ACI 201.2R-92-

(Reapproved 1997)

Guide to Durable Concrete

Stanley J. Bias, Jr.W. Barry Butler

Ramon L. CarrasquilloPaul D. CarterBoguslaw ChojnackiKenneth C. ClearWilliam A. Cordon

Be.rnard Erlin*Emery FarkasPer FidjestolJohn F . G ibbonsEugene D. Hill, Jr.J e n s H o l mR. Douglas Hooton

Reported by ACI Committee 201

Cameron Maclnnis*Chairman

John M. JaffePaul Klieger’Joseph F. LamondTorbjorn J. Larsen

Kenneth R. LauerStella L. MarusinTimothy B. MastersBryant Mather*

Katharine Mather*’Tarun R. NaikHoward Newlon, Jr.

Michel PigeonRobert E:Price*Dr. RasheeduuafarThomas J. Reading*

Mauro J. ScaliS e c r e t a r y

*Members of subcommittee who prepared this report.‘Deceased.

This guide is a complete revision of the committee document “GuideIO Durable Concrete, ” ACI 201.2R-77 (Reapproved 1982). which ap-Frs in Part I of the ACI Manual of Concrete Practice. This new

document represents major revisions in a number of areas, which re-flect an increase in the knowledge of the numerous factors influenc-ing concrete durability since 1977.

In general, separate chapters are devoted to specific types of con-crete deterioration. Each chapter contains a discussion of the mech-

anisms involved and the recommended requirements for individualcomponents of the concrete, quality considerations for concretemixtures, construction procedures, and influences of the exposureenvironment, all important considerations to assure concrete durabil-ity. Some guidance as to repair techniques is also provided.

“Water-cement ratio” is used throughout this document rather

than the newer term, “water-cementitious materials ratio, ” since therecommendations are based on data referring to water-cement ratio.If cementitious materials other than portland cement have been in-cluded in the concrete, judgment regarding required water-cementratios have been based on the use of that ratio. This does not implythat new data demonstrating concrete performance developed using

portland cement and other cementitious materials should not be re-ferred to in terms of water-cementitious materials. Such information,if available, will be included in future revisiolrs.

AC1 Committee Rewns. Guides, Standard practices. and Commentaries are intended forguidance in planning.‘designmg. executing. and inspecting constructton. This document isIntended for the use of indw~duals who arc competent to evaluate the significanceand limitations of Its content and recommendations and who will accept responsibilityfor the application of the material it contains. The American Concrete Institute disclaimsanv and all resoonsibibtv fo’r the stat4 orincioles. The institute shall not bc liable for anylo& or damag~arising ttkreftom. . .

Reference to this document shall not be made in contract documents. If items found inthis document are desired by the Architect/Engineer to be a part of the contract docu-ments. they shall be restated in mandatory language for incorporation by the Architect/Engineer.

Charles F. ScholerHannah Schell

Jan P. SkalnyPeter SmithGeorge V. TeodoruJ. Dearle Thorpe

P. J. TikalskyClaude B. Trusty, Jr.George J. VentaGary L. VondranOrville R. WernerDavid A. Whiting*J. Craig Williams*

Byron I. Zolin

Keywords: abrasion; abrasion nsis~ance; acid resistance; adhesives; admix-tures; l ggmgala: air cnlrainment; alkali-aggregate reactions; l lkali-carbonalcreactions; bridge decks; calcium chloride; c e m e n t - a g g r e g a t e reaclions; cementpastes; chemical analysis; chcmicrl atlack; c h l o r i d e s ; coat ings; cold weatherconstruction; conctelc dunbilily; concrete pavements: corrosion; corrosion re-

s&am; cracking (fracturing); curing; damage; deicers; deterioration; dunbil-ily; epoxy resins; floors; fly ash; freeze-thaw dunbilily; freezing; hot weatherconstruction; mix proportioning; petrography; plastics, polymers, and resins;pozzolans; prolcclive coatings. protectors; reinforced concrete; reinforcingsteels; repairs: silica fume; skid resistance; spalling; strength; sulfalc l ll8ck;temperature; water-cement ratio; waler-cementitious ratio; waterproof coat-ings; water-reducing agents.

IntroductionCONTENTS

Chapter 1 -Freezing and thawingI. I -General

I .2-Mechanisms of frost actionI .3-Ice-removal agentsI .4-Recommendations for durable structures

AC1 201.28-92 supersedes AC1 201.2R-77, Rupproved 1982effective FebruaryI. 1992. The document has undergone extensive revision. Chapter 7 has beencompletely revised and major revisions have also been made in Chapters 2. 3. 4.and 5.

Copyright 0 1991. American Concrete Institute.A l l righrs reserved in&din8 tighls of reproduction and use in any form or by

any means, including the making of copies by any photo process. or by any elec-ITOIUC or meehoial device, printed. written. or on1 or recording for sound orvisual reproduction or for use in any knowledge or retr ieval system or deviceunless permission in writin is obtained from the copyright proprietors.

201.2R.2 MANUAL OF CONCRETE PRACTICE

Chapter 2-Aggressive chemical exposure2.1-General2.2-Sulfate attack2.3--Seawater exposure2.4-Acid attack2.S-Carbonation

Chapter 3-Abrasion3.1--Introduction3.2-Testing concrete for resistance to abrasion3.3-Factors affecting abrasion resistance of concrete3.4-Recommendat ions for obtaining abrasion-resistant concrete

surfaces3.5-Improving wear resistance of existing floors3.dStudded tire and tire chain wear on concrete3.7-Skid resistance of pavements

Chapter 4-Corrosion of metals and othermaterials embedded in concrete

4. I-Introduction4.2-Principles of corrosion4.3-Effects of concrete-making components4.4-Concrete qual i ty and cover over steel4.5-Positive protective systems4.6-Corrosion of materials other than steel4.7-Summary comments

Chapter 5-Chemical reactions of aggregates5. l-Types of reactions5.2-Alkali-silica reaction5.3-Alkali-carbonate reaction5.4-Preservation of concrete containing reactive aggregate5.5-Recommendations for future studies

Chapter 6-Repair of concrete6.1 -Evaluation of damage and selection of repair method6.2-Types of repairs6.3-Preparations for repair6.4-Bonding agents6.5-Appearance6.6-Curing6.7-Treatment of cracks

Chapter I-Use of protectivebarrier systems toenhance concrete durability

7. I-Characteristics of a protective-barrier system7.2-Elements of a protective-barrier system7.3-Guide for selection of protective-barrier systems7.4-Moisture in concrete and effect on barrier adhesion7.5-Influence of ambient conditions on adhesion7.6-Encapsulat ion of concrete

Chapter a-References8. I -Recommended references8.2-Cited references

Durability of hydraulic-cement concrete is defined asits ability to resist weathering action, chemical attack,abrasion, or any other process of deterioration. Dura-ble concrete will retain its original form, quality, and

INTRODUCTION

_ .serviceability when exposed to its environment. Someexcellent general references on the subject are available(Klieger 1982; Woods 1%8).

This guide discusses in some depth the more impor-tant causes of concrete deterioration, and gives recom-mendations on how to prevent such damage. Chaptersare included on freezing and thawing, aggressive chem-ical exposure, abrasion, corrosion of metals, chemicalreactions of aggregates, repair of concrete, and the useof protective-barrier systems to enhance concrete dura-bility. The fire resistance of concrete is not covered,since it is included in the mission of AC1 Committee216 (AC1 216R); nor is cracking, which is included inthe mission of AC1 Committee 224 (see AC1 224R andAC1 224.1 R).

It has been well established for many years thatfreezing and thawing action in the temperate regions ofthe world can cause severe deterioration of concrete.Increased use of concrete in countries having hot cli-mates during recent decades has drawn attention to thefact that deleterious chemical processes such as corro-sion and alkali-aggregate reactions are aggravated byhigh temperatures. Also, the combined effects of coldwinter and hot summer exposures should receive atten-tion in proportioning and making of durable concrete.

It should be kept in mind that water i; required forthe chemical and most physical processes to take placein concrete, both the desirable ones and the deleteri-ous. Heat provides the activation energy that makes theprocesses proceed. The integrated effects of moisture,heat, and other environmental elements are importantand should be considered and monitored. Selecting ap-propriate materials of suitable composition, and proc-essing them correctly under the existing environmentalconditions, is therefore essential to achieve concretesresistant to deleterious effects of water, aggressive so-lutions, and extremes of temperature.

Freezing and thawing damage is a serious problem.The mechanisms involved are now fairly well under-stood. The damage is greatly accelerated, particularly inpavements, by the use of deicing salts, often resultingin severe scaling at the surface. Fortunately, concretemade with good quality aggregates, a low water-cementratio, and a proper air-void system, which is allowed todevelop proper maturity before being exposed to severefreezing and thawing, will be highly resistant to suchaction.

Sulfates in soil, ground water, or seawater will be re-sisted by using suitable cementitious materials and aproperly proportioned concrete mixture which has beensubjected to proper quality control.

Quality concrete will resist occasional exposure tomild acids, but no concrete offers good resistance toattack by strong acids or compounds that may convertto acids; special protection is necessary in these cases.

GUIDE TO DURABLE CONCRETE 201.2~.3

Abrasive action may sometimes cause concrete sur-faces to wear away. Wear can be a particular problemin industrial floors. In hydraulic structures, particles ofsand or gravel in flowing water can erode surfaces. Theuse of high-quality concrete and, in extreme cases, avery hard aggregate will usually result in adequate du-rability under these exposures. The use of studded tireson automobiles has caused serious wear in concretepavements; conventional concrete will not withstandthis action.

The spalling of concrete in bridge decks has becomea serious problem. The principal cause is reinforcingsteel corrosion, which is largely due to the use of deic-ing salts. The formation of the corrosion products pro-duces an expansive force which causes the concrete tospa11 above the steel. Ample cover over the steel and useof a low-permeability, air-entrained concrete will as-sure good durability in the great majority of cases, butmore positive protection, such as epoxy-coated rein-forcing steel, cathodic protection, chemical corrosioninhibitors, or other means, is needed for very severeexposures.

Although aggregate is commonly considered to be aninert filler in concrete, such is not always the case. Cer-tain aggregates can react with alkalies in cement, caus-ing expansion and deterioration. Care in the selectionof aggregate sources, and the use of low-alkali cementand/or preiested pozzolans or ground slag will allevi-ate this problem.

The final chapters of this report discuss the repair ofconcrete that has not withstood the forces of deterio-ration, and the use of protective-barrier systems to en-hance durability.

The committee wishes to stress that the use of goodmaterials and good mixture proportioning will not as-sure durable concrete. Good quality control and work-manship are also absolutely essential to the productionof durable concrete. Experience has shown that twoareas should receive special attention: 1) control of en-trained air and 2) finishing of slabs. The ACI Manualof Concrete Inspection (AC1 3 11.1 R) describes goodconcrete practices and inspection procedures. AC1302.1 R describes in detail proper practice for consoli-dating and finishing floors and slabs in Chapter 7. AC1316R reviews pavement installation. AC1 330R dis-cusses parking lot concrete. AC1 332R covers residen-tial concrete, including driveways and other flatwork.

CHAPTER l-FREEZING AND THAWING1 .l -General

Exposing damp concrete to freezing and thawing cy-cles is a severe test that only good quality concrete sur-vives without impairment. Air-entrained concrete whichis properly proportioned with quality materials, manu-factured, placed, finished, and cured resists cyclicfreezing for many years.

It should be recognized, however, that under ex-tremely severe conditions even quality concrete maysuffer damage from cyclic freezing, e.g., if it is kept in

a state of nearly complete saturation. This situtationmay be created when cold concrete is exposed towarmer, moist air on one side and evaporation is insuf-ficient or restricted on the cold side, or when the con-crete is subjected to a head of water for a period oftime prior to freezing.

A general discussion on the subject of frost action inconcrete is provided by Cordon (1966).

1.2--Mechanisms of frost actionPowers and his associates conducted extensive re-

search on frost action in concrete from 1933 to 1961.They were able to develop reasonable hypotheses to ex-plain the rather complex mechanisms.

Hardened cement paste and aggregate behave quitedifferently when subjected to cyclic freezing, and areconsidered separately.

1.2.1 Freezing in cement paste-In his early papers,Powers (1945, 1954, 1955, and 1956) attributed frostdamage in cement paste to stresses caused by hydraulicpressure in the pores, the pressure being due to resis-tance to movement of water away from the regions offreezing. It was believed that the magnitude of thepressure depended on the rate of freezing, degree ofsaturation and coefficient of permeability of the paste,and the length of the flow-path to the nearest place forthe water to escape. The benefits of entrained air wereexplained in terms of the shortening of flow-paths toplaces of escape. Some authorities still accept this hy-pothesis.

Later studies by Powers and Helmuth producedstrong evidence that the hydraulic pressure hypothesiswas not consistent with experimental results (Powers1956, 1975; Helmuth 196Oa, 1960b; Pickett 1953). Theyfound that during freezing of cement paste most of thewater movement is toward, not away from, sites offreezing, as had been previously believed. Also, the di-lations (expansions) during freezing generally decreasedwith increased rate of cooling. Both of these findingswere contrary to the hydraulic pressure hypothesis, andindicated that a modified form of a theory previouslyadvanced by Collins (1944) (originally developed to ex-plain frost action in soil) is applicable.

Powers and Helmuth pointed out that the water incement paste is in the form of a weak alkali solution.When the temperature of the concrete drops below thefreezing point, there is an initial period of supercool-ing, after which ice crystals will form in the larger cap-illaries. This results in an increase in alkali content inthe unfrozen portion of the solution in these capillar-ies, creating an osmotic potential which impels water inthe nearby unfrozen pores to begin diffusing into thesolution in the frozen cavities. The resulting dilution ofthe solution in contact with the ice allows furthergrowth of the body of ice (ice-accretion). When thecavity becomes full of ice and solution, any further ice-accretion produces dilative pressure which can cause thepaste to fail. When water is being drawn out of un-frozen capillaries, the paste tends to shrink, (Experi-ments have verified that shrinkage of paste, or con-

201.2R-4 MANUAL OF CONCRETE PRACTICE

Crete, occurs during part of the freezing cycle.)According to Powers, when the paste contains en-

trained air, and the average distance between air bub-bles is not too great, the bubbles compete with the capillaries for the unfrozen water and normally win thiscompetition. For a better understanding of the mecha-nisms involved, the reader is directed to the referencescited previously. Many researchers now believe thatstresses resulting from osmotic pressure cause most ofthe frost damage to cement paste.

In recent years, Litvan (1972) has further studiedfrost action in cement paste. Litvan believes that thewater adsorbed on the surface or contained in thesmaller pores cannot freeze due to the interaction be-tween the surface and the water. Because of the differ-ence in vapor pressure of this unfrozen and super-cooled liquid and the bulk ice in the surroundings ofthe paste system, there will be migration of water to lo-cations where it is able to freeze, such as the largerpores or the outer surface. The process leads to partialdesiccation of the paste and accumulation of ice increvices and cracks. Water in this location freezes, pry-ing the crack wider, and if the space fills with water inthe next thaw portion of the cycle, further internalpressure and crack opening results. Failure thus occurswhen the required redistribution of water cannot takeplace in an orderly fashion either because the amountof water is too large, i.e., high water-cement ratio forthe same level of saturation, the available time is tooshort (rapid cooling), or the path of migration is toolong (lack of entrained air bubbles). Litvan believes thatin such cases, the freezing forms a semi-amorphoussolid (noncrystalline ice) resulting in great internalstresses. Additional stresses can be created by the non-uniform moisture distribution.

There is general agreement that cement paste of ade-quate strength and maturity can be made completelyimmune to damage from freezing by means of en-trained air, unless special exposure conditions result infilling of the air voids. However, air entrainment alonedoes not preclude the possibility of damage of concretedue to freezing, since freezing phenomena in aggregateparticles must also be taken into consideration.

1.2.2 Freezing in aggregate particles-Most rockshave pore sizes much larger than those in cement paste,and Powers (1945) found that they expel water duringfreezing. The committee believes that the hydraulicpressure theory, previously described for cement paste,plays a major role in most cases.

Dunn and Hudec (1965) advanced the “ordered wa-ter” theory, which states that the principal cause of de-terioration of rock is not freezing but the expansion ofadsorbed water (which is not freezable); specific casesof failure without freezing of claybearing limestone ag-gregates seemed to support this conclusion. This, how-ever, is not consistent with the results of research byHelmuth (1961) who found that adsorbed water doesnot expand, but actually contracts during cooling.Nevertheless, Helmuth agrees that the adsorption oflarge amounts of water in aggregates with a very fine

pore structure can disrupt concrete through ice forma-tion.

The size of the coarse aggregate has been shown to bean important factor in its frost resistance. Uerbeck andLandgren (1960) have demonstrated that, when uncon-fined by cement paste, the ability of natural rock towithstand freezing and thawing without damage in-creases with decrease in size, and that there is a criticalsize below which rocks can be frozen without damage.They showed that the critical size of some rocks can beas small as a quarter of an inch (6 mm). However, someaggregates (e.g., granite, basalt, diabase, quartzite,marble) have capacities for freezable water so low thatthey do not produce stress when freezing occurs undercommonly experienced conditions-regardless of theparticle size.

Various properties related to the pore structurewithin the aggregate particles, such as absorption, po-rosity, pore size, and pore distribution or permeability,may be indicators of potential durability problemswhen the aggregates are used in concrete which will be-come saturated and freeze in service. Generally, it is thecoarse aggregate particles with relatively high porosityor absorption values, caused principally by

pm, that aremost easily saturated and contribute to deterioration ofconcrete individual popouts. Larger pores usually donot get completely filled with water, and thereforedamage is not caused on freezing. Water in very finepores may not freeze as readily (AC1 221R).

Fine aggregate is generally not a problem since theparticles are small enough to be below the critical sizefor the rock type and the entrained air in the surround-ing paste can provide an effective level of protection(Gaynor 1967).

The role of entrained air in alleviating the effect offreezing in coarse aggregate particles is minimal.

1.2.3. Overall effects in concrete-Without entrainedair, the paste matrix surrounding the aggregate parti-cles may fail when it becomes critically saturated and isfrozen. However, if the matrix contains an appropriatedistribution of entrained-air voids characterized by aspacing factor less than about 0.008 in. (0.20 mm),freezing does not produce destructive stress (Verbeck1978).

There are some rocks which contain practically nofreezable water. Air-entrained concrete made with anaggregate composed entirely of such rocks will with-stand freezing for a long time even under continuouslywet exposures. This time may be shortened if the airvoids fill with water and solid matter.

If absorptive aggregates (such as certain cherts andlightweight aggregates) are used and the concrete is in acontinuously wet environment, the concrete will prob-ably fail if the coarse aggregate becomes saturated(Klieger and Hanson l%l). The internal pressure de-veloped when the particles expel water during freezingruptures the particles and the matrix. If the particle isnear the concrete surface, a popout can result.

Normally, aggregate in concrete is not in a critical

medium-sized pore spaces in the range of 0.1 to 5

GUIDE TO DURABLE CONCRETE 201.2R-5

state of saturation near the end of the construction pe-riod because of desiccation produced by the chemicalreaction during hardening (self-desiccation of the ce-ment paste) and loss by evaporation. Therefore, if anyof the aggregate ever becomes critically saturated, it willbe by water obtained from an outside source. Struc-tures so situated that all exposed surfaces are kept con-tinuously wet, and yet are periodically subject to freez-ing, are uncommon. Usually concrete sections tend todry out during dry seasons when at least one surface isexposed to the atmosphere. That is why air-entrainedconcrete generally is not damaged by frost action evenwhere absorptive aggregate is used.

Obviously, the drier the aggregate is at the time theconcrete is cast, the more water it must receive to reachcritical saturation, and the longer it will take. This is animportant consideration, because the length of the wetand cold season is limited. It may prove a disadvantageto use gravel directly from an underwater source, espe-cially if the structure goes into service during the wetseason or shortly before the beginning of winter.

Some kinds of rock, when dried and then placed inwater, are able to absorb water rapidly and reach satu-ration quickly; they are described as the readily satura-ble type. This type, even when dry at the start, mayreach high levels of saturation while in a concretemixer, and might not become sufficiently dried by self-desiccation; hence, with such a material, trouble is inprospect if there is not a sufficiently long dry periodbefore the winter season sets in. A small percentage ofreadily saturable rocks in an aggregate can cause seri-ous damage. Rocks which are difficult to saturate,which are generally coarse grained, are less likely tocause trouble. Obviously, data on the proneness to sat-uration of each kind of rock in an aggregate could beuseful.1.3~Ice-removal agents

When the practice of removing ice from concretepavements by means of salt (sodium chloride and/orcalcium chloride) became common, it was soon learnedthat these materials caused or accelerated surface dis-integration in the form of pitting or scaling. (Thesechemicals also accelerate the corrosion of reinforce-ment, which can cause the concrete to spall, as de-scribed in Chapter 4.)

