire William F. Van SiserenMichael P. Gillen Joseph A. Lehmann Michael C. Wallrap

CONTENTSChapter 1—Int roduction, p. 222R-1

1.1—Background1.2—Scope1.3—References

ion
-l
Chapter 2—Mechanism of cor rosion of steel inconcrete, p. 222R-3

2.1—Introduction2.2—Principles of corrosion2.3—Effects of the concrete environment on corros2.4—References

d

ride

lo-

-

Chapter 3—P rotection against cor rosion in n ewconstruction, p. 222R-11

3.1—Introduction3.2—Design and construction practices3.3—Methods of excluding external sources of chlo

ion from concrete3.4—Methods of protecting reinforcing steel from ch

ride ion

ACI 222R-96 replaces ACI 222R-89 and became effective May 23, 1996.Copyright (c) 1997, American Concrete Institute. All rights reserved including

rights of reproduction and use in any form or by any means, including the making ofcopies by any photo process, or by any electronic or mechanical device, printed orwritten or oral, or recording for sound or visual reproduction or for use in any knowl-edge of retrieval system or device, unless permission in writing is obtained from thecopyright proprietors.

1

222R-2 ACI MANUAL OF CONCRETE PRACTICE

The art.onsrce-ary

pre-oc-

3.5—Corrosion control methods3.6—References

Chapter 4—P rocedures for identifying cor rosionenvi ronments and active cor rosion in concrete,p. 222R-21

4.1—Introduction4.2—Methods of evaluation4.3—References

ons, re--

low andrro- inon-

Chapter 5—Remedial measures, p. 222R-235.1—Introduction5.2—General5.3—Applicability5.4—The remedies and their limitations5.5—Summary5.6—References

n asridestruc-

d inothcor-

Chapter 6—References to documents of standa rd-producing o rganizations, p. 222R-28

Appendix A—Standa rd Test Method for Water-Soluble Chloride Available for Cor rosion ofEmbedded Steel in Mo rtar and Concrete Using theSoxhlet Extracto r, p. 222R-28

el-t inlm. Inme

ricaro- prothe ofuatserted

hasf it

d thentanuofthe

artic facoure ond-

air

t begn,

di-ellry- arecon-ste, duealsoMixea-

callyl the

usehanding

udent-

teelf thebe ofpre-uld

steelsedible.on-son-ix in-

onome

CHAPTER 1—INTRODUCTION1.1—Background

Concrete normally provides reinforcing steel with exclent corrosion protection. The high alkaline environmenconcrete results in the formation of a tightly adhering fiwhich passivates the steel and protects it from corrosionaddition, concrete can be proportioned to have a low perability which minimizes the penetration of corrosion-inducingsubstances. Low permeability also increases the electresistivity of concrete which impedes the flow of electchemical corrosion currents. Because of these inherenttective attributes, corrosion of steel does not occur in majority of concrete elements or structures. Corrosionsteel, however, can occur if the concrete is not of adeqquality, the structure was not properly designed for the vice environment, or the environment was not as anticipaor changes during the service life of the concrete.

The corrosion of metals, especially steel, in concretereceived increasing attention in recent years because owidespread occurrence in certain types of structures anhigh cost of repairs. The corrosion of steel reinforcemwas first observed in marine structures and chemical mfacturing plants.1.1,1.2,1.3 More recently, numerous reports its occurrence in bridge decks, parking structures, and ostructures exposed to chlorides has made the problem pularly prominent. The consequent extensive research ontors contributing to steel corrosion has increased understanding of corrosion, especially concerning the rolchloride ions. It is anticipated that the application of the fiings of this research will result in fewer instances of corrosionin new concrete structures and improved methods of reping corrosion-induced damage in existing structures.

For these improvements to occur, the information musdisseminated to individuals responsible for the desi

-

l

-

e-

se

-

r--

f

-

construction, and maintenance of concrete structures. main purpose of this report is to present the state of theWhile several metals may corrode under certain conditiwhen embedded in concrete, the corrosion of steel reinfoment is of the greatest concern, and, therefore, is the primsubject of the report.

Chloride ions are considered to be the major cause of mature corrosion of steel reinforcement. Corrosion can cur in some circumstances in the absences of chloride ihowever. For example, carbonation of concrete results induction of its alkalinity, thereby permitting corrosion of embedded steel. Carbonation, however, is usually a sprocess in concrete which has a low water-cement ratiocarbonation-induced corrosion is not as common as cosion induced by chloride ions. Chloride ions are commonnature and small amounts are usually unintentionally ctained in the mix ingredients of concrete.

Chloride ions also may be intentionally added, most oftea constituent of accelerating admixtures. Dissolved chloions also may penetrate unprotected hardened concrete in tures exposed to marine environments or to deicing salts.

The rate of corrosion of steel reinforcement embeddeconcrete is strongly influenced by environmental factors. Boxygen and moisture must be present if electrochemical rosion is to occur. Reinforced concrete with significant graents in chloride ion content is vulnerable to macroccorrosion, especially if subjected to cycles of wetting and ding. Other factors that affect the rate and level of corrosionheterogeneities in the concrete and the steel, pH of the crete pore water, carbonation of the portland cement pacracks in the concrete, stray currents, and galvanic effectsto contact between dissimilar metals. Design features play an important role in the corrosion of embedded steel. proportions, depth of cover over the steel, crack control msures, and implementation of measures designed specififor corrosion protection are some of the factors that controonset and rate of corrosion.

Deterioration of concrete due to corrosion results becathe products of corrosion (rust) occupy a greater volume tthe steel and exert substantial stresses on the surrounconcrete. The outward manifestations of the rusting inclstaining, cracking, and spalling of the concrete. Concurrely, the cross section of the steel is reduced. With time, structuraldistress may occur either by loss of bond between the sand concrete due to cracking and spalling or as a result oreduced steel cross-sectional area. This latter effect can special concern in structures containing high strength stressing steel in which a small amount of metal loss copossibly induce tendon failure.

The research on corrosion to date has not produced aor other type of reinforcement that will not corrode when uin concrete and that is both economical and technically feasHowever, research has pointed to the need for quality ccrete, careful design, good construction practices, and reaable limits on the amount of chloride in the concrete mingredients. Other measures that are being investigatedclude the use of corrosion inhibitors, protective coatingsthe steel, and cathodic protection. There has been s


fromy a

success in each of these areas though problems resultingcorrosion of embedded metals are far from eliminated.

ntro emctingures dis-ssist in- and

tion

ium,n ofter,

upply

1.2—Scope This report discusses the factors that cause and co

corrosion of steel in concrete, measures for protectingbedded metals in new construction, techniques for detecorrosion in structures in service, and remedial procedConsideration of these factors, and application of thecussed measures, techniques, and procedures should areducing the occurrence of corrosion and result, in mosstances, in the satisfactory performance of reinforcedprestressed concrete elements.

in-

-

tial,edox

o the is

ors

1.3—References 1.1. Tremper, Bailey; Beaton, John L; and Stratfull, R. F. “Causes and

Repair of Deterioration to a California Bridge Due to Corrosion of Reforcing Steel in a Marine Environment II: Fundamental Factors CausingCorrosion,” Bulletin No. 182, Highway Research Board, Washington, D.C.,1958, pp. 18-41.

1.2. Evans. U. R., The Corrosion and Oxidation of Metals: Scientific Principles and Practical Applications, Edward Arnold, London, 1960, 303 pp.

1.3. Biczók, Imre, Concrete Corrosion and Concrete Protection, 3rdEdition, Académiai Kiadó, Budapest, 1964.

m-ion m ie o

xam

cretmixsion

tive is

andntentr ofyed,in-

e ex-ng.are

the

CHAPTER 2—MECHANISM OF CORROSION OFSTEEL IN CONCRETE

2.1—Int roduction This chapter is divided into two sections. In the first, e

phasis is placed on the processes responsible for corrossteel reinforcement in concrete. The corrosion mechanisgenerally accepted to be electrochemical in nature. Somthe major causes of corrosion of steel in concrete are eined and related phenomena are discussed.

In the second part, a discussion is given on the convariables that influence corrosion, including concrete proportions, quality, and cover, and the effects of corroinhibiting admixtures and carbonation.

omis

hede,odeonet

e o bods a

zed

atednd i

,n,xy-

hodicstate. im-uldion

on-yondased

be-trib-nd,sionwev-nt.ldomously

ive

2.2—Principles of cor rosion2.2.1 Corrosion—An electrochemical process—Al-

though iron can corrode by chemical attack, the most cmon form of corrosion in an aqueous medium electrochemical.2.1 The corrosion process is similar to taction which takes place in a flashlight battery. An anowhere electrochemical oxidation takes place, a cathwhere electrochemical reduction occurs, an electrical cductor, and an aqueous medium must be present. Any msurface on which corrosion is taking place is a compositanodes and cathodes electrically connected through theof the metal itself. Reactions at the anodes and cathodebroadly referred to as “half-cell reactions.”

At the anode, which is the negative pole, iron is oxidito ferrous ions.

Fe Fe++ + 2e-

E˚ = - 0.440 Standard Redox Potential (2-1)

The Standard Redox Potential is the potential generwhen the metal is connected to a hydrogen electrode a

l-

.

t in

one method of expressing electromotive forces. The Fe++ inEq. (2-1) is subsequently changed to oxides of iron bnumber of complex reactions. The volume of the reacproducts is several times the volume of the iron.

At the cathode, reduction takes place. In an acid medthe reaction taking place at the cathode is the reductiohydrogen ions to hydrogen. However, as will be shown laconcrete is highly basic and usually has an adequate sof oxygen, so the cathodic reaction is

1/2 H2O + 1/4 O2 + e- OH -

E˚ = 0.401 Standard Redox Potential (2-2)

The corroding iron piece has an open circuit potenalso called a rest potential, related to the Standard RPotentials of the reactions in Eq. (2-1) and (2-2), to the com-position of the aqueous medium, the temperature, and t“polarizations” of these half-cells. The rate of corrosionrelated to the “polarizations” as will be discussed later.

ofsf-

e

-

,-alfyre

2.2.2 Availability of oxygen in concrete—Although theavailability of oxygen is one of the main controlling factfor corrosion of steel, there appears to be little quantitainformation on its effect in concrete. Some informationshown, however, in Fig. 2.1, where the rate of oxygen diffusionthrough water-saturated concrete of varying quality thickness is illustrated.2.2 The rate of oxygen diffusiothrough concrete is also significantly affected by the exto which the concrete is saturated with water. A numbeinvestigations indicate that if the steel passivity is destroconditions will be conducive to the corrosion of steel reforcement in those parts of a concrete structure that arposed to periods of intermittent wetting and dryiInvestigations also indicate that, although chlorides present, the rate of steel corrosion will be very slow ifconcrete is continuously water-saturated.2.3 In wet concretedissolved oxygen will primarily be diffusing in solutiowhile in a partly dry concrete the diffusion of gaseous ogen is much faster. For oxygen to be consumed in a catreaction, however, the oxygen has to be in a dissolved Since it is the concentration of dissolved oxygen that isportant, all factors affecting the solubility of oxygen shoalso affect its availability. The effect of salt on the corrosrate has been demonstrated by Griffin and Henry2.4 (seeFig. 2.2). Corrosion increased as the sodium chloride ccentration increased until a maximum was reached. Bethis, the rate of corrosion decreased despite the increchloride ion concentration. This change in relationshiptween corrosion and sodium chloride concentration is atuted to the reduced solubility and diffusivity of oxygen, atherefore, the availability of oxygen to sustain the corroprocess. This represents corrosion in a salt solution; hoer, the availability of oxygen in wet concrete may be differe

Problems due to corrosion of embedded steel have sebeen observed in concrete structures that are continusubmerged, even as in seawater.

s2.2.3The importance of chloride ions—As will be discussed

later, concrete can form an efficient corrosion-prevent


h mor-
Fig. 2.1—Effect of water-cement ratio and thickness on diffusion of oxygen througtar and concrete2.2
suchts ofg ef-

awa-xingo sea-ntaint are

eces-

in-e
Fig. 2.2—Effect of concentration of sodium chlorid
on corrosion rate2.4

elln- andothe thehlo

y ofg adatesrine

ride

epas-pos-teel ionslyrate.n

environment for embedded steel. However, it is wdocumented2.5-2.7that the intrusion of chloride ions in reiforced concrete can cause steel corrosion if oxygenmoisture are also available to sustain the reaction. No common contaminant is documented as extensively inliterature as a cause of corrosion of metals in concrete. Cride ions may be introduced into concrete in a varietways. Some are intentional inclusion as an acceleratinmixture; accidental inclusion as contaminants on aggregor penetration by deicing salts, industrial brines, maspray, fog, or mist.

ce-olidnt is

etal

he

2.2.3.1Incorporation in concrete during mixing. One ofthe best known accelerators of the hydration of portlandment is calcium chloride. Generally, up to 2 percent scalcium chloride dihydrate based on the weight of ceme

added. Chlorides may be contained in other admixtures as some water-reducing admixtures where small amounchloride are sometimes added to offset the set-retardinfect of the water reducer.

In some cases, where potable water is not available, seter or water with a high chloride content is used as the miwater. In some areas of the world, aggregates exposed twater (or that were soaked in seawater at one time) can coa considerable quantity of chloride salts. Aggregates thaporous can contain larger amounts of chloride.

2.2.3.2Diffusion into mature concrete. Chlorides canpermeate through sound concrete, i.e., cracks are not nsary for chlorides to enter the concrete.2.8 Approximatedeterminations of the diffusion coefficients for chlorideconcrete have been published,2.9 but largely the literature neglects the interaction between concrete and chloride.

r

-

-;

2.2.3.3Electrochemical role of free chloride ions. Thereare three modern theories to explain the effects of chloions on steel corrosion:

(a) The Oxide Film Theory—Some investigators believthat an oxide film on a metal surface is responsible for sivity and thus protection against corrosion. This theory tulates that chloride ions penetrate the oxide film on sthrough pores or defects in the film easier than do other(e.g., SO4

-). Alternatively, the chloride ions may colloidaldisperse the oxide film, thereby making it easier to penet

(b)The Adsorption Theory—Chloride ions are adsorbed othe metal surface in competition with dissolved O2 or hydrox-yl ions. The chloride ion promotes the hydration of the mions and thus facilitates the dissolution of the metal ions.

(c) The Transitory Complex Theory—According to thistheory, chloride ions compete with hydroxyl ions for t


ofm

lecatesionsrve

ightste rus

tod tur- as

ducaes.

ferrous ions produced by corrosion. A soluble complexiron chloride forms.2.9 This complex can diffuse away frothe anode destroying the protective layer of Fe(OH)2, per-mitting corrosion to continue. Some distance from the etrode, the complex breaks down, iron hydroxide precipitand the chloride ion is free to transport more ferrous from the anode. Evidence for this process can be obsewhen concrete with active corrosion is broken open. A lgreen semisolid reaction product is often found near the which, on exposure to air, turns black and subsequentlyred in color.

Since corrosion is not stifled, more iron ions continuemigrate into the concrete away from the corrosion site anreact with oxygen to form higher oxides that result in a fofold volume increase. It is the expansion of iron oxidesthey are transformed to higher oxidation states that prointernal stress, which eventually cracks the concrete. Formtionof iron chloride complexes may also lead to disruptive forc

incellenton-s car

2.2.4Corrosion rate and pH—As illustrated in Fig. 2.3, thecorrosion rate of iron is reduced as the pH increases. Sconcrete has a pH higher than 12.5, it is usually an excemedium for protecting steel from corrosion. Only under cditions where salts are present or the concrete cover habonated does the steel become vulnerable to corrosion.

-ctri-the

eouat ativeon , ar,d to

rat, threte

phe

wero. Alver fac

ssing

sionlo-

l-

ble

-

d

elt

o

e

Fig. 2.3—Effect of pH on corrosion of iron in aeratedsoft water at room temperature2.1

s

ine

ee

-

2.2.5 Pourbaix diagrams—Pourbaix2.10 devised a compact summary of thermodynamic data in the form of elecal potential versus pH diagrams, which includes electrochemical and corrosion behavior of iron in an aqumedium. Fig. 2.4 indicates that iron is in a passive state pH in the range of 8 to 13. This accounts for the protecproperties of concrete and the absence of steel corrosiconcrete when no chloride ions, or only small amountspresent. A careful inspection of Fig. 2.4 indicates, howeverthat corrosion may begin if the pH of the system is raiseabove about 13. In this case, a soluble ferrite, HFeO2

-, forms.Any mechanism where iron dissolves at an appreciablein concrete can be considered serious corrosion. Thusaddition of materials that further increase the pH of concmay be detrimental; however, the occurrence of this nomenon in concrete has not been confirmed.

-

-

e

-

Fig. 2.4—Pourbaix diagram for iron2.10

riateon in gen-

osedcingsed

thode.ly ex-

2.2.6 High-strength steels and other metals—There is lit-tle information in the literature suggesting that the nehigh-strength steels are either more or less subject to corrsionthan the previous “standard” and lower strength steelsthough information is lacking, it is expected that whatedifferences might exist would be secondary to the majortors influencing corrodibility of the steel.

Studies have been made on the corrosion of prestresteel in concrete2.11 that demonstrate the hazards of addchlorides to concrete containing prestressed steel.

Aluminum corrodes in concrete and the rate of corrois higher if the aluminum is in contact with steel and chrides or if alkalies are present.2.12 Lead and tin (e.g., tin soder) can corrode similarly.

At pH about 12.5, zinc reacts rapidly to form soluzincates and hydrogen gas is liberated

d

Zn + OH- + H2O -> HZnO2- + H2 ↑ (2-3)

If galvanized steel is to be used in concrete, appropmeasures should be taken to prevent hydrogen evolutithe fresh concrete. A small amount of a chromate salt iserally used.2.13

In submerged concrete structures having freely expsteel components in metallic connection with the reinforsteel, galvanic cells may develop with the freely exposteel forming the anode and the embedded steel the caThis may cause an increased corrosion rate on the freeposed steel.


notceprine

ntd

bri-

n

siv-ar-

the fur-or-s

ons

Copper, chromium, nickel, and silver generally do corrode in concrete. However, they are somewhat sustible to corrosion when the concrete is located in a maenvironment.2.14

chcenell

hifttialo-le

ionly

ally poe-

plycennbe

en-c-

sis-anypo-

ce

ts the

ele o

e ofon-henrentedater

f atial

pHn al-kly du-

odiccur. byrent byde as- andur- cir-ichex- the

witho be

cally

ingion,ete

2.2.7 Polarization of half-cells—The equilibrium of ahalf-cell is characterized by its reversible potential whidepends on the Standard Redox Potential, on the contrations (activities) of the species participating in the half-creaction (e.g., OH-, Fe++, etc.), and on the temperature.

When a current flows through the half-cell, there is a sof its measured potential away from the reversible potenThis shift is called polarization. For a given current, the plarization is large for half-cell reactions that are irreversibor nearly so. For instance, the oxygen half-cell react[Eq. (2-2)] is almost irreversible and can have a relativehigh degree of polarization.

Some corrosion processes, although thermodynamicmore favored than others as indicated by the reversibletentials of the half-cells, are actually slower in practice bcause of polarization effects.

There are three general kinds of polarization that apto the anode as well as to the cathode. These are contration polarization, ohmic polarization, and activatiopolarization. These three kinds of polarization can present simultaneously.

(a) Concentration polarization occurs when the conctration of the electrolyte changes in the vicinity of the eletrode. An example of this would be depletion of oxygenconcentration at the cathode.