The mechanism by which deicing agents damageconcrete is fairly well understood. It is generally agreedthat the action is primarily physical rather than chemi-cal. The mechanism involves the development of dis-ruptive osmotic and hydraulic pressures during freez-ing, principally in the paste, similar to ordinary frostaction, which is described in Section 1.2. It is, how-ever, more severe.

The concentration of deicer in the concrete plays animportant role-in the development of these pressures.Verbeck and Klieger (1957) showed that scaling of theconcrete is greatest when ponded with intermediateconcentrations (3 to 4 percent) of deicing solutions.Similar behavior was observed for the four deicerstested: calcium chloride, sodium chloride, urea, and

ethyl alcohol. Browne and Cady (1975) drew similarconclusions. Litvan’s findings (1975, 1976) were con-sistent with the studies just mentioned. He furtherconcluded that deicing agents cause a high degree ofsaturation in the concrete, and that this is mainly re-sponsible for their detrimental effect. Salt solutions (ata given temperature) have a lower vapor pressure thanwater; therefore, little or no drying takes place betweenwetting (see Section 1.2.3) and cooling cycles. The useof ASTM C 672 will demonstrate the acceptability orfailure of a given concrete mixture.

The benefit from entrained air in concrete exposed todeicers is explained in the same way as for ordinaryfrost action. Laboratory tests and field experience haveconfirmed that air entrainment greatly improves resis-tance to deicers and is essential under severe condi-tions. It is now possible to consistently build scale-re-sistant pavements.

1.4--Recommendations for durable structuresConcrete that will be exposed to a combination of

moisture and cyclic freezing requires the following:1. Design of the structure to minimize exposure to

moisture.2. Low water-cement ratio.3. Appropriate air entrainment.4. Quality materials.5. Adequate curing before first freezing cycle.6, Special attention to construction practices.

These items are described in detail in the following par-agraphs.

1.4.1 Exposure to moisture-Since the vulnerabilityof concrete to cyclic freezing is influenced so greatly bythe degree of satuaration of the concrete, every precau-tion should be taken to minimize water uptake. Muchcan be accomplished along these lines by careful initialdesign of the structure.

The geometry of the structure should promote gooddrainage. Tops of walls and all outer surfaces should besloped. Low spots conducive to the formation of pud-dles should be avoided. Weep holes should not dis-charge over the face of exposed concrete. Drainagefrom higher ground should not flow over the top orfaces of concrete walls (Miesenhelder 1960).

Joints not necessary for volume change controlshould be eliminated and provisions for drainageshould be made. Drip beads can prevent water fromrunning under edges of structural members. “Watertraps” or reservoirs, such as may result from extendingdiaphragms to the bent caps of bridges, should beavoided during design.

Even though it is seldom possible to keep moisturefrom the underside of slabs on grade, subbase founda-tions incorporating the features recommended in AC1316R will minimize moisture buildup. Care should alsobe taken to minimize structural cracks which may col-lect or transmit water.

Extensive surveys of concrete bridges and otherstructures have shown a striking correlation betweenfreezing and thawing damage of certain portions, and

201.2R-6 MANUAL OF CONCRETE PRACTICE

excessive exposure to moisture of these portions due tothe structural design (Callahan, Lott, and Kesler 1970;Jackson 1946; Lewis 1956).

1.4.2 Water-cement ratio---Frost-resistant normal-weight concrete should have a water-cement ratio notexceeding the following: thin sections (bridge decks,railings, curbs, sills, ledges, and ornamental works) andany concrete exposed to deicing salts-0.45;* all otherstructures-0.50.*

Because the degree of absorption of some light-weight aggregates may be uncertain, it is impracticableto calculate the water-cement ratio of concretes con-taining such aggregates. For these concretes, a 2%daycompressive strength of at least 4000 psi (27.6 MPa)should be specified.

1.4.3 Entrained air--Too little entrained air will notprotect cement paste against freezing and thawing. Toomuch air will unduly penalize the strength. Recom-mended air contents of concrete are given in Table1.4.3.

Note that air contents are given for two conditions ofexposure-severe and moderate. These values provideabout 9 percent of air in the mortar fraction for severeexposure, and about 7 percent for moderate exposure.

Air-entrained concrete is produced through the use ofan air-entraining admixture added at the concretemixer, and air entraining cement, or both if necessary.The resulting air content depends on many factors in-cluding the properties.of the materials being used (ce-ment, chemical admixtures, aggregates, pozzolans,etc.), mixture proportions, types of mixer, mixing time,and temperature. Where an air-entraining admixture isused, the dosage is varied as necessary to give the de-sired air content. This is not possible where an air-en-training cement alone is used, and occasionally the aircontent will be inadequate or excessive. Nevertheless,this is the most convenient method for providing someassurance of protection from cyclic freezing on smalljobs where equipment to check the air content is notavailable. The preferred procedure is to use an air-en-training admixture.

Samples for air content determination should betaken as close to the point of placement as feasible.Frequency of sampling should be as specified in ASTMC .94. For normal-weight concrete, the following testmethods may be used: volumetric method (ASTMC 173), pressure method (ASTM C 231), or the unitweight test (ASTM C 138). The unit weight test (ASTMC 138) can be used to check the other methods. Forlightweight concrete, the volumetric method (ASTMC 173) should be used.

The air content and other characteristics of air-voidsystem in hardened concrete may be determined micro-scopically (ASTM C 457). AC1 212.3R lists the air voidcharacteristics required for durability. ASTM C 672provides a method to assess the resistance of concreteto deicer scaling.

*See Chapter 4 for further information when corrosion considerations arise.

Table 1.4.3 - Recommended air contents forfrost-resistant concrete

*A reasonable tolerance for ah content in f ield construction is f I’/, per-cant.

TOutdoor exposure in a cold climate where the concrete may be in almostcontinuous contact with moisture prior to freezing, or where deicing salts areused. Examples are pavements, bridge decks, sidewalks, and water tanks.

$ Outdoor exposure in a cold climate where the concrete will be only occa-sionally exposed to moisture prior to freezing, and where no deicing salts willbe used. Examples are certain exterior walls, beams, girders, and slabs not indirect contact with soil.

SThese air contents apply to the whole as for the preceding aggregate sizes.When testing these concretes, however, aggregate larger than 1 Vr in. (37.5 mm)is removed by handpicking or sieving and the air content is determined on theminus 1% in. (37.5 mm) fraction of the mixture. (The field tolerance applies tothis value.) From this the air content of the whole mixture is computed.

There is conflicting opinion on whether air contents lower than those givenin the table should be permitted for high strength (more about 5500 psi) (37.8MPa) concrete. This committee believes that where supporting experience andexperimental data exist for particular combinations of materials, constructionpractices and exposure, the air contents may be reduced by approximately 1Percent. (For nominal maximum aggregate sizes over I % in. (37.5 mm), thisreduction apphes to the minus I ‘A in. (37.5 mm) fraction of the mixture.

1.4.4 Materials1.4.4.1 Cementitious materials-The different

types of portland and blended hydraulic cements, whenused in properly porportioned and manufatured air-en-trained concrete, will provide similar resistance to cy-clic freezing. Cement should conform to ASTM C 150or C 595.

Most fly ashes and natural pozzolans when used asadmixtures have little effect on the durability of con-crete provided that the air content, strength, and mois-ture content of the concrete are similar. However, asuitable investigation should be made before using un-proven materials. Fly ashes and natural pozzolansshould conform to ASTM C 618. Ground granulatedblast-furnace slag should conform to ASTM C 989. Incontinental European countries (Belgium, the Nether-lands, France, and Germany) blast-furnace slag ce-ments have been used successfully for over a century inconcretes exposed to severe freezing and thawing envi-ronments, including marine exposures.

1.4.4.2 Aggregates-Natural aggregates shouldmeet the requirements of ASTM C 33, although thiswill not necessarily assure their durability. Lightweightaggregates should meet the requirements of ASTMC 330. These specifications provide many requirementsbut leave the final selection of the aggregate largely upto the judgment of the concrete engineer. If the engi-neer is familiar with the field performance of the ag-gregate proposed, his judgment may be quite adequate.In some situations, it is possible to carry out field serv-ice record studies to arrive at a basis for acceptance orrejection of the aggregate. When this is not feasible,heavy reliance must be placed on cautious interpreta-tions of laboratory tests.

GUIDE TO DURABLE CONCRETE 201.2R-7

Laboratory tests on the aggregate include absorp-tion, specific gravity, soundness, and determination ofthe pore structure. Descriptions of the tests, and opin-ions on their usefulness, have been published (Newlon1978; Buth and Ledbetter 1970). Although these dataare useful, and some organizations have felt justified insetting test hmits on aggregates, it is generally agreedthat principal reliance should be placed on tests onconcrete made with the aggregate in question.

Petrographic studies of both the aggregate (Mielenz1978) and concrete (Erlin 1966; K. Mather 1978a) areuseful for evaluating the physical and chemical charac-teristics of the aggregate and concrete made with it.

Laboratory tests on concrete include the rapid freez-ing and thawing tests (ASTM C 666), in which the du-rability of the concrete is measured by the reduction indynamic modulus of elasticity of the concrete. ASTMC 666 permits testing by either Procedure A - freezingand thawing in water, or Procedure B - freezing in airand thawing in water.

The results of tests using ASTM C 666 have beenwidely analyzed and discussed (Arni 1966; Buth andLedbetter 1970; AC1 221R; Transportation ResearchBoard 1959). These tests have been criticized becausethey are accelerated tests and do not duplicate condi-tions in the field. It has been pointed out that test spec-imens are initially saturated, which is not normally thecase for field concretes at the beginning of the winterseason. Furthermore, the test methods do not realisti-cally duplicate the actual moisture conditions of theaggregates in field concretes. The rapid methods havealso been criticized because they require cooling ratesgreater than those encountered in the field. Also, thesmall test specimens used are unable to accommodatelarger aggregate sizes proposed for use, which may bemore vulnerable to popout and general deteriorationthan smaller sizes. The presence of a piece of popoutproducing aggregate in the central portion of the rela-tively small test specimens can cause some of thesespecimens to fail, whereas the popout material wouldonly cause superficial surface defects in in-service con-crete (Sturrup et al. 1987).

It is generally conceded that while these various testsmay classify aggregates from excellent to poor in ap-proximately the correct order, they are unable to pre-dict whether a marginal aggregate will give satisfactoryperformance when used in concrete at a particularmoisture content and subjected to cyclic freezing expo-sure. The ability to make such a determination wouldbe of great economic importance in many areas wherehigh-grade aggregates are in short supply, if the use oflocal marginal aggregates could be permitted. Despitethe shortcomings of ASTM C 666, many agencies be-lieve that this is the most reliable indicator of the rela-tive durability of an aggregate (Sturrup et al. 1987).

Because of these objections to ASTM C 666, a dila-tion test was conceived by Powers (1954) and furtherdeveloped by others (Harman, Cady, and Bolling 1970;Tremper and Spellman 1961). ASTM C 671 requiresthat air-entrained concrete specimens be brought ini-

tially to the moisture condition expected for the con-crete at the start of the winter season, this moisturecontent preferably having been determined by fieldtests. The specimens are then immersed in water andperiodically frozen at the rate to be expected in thefield. The increase in length (dilation) of the specimenduring the freezing portion of the cycle is accuratelymeasured. ASTM C 682 assists in interpreting the re-sults.

Excessive length change in this test is an indicationthat the aggregate has become critically saturated andvulnerable to’damage. If the time to reach critical sat-uration is less than the duration of the freezing seasonat the job site, the aggregate is judged unsuitable foruse in that exposure. If it is more, it is judged that theconcrete will not be vulnerable to cyclic freezing.

The time required for conducting a dilation test maybe greater than that required to perform a test byASTM C 666. Also, the test results are very sensitive tothe moisture content of the aggregate and concrete.Results to date, however, are fairly promising. Al-though most agencies are continuing to use ASTMC 666, results from ASTM C 671 may turn out to bemore useful (Philleo 1986).

When a natural aggregate is found to be unaccepta-ble by service records or tests or both, it may some-times be improved by removal of lightweight, soft, orotherwise inferior particles.

1.4.4.3 Admixtures-Air-entraining admixturesshould conform to ASTM C 260. Chemical admixturesshould conform to ASTM C 494. Admixtures for flow-ing concrete should conform to ASTM C 1017.

Some mineral admixtures (including pozzolans) andaggregates containing large amounts of fines may re-quire a larger amount of air-entraining admixture todevelop the required amount of entrained air.

Detailed guidance on the use of admixtures is pro-vided by AC1 212.313.

1.4.5 Maturity-Air-entrained concrete should beable to withstand the effects of freezing as soon as itattains a compressive strength of about 500 psi (3.45MPa) provided that there is no external source ofmoisture. At a temperature of 50 F (10 C), most well-proportioned concrete will reach this strength sometime during the second day.

Before being exposed to extended freezing while crit-ically saturated (see ASTM C 666), the concrete shouldattain a compressive strength of about 4000 psi (27.6MPa). A period of drying following curing is advisa-ble. For moderate exposure conditions, a strength of3000 psi (20.7 MPa) should be attained (Kleiger 1956).

tion practices-Good constructionpractices are essential when durable concrete is re-quired.

Particular attention should be given to the construc-tion of pavement slabs that will be exposed to deicingchemicals because of the problems inherent in obtain-ing durable slab finishes, and the severity of the expo-sure. The concrete in such slabs should be adequatelyconsolidated; however, overworking the surface, over-

201.2R-I) MANUAL OF CONCRETE PRACTICE

finishing, and the addition of water to aid in finishingmust be avoided. These activities bring excessive mor-tar or water to the surface, and the resulting laitance isparticularly vulnerable to the action of deicing chemi-cals. These practices may also remove entrained airfrom the surface region. This is of little consequence ifonly the larger air bubbles are expelled, but durabilitycan be seriously affected if the small bubbles are re-moved. Timing of finishing is critical (see AC1 302.lR).

Prior to the application of any deicer, pavementconcrete should have received some drying, and thestrength level specified for the opening of traffic shouldbe considered in the scheduling of late fall paving. Insome cases, it may be possible to employ methods otherthan ice-removal agents, such as abrasives, for controlof slipperiness where the concrete may still be vulnera-ble.

Where lightweight concrete is proposed, care shouldbe exercised not to wet the aggregate excessively priorto mixing. Saturation by vacuum or thermal means(where necessary for pumping, for example) may bringlightweight aggregates to a moisture level at which theabsorbed water will cause concrete failure when it iscyclically frozen unless the concrete has the opportu-nity to dry out before freezing. Additional details andrecommendations are given in a publication of the Cal-ifornia Department of Transportation (1978).

CHAPTER 2 - AQQRESSIVE CHEMICALEXPOSURE

2.1 -GeneralConcrete will perform satisfactorily when exposed to

various atmospheric conditions, to most waters and

soils containing aggressive chemicals, and to manyother kinds of chemical exposure. There are, however,some chemical environments under which the useful lifeof even the best concrete will be short, unless specificmeasures are taken. An understanding of these condi-tions permits measures to be taken to prevent deterio-ration or reduce the rate at which it takes place.

Concrete is rarely, if ever, attacked by solid, drychemicals. To produce significant attack on concrete,aggressive chemicals must be in solution and abovesome minimum concentration. Concrete which is sub-jected to aggressive solutions under pressure on oneside is more vulnerable than otherwise, because thepressures tend to force the aggressive solution into theconcrete.

Comprehensive tables have been prepared by AC1Committee 515 (515.1R) and the Portland Cement As-sociation (1968) giving the effect of many chemicals onconcrete. Biczok (1972) gives a detailed discussion ofthe deteriorating effect of chemicals on concrete, in-cluding data both from Europe and the United States.

The effects of some of the more common chemicalson the deterioration of concrete are summarized in Ta-ble 2.1. Provided that due care has been taken in selec-tion of the concrete materials and proportioning of theconcrete mixture, the most important factors which in-fluence the ability of concrete to resist deterioration areshown in Table 2.2. Therefore Table 2.1 should beconsidered as only a preliminary guide.

Consult appropriate AC1 Standards and CommitteeReports for the procedures to be followed for the typeof construction involved.

Major areas of concern are exposure to sulfates, sea-

Table 2.1 - Effect of commonly used chemicals on concreteRate of

attack atambient

temperature

Rapid

Moderate

Slow

Negligible

HydrochloricHydrochloricNi tr icSulfur ic

Phosphoric

Carbonic

*The effect of potassium hydroxide iisr‘Avoid siliceous aggregates bccm~r I

similar to that of sodium hydroxide.h ey arc attacked by strong solutions OF sodium hydroxide.

Organicacids

AceticFormicLactic

Alkalinesolutions

saltsolutions

Aluminumchloride

Ammoniumnitrate

Ammoniumsulfate

Tannic Sodium*hydroxide

> 20 percent’

Sodiumsulfate

Magnesiumsulfate

Calciumsulfate

Sodium*hydroxide

10-;otipee;ent’

hypochlorite

Ammoniumchloride

MagnesiumchlorideSodiumcyanide

Miscellaneous

-

Bromine (gas)Sulfite liquor

fpr;Er (gas)Softwater

Ammonia(liquid)

GUIDE TO DURABLE CONCRETE 201.2Fb9

Table 2.2 - Factors influencing chemical attackon concrete

Factors which accelerateor aggravate attack

1. High porosity due to:i. High water absorptionii. Permeabitity

iii. Voids

2. Cracks and separations due to:

i. Stress concentrationsii. Thermal shock

3. Leaching and liquidpenetration due to:

i. Flowing liquid’ii. Ponding

iii. Hydraulic pressure

*The mixture proportions and the initCrete determine its homogeneity and der

IPoor curing procedures result in flaw.

ialIsit:1s a

Factors which mitigateor delay attack

I. Dense concrete achieved by:i. Proper mixture

proportioning*ii. Reduced unit water

contentiii. Increased cementitious

material contentiv. Air entrainmentv. Adequate consolidation

vi. Effective curing+

!. Reduced tensile stress inconcrete by:’

i. Using tensilereinforcement ofadequate size, correctlylocated

ii. Inclusion of pozzolan(to suppress temperaturerise)

iii. Provision of adequatecontraction ioints

I. Structural design

i. To minimize areas ofcontact and turbulence

ii. Provision of membranesand protective-barriersystem(s) ** to reducepenetration

mixing and processing of fresh con-Y.nd cracks.

‘Keslstance to crackmg depends on strength and stram capacity.*Movement of water-carrying deleterious substances increases reactions which

depend on both the quantity and velocity of flow.**Concrete which will be frequently exposed to chemicals known to produce

rapid deterioration should be protected with a chemically resistant protective-barrier system.

water, and acids, and carbonation. These are discussedin detail in the following sections.

2.2-Sulfate attack2.2.1 Occurrence-Naturally occurring sulfates of

sodium, potassium, calcium, or magnesium which canattack concrete are sometimes found in soil or dis-solved in groundwater adjacent to concrete structures.When evaporation takes place from an exposed face,the sulfates may accummulate at that face, thus in-creasing their concentration and potential for causingdeterioration. Sulfate attack has occurred at variouslocations throughout the world, and is a particularproblem in arid areas, such as the northern GreatPlains area of the United States, the prairie provincesof Canada (Swenson 1968; Reading 1975), parts of thewestern United States (U.S. of Bureau of Reclamation1975) and the Middle East (French and Pool 1976). Sixuseful papers are given in Klieger 1982.

The water used in concrete cooling towers can also bea potential source of sulfate attack because of thegradual build-up of sulfates from evaporation, partic-ularly where such systems use relatively small amountsof make-up Gater. Sulfates are also present in ground-water and in fill containing industrial waste productssuch as cinders.

2.2.2 Mechanism-As Lea (1971), Mehta (1976),and others pointed out, there are apparently two chem-ical reactions involved in sulfate attack on concrete:

1. Combination of sulfate with calcium ions liber-ated during the hydration of the cement to form gyp-sum (CaSO, * 32H20).

2. Combination of sulfate ion and hydrated calciumaluminate to form calcium sulfoaluminate (ettringite)(3CaO*Al,O,-3CaSO,.3H,O).

Both of these reactions result in an increase in solidvolume. The formation of ettringite is the cause ofmost of the expansion and disruption of concretescaused by sulfate solutions.

The chemical deterioration of concrete in seawater,which contains sulfates, has concerned concrete tech-nologists for generations (Lea 1971; Gjorv 1957; Idorn1958, 1967). It is generally thought that seawater, whichalso contains chlorides, is not as severe an exposure assulfate groundwater. Seawater is discussed in greaterdetail in a later section of this chapter.

2.2.3 Recommendations-Protection against sul-fate attack is obtained by using a dense, quality con-crete with a low water-cement ratio, and concrete-mak-ing ingredients appropriate for producing concretehaving the needed sulfate resistance. Air entrainment isof benefit insofar as it reduces the water-cement ratioand, hence, the permeability (Verbeck 1968). Propercuring, in accordance with AC1 308, is essential toachieve minimal permeability.

There is reasonably good correlation between thesulfate resistance of cement and its calculated trical-cium aluminate (C,A) content (B. Mather 1968). Ac-cordingly, ASTM C 150 includes Type V (sulfate-re-sisting) cement for which a maximum of 5 percentcalculated C,A is permitted and Type II (moderatelysulfate-resisting) cement for which the calculated C,A islimited to 8 percent. There is also some evidence thatthe alumina in the aluminoferrite phase of portland ce-ment may participate in delayed sulfate attack. HenceASTM C 150 provides that in Type V cement the C,AF+ 2C,A must not exceed 25 percent, unless the alter-nate requirement based on the use of the performancetest (ASTM C 452) is invoked. In the case of Type Vcement, the sulfate expansion test (ASTM C 452) maybe used in lieu of the chemical requirements (K. Mather1978b).

Recommendations for the type of cement and water-cement ratio for normal-weight concrete which will beexposed to sulfates in soil, groundwater, or seawaterare given in Table 2.2.3. The values also apply to areasin the splash or spray zone. These values are also appli-cable to structural lightweight concrete except that themaximum water-cement ratios of 0.50 and 0.45 shouldbe replaced by specified 28-day strengths of 3750 and4250 psi (25.8 and 29.4 MPa), respectively.

Studies have shown that some pozzolans and groundgranulated iron blast-furnace slags, used either inblended cement or ded separately to the concrete inthe mixer increase the life expectancy of concrete insulfate exposure considerably. In addition to the bene-ficial effects that many slags and pozzolans have on thepermeability of concrete (Bakker 1980; Mehta 1981)they also combine with the alkalies and calcium hy-

201,2R-10 MANUAL OF CONCRETE PRACTICE

droxide released during the hydration of the cement(Vanden Bosch 1980; Roy and Idorn 1982; Idom andRoy 1986) and thus reduce the potential for gypsumformation (Biczok 1972; Lea 1971; Mehta 1976; Kal-ousek, Porter, and Benton 1972).

Table 2.2.3 requires a suitable pozzolan or slag alongwith Type V cement in very severe exposures. Researchhas indicated that some pozzolans and slags may be ef-fective in improving the sulfate resistance of concretemade with Type I and Type II cement. However, somepozzolans, especially some Class C fly ashes, decreasethe sulfate resistance of mortars in which they are used(K. Mather 1981, 1982). Good results were obtainedwhen the pouolan was a fly ash meeting the require-ments of ASTM C 618 Class F (Dikeou 1975; Dunstan1976). Slag should meet ASTM C 989.

Experience in Europe and South Africa shows excel-lent durability of concrete in seawater when ground slagis used in blends with portland cements of Type I orIII. The often low-alkali contents of Type V cementsare in fact disadvantageous for the activation of theslag fraction (or pozzolan) in the blended cements (Royand Idorn 1982; Idorn and Rostam 1983; Bhatty andGreening 1978). Actual tests of the combination shouldbe made using accelerated test procedures such as thosedescribed by Dikeou (1975); information from long-term field performance in structures of field exposurestations should also be considered where available.

Calcium chloride in concrete reduces its resistance toattack by sulfate (U.S. Bureau of Reclamation 1975)and its use should be prohibited in concrete exposed tosevere and very severe environments (Table 2.2.3).

It is recognized that these recommendations are con-servative, and area intended to insure long service lifeof concrete. Less stringent requirements are appropri-ate when justified by local experience or for shorter de-sired service life.

2.3-Seawater exposure2.3.1 Seawater may be encountered with a range of

concentrations of dissolved salts, though always with aconstant proportion of the constituents to one another.

The concentration is lower in the colder and temper-ate regions than in the warm seas, and especially highin shallow coastal areas with excessive diurnal evapo-ration rates.

Where concrete structures are placed on reclaimedcoastal areas with the foundations below the salineground water levels, capillary suction and evaporationmay cause supersaturation and crystallization in theconcrete above ground, resulting both in chemical at-tack on the cement paste (sulfate), and in aggravatedcorrosion of steel (chlorides).

In tropical climates these combined deleterious ef-fects may cause severe defects in concrete in the courseof a very few years.

2.3.2 The reaction of mature concrete with the sul-fate ion in seawater is similar to that with sulfate ion infresh water or leached from soils, but the effects are

Table 2.2.3 - Recommendetlons for normalweight concrete subject to sulfate attack

Severe 0.20-2.00 1500-10.000very atvtrt over 2.00 Over 10.000

*Sulfate expressed as SO, is ~&II* to @fate expressed as SO, as in reportsof chemical analysis of cement as SO, x I .Z = SO,.

‘When chlorides or other dcpsssivating agents are present in addition to sul-fate, a lower water-cement ratio may be necessary to reduce corrosion poten-tial of embedded items. Refer to Chapter 4.

:Or a blend of Type I cement and a ground granulated blast furnace slag ora oozxolan that has been determined by tests to give equivalent sulfate r&s-tance.

‘Or a bknd of Type II cement and a ground granulated blast furnace slag ora pouolan that has been determined by tests to give equivalent sulfate rcsis-tance.

**Use a poxxolan or slag that has been determined by tests to improve sul-fate resistance when used in concrete containing Type V cement.

different (B. Mather 1966). The concentration of sul-fate ion in seawater can be increased to high levels bycapillary action and evaporation under extreme cli-matic conditions. The presence of chloride ions, how-ever, alters the extent and nature of the chemical reac-tion so that less expansion is produced by a cement ofgiven calculated C,A content than would be expected ofthe same cement in a fresh water exposure where thewater has the same sulfate ion content. The perform-ance of concretes continuously immersed in seawatermade with ASTM C 150 cements having C,A contentsas high as 10 percent have proven satisfactory providedthe permeability of the concrete is low (Browne 1980).The Corps of Engineers (1985) permits, and the Port-land Cement Association recommends, up to 10 per-cent calculated C,A for concrete that will be perma-nently submerged in seawater if the water-cement ratiois kept below 0.45 by mass.

Verbeck (1968) and Regourd et al. (1980) showed,however, that there may be considerable difference be-tween the calculated and the measured clinker compo-sition of cement, especially as far as C,A and C,AF areconcerned. Therefore, the interrelation between themeasured C,A content and the seawater resistance maybe equally uncertain.

2.3.3 The requirement for low permeability is essen-tial not only to delay the effects of sulfate attack, butalso to afford adequate protection to reinforcementwith the minimum concrete cover as recommended byAC1 Committee 357 (357.1 R) for exposure to seawater.The required low permeability is attained by using con-crete with a low water-cement ratio, well consolidatedand adequately cured.

The permeability of concrete made with appropriateamounts of suitable ground blast-furnace slag or poz-zolan can be as low as 100th or 1000th that of com-parable concrete of equal strength made without slag orpozzolan (Bakker 1980). The satisfactory performanceof concretes containing ground slag in a marine envi-

GUIDE TO DURABLE CONCRETE 201.2Fb11

ronment has been described (B. Mather 1981; VandenBosch 1980; Lea 1971).

Careful attention must also be given to structuralaspects with properly designed and constructed joints toassure that cracking is minimized to prevent the expo-sure of reinforcement. Additionally, concrete shouldreach a maturity equivalent of not less than 5000 psi (35MPa) at 28 days when fully exposed to seawater.

Conductive coatings applied at the time of construc-tion as part of a cathodic protection system may pro-vide additional protection for concrete which ispartially submerged or reaches down to saline ground-water. Silane coatings, which are water-repellent, haveshown excellent protection characteristics.

Caution is required with the application of coatingswhich significantly restrict evaporation of free waterfrom the interior of concrete and thus reduce resistanceto freezing and thawing.

Marine structures often involve thick sections andrather high cement factors. Such concrete may need tobe treated as “mass concrete,” i.e., concrete in whichthe effect of the heat of hydration needs to be consid-ered. When this is the case, the recommendations ofAC1 207.1R, 207.2R, and 224R should be followed.2.4-Acid attack

In general, portland cement does not have good re-sistance to acids, although some weak acids can be tol-erated, particularly if the exposure is occasional.

2.4.1 Occurrence-The products of combustion ofmany fuels contain sulfurous gases which combine withmoisture to form sulfuric acid. Also, sewage may becollected under conditions which lead to acid forma-tion.

Water draining from some mines, and some indus-trial waters, may contain or form acids which attackconcrete.

Peat soils, clay soils, and alum shale may containiron sulfide (pyrite) which, upon oxidation, producessulfuric acid. Further reaction may produce sulfatesalts, which can produce sulfate attack (Hagerman andRoosaar 1955; Lossing 1966; Bastiansen, Mourn, andRosenquist 1957; Mourn and Rosenquist, 1959).

Mountain streams are sometimes mildly acidic, dueto dissolved free carbon dioxide. Usually these watersattack only the surface if the concrete is of good qual-ity and has a low absorption. However, some mineralwaters containing large amounts of either dissolvedcarbon dioxide or hydrogen sulfide, or both, can seri-ously damage any concrete (RILEM 1962; Thornton1978). In the case of hydrogen sulfide, bacteria thatconvert this compound to sulfuric acid may play animportant role (RILEM 1962).

Organic acids from farm silage, or from manufac-turing or processing industries such as breweries, dair-ies, canneries; and wood-pulp mills, can cause surfacedamage. This may be of considerable concern in thecase of floors, even where their structural integrity isnot impaired.

2.4.2 Mechanism--The deterioration of concrete byacids is primarily the result of a reaction between these

chemicals and the calcium hydroxide of the hydratedportland cement. (Where limestone and dolomitic ag-gregates are used, they are also subject to attack by ac-ids.) In most cases, the chemical reaction results in theformation of water-soluble calcium compounds whichare then leached away by the aqueous solutions (Biczok1972). Oxalic and phosphoric acid are exceptions, be-cause the resulting calcium salts are insoluble in waterand are not readily removed from the concrete surface.

In the case of sulfuric acid attack, additional or ac-celerated deterioration results because the calcium sul-fate formed will affect concrete by the sulfate attackmechanism described in Section 2.2.2.

If acids, chlorides, or other aggressive or salt solu-tions are able to reach the reinforcing steel throughcracks or pores in the concrete, corrosion of steel canresult (see Chapter 4) which will in turn cause crackingand spalling of the concrete.

2.4.3 Recommendations-A dense concrete with lowwater-cement ratio may provide an acceptable degree ofprotection against mild acid attack. Certain pozzolanicmaterials, and silica fume in particular, increase the re-sistance of concrete to acids (Sellevold and Nilson1987). In all cases, however, exposure time to acidsshould be minimized, if possible, and immersion shouldbe avoided.

No hydraulic-cement concrete, regardless of its com-position, will long withstand water of high acid con-centration (pH of 3 or lower). In such cases, an appro-priate protective-barrier system or treatment must beused. AC1 515.1R gives recommendations for barriersystems to protect concrete from various chemicals.Chapter 7 of this guide discusses the general principlesinvolved in the use of such systems.

PS-Carbonation2.5.1 When concrete or mortar is exposed to carbon

dioxide, a reaction producing carbonates takes placewhich is accompanied by shrinkage.

Virtually all the constituents of hydrated portlandcement are susceptible to carbonation. The results canbe either beneficial or harmful depending on the time,rate, and extent to which they occur and the environ-mental exposure. On the one hand, intentional carbon-ation during production may improve the strength,hardness, and dimensional stability of concrete prod-ucts. In other cases, however, carbonation may result indeterioration and a decrease in the pH of the cementpaste leading to corrosion of reinforcement near thesurface. Exposure to carbon dioxide during the hard-ening process may affect the finished surface of slabs,leaving a soft, dusting, less wear-resistant surface.During the hardening process, the use of unventedheaters, or exposure to exhaust fumes from equipmentor other sources, can produce a highly porous surfacesubject to further chemical attack.

The source of the carbon dioxide can be either theatmosphere or water carrying dissolved CO*.

2.5.2 Atmospheric Carbonation-Reaction of hyrdrated portland cement with carbon dioxide in the air

201*2R-12 MANUAL OF CONCRETE PRACTICE

is generally a slow process (Ludwig 1980). It is highlydependent-on the relative humidity of the environment,temperature, permeability of the concrete, and concen-tration of CO1. Highest rates of carbonation occurwhen the relative humidity is maintained between 50and 75 percent. Below 25 percent relative humidity, thedegree of carbonation that takes place is considered in-significant (Verbeck 1958). Above 75 percent relativehumidity, moisture in the pores restricts CO, penetra-tion.

Relatively permeable concrete undergoes more rapidand extensive carbonation than dense, well-consoli-dated, and cured concrete. Lower water-cement ratiosand good consolidation also serve to reduce permeabil-ity and restrict carbonation to the surface. Concrete inindustrial areas with higher concentrations of CO1 inthe air is more susceptible to attack.

2.5.3 Carbonation by ground w&r-Carbon diox-ide absorbed by rain enters the ground water as car-bonic acid. Additional carbon dioxide together withhumic acid can be dissolved from decaying vegetationresulting in high levels of free CO,. While such watersare usually acid, the aggressivity cannot be determinedby pH alone. Reaction with carbonates in the soil pro-duce an equilibrium with calcium bicarbonate whichcan result in solutions with a neutral pH but containingsignificant amounts of aggressive CO, (Lea 1971).

The rate of attack, similar to that by CO2 in the at-mosphere, is dependent upon the quality of the con-crete and concentration of the aggressive carbon diox-ide. There is no consensus at this time as to limitingvalues because of widely varying conditions in under-ground construction. It has been concluded in somestudies, however, that water containing more than 20parts per million (ppm) of aggressive CO, can result inrapid carbonation of the hydrated cement paste. On theother hand, freely moving waters with 10 ppm or lessof aggressive CO, may also result in significant carbon-ation (Terzaghi 1948 and 1949).

CHAPTER 3-ABRASION3.1 -Introduction

The abrasion resistance of concrete is defined as the“ability of a surface to resist being worn away by rub-bing and friction” (AC1 116R). Abrasion of floors andpavements may result from production operations, orfoot or vehicular traffic; abrasion resistance is there-fore of concern in industrial floors (Love11 1928). Windor waterborne particles can also abrade concrete sur-faces (Price 1947). There are instances where abrasionis of little concern structurally, yet there may be a dust-ing problem which can be quite objectionable in somekinds of service. Abrasion (erosion) of concrete in hy-draulic structures is discussed only briefly in this guide;the subject is treated in detail in AC1 210R.

3.2-Testing concrete for resistance to abrasionResearch to develop meaningful laboratory tests on

concrete abrasion has been underway for more than a

century. The problem is complicated because there areseveral different types of abrasion, and no single testmethod has been found which is adequate for all con-ditions. Four general areas should be considered (Prior1966):

1. Floor and slab construction. Table 1. l-FloorClassifications, AC1 302.1R. Classes of wear are desig-nated and special considerations required for good wearresistance indicated. (Table 1.1 is reproduced here asTable 3.2.)

2. Wear on concrete road surfaces due to heavytrucks and automobiles with studded tires or chains(attrition, scraping, and percussion).

3. Erosion of hydraulic structures such as dams,spillways, tunnels, bridge piers, and abutments, due tothe action of abrasive materials carried by flowing wa-ter (attrition and scraping).

4. Cavitation action on concrete in dams, spillways,tunnels, and other water-carrying systems causes ero-sion where high velocities and negative pressures arepresent. This damage can best be corrected by changesin design which are not covered in this guide.

ASTM C 779 covers three operational procedures forevaluating floor surfaces. Procedure A-revolving discs(based on Schuman and Tucker 1939); Procedure B-dressing wheels; and Procedure C-ball bearings.

Each method has been used to develop informationon wear resistance. Prior (1966) commented that themost reliable method uses revolving discs. Reproduci-bility of abrasion testing is an important factor in se-lecting the test method. Replication of results is neces-sary to avoid misleading data from single specimens.

The quality of the concrete surface, the aggregatesused that are abraded during the test procedure, theprocedure itself, and the care and selection of represen-tative samples will affect test results. Samples that arefabricated in the laboratory must be made in identicalfashion for proper comparison and the selection of sitesfor field testing must be made with extreme care toprovide representative results.

It is not possible to set precise limits for abrasion re-sistance of concrete. It is necessary to rely on relativevalues based on test results that will provide a predic-tion of service.

Underwater abrasion presents special demands fortest procedures. A Standard Test Method for AbrasionResistance of Concrete (underwater method), ASTMC 1138, utilizes agitation of steel balls in water to de-termine abrasion resistance.

3.3-Factors affecting abrasion resistance ofconcrete

The abrasion resistance of concrete is a progressivephenomenon. Initially, resistance is closely related tostrength at the wearing surface and floor wear is bestjudged on this basis. Resistance can be modified by theuse of shakes and toppings, finishing techniques, andcuring procedures.

As the paste wears, the fine and coarse aggregate areexposed and abrasion and impact will cause additional

GUIDE TO DURABLE CONCRETE

Table 3.2-Floor classifications (Table 1.1 in ACI 302.1 R)

201.2&13

Usualtraffic

Concrete finishingtechnique

(see Chapter 7)

Medium steeltrowel

Specialconsiderations

Grade for drainage;make plane for tile

Class UseResidential ortile covered

Light foot

Foot Nonslip aggregate;mix in surface

Steel trowel;special finish fornonslip

Offices,churches,schools,hospitals

Ornamentalresidential

Color shake,special

Steel trowel,color, exposedaggregate; wash ifaggregate is to beexposed

Float, trowel, andb room

SingleCourse

Light foot andpneumat icwheels

Foot andpneumat icwheels*

Drives, garagefloors andsidewalks forresidences

Light industrialcommercial

Crown; pitch;joints; airentrainment

Careful curing Hard steel troweland brush fornonslip

irei$ metallic or

aggregate, floatand trowel

Foot andwheels -abrasive wear+

Single-courseindustrial,integraltoodna

Careful curing

6

7

asive cobd f o r Clolithic

Bonded two-course heavyindustrial

Base:Textured surfaceand bond

Surface leveled byscreeding

Foot and hardwheel vehicles- severeabrasion

Classes3, 4, 5, 6

,itions on floor sumES 4 and 5 floors..face treatment is r

Topping:Special aggregate,and/or mineral ormetallic surface

Special powerfloats withrepeated steeltroweling

TwoCourse

*Underwill be reqaggregate I

treatment

Unbondedtoppings

Mesh reinforcing;bonded breaker onold concretesurface; minimumthickness 2% in.(nom. 64 mm) 1_

td a higher quality surface. . . .the exposure will t. . K much more severe ar.̂ .ter tnese conamons a uass 0 two-course ttoor or a mmerat or metartlcmmended.

degradation that is related to aggregate-to-paste bondstrength and hardness of the aggregate.

Tests (Scripture, Benedict, and Bryant 1953; Witteand Backstrom 1951) and field experience have gener-ally shown that compressive strength is the most im-portant single factor controlling the abrasion resistanceof concrete, with abrasion resistance increasing withincrease in compressive strength. Since abrasion occursat the surface, it is critical that the surface strength bemaximized.

Reliance should not be placed solely on test‘ cylindercompressive strength results but careful inspectionshould be given the installation and finishing of thefloor surface (Kettle and Sadegzaden 1987).

With a given concrete mixture, compressive strengthsat the surface can be improved by:

1. Avoiding segregation.2. Eliminating bleeding.3. Properly timed finishing.4. Minimizing surface water-cement ratio, i.e., for-

bid any water addition to the surface to aid finishing.5. Hard toweling of the surface.6. Proper curing procedures.

Proportioning of the mixture for optimum compres-sive strengths at an economic level will include use of aminimum water-cement ratio and proper aggregate sizefor the strength requirements.