(b) Ohmic polarization occurs because of the ohmic retance of the electrolyte (e.g., moist concrete) and of films on the electrode surface. This produces an ohmic tential drop in accordance with Ohm’s law (IR drop).

(c) Activation polarization occurs due to kinetic hindranof the rate controlling step of the electrode reaction.

Tafel2.1 has shown experimentally that for large currenin the absence of concentration and ohmic polarization,measured polarization η, which is the activation polarization, isdirectly proportional to the logarithm of the current density i

η = a + blog i

where a and b are the so-called Tafel intercept and Tafslope parameters, respectively. These parameters can btained by plotting η versus i on semilogarithmic paper. The

Tafel intercept parameter a is related to the exchange curredensity io, which is the equilibrium current flowing back anforth through the electrode electrolyte interface at equilium and a measure of the reversibility of the reaction.

The Tafel slope parameter b, on the other hand, gives ainsight according to modern electrode reaction theory2.15

into the mechanism of the electrode reaction.

-

-

.

-

2.2.8 Passivity and transpassivity—Specific to the anodichalf-cell in the corrosion process are the concepts of pasity and transpassivity. Passivity of a metal is generally chacterized by a thin and tightly adherent oxide film on metal surface, which tends to protect the metal againstther corrosion. The exact composition of the thin and nmally invisible film has been difficult to determine. It seemclear, however, that it is made up of chemical combinatiof oxygen and is simply called an oxide film.

When a potential is applied to an iron electrode, the ratcurrent flow depends on the state of passivity. Iron in ccrete is generally passive, such that little current flows wa potential is progressively increased, eventually the curwill flow. This is because, at this point, oxygen is evolvand the electrode reaction involves the electrolysis of w(see Curve I, Fig. 2.5).

2H2O -> 4H+ + O2 + 4e- (2-4)

This phenomenon is called “transpassivity.” The use ocorrosion testing procedure involving an impressed potenhas been reported.2.17 Grimes et al.2.18 found that when ahigh anodic voltage was applied to steel in concrete, thearound the steel was changed to values ranging betweemost 0 and 4. He also found that a liquid pool formed quicaround the reinforcing steel. Neither steel nor concrete israble in a low pH (acidic) environment.

Fig. 2.5—Iron electrode in concrete2.16

-

b-

2.2.9 Types of corrosion-controlling mechanisms—Aspreviously mentioned, it is necessary to have both a cathand an anodic reaction for a corrosion process to ocHowever, different corrosion situations can be visualizedplotting, on the same graph, the log of the absolute cur(I) versus potential (E) curves which would be obtainedpolarizing each half-cell with an auxiliary counter electroin absence of the other half-cell. For this discussion, wesume (a) no IR drop between anodic and cathodic areas(b) simple algebraic addition of the anodic and cathodic crents. Under these conditions, the rest potential or opencuit potential of the corroding sample will be that at whthe two i versus E curves intercept. At this point no net, ternal current flows, and the absolute value of current atintercept is equal to the corrosion current.

If the cathodic process is the slower process (the one the larger polarization), the corrosion rate is considered tcathodically controlled (Fig. 2.6). Conversely, if the anodicprocess is slower, the corrosion rate is said to be anodicontrolled (Fig. 2.7).

In concrete, one or two types of corrosion-rate-controllmechanisms normally dominate. One is cathodic diffuswhere the rate of oxygen diffusion through the concr


thcamplalco

prcupli

Fig. 2.6—Cathodic control

Fig. 2.7—Anodic control

de-s in

everbe ofcedif-inus

then re-od,ly to

atngent-ases

determines the rate of corrosion. In Fig. 2.8, cathodic diffusioncontrol is shown for two different rates of oxygen diffusion.

The other type of controlling mechanism involves the velopment of a high resistance path. When steel corrodeconcrete, anodic and cathodic areas may be as much as sfeet apart; therefore, the resistance of the concrete may great importance. Fig. 2.9 illustrates two cases of resistancontrol where the potential available for corrosion is the ference in the potential between anode and cathode mthe relevant IR loss.

2.2.10Stray current corrosion—Stray electric currents areose that follow paths other than the intended circuit. Theyn greatly accelerate the corrosion of reinforcing steel. Theost common sources of these are electric railways, electro-ating plants, and cathodic protection systems. Kondo et.2.9 reported corrosion of reinforcing steel embedded inncrete used in an electric railroad.2.2.11 Cathodic protection—The principle of cathodicotection is to change the potential of a metal to reduce therrent flow and thereby the rate of corrosion. This is accom-shed by the application of a protective current at a higher

Fig. 2.8—Cathodic diffusion control

Fig. 2.9—Resistance control

al

voltage than that of the anodic surface. The current flows to the original anodic surface resulting in cathodicactions occurring there. The difficulties in using this methhowever, are to determine the correct potential to appthe system and to make sure that it is applied uniformly.

e saltsty inbil-, ac-y belfide

2.2.12 Effects of other salt ions—It has been reported thsulfate2.20 and carbonate2.21 salts can also cause reinforcisteel to corrode. However, this has not been well documed. Although the concrete may have cracked in certain cand the exposed steel may have appeared rusted, thesmay not have been the primary cause. Their low solubilia high calcium ion environment would reduce their availaity. However, certain soluble salts, such as perchloratesetates, and salts of halogens other than chlorine macorrosive to steel embedded in concrete. Hydrogen suhas also been cited as a cause of corrosion.2.22

stressuced

2.2.13 Stress corrosion cracking—Stress corrosion2.23 isdefined as the process in which the damage caused by and corrosion acting together greatly exceeds that prodwhen they act separately.


sionsio

ualctil-ith

genalts nothetedrete

ent is in

glye af thewason-hen

ined

e-tial-dy,

wasnt byties

um-nity.on inspe-pos-

uchtion

d ind ce-

al- of

isith

dis-

lu-

In stressed steel, a small imperfection caused by corrocan lead to a serious loss in tensile strength as the corrocontinues at the initial anode area.

Another form of corrosion that is related to stress corrosioncracking is intergranular corrosion. In this case, a gas, usly hydrogen, is absorbed in the iron, causing a loss of duity and cracking. Hydrogen cracking in connection wcathodic protection will be discussed in Chapter 5. Other ma-terials that may cause intergranular corrosion are hydrosulfide and high concentrations of ammonia and nitrate sThe mechanism of how this type of corrosion proceeds isfully understood; however, it is believed that it involves treduction in the cohesive strength of the iron. Documenexamples of stress corrosion cracking of steel in conchave not been found in the literature.

s,hy-dater

m- the ox

oft.

tionpes of

ent de-ride per-sell

alci-s a

ith

of

her

2.3—Effects of the concrete e nvi ronment oncor rosion

2.3.1 Portland cement—When portland cement hydratethe silicates react with water to produce calcium silicate drate and calcium hydroxide. The following simplifieequations give the main reactions of portland cement with w

2(3CaO ⋅ SiO2) + 6H2O → 3CaO ⋅ 2SiO2 ⋅ 3H20 +3Ca(OH)2 (2-5)

2(2CaO ⋅ SiO2) + 4H2O → 3CaO ⋅ 2SiO2 ⋅ 3H2O + Ca(OH)2 (2-6)

As previously mentioned, the high alkalinity of the cheical environment normally present in concrete protectsembedded steel because of the formation of a protectiveide film on the steel. The integrity and protective qualitythis film depends on the alkalinity (pH) of the environmen

Differences in the types of cement are a result of variain composition or fineness or both, and as such, not all tyof cement have the same ability to provide protection

n

-

.

-

embedded steel. According to Pressler et al.2.24 a well-hy-drated portland cement may contain from 15 to 30 perccalcium hydroxide by weight of the original cement. Thisusually sufficient to maintain a solution at a pH about 13the concrete independent of moisture content.

Rosenqvist2.25 has described an example of exceedinrapid steel corrosion in a tropical concrete wharf wherpozzolan was added to the portland cement. Corrosion oreinforcement was observed shortly after construction completed. Measurements of pH in the extract from the ccrete adjacent to the steel gave values from 5.7 to 8.5. Wthe pozzolan was omitted, an increase in pH was obtaand no damage was observed.

Other research2.26 showed that a pozzolan did not adversly affect the performance of steel in concrete that was parly immersed in a saturated salt solution. In this latter stuthe pozzolan studied was a calcined volcanic tuff that added to the concrete at dosages of 16 and 32 perceweight of the cement. It may be that the protective qualiof the pozzolan are related to its source or generic type.

The use of blended cements might, under certain circstances, be detrimental because of a reduction in alkaliHowever, blended cements can give a substantial reductipermeability and also an increase in electrical resistivity ecially where a reduction in the water-cement ratio is made sible. Also, such blended cements may give concrete as mas two to five times higher resistance to chloride penetrathan concrete made with portland cements.2.27 The effectswould be beneficial as far as corrosion is concerned ansome circumstances the benefits associated with blendements more than offset the adverse effects.

Even for cements with the highest reserve basicity, thekalinity may be reduced in a number of ways. Reductionalkalinity by leaching of soluble alkaline salts with wateran obvious process. Partial neutralization by reaction wcarbon dioxide (carbonation), as present either in air or solved in water, is another common process.

The silicates are the major components in portland cemimparting strength to the matrix. No reactions have beentected between chloride ions and silicates. Calcium chloaccelerates the hydration of the silicates when at least 1cent by weight is added.2.28 Calcium chloride seems to act aan accelerator in the hydration of tricalcium silicate as was to promote the corrosion of steel.

Also present in portland cement are C3A and an alumino-ferrite phase reported as C4AF. The C3A reacts rapidly in thecement system to cause flash set unless it is retarded. Cum sulfate is used as the retarder. Calcium sulfate formcoating of ettringite (C3A ⋅ 3CaSO4 ⋅ 32H2O) around the alu-minate grains thereby retarding their reactivity (Fig. 2.10).

Calcium chloride also forms insoluble reaction products wthe aluminates in cement (see Fig. 2.11). The most commonlynoted complex is C3A ⋅ CaCl2 ⋅ xH2O, Friedel’s salt (Table 2.1).The rate of formation of this material is slower than thatettringite (compare Fig. 2.10 with Fig. 2.11). The chloride-aluminate complex forms after ettringite and prevents furtreactions of sulfate with the remaining aluminates.2.29

Fig. 2.10—Rate of reaction of gypsum with tricalcium aminate in portland cement2.29


n pur-

Table 2.1—Comparison of p roduce rs analysis ofeight different po rtland cements with total amount ofcalcium chloride that reacted with each cement 2.29

1/3 SO3 content* C3A + C4AF - 1/3SO3Amount of calcium

chloride reacted0.20 1.43 1.190.05 1.42 1.370.30 1.10 1.180.07 1.62 1.450.72 0.69 0.630.25 1.03 1.080.26 1.05 1.070.23 1.15 1.05

Average 1.18 1.13*Data expressed in molar equivalents rather than usual percent for compariso

poses. Wide differences in SO3 content obtained intentionally.

uceg toan-uthave

d inon-

ecttions codgerou

taines.t-siontureging

ofrete wiliox-

e isres in

lev-e

of

the

-sug

ith
Fig. 2.11—Rate of reaction of calcium chloride wportland cement2.29
The chemical combining of C3A with chlorides is fre-quently referred to as a beneficial effect in that it will redthe rate of chloride penetration into concrete. AccordinMehta,2.30 a chemical binding of penetrating chlorides cnot be expected unless the C3A content is higher than abo8 percent. However, recent experimental observations shown that as much as 8.6 percent C3A does not effectivelyreduce chloride penetration when concrete is immerseseawater2.31 and other studies have found more corrosiinduced distress associated with high C3A contents.2.32

2.3.2 Aggregate—The aggregate generally has little effon the corrosion of steel in concrete. There are excepThe most serious problems arise when the aggregatestain chloride salts. This can happen when sand is drefrom the sea or taken from seaside or arid locations. Poaggregates can absorb considerable quantities of salt.

Care should be exercised when using admixtures coning chloride in combination with lightweight aggregatHelms and Bowman2.33 found that reinforcing steel in lighweight concrete was particularly susceptible to corrowhen exposed to moisture and changes in temperaLightweight aggregate containing sulfides can be damato high-strength steel under stress.

entct tot sol-

ity

2.3.3 Water—The effect of moisture content or degreewater saturation on the rate of oxygen diffusion into conchas already been discussed. A high moisture contentalso substantially reduce the rate of diffusion of carbon dide and hence the rate of carbonation of the concrete.

An important effect of the moisture content of concretits effect on the electrical resistivity of the concrete. Progsive drying of initially water-saturated concrete resultsthe electrical resistivity increasing from about 7 × 103

ohm cm to about 6000 × 103 ohm cm (Fig. 2.12).2.34Field ob-servations indicate that when the resistivity exceeds a el of 50 to 70 × 103 ohm cm, steel corrosion would bnegligible.2.35 Other authors quote values of resistivity10 x 103 ohm cm2.36 and 12 × 103 ohm cm2.37 abovewhich corrosion induced damage is unlikely even in presence of chloride ions, oxygen, and moisture.

.n-ds

-

.

l

-

Fig. 2.12—Effect of water saturation on the resistivof concrete2.34

rtednoushite,

2.3.4 Corrosion inhibiting admixtures—Numerous chemical admixtures, both organic and inorganic, have been gested as specific inhibitors of steel corrosion.2.13 Some of the

-

admixtures, however, may retard time of setting of the cemor be detrimental at later ages. Many would be subjeleaching and hence less effective in concrete that has losuble material by leaching. Among those compounds repoas inorganic inhibitors are potassium dichromate, stanchloride, zinc and lead chromates, calcium hypophosp


stcanly o

ffe

sgtep

ex-lessover inum

er sat- the

f

n

retel

edies

an

sodium nitrite, and calcium nitrite. Organic inhibitors suggeed have included sodium benzoate, ethyl aniline, and mertobenzothiazole. With some inhibitors, inhibition occurs oat addition rates sufficiently high to counteract the effectschlorides. Sodium nitrite has been used with apparent etiveness in Europe.2.38 Calcium nitrite2.39 is being studied inthe United States since several side effects of the sodiumwould be avoided. Some of the side effects are low strenerratic times of setting, efflorescence, and enhanced suscbility to the alkali-aggregate reaction.2.40

-igh

ofsesioe vateizinr i

turesLon-

s

uit

2.3.5 Concrete quality—Concrete will offer more protection against corrosion of embedded steel if it is of a hquality. A low water-cement ratio will slow the diffusion chlorides, carbon dioxide, and oxygen and also the increastrength of the concrete may extend the time before corroinduced stresses cause cracking of the concrete. The porume and permeability can be reduced by lowering the wcement ratio. The type of cement or use of superplasticand mineral admixtures may also be an important factocontrolling the permeability and the ingress of chlorides.2.31

sib ses

rmmi-

e dbot resit

re

l

non-

1

teus

f

2.3.6 Thickness of concrete cover over steel—The amountof concrete cover over the steel should be as large as posconsistent with good structural design, the severity of thevice environment, and cost. The effect of cover thicknesmore than a simple linear relationship. Considering the nodiffusion of an electrolyte into a porous solid without checal reaction, a relationship such as shown in Fig. 2.13 wouldbe anticipated. However, in the case of cement paste, thfusion of chloride ions into the paste is accompanied by physical adsorption and chemical binding. These effectsduce the concentration of chloride ion at any particular and hence the tendency for inward diffusion is furtherduced. This is also illustrated in Fig. 2.13.

Fig. 2.13—Gradient of total chloride concentration depthdepends on whether chemical reaction occurs with cement2.41

n-d re n

aliton-

2.3.7 Carbonation—Carbonation occurs when the cocrete reacts with carbon dioxide from the air or water anduces the pH to about 8.5. At this low pH the steel islonger passive and corrosion may occur. For high-qu

concrete, in situations where the rate of carbonation istremely slow, carbonation is normally not a problem uncracking of the concrete has occurred or the concrete cis defective or very thin. Carbonation is not a problemvery dry concrete or in water-saturated concrete. Maximcarbonation rates are observed at about 50 percent waturation. A more complete discussion of carbonation andcorrosion of steel in carbonated concrete is given in Refer-ences 2.17 and 2.21.

-p-

fc-

alth,ti-

inn-ol-r-gn

le,r-isal

if-h-

e-

-oy

2.4—References2.1. Uhlig, Herbert H., Corrosion and Corrosion Control, 2nd Edition,

John Wiley & Sons, New York, 1971, 419 pp.2.2. Gjørv, O. E.; Vennesland, Ø; and El-Busaidy, A. H. S., “Diffusion of

Dissolved Oxygen through Concrete,” Paper No. 17, NACE Corrosion 76,National Association of Corrosion Engineers, Houston, Mar. 1976, 13 pp.

2.3. Shalon, R., and Raphael, M., “Influence of Sea Water on Corrosionof Reinforcement,” ACI JOURNAL, Proceedings V. 55, No. 8, Feb. 1959, pp.1251-1268.

2.4. Griffin, Donald F., and Henry, Robert L., “Effect of Salt in Concreteon Compressive Strength, Water Vapor Transmission, and Corrosion oReinforcing Steel,” Proceedings, ASTM, V. 63, 1963, pp. 1046-1078.

2.5. Stratfull, R. F.; Jurkovich, W. J.; and Spellman, D. L., “CorrosioTesting of Bridge Decks,” Transportation Research Record No. 539, Trans-portation Research Board, 1975, pp. 50-59.

2.6. Browne, Roger D., “Mechanisms of Corrosion of Steel in Concin Relation to Design, Inspection, and Repair of Offshore and CoastaStructures,” Performance of Concrete in Marine Environment, SP-65,American Concrete Institute, Detroit, 1980, pp. 169-204.

2.7. Aziz, M. A., and Mansur, M. A., “Deterioration of Marine ConcretStructures with Special Emphasis on Corrosion of Steel and Its Reme,”Corrosion of Reinforcement in Concrete Construction, Ellis Horwood Lim-ited, Chichester, 1983, pp. 91-99.

2.8. Gjørv, Odd E., “Durability of Concrete Structures in the OceEnvironment,” Proceedings, FIP Symposium on Concrete Sea Struc(Tbilisi, Sept. 1972), Federation Internationale de la Precontrainte, don, 1973, pp. 141-145.

2.9. Foley, R. T., “Complex Ions and Corrosion,” Journal, Electrochemi-cal Society, V. 122, No. 11, 1975, pp. 1493-1549.

2.10. Pourbaix, M., Atlas of Electrochemical Equilibrium in AqueouSolutions, Pergamon Press Limited, London, 1976.

2.11. Monfore, G. E., and Verbeck, G. J., “Corrosion of Prestressed Wire inConcrete,” ACI JOURNAL, Proceedings V. 57, No. 5, Nov. 1960, pp. 491-516.

2.12. Monfore, G. E., and Ost, Borje, “Corrosion of Aluminum Condin Concrete,” Journal, PCA Research and Development Laboratories, V. 7,No. 1, Jan. 1967, pp. 10-22.

2.13. Boyd, W. K., and Tripler, A. B., “Corrosion of Reinforcing SteeBars in Concrete,” Materials Protection, V. 7, No. 10, 1968, pp. 40-47.

2.14. Baker, E. A.; Money, K. L.; and Sanborn, C. B., “Marine CorrosioBehavior of Bare and Metallic-Coated Steel Reinforcing Rods in Ccrete,” Chloride Corrosion of Steel in Concrete, STP-629, ASTM, Phila-delphia, 1977, pp. 30-50.

2.15. Stern, M., and Geary, A. L., “Electrochemical Polarization No. Theoretical Analysis of the Shape of Polarization Curves,” Journal, Elec-trochemical Society, V. 104, No. 1, 1957, pp. 56-63.

2.16. Rosenberg, A. M., and Gaidis, J. M., “The Mechanism of NitriInhibition of Chloride Attack on Reinforcing Steel in Alkaline AqueoEnvironments,” Materials Performance, V. 18, No. 11, 1979, pp. 45-48.

2.17. “Corrosion of Reinforcement and Prestressing Tendons: A State othe Art Report,” Materials and Structures/Research and Testing (RILEM,Paris), V. 9, No. 51, May-June 1976, pp. 187-206.

2.18. Grimes, W. D.; Hartt, W. H.; and Turner, D. H., “Cracking of Con-crete in Sea Water Due to Embedded Metal Corrosion,” Corrosion, V. 35,No. 7, 1979, pp. 309-316.

2.19. Kondo, Yasuo; Takeda, Akihiko; and Hideshima, Setsuji, “Effectof Admixtures on Electrolytic Corrosion of Steel Bars in Reinforced Ccrete,” ACI JOURNAL, Proceedings V. 56, No. 4, Oct. 1959, pp. 299-312.


th

nd

-

de

h

f-

p.

of

d

and

n-

con-ided

the

rface ion

ce-

usets

es,m-so-thent

the

andionof

on-onght-

f

rt-h aod

entng

or-otss

ub-

in sidedesingon-

ini-tes,ast

ral-roic.

2.20. Eickemeyer, J., “Stress Corrosion Cracking of a High-Streng

Steel in Saturated Ca(OH)2 Solutions Caused by Cl and SO4 Additions,”Corrosion Science, V. 18, No. 4, 1978, pp. 397-400.

2.21. Hamada, M., “Neutralization (Carbonation) of Concrete aCorrosion of Reinforcing Steel,” Proceedings, 5th International Sym-posium on the Chemistry of Cement (Tokyo, 1968), Cement Association of Japan, Tokyo, 1969, V. 3, pp. 343-360.

2.22. Treadaway, K. W. J., “Corrosion of Prestressed Steel Wire inConcrete,” British Corrosion Journal (London), V. 6, 1971, pp. 66-72.

2.23. Sluijter, W. L., and Kreijger, P. C., “Potentio Dynamic Polar-ization Curves and Steel Corrosion,” Heron (Delft), V. 22, No. 1, 1977,pp. 13-27.

2.24. Weise, C. H., “Determination of the Free Calcium HydroxiContents of Hydrated Portland Cements and Calcium Silicates,” Analyt-ical Chemistry, V. 33, No. 7, June 1961, pp. 877-882. Also, ResearchDepartment Bulletin No. 127, Portland Cement Association.

2.25. Rosenqvist, I. T., “Korrosjon av Armeringsstal i Betong,”Teknisk Ukeblad (Oslo), 1961, pp. 793-795.

2.26. Spellman, D. L., and Stratfull, R. F., “Concrete Variables andCorrosion Testing,” Highway Research Record No. 423, HighwayResearch Board, 1973, pp. 27-45.

2.27. Page, C. L.; Short, N. R.; and El Tarras, A., “Diffusion of Chlo-ride Ions in Hardened Cement Pastes,” Cement and Concrete Research,V. 11, No. 3, May 1981, pp. 395-406.

2.28. Ramachandran, V. S., Calcium Chloride in Concrete, AppliedScience Publishers, London, 1976, 216 pp.

2.29. Rosenberg, Arnold M., “Study of the Mechanism througWhich Calcium Chloride Accelerates the Set of Portland Cement,” ACIJOURNAL, Proceedings V. 61, No. 10, Oct. 1964, pp. 1261-1270.

2.30. Mehta, P. K., “Effect of Cement Composition on Corrosion oReinforcing Steel in Concrete,” Chloride Corrosion of Steel in Concrete, STP-629, ASTM, Philadelphia, 1977, pp. 12-19.

2.31. Gjørv, O. E., and Vennesland, Ø., “Diffusion of Chloride Ionsfrom Seawater into Concrete,” Cement and Concrete Research, V. 9,No. 2, Mar. 1979, pp. 229-238.

2.32. Stratfull, R. F., Discussion of “Long-Time Study of CementPerformance. Chapter 12—Concrete Exposed to Sea Water and FreshWater,” by I. L. Tyler, ACI JOURNAL, Proceedings V. 56, Part 2, Sept.1960, pp. 1455-1458.

2.33. Helms, S. B., and Bowman, A. L., “Corrosion of Steel in Light-weight Concrete Specimens,” ACI JOURNAL, Proceedings V. 65, No.12, Dec. 1968, pp. 1011-1016.

2.34. Gjφrv, O. E.; Vennesland, Ø.; and El-Busaidy, A. H. S., “Elec-trical Resistivity of Concrete in the Oceans,” OTC Paper No. 2803, 9thAnnual Offshore Technology Conference, Houston, May 1977, p581-588.

2.35. Tremper, Bailey; Beaton, John L.; and Stratfull, R. F., “Causesand Repair of Deterioration to a California Bridge Due to CorrosionReinforcing Steel in a Marine Environment II: Fundamental FactorsCausing Corrosion,” Bulletin No. 182, Highway Research Board, Wash-ington, D.C., 1958, pp. 18-41.

2.36. Browne, R. D., “Design Prediction of the Life for ReinforceConcrete in Marine and Other Chloride Environments,” Durability ofBuilding Materials (Amsterdam), V. 1, 1982, pp. 113-125.

2.37. Cavalier, P. G., and Vassie, P. R., “Investigation and Repair ofReinforcement Corrosion in a Bridge Deck,” Proceedings, Institution ofCivil Engineers (London), Part 1, V. 70, 1981, pp. 461-480.

2.38. Woods, Hubert, Durability of Concrete Construction, ACIMonograph No. 4, American Concrete Institute/Iowa State UniversityPress, Detroit, 1968, p. 102.

2.39. Rosenberg, A. M.; Gaidis, J. M.; Kossivas, T. G.; and Previte,R. W., “A Corrosion Inhibitor Formulated with Calcium Nitrite for Usein Reinforced Concrete,” Chloride Corrosion of Steel in Concrete, STP-629, ASTM, Philadelphia, 1977, pp. 89-99.

2.40. Craig, R. J., and Wood, L. E., “Effectiveness of CorrosionInhibitors and Their Influence on the Physical Properties of PortlCement Mortars,” Highway Research Record No. 328, HighwayResearch Board, 1970, pp. 77-88.

2.41. Verbeck, George J., “Mechanisms of Corrosion of Steel in Cocrete,” Corrosion of Metals in Concrete, SP-49, American ConcreteInstitute, Detroit, 1975, pp. 21-38.

CHAPTER 3—PROTECTION AGAINSTCORROSION IN NEW CONSTRUCTION

3.1—IntroductionMeasures which can be taken in reinforced concrete

struction to protect the steel against corrosion can be divinto three categories:

(a) Design and construction practices that maximizeprotection afforded by the portland cement concrete.

(b) Treatments that penetrate or are applied on the suof the reinforced concrete member to exclude chloridefrom the concrete.

(c) Techniques that prevent corrosion of the reinforment directly.

In the last case, two approaches are possible: tocorrosion-resistant reinforcing steel or to nullify the effecof chloride ions on unprotected reinforcement.

3.2—Design and construction practicesThrough careful design and good construction practic

the protection provided by portland cement concrete to ebedded reinforcing steel can be optimized. It is not the phistication of the structural design that determines durability of a concrete member in a corrosive environmebut the detailing practices.3.1 The provision of adequatedrainage and a method of removing drainage water fromstructure are particularly important.

In reinforced concrete members exposed to chlorides subjected to intermittent wetting, the degree of protectagainst corrosion is determined primarily by the depth cover to the reinforcing steel and the permeability of the ccrete.3.2-3.6Estimates of the increase in corrosion protectiprovided by an increase in cover have ranged between slily more than a linear relationship3.3,3.7 to as much as thesquare of the cover.3.8 The time to spalling is a function othe ratio of cover to bar diameter,3.8 the reinforcement spac-ing, and the concrete strength. Although conventional poland cement concrete is not impermeable, concrete witvery low permeability can be made through the use of goquality materials, a minimum water-cement ratio consistwith placing requirements, good consolidation and finishipractices, and proper curing.

In concrete that is continuously submerged, the rate of crosion is controlled by the rate of oxygen diffusion that is nsignificantly affected by the concrete quality or the thickneof cover.3.9 However, as noted in Chapter 2, corrosion ofembedded steel is a rare occurrence in continuously smerged concrete structures.

Placing limits on the allowable amounts of chloride ion concrete is an issue still under active debate. On the oneare the purists who would like to see essentially no chloriin concrete. On the other are the practitioners, includthose who must produce concrete under cold weather cditions, precast concrete manufacturers who wish to mmize curing times, producers of chloride-bearing aggregaand some admixture companies, who would prefer the lerestrictive limit possible. Since chlorides are present natuly in most concrete-making materials, specifying a zechloride content for any of the mix ingredients is unrealist

222R-12 ACI COMMITTEE REPORT

enrid c

ac-sa ex to

ndure

sigcotal

tharefteodhl

isheree,xp

dentncionf th

oexheide exrede ofin

d bn-edor

nt toast 8mnt in

acided

r, inre-

bleid-lo-h-des,enedre-lo-ates

centgre-g-ere iss ined tois is fromient

06ndggest-tain) de-

sentreteix,

ntis a

ni-ation

e ofxistationoorandtorsre,

ent.dif-ix theride with insioncath-

ty ofcip-

dse in

However, it is also known that wherever chloride is presin concrete, the risk of corrosion increases as the chlocontent increases. When the chloride content exceeds atain value (termed the “chloride corrosion threshold”), unceptable corrosion may occur provided that other necesconditions, chiefly the presence of oxygen and moisture,ist to support the corrosion reactions, It is a difficult taskestablish a chloride content below which the risk of corrosion isnegligible, which is appropriate for all mix ingredients aunder all exposure conditions, and which can be measby a standard test.

Three different analytical values have been used to denate the chloride content of fresh concrete, hardened crete, or any of the concrete mixture ingredients: (a) to(b) acid-soluble, and (c) water-soluble.

The total chloride content of concrete is measured bytotal amount of chlorine. Special analytical methods necessary to determine it, and acid-soluble chloride is omistakenly called total chloride. The acid-soluble methis the test method in common use and measures that cride that is soluble in nitric acid. Water-soluble chloridechloride extractable in water under defined conditions. Tresult obtained is a function of the analytical test proceduparticularly with respect to particle size, extraction timand temperature, and to the age and environmental esure of the concrete.

It is also important to distinguish clearly between chloricontent, sodium chloride content, calcium chloride conteor any other chloride salt content. In this report, all referees to chloride content pertain to the amount of chloride (Cl-) present. Chloride contents are expressed in terms omass of cement unless stated otherwise.

Lewis3.10reported that, on the basis of polarization testssteel in saturated calcium hydroxide solution and water tracts of hydrated cement samples, corrosion occurred wthe chloride content was 0.33 percent acid-soluble chloror 0.16 percent water-soluble chloride based on a 2 hrtraction in water. However, it has been shown that the powater in many “typical” portland cement concretes mausing relatively high-alkali cements is a strong solutionsodium and potassium hydroxides with a pH approach14. Since the passivity of embedded steel is determinethe ratio of the hydroxyl concentration to the chloride cocentration,3.11 the amounts of chloride that can be toleratin concrete are higher than those that will cause pitting crosion in a saturated solution of calcium hydroxide.3.12

Work at the Federal Highway Administratiolaboratories3.5 showed that for hardened concrete subjecexternally applied chlorides, the corrosion threshold w0.20 percent acid-soluble chlorides. The average contenwater-soluble chloride in concrete was found to be 75 topercent of the content of acid-soluble chloride in the saconcrete. This corrosion threshold value was subsequeconfirmed by field studies of bridge decks including thoseCalifornia3.13 and New York.3.14 These investigations showthat, under some conditions, a chloride content of as little0.15 percent water-soluble chloride (or 0.20 percent asoluble chloride) is sufficient to initiate corrosion of embedd

teer-

ry-

d

-n-,

e

n

o-

,

o-

,-

e

f-n,--

gy

-

of0ely

s-

steel in concrete exposed to chlorides in service. Howevedetermining a limit on the chloride content of the mix ingdients, several other factors need to be considered.

As noted in the figures already given, the water-soluchloride content is not a constant proportion of the acsoluble chloride content. It varies with the amount of chride in the concrete,3.10the mix ingredients, and the test metod. All the materials used in concrete contain some chloriand, in the case of cement, the chloride content in the hardconcrete varies with cement composition. Although agggates do not usually contain significant amounts of chride,3.15 there are exceptions. There are reports of aggregwith an acid-soluble chloride content of more than 0.1 perof which less than one-third is water-soluble when the aggate is pulverized.3.16The chloride is not soluble when the agregate is placed in water over an extended period and thno difference in the corrosion performance of structuresouthern Ontario made from this aggregate when comparother chloride-free aggregates in the region. However, thnot always the case. Some aggregates, particularly thosearid areas or dredged from the sea, may contribute sufficchloride to the concrete to initiate corrosion. A limit of 0.percent acid-soluble chloride ion in the combined fine acoarse aggregate (by mass of the aggregate) has been sued with a further proviso that the concrete should not conmore than 0.4 percent chloride (by mass of the cementrived from the aggregate.3.17

There is thought to be a difference in the chloride corrosionthreshold value depending on whether the chloride is prein the mix ingredients or penetrates the hardened concfrom external sources. When chloride is added to the msome will chemically combine with the hydrating cemepaste. The amount of chloride forming chloroaluminates function of the C3A content of the cement.3.18 Chloride add-ed to the mix will also tend to be distributed relatively uformly and, therefore, has less tendency to cause the creof concentration cells.

Conversely, when chloride permeates from the surfachardened concrete, uniform chloride contents will not earound the steel because of differences in the concentrof chlorides on the concrete surface resulting from pdrainage, for example, local differences in permeability, variations in the depth of cover to the steel. All these facpromote differences in the environment (oxygen, moistuand chloride content) along a given piece of reinforcemFurthermore, most structures contain reinforcement at ferent depths, and, because of the procedures used to fsteel, the steel is electrically connected. Thus, when chlopenetrates the concrete, some of the steel is in contactchloride-contaminated concrete while other steel ischloride-free concrete. This creates a macroscopic corrocell that can possess a large driving voltage and a large ode to small anode ratio which accelerates the rate of corrosion.In seawater, it has been suggested that the permeabilithe concrete to chloride penetration is reduced by the preitation of magnesium hydroxide.3.19

In laboratory studies3.20 in which sodium chloride was addeto the mix ingredients, it was found that a substantial increa


rider

dtoila

conumwasso-e-tially.l-

.06rcer reureretesol- thehestrucion

t,heide

nor forisiofor

terbe-l reighresor-re-

reaera

singane oedht ok o

on- C

orer-alscol-meet

ayates,

-oestestto thegre- orthe

osedpos- of

and aould thets. Intruc-

chlo-n of

elec-tions val-

inu-ion-sent. in-

o-eriors or

ag-foreage.

n usetentdg-ng.dgeal-

lo-ut 5n-ced 10serva-

dif-

g at the

corrosion rate occurred between 0.4 and 0.8 percent chloalthough the moisture conditions of the test specimens wnot clearly defined. Other researchers have suggeste3.21

that the critical level of chloride in the mix ingredients initiate corrosion is 0.3 percent and that this has a simeffect to 0.4 percent chloride penetrating the hardened crete from external sources. In studies in which calcichloride was added to portland cement concrete, it found3.22 that the chloride ion concentration in the pore lution remained high during the first day of hydration. It dclined gradually, but it appeared that substanconcentrations of chloride ion remained in solution indefinite

Chloride limits in national codes vary widely. ACI 318 alows a maximum water-soluble chloride ion content of 0percent in prestressed concrete, 0.15 percent for reinfoconcrete exposed to chloride in service, 1.00 percent foinforced concrete that will be dry or protected from moistin service, and 0.30 percent for all other reinforced concconstruction. The British Code, CP 110, allows an acid-uble chloride ion content of 0.35 percent for 95 percent oftest results with no result greater than 0.50 percent. Tvalues are largely based on an examination of several stures in which it was found there was a low risk of corrosup to 0.4 percent chloride added to the mixture.3.23 However,corrosion has occurred at values less than 0.4 percen3.24

particularly where the chloride content was not uniform. TNorwegian Code, NS 3474, allows an acid-soluble chlorcontent of 0.6 percent for reinforced concrete made with mal portland cement but only 0.002 percent chloride ionprestressed concrete. Both these codes are under revOther codes have different limits, though the rationale these limits is not well established.

Corrosion of prestressing steel is generally of greaconcern than corrosion of conventional reinforcement cause of the possibility that corrosion may cause a locaduction in cross section and failure of the steel. The hstresses in the steel also render it more vulnerable to stcorrosion cracking and, where the loading is cyclic, to crosion fatigue. However, most examples of failure of pstressing steel that have been reported3.24-3.26have resultedfrom macrocell corrosion reducing the load carrying aof the steel. Nevertheless, because of the greater vulnbility and the consequences of corrosion of prestressteel, chloride limits in the mix ingredients are lower thfor conventional concrete. Based on the present statknowledge, the following chloride limits in concrete usin new construction, expressed as a percentage by weigportland cement, are recommended to minimize the rischloride-induced corrosion.

Category Chloride limit for new construction

Acid-soluble Water-solubleTest method ASTM C 1152 ASTM C 1218 Soxhlet*

Prestressed concrete 0.08 0.06 0.06Reinforced concrete

in wet conditions 0.10 0.08 0.08

Reinforced concretein dry conditions 0.20 0.15 0.15

*The Soxhlet Method of Test is described in the appendix to this report.

e,e

r-

d-

e-

-

n.

-

s-

-

f

ff

Normally concrete materials are tested for chloride ctent using either the acid-soluble test described in ASTM1152, “Acid-Soluble Chloride in Mortar and Concrete,” water-soluble test described in ASTM C 1218, “WatSoluble Chloride in Mortar and Concrete.” If the materimeet the requirements given in either of the relevant umns in the previous table they are acceptable. If they neither of the relevant limits given in the table then they mbe tested using the Soxhlet test method. Some aggregsuch as those discussed previously,3.16 contain a considerable amount of chloride that is sufficiently bound that it dnot initiate or contribute towards corrosion. The Soxhlet appears to measure only those chlorides that contribute corrosion process, thus permitting the use of some aggates that would not be allowed if only the ASTM C 1152ASTM C 1218 tests were used. If the materials fail Soxhlet test, then they are not suitable.

For prestressed and reinforced concrete that will be expto chlorides in service, it is advisable to maintain the lowest sible chloride levels in the mix to maximize the service lifethe concrete before the critical chloride content is reached high risk of corrosion develops. Consequently, chlorides shnot be intentionally added to the mix ingredients even ifchloride content in the materials is less than the stated limimany exposure conditions, such as highway and parking stures, marine environments, and industrial plants where rides are present, additional protection against corrosioembedded steel is necessary.