Consideration must be given to the quality of the ag-gregate in the surface region (Scripture, Benedict, andBryant 1953; Smith 1958). The service life of someconcrete, such as warehouse floors subjected to abra-sion by steel or hard rubber wheeled traffic, may begreatly lengthened by the use of a specially hard ortough aggregate.

High-quality aggregates may be employed either bythe dry-shake method or as part of a high-strength top-ping mixture. If abrasion is the principal concern, ad-dition of high-quality quartz, traprock, or emery ag-gregates properly proportioned with cement willincrease the wear resistance by improving the compres-sive strength at the surface. For additional abrasion re-sistance, a change to a blend of metallic aggregate andcement will increase the abrasion resistance further andprovide additional surface life.

The use of two-course floors employing a high-strength topping is generally limited to floors where

201.2Rd4 MANUAL OF CONCRETE PRACTICE

both abrasion and impact are destructive effects at thesurface. While providing excellent abrasion resistance,a two-course floor will generally be more expensive andis justified only when impact is a consideration. Addi-tional impact resistance can be obtained by using atopping containing portland cement and metallic ag-gregate.

A key element in production of a quality floor sur-face is curing (Prior 1966; AC1 302. IR; AC1 308). Sincethe uppermost part of the surface region is that portionabraded by traffic, maximum strength must be achievedat the surface. This is partially accomplished throughproper timing of the finishing operation, hand trowel-ing, and adequate curing.

3.40Recommendations for obtaining abrasion=resistant concrete surfaces

The following measures will result in abrasion-resis-tant concrete surfaces.

3.4.1 Choose the appropriate concrete strength byreference to AC1 302.1R, Table 5.2.1 - “Recom-mended slump and strength for each class of concretefloor.” Floor classes relate to those described in Table1.1 - “Floor classifications.” Compressive strengthlevels may be attained in a number of ways.

1. A low water-cement ratio at the surface. Use ofwater-reducing admixtures, a mixture proportioned toeliminate bleeding, and timing of finishing to avoid theaddition of water during troweling. Vacuum dewater-ing may be a valid option.

2. Proper grading of fine and coarse aggregate(meeting ASTM C 33). The maximum size of coarseaggregate should be chosen for optimum workabilityand minimum water content.

3. Use the lowest slump consistent with properplacement and consolidation meeting the requirementsof “Guide for Consolidation of Concrete,” AC1 309R.Proportion the mixture for the desired slump and toachieve the required strength.

4. Air contents should be consistent with exposureconditions. For indoor floors not subjected to freezingand thawing, air contents of 3 percent or less are pref-erable. In addition to a detrimental effect on strengths,high air contents can cause blistering if finishing is im-properly timed. Entrained air should not be used whenusing dry shakes unless special precautions are fol-lowed.

3.4.2 Two-course floors-High-strength toppings inexcess of 6000 psi (40 MPa) will provide increased ab-rasion resistance using locally available aggregate. Nor-mally, the nominal maximum aggregate size in a top-ping is 12.5 mm (l/2 in.).

3.4.3 Special concrete aggregates-selection of ag-gregates for improved strength at a given water-cementratio will also improve abrasion resistance. These arenormally applied as a dry shake or in a high-strengthtopping.

3.4.4 Proper finishing procedures-Delay floatingand troweling until the concrete has lost its surface wa-

ter sheen. It may be necessary to remove free waterfrom the surface to permit proper finishing before thebase concrete hardens. Do not finish concrete withstanding water since this will radically reduce the com-pressive strength at the surface. The delay period willvary greatly depending on temperature, humidity, andthe movement of air. More complete finishing recom-mendations are included in AC1 302.1 R

3.4.5 Vacuum dewatering-Vacuum dewatering is amethod for removing water from concrete immediatelyafter placment (New Zealand Portland Cement Associ-ation 1975). While this permits a reduction in water-ce-ment ratio, the quality of the finished surface is stillhighly dependent on the timing of finishing and subse-quent actions by the contractor. Care should be takento insure that proper dewatering is accomplished at theedges of the vacuum mats. Improperly dewatered areaswill be less resistant to abrasion due to higher water-ce-ment ratios.

3.4.6 Special dry shakes and toppings-When severewear is anticipated, the use of special dry shakes ortoppings should be considered. For selection, the rec-ommendations found in AC1 302.1 R should be fol-lowed.

3.4.7 Proper curing procedures-For most concretefloors, curing by keeping the concrete continuously wetis the most effective method of producing hard, densesurfaces. Water curing, however, may not be the mostpractical method. Curing compounds, which seal mois-ture in the concrete, may be used.

Water curing may be by sprays, damp burlap, orcotton mats. Water-resistant paper or plastic sheets aresatisfactory provided the concrete is first sprayed withwater and then immediately covered with the sheets,with the edges overlapped and sealed with water-resis-tant tape.

Curing compounds should meet ASTM C 309 at thecoverage rate used, and should be applied in a uniformcoat immediately after concrete finishing. The com-pound should be covered with scuff-resistant paper ifthe floor is subjected to traffic before curing is com-pleted. More complete information is found in AC1308.

Wet curing is recommended for concrete of low wa-ter-cement ratio (to supply additional water for hydra-tion), where cooling of the surface is desired, whereconcrete will later be bonded, or where liquid harden-ers will be applied. It should also be required for areasto receive paint or floor tile unless the curing com-pound is compatible with these materials. Curingmethods are described in detail in AC1 308, Unventedsalamanders or other unvented heaters producing car-bon dioxide should not be used during cold-weatherfloor construction. Also of concern are finishing ma-chines and vehicles used in the area or other CO2sources such as welding unless the building is well ven-tilated. Under certain conditions carbon dioxide willadversely affect the fresh concrete surface during theperiod between placment and the application of a cur-ing compound that will seal out the air. The severity of

GUIDE TO DURABLE CONCRETE 201.2%15

the effect is dependent on’concentration of the CO, inthe atmosphere, humidity, temperature, and length ofexposure of the concrete surface to the air (Kauer andFreeman 1955). Carbonation will destroy the abrasionresistance of the surface to varying depths dependingupon the depth of carbonation. The only resourse is togrind the floor and remove the offending soft surface.

3.5-Improving wear resistance of existing floorsLiquid surface treatments (hardeners) may some-

times be used to improve the wear resistance of floors(Smith 1956). Magnesium or zinc fluosilicate, or so-dium silicate, are most commonly used. Their principalbenefit is in reducing dusting. They may also slightlyresist deterioration by some oils and chemicals comingin contact with the concrete. Liquid hardeners are mostuseful on older floors which have started to abrade ordust, as a result of poor quality concrete or poor con-struction practices, such as finishing while bleed wateris on the surface, and/or inadequate curing. In suchcases, they serve a useful purpose in prolonging theservice life of the floor. New floors, properly cured,should be of such quality that treatments with liquidhardeners should not be required, except where evenslight dusting cannot be tolerated, i.e., an in power-house floors.

Liquid hardeners should not be applied to new floorsuntil they are 28 days old to allow time for calcium hy-droxide to be deposited at the surface. Magnesium orzinc fluosilicate,* and sodium silicate liquid surfacetreatments react chemically with hydrated lime (cal-cium hydroxide) which is available at the surface ofuncured concrete. This lime is generated during cementhydration and, in inadequate curing conditions, is sus-pended in the pore water and is deposited on the con-crete surface as the water evaporates. Proper curing re-duces or eliminates these surface or near-surface limedeposits (National Bureau of Standards 1939). Thefloor should be moist-cured for 7 days and then al-lowed to air-dry during the balance of the period. Cur-ing compounds should not be used if hardeners are tobe applied, because they reduce the penetration of theliquid into the concrete. The hardener should be ap-plied in accordance with the manufacturer’s instruc-tions.

3.6-Studded tire and tire chain wear onconcrete

Tire chains and studded snow tires cause considera-ble wear to concrete surfaces, even where the concreteis of good quality. Abrasive materials such as sand areoften applied to the pavement surface when roads areslippery. However, experience from many years’ use ofsand in winter indicates that this causes little wear if theconcrete is of good quality and the aggregates are wear-resistant.

‘Fluosilicates have toxic effects on workers and the environment and mustbe handled with care.

In the case of tire chains, wear is caused by a flailingand scuffing action as the rotating tire brings the metalin contact with the concrete surface. Fortunately, theuse of chains is limited mainly to roads in the snow beltof mountain areas, and even there they are used onlywhen essential.

Studded snow tires have caused widespread and seri-ous damage, even to high-quality concrete. In this casethe damage is due to the dynamic impact of the smalltungsten carbide tip of the studs, of which there areroughly 100 in each tire. One laboratory study showedthat studded tires running on surfaces to which sandand salt were applied caused 100 times as much wear asunstudded tires (Krukar and Cook 1973). Fortunately,the use of studded tires has been declining for a num-ber of years.

Wear caused by studded tires is usually concentratedin the wheel tracks. Ruts from % to L/z in. (6 to 12 mm)deep may form in a single winter in regions where ap-proximately 30 percent of passenger cars are equippedwith studded tires and traffic is heavy (Smith andSchonfeld 1970). More severe wear occurs where vehi-cles stop, start, or turn (Keyser 1971).

Investigations have been made, principally in Scan-danavia, Canada, and the United States, to examine theproperties of existing concretes as related to studdedtire wear (Smith and Schonfeld 1970, 1971; Keyser1971; Preus 1973; Wehner 1966; Thurmann 1969). Insome cases there was considerable variability in thedata, and the conclusions of the different investigatorswere not in agreement. However, most found that ahard coarse aggregate and a high-strength mortar ma-trix are somewhat beneficial in resisting abrasion.

Another investigation was aimed at developing morewear-resistant types of concrete overlays (Preus 1971).It was concluded that polymer cement and polymer-flyash concretes provided better resistance to wear, al-though at the sacrifice of skid resistance. Steel fibrousconcrete overlays were also tested and showed reducedwear as compared with sections of regular concrete.Although these results are fairly promising, no “af-fordable” concrete surface has yet been developedwhich will provide a wear life, when studded tires areused, approaching that of normal surfaces underrubber tire tire wear.

A report (Transportation Research Board 1975)summarizes available data on pavement wear, and onthe performance and winter accident record while stud-ded tires have been in use.

3.7-Skid resistance of pavementsThe skid resistance of concrete pavement depends on

its surface texture. Two types of texture are involved1. Macro (large-scale)-texture resulting from surface

irregularities “built in” at the time of construction.2. Micro (small-scale)-texture resulting from the

hardness and type of fine aggregate used.The micro-texture is the more important at speeds of

less than about 50 mph (80 km/hr) (Kummer andMeyer 1967; Murphy 1975; Wilk 1978). At speeds

201.2R-16 MANUAL OF CONCRETE PRACTICE

greater than 50 mph (80 km/hr) the marco-texture be-comes quite important because it must be relied on toprevent hydroplaning.

The skid resistance of concrete pavement initially de-pends on the texture built into the surface layer (Dahir1981). In time, rubber-tired traffic abrades the immodiate surface layer; removing the beneficial macro-tex-ture and eventually exposing the coarse aggregate par-ticles. The rate at which this will occur and the conse-quences on the skid resistance of the pavement dependon the depth and quality of the surface layer and therock types in the fine and coarse aggregate.

Fine aggregates containing significant amounts ofsilica in the larger particle sizes will assist in slowingdown the rate of wear and maintaining the micro-tex-ture necessary for satisfactory skid resistance at thelower speeds. Certain rock types, however, polish un-der rubber-tire wear. These include very fine-texturedlimestones, doiomites, and serpentine; the finer thetexture, the more rapid the polishing. Where both thefine and coarse aggregate are of this type there may bea rapid polishing of the entire pavement surface with aberious reduction in skid resistance. Where only thecoarse aggregate is of the polishing type, the problem isdelayed until the coarse aggregate is exposed by wear.On the other hand, if the coarse aggregate is, for ex-ample, a coarse-grained silica or vesicular slag, the skidresistance may be increased when it is exposed.

The macro-texture, quite important because it willprevent hydroplaning, is accomplished by constructinggrooves in the concrete-either before hardening or bysawing after the concrete has sufficient strength to pro-vide channels for the escape of water otherwise trappedbetween the tire and pavement. It is vital that the “is-land” between the grooves be particularly resistant toabrasion and frost action. A high-quality concrete,properly finished and cured, possesses the required du-rability.

CHAPTER 4-CORROSION OF METALS ANDOTHER MATERIALS EMBEDDED IN CONCRETE

4.1 -introductionThe conditions causing corrosion of reinforcing steel

and prestressing steel are of vital importance. Theavoidance of these conditions is, of course, necessary ifconcrete containing steel is to have the intended ion-gevity. The purpose of this chapter is to summarize themechanisms and conditions of corrosion, and methodsand techniques for circumventing corrosion.

Concrete usually provides protection against rustingof adequately embedded steel because of the highly ai-kaiine environment of the portiand cement paste. Theadequacy of that protection is dependent upon theamount of concrete cover, the quality of the concrete,the details of the construction, and the degree of expo-sure to chlorides from concrete-making componentsand external sources.

AC1 222R, on corrosion of steel in concrete, details:1) mechanisms of corrosion, 2) protection against cor-

rosion in new construction, 3) methods for identifyingcorrosive environments, 4) techniques for identifyingsteel undergoing active corrosion, and 5) remedialmeasures and their limitations. That report should beconsulted.

4.2-Principles of corrosion4‘2.1 Corrosion of steel in concrete is usually an

electrochemical process that requires the developmentof an anode where oxidation takes place and a cathodewhere reduction takes place. At the anode, electrons areliberated and ferrous ions are formed (Fe * Fe+ + +Ze-); at the cathode hydroxyi ions are liberated (HZ0+M& + 2e- G= 2(OH-). The ferrous ions may sub-sequently combine with oxygen or the hydroxyi ionsand produce various forms of rust.

Steel in concrete is usually protected against corro-sion by the high pH of the surrounding Portland ce-ment paste. Uncarbonated cement paste has a mini-mum pH of 12.5, and steel will not corrode at that pH.If the pH is lowered (e.g., pH 10 or less) corrosion mayoccur. Carbonation of the portiand cement paste canlower the pH to levels of 8 to 9, and corrosion may en-sue. When moisture and a supply of oxygen are pres-ent, the presence of water-soluble chloride ions, abovethreshold levels of 0.2 percent (0.4 percent calciumchloride) by mass of Portland cement can acceleratecorrosion under many circumstances (AC1 222R).Chloride in concrete is frequently referred to as cai-cium chloride (dihydrate, anhydrous, and flake andpellet forms), or chloride (Cl -). The basic reference tochloride, particularly with respect to corrosion, is chio-ride (Cl -) as percent by mass of Portland cement. Forchloride used as an admixture, the usual references areto flake calcium chloride (contains 20 to 23 percent wa-ter) as a 1 or 2 percent addition by mass of portiandcement. The actual amount of calcium chloride in dif-ferent formulations is shown in Table 4.2.1.

Corrosion can be induced if the concentration ofoxygen, water, or chloride differs at various locationsalong a steel bar or electrically-connected steel system.Other driving forces include couplings of differentmetals (galvanic corrosion), and stray electrical cur-rents such as caused by DC current of electric railways,electroplating plants, and cathodic systems used toprotect other steel systems (e.g., pipe).

In each of the preceding situations, a strong eiectro-iyte (e.g., chloride) and moisture are needed to pro-mote the corrosion, or at least cause it to occur rapidly

Table 4.21 -Chloride dataCalcium chlor ide

compound CaCl*, percent Cl-, percent

77/;aE,percent CaCI,

9O‘per&nt CaCI, (anhydrous)94 to 97 percent CaCI,

(anhydrous)29 percent CaCI, solution

GUIDE TO DURABLE CONCRETE 201.2R-17

(in years instead of decades). If steel in contact with theconcrete is not fully encased by it (e.g., decking, doorjams, posts), even trace amounts of chloride can trig-ger and accelerate corrosion if moisture and oxygen arepresent .

There has been a great deal of discussion about thesignificance of chloride introduced into the concretewhen it was made versus chloride that enters the con-crete from the environment. The former has been called“domestic” chloride, and the latter “foreign” chlo-ride. Domestic chloride, for example, might include achloride component of set-accelerating admixtures, ofwater-reducing admixtures, of aggregates, or of cemen-titious materials.

If there is uniform distribution of chlorides, corro-sion may be minimal. However, even if there is a uni-form distribution of chlorides, significant corrosion canresult because of differences in oxygen and moisturecontents, or because of other factors. Further, in thecase of a calcium chloride addition, even if the chlorideis initially uniformly distributed, a nonuniform distri-bution eventually may result, due to movement of wa-ter (containing chloride in solution). Additionally, someof the domestic chloride can become chemically fixedby reactions with calcium aluminate components of theportland cement forming calcium chloroaluminate hy-drates, or chloride once chemically bound can becomeunbound because of carbonation.

Based upon a review of literature on the relationshipof chloride concentrations and corrosion of fully em-bedded steel, AC1 Committee 222 recommends the fol-lowing maximum acid-soluble chloride ion contents,expressed as percent by mass of the cement, as a meansof minimizing the risk of corrosion: (a) prestressedconcrete-O.08 percent; (b) reinforced concrete-O.20percent .

Committee 222 also comments that because some ofthe concrete-making materials contain chlorides thatmay not be released into the concrete, documentationon the basis of past good performance may provide abasis for permitting higher chloride levels. The sug-gested levels provide a conservative approach that isnecessary because of the conflicting data on chloridethreshold levels and the effect of different exposure en-vironments. The conservative approach is also recom-mended because many exposure conditions, such asbridge decks, garages, and concretes in a marine envi-ronment, allow the intrusion of foreign chlorides. Ininstances where foreign chlorides are present, concreteshould be made with admixtures and other concrete-making components that contain only trace amounts ofchloride, or none at all.

There have been instances of corrosion in relativelydry exposures, such as inside buildings, where the con-crete was made with calcium chloride additions withinthe 1 to 2 percent levels usually deemed satisfactory forconcrete that will stay dry (Erlin and Hime 1976). Inthese circumstances, concrete drying has been very slowbecause of thick sections or the use of tiles and otherbarriers to prevent loss of water by evaporation.

4.3-Effects of concrete-making components4.3.1 Portland cement, ground granulated blast-fur-

nace slag, and pozzolans-The high pH of concrete re-sults largely from the presence of calcium hydroxide,liberated when the portland cement hydrates, andwhich constitutes from about 15 to 25 percent of theportland cement paste. Because the pH of a saturatedsolution of calcium hydroxide is 12.5, that is the mini-mum pH of uncarbonated paste. A higher pH can re-sult because of the present of sodium and potassiumhydroxide.

The tricalcium-aluminate component of portland ce-ment can react with chloride to form calcium chloro-aluminate hydrates which chemically tie up some of thechloride. Studies on the durability of concrete in a sea-water exposure showed that when cement having 5 to 8percent tricalcium aluminate (C,A) was used there wasless cracking due to steel corrosion than when cementhaving a C,A content less than 5 percent was used(Verbeck 1968). However, it is principally the domesticchloride that reacts, especially during the initial week orso of cement hydration. Subsequent carbonation of thepaste (usually restricted to shallow surface regions andcracks) may result in the liberation of some of thatchemically bound chloride.

The chloride content of portland cement, fly ash, andsilica fume is typically very low. However, slag mayhave a significant chloride content if quenched with saltwater.

4.3.2 Aggregates-Aggregates can contain chloridesalts, particularly those aggregates associated with seawater and those whose natural sites are in ground wa-ters containing chloride. There have been reported in-stances (Gaynor 1985) where quarried stone, gravels,and natural sand contained small amounts of chloridethat have provided concrete with chloride levels thatexceed the maximum permissible levels previously de-scribed. For example, coarse aggregate containing 0.06percent chloride, when used in amounts of 1800 lb/yd’(815 kg/m’) of concrete, and with a cement content of576 lb/yd’ (261 kg/m’), will result in 0.2 percent chlo-ride by mass of cement. That level is the upper limitrecommended in AC1 222R. However, not all of thatchloride will necessarily become available to the paste.Thus, AC1 222R indicates that higher levels may betolerable if past performance has shown that the higherchloride content have not caused corrosion.

4.3.3 Mixing water-Potable mixing water can con-tain small amounts of chloride, usually at levels from20 to 100 ppm. Such amounts are considered insignifi-cant. For example, for a concrete mixture containing576 lb (261 kg) of portland cement per cubic yard anda water-cement ratio of 0.5, the resulting chloride levelwould only be from 0.001 to 0.005 percent by mass ofportland cement. Reclaimed wash water, however, maycontain significant amounts of chloride depending onthe chloride content of the original concrete mixtureand the water used for washing.