Since moisture and oxygen are always necessary for trochemical corrosion, there are some exposure condiwhere the chloride levels may exceed the recommendedues and corrosion will not occur. Concrete which is contously submerged in seawater rarely exhibits corrosinduced distress because there is insufficient oxygen preSimilarly, where concrete is continuously dry, such as theterior of a building, there is little risk of corrosion from chlride ions present in the hardened concrete. However, intlocations that are wetted occasionally, such as kitchenlaundry rooms or buildings constructed with lightweight gregate that is subsequently sealed (e.g., with tiles) bethe concrete dries out, are susceptible to corrosion damWhereas the designer has little control over the change iof a building or the service environment, the chloride conof the mix ingredients can be controlled. Estimates or juments of outdoor “dry” environments can be misleadiStratfull3.27 has reported the case of approximately 20 bridecks containing 2 percent calcium chloride built by the Cifornia Department of Transportation. The bridges werecated in an arid area where the annual rainfall was aboin., most of which fell at one time. Within 5 years of costruction, many were showing signs of corrosion-induspalling and most were removed from service withinyears. For these reasons, the committee believes a contive approach is necessary.

The maximum chloride limits suggested in this report fer from those suggested by ACI Committee 2013.2 andthose contained in the ACI Building Code. When makincomparison between the figures, it should be noted tha


theidesconf coes,nt

mp the

es

retetrce-ng or re-en

m-et-een

dis-n is

retes

re-on-

lowlow-sis-ter-ent

theent.) oftectpli-

mmr 90 ba-0tio

.45.ionron-otherions

con-r is

d ofsid-

l is

resd ao thef the

ce for

figures in this report refer to acid-soluble chlorides while other documents make reference to water-soluble chlorAs noted previously, Committee 222 has taken a more servative approach because of the serious consequences orosion, the conflicting data on corrosion threshold valuand the difficulty of defining the service environmethroughout the life of a structure.

Various nonferrous metals and alloys will corrode in daor wet concrete. Surface attack of aluminum occurs in

.-r-

presence of alkali-hydroxide solutions. Anodizing providno protection.

Much more serious corrosion can occur if the conccontains chloride ions, particularly if there is electrical (meal-to-metal) contact between the aluminum and steel reinfoment, because a galvanic cell is created. Serious crackisplitting of concrete over aluminum conduits has beenported.3.28,3.29 Certain organic protective coatings have berecommended3.30 when aluminum must be used and it is ipracticable to avoid contamination by chlorides. Other mals such as zinc, nickel, and cadmium, which have bevaluated for use as coatings for reinforcing steel, arecussed elsewhere in this chapter. Additional informatiocontained in Reference 3.31.

Where concrete will be exposed to chloride, the concshould be made with the lowest water-cement ratio consitentwith achieving maximum consolidation and density. The ef-fects of water-cement ratio and degree of consolidation on therate of ingress of chloride ions are shown in Fig. 3.1 and 3.2.Concrete with a water-cement ratio of 0.40 was found tosist penetration by deicing salts significantly better than ccretes with water-cement ratios of 0.50 and 0.60. A water-cement ratio is not, however, sufficient to insure permeability. As shown in Fig. 3.2, a concrete with a watercement ratio of 0.32 but with poor consolidation is less retant to chloride ion penetration than a concrete with a wacement ratio of 0.60. The combined effect of water-cemratio and depth of cover is shown in Fig. 3.3, which illus-trates the number of daily applications of salt before chloride content reached the critical value (0.20 percacid-soluble) at the various depths. Thus, 40 mm (1.5 in0.40 water-cement ratio concrete was sufficient to proembedded reinforcing steel against corrosion for 800 apcations of salt. Equivalent protection was afforded by 70 (2.75 in.) of concrete with a water-cement ratio of 0.50 omm (3.5 in.) of 0.60 water-cement ratio concrete. On thesis of this work, ACI 201.2R recommends a minimum of 5mm (2 in.) cover for bridge decks if the water-cement rais 0.40 and 65 mm (2.5 in.) if the water-cement ratio is 0Even greater cover, or the provision of additional corrosprotection treatments, may be required in some enviments. These recommendations can also be applied to reinforced concrete components exposed to chloride and intermittent wetting and drying.

Even where the recommended cover is specified, struction practices must insure that the specified coveachieved. Conversely, placing tolerances, the methoconstruction, and the level of inspection should be conered in determining the specified cover.

The role of cracks in the corrosion of reinforcing steecontroversial. Two schools of thought exist.3.32,3.33 Oneviewpoint is that cracks reduce the service life of structuby permitting more rapid penetration of carbonation anmeans of access of chloride ions, moisture, and oxygen treinforcing steel. The cracks thus accelerate the onset ocorrosion processes and, at the same time, provide spathe deposition of the corrosion products. The other viewpointis that while cracks may accelerate the onset of corrosion, such

Fig. 3.1—Effect of water-cement ratio on salt penetration3.5

Fig. 3.2—Effect of inadequate consolidation on saltpenetration3.5

Fig. 3.3—Effect of water-cement ratio and depth of coveron relative time to corrosion3.5


allyid

fent o

thein,enter-bant, foodcoacnce

idta-racntratio

beckowpenionit reankin

tingionusananor, thelin

n-es,

isfyys-se-ms

dif-ul-alore

r-ip,at

ome-tter be-ereep-calas the of-ion,

the ofde-hase

ered

thencyis-

lis-

corrosion is localized. Since the chloride ions eventupenetrate even uncracked concrete and initiate more wspread corrosion of the steel, the result is that after ayears’ service there is little difference between the amoucorrosion in cracked and uncracked concrete.

The differing viewpoints can be partly explained by fact that the effect of cracks is a function of their origwidth, intensity, and orientation. Where the crack is perpdicular to the reinforcement, the corroded length of incepted bars is likely to be no more than three diameters.3.33 Cracks that follow the line of a reinforcemebar (as might be the case with a plastic shrinkage crackexample) are much more damaging because the corrlength of bar is much greater and the resistance of the crete to spalling is reduced. Studies have shown that crless than about 0.3 mm (0.012 in.) wide have little influeon the corrosion of reinforcing steel.3.8Other investigationshave shown that there is no relationship between crack wand corrosion.3.34-3.36 Furthermore, there is no direct reltionship between surface crack width and the internal cwidth. Consequently, it has been suggested that the coof surface crack widths in design codes is not the most rnal approach.3.37

For the purposes of design, it is useful to differentiatetween controlled and uncontrolled cracks. Controlled craare those which can be reasonably predicted from a knedge of section geometry and loading. For cracking perdicular to the main reinforcement, the necessary conditfor crack control are that there be sufficient steel so that mains in the elastic state under all loading conditions, that at the time of cracking, the steel is bonded (i.e., cracmust occur after the concrete has set).

Examples of uncontrolled cracking are cracks resulfrom plastic shrinkage, settlement, or an overload conditUncontrolled cracks are frequently wide and usually caconcern, particularly if they are active. However, they cnot be dealt with by conventional design procedures, measures have to be taken to avoid their occurrence they are unavoidable, to induce them at places where are unimportant or can be conveniently dealt with, by seafor example.

shlorid ba

ar

temerla

ay de-pli-

ringsese.o-

od.te is

odaredand pri-cti-ple,r-

3.3—Methods of excluding external sources ofchloride ion from concrete

3.3.1 Waterproof membranes—Waterproof membranehave been used extensively to minimize the ingress of cride ions into concrete. Since external sources of chloions are waterborne, a barrier to water will also act as arier to any dissolved chloride ions.

The requirements for the ideal waterproofing systemstraightforward;3.38 it should:

(a) Be easy to install,(b) Have good bond to the substrate,(c) Be compatible with all the components of the sys

including the substrate, prime coat, adhesives, and ov(where used),

e-wf

-

r

redn-ks

h

kol-

-sl--s-dg

.e-dify

g

-er-

e

y

(d) Maintain impermeability to chlorides and moisture uder service conditions, especially temperature extremcrack bridging, aging, and superimposed loads.

The number of types of products manufactured to satthese requirements makes generalization difficult. Any stem of classification is arbitrary, though one of the most uful is the distinction between the preformed sheet systeand the liquid-applied materials.3.38 The preformed sheetsare manufactured under factory conditions but are often ficult to install, usually require adhesives, and are highly vnerable to the quality of the workmanship at criticlocations in the installation. Although it is more difficult tcontrol the quality of the liquid-applied systems, they aeasier to apply and tend to be less expensive.

Given the different types and quality of available wateproofing products, the differing degrees of workmanshand the wide variety of applications, it is not surprising thlaboratory3.39-3.41 and field3.14,3.42,3.43 evaluations ofmembrane performance have also been variable and stimes contradictory. Sheet systems generally perform bethan liquid-applied systems in laboratory screening testscause quality of workmanship is not a factor. Although thhas been little uniformity in both methods of test or acctance criteria, permeability (usually determined by electriresistance measurements) has generally been adopted most important criterion. However, some membranes dofer substantial resistance to chloride and moisture intruseven when pinholes or bubbles occur in the membrane.3.41

Field performance has been found to depend not only ontype of waterproofing material used, but also on the qualityworkmanship, weather conditions at the time of installation, sign details, and the service environment. Experience ranged from satisfactory3.43 to failures that have resulted in thmembrane having to be removed.3.44,3.45

Blistering, which affects both preformed sheets and liquid-applied materials, is the single greatest problem encountin applying waterproofing membranes.3.46 It is caused by theexpansion of entrapped gases, solvents, and moisture inconcrete after application of the membrane. The frequeof blisters occurring is controlled by the porosity and moture content of the concrete3.47and the atmospheric conditions.Water or water vapor is not a necessary requirement for bter formation, but is often a contributing factor. Blisters malso result from an increase in concrete temperature or acrease in atmospheric pressure during or shortly after apcation of membranes. The rapid expansion of vapors duthe application of hot-applied products sometimes caupunctures (which are termed “blowholes”) in the membran

Membranes can be placed without blisters if the atmspheric conditions are suitable during the curing periOnce cured, the adhesion of the membrane to the concreusually sufficient to resist blister formation. To insure goadhesion, the concrete surface must be carefully prepand be dry and free from curing membranes, laitance, contaminants such as oil drippings. Sealing the concreteor to applying the membrane is possible, but rarely pracal.3.48 Where the membrane is to be covered (for examwith insulation or a protective layer), the risk of blister fo


een

y inhee d con theunc

emi-

ofhe-e ap-of

er-been

mation can be reduced by minimizing the delay betwplacement of the membrane and the overlay.

Venting layers have been used in Europe, but rarelNorth America, to prevent blister formation by allowing tvapor pressures to disperse beneath the membrane. Thadvantages of using venting layers are that they requiretrolled debonding of the membrane, leakage throughmembrane is not confined to the immediate area of the pture, and they increase cost.

-ith aave) i

es o thedry-atermeoidtion

ion

ini-hemield in-

expron arstanpro-to ated

loridave

toll bee dise an

defi- con

act

ap-erlayr sev-weenrna- and then tol-

thecon- thece tole us-

firsted toface.first-ove.facethees

con-

-paxi-been

fcon-he

3.3.2Polymer impregnation—Polymer impregnation consists of filling some of the voids in hardened concrete wmonomer and polymerizing in situ. Laboratory studies hdemonstrated that polymer impregnated concrete (PICstrong, durable, and almost impermeable. The propertiPIC are largely determined by the polymer loading inconcrete. Maximum polymer loadings are achieved by ing the concrete to remove nearly all the evaporable wremoving air by vacuum techniques, saturation with a monounder pressure, and polymerizing the monomer in the vof the concrete while simultaneously preventing evaporaof the monomer.

In the initial laboratory studies on PIC, polymerizatwas accomplished by gamma radiation.3.49 However, this isnot practical for use in the field. Consequently, chemicaltiators, which decompose under the action of heat or a cical promoter, have been used exclusively in fapplications. Multifunctional monomers are often used tocrease the rate of polymerization.

The physical properties of PIC are determined by thetent to which the ideal processing conditions are commised. Since prolonged heating and vacuum saturatiodifficult to achieve, and increase processing costs subtially, most field applications have been aimed toward ducing only a surface polymer impregnation, usually depth of about 25 mm (1 in.). Such partially impregnaslabs have been found to have good resistance to chpenetration in laboratory studies, but field applications hnot always been satisfactory.3.5,3.50

There have been a few full-scale applications of PICprotect reinforcing steel against corrosion and it must sticonsidered largely an experimental process. Some of thadvantages of PIC are that the monomers are expensivthat the processing is lengthy and costly. The principal ciency identified to date has been the tendency of thecrete to crack during heat treatment.

r-been

lowxce

po-lem

t the

ingen-

icaltenttte isard-r andc-

nd-o the Thecon-ixer

3.3.3Polymer concrete overlays—Polymer concrete ovelays consisting of aggregate in a polymer binder have placed experimentally.

Most monomers have a low tolerance to moisture andtemperatures; hence the substrate must be dry and in eof 4 C (40 F). Improper mixing of the two (or more) comnents of the polymer has been a common source of probin the field. Aggregates must be dry so as not to inhibipolymerization reaction.

Workers should wear protective clothing when workwith epoxies and some other polymers because of the pottialfor skin sensitization and dermatitis.3.51Manufacturers’ rec

is--

-

ommendations for safe storage and handling of the chcals must be followed.

A bond coat of neat polymer is usually applied aheadthe polymer concrete. Blistering, which is a common pnomenon in membranes, has also caused problems in thplication of polymer concrete overlays. A number applications were reported in the 1960s.3.52,3.53 Many lastedonly a few years. More recently, experimental polymer ovlays based on a polyester-styrene monomer have placed, using heavy-duty finishing equipment to compand finish the concrete.3.54

sf

,rs

-

--e-

e

-

3.3.4 Portland cement concrete overlays—Portland ce-ment concrete overlays for new reinforced concrete areplied as part of a two-stage construction process. The ovmay be placed before the first-stage concrete has set oeral days later, in which case a bonding layer is used betthe two lifts of concrete. The advantage of the first altetive is that the overall time of construction is shortenedcosts minimized. In the second alternative, cover toreinforcing steel can be assured, and small constructioerances achieved because dead load deflections fromoverlay are very small. No matter which sequence of struction is employed, materials can be incorporated inoverlay to provide superior properties, such as resistansalt penetration and wear and skid resistance, than possibing single stage construction.

Where the second stage concrete is placed after thestage has hardened, sand or water blasting is requirremove laitance and to produce a clean, sound surResin curing compounds should not be used on the stage construction because they are difficult to remEtching with acid was once a common means of surpreparation.3.55,3.56 but is now rarely used because of possibility of contaminating the concrete with chloridand the difficulty of disposing of the runoff.

Several different types of concrete have been used ascrete overlays including conventional concrete,3.57 concretecontaining steel fibers,3.57 and internally sealed concrete.3.57,3.60 However, two types of concrete, low-slumand latex-modified concrete, each designed to offer mmum resistance to penetration by chloride ions, have used most frequently.
d
-

ss

s

3.3.5Low-slump concrete overlays—The performance olow-slump concrete is dependent solely on the use of ventional materials and good quality workmanship. Twater-cement ratio is reduced to the minimum pract(usually about 0.32) through the use of a high cement con(over 470 kg/m3 or 800 lb/yd3) and a water content sufficiento produce a slump less than 25 mm (1 in.). The concreair-entrained and a water-reducing admixture or mild reter is normally used. The use of such a high cement factolow workability mixture dictates the method of mixing, plaing, and curing the concrete.

Following preparation of the first-stage concrete, a boing agent of either mortar or cement paste is brushed intbase concrete just before the application of the overlay.base concrete is not normally prewetted. The overlay crete is mixed on site, using either a stationary paddle m


noon-reeiallyucha-

rmeitscle-urladamthendsr re the weter.pairr. pro. Ineen

butionp-ugh

ired

rt-ngeshesi-

be-restures

on-ge

ht toountsostherond toatercrete, theon,flows.ilmdoes

tex-

or a mobile continuous mixer, because truck mixers aresuited to producing either the quantity or consistency of ccrete required. The concrete must be compacted and sced to the required surface profile using equipment specdesigned to handle stiff mixtures. Such machines are mheavier and less flexible than conventional finishing mchines and have considerable vibratory capacity. The peability of the concrete to chloride ions is controlled by degree of consolidation, which is often checked with a nuar density meter as concrete placement proceeds. Wet bis placed on the concrete as soon as practicable without aging the overlay (usually within 20 min of placing), and wet curing is continued for at least 72 hr. Curing compouare not used, since not only is externally available watequired for more complete hydration of the cement, butthin overlay is susceptible to shrinkage cracking and theburlap provides a cooling effect by evaporation of the wa

Low-slump concrete was originally proposed as a rematerial for concrete pavements3.55 and was developed fopatches and overlays on bridges3.61-3.63 in the early 1960sMore recently, concrete overlays have been used as atection against reinforcing steel corrosion in new bridgesgeneral, the performance of low-slump overlays has bgood.3.64-3.66 Local bond failures have been reported, these have been ascribed to inadequate surface preparat3.64

or premature drying of the grout.3.66 The overlays are suscetible to cracking, especially on continuous structures, thothis is a characteristic of all rigid overlays.

theus-oly-ce ality,tancble

ar-aguvenienen oight

eteod-

r pratio

ausr th

ring

tha

d toost and

ineroughport-

d oc-esend it

.

entd by-sive of ateelgen,

sell

t hasre for

3.3.6Latex-modified concrete overlays—Latex-modifiedconcrete is conventional portland cement concrete withaddition of a polymeric latex emulsion. The water of spension in the emulsion hydrates the cement and the pmer provides supplementary binding properties to produconcrete having a low water-cement ratio, good durabigood bonding characteristics, and a high degree of resisto penetration by chloride ions, all of which are desiraproperties in a concrete overlay.

The latex is a colloidal dispersion of synthetic rubber pticles in water. The particles are stabilized to prevent colation, and antifoaming agents are added to preexcessive air entrapment during mixing. Styrene-butadlatexes have been used most widely. The rate of additiothe latex is approximately 15 percent latex solids by weof the cement.

The construction procedures for latex-modified concrare similar to those for low-slump concrete with minor mifications. The principal differences are:

1. The base concrete must be prewetted for at least 1 hor to placing the overlay, because the water aids penetrof the base and delays film formation of the latex.

2. A separate bonding agent is not always used, becsometimes a portion of the concrete itself is brushed ovesurface of the base.

3. The mixing equipment must have a means of stoand dispensing the latex.

4. The latex-modified concrete has a high slump so conventional finishing equipment can be used.

t

d-

-

p-

-

t

-

e

-t

f

5. Air entrainment of the concrete is believed not requfor resistance to freezing and thawing.

6. A combination of moist curing to hydrate the poland cement and air drying to develop the film formiqualities of the latex are required. Typical curing timare 24 hr wet curing, followed by 72 hr of dry curing. Tfilm-formation property of the latex is temperature sentive and film strengths develop slowly at temperatureslow 13 C (55 F). Curing periods at lower temperatumay need to be extended and application at temperaless than 7 C (45 F) is not recommended.

Hot weather causes rapid drying of the latex-modified ccrete, which makes finishing difficult and promotes shrinkacracking. Some contractors have placed overlays at nigavoid these problems. The entrapment of excessive amof air during mixing has also been a problem in the field. Mspecifications limit the total air content to 6.5 percent. Higair contents reduce the flexural, compressive, and bstrengths of the overlay. Furthermore, the permeabilitychloride ions increases significantly at air contents grethan about 9 percent. Where a texture is applied to the conas, for example, grooves to impart good skid resistancetime of application of the texture is crucial. If applied too sothe edges of the grooves collapse because the concrete If the texturing operation is delayed until after the latex fforms, the surface of the overlay tears and, since the film not reform, cracking often results.

High material costs and the superior performance of lamodified concrete in chloride penetration tests have lelatex-modified concrete overlays being thinner than mlow-slump concrete overlays. Typical thicknesses are 4050 mm (1.5 and 2 in.).

Although latex-modified overlays were first used 1957,3.67 the majority of installations were placed aft1975. Performance has been generally satisfactory, thextensive cracking and some debonding have been reed,3.68 especially in overlays 20 mm (0.75 in.) thick thatwerenot applied at the time of the original deck construction. Themost serious deficiency reported has been the widespreacurrence of shrinkage cracking in the overlays. Many of thcracks have been found not to extend through the overlay ais uncertain whether this will impair long-term performance

i-n

ee

3.4—Methods of protecting reinforcing steel fromchloride ion

The susceptibility to corrosion of mild steel reinforcemin common use is not thought to be significantly affecteits composition, grade, or the level of stress.3.69 Consequently, to prevent corrosion of the reinforcing steel in a corroenvironment, either the reinforcement must be madenoncorrosive material, or conventional reinforcing smust be coated to isolate the steel from contact with oxymoisture, and chlorides.

t

3.4.1 Noncorrosive steels—Natural weathering steelcommonly used for structural steelwork do not perform win concrete containing moisture and chloride3.2 and are notsuitable for reinforcement. Stainless steel reinforcemenbeen used in special applications, especially as hardwa


ut it inval re

areev

esuasi

ce it

enhan-ringf theeen fre-

attaching panels in precast concrete construction, bmuch too expensive to replace mild-steel reinforcemenmost applications. Stainless-steel-clad bars have been eated in the FHWA time-to-corrosion studies, and found toduce the frequency of corrosion-induced cracking compto black steel in the test slabs, but did not prevent it. Hower, it was not determined whether the cracking was the rof corrosion of the stainless steel or corrosion of the bsteel at flaws in the cladding.

ntrifi- poium

am-s thct thy eis nr inf un

s p butblebeeg o-lerhigh

rs i hav

e r no

dn-. In ob

uld ated

entydely wit

oxyxylantbeeeg-in-xy-

pre- bar

. Ac-icks ex-

e noteler-e thewasted.ts,

ingouldicallygedoxy,entere

ea of

medondatedd re-

thewith

- em-and

n in-.

arilyrites,estedults

ro-crete,

orective

3.4.2Coatings—Metallic coatings for steel reinforcemefall into two categories, sacrificial and noble or nonsaccial. In general, metals with a more negative corrosiontential (less noble) than steel, such as zinc and cadmgive sacrificial protection to the steel. If the coating is daged, a galvanic couple is formed in which the coating ianode. Noble coatings such as copper and nickel protesteel only as long as the coating is unbroken, since anposed steel is anodic to the coating. Even where steel exposed, macrocell corrosion of the coating may occuconcrete through a mechanism similar to the corrosion ocoated steel.

Nickel,3.70,3.71 cadmium,3.72 and zinc3.70,3.73,3.74have allbeen shown to be capable of delaying, and in some caseventing, the corrosion of reinforcing steel in concrete,only zinc-coated (galvanized) bars are commonly availaResults of the performance of galvanized bars have conflicting, in some cases extending the time-to-crackinlaboratory specimens,3.75 in others reducing it3.76 and sometimes giving mixed results.3.77 It is known that the zinc wilcorrode in concrete3.71,3.78 and that pitting can occur undconditions of nonuniform exposure in the presence of chloride concentrations.3.79

Field studies of galvanized bars in service for many yeaeither a marine environment or exposed to deicing saltsfailed to show any deficiencies in the concrete.3.74 However, inthese studies, chloride ion concentrations at the level of thinforcing steel were low, such that the effectiveness couldbe established conclusively. Marine studies3.80 and acceleratefield studies3.81 have shown that galvanizing will delay the oset of delaminations and spalls but will not prevent themgeneral, it appears that only a slight increase in life will betained in severe chloride environments.3.82 When galvanizedbars are used, all bars and hardware in the structure shocoated with zinc to prevent galvanic coupling between coand uncoated steel.3.82

Numerous nonmetallic coatings for steel reinforcemhave been evaluated,3.83-3.86 but only fusion-bonded epoxpowder coatings are produced commercially and wiused. The epoxy coating isolates the steel from contactoxygen, moisture, and chloride.

The process of coating the reinforcing steel with the epconsists of electrostatically applying finely divided epopowder to thoroughly cleaned and heated bars. Many poperate a continuous production line and many have constructed specifically for coating reinforcing steel. Intrity of the coating is monitored by a holiday detector stalled directly on the production line. The use of epo

s

u--d-lts

-,

ee

x-ot

-

re-

.nf

ne

e-t

coated reinforcing steel has increased substantially sinwas first used in 1973.

The chief difficulty in using epoxy-coated bars has bepreventing damage to the coating in transportation and dling. Cracking of the coating has also been observed dufabrication where there has been inadequate cleaning obar prior to coating or the thickness of the coating has boutside specified tolerances. Padded bundling bands,quent supports, and nonmetallic slings are required to vent damage during transportation. Coated tie wires andsupports are needed to prevent damage during placingcelerated time-to-corrosion studies have shown that nand cuts in the coating do not cause rapid corrosion of theposed steel and subsequent distress in the concrete.3.87 Itshould be noted, however, that the damaged bars werelectrically connected to uncoated steel in the early accated tests. Subsequent tests3.88showed that even in the casof electrical coupling to large amounts of uncoated steel,performance of damaged and nonspecification bars good, but not as good as when all the steel was coaConsequently, for long life in severe chloride environmenconsideration should be given to coating all the reinforcsteel. If only some of the steel is coated, precautions shbe taken to assure that the coated bars are not electrcoupled to large quantities of uncoated steel. A damacoating can be repaired using a two-component liquid epbut it is more satisfactory to adopt practices that prevdamage to the coating and limit touch-up only to bars whthe damage exceeds approximately 2 percent of the arthe bar.

Studies have demonstrated that epoxy-coated, deforreinforcing bars embedded in concrete can have bstrengths and creep behavior equivalent to those of uncobars.3.89,3.90Another study3.91 reported that epoxy-coatereinforcing has less slip resistance than normal mill scaleinforcing, although, for the particular specimens tested,epoxy-coated bars attained stress levels compatible ACI development requirements.

-

be

h

sn

3.5—Corrosion control methods3.5.1Chemical inhibitors—A corrosion inhibitor is an ad

mixture to the concrete used to prevent the corrosion ofbedded metal. The mechanism of inhibition is complex there is no general theory applicable to all situations.

The effectiveness of numerous chemicals as corrosiohibitors for steel in concrete3.69,3.92-3.94 has been studiedThe compound groups investigated have been primchromates, phosphates, hypophosphites, alkalies, nitand fluorides. Some of these chemicals have been suggas being effective; others have produced conflicting resin laboratory screening tests.

Many inhibitors that appear to be chemically effective pduce adverse effects on the physical properties of the consuch as a significant reduction in compressive strength. Mrecently, calcium nitrite has been reported to be an effecorrosion inhibitor3.95 and studies are continuing.3.88


wa-und

f
Admixtures used to prevent corrosion of the steel by “terproofing” the concrete, notably silicones, have been foto be ineffective.3.92
oraventlys

t incctio1.0

al,re

e

ing

etel

3.5.2 Cathodic protection—Although cathodic protectionis a viable method of protecting reinforcing steel against crosion in new construction, most installations to date hbeen to arrest corrosion in existing structures. Consequethe principles and performance of cathodic protection stems are described in Chapter 5.

It should be noted, however, that the reinforcemenmany offshore structures is connected to the cathodic protetionsystem used on the exposed steel. This results in proteof the reinforcement and current densities of 0.5 to mA/m2 (0.05 to 0.1 mA/ft2) have been reported.3.96 Thus ca-thodic protection of the reinforcement, though unintentionhas been applied in several of the largest offshore structu

-

on-

te

-

n-ut-

g

Cor

n

f,

te

d

n-.

er

,

ruc-

ts

n-.r-

arch

nnd

.

se

-

3.6—References3.1. “Durability of Concrete Bridge Decks,” NCHRP Synthesis No. 57,

Transportation Research Board Washington, D.C., 1979, 60 pp.3.2. ACI Committee 201, “Guide to Durable Concrete” (ACI 201.2R77)

(Reaffirmed 1982), American Concrete Institute, Detroit, 1977, 37 pp.3.3. Beaton, J. L., and Stratfull, R. F., “Environmental Influence on Cor

rosion of Reinforcing in Concrete Bridge Substructures,” HighwayResearch Record No. 14, Highway Research Board, 1963, pp. 60-78.

3.4. Ost, Borje, and Monfore, G. E., “Penetration of Chloride into Ccrete,” Journal, PCA Research and Development Laboratories, V. 8, No. 1,Jan. 1966, pp. 46-52.

3.5. Clear, K. C., “Time-to-Corrosion of Reinforcing Steel in ConcreSlabs, V. 3: Performance after 830 Daily Salt Applications,” ReportNo. FHWA-RD-76-70, Federal Highway Administration, Washington,D.C., 1976, 64 pp.

3.6. Stark, David, “Studies of the Relationships between Crack Patterns,Cover over Reinforcing Steel and the Development of Surface Spalls inBridge Decks,” Special Report No. 116, Highway Research Board, Wash-ington, D.C., 1971, pp. 13-21. Also, Research and Development BulletinNo. RD020E, Portland Cement Association.

3.7. Cornet, I., “Protection with Mortar Coatings,” Materials Protection,V. 6, No. 3, 1967, pp. 56-58.

3.8. Atimay, E., and Ferguson, P. M., “Early Chloride Corrosion of Reinforced Concrete—A Test Report,” Materials Performance, V. 13, No. 12,1974, pp. 18-21.

3.9. Gjørv, O. E.; Vennesland, Ø.; and El-Busaidy, A. H. S., “Diffusion ofDissolved Oxygen through Concrete,” Paper No. 17, NACE Corrosion 76,National Association of Corrosion Engineers, Houston, Mar. 1976, 13 pp.

3.10. Lewis, D. A., “Some Aspects of the Corrosion of Steel in Cocrete,” Proceedings, 1st International Congress on Metallic Corrosion, Bterworths, London, 1962, pp. 547-552.

3.11. Gouda, V. K., “Corrosion and Corrosion Inhibition of ReinforcinSteel I: Immersed in Alkaline Solutions,” British Corrosion Journal (Lon-don), V. 5, 1970, pp. 198-203.

3.12. Arup. H., “Recent Progress Concerning Electrochemistry and rosion of Steel in Concrete,” ARBEM Symposium, Paris, Oct. 1982.

3.13. Stratfull, R. F.; Jurkovich, W. J.; and Spellman. D. L., “CorrosioTesting of Bridge Decks,” Transportation Research Record No. 539, Trans-portation Research Board, 1975, pp. 50-59.

3.14. Chamberlin, William P.; Irwin, Richard J.; and Amsler, Duane E.,“Waterproofing Membranes for Bridge Deck Rehabilitation,” ResearchReport No. 52 (FHWA-NY-77-59-1), New York State Department oTransportation/Federal Highway Administration, Washington, D.C., 197743 pp.

3.15. Hime, William D., and Erlin, Bernard, “Chloride Free Concrete,”ACI JOURNAL, Proceedings V. 74, No. 10, Oct. 1977, p. N7.

3.16. Rogers. C., and Woda, G., “The Chloride Ion Content of ConcreAggregate from Southern Ontario,” Report No. EM-17, Ontario Ministryof Transportation and Communications, Downsview, 1977.

3.17. “Impurities in Aggregates for Concrete,” Advisory Note No. 18,Cement and Concrete Association, London, 1970, 8 pp.

-

y,-

n

s.

-

3.18. Roberts, M. H., “Effect of Calcium Chloride on the Durability oPretensioned Wire in Prestressed Concrete.” Magazine of ConcreteResearch (London), V. 14, No. 42, Nov. 1962, pp. 143-154.

3.19. Haynes, Harvey H., “Permeability of Concrete in Sea Water,” Per-formance of Concrete in Marine Environment, SP-65, American ConcretInstitute, Detroit. 1980, pp. 21-38.

3.20. Locke, C. E., and Siman, A., “Electrochemistry of ReinforcSteel in Salt-Contaminated Concrete,” Corrosion of Reinforcing Steel inConcrete, STP-713, ASTM, Philadelphia, 1980, pp. 3-16.

3.21. Browne, Roger D., “Mechanisms of Corrosion of Steel in Concrin Relation to Design, Inspection, and Repair of Offshore and CoastaStructures,” Performance of Concrete in Marine Environment, SP-65,American Concrete Institute, Detroit, 1980, pp. 169-204.

3.22. Diamond, S., and Lopez-Flores, F., “Fate of Calcium Chloride Dis-solved in Concrete Mix Water,” Journal, American Ceramic Society, V. 64,No. 11, Nov. 1981, pp. C 162-164.

3.23. “The Role of Calcium Chloride in Concrete,” Concrete Construction,V. 21, No. 2, Feb. 1976, pp. 57-61.

3.24. Cornet, I., “Corrosion of Prestressed Concrete Tanks,” MaterialsProtection, V. 3, No. 1, Jan. 1964, pp. 9-100.

3.25. Peterson, Carl A., “Survey of Parking Structure Deterioration anDistress,” Concrete International: Design & Construction, V. 2, No. 3,Mar. 1980. pp. 53-61.

3.26. Monfore, G. E., and Verbeck, G. J., “Corrosion of Prestressed Wire inConcrete,” ACI JOURNAL, Proceedings V. 57, No. 5, Nov. 1960, pp. 491-516.

3.27. A Manual for the Corrosion Control of Bridge Decks, ContractNo. DTFH-61-C-00016, Stratfull, R. F., ed., U.S. Department of Trans-portation, Washington, D.C.

3.28. “Conduit in Concrete,” Engineering Journal (Montreal), V. 38,No. 10, 1955, pp. 1357-1362.

3.29. “Spalled Concrete Traced to Conduit,” Engineering News-Record,V. 172, No. 11, Mar. 12, 1964, pp. 28-29.

3.30. McGeary, Frank L., “Performance of Aluminum in Concrete Cotaining Chlorides,” ACI JOURNAL, Proceedings V. 63, No. 2, Feb, 1966. pp247-266.

3.31. Erlin, B., and Woods, H., “Corrosion of Embedded Materials Oththan Reinforcing Steel,” Significance of Tests and Properties of Concreteand Concrete-Making Materials, STP-169B, ASTM, Philadelphia, 1978pp. 300-319.

3.32. Manning, D. G., “Corrosion Resistant Design of Concrete Sttures,” Proceedings, Canadian Structural Concrete Conference, Universityof Toronto, 1981, pp. 199-223.

3.33. Beeby, A. W., “Corrosion of Reinforcing Steel in Concrete in IRelation to Cracking,” The Structural Engineer (London), V. 56A, No. 3,1978, pp. 77-81.

3.34. Tremper, Bailey, “The Corrosion of Reinforcing Steel in Cracked Cocrete,” ACI JOURNAL, Proceedings V. 43, No. 10, June 1947, pp. 1137-1144

3.35. Martin, H., and Schiessel, P., “The Influence of Cracks on the Corosion of Steel in Concrete,” Preliminary Report, RILEM InternationalSymposium on the Durability of Concrete, Prague, 1969, V. 2.

3.36. Raphael, M., and Shalon, R., “A Study of the Influence of Climateon Corrosion of Reinforcement,” Proceedings, RILEM Symposium onConcrete and Reinforced Concrete in Hot Countries, Building ReseStation, Haifa, 1971, pp. 77-96.

3.37. Beeby, A. W., “Concrete in the Oceans—Cracking and Corrosio,”Technical Report No. 1, CIRIA/UEG, Construction Industry Research aInformation Association/Department of Energy, London, 1978.

3.38. Manning, David G., and Ryell, John, “Durable Bridge Decks,”Report No. RR203, Ontario Ministry of Transportation and Communications,Downsview, 1976, 67 pp.

3.39. Boulware, R. L., and Elliott, A. L., “California Seals Salt-DamagedBridge Decks,” Civil Engineering-ASCE, V. 41, No. 10, Oct. 1971, pp. 42-44

3.40. Van Til, C. J.; Carr, B. J.; and Vallerga, B. A., “Waterproof Mem-branes for Protection of Concrete Bridge Decks: Laboratory Pha,”NCHRP Report No. 165, Transportation Research Board, Washington,D.C., 1976, 70 pp.

3.41. Frascoia, R. I., “Vermont’s Experience with Bridge Deck Protective Systems,” Chloride Corrosion of Steel in Concrete, STP-629, ASTM,Philadelphia, 1977, pp. 69-81.

3.42. Spellman, D. L., and Stratfull, R. F., “Bridge Deck Membranes—Evaluation and Use in California,” Report No. CA-DOT-TL-5116-9 73-38,California Department of Transportation, Sacramento, 1973.


ems-

ck,.

te-d

y,”

en

ical

ique

n,

es

nd

on

on

wit6,

on-g-

on

onndi

.te,”h-

ete

, 73

ort

ent

ge of

ith

ete

es-r-

d

ys,t

on-i-

.,

ro-on-

ing

of

ized

ized

ng

42

a-

.sess,

um

arty,

dge

n-

etal

llic

on-i-

ndrc-

ept.on,

rems,”h-

ein-,

3.43. Frascoia, R. I., “Evaluation of Bridge Deck Membrane Systand Membrane Evaluation Procedures,”Report No. 77-2, Vermont Department of Highways, Montpelier, 1977.

3.44. Corkill, J. T., “A Field Study of the Performance of Bridge DeWaterproofing Systems in Ontario,”Proceedings, 1975 Annual ConferenceRoads and Transportation Association of Canada, 1975, V. 2, pp. 79-100

3.45. “Membrane Waterproofing for Bridge Decks,”Final Report, Ore-gon Department of Transportation, Salem, 1977.

3.46. MacDonald, M. D., “Waterproofing Concrete Bridge Decks: Marials and Methods,”TRRL Report No. LR636, Transport and RoaResearch Laboratory, Crowthorne, Berkshire, 1974, 32 pp.

3.47. Legvold, T. L., “Bridge Deck Waterproofing Membrane StudReport No. R-262, Iowa State Highway Commission, Ames, 1974.

3.48. Meader, A. L., Jr.; Schmitz, C. G.; and Henry, J. E., “Developmof a Cold Poured Bridge Deck Membrane System,”Chloride Corrosion ofSteel in Concrete, STP-629, ASTM, Philadelphia, 1977, pp. 164-177.

3.49. Steinberg, M., et al., “Concrete-Polymer Materials—First TopReport,” BNL 50134 (T-509) andUSBR General Report No. 41, BrookhavenNational Laboratory/U.S. Bureau of Reclamation, Denver, 1968.

3.50. Smoak, W. G., “Development and Field Evaluation of a Technfor Polymer Impregnation of New Concrete Bridge Deck Surfaces,”ReportNo. FHWA-RD-76-95, Federal Highway Administration, WashingtoD.C., 1976.

3.51. Tremper, Bailey, “Repair of Damaged Concrete with Epoxy Rins,” ACI JOURNAL, Proceedings V. 57, No. 2, Aug. 1960, pp. 173-182.

3.52. McConnell, W. R., “Epoxy Surface Treatments for PortlaCement Concrete Pavements,”Epoxies with Concrete, SP-21, AmericanConcrete Institute, Detroit, 1968, pp. 9-17.