4.3.4 Admixtures other than those composed princi-pally of calcium chloride and contributing less than 0.1

201.2R.lb MANUAL OF CONCRETE PRACTICE

percent chloride ions by muss of cement--Some water-reducing admixtures contain chloride to improve ad-mixture performance, but contribute only smallamounts of chloride to the concrete when they areadded at recommended rates. Normal setting admix-tures that contribute less than 0.1 percent chloride bymass of cement are most common and their use shouldbe evaluated based on an application basis. If chlorideions in the admixture are less than 0.01 percent by massof cementitious material, such contribution representsan insignificant amount and may be considered innoc-uous.

Accelerating admixtures, other than those based oncalcium chloride, have been used in concrete with varying success. Accelerators that do not contain chlorideshould not be assumed to be noncorrosive. Materialsmost commonly used are calcium formate, sodium thi-ocyanate, calcium nitrate, and calcium nitrite. It isgenerally accepted that formates (Helm 1987) are non-corrosive in concrete.

Calcium nitrite is the only accelerating chemical ret-ommended by an admixture manufacturer as a corro-sion inhibitor. Laboratory studies have demonstratedthat it will delay the onset of corrosion or reduce therate after it has been initiated (Berke 1985; Berke andRoberts 1989). The ratio of chloride ions to nitrite ionsis important. Studies (Berke 1987) show that calciumnitrite can provide corrosion protection even at chlorideto nitrite ratios exceeding 1.5 to 1.0 by weight. Al-though dosage rates can be varied, 40 to 170 fl oz per100 lb (26 to 110 mL/kg) of cement is the most com-mon range. An extensive review of calcium nitrite’s usein concrete was compiled by Berke and Rosenberg(1989). It documents the effectiveness of calcium nitriteas n corrosion inhibitor for steel, galvanized steel, andnluminum in concrete.

Structures subjected to deicing salt applicatiousshould be designed to limit penetration of chlorides tothe reinforcing steel. If the accelerating effect fromcalcium nitrite is undesirable, use of a retarder isrecommended. Increased air-entraining agent may beuecessnry when calcium nitrite is used.

At high dosages sodium thiocyanate has been re-ported to promote corrosion (Manns and Eichler 1982).The threshold dosage at which it will initiate corrosionis between 0.75 and 1.0 percent by mass of cement(Mnnns and Eichler 1982, Nmai and Corbo 1989).

4.4.Concrete quality and cover over steelOne cause of chloride intrusion into concrete is

cracks. These cracks allow infiltration by chlorides at amuch faster rate than by the slower diffusion proc-esses, and establish chloride concentration cells that caninitiate corrosion. To minimize crack formation, con-crete should always be made with the lowest practicalwater-cement ratio commensurate with workability re-quirements for proper consolidation. Quality concrete

will have decreased water permeability and absorption,increased resistance to chloride intrusion, and reducedrisk of corrosion.

When concrete is kept moderately dry, corrosion ofsteel can be minimized. For example, if concrete con-taining as much as 2 percent flake calcium chloride isallowed to dry to a maximum relative humidity of 50 to60 percent, embedded steel should either not corrode,or corrode at an inconsequential rate (Tutti 1982).

4.4.1 Cover over steel-Extensive tests (Clear 1976;Pfeifer, Landgren, and Zoob 1987; Marusin and Pfei-fer 1985) have shown that 1 in. (25 mm) cover over baresteel bars is inadequate for severe corrosion environ-ments even if the concrete has a water-cement ratio aslow as 0.30. Tests have also shown that the chloridecontent in the top % in. (12 mm) of concrete can bevery high compared to those at depths of 1 to 2 in. (25to 50 mm) even in concrete of high quality, e.g., water-cement ratio of 0.30. As a result, adequate cover formoderate-to-severe corrosion environments should be aminimum of I I/ in. (38 mm) and preferably at least 2in. (50 mm).

4.4.2 Concrete permeability and electrical resistiv-ity-The permeability of concrete to water and chlo-ride is the major factor affecting the process of corro-sion of embedded metals. *

While the surface regions of exposed concrete struc-tures will have high or low electrical conductivity val-ues (depending upon the wetting and drying conditionsof the environment) the interior of concrete usually re-quires extensive drying to achieve low electrical con-ductivity. Tests sponsored by the Federal Highway Ad-ministration (Pfeifer, Landgren, and Zoob 1987) showthat 7 to 9 in. (178 to 220 mm) thick reinforced con-crete slabs with water-cement ratios ranging from 0.30to 0.50 have essentially equal initial AC electrical resis-tance values between the top and bottom reinforcingbar mats at 28 days. Similar AC-resistance tests onconcrete made with silica fume at water-cement plussilica fume ratios of 0.20 show extremely high initialelectrical resistance values when compared to the con-cretes having water-demerit ratios of 0.30 to 0.50. Thehigh electrical resistance values increased the resistanceto steel corrosion. The high electrical resistance of sil-ica fume concrete can be due to densification of thepaste microstructure.

4.4.3 Water-cement ratio and concrete cover oversteel-Generally, low water-cement ratios will produceless permeable concrete, and thus provide greater pro-tection against corrosion. In severe, long-term, accel-erated salt water exposure tests of reinforced concreteslabs with 1 in. (25 mm) of cover over the steel, con-cretes with water-cement ratios of 0.30, 0.40, and 0.50each developed corrosion activity, the concrete havingthe 0.50 water-cement rati eveloping the most severecorrosion currents and degree of rusting of the steel.These tests show that 1 in. (25 mm) of cover is inade-quate for concrete made with commonly specified wa-ter-cement ratios when exposure is to water that con-tains chlorides. These same laboratory tests show that

GUIDE TO DURABLE CONCRETE 201.2R-19

2 in. (50 mm) and 3 in. (75 mm) of cover provide ad-ditional corrosion protection because chloride ionscould not permeate the concrete in sufficient amountsto exceed the threshold value for triggering corrosion(Marusin and Pfeifer 1985). Long-term field studies,however, have shown that concretes made with a 0.5water-cement rat&, with 2 to 3 in. of concrete coverwill not, under certain circumstances, protect steel fromcorroding.

Numerous test programs have shown that concretemade with a water-cement ratio of 0.40 and adequatecover over the steel performs significantly better thanconcretes made with water-cement ratios of 0.50 and0.60; recent tests show that concrete having a water-ce-ment ratio of 0.32 and adequate cover over the steelwill perform even better. Chloride ion permeability toa 1 in. (25 mm) depth is about 400 to 600 percentgreater for concrete made with water-cement ratios of0.40 and 0.50 than for concrete made with a water-ce-ment ratio of 0.32.

Based upon the preceding information, the water-ce-ment ratio of concrete that will be exposed to sea orbrackish water, or be in contact with more than mod-erate amounts of chlorides, should be as low as possi-ble and preferably less than 0.40. If.this water-cementratio cannot be achieved, a maximum water-cement ra-tio of 0.45 may be used provided that the thickness ofcover over the steel is increased. For severe marine ex-posure a minimum concrete cover of 3 in. (75 mm)should be used. AASHTO recommends 4 in. (100 mm)of cover for cast-in-place concrete, and 3 in. (75 mm)of cover for precast piles. These recommended water-cement ratios apply for all types of portland cement.

For trial mixture purposes, AC1 211.1 can be used todetermine the cement factor required for obtaining agiven water-cement ratio.

A low water-cement ratio does not by itself assureconcrete of low permeability. For example, “no-fines”concrete can have a low water-cement ratio and yet behighly permeable, as evidenced by the use of such con-crete to produce porous pipe. Thus, in addition to thelow water-cement ratio, the concrete must be properlyproportioned, and well consolidated to produce a con-crete of low permeability.

Salts applied in ice-control operations will be ab-sorbed by the concrete. To reduce the likelihood ofcorrosion, a minimum cover of 2 in. (50 mm) and a lowwater-cement ratio (0.40 maximum) are desirable. Itshould be noted that because of construction toler-ances, a design cover of at least 2.6 in. (65 mm) isneeded to obtain a minimum cover of 2 in. (50 mm)over 90 to 95 percent of the reinforcing steel (Van Dav-eer and Sheret 1975). Nondestructive techniques such asmagnetic devices (pachometer) and radar are availablefor determinjng the depth of cover over reinforcingsteel in hardened concrete (Clear 1974a; Van Daveerand Sheret 1975).

4.4.4 Mixrure proportions-Low water-cement ratiosdecrease concrete permeability, which results in greaterresistance to chloride intrusion. In seawater exposure

studies of reinforced concrete where cover over the steelwas nominally 1 I/ in. (37 mm), a water-cement ratio of0.45 provided good corrosion protection, a water-ce-ment ratio of 0.53 provided an intermediate degree ofprotection, and a water-cement ratio of 0.62 providedlittle protection (Verbeck 1968). Tests of concrete slabsat equal cement contents, which were salted daily, in-dicated that water-cement ratios of 0.40 provided sig-nificantly better corrosion protection than water-ce-ment ratios of 0.50 and 0.60 (Clear and Hay 1973).Based upon these studies, the water-cement ratio forconcrete exposed to brackish or seawater, or in contactwith chlorides from other sources, should not exceed0.40. Any means of decreasing the permeability of con-crete, such as by the use of high-range water reducers,pozzolans, and silica fume, will prolong the onset ofcorrosion.

Exposure of concrete at inland sites, i.e., sites so farinland that no salt comes from the sea, has not gener-ally been recognized as constituting a corrosion prob-lem except where exposed to brakish water or wheredeicing salts are used. Severe corrosion of bridge andparking structures has occurred.

4.4.5 Workmanship-Good workmanship is vital forsecuring uniform concrete and concrete of low perme-ability. For low-slump concrete, segregation and hon-eycombing can be avoided by good consolidation prac-tices. Because low-slump concrete is often difficult toconsolidate, a density monitoring device is helpful forinsuring good consolidation (Honig 1984).

4.4.6 Curing-Permeability is reduced by good cur-ing because of increased hydration of the cement. Atleast 7 days of uninterrupted moist-curing or mem-brane-curing should be specified, Prevention of the de-velopment of excessive early thermal stresses is also im-portant (Acker, Fourier, and Malier 1986).

4.4.7 Drainage-Particular attention should be givento design details to insure that water will drain and notpond on surfaces.

4.4.8 Exposed items-Careful attention should begiven to partially embedded and partially exposeditems, such as bolts, that are exposed directly to corro-sive environments. The resistance of these items to thecorrosive environment should be investigated and thecoupling of dissimilar metals avoided. Concrete shouldbe carefully placed around embedded items so that it iswell consolidated and does not create paths that willpermit corrosive solutions to easily reach the concreteinterior.

4.5-Positive protective systemsCosts of repairing corrosion-caused damage are very

high. Many protective systems have been proposed,some of which have been shown to be effective whileothers have failed. It is beyond the scope of this guideto discuss all possible systems. However, the most suc-cessful systems are listed in the following paragraphs.

1. Overlays and patches of very low water-cementratio (0.32) using conventional low-slump concrete, la-tex-modified concrete overlays (Clear and Hay 1973;

201.2R-20 MANUAL OF CONCRETE PRACTICE

Federal Highway Administration 1975c), concrete con-taining silica fume, and concrete containing high-rangewater reducing admixtures.

2. Epoxy-coated reinforcing steel (Clifton, Beeghly,and Mathey 1974; Federal Highway Administration197Sa).

3, “Waterproof” membranes (Van Til, Carr, andVallerga 1976).

4. Surface protective-barrier systems produced fromselect silanes, siloxanes, epoxies, polyurethanes, andmethacrylates (Van Daveer and Sheret 1975).

5. Cathodic protection.6. Polymer impregnation (Smock 1975).7. Replacement of the existing concrete with concrete

containing a corrosion inhibitor.General information on repairs of concrete and use

of protective-barrier systems are given in Chapters 6and 7 of this report.

4.6Corrosion of materials other than steel4.6.1 Aluminum-Corrosion of aluminum embed-

ded in concrete can occur and can crack the concrete.Conditions conducive to corrosion are created if theconcrete contains steel in contact with the aluminum,chlorides are present in appreciable concentrations, orthe cement is high in alkali content (Woods 1968). In-creasing ratios of steel area (when the metals are cou-pled), particularly in the presence of appreciableamounts of chloride, increases corrosion of the alumi-num. Additionally, hydrogen gas evolution may occurwhen fresh concrete contacts aluminum and this mayincrease the porosity of the concrete and therefore thepenetration of future corrosive agents. Some aluminumalloys are more susceptible to this probtem than others.Corrosion inhibitors (e.g., calcium nitrite) have beenshown to improve the corrosion resistance of alumi-num in concrete (Berke and Rosenberg 1989).

4.6.2 Lead--Lead in damp concrete can be attackedby the calcium hydroxide in the concrete and may bedestroyed in a few years. Contact of the lead with re-inforcing steel can accelerate the attack. It is recom-mended that a protective plastic, or sleeves which areunaffected by damp concrete, be used on lead to beembedded in concrete. Corrosion of embedded lead isnot likely to damage the concrete.

4.6.3 Copper and copper alloys-Copper is not nor-mally corroded by concrete, as evidenced by the wide-spread and successful use of copper waterstops and theembedment of copper pipes in concrete for many years(Erlin and Woods 1978). However, corrosion of copperpipes has been reported where ammopia is present.Also, there have been reports that small amounts ofammonium and possibly of nitrates can cause stresscorrosion cracking of embedded copper. It should fur-ther be noted that unfavorable circumstances are cre-ated if the concrete also contains steel connected to thecopper. In this case it is the steel which will corrode.

4.6.4 Zinc-Zinc reacts with alkaline materials suchas those found in concrete. However, zinc in the form

of a galvanizing coating on reinforcing steel is some-times intentionally embedded in concrete. Availabledata are conflicting as to the benefit, if any, of thiscoating (Stark and Perenchio 1975; Hill, Spellman, andStratfull 1976; Griffin 1969; Federal Highway Admin-istration 1976). A chromate dip on the galvanized barsor the use of 400 ppm of chromate in the mixing wateris recommended to prevent hydrogen evolution in thefresh concrete. Caution should be exercised when usingchromium salts because of possible skin allergies. Ad-ditionally, users are cautioned against permitting gal-vanized and black steel to come in contact with eachother in a structure, since theory indicates that the useof dissimilar metals can cause galvanic corrosion. Cor-rosion inhibitors, such as calcium nitrite, have been

shown to improve the corrosion resistance of zinc inconcrete (Berke and Rosenberg 1989).

Some difficulty has been experienced with the corro-sion and perforation of corrugated galvanized sheetsused as permanent bottom forms for concrete roofs andbridge decks. Such damage has been confined largely toconcrete containing appreciable amounts of chlorideand to areas where chloride solutions are permitted todrain directly onto the galvanized sheet.

4.6.5 Other metals-Chromium and nickel alloyedmetals generally have good resistance to corrosion inconcrete, as do silver and tin. However, the corrosionresistance of some of these metals may be adversely af-fected by the presence of soluble chlorides in seawateror deicing salts. Special circumstances might justify thecost of Monel, or Type 316 stainless steel in marine lo-cations, if data are available to document their supe-rior performance in concrete containing moisture andchlorides or other electrolytes. However, the 300 Seriesstainless steels are susceptible to stress corrosion crack-ing when the temperature is over 140 F (60 C) andchloride solutions are in contact with the material.Embedded natural weathering steels generally do notperform well in concrete containing moisture and chlo-ride. Weathering steels adjoining concrete may dis-charge rust and cause staining of concrete surfaces.

4.6.6 Plastics-Plastics are being used increasingly inconcrete as pipes, shields, waterstops, chairs, etc., aswell as a component in the concrete. Many plastics areresistant to strong alkalies and therefore would be ex-pected to perform satisfactorily in concrete, However,because of the great variety of plastics and materialscompounded with them, specific test data should bedeveloped for each intended use. Special epoxies havebeen used successfully as reinforcing bar coatings andwill be discussed later in this guide.

4.6.7 Wood-Wood has been widely used in oragainst mortars and concretes. Such use includes theincorporation of sawdust, wood pulp, and wood fibersin the concrete as well as the embedment of timber.

The use of untreated sawdust, wood chips, or fibersusually results in slow setting and low strength con-crete. The addition of hydrated lime equal to ‘/, to 1/2the volume of cement is usually effective in minimizingthese problems. The further addition of up to 5 percent

GUIDE 7’0 DURABLE CONCRETE 201.2R-21

of calcium chloride dihydrate by weight of cement hasalso helped to minimize these problems. However, cal-cium chloride in such amounts can cause corrosion ofembedded metals and can have adverse effects on theconcrete itself.

Another problem with such concrete is the high vol-ume change, which occurs even with changes in atmos-pheric humidity, This volume change may lead tocracking and warping.

The embedment of lumber in concrete has sometimesresulted in leaching of the wood by calcium hydroxidewith subsequent deterioration. Softwoods, preferablywith a high resinous content, are reported to be mostsuitable for such use.

4.7-Summary commentsPortland cement concrete can provide excellent cor-

rosion protection to embedded steel. When corrosionoccurs, costs of repairs can be exceedingly high. Theuse of high quality concrete, adequate cover over thesteel, and good design are prerequisites if corrosion isto be minimized.

AC1 222R provides a current summary of the causesand mechanisms of corrosion of steel. It includes in-formation on how to protect against corrosion in newstructures as well as procedures for identifying corro-sive environments; it also describes some remedialmeasures for existing situations where corrosion is oc-curring.

CHAPTER 5-CHEMICAL REACTIONS OFAGGREGATES

5.1 -Types of reactionsChemical reactions of aggregates in concrete can af-

fect the performance of concrete structures. Some re-actions may be beneficial; others may result in seriousdamage to the concrete by causing abnormal internalexpansion which may produce cracking, displacementof elements within larger structural entities, and loss ofstrength (Woods 1968).

5.1.1 Alkali-silica reaction-The reaction that hasreceived greatest attention and which was the first to berecognized involves a reaction between the OH- ion as-sociated with the alkalies (Na,O and K,O) from thecement and other sources, with certain siliceous con-stituents that may be present in the aggregate. Thisphenomenon was referred to as “alkali-aggregate reac-tion,” but is more properly designated as “alkali-silicareaction.” The earliest paper discussing alkali-silica re-action is that by Stanton (1940).

Deterioration of concrete involving certain sand-gravel aggregates has occurred in Kansas, Nebraska,and eastern Wyoming (Gibson 1938; Lerch 1959). Be-cause early studies showed no consistent relationshipbetween the distress and alkali content of the cement,this deterioration was called “cement-aggregate reac-tion” to differentiate it from alkali-silica reaction.Subsequent research indicated that this phenomenon isalkali-silica reaction (Hadley 1964).

There are reports mentioning structural repairs or re-placements due to malfunction or concern for safety,for instance at the Matilija dam of the U.S. Bureau ofReclamation and the Jersey New Waterworks Dam(Coombes, Cole, and Clarke 1975; Coombes 1976).Also, the bascule piers of the Oddeesund Bridge andthe Vilsund Bridges, Jutland, Denmark, were thor-oughly repaired due to malfunction in service (DanishNational Institute of Building Research, 1956-65). Mis-alignment of machinery and malfunction of opera-tional structural elements in a dam in India was re-ported by Visvesvaraya, Rajkumar, and Mullick (1987).Repairs to rather new highway bridges and pavementsand at a nuclear power plant near Cape Town, SouthAfrica were reported by Oberholster (1981). The com-bined effects of thermal expansion on a dam face andexpansive alkali-silica reactions in the concrete massnecessitated repairs in Fontana Dam (Abraham andSloan 1979).

In North Germany, the Lachswehrbrucke in Lubeckwas removed in 1969 about 1 year after constructiondue to severe cracking caused by alkali-silica reaction.This case received intensive public interest, although notechnical report was issued. A comprehensive report onalkali-silica reaction in Germany was published in 1973(Verein Deutscher Zementwerke 1973); other studies aredescribed by Lenzner (1981).

In mass concrete gravity dams, concern about thestructural integrity may become justified within thestipulated lifetime. Experience shows that the warningsignals, cracking and gross expansions, often allow fortimely remedial work.

In reinforced concrete, the reinforcement contributesconsiderable resilience against decline of the structuralsafety, but it cannot prevent malfunctions caused byexpansion and displacement of structural members.

In some structures, concrete spalling may causesafety risks, for instance, on airport runways and onbridges over highways. In other cases the public atti-tude or esthetic concern may necessitate remedial work.

One well-documented case of damage to concrete inservice where low-alkali cement was used with alkali-reactive aggregate was described by Hadley (1968). Itdeals with pavements in a region of very hot, dry sum-mers where there was migration and concentration ofalkalies as moisture moved through the pavement toevaporate at the top surface. The same concern applieswhere concrete structures are exposed to additional al-kalies in a marine environment or by the application ofdeicing salts based on sodium chloride. Additional in-stances of damage to concrete by expansion due to al-kali-silica reaction where the cement is believed to havehad an alkali content below 0.60 percent Na,O equiva-lent have been reported (Stark 1978).

The effects of temperature must also be taken intoconsideration. The chemical reactions are acceleratedby increased temperatures. At low temperatures, thereactions may become dormant.

5.1.2 Alkali-carbonate rock reaction-It has alsobeen clearly demonstrated that certain carbonate rocks

201.2R.22 MANUAL OF CONCRETE PRACTICE

participate in reactions with alkalies which, in some in-stances, produce detrimental expansion and cracking.Detrimental reactions are usually associated with argil-laceous dolomitic limestones which have somewhat un-usual textural characteristics (Hadley 1964). This reac-tion is designated as the alkali carbonate-rock reaction.It has been extensively studied in Canada, where it wasoriginally recognized (Swenson 1957; Swenson and Gil-lott 1960; Feldman and Sereda 1961; Gillott and Swen-son 1969; Gillott 1963; Swenson and Gillott 1967) andin the United States (Sherwood and Newlon 1964;Newlon and Sherwood 1964; Newlon, 0201, and Sher-wood 1972; Newlon, Sherwood, and 0201 1972; Walker1974; 0201 and Newlon 1974).

In addition to the detrimental expansive alkali-car-bonate reaction, another phenomenon associated withsome carbonate rocks occurs in which the peripheralzones of the aggregate particles in contact with cementpaste are modified and develop prominent rims withinthe particle and extensive alteration of the surroundingpaste (Hadley 1964; Newlon and Sherwood 1964;Bisque and Lemish 196Oa, 196Ob; Lemish and Moore1964; Hiltrop and Lemish 1960). Some rims, whenetched with dilute acid appear in positive relief, whileothers exhibit negative relief; hence the terms “positiverims” and “negative rims.” As contrasted with alkali-carbonate reactions which cause detrimental expansionand cracking, it is doubtful that the rim-forming alkali-carbonate reaction is by itself a deleterious reaction(Buck and Dolch 1976).

Some recent cases of very large structural expansionand consequent distress were reported by Grattan-Bel-lew (1987).

5.1.3 Other reactions involving aggregate-Otherdamaging chemical reactions invo1vin.g aggregates in-clude the oxidation or hydration of certain unstableoxides, sulfates, or sulfides that occur after the aggre-gate is incorporated into the concrete. Examples in-clude the hydration of anhydrous magnesium oxide,calcium oxide, or calcium sulfate, or the oxidation ofpyrite (Mielenz 1964). Apparently sound dolostone ag-gregate that has been found to be stable in concrete atnormal temperatures may deteriorate due to oxidationof small amounts of pyrite when used at elevated tem-peratures (Soles 1982). Metallic iron may occur as acontaminant in aggregate and subsequently be oxi-dized. Still other reactions may result from organic im-purities, i.e., humus, sugar, etc. (Hansen 1964). Usersof aggregate should be aware of these possibilities andemploy corrective measures where necessary. Carefultesting and examination of the aggregates will usuallyindicate the presence of such reactive impurities andtheir use in concrete can be avoided.

The alkali-silica and alkali-carbonate reactions aremore important than the others and will be discussed indetail in the following section.

5.2-Alkalidica reactlon5.2.1 Occurrence-A map (Mielent 1978) and data

(Meissner 1941; Hinds and Tuthill 1941; Kammer and

Carlson 1941; Dolar-Mantuani 1969; Buck and Mather1%9; Brown 1955; K. Math& 1973; Duncan et al. 1973;Duncan, Gillott, and Swenson 1973; Duncan, Swen-son, and Gillott 1973; Gogte 1973; Halldorsson 1975)are available showing areas known to have natural ag-gregates suspected of or known to be capable of alkali-silica reaction. Most of these references refer to NorthAmerica; however, the available evidence (Halldorsson1975) suggests that similar considerations are applica-ble elsewhere. Cases have been reported from Den-mark, Iceland, Sweden, Germany, France, Britain, It-aly, Cyprus, Turkey, Chile, Argentina, Brazil, India,Japan, New Zealand, Australia, East, West, and SouthAfrica, and other countries (Halldorrson 1975; Dia-mond 1978a; Oberholster 1981; Idorn and Rostam1983; Grattan-Bellew 1987).

At one time it appeared that the greatest abundanceof alkali-silica reactive rocks in the United States was inthe western half of the country. This is probably stillcorrect, for the quickly developing alkali-silica reactionwhich was the first to be recognized (Stanton 1940;Meissner 1941; Hinds and Tuthill 1941; TransportationResearch Board 1958). However, there is also a slowlydeveloping type (Kammer and Carlson 1941).

The aggregate constituents recognized as reactive in1958 are shown in Table 5.2.1. (Transportation Re-

Table 5.2.1 -Deleteriously reactive siliceousconstituents that may be present in aggregates

Reactive substance IChemical composition1 Physical character

10cryptocrystalline;

(b) Crystalline. but

I in tenselyfractured.strained, &d/orinclusion-filled

olcanic rocks or

The most important deleteriously alkali-reactive rocks (that is,rocks containing excessive amounts of one or more of the substanceslisted above) are as follows:

Opaline chertsChalcedonic chertsQuartzose chertsSiliceous limestonesSiliceous dolomitesRhyolites and tuffsDacites and tuffs

Andesita and tuffsSiliceous shalesPhyllitesOpaline concretionsFractured, strained, and inclusion-

filled quartz and quartsites

b eNote: A rock may be classifii as, for exmnple, a “sil iceous limestone” &innocuous if its siliceous constituents are other thrn those indicated above.

GUIDE TO DURABLE CONCRETE 201.2%23

search Board 1958). Since l958, other rocks have beenrecognized as reactive. These include argillites, gray-wackes (Dolar-Mantuani 1969), quartzites (Duncan etal. 1973; Duncan, Gillott, and Swenson 1973; Duncan,Swenson, and Gillott 1973), schists (Gogte 1973), aswell as fractured and strained quartz, recognized as re-artive by Brown (1955) and granite gneiss (K. Matheri373). Such strained quartz is typically characterized byundulatory extinction (Gogte 1973). Several of theserocks-including granite gneisses, metamorphosedsubgraywackes, and some quartz and quartzite grav-els-appear to react slowly even with high-alkali ce-ment, the reactivity not having been recognized untilthe structures were over 20 years old (Buck and Mather1969; Brown 1955; K. Mather 1973). Stark and Bhatty(1986) have shown that reactive aggregates can becaused to react by alkali derived from rocks and min-erals that may not themselves be alkali-silica reactivebut that can yield alkali by leaching.

In the evaluation of the ages at which reactivity hasbeen recognized in structures, one must also recognizethe uncertainty of the time of recognition of reaction,the influence of ambient temperatures and humidity,the alkali-silica ratio of the reacting system, and theconcentrations of reactive aggregates.

In South Africa, deleterious reactions with gray-wacke have not been particularly slow (Oberholster1981).

Lightweight aggregates, which often consist predom-inantly of amorphous silicates, would appear to havethe potential for being reactive with cement alkalies.However, no case histories of distress of lightweightconcrete caused by alkali reaction have been reportedso far as is known to AC1 Committee 213. An unpub-lished account of elongation of a lightweight concretebridge-deck may be the exception to this rule but it hasnot been adequately documented.

5.2.2 Mechanisms--Alkali-silica reaction can causeexpansion and cracking of concrete structures andpavements. The phenomenon is complex and varioustheories have been advanced to explain field and labo-ratory evidence (K. Mather 1973; Gogte 1973; Hansen1944; Powers and Steinour 1955; Diamond 1975, 1976).Unanswered questions remain. Silica can be dissolvedin solutions of high pH. The initial reaction product atthe surface will be a nonswelling calcium-alkali silica gelapproaching C-S-H. For reaction to continue safely,the amount of reactive material must either be negligi-ble or more than a pessimum* amount, dependingon the amount of alkali and the fineness of reactivematerial. Formation of the nonexpansive product is de-sirable and will occur if the reactive particles presentsufficient surface for reaction, that is, if the reactiveparticles are sufficiently numerous or sufficiently fine.Alkali-silica reactive materials of high Fineness are infact pozzolanic materials and blast-furnace slag prop-erly made and used may transform the reactions to be-come beneficial (Pepper and Mather 1959; Idorn and

*P&mum means worst, the opposite of optimum.

Roy 1986). If the amount of alkali is large with respectto the reactive aggregate surface, interior alkali-silicagel with unlimited expansive potential will form, im-bibe water, and exert potentially destructive force.

5.2.3 Laboratory tests for alkali-silica reactivity-Laboratory tests should be made on aggregates fromnew sources and when service records indicate thatreactivity may be possible. The most useful laboratorytests are:

a. Petrographic examination (ASTM C 295)-Thisdocument provides a standard practice for the petro-graphic examination of aggregates (Mielenz 1978). Thetypes of reactive aggregate constituents involved in al-kali-aggregate reaction are listed in Table 5.2.1, andprocedures for recognizing these constituents have beendescribed (Kammer and Carlson 1941; B. Mather 1948;Brown 1955; Diamond 1975, 1976). Recommendationsare available that show the amounts of reactive miner-als, as determined petrographically, that can be toler-ated (B. Mather 1948; Mielenz 1958; Corps of Engi-neers 1985). These procedures apply principally to themore extensively studied reactive constituents.

The reactive rocks and minerals that have been morefrequently encountered in recent years appear to havelarger pessimum proportions and are harder to recog-nize in petrographic examination. Highly deformedquartz with deformation lamellae appears characteris-tic of the reactive quartz-bearing rocks. Relativelycoarse-grained micas (Duncan et al. 1973; Duncan,Gillott, and Swenson 1973; Duncan, Swenson, and Gil-lott 1973) have also been regarded as reactive constitu-ents; fine-grained micas are reactive in argillites (Dolar-Mantuani 1969). The pessimum proportion conceptdoes not appear to apply for reactive coarse aggregatesbecause the reactivity may be partial. In general, theconcept is difficult to apply in engineering practice, be-cause mineral composition of aggregates of mixed rocktypes cannot be monitored practically and economi-cally, and also because the effects of the particle sizeand alkali concentration are inseparably governing pa-rameters.

b. Mortar-bar test for potential reactivity (ASTMC 227)-This method is the one most generally reliedon to indicate potential alkali reactivity. Acceptancecriteria are given in the appendix to ASTM C 33 forevaluating results of tests made using ASTM C 227.The procedure is useful not only for the evaluation ofaggregates, but also for the evaluation of specific ag-gregate-cement combinations. Particular care must betaken to insure that the bars are never allowed to losemoisture. From the results of Duncan et al. (1973);Duncan, Gillott, and Swenson (1973); and Duncan,Swenson, and Gillott (1973) it may be expected thatcertain metamorphic siliceous rocks will not reliablydevelop an expansive reaction in storage at 100 F (38C). More elevated temperatures and/or longer periodsin test, probably 1 to 3 years, will be required to de-velop evidence of reactivity. This prolongation of test-ing time makes it particularly desirable to employ pet-rographic criteria that will allow identification of these

201*2R-24 MANUAL OF CONCRETE PRACTICE

rocks. Studies in recent years suggest that the mortar-bar test is not always able to insure safe determinationof the expansive reactivity of aggregates in field con-crete (Oberholster and Davies 1986).

Variations on ASTM C 227 have been tried in sev-eral different laboratories in different countries with theintent of inducing meaningful results more quickly, es-pecially for aggregates that are slowly reactive. Onesuch procedure involves storage in sodium chloride so-lution (Chatterji 1978).

c. Chemical test for potential reactivity (ASTMC 289)-This method is used primarily for a quickevaluation with results being obtainable in a few daysas compared with 3 to 12 months with the mortar-bartest. Care must be exercised in interpreting the resultsof this test. Criteria for interpretation are given in theAppendix to ASTM C 33. Transportation ResearchBoard Special Report No. 31 (1958) and Chaiken andHalstead (1960) give more details concerning interpre-tation of the results. Some of the more recently studiedreactive rocks fall into a region below the end of thecurve (Fig. 2, ASTM C 289) so that the results cannotbe easily interpreted using the criteria given in the stan-dard.

The test, in effect, measures the pozzolanic reactivityof the suspected aggregate at about the maximum tem-perature found in much concrete during the initial cur-ing phase. Thus it emphasizes the essential identity ofthe alkali-silica reaction and the pozzolanic reaction.This test method has given questionable results when’evaluating lightweight aggregates, and it is therefore notrecommended for this purpose (Ledbetter 1973).

5.2.4 Criteria for judging reactivity-When availa-ble, the field performance record of a particular aggre-gate, if it has been used with cement .of high-alkalicontent, is the best means for judging its reactivity(Mielenz 1958). If such records are not available, themost reliable criteria are petrographic examination withcorroborating evidence from the mortar-bar test (Corpsof Engineers 1985), sometimes supplemented by testson concrete. The chemical test results should also beused in conjunction with results of the petrographic ex-amination and mortar-bar test. It is strongly recom-mended that reliance not be placed upon the results ofonly one kind of test in any evaluation (Corps of En-gineers 1985).

5.2.5 Recommended procedures to be used with al-kali-reactive aggregates-If aggregates are shown byservice records or laboratory examination to be poten-tially reactive, they should not be used when the con-crete is to be exposed to seawater or other environ-ments where alkali is available to enter the concrete insolution from an external source (Transportation Re-search Board 1958). When reactive aggregates must beused, this should be done only after thorough tests, andpreferably after service records have established thatwith appropriate limits on the alkali content of the ce-ment, or with the use of appropriate amounts of an ef-fective pozzolan or slag, or both, satisfactory servicecan be anticipated (Pepper and Mather 1959). In cases

where alkali from the environment is not involved andthere are no nonreactive materials available economi-cally, reactive materials may be used provided the fol-lowing safeguards are employed: ’

a. Low-alkali cement-Specify a low-alkali cement(maximum of 0.60 percent equivalent Na,O). Prohibitthe use of seawater or alkali soil water as mixing waterand avoid the addition of sodium or potassium chlo-ride. Beware of the risk of migration of alkalis by dif-fusion in concrete.

b. Pozzolan or slag-Alternatively, use a suitablepozzolanic material meeting the relevant requirementsof ASTM C 618, or blast furnace slag meeting the re-quirements of ASTM C 989. Pozzolans should betested in accordance with ASTM C 441 to determinetheir effectiveness in preventing excessive expansion dueto the alkali-aggregate reaction. The criterion of 75percent reduction in mortar-bar expansion, based on anarbitrary cement-to-pozzolan ratio, merely provides abasis for comparison. Pepper and Mather (1959)showed that many pozzolans would need to be used athigher proportions to achieve 75 percent reduction inexpansion of a Pyrex mixture with a cement having a1.0 percent Na,O equivalent. Pozzolans (natural, flyash, silica fume) when tested in a similar manner mustshow mortar-bar expansions less than 0.020 percent at14 days. Fortunately, most reactive aggregates are lessreactive than Pyrex.

Whenever the use of pozzolanic materials is consid-ered, it should be remembered that if these materialsincrease water demand, they may cause increased dry-ing shrinkage in concrete exposed to drying. Increased‘water demand results from high fineness and poor par-ticle shape. Usually, well granulated and ground blastfurnace slag will improve the workability of concrete.The rate of strength development in correctly propor-tioned concrete made with a pozzolan or slag can equalor exceed that of portland cement concretes at 28 days.

5.2.6 Cement-aggregate reaction-Sand-gravel ag-gregates in Kansas, Nebraska, and Wyoming have beeninvolved in concrete deterioration described as due to“cement-aggregate reaction” (Gibson 1938; Lerch1959; Hadley 1968) and differentiated from alkali-silicareaction because of lack of clear-cut dependence onlevel of alkali content of cement. It is now known(Hadley 1968) that the reaction is alkali-silica reaction.Evaporation at the surface of the concrete causes anincrease in alkali concentration in the pore fluids nearthe drying surface. Under these and other comparableconditions, even a low-alkali cement may cause objcc-tionable deterioration, particularly near the surface.Special tests, such as ASTM C 342, have been devisedto indicate potential damage from this phenomenon.Petrographic examination (ASTM C 295) and mortarbars (ASTM C 227). with results interpreted as de-scribed by Hadley (1%8) are regarded as more reliable.

The use of these potentially deleteriously reactivesand-gravel aggregates should be avoided where possi-ble. However, if they must be used, a suitable pozzolan

GUIDE TO DURABLE CONCRETE 201.2R-25

or blast furnace slag that does not increase dryingshrinkage and 30 percent or more (by mass) of non-reactive limestone coarse aggregate should be used.Concrete tests should be used to determine whether theresulting combination is satisfactory (TransportationResearch Board 1958; Powers and Steinour 1955), andwhether the limestone is frost resistant in air-entrainedconcrete in the grading in which it is used.

5.3-Alkali-carbonate reaction5.3.1 Occurrence-Certain carbonate-rock aggre-

gates, usually dolomitic, have been found to be reac-tive in concrete structures in Canada (Ontario) and inthe United States (Illinois, Indiana, Iowa, Michigan,Missouri, New York, South Dakota, Virginia, Tennes-see, and Wisconsin). Both quarried aggregates andgravels containing particles from the same formationmay be reactive..

5.3.2 Mechanism-More than one mechanism to ex-plain alkali-carbonate reactivity has been proposed(Hadley 1964; Gillott and Swenson 1969; Gillot 1963a;Sherwood and Newlon 1964; Newlon, 0~01, and Sher-wood 1972). It is clear that when dedolomitizationleading to the formation of brucite [Mg(OH),] occurs,there is a regeneration of the alkali. This is a featurethat is different from alkali-silica reactivity, in whichthe alkali is combined in the reaction product as the re-action proceeds. The presence of clay minerals appearssignificant in some cases and their swelling, whenopened to moisture by dedolomitization, is the basis forone of the possible explanations of the reaction (Gillott1963a).

Rim growth is not unusual in many carbonate rocks,and it has been reported as associated with distress inpavements in Iowa (Welp and De Young, 1964). How-ever, this is not always the case. The nature of rim for-mation is not fully understood (Hadley 1964). It is,however, associated with a change in the distribution ofsilica and carbonate between the aggregate particle andthe surrounding cement paste, the rims appearing toextend concentrically deeper into the aggregate withtime.

The affected concrete is characterized by a networkof pattern or map cracks, usually most strongly devel-oped in areas of the structure where the concrete has aconstantly renewable supply of moisture, such as closeto the waterline in piers, from earth behind retainingwalls, from beneath road or sidewalk slabs, or by wickaction in posts or columns. A feature of the alkali-car-bonate reaction that distinguishes it from the alkali-sil-ica reaction is the general absence of silica-gel exuda-tions at cracks. Additional signs of the severity of thereaction are closed expansion joints with possiblecrushing of the adjacent concrete (Hadley 1964; Swen-son and Gillott 1964).

5.3.3 Identification by laboratory testsa. Petrographic examination of aggregate (ASTM

C 29.5)-Such examination may be used to identify thefeatures of the rock as listed by Hadley (1964), andmodified by Buck (1969) and Dolar-Mantuani (1964,

1971). While it is generally true that reactive rocks canbe characterized as having dolomite rhombs from 1 to200 pm in maximum dimension in a background offiner calcite and insoluble residue, the presence of all orany dolomite in a fine-grained carbonate rock makes itdesirable to conduct the rock-cylinder test (ASTMC 586). This is recommended whether or not the tex-ture is believed to be typical, and whether or not insol-uble residue including clay amounts to a substantialportion of the aggregate. As expansive rocks are rec-ognized from more areas, the more variable the tex-tures and composition appear to be.

b. Rock-cylinder test (ASTM C 586)-The rock-cyl-inder test was first adopted by ASTM in 1966 based onwork by Hadley (1964). It is discussed by Walker(1978). It should be used as a screening test.

c. Expansion of concrete prisms-The prisms aremade with job materials and stored at 100 percent rel-ative humidity at 73 F (23 C) (Swenson and Gillott1964), or to accelerate the reaction, they may be madewith additional alkali or stored at elevated temperatureor both (Smith 1964, 1974; Gillott 1963a; Rogers 1986).Swenson and Gillott (1964) reported that such testsshowed that expansion of concrete with highly reactivecarbonate rock could be reduced to “safe” values onlyif the alkali content of the cement is below 0.45 or 0.40percent as Na,O equivalent. Thus, they stated, “thenormally accepted maximum of 0.60-percent alkali in‘low-alkali cement’ is not adequate.”