3.53. Santucci, L. E., “Polyester Overlays for Portland Cement Ccrete Surfaces,”Highway Research Record No. 14, Highway ResearchBoard, 1963, pp. 44-56.

3.54. Jenkins, J. C.; Beecroft, G. W.; and Quinn, W. J., “Polymer Ccrete Overlays: Interim Users Manual,”ReportNo. FHWA-TS-78-218,Federal Highway Administration, Washington, D.C., 1977.

3.55. Felt, Earl J., “Resurfacing and Patching Concrete PavementsBonded Concrete,”Proceedings, Highway Research Board, V. 35, 195pp. 444-469.

3.56. Westall, William G., “Bonded Resurfacing and Repairs of Ccrete Pavements,”Bulletin No. 260, Highway Research Board, Washinton, D.C., 1960, pp. 14-24.

3.57. McKeel, W. T., Jr., and Tyson, S. S., “Two-Course Bonded Cstruction and Overlays,” ACI JOURNAL, Proceedings V. 72, No. 12, Dec.1975, pp. 708-713.

3.58. Tyson, S. S., “Two-Course Bonded Concrete Bridge Deck Cstruction—Interim Report No. 2: Concrete Properties and Deck Cotion Prior to Opening to Traffic,”Report No. VHTRC-R3, VirginiaHighway and Transportation Research Council, Charlottesville, 1976

3.59. Jenkins, G. H., and Butler, J. M., “Internally Sealed ConcreReportNo. FHWA-RD-75-20, Federal Highway Administration, Wasington, D.C., 1975, 106 pp.

3.60. Clear, K. C., and Forster, S. W., “Internally Sealed ConcrMaterial Characterization and Heat Treating Studies,”ReportNo. FHWA-RD-77-16, Federal Highway Administration, Washington, D.C., 1977pp.

3.61. Hilton, N., “A Two-Inch Bonded Concrete Overlay for the PMann Bridge,”Engineering Journal (Montreal), May 1964, pp. 39-44.

3.62. O’Connor, E. J., “Iowa Method of Partial-Depth Portland CemResurfacing of Bridge Decks,”Chloride Corrosion of Steel in Concrete,STP-629, ASTM, Philadelphia, 1977, pp. 116-123.

3.63. Bukovatz, J. E.; Crumpton, C. F.; and Worley, H. E., “BridDeck Deterioration Study, Final Report,” State Highway CommissionKansas, Topeka, 1973.

3.64. Manning, D. G., and Owens, M. J., “Ontario’s Experience wConcrete Overlays for Bridge Decks,”RTAC Forum(Ottawa), V. 2, No.1, 1979, pp. 31-37.

3.65. Bergen, J. V., and Brown, B. C., “An Evaluation of ConcrBridge Deck Resurfacing in Iowa,”Special Report, Iowa State HighwayCommission, Ames, 1975.

3.66. Tracy, R. G., “Bridge Deck Deterioration and Restoration Invtigation No. 639,”Interim Report, Minnesota Department of Transpotation, St. Paul, Dec. 1976.

t

-

-

-

h

-

--

:

3.67. Cardone, S. M.; Brown, M. G.; and Hill, A. A., “Latex-ModifieMortar in the Restoration of Bridge Structures,”Bulletin No. 260, High-way Research Board, Washington, D.C., 1960, pp. 1-13.

3.68. Bishara, A. G., “Latex Modified Concrete Bridge Deck OverlaField Performance Analysis,”Report No. ODOT 2895, Ohio Departmenof Transportation, Columbus, 1978.

3.69. Verbeck, George J., “Mechanisms of Corrosion of Steel in Ccrete,”Corrosion of Metals in Concrete, SP-49, American Concrete Insttute, Detroit, 1975, pp. 21-38.

3.70. Tripler, Arch B.; White, Earl L.; Haynie, F. H.; and Boyd, W. K“Methods of Reducing Corrosion of Reinforcing Steel,”NCHRP ReportNo. 23, Highway Research Board, Washington, D.C., 1966, 22 pp.

3.71. Baker, E. A.; Money, K. L.; and Sanborn, C. B., “Marine Corsion Behavior of Bare and Metallic-Coated Reinforcing Rods in Ccrete,” Chloride Corrosion of Steel in Concrete, STP-692, ASTM,Philadelphia, 1977, pp. 30-50.

3.72. Bird, C. E., and Strauss, F. J., “Metallic Coating for ReinforcSteel,”Materials Protection, V. 6, July 1967, pp. 48-52.

3.73. Cornet, I.; Ishikawa, T.; and Bresler, B., “The MechanismSteel Corrosion in Concrete Structures,”Materials Protection, V. 7, No.3, Mar. 1968, pp. 44-47.

3.74. Cook, A. R., and Radtke, S. F., “Recent Research on GalvanSteel for Reinforcement of Concrete,” Chloride Corrosion of Steel inConcrete, STP-629, ASTM, Philadelphia, 1977, pp. 51-60.

3.75. Cornet, I., and Bresler, B., “Corrosion of Steel and GalvanSteel in Concrete,”Materials Protection, V. 5, Apr. 1966, pp. 69-72.

3.76. Griffin, Donald F., “Effectiveness of Zinc Coating on ReinforciSteel in Concrete Exposed to a Marine Environment,”Technical Note No.N-1032, Naval Civil Engineering Laboratory, Port Hueneme, 1969,pp.

3.77. Hill, George A.; Spellman, D. L.; and Stratfull, R. F., “Labortory Corrosion Tests of Galvanized Steel in Concrete,”TransportationResearch RecordNo. 604, Transportation Research Board, 1976, pp. 25-30

3.78. Pourbaix, M., Atlas of Electrochemical Equilibria in AqueouSolutions, translated from the French by J. A. Franklin, Pergamon PrNew York, 1966, pp. 409-410.

3.79. Unz, M., “Performance of Galvanized Reinforcement in CalciHydroxide Solution,” ACI JOURNAL, Proceedings V. 75, No. 3, Mar.1978, pp. 91-99.

3.80. Sopler, B., “Corrosion of Reinforcement in Concrete—PSeries D,”Report No. FCB 73-4, Norwegian Institute of TechnologUniversity of Trondheim, 1973.

3.81. Arnold, C. J., “Galvanized Steel Reinforced Concrete BriDecks: Progress Report,” Report No. FHWA-MI-78-R1033, FederalHighway Administration, Washington, D.C., 1976.

3.82. Clear, K. C., “Time-to-Corrosion of Reinforcing Steel in Cocrete Slabs, V. 4: Galvanized Reinforcing Steel,”Report No. FHWA-RD-82-028, Federal Highway Administration, Washington, D.C., 1981.

3.83. Castleberry, J. R., “Corrosion Prevention for Concrete and MReinforcing in the Construction Industry,”Materials Protection, V. 7, Mar.1968, pp. 21-28.

3.84. Clifton, J. R.; Beeghley, H. F.; and Mathey, R. G., “NonmetaCoatings for Concrete Reinforcing Bars,”ReportNo. FHWA-RD-74-18,Federal Highway Administration, Washington, D.C., 1974, 87 pp.

3.85. Backstrom, T. E., “Use of Coatings on Steel Embedded in Ccrete,”Corrosion of Metals in Concrete, SP-49, American Concrete Insttute, Detroit, 1975, pp. 103-113.

3.86. Pike, R. G.; Hay, R. E.; Clifton, J. R.; Beeghly, H. F.; aMathey, R. G., “Nonmetallic Protective Coatings for Concrete Reinfoing Steel,” Transportation Research Record No. 500, TransportationResearch Board, 1974, pp. 36-44.

3.87. Clear, K. C., “FCP Annual Progress Report—Year Ending S30, 1978, Project 4B,” Federal Highway Administration, WashingtD.C., 1978.

3.88. Virmani, Y. P.; Clear, K. C.; and Pasko, T. J., Jr., “Time-to-Corrosionof Reinforcing Steel in Concrete Slabs, V. 5: Calcium Nitrite Admixtuor Epoxy-Coated Reinforcing Bars as Corrosion Protection SysteReport No. FHWA-RD-83-012, Federal Highway Administration, Wasington, D.C., 1983, 71 pp.

3.89. Mathey, Robert G., and Clifton, James R., “Bond of Coated Rforcing Bars in Concrete,”Proceedings, ASCE, V. 102, ST1, Jan. 1976pp. 215-229.


D.,

oxy-

ingtion-

ibi-mend,

e,”e,

, R.in

ofte in

or-retee

gnt of

ansis- in amag- bart arenre-ich a

d tocingingerio-

3.90. Clifton, James R.; Mathey, Robert G.; and Anderson, Erik“Creep of Coated Reinforcing Bars in Concrete,”Proceedings, ASCE,V. 105, ST10, Oct. 1979, pp. 1935-1947.

3.91. Johnston, D. W., and Zia, P., “Bond Characteristics of EpCoated Reinforcing Bars,”Report No. FHWA-NC-82-002, Federal Highway Administration, Washington, D.C., 1982.

3.92. Clear, K. C., and Hay, R. E., “Time-to-Corrosion of ReinforcSteel in Concrete Slabs, V. 1: Effect of Mix Design and ConstrucParameters,” Report No. FHWA-RD-73-32, Federal Highway Administration, Washington, D.C., 1973, 103 pp.

3.93. Craig, R. J., and Wood, L. E., “Effectiveness of Corrosion Inhtors and Their Influence on the Physical Properties of Portland CeMortars,” Highway Research RecordNo. 328, Highway Research Boar1970, pp. 77-88.

3.94. Griffin, D. F., “Corrosion Inhibitors for Reinforced ConcretCorrosion of Metals in Concrete, SP-49, American Concrete InstitutDetroit, 1975, pp. 95-102.

3.95. Rosenberg, A. M.; Gaidis, J. M.; Kossivas, T. G.; and PreviteW., “A Corrosion Inhibitor Formulated with Calcium Nitrite for Use Reinforced Concrete,”Chloride Corrosion of Steel in Concrete, STP-629,ASTM, Philadelphia, 1977, pp. 89-99.

3.96. Fidjestol, P.; Askheim, N. E.; and Roland, B., “Location Potential Corrosion Areas in Concrete Marine Structures,” Concrethe Oceans, Phase II, Plc, Final Report.

oron thmso altiocifhism-eri-allly

e. I for

edret

it iruc

kov

rivemi-nlo-ing fo

s inr are

tctureange so-

lam- au-s ofg ga-red.ting

byellon-.l is

retents

air:h

.osseliedayed

y ating

CHAPTER 4—PROCEDURES FOR IDENTIFYINGCORROSIVE ENVIRONMENTS AND ACTIVE

CORROSION IN CONCRETE

4.1—IntroductionThere are many approaches available for identifying c

rosive environments and active corrosion of steel in ccrete. Generally, a visual inspection of the structure andenvironment in which it serves is the first step in any exaination. Visual inspections may range from a simple curry examination to those that are very detailed, whereincracks and other visual evidences of physical deterioraare plotted on scaled diagrams of the structure and speinformation is gathered on environmental exposure. Ttype of inspection may also include taking a limited nuber of cores to be examined visually for evidence of detoration due to corrosion. The detailed type of visuinspection is time-consuming and costly, and generaonly useful for research studies of structure performancdoes not develop the type of information that is requiredscheduling of maintenance.

There are several techniques and tools that can be usmore specifically delineate areas of deteriorated concand areas of potential or active corrosion of steel.4.1-4.6 How-ever, an inspection program can become expensive if necessary to survey more than a nominal number of sttures. For purposes of planning a maintenance or rehabilitationprogram, techniques such as suggested by Stratfull, Jurich, and Spellman,4,5 or Manning,4.7,4.8 should be used tominimize expenses. These techniques have been defrom experience and include judicious use of visual exanations together with collection of specific information othe extent of physical deterioration, active corrosion, chride ion contamination, and depth of cover over reinforcsteel. The references noted previously should be studiedmore detailed information on these techniques.

4.2—Methods of evaluationCertain tools are used for identifying and quantifying c

rosive environments, extent of active corrosion, and concdeterioration.4.9 Following is a brief description of thestools, together with their purpose and limitations.

t

4.2.1 Pachometer—This tool is used to locate reinforcinsteel embedded in concrete, and to determine the amoucover over the steel. It is battery-operated and contains a trtorized oscillator that establishes an elecromagnetic fieldsearch coil. In the presence of a steel reinforcing bar, the netic field is distorted. By calibration, the distance from themay be read from the meter dial. Two styles of equipmenavailable. The first is a handheld device4.10 and the second is aautomated device4.11 mounted on wheels. Automatic data cording equipment is added to faciliate the speed with whsurvey can be conducted.

The knowledge of cover depth is essential if it is desireobtain samples of the concrete at the level of the reinforsteel for chloride ion analysis. It is also useful in determinthe potential for corrosion and subsequent concrete detration since it has been well established that structurecorrosive environments with inadequate concrete covesubject to early deterioration.

--e--lnic

4.2.2Delamination detectors—There are many tools thamay be used to detect delaminations or subsurface fraplanes parallel to the concrete surface. These devices rfrom simple chain drags or lightweight hammers to morephisticated devices such as the Delamtect.4.3,4.9 Almost anysounding device can be used to locate hollow areas or deinations caused by corrosion of the reinforcing steel. Thetomated Delamtect is useful for surveying large numberbridge decks or other horizontal surfaces such as parkinrage floors if a record of the area of delamination is desiHowever, the simpler chain drag is adequate for locadelaminated areas during repair operations.

t

toe

s-

-

d

r

4.2.3 Electrical potential measuring equipment—Thisequipment, adapted for use with reinforced concrete Stratfull,4,2 consists of a copper-copper sulfate half-c(CSE), a high-impedance voltmeter, and lead wires to cnect the half-cell and the reinforcing steel to the voltmeter4.9

The equipment is used to determine if the reinforcing steein a passive or active state relative to active corrosion.4.2 De-tails of the test method are given in ASTM C 876.

The accuracy of the method is good when proper concprewetting is used. The significance of the measuremecan be summarized as follows for structures exposed to

Potentials more negative than -0.35 V CSE: Very higprobability of the presence of active corrosion.Potentials more positive than -0.20 V CSE: Very highprobability of no corrosion.Potentials in the range of -0.20 V to -0.35 V CSE:Uncertainty as to whether or not corrosion is present

It is in the uncertain range that potential differences acra structure, and other detection methods, must often be ron to deduce the probable condition. In Federal HighwAdministration studies, potential differences rarely exce100 mV when corrosion was not active, or was active onlextremely low rates. In reinforced concrete undergo


om-

ativith

) geencl dif-ely re ohlyatese. Indingsults verquitselytion po-s.

ble,mayub-ndi-er,the

videuseteelion

ance ofrical the haveigh-d

significant corrosion, the potential differences were cmonly over 200 mV.4.12,4.13

It should also be noted that a potential value more negthan -0.35 V CSE on actively corroding field structures (wmany electrically interconnected anodes and cathodeserally provides information on only the presence or absof corrosion and not on corrosion rate. It is the potentiaference between the anode and cathode that most closlates to corrosion rate rather than simply the magnitudthe anode potential. A common example in which hignegative potentials are not indicative of high corrosion ris a totally water-saturated reinforced concrete structursuch a structure, oxygen availability to the noncorrosteel is severely restricted and cathodic polarization reThis drives both the anode and the cathode potentials tonegative values, and yet corrosion rate is most often low. By careful measurement of potentials on a clospaced grid pattern, high versus low corrosion rate situacan be identified by studying potential differences. Largetential differences generally indicate high corrosion rate

n arecivech-e tine

enshloubl.

rideoubleent(80 antheandsinn i

enssees reretr Foatebe

cre

ingr coll a be

thanigh-inh asoistteel forrast,l innlytedess,hee, onorc-cateheflow

ivetain-obes in-ntialarac-ed

spe-se ofct ofided

resis-ctriclate,

theetted

4.2.4 Chloride analysis—Measurements of chloride ioconcentration at the level of reinforcing steel in concretemade to determine if any environment exists that is conduto corrosion of the steel. Two wet chemical analysis teniques are used to isolate chloride from the concrete, ondetermine acid-soluble chloride and the other to determwater-soluble chloride. As discussed in Chapter 3, the chlorideion content measured by the water-soluble test is very stive to the test procedures. Consequently, all values of cride ion content in this report are referenced to the acid-soltest described in ASTM C 114 and also in AASHTO T 260

The preferred method of sample procurement for chlomeasurement is to obtain concrete in powdered form withthe aid of liquid coolants that could leach out water-soluchloride. This can be done by using impact drilling equipmand collecting the pulverized material. Alternatively, a 3-in. mm) diameter or larger core can be obtained by wet coringthen extracting samples from the interior by dry sawing or opulverizing methods. Measurements for acid-soluble water-soluble chloride can be made on this type of sample uthe standard test procedures. Additional guidance is giveReferences 4.14 and 4.15.

In many existing concrete structures, the exact cemcontent is not known. Thus, chloride levels can be exprein terms of percent by weight of concrete, or, sometimpounds of chloride per cubic yard of concrete. The latterquires an assumed or measured unit weight of the concA table in Reference 4.7 gives the conversion formulas fothe various methods of expressing chloride in concrete.greater precision, nonevaporable water contents (wchemically combined through cement hydration) can measured on each powdered sample from a given conand used as correction factors for aggregate induced distortionsin measured chloride levels.

In any interpretation of chloride data, sound engineerjudgment must be used to assess the actual potential forosion. As stated earlier, free moisture and oxygen as wechloride must be available to induce corrosion. If it can

e

n-e

concluded that either moisture or oxygen is not availathere would be no corrosion threshold. Such conditions prevail, for example, in concrete that is continuously smerged or in internal members in buildings where air cotioning units maintain constantly low humidities. Howevthe difficulty of assessing the possibility of corrosion in service environment is discussed more fully in Chapter 3.
e-f
.y

e

s

o

i--e

t

dr

gn

td,-e.

rr

4.2.5 Rate of corrosion probes—Two basic types ofprobes are available for embedment into concrete to proan indication of rate of corrosion. One type involves the of two or three electrically isolated short sections of swire or reinforcing steel and the use of linear polarizattechniques to estimate instantaneous corrosion rates.4.17 Thesecond, more widely used device is the electrical resistrate-of-corrosion probe which provides cumulative ratecorrosion data from periodic measurements of the electresistance of a steel wire or hollow cylinder embedded inconcrete. Experiences to date with use of these probesbeen conflicting. However, based on recent Federal Hway Administration studies,4.12 this appears to be relatemore to the rate of corrosion process of steel in concreteto inaccuracies of the devices themselves. The Federal Hway Administration studies indicate that current flow withphysically separated, macroscopic corrosion cells, sucthe case of large quantities of steel in chloride-free mconcrete in close proximity and electrically coupled to sin chloride-bearing concrete, are primarily responsiblethe very high rates of corrosion on bridge decks. In contmicroscopic self-corrosion of a small section of steechloride-contaminated concrete most often results in orelatively low corrosion rates. Since the electrically isolalinear polarization devices only simulate this latter procvalid predictions of the overall effect of corrosion on tstructure are not possible. The electrical resistance probthe other hand, can be electrically coupled with the reinfing steel in the structure and thus, potentially, can indimacrocell activity. However, this is possible only if tprobe becomes the anode of a macrocell and current within this macrocell is typical of that of a macrocell acton the reinforcing steel. These are very important uncerties and have generally limited the use of resistance prto research and field evaluation efforts in which specialstallation procedures are required, and electrical poteand current measurements can be made to define the chteristics of the probe-reinforcing bar-macrocell. Continustudy of these devices is needed, as well as lower cost optionssuch as short sections of reinforcing steel installed in a cific manner, and current flow measurements. To date, usuch procedures in the field to aid in studies on the efferehabilitation procedures on corrosion rate have provexcellent results.

te

r-s

4.2.6 Electrical resistance measuring equipment—Theprimary use of these measurements is to determine the tance of waterproof membranes that are made from dielematerials. This equipment consists of a copper contact psponges, ohmmeter, and lead wire.4.1 One terminal of theohmmeter is connected to the reinforcing steel, whileother terminal is connected to the copper plate. The w


o fad a

int.n thro-eny a

whe

ofon-e reters for

ofenv de-ly to

usech-

heirs.resingnmen theobes

d

te

ent

n

ge

, con-edetely

ent-teelcom-oat-ansur-com-

sponges are fastened to the bottom of the copper plate tcilitate contact. The electrical resistance can be measureany point by moving the copper contact plate to that poThe resistance has been related to the number of holes imembrane and its permeability to water. This technique pvides a nondestructive method to evaluate the environmand potential for corrosion in concrete decks protected bmembrane system. Advantages and cautions needed using these measurements are discussed in Reference 4.7.The technique has not proven satisfactory for evaluationepoxy resin coatings on reinforcing steel embedded in ccrete. Some work has been done on measurement of thsistance of these coatings using alternating current meHowever, the state of the art is not sufficiently advancedgeneral use as an evaluation technique.

Other techniques and tools to determine the conditionthe concrete and to detect the presence of a corrosive ronment or active corrosion have been studied to varyinggrees. Ultrasonic methods have been used successfulinvestigate the quality and condition of concrete.

To a lesser extent, infrared and radar scans have beento record the condition of concrete. Trial uses of these teniques on bridge decks have been encouraging,4.18 but moredevelopment work is needed, especially to determine tvalue in assessing the condition of asphalt-covered deck

A study is underway to develop nondestructive procedufor direct measurement of the rate of corrosion of reinforcsteel in concrete.4.12 Three electrode-linear-polarizatiomeasurements are encouraging. Because the measureare made directly on the in situ reinforcing steel, many ofproblems discussed under the section on corrosion prshould not be present.

in

2.a-

in

of

thesign

as an and

ture.

em-n en-

en-

.i-

romer-

cor- ory beugher

, or

4.3—References4.1. Spellman, Donald L., and Stratfull, Richard F., “An Electrical

Method for Evaluating Bridge Deck Coatings,” Highway Research RecordNo. 357, Highway Research Board, 1971, pp. 64-71.

4.2. Stratfull, R. F., “Half-Cell Potentials and the Corrosion of Steel Concrete,” Research Report No. CA-HY-MR-51 16-7-72-42, CaliforniaDepartment of Transportation, Sacramento, 1972.

4.3. Moore, William M.; Swift, Gilbert; and Milberger, Lionel J., “AnInstrument for Detecting Delamination in Concrete Bridge Decks,” High-way Research Record No. 451, Highway Research Board, 1973, pp. 44-5

4.4. Clear, K. C., “Evaluation of Portland Cement Concrete for Permnent Bridge Deck Repair,” Report No. FHWA-RD-74-5, Federal HighwayAdministration, Washington, D.C., 1974, 48 pp.

4.5. Stratfull, R. F.; Jurkovich, W. J.; and Spellman, D. L., “CorrosionTesting of Bridge Decks,” Research Report No. CA-DOT-TL-5116-12-75-10,California Department of Transportation, Sacramento, 1975.

4.6. Ross, Joseph E., “Bridge Deck Deterioration Study,” ResearchReport No. 85, Louisiana Department of Highways, Baton Rouge, 1975.

4.7. “Durability of Concrete Bridge Decks,” NCHRP Synthesis No. 57,Transportation Research Board, Washington, D.C., 1979, 60 pp.

4.8. Manning, David G., and Holt, Frank B., “Detecting Delamination Concrete Bridge Decks,” Concrete International: Design & Construction,V. 2, No. 11, Nov. 1980, pp. 34-41.

4.9. Clear, K. C., “Permanent Bridge Deck Repair,” Public Roads, V. 39,No. 2, 1975, pp. 53-62.

4.10. “Durability of Concrete Bridge Decks—A Cooperative Study,”Report No. 2 (EBO44E), Michigan State Highway Department/BureauPublic Roads/Portland Cement Association, Skokie, 1965, 107 pp.

4.11. “Rolling Pachometer, Operating Manual and Specification,” Office ofDevelopment, Federal Highway Administration, Washington, D.C., 1975.

-t

e

t

n

-.

i-

4.12. Clear, K. C., “FCP Annual Progress Report—Year Ending Sept.30, 1981, Project 4K: Cost Effective Rigid Concrete Construction anRehabilitation in Adverse Environments,” Federal Highway Administration,Washington, D.C., 1979.

4.13. Clear, K. C., “Time-to-Corrosion of Reinforcing Steel in ConcreSlabs, V. 4: Galvanized Reinforcing Steel,” Report No. FHWA-RD-82-028,Federal Highway Administration, Washington, D.C., 1981.

4.14. Berman, H. A., “Determination of Chloride in Hardened CemPaste, Mortar, and Concrete,” Report No. FHWA-RD-72-12, Federal High-way Administration, Washington, D.C., 1972, 22 pp.

4.15. Clear, K. C., and Harrigan, E. T., “Sampling and Testing for Chlo-ride Ion in Concrete,” Report No. FHWA-RD-77-85, Federal HighwayAdministration, Washington, D.C., 1977, 25 pp.

4.16. Clear, K. C., and Hay, R. E., “Time-to-Corrosion of ReinforcingSteel in Concrete Slabs, V. 1: Effect of Mix Design and ConstructioParameters,” Report No. FHWA-RD-73-32, Federal Highway Administration,Washington, D.C., 1973, 103 pp.

4.17. Lankard, D. R.; Slater, J. E.; Hedden, W. A.; and Niesz, D. E.,“Neutralization of Chloride in Concrete,” Report No. FHWA-RD-76-60,Federal Highway Administration, Washington, D.C., 1975, 143 pp.

4.18. Holt, F. B., and Manning, D. G., “Detecting Concrete BridDeck Delaminations with Infrared Thermography,” Public Works, Mar.1979, pp. 66-69.

d

ts

CHAPTER 5—REMEDIAL MEASURES5.1 —Introduction

This chapter discusses measures available to arresttrol, or minimize corrosion activity on an existing reinforcconcrete structure after it is found unnecessary to complreplace the structure.

Some of the normally accepted and commonly implemed remedial measures for controlling corrosion on sstructures are not applicable to the reinforced concrete posite structure. For example, application of protective cings is an effective tool for controlling corrosion on existing steel structure. Obviously, coating of the steel faces to be embedded in concrete would have to be acplished prior to or during the construction phase. Thususe of protective coatings for steel reinforcement is a demeasure and has been discussed in Chapter 3.

Chemical inhibitors are commonly used to control corrosionin a closed system. Inhibitors would need to be introduced admixture during the concrete batching and mixing phaseare not applicable to the existing reinforced concrete struc

5.2—GeneralRemedial measures for controlling corrosion of steel

bedded in portland cement concrete use sound corrosiogineering principles directed toward:

1. Insulating the concrete surfaces from the corrosivevironment.

2. Modifying the environment to make it less corrosive3. Actively controlling the electron flow within the env

ronment so that no metal is lost from the structure.4. Applying a combination of the previous techniques.The reinforced concrete structure may be insulated f

the corrosive environment through application of an impmeable, dielectric barrier between the structure and therosive environment. The barrier may be a coatingmembrane applied to the surface of the concrete, maformed as an integral part of the concrete matrix thropolymer impregnation, or may be an overlay of polymconcrete, low-slump concrete, latex-modified concrete


detr.ithedes.n ocal

n

ionersks,

nsu-

internally-sealed concrete. These have been discussed inin Chapter 3 and will only be mentioned briefly in this chapte

The environment may be altered to reduce corrosion eby removingdetrimental constituents (such as the chloriion), or by removing or neutralizing stray current source

Cathodic protection can be used to control the directioelectron flow within the steel-environment electrochemicircuit to stop corrosion of the reinforcing steel.

tostanyc-

haounfooy bore, arrag

thad it surs nobur-y oive discre

oftin-

s thes tas tm

lly.

ss-seful

ov-ro-alsydro-

gen berideinate-

uit-ode.) inidere itt atre-. In

con-Dis-ime-rent, inthe

can

5.3—ApplicabilityAll reinforced concrete structures are susceptible

corrosion. Although bridge decks5.1-5.3 are perhaps the monotorious examples today, the literature contains mreferences5.4-5.6 to other types of reinforced concrete strutures where corrosion of the reinforcing steel is being orbeen experienced. These include buildings, caissons, fdations, parking garages, piers, piles, pipes, silos, tower ings, and water tanks. Some of these structures matotally or partially buried in soil. Others, such as offshplatforms, water tanks, and the internal surfaces of pipeexposed to aqueous solutions. Bridge decks, parking gaes, and buildings are exposed to the atmosphere.

If the structure is buried or permanently underwater suchthe concrete surfaces are not accessible for treatment, animpractical to expose them, treatment of the surfaces withface coatings or membranes or by application of overlays iapplicable. Similarly, if the structure to be maintained is a ied pipeline or an offshore platform exposed in a large bodwater, modifying the environment to make it less corroswould not be a practical solution. Thus, not all the remediescussed here are applicable to all types of reinforced constructures in the various environments.

Cathodic protection is by far the most versatile methodcorrosion control since it is applicable to any electrically conuous structure within a suitable electrolyte. In as much asteel embedded in concrete, and not the concrete, requirprotection from metallic corrosion, damp concrete serves suitable electrolyte and even structures exposed to the asphere, such as bridge decks, can be protected cathodica

ly theem

conesecrerrierntscor-cor-aver a

mayer, inat-, oxsio

thusm-

luat-andthelity

ron-chlo-caner-ys re-cur-icalmere

5.4—The remedies and their limitations1. Insulative remedies: The insulative methods current

employed to isolate reinforced concrete structures fromcorrosive environment include surface coatings and mbranes, polymer impregnation and overlays of polymer crete, low-slump concrete, or latex-modified concrete. Thmethods are suitable only when the surfaces of the constructure are exposed for treatment. Ideally, these bawould prevent continued intrusion of harmful contaminaand the availability of oxygen or moisture to sustain the rosion reactions. However, in existing structures active rosion is already underway and harmful species hcontaminated the concrete. Insulative methods used aftetive corrosion is experienced do not stop corrosion but mitigate the effects of the corrosion processes. Howevinsulative methods are used without initially decontamiing the concrete, sufficient amounts of the contaminantsygen, and moisture may be entrapped such that corro

ail

r

f

s-

t-e

e-

t is-t

f

-te

eheao-

--

tes

c-

f

-n

progresses until structural integrity is threatened. These isu-lative measures, when used without initial decontamination,should be considered as only a temporary remedy.5.7

Also with insulative remedies an erroneous expectatis that a perfect seal is attainable in practice. All barriwill contain discontinuities such as pinholes, breacracks, poor seams, or other defects that will allow intrusion ofcontaminants in localized areas. Nevertheless, these ilative methods can substantially reduce the rate of intrusionof harmful contaminants and retard the corrosion procees. In many cases, they can successfully extend the ulife of a structure.

2. Modification of the environment: Methods available forrendering the environment less corrosive include the remal or elimination of harmful constituents from the electlyte. These constituents could be in the form of chemicsuch as water and chlorides, gases such as oxygen or hgen sulfide, or electrical currents.

Water can often be eliminated by facilitating drainaaway from rather than through a structure. Chlorides caeliminated by a process known as electrochemical chloremoval. This process has been used to decontamstructures such as bridge decks5.8 in research studies. The removal is accomplished electrochemically, by using a sable electrolyte, an ion exchange resin, and a noble anThe reinforcing steel is the cathode (negatively chargedthis electrochemical circuit. The negatively charged chlorion (Cl-) is attracted to the positively charged anode wheis trapped by the exchange resin. Although a field tesMarysville, Ohio, on a bridge deck displayed promising sults, the method remains in its initial development stagethis test, up to 90 percent of the chlorides present in the crete above the top mat of reinforcement was removed. advantages are that the method is expensive and tconsuming and requires the application of high direct curvoltage that generates heat (around 200 F [90 C]) whichturn, can induce cracking of the concrete. In addition, permeability of the concrete was increased.

Harmful gases such as oxygen and hydrogen sulfidebe stripped by chemical processes from the electrolyte, making it less corrosive. This method is applicable predoinantly for structures exposed in aqueous solution.

Deep polymer impregnation5.9 of the critically contami-nated concrete around the reinforcing steel is being evaed. The method ties up the existing contaminants prevents intrusion of additional contaminants. Although theory is realistic, the practicality and economic feasibiof deep polymerization are not established.

Corrosion of steel in concrete can be caused by envimental factors other than chemical constituents such as rides, moisture, or oxygen. Stray electrical currents result in corrosion by electrolysis, i.e., cathodic interfence.5.5,5.10In corrosion by electrolysis, direct current strafrom an exterior source and is collected by the steel in ainforced concrete structure. Inasmuch as the collected rent must return to its source to complete the electrochemcircuit, the current is discharged from the structure at solocations. At the point of current discharge from the structu


com systyperuc-igat- thes. Inicallyrenc via trucin thctrohic

th-

-edlineotht th

do- any

tials

em-oderomo beodicodic

heredic

ancedr-

ic ar-eir

mi-trodere toed isum

ctivelect-f ca-ass

umi-rodeinc,amp

lt-node be-zinc

es-nceccor- the

s

to the electrolyte, metal loss is experienced. The most mon sources of stray currents include cathodic protectiontems, electrified railways, and electroplating plants. This of corrosion is most commonly manifested in grounded sttures, i.e., those in contact with the earth. A method of miting this type of corrosion, implemented for many years inburied pipe industry, is the installation of resistance bondresistance bonding, the structure being affected is electrconnected through a resistor to the source of the interfecurrents. In this manner, the current returns to its sourcemetallic path such that no metal loss from the affected sture occurs. Another method uses galvanic anodes to dracollected current. Collected current is passed on to the elelyte and back to its source from the surface of the anode wcorrodes, rather than the structure.

Examples of stray current intrusion and mitigation meods are shown in Fig. 5.1 to 5.4.

3. Active control of electron flow—From the Pourbaix diagram for iron (Fig. 2.4), it can be seen that steel embeddin concrete is normally passivated due to the highly alka(high pH) concrete environment. The diagram shows aner area wherein no steel corrosion occurs. This area alower portion of the diagram is labeled immunity. In this main, the potential of the steel is more negative than innaturally occurring condition, regardless of pH.

The method of providing the highly negative steel potenrequired for immunity is referred to as cathodic protection. Incathodically protecting a structure, a favorable electrochical circuit is established by installing an external electrin the electrolyte and passing current (conventional) fthat electrode through the electrolyte to the structure tprotected. This current polarizes the potential of the cathsurfaces (relatively positive) on the steel to that of the an

Fig. 5.1—Underground pipe crossing Layout Line A icathodically protected

Fig. 5.2—Elevation showing current flow patterns

Fig. 5.3—Mitigation with galvanic anodes (current returnsvia low resistance anode)

Fig. 5.4—Mitigation with resistance bond (current returnsvia metallic path)

--

ea-e-

h

-e

(more negative) surfaces. When this is accomplished, tis no current flow between the formerly anodic and cathosurfaces and corrosion is arrested. This represents a balor equilibrium condition. In normal practice, sufficient curent is passed to the surfaces so that the formerly anodeas will be receiving current from the electrolyte and thpotential will be shifted to the more negative direction.

There are two ways in which the protective electrochecal circuit can be established. One method uses an elecmade of a metal or alloy more negative than the structube protected. For example, if the structure to be protectconstructed of steel, either magnesium, zinc, or aluminmay be coupled with the structure. Inasmuch as a protegalvanic cell is set up between the steel and the alloy seed, this method is known as the galvanic anode method othodic protection. Also, since the galvanic anodes pcurrent to the electrolyte, they corrode or sacrifice themselvesto protect the structure. Hence, magnesium, zinc, and alnum are termed sacrificial anodes. Sacrificial anodes corat relatively high rates. Corrosion rates for magnesium, zand aluminum are of the order of 17, 26, and 12 lb per year, respectively.5.11,5.12

The high consumption rates, as well as low-driving voage, are the primary disadvantages of the galvanic amethod of cathodic protection. The open circuit potentialtween steel and magnesium is on the order of 1 V, while and aluminum are somewhat less.5.11,5.12 Thus, with thismethod, it is imperative that a low-resistance circuit betablished by installation of many anodes in a low-resistamedia. The anodes installed should also be sized in adance with their respective consumption rates to providenecessary design life.


calur-ce rental-

thelow betedhesnd o

on-

ernal oflne

be

m are

ied

inade

resntly in-

incanuch

of ato re-ibu-

eteerss the

ub-tive

elec-ore 0.3hat0.5

tmo-mayient

eelft

Table 5.1—Comparison of electrochemical circuitcharacteristics

Galvanic or sacrificial anode Impressed current1 No external power required External power required

2 Fixed, small driving voltage Voltage variable overwide range

3 Limited, small current output Current variable overwide range

4 Interference of adjacentstructure not likely Interference can result

5 Overprotection not likely Overprotection can result6 Anodes rapidly consumed Anodes slowly consumed

7 Sensitive to temperature andmoisture conditions

x

The other way in which the favorable electrochemicircuit can be established is by introducing electrical crent from an external source. Because an outside sourcurrent is used, this method is termed impressed curcathodic protection. This method also requires the inslation of an external electrode in the electrolyte with structure to be protected. However, since the current fis not dependent on the favorable potential differencetween the electrode and the structure to be protecmore noble materials can be selected for the anode. Tmaterials include high-silicon cast iron, graphite, aeven more noble materials such as platinized-titaniumplatinized-niobium. These metals corrode or are csumed very slowly, less than 1 lb per amp year.5.11,5.12

oft

-,e

r

These anodes are coupled to the structure via the extsource of electrical power. This source can be in the formbatteries, thermoelectric generators, generators, or photovotaiccells. Most commonly, however, alternating current livoltage is converted to direct current by a rectifier.

The two ways in which cathodic protection can achieved are shown in Fig. 5.5 and 5.6. A comparison of therespective advantages and disadvantages of the systeshown in Table 5.1.

Successful application of cathodic protection to burpipe were documented as early as 1946.5.13,5.14 The initialapplication of cathodic protection to bridge decks was1974 and other applications have subsequently been mwith encouraging results.

The cathodic protection of reinforced concrete structuis thus proven technology and the problems being curreencountered deal with criteria of protection, design, andspection of the installation.

In protecting buried structures or structures exposedwater or in soils, low-resistance electrochemical circuits normally be established. However, on other structures sas bridge decks, a highly conductive overlay consisting coke breeze-asphalt mixture or closely spaced anodes duce the circuit resistance and to promote uniform distrtion of current to all reinforcement is required.5.15,5.16 Atypical arrangement is shown in Fig. 5.7.