Comparison is usually made with the expansion ofprisms containing a nonreactive control aggregate.ASTM C 1105 to measure length change of concretedue to alkali-carbonate rock reaction was adopted in1989 and a Canadian standard (CSA A23.2-14A) usingconcrete specimens is available.

d. Petrographic examination of the concrete-Thiscan confirm the types of aggregate constituents presentand their characteristics. Distress that has occurred inthe aggregate and surrounding matrix, such as micro-and macro-cracking, may be observed. Reaction rimsmay be observed in certain aggregate particles and maybe identified as negative or positive by acid etching.Their presence does not necessarily signify harmful re-sults. Secondary deposits of calcium carbonate, cal-cium hydroxide, and ettringite may be found in voidswithin the concrete. Deposits of silica, hardened or ingel form, associated with the suspect aggregate parti-cles will not usually be found (Hadley 1964).

e. Other laboratory tests-Alkali-carbonate reactionmay be identified by visual observation of sawed orground surfaces. X-ray examination of reaction prod-ucts is also sometimes useful. ASTM C 227, ASTMC 289, and ASTM C 342, which are applicable to al-kali-silica reaction, are not applicable to alkali-carbon-ate reactivity.

5.3.4 Criteria for judging reactivity-Definitive cor-relations between expansions occurring in the labora-tory in rock cylinders or concrete prisms and deleteri-ous field performance have not yet been established.The factors involved are complex and include the

201.2fb26 MANUAL OF CONCRETE PRACtlCE

heterogeneity of the rock, coarse aggregate size, per-meability of the concrete, and seasonal changes in en-vironmental conditions in service. The principal envi-ronmental conditions include availability of moisture,level of temperature, and possibly the use of sodiumchloride as a deicing chemical.

Cracking is usually observed in concrete prisms at anexpansion of about 0.05 percent. Experience in Ontario(Rogers 1986) indicates that if concrete prisms madeaccording to the Canadian Standards Association TestMethod (CSA A23.2-14A) do not show expansiongreater than 0.02 percent after 1 year, harmful reactiv-ity is unlikely. Slightly less restrictive criteria has beensuggested elsewhere (Swenson and Gillott 1964; Smith1974).

It is not certain that rapid determination of potentialreactivity can always be made by using the rock-cylin-der test, since some rocks showing an initial contrac-tion may develop considerable expansion later (Dolar-Mantuani 1964; Missouri Highway Department 1%7a).No universal correlation exists between the expansionof rock cylinders and concrete in service, though it mayexist between the expansion of rock cylinders and theexpansion of concrete prisms stored in the laboratory(Hadley 1964; Newlon and Sherwood 1964; MissouriHighway Department 1%7a; Rogers 1986).

Expansions greater than 0.10 percent in the rock cyl-inders are usually taken as a warning that further testsshould be undertaken to determine expansion of theaggregate in concrete. Fortunately, many carbonaterocks that expand in rock cylinders do not expand inconcrete.

5.3.5 Recommended procedures to minimize alkali-carbonate reactivity-Procedures that can be employedto minimize the effects of the reaction include:

a. Avoiding reactive rocks by selective quarrying(Bisque and Lemish 1960a; Smith 1964; Gillott 1%3a).

b. Dilution with nonreactive aggregates, or use of asmaller nominal maximum size (Newlon and Sherwood1964; Swenson and Gillott 1964).

c. Use of very low-alkali cement (less than 0.6 per-cent Na,O equivalent [see Section 5.3.3(c)]. This willprevent harmful expansions in most cases (Swenson andGillott 1964; Missouri Highway Department 1967b);however, in pavements where sodium chloride is usedas a deicing chemical, this cannot be taken as certain(Smith 1964; Missouri Highway Department 1%7b).

Avoiding reactive rocks by selective quarrying is thesafest and usually the most economical procedure tominimize alkali-carbonate reactivity. It should be notedthat pozzolans serve only as a diluent and are not ef-fective in mitigating alkali-carbonate reactions.

BA-Preservation of concrete containing reactiveaggregate

There are no known methods of adequately preserv-ing existing concrete which contains the elements thatcontribute to the potentially deleterious chemical reac-tions. Water or moisture is partly involved in at least

two of these reactions. The destructive effects of freez-ing and thawing are more pronounced after the initialstages of destruction by these chemical reactions,Therefore, any practicable means of decreasing the ex-posure of such concrete to water may extend its usefullife. It has been reported that in Iceland, treatment ofvertical concrete surfaces with monosilanes is benefi-cial.

5.5-Recommtlndations for future studiesCurrent criteria employed in the United States that

provide a basis for separating aggregates into reactiveand nonreactive, while generally effective in preventingrecurrences of catastrophic destruction of concretestructures, are now seen to be inefficient in two ways.First, they have often caused more severe precautionsto have been taken than were justified. One example islimiting the calculated cement alkalies to 0.60 percentNa,O equivalent when a higher maximum would surelyhave been safe in some cases. Second, they have some-times permitted alkali-silica reaction to occur to a de-gree causing notable cracking when aggregates errone-ously classed as nonreactive were used with cementscontaining more than 0.60 percent Na,O equivalent.

It is concluded that new research (B. Mather 1975),or a reinterpretation of the results of previous research,is needed to better characterize the following relevantparameters:

a. Degree and rate of aggregate reactivity.b. Influence of concrete mixture proportions, espe-

cially unit cement content.c. Influence of environment on the concrete, espe-

cially temperature and humidity.d. Influence of dimensions of structures, the struc-

tural features, and the stress transfer system on crack-ing developed by alkali-silica reactions.

Additional future research should address optimiza-tion of the use of pouolans and slag, and methods ofdecreasing exposure to water of concrete made with re-active aggregate.

CHAPTER @-REPAIR OF CONCRETEDetailed coverage of concrete repairs falls within the

mission of AC1 Committee 546. This chapter willtherefore give only a brief, general coverage of the sub-ject, with emphasis on the durability aspect. See alsoAC1 224.1R. “Causes, Evaluation and Repair ofCracks in Concrete Structures.”

ill-Evaluation of damage and selection ofrepair method

To evaluate objectively the damage to a structure, itis necessary to determine what caused the damage. Thedamage may be the result of poor design, faulty work-manship, mechanical abrasive action, cavitation orerosion from hydraulic action, leaching, chemical at-tack, chemical reaction inherent in the concrete mix-ture, exposure to deicing agents, corrosion of embed- : i

GUIDE TO DURABLE CONCRETE 2 0 1 . 2 1 - 2 7

ded metal, or other lengthy exposure to an unfavorableenvironment. Guidance for examining and samplinghardened concrete in construction may be found inASTM C 823.

Whatever may have been the cause, it is essential toestablish the extent of the damage, and determine if themajor portion of the structure is of suitable quality onwhich to build a sound repair. Based on this informa-tion, the type and extent of the repair are chosen. Thisis the most difficult step-one which requires a thor-ough knowledge of the subject and mature judgment bythe engineer. If damage is the result of moderate expo-sure of what was an inferior concrete in the first place,then replacement by good quality concrete should as-sure lasting results. On the other hand, if good qualityconcrete was destroyed, the problem becomes morecomplex. In that case, a very superior quality of con-crete is required, or the exposure conditions must bealtered.

The repair of spalls from reinforcing bar corrosion(see Section 4) requires a more detailed study. Simplyreplacing the deteriorated concrete and restoring theoriginal cover over the steel will not solve the problem.Also, if the structure is salt-contaminated, the electro-lytic conditions will be changed by the application ofnew concrete, and the consequences of these changedconditions must be considered before any repairs areundertaken.

6.2-Types of repairs6.2.1 Concrete replacement-The concrete replace-

ment method consists of replacing defective concretewith concrete of suitable proportions and consistency,so that it will become integral with the base concrete.

Concrete replacement is the desired method if thereis honeycomb in new construction or deterioration ofexisting concrete which goes entirely through the wallor beyond the reinforcement, or if the quantity is large.For new work, the repairs should be made immediatelyafter stripping the forms (Tuthill 1960; U.S. Bureau ofReclamation 1975). Considerable concrete removal isalways required for this type of repair. Excavation ofaffected areas should continue until there is no ques-tion that sound concrete has been reached. Additionalchipping may be necessary to accommodate the repairmethod selected and shape the cavity properly.

Concrete for the repair should generally be similar tothe old concrete in nominal maximum size of aggregateand water-cement ratio, provided durability is not sac-rificed. Color is important in some exposed concrete.

Forming will usually be required for large repairs invertical surfaces.

6.2.2 Dry #a&-The dry pack method consists oframming a very stiff mixture into place in thin layers.It is suitable for filling form tie-rod holes and narrowslots, and for repairing any cavity which has a rela-tively high ratio of depth of area. Practically noshrinkage will occur with very stiff mixtures, and theydevelop strength equalling or exceeding that of the par-

ent concrete. The method does not require any specialequipment, but cement finishers must be trained in thistype of repair if the results are to be satisfactory (U.S.Bureau of Reclamation 1975).

6.2.3 Preplaced-aggregate concrete-Preplaced-ag-gregate concrete may be used advantageously for cer-tain types of repairs. It bonds well to concrete and haslow drying shrinkage. It is also well adapted to under-water repairs. This is a specialized process which is de-scribed in AC1 304R.

6.2.4 Shotcrete-Properly applied shotcrete has ex-cellent bond with new or old concrete and is frequentlythe most satisfactory and economical method of mak-ing shallow repairs. It is particularly adapted to verticalor overhead surfaces where it is capable of supportingitself without a form, without sagging or sloughing.Shotcrete repairs generally perform satisfactorily whererecommended procedures of AC1 506R are followed.Simplified equipment has been developed for use insmall repairs (U.S. Bureau of Reclamation 1975).

6.2.5 Repair of scaled areas and spalls in slabs-Scaling of concrete pavement surfaces is not unusualwhere they are subject to deicing salts, particularly ifthe concrete is not adequately air-entrained. Such areasmay be satisfactorily repaired by a thin concrete over-lay provided the surface of the old concrete is sound,durable, and clean (Felt 1956, 1960). A minimum over-lay thickness of about 1% in. (38 mm) is needed forgood performance (AC1 316R). The temperature of theunderlying slab should be as close as possible to that ofthe new concrete.

Spalls may occur adjacent to pavement joints orcracks. Spalls usually are several inches in depth, andeven deeper excavation may be required to remove allconcrete which has undergone some slight degree ofdeterioration. They may be repaired by methods simi-lar to those used for scaled areas.

Numerous quick-setting patching materials, some ofwhich are proprietary, are available. Information onthe field performance of these materials is given byFHWA (1975~).

6.3-Preparations for repairSawcuts around the perimeter of a repair are usually

advisable, particularly in the case of slabs, to eliminatefeather edges. If practicable, the sawcuts should bemade at a slight angle so that the width at the base ofthe patch is greater than at the surface, thereby provid-ing some keying action.

All deteriorated or defective concrete must be re-moved; in the case of slabs, suitable mechanical or hy-draulic scarification equipment should be used. Next,the surfaces of the concrete must be thoroughlycleaned, preferably by wet sandblasting.

Special measures must be taken where chlorides are afactor in the deterioration, as described in Chapter 4.

The bonding surface should have been previously wetdown, and should be damp at the time of patching. Thesurface should be carefully coated with a layer of mor-

201.2R-2e MANUAL OF CONCRETE PRACTICE

tar about ‘/s in, (3 mm) thick, or with another suitablebonding agent (see Section 6.4). If an epoxy bonding’agent is used, the repair surface should be dried beforethe epoxy is applied. The repair should proceed imme-diately after application of the bonding agent, unlessdirected otherwise by the manufacturer of a propri-etary material.

6.4-Bonding agentsBonding layers are generally used to establish unity

between fresh concrete or mortar and the parent con-crete. Sand-cement mortar or neat cement paste hasmost commonly been used in the past. Many reports inthe literature testify to the success of these treatmentswhere recommended practices have been followed.Bonding agents may also be used for additional insur-ance.

Epoxy resins are sometimes used as bonding agents(AC1 503.2). These materials develop a bond havinggreater tensile and shear strength than concrete. Theyare resistant to most chemicals and some formulationsare highly water-resistant. It is not possible to have ac-ceptable results when the concrete is brought to afeather edge. Better results are obtained if a VI in. (20mm) minimum thickness is maintained. These are somedisadvantages in using epoxy resin, such as toxicity andshort pot life. A number of failures of epoxy coatingshave been reported. They have been ascribed to differ-ences in thermal and tensile properties, and moduli ofelasticity of the two materials. These studies are con-tinuing. For most effective results, epoxy bondingagents should not be applied in layers thicker than x6 in.(5 mm). Birdbaths and puddles must be avoided (seeAC1 503.2). Types and grades of epoxies for varioususes are given in ASTM C 88 1.

Other types of bonding agents are available. Certainlatexes, supplied as emulsions or dispersions, improvethe bond and have good crack resistance. Polyvinyl ac-etates, styrene butadienes, and acrylics are among thoseused. However, polyvinyl acetates should not be used,except in dry service conditions (ASTM 1059). Latexesmay be used either as a bonding layer or added to theconcrete or mortar during mixing. The substrate shouldbe wetted down with water prior to placing latex mod-ified concrete.

i&B-Appearance‘Unless proper attention is given to all of the factors

influencing the appearance of concrete repairs, they arelikely to be unsightly. For concrete where appearance isimportant, particular care should be taken to insurethat the texture and color of the repair will match thesurrounding concrete. A proper blend of white cementwith the job cement, or the careful use of pigments, willenable the patch to come close to matching the color ofthe original concrete. A patch on a formed concretesurface should never be finished with a steel trowel,since this results in a dark color which is impossible toremove.

&B-CuringAll conventional concrete or mortar for repairs must

be moist-cured according to the recommendations ofAC1 308. Latexes may require special curing. Epoxyresins require no moist-curing.

&t-Treatment of cracksThe decision of whether a crack should be repaired to

restore structural integrity or merely sealed is depend-ent on the nature of the structure and the cause of thecrack, and upon its location and extent. If the stresseswhich caused the crack have been relieved by its occur-ence, the structural integrity can be restored with someexpectation of permanance. However, in the case ofworking cracks (such as cracks caused by foundationmovements, or cracks which open and close from tem-perature changes), the only satisfactory solution is toseal them with a flexible or extensible material.

Thorough cleaning of the crack is essential beforeany treatment takes place. All loose concrete, old jointsealant, and other foreign material must be removed.The method of cleaning is dependent upon the size ofthe crack and the nature of the contaminants. It mayinclude any combination of the following: compressedair, wire brushing, sandblasting, routing, or the use ofpicks or similar tools.

Restoration of structural integrity across a crack hasbeen successfully accomplished using pressure injectionof low viscosity epoxies (Chung 1975; Stratton andMcCollum 1974) and other monomers (Kukacka et al.1974) which polymerize in situ and rebond the parentconcrete.

Sealing of cracks without restoration of structuralintegrity requires the use of materials and techniquessimilar to those used in sealing joints. A detailed dis-cussion of the types of joint sealants available andmethods of installation is contained in AC1 504R. Sincecracks are generally narrower than joints, some modi-fication in procedure, such as widening the crack witha mechanical router or the use of a low viscosity mate-rial, is often necessary.

CHAPTER 7-USE OF PROTECTIVE~BARRIERSYSTEMS TO ENHANCE CONCRETE

DURABILITY7.1 -Characteristics of a protective-barriersystem

Protective-barrier systems are used to protect con-crete from degradation by chemicals and subsequentloss of structural integrity, to prevent staining of con-crete, or to protect liquids from being contaminated bythe concrete.

A protective-barrier system consists of the barriermaterial, the concrete surface it is to protect, the con-crete structure, and the foundation. The quality of theconcrete, especially at and near the surface, will influ-ence performance of the system because it affects theability of the barrier material to perform as expected.The elements of a protective-barrier system are shown

GUIDE TO DURABLE CONCRETE 201.2R-29

@ Concrete to depth @of 1/4 in. (6 ~1)~. / E%Ei

interface

@ Barrier material

Fig. 7. l-Elements of a protective-barrier system

in Fig. 7.1 and the role of each is explained in Section7.2. An understanding of these elements is essential toobtain optimum performance from protective-barriersystems.

7.2-Elements of a protective barrier-system7.2.1 Barrier material-To be effective in protecting

concrete, a barrier material should have certain basicproperties, as follows:

a. When the barrier material is exposed to chemicalsfrom the environment, the chemicals should not causeswelling, dissolution, cracking, or embrittlement of thebarrier material. Also, the chemicals should not per-meate or diffuse through the barrier to destroy the ad-hesion between it and concrete.

b. The abrasion resistance must be adequate to pre-vent the barrier material from being removed duringnormal service.

c. The adhesive bond strength of a nonbituminousbarrier to the concrete should be at least equal to thetensile strength of the concrete at the surface. Obvi-ously, this bond will also be affected by the cleanlinessof the interface when the barrier material is being ap-plied.

7.2.2 Concrete-barrier interface-Most nonbitumi-nous barrier materials specifically formulated for useover concrete develop and maintain an adhesive bondstrength greater than the tensile strength of the con-crete, provided that the surface is properly prepared.The surface should be free of loose particles, dirt, dust,oil, waxes, and other chemicals that prevent adhesion.Moisture within the concrete may affect the ability of abarrier system to adhere to the surface if water vapordiffusing out of the concrete condenses at the concrete-barrier interface before the barrier has had an oppor-tunity to cure. This problem is discussed in detail inSection 7.4.

7.2.3 Concrete to u depth of ‘/r in, (6 mm)-Perhapsthe most critical part of a nonbituminous barrier sys-

tern is the first % in. (6 mm) of concrete. When a fail-ure occurs, a thin layer of concrete up to ‘/ in. (6 mm),but usually less than ‘/s in. (3 mm) thick, generally ad-heres to the underside of the barrier material. Thismeans that the concrete failed because the internalstresses in the barrier material were greater than thetensile strength of the concrete near the interface. Thesestresses may derive from the following:

1. Shrinkage and polymerization develop stresseswhen the barrier material is cured. This is common toall two-component polymeric materials cured by achemical reaction between the resin and curing agent.

2. Differential volume change in the concrete and thebarrier due to a difference in linear coefficient of ther-mal expansion coupled with a change in temperature.All polymeric barriers have a much higher linear coef-ficient of thermal expansion than concrete. A granularfiller is usually added to the barrier material so that itsthermal coefficient will be closer to that of concrete.

A barrier system should have a low modulus of elas-ticity to prevent stresses from being greater than thetensile strength of concrete over the range of tempera-ture expected for its use. Weak surface concrete can re-sult from use of too high a water-cement ratio, over-working during finishing, the presence of laitance onthe surface, or improper curing. As a result, the con-crete may fail due to the stresses imposed on it even bya low modulus barrier system. Removal of weak sur-face material is essential for satisfactory performanceof these barrier systems. Procedures to accomplish thisare given in Section 3.4 of AC1 515.1R.

7.2.4 Concrete structures-Any cracks in the con-crete, including those that occur before and after appli-cation of the barrier, will reflect through the barrier ifthe concrete is subject to movement from temperaturechanges or from load application. This concrete move-ment can destroy the ability of the barrier to provideprotection for the concrete. A poor quality concreteslab with high permeability may allow groundwater totravel through the concrete so rapidly that the surfacewill never dry sufficiently to allow the barrier to de-velop good adhesion, or it may push the barrier mate-rial away from the concrete.

7.2.5 Foundation conditions-A dimensionally un-stable base or one that does not have sufficient sup-porting strength can cause cracks in the concrete whichare detrimental to these barriers, as discussed previ-ously. Also, the availability and amount of groundwa-ter is a major factor in the success of a barrier. The useof a barrier to water on the exterior surfaces of tanksand tunnels, for example, will retard the entry of waterinto the concrete and is required when an interior pro-tective barrier system is to be applied.