The criteria for protection of steel embedded in concrare not clearly defined. Most commonly, corrosion engineuse the potential compared to a standard reference cell asole criterion. The criterion for steel that is buried or smerged is normally accepted as -0.85 V, or more negathan a copper/CSE (copper-copper sulfate reference trode). However, steel embedded in concrete exhibits mnoble potentials than exposed steel in the order of 0.2 toV more positive. Therefore, some investigators claim tprotection is provided at lower potentials, in the order of -V with CSE reference.5.17

For steel embedded in concrete exposed to the asphere, research has indicated that the -0.85-V criterion not be attainable. Quite possibly the result may be sufficcurrent to cause concern about lack of bond.5.18

The possibility of the loss of bond of the reinforcing stis related to high current densities, at least 25 mA/2.

Fig. 5.5—Galvanic cathodic protection (buried pipe)

Fig. 5.6—Impressed current cathodic protection (buried pipe)

Fig. 5.7—Cathodic protection circuit using conductive mias secondary anode


tion

nles

ac-proyn of pro thael ine. Ies oomn ond-losebed

blyob-

ygeithf 3

allromuldd tht re-s def-ingte. Ithichol. th

d toatedvaricell cri-ut o

pedl betrol neavelyby a benifi-nt

po

Thes th

re- Thiss be-ted rate, if of anybe

hy-

de to hy-med thisis de-therro-l, it

ilityrogents

eld

to

ur- Vm-tion

edicor-trolngn isn theini-

on-sive re-tro-

vel-preg- asin-

hasting

to bes as

However, it would be most unusual for a cathodic protecsystem, typically designed to operate at 2 mA/ft2 of steel sur-face, to operate in excess of 25 mA/ft2 for sufficient time(several years) to cause deterioration in bond strength uthe potential criterion was applied inappropriately.

Corrosion of steel in concrete is controlled by oxygencess. Polarization of the steel is controlled by cathodic tection. As mentioned in Chapter 2, concrete is a veralkaline medium and cathodic reaction is the reductiooxygen to hydroxides. The same is true for the cathodictection currents. If the current reduces the oxygen fasterit can be replenished, when cathodically protecting steconcrete, the steel will polarize to a more negative valuthe oxygen supply is great, then to obtain greater degrepolarization, the current supply must be increased. In sbridge decks, the current required to obtain the criterio-0.85 V would be such that there would be fear for disboment even though the half-cell potential was not even cto that value. Thus, cathodic protection of concrete emded steel is not necessarily a standard procedure.

For concrete that is buried or submerged, probamoisture-saturated, the -0.85 V CSE criterion is easily tained at current densities as low as 25 µ A/ft2. For bridgedecks, where the concrete is comparatively dry and oxis abundant, the criteria may be -0.85 V if obtainable wreasonable, current density (probably a maximum omA/ft2 of deck surface). If not, a shift of 400 mV for bridge deck half-cell potentials is a criterion developed fthe statistical distribution of half-cell potentials that cochange the least negative potential to equal or exceemost negative half-cell potential, or to provide the currenquired to achieve the cathodic protection requirements atermined by the E-log I curve.5.18 In this latter case, the halcell potential of the bridge deck can vary wildly dependon the moisture content and temperature of the concrealso eliminates the need of a permanent half-cell wwould be required in the case of half-cell potential contr

The theory behind the constant current determined byE-log I criteria is that the potential of steel is directly relateoxygen reduction. The corrosion of the steel is directly relto the corrosion current. Thus, with a constant current, ations in oxygen reduction will cause variations in the half-potential of the steel. Conversely, if the cathodic protectionteria is a constant half-cell potential, then the current outpthe system will vary as the oxygen supply varies.

When using the half-cell potential criterion as develothrough the E-log I method, there is a risk that there wiltimes when the cathodic system will not completely conthe corrosion of the steel. For example, if the concrete issaturation, the steel can usually be polarized with relatismall current densities. Then if the rectifier is regulated half-cell potential and the concrete dries so that oxygencomes abundant, and the polarized potential drifts sigcantly less negative, it is likely that there will be insufficiecurrent capacity to raise the potential to the protectivetential value.

Corrosion is caused by the flow of electrons or current.differences in half-cell potentials is the voltage that cause

s

-

-n

ff

ef

-

n

e

-

current to flow. Once the steel is made cathodic in that itceives current, the current causes oxygen to be reduced.same amount of current may reduce oxygen faster than it iing replenished and result in polarization with an associapotential change. If the oxygen is replenished at the sameas it is reduced, no additional polarization will result. Thusthe amount of current for cathodic protection will make allthe steel cathodic and oxygen reduction is taking place,greater amount of cathodic protection current will simply wasted on reducing oxygen.

In addition to disbondment, overprotection can result in drogen embrittlement.5.6,5.19 In Chapter 2 it was shown that inacid environments hydrogen ions are reduced at the cathoatomic hydrogen which, in turn, combine to form gaseousdrogen. When overprotection results, hydrogen gas is forat a faster rate than can be diffused through the coating, ininstance, concrete. When this occurs, gaseous pressure veloped at the steel-coating interface which tends to eispall the coating (disbondment) or to diffuse as atomic hydgen into the metal. When hydrogen diffuses into the metafurther strains the metal lattice, resulting in reduced ductand toughness. These phenomena are referred to as hydembrittlement. Normally, hydrogen embrittlement affechigh-strength steels only, generally those having yistrengths of 90 ksi (620 MPa) or higher5.20,5.21 and is conse-quently a potential problem in applying cathodic protectionprestressed concrete elements.

Because of the adverse effects possible from overprotection,polarized potentials (determined immediately after the crent has been interrupted) are normally limited to -1.10CSE to avoid the possibility of disbonding and hydrogen ebrittlement problems.5.22 In addition, protection above thalevel would require more current and a costlier installatwithout achieving additional protection from corrosion.

e

-

f

r

-

-

e

5.5—SummaryRemedies for controlling corrosion on existing reinforc

concrete structures use sound corrosion engineering princplesdirected at insulation of the reinforced concrete from the rosive environment, alteration of the environment, or conof electrical current flow within the environment. In dealiwith the reinforced concrete composite, when corrosiodetected, the deleterious contaminants are already withiconcrete matrix. Insulative measures, although they mmize the rate of corrosion or the intrusion of additional ctaminants, entrap the existing quantities of these corrocontaminants. Their effectiveness can be improved bymoval of contaminants prior to sealing such as by elecchemical chloride removal.

Many of the proposed remedies are in the early deopment stage. Such approaches as deep polymer imnation and corrosion inhibitors have not been provenpracticable methods of corrosion control on existing reforced concrete structures.

Of the remedies discussed, only cathodic protectionproven to be capable of stopping corrosion on an exisstructure. The technology is proven and has been foundcost-effective. The design procedures for such structure


e dsueve

tion

aw

buried pipe and water tanks are well established, but thsign criteria for structures exposed to the atmosphere, as bridge decks and parking structures, are still in the dopmental stages.

rchfor

ete

s

o-

tial

tee,

geh

idge

it,

f

an

r-n-

is of

e

ing

tion

asis.

5.6—References5.1. Godfrey, Kneeland A., “Bridge Decks,” Civil Engineering—ASCE,

V. 45, No. 8, Aug. 1975, pp. 60-65.5.2. Manning, David G., and Ryell, John, “Durable Bridge Decks,”

Report No. RR203, Ontario Ministry of Transportation and Communications,Downsview, 1976, 67 pp.

5.3. Arnold, C. J., “Bridge Decks in Michigan: A Summary of Reseaand Performance,” Conference on Federally Coordinated Program Research and Development, Pennsylvania State University, UniversityPark, Sept. 1976.

5.4. Griffin, D. F., “Corrosion of Reinforced Concrete in Marine Envi-ronments,” Materials Protection, V. 4, No. 11, Nov. 1965, pp. 8-11.

5.5. Ellis, W. J., “Corrosion of Cement Mortar Coated Pipelines,” Amer-ican Water Works Association Meeting, Honolulu, Oct. 1974.

5.6. Philips, E., “Survey of Corrosion of Prestressing Steel in ConcrWater-Retaining Structures,” Technical Paper No. 9, Australian WaterResources Council, Canberra, 1975.

5.7. Frascioia, R. I., “Waterproofing Membranes—Are Their ProblemInsurmountable,” FCP Project 4B Review, State College, Pennsylvania,Sept. 1976.

5.8. Slater, John E.; Lankard, David R; and Moreland, Peter J., “Electrchemical Removal of Chlorides from Concrete Bridge Decks,” MaterialsProtection, V. 15, No. 11, Nov. 1976, pp. 21-26.

5.9. Hay, R. E., “The Bridge Deck Problem—An Analysis of PotenSolutions,” Public Roads, V. 39, No. 4, Mar. 1976, pp. 142-147.

5.10. Mudd, O. C., “Control of Pipe-Line Corrosion,” Corrosion, V. 1,No. 12, Dec. 1945, pp. 192-218, and V. 2, No. 3, Mar. 1946, pp. 25-58.

5.11. Corrosion Prevention and Control Manual, Navdocks MO-306,Department of the Navy, Bureau of Yards and Docks, Washington, D.C.,June 1964.

5.12. Peabody, A. W., “Control of Pipeline Corrosion,” National Associ-ation of Corrosion Engineers, Houston, 1967.

5.13. Henderson, D., “Coated Pipe and Cathodic Protection,” ConsultingEngineer, Mar. 1962.

5.14. Deskins, R. L., “Cathodic Protection of a Mortar Coated SWater Distribution System,” Materials Protection, V. 5, No. 9, Sept. 1966pp. 35-37.

5.15. Stratfull, R. F., “Experimental Cathodic Protection of a BridDeck,” Transportation Research Record No. 500, Transportation ResearcBoard, 1974, pp. 1-15.

5.16. Fromm, H. J., “Cathodic Protection of Rebar in Concrete BrDecks,” Materials Performance, V. 16, No. 11, Nov. 1977, pp. 21-29.

5.17. Robinson, R. C., “Cathodic Protection of Steel in Concrete,” Cor-rosion of Metals in Concrete, SP-49, American Concrete Institute, Detro1975, pp. 83-93.

5.18. Stratfull, Richard F., “Criteria for the Cathodic Protection oBridge Decks,” Corrosion of Reinforcement in Concrete Construction, EllisHorwood, Chichester, 1983, pp. 287-331.

5.19. Dykmass, M. J., “Corrosion of Prestressing Steel in ConcreteHow This Can Be Minimized or Prevented,” National Association of Cor-rosion Engineers Western Regional Conference, San Diego, Sept. 1976.

5.20. Uhlig, Herbert H., Corrosion and Corrosion Control, John Wiley& Sons, New York, 1963.

5.21. Fontana, M. G., and Greene, N. D., Corrosion Engineering, 2ndEdition, McGraw-Hill Book Co., New York, 1978, 448 pp.

5.22. Scott, G. N., “The Corrosion Inhibitive Properties of Cement Motar Coatings,” National Association of Corrosion Engineers Annual Covention, Kansas City, Mar. 1962.

nth

CHAPTER 6—REFERENCES TO DOCUMENTS OFSTANDARD-PRODUCING ORGANIZATIONS

The documents of the various standards-producing orgazations referred to in this document are listed below witheir serial designation.

e-chl-

l

American Association of State Highway and TransportaOfficials

T 260 Sampling and Testing for Total ChlorideIon Content in Concrete and Concrete R

Materials

American Concrete Institute201.2R Guide to Durable Concrete318 Building Code Requirements for

Reinforced Concrete

ASTMC 114 Standard Method for Chemical Analys

Hydraulic CementC 876 Standard Test Method for Half-Cell Potentials of Reinforcing Steel in ConcreteC 1152 Standard Test Method for Acid-Soluble

Chloride in Mortar and ConcreteC 1218 Standard Test Method for Water-Solubl

Chloride in Mortar and Concrete

These publications may be obtained from the followorganizations:

American Association of State Highway and TransportaOfficials444 North Capitol St., NWSuite 225Washington, D.C. 20001

American Concrete InstituteP.O. Box 9094Farmington Hills, MI 48333-9094

ASTM100 Barr Harbor DriveWest Conshohocken, PA 19428-2959

This report was submitted to letter ballot of the committee on an item-by-item bThe committee consists of 17 members. All items were approved by the necessarytwo-thirds vote.

d

i-

Appendix A—ACI 222.1-96

Provisional Standard Test Method for Water-Soluble Chloride Available for Corrosion of

Embedded Steel in Mortar and Concrete Using theSoxhlet Extractor

Reported by ACI Committee 222

Some water-soluble chlorides, primarily in certain aggregates, do not inducecorrosion of embedded reinforcing steel since these chlorides are bound withinthe aggregate. Currently, available test methods cannot distinguish between thewater-soluble chlorides that support corrosion and those that do not. This testmethod detects only water-soluble chlorides that contribute to the corrosion ofthe reinforcing steel.


tes;
Keywords: Water-soluble chlorides; corrosion; steel; mortar; concreSoxhlet Extractor.
am-, or

e fo

1—Scope1.1—This test method provides procedures for the s

pling and analysis of hydraulic-cement mortar, concreteaggregate for chloride that is water-soluble and availablthe corrosion reaction under the conditions of the test.

ll of th ap ap

1.2—This test method does not purport to address athe safety problems, if any, associated with its use. It isresponsibility of the user of this test method to establishpropriate safety and health practices and determine theplicability of regulatory limitations prior to use.

nttalsteexientabl node

ridechloss to

tion.foreralldelyld bsul

, the con-era- in therous

2—Significance and Use2.1—Water-soluble chloride, when present in sufficie

amounts, may initiate or accelerate the corrosion of mesuch as steel embedded in or contacting a cement sysuch as mortar, grout, or concrete. Other test methods for the determination of water-soluble chloride in a cemsystem.* However, some aggregates contain a consideramount of chloride that is bound in the aggregate and isavailable for the corrosion reaction. The test method scribed in ASTM C 1218 measures a portion of the chlocontained in these aggregates. However, the amount of ride measured is very dependent on the degree of finenewhich the aggregates are ground during sample prepara†

The problem with the ASTM C 1218 test method is theretwofold: the test measures chlorides that are not geneavailable for the corrosion reaction, and the test gives wivariable results. The test method described herein shouused when chloride-bearing aggregates influence the reobtained using ASTM C 1218.

i-ateigh eridee

mple.id iseats. ac-

vesive

2.2—Sulfides are known to interfere with the determnation of chloride content. Blast-furnace slag aggregand cement contain sulfide sulfur in concentrations henough to cause significant interference and produceroneous test results. Treatment with hydrogen peroxas discussed in ASTM C 114,‡ shall be used to eliminatsuch interference.

b- C

nd

nce,

tion.

3—Apparatus3.1—Sampling Equipment: The apparatus required for o

taining samples by coring or sawing is described in ASTM42.§ Sampling by drilling is not applicable for this test ashall not be used.

theer o

ex-

as-ride or

3.2—Sample Processing Apparatus:3.2.1 Samples too large to fit in the sample holder of

Soxhlet shall be reduced in size by means of a jaw crushby hammering.

3.2.2 Extract chlorides from the sample using a Soxhlettractor, a schematic of which is shown in Fig. A1. The Soxhlet

*ASTM Standard Test Method C 1218-92, Standard Test Method for Water-SolubleChloride in Mortar and Concrete.

†For more information see “The Determination of the Chloride Content of Con-crete” by Brian B. Hope, John A. Page, and John S. Poland, Cement and ConcreteResearch, V. 15, No. 5, Pergamon Press, New York, Sept. 1985, pp. 863-870.

‡ASTM Test Method C 114, Test Methods for Chemical Analysis of Hydraulic Cement.§ASTM C 42, Test Method for Obtaining and Testing Drilled Cores and Sawn

Beams of Concrete.

r

e--

smst

et-

-Fig. A1—Soxhlet extraction apparatus

y

ets

s

-,

extractor consists of a heater, a lower flask to hold watersample compartment, and a condenser. The extractortains approximately 100 ml of distilled water in the lowflask. Heat is applied to this flask; vapor from the boiling wter passes to the condenser; and the condensate collectssample compartment. The sample is contained in a poholder and the hot condensate collects around the saWhen the condensate reaches a critical height, the liqusiphoned back into the lower flask and the process repThe nonvolatile components extracted from the samplecumulate in the lower flask, while each extraction involfresh hot distillate. The heat input shall be sufficient to gan extraction cycle about every 20 min. For conveniesuitable commercial equipment is available.**

4—Reagents4.1—The reagents required for the chloride determina

are given in the test method for chloride of ASTM C 114

days

TMay

r5— Sampling

5.1—Reduce the size of a minimum 300 gm samplespecified in Section 6, and divide this sample to a minimum 30 gm representative sample for use in the chlodetermination. If the sample is taken from concretemortar then the concrete or mortar shall be at least 7 old before sampling.

Note 1—Concrete cores taken in accordance with ASC 42 or concrete cylinders cast from the proposed mix m

** Suitable Soxhlet extraction equipment is available from Fisher Scientific (catalogNo. 09-551A) and other manufacturers.


00encnot

.0 theTM

be cut longitudinally or laterally to provide the required 3gm sample representative of the core or cylinder. Experihas shown that the cooling water from core cutting will dissolve a significant amount of chloride.

sh-pow-me

n no

et,no

001

6— Sample Preparation6.1—Using the jaw crusher or hammer, reduce the sample

so that it fits the sample holder using the minimum cruing necessary. The sample shall not be crushed to a der since this would release chloride bound in soaggregates which, as previously discussed, are knowto contribute to corrosion.

lo-
7—Procedure
7.1—A single test shall consist of determination of chride contents of three individual 30 gm samples.

gcto de

7.2—Weigh each sample (30 g ± 5 g) to the nearest 0.01and place in the porous sample holder of a Soxhlet extraAdd a wad of glass wool. Place approximately 100 ml ofionized water in the lower flask.

wa- Tu

eive

7.2.1 Assemble the condenser complete with cooling ter supply pipes to the extractor and place on the heater.on both the heater and condenser cooling water and allowtraction to continue for 24 hr; adjust the heating rate to ga cycle about every 20 min.

theasks toe at toica-dd

the

ro-

7.2.2 At the conclusion of the extraction stage, transfersolution to a 500 ml volumetric flask. Rinse the Soxhlet flthree times with distilled water, transferring the washingthe 500 ml volumetric flask; add distilled water to producvolume of 500 ml. With a pipette transfer a 25 ml aliquoa 250 ml conical flask. Add 3 drops of methyl orange indtor (prepared in accordance with ASTM C 114) and a

edilute (1+1) nitric acid until the solution is acidified. Add 3± 0.1 ml of hydrogen peroxide (30 percent solution) tosolution. Proceed in accordance with the reference ASC 114, starting with the procedure specified in Section19.5.3 and continuing to the end of Section 19.5.8.

7.2.3 Make a blank determination by using the Soxhlcomplete with thimble and glass wool, but containing sample of cementitious material.

t

r.-

rnx-

8—Calculation8.1—Calculate percent of chloride to the nearest 0.

percent as the average chloride content of the triplicate samples,each calculated as follows

(1)

whereV1 = ml of 0.05 N AgNO3 solution used for titration of

the sample (equivalence point)Vb = ml of 0.05 N AgNO3 solution used for titration of

the blank (equivalence point)N = normality of 0.05 N AgNO3 solution, calculated to

±0.001M = mass of concrete or mortar sample, gV2 = volume of the 25 ml aliquot determined to ±0.1 ml

(larger or smaller aliquots may be used depending onchloride concentrations present)

Chloride, percent 3.5453 V1 Vb–( ) NM----- 500

V2---------×××=

8.2—Sufficient data are not available at this time to pvide precision and bias statements.