7.3~Guide for selection of protective-barriersystems

7.3.1 Categories of service--Selection of a barriersystem which provides optimum performance at thelowest cost (on cost per year basis) is complicated be-

201.2R-30 MANUAL OF CONCRETE PRACTICE

Table 7.3.1-Protective barrier systems=Qeneral cate9orles (from ACI515.1 R-79)

Severityof chemical

environment

Mild

Intermediate

Severe

Totalnominalthickness

range

Under4 0 mil(1 mm)

Typicalprotective

barrier systems

Polyvinyl butyral.polyurethane. epoxy,acrylic, chlorinated rubber,styrene-acrylic copolymer

Asphalt . coal tar,chlorinated rubber, epoxy,polyurethane, vinyl,neoprene, coal-tar epoxy,coal-tar urethane

125 to375 mil

(3 to 9 mm)

Sand-filled epoxy, sand-filled polyester. sand-filledpolyurethane. bituminousmaterials

20 to Glass-reinforced epoxy,250 mil glass-reinforced polyester,

(S to6mm) precured neoprene sheet,plasticized PVC sheet

Over250 mil(6 mm)

epoxy(b) Asphal t membrane

covered with acid-proofbrick using a chemical-resistant mortar

cause there are so many systems available. To help inthe selection proceSs, protective-barrier systems are di-vided into three general categories accprding to the se-verity of the chemical service conditions; these are mild,intermediate, and severe (see Table 7.3.1).

1. Mild chemical service conditions-Typical exam-ples of mild service conditions are exposure to water,chemical solutions with a pH as low as 4, deicing salts,freezing and thawing cycles, and high purity water.Thicknesses of protective barriers are under 40 mil (1mm). Some of the generic types of materials used forthe barriers include polyvinyl, butyral, acrylics, epoxy,polyurethane, chlorinated rubber, asphalt, coal tar, andvinyls.

2. Intermediate chemical service conditions-Typicalexamples of intermediate service conditions are inter-mittent exposure to dilute acids in chemical, dairy, andfood-processing plants, and abrasion in combinationwith chemicals. Thicknesses of protective barriers are125 to 375 mil (3 to 9 mm). Typical barrier materialsinclude sand-filled epoxy, sand-filled polyester, sand-filled polyurethanes, and bituminous formulations.

3. Severe chemical service conditions-Examples ofsome severe chemical service conditions are dilute orconcentrated mineral or organic acids, and strong al-kali solutions. Thicknesses of these barriers are typi-cally 20 to 250 mil (% to 6 mm), or over 250 mil (6mm) in some cases. Some of the barrier systems are

Typical but notexclusive uses of

protective systemsin order of severity

l Pr;;ection against deicing

l Improve freeze-thawresistance

l Prevent staining of concretel Use for high-purity wattr

s e r v i c e

l Protect concrete in contactwith chemical solutionshaving a pH as low as 4,depending on the chemical

l Protect concrete fromabrasion and intermittentexposure to dilute acids inchemical, dairy, and foodprocessing plants

l Protect concrete tanks andfloors during continuousexposure to dilute mineral,(pH is below 3) organicacids, salt solutions, strongalkalies

* Protect concrete tanks duringcontinuous or intermittentimmersion, exposure towater, dilute acids, strongalkalies, and salt solutions

l Protect concrete fromconcentrated acids orcombinations of acids andsolvents

glass-reinforced epoxy or polyester, precured neoprenesheet, plasticized polyvinyl chloride (PVC) sheet, as-phalt membrane covered with acid-resistant brick usinga chemical-resistant mortar, and a sand-filled epoxysystem top coated with an epoxy barrier.

A description of the materials listed in Table 7 isfound in AC1 515.1R.

7.3.2 Factors affeCng selection-Selection of a bar-rier to protect concrete for a specific chemical servicerequires an awareness of the following items:

1. The barrier material must be resistant to deterio-ration or degradation by the chemicals to which it willbe exposed at the operating temperature.

2. The barrier material must resist the diffusion orpermeation of the chemical through it. Adhesion of thebarrier material to the concrete surface can be ad-versely affected by this phenomenon-especially whenthe diffusing material is acidic.

Chemical resistance and permeation resistance aretwo separate properties. A chemical such as hydro-diloric acid can permeate through various plastic andrubber barrier materials to cause loss of adhesion with-out any indication that the chemical has degraded thebarrier materials.

3. The temperature of the .chemicals contacting thebarrier material will affect performance. Each materialhas its own characteristic maxiinum operating temper-ature for a given chemical environment. Thermal shock

GUIDE TO DURABLE CONCRETE 201.2R-31

caused by rapid changes in temperature can crack somebarrier materials or result in loss of bond between thebarrier and the concrete.

7.3.3 Selection and testing of barrier material--Thereis no guarantee that materials made by different man-ufacturers will perform similarly, even when classifiedas the same generic type. They vary in the types andamounts of ingredients, so their performance will alsovary. In addition, the application characteristics, suchas ease of applying the material to concrete, sensitivityto moisture on a concrete surface, or a very limitedtemperature application range, will affect perform-ance.

The thickness of the barrier required will depend onthe severity of aggressiveness of the environment. Bar-rier selection must be based on testing or past experi-ence. If tests are to be conducted, the entire barriersystem should be applied to concrete specimens beforeexposing them to the actual environment or one thatsimulates as closely as possible this environment. If se-lection must be made before tests of sufficient duration(as agreed between manufacturer and user) can be con-ducted, the barrier supplier should be asked to supplyfully documented case histories where his system hasprotected concrete under the same or similar environ-mental conditions. The selection of a reliable barriermanufacturer and applicator is as important as the se-lection of the barrier itself. AC1 515.1R provides back-ground information on this subject.

7.4--Moisture in concrete and effect on barrieradhesion

Concrete should be dry before the barrier material isapplied. Not only is surface moisture objectionable, butmoisture within the concrete may also affect the abilityof a coating to adhere to the surface. There are no pre-cise guidelines to indicate when moisture will be aproblem although a qualitative test is described in Sec-tion 7.4.1. A brief explanation of how the moisture inconcrete can affect the adhesion of a barrier follows.

Poor barrier adhesion to the concrete can result ifwater vapor diffuses out to the concrete surface. A sur-face that is too damp may produce voids in the barriermaterial and lead to blistering or peeling after it hascured to a hard film. The following factors should beconsidered in determining whether or not this will be aproblem:

1. The rate of vapor transmission through and fromthe concrete.

2. The amount of moisture remaining in the concreteat any given stage.

3. The ability of the coating to breathe, and there-fore, allow moisture to pass through itself.

4. The temperature differential between the concretesurface and ambient air temperature while the coatingis curing. If the concrete temperature is below that ofthe dewpoint of the surrounding air, moisture will con-dense on the surface.

5. The ability of the material to displace moisturefrom the surface.

7.4.1 Dryness of surface-test method-For somebarrier systems, a qualitative moisture test for normalweight concrete is recommended by AC1 Committee503. Moisture content is considered excessive if mois-ture collects at the bond line between the concrete andthe barrier material before the barrier has cured. Thismay be evaluated by taping a 4 by 4 ft (1.2 by 1.2 m)clear polyethylene sheet to the concrete surface and de-termining the time required for moisture to collect onthe underside of the sheet. Also the ambient condi-tions, i.e., sunlight, temperature, and humidity, duringthe test should simulate, as much as practicable, theconditions existing during application and curing of thebarrier. The time for moisture to collect should becompared with the time required for the barrier ‘mate-rial to cure-a value that should be supplied by thematerial manufacturer. If it cures in a time that is lessthan that required for moisture to collect, it may beconcluded that the concrete is adequately dry.

7.5-Influence of ambient conditions onadhesion

For concrete surfaces exposed to the sun it has beenfound that there is better adhesion between the con-crete surface and the barrier material when it is appliedin the afternoon. It is reasoned that exposure to sunand air for at least 6 hr results in a lower surface mois-ture condition. A secondary benefit of applying a bar-rier material in the afternoon is that the surface willnormally have reached its maximum temperature sothat no further expansion of air in the concrete poresand outgassing will occur. This will eliminate the ten-dency for expanding air to cause blistering of the ap-plied barrier material while it is curing.

7.6-Encapsulation of concreteEncapsulation of concrete is a special problem. This

can occur when a concrete slab-on-grade receives a va-por barrier on the underside and is subsequently cov-ered on top with a barrier system. Water can be trappedin the concrete, making it more susceptible to damageby freezing and thawing. In addition, if the concrete isencapsulated during a relatively cool day and then issubjected to higher ambient temperatures, the in-creased vapor pressure of the trapped water could causeloss of adhesion of the barrier material. The use of abreathing barrier can minimize the problem.

CHAPTER 6 - REFERENCES6.1 -Recommended references

The documents of the various standards-producingorganizations referred to in this document are listedbelow with their serial designations.

American Concrete Institute116R Cement and Concrete Terminology-SP 19201.1R Guide for Making a Condition Survey of

Concrete in Service

201.2R-32 MANUAL OF CONCRETE PRACTICE

201.3R

207.1R207.2R

210R211.1

212.3R213R

216R

221R

222R224R224.lR

302.lR

304R

308309R3ll.lR316R

32S.lR330R

332R

345

345.lR357.lR

503R503.2

504R

506R515.lR

546.lR

ASTMc 33c 88

c 94C 138

c 150

Guide for Making a Condition Survey ofConcrete PavementsMass ConcreteEffect of Restraint, Volume Change, and Re-inforcement on Cracking of Mass ConcreteErosion of Concrete in Hydraulic StructuresStandard Practice for Selecting Proportionsfor Normal, Heavyweight, and Mass Con-cre teChemical Admixtures for ConcreteGuide for Structural Lightweight AggregateConcreteGuide for Determining the Fire Endurance ofConcrete ElementsGuide for Use of Normal Weight Aggregatesin ConcreteCorrosion of Metals in ConcreteControl of Cracking in Concrete StructuresCauses, Evaluation, and Repair of Cracks inConcrete StructuresGuide for Concrete Floor and Slab Construc-tionGuide for Measuring, Mixing, Transporting,and Placing ConcreteStandard Practice for Curing ConcreteGuide for Consolidation of ConcreteAC1 Manual of Concrete Inspection-SP 2Recommendations for Construction of Con-crete Pavements and Concrete BasesDesign of Concrete Overlays for PavementsGuide for the Design and Construction ofConcrete Parking LotsGuide to Residential Cast-in-Place ConcreteConstructionStandard Practice for ‘Concrete HighwayBridge Deck ConstructionRoutine Maintenance of Concrete BridgesState-of-the-Art Report on Offshore Con-crete Structures for the ArcticUse of Epoxy Compounds with ConcreteStandard Specification for Bonding PlasticConcrete to Hardened Concrete with a Multi-Component Epoxy AdhesiveGuide to Sealing Joints in Concrete Struc-turesGuide to ShotcreteGuide to the Use of Waterproofing, Damp-proofing, Protective, and Decorative BarrierSystems for ConcreteGuide for Repair of Concrete Bridge Super-structures

Specification for Concrete AggregatesTest Method for Soundness of Aggregates byUse of Sodium Sulfate or Magnesium SulfateSpecification for Ready-Mixed ConcreteTest Method for Unit Weight, Yield, and AirContent (Gravimetric) of ConcreteSpecification for Portland Cement

c 173

C 227

C231

c260

C 289

C 295

c309

c 330

C 342

C441

C 452

c 457

c494

C 586

c 595c 618

C666

C 671

C 672

C 682

c 779

C 823

c 881

C 989

c 1012

Test Method for Air Content of FreshlyMixed Concrete by the Volumetric MethodTest Method for Potential Alkali Reactivityof Cement-Aggregate Combinations (Mortar-Bar Method)Test Method for Air Content of FreshlyMixed Concrete by the Pressure MethodSpecification for Air-Entraining Admixturesfor ConcreteTest Method for Potential Reactivity of Ag-gregates (Chemical Method)Guide for Petrographic Examination of Ag-gregates for ConcreteSpecification for Liquid Membrane-FormingCompounds for Curing ConcreteSpecification for Lightweight Aggregates forStructural ConcreteTest Method for Potential Volume Change ofCement-Aggregate CombinationsTest Method for Effectiveness of MineralAdmixtures or Ground Blast-Furnace Slag inPreventing Excessive Expansion of ConcreteDue to the Alkali-Silica ReactionTest Method for Potential Expansion ofPortland Cement Mortars Exposed to SulfateTest Method for Microscopical Determina-tion of Parameters of the Air-Void System inHardened ConcreteSpecification for Chemical Admixtures forConcreteTest Method for Potential Alkali Reactivityof Carbonate Rocks for Concrete Aggregates(Rock-Cylinder Method)Specification for Blended Hydraulic CementsSpecification for Fly Ash and Raw or Cal-cined Natural Pozzolan for Use as a MineralAdmixture in Portland Cement ConcreteTest Method for Resistance of Concrete toRapid Freezing and ThawingTest Method for Critical Dilation of Con-crete Specimens Subjected to FreezingTest Method for Scaling Resistance of Con-crete Surfaces Exposed to Deicing ChemicalsPractice for Evaluation of Frost Resistance ofCoarse Aggregates in Air-Entrained Concreteby Critical Dilation ProceduresTest Method for Abrasion Resistance of Hor-izontal Concrete SurfacesPractice for Examination and Sampling ofHardened Concrete in ConstructionsSpecification for Epoxy-Resin-Base BondingSystems for ConcreteSpecification for Ground Granulated Blast-Furnace Slag for Use in Concrete and Mor-tarsTest Method for Length Change of Hydrau-lic-cement Mortars Exposed to a Sulfate So-lution

GUIDE TO DURABLE CONCRETE 201.2R-33

C 1017 Specification for .Chemical Admixtures forUse in Producing Flowing Concrete

C 1059 Specification for Latex Agents for BondingFresh to Hardened Concrete

C 1105 Test Method for Length Change of ConcreteDue to Alkali Carbonate Rock Reaction

C 1138 Test Method for Abrasion of Concrete (Un-derwater Method)

Canadian Standards AssociationCSA A23.2-14A Test for Alkali-Aggregate

Reaction

The preceding references are available from

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

ASTM100 Barr Harbor Dr.West Conshohocken, Pa. 194282959

Canadian Standards Association178 Rexdale Blvd.Rexdale, Ontario, CanadaM9W lR3

8.2-Cited referencesAbraham, Thomas J., and Sloan, Richard C., 1979. “Analysis and

Repair of Cracking in TVA’s Fontana Dam Caused by Temperatureand Concrete Growth,” Proceedings, International Congress onLarge Dams, New Delhi, International Committee on Large Dams,Paris, V. 2, pp. l-24.

Acker, P.; Foucrier, C.; and Malier, Y., 1986. “Temperature-Re-lated Mechanical Effects in Concrete Elements and Optimization ofthe Manufacturing Process,” Concrete ut Ear/y Ages, SP-95 Ameri-can Concrete Institute, Detroit, pp. 33-47.

Arni, H.T., 1966. “Resistance to Weathering-Hardened Con-crete, Significance of Tests and Properties of Concrete and Concrete-Making Materials,” STP-169A. American Society for Testing andMaterials, Philadelphia. pp. 261-274.

Bakker, Robert, 1980. “On the Cause of Increased Resistance ofConcrete Made from Blast Furnance Cement to the Alkali-Silica Re-action and to Sulfate Corrosion,” Thesis, RWTH, Aachen, 118 pp.(Translated from the German, Den s’Hertogenbosch, The Nether-lands)

Bastiensen, R.; Mourn, J.; and Rosenquist, I.. 1957. “Some Inves-tigations of Alum Slate in Construction” (Bidragfil Belysning av visseBygningstekniske Problemer ved Osloomradets Alunskifere), Publi-cation No. 22, Norwegian Geotechnical Institute, Oslo, 69 pp. (inNorwegian)

Berke, N. S., 1987. “Effect of Calcium Nitrite and Mix Design onthe Corrosion Resistance of Steel in Concrete (Part 2, Long-Term),”NACE Corrosion 87, Paper No. 132, National Association of Cor-rosion Engineers, Houston.

Berke, N. S., 1985. “Effects of Calcium Nitrite and Mix Design onthe Corrosion Resistance of Steel in Concrete (Part I),” NACE Cor-rosion 85, Paper No. 273, National Association of Corrosion Engi-

neers, Houston.Berke, N. S., and Roberts, L. R., 1989. “Use of Concrete Admix-

tures to Provide Long-Term Durability from Steel Corrosion,” Su-perplasticizers and Other Chemical Admixtures in Concrete, SP-I 19,American Concrete Institute, Detroit, pp. 383403.

Berke, N. S., and Rosenberg, A., 1989. “Technical Review of Cal-cium Nitrite Corrosion Inhibitor in Concrete,” Transportation Re-search Record 1211, Transportation Research Board, Washington,D.C. , p . 18.

Berke, Neal S.; Shen, Ding Feng; and Sundberg, Kathleen M.,1990. “Comparison of the Polarization Resistance Technique to theMacrocell Corrosion Technique,” Corrosion Rates of Steel in Con-crete, STP-1065, ASTM, Philadelphia, pp. 38-51.

Berman, H. A., 1972. “Determination of Chloride in HardenedCement Paste, Mortar and Concrete,” Report No. FHWA-RD-72-12,Federal Highway Administration, Washington, D.C.. Sept.

Bhatty, M. S. Y., and Greening, N. R., 1978. “Interaction of Al-kalies with Hydrating and Hydrated Calcium Silicates,” Proceed-ings, 4th International Conference on the Effects of Alkalies in Ce-ment and Concrete, Publication No. CE-MAT-l-78, School of CivilEngineering, Purdue University, West Lafayette, pp. 87-l 12.

Biczok, Imre, 1972. Concrete Corrosion-Concrete Protection, 8thEdition, Akademiai Kiado, Budapest, 545 pp.

Bisque, R. E., and Lemish, J., I96Oa. “Silicification of CarbonateAggregates in Concrete,” EuNetin No. 239, Highway (Transpor ta-tion) Research Board, pp. 41-55.

Bisque, R. E., and Lemish, J., 1960b. “Effect of Illitic Clay on theChemical Stability of Carbonate Aggregates,” Bulletin No. 275,Highway (Transportation) Research Board, pp. 32-38.

Blaine, R. L.; Arni, H. T.; and Evans, D. N., 1966. “lnterrela-tions Between Cement and Concrete Properties,” Part 2, Section 4,Variables Associated with Expansion in the Potential Sulfate Expan-sion Test, Building Science Series 5, National Bureau of Standards,Washington, D.C., pp. l-26.

Brown, L. S., 1955. “Some Observations on the Mechanics of Al-kali-Aggregate Reaction,” ASTM Bulletin No. 205, p. 40.

Browne, Frederick P., and Cady. Philip D., 1975. “Deicer ScalingMechanisms in Concrete,” Durability of Concrete, SP-47, AmericanConcrete Institute, Detroit, pp. 101-l 19.

Browne, Roger D., 1980. “Mechanisms of Corrosion of Steel inConcrete in Relation to Design, Inspection. and Repair of Offshoreand Coasta l St ructures ,” Performance of Concrete in Marine Envi-ronment, SP-65, American Concrete Institute, Detroit, pp. 169-204.

Buck, A. D., 1%9. “Potential Alkali Reactivity of Carbonate Rockfrom Six Quarries,” Miscellaneous Paper No. C-69-15, U.S. ArmyEngineer Waterway Experiment Station, Vicksburg, 22 pp.

Buck, Alan D., and Dolch, W. L., 1976. “Investigation of a Re-action Involving Nondolomitic Limestone Aggregate in Concrete,”AC1 JOURNAL, Proceedings V. 63, No. 7, July, pp. 755-766.

Buck, Alan D., and Mather, Katharine, 1969. “Concrete Coresfrom Dry Dock No. 2, Charleston Naval Shipyard, S.C.,” Misce//a-neous Paper No. C-69-6, U.S. Army Engineer Waterways Experi-ment Station, Vicksburg, 59 pp.

Buth, Eugene, and Ledbetter, W. B., 1970. “Influence of the De-gree of Saturation of Coarse Aggregate on the Resistance of Struc-tural Lightweight Concrete to Freezing and Thawing,” Highwuy Re-

seurch Record No. 328, Highway (Transportation) Research Board,pp. l-13.

California Department of Transportation, 1978. California Test528, “Test for Freeze-Thaw Resistance of Aggregates in Air-En-trained Concrete (Powers Procedure).”

Callahan, Joseph P.; Lott, James L.; and Kesler, Clyde E.,1970.

%, STlO, pp. 2021-2036.Calleja, J., 1980. “Durability of Cements and Concretes, Proceed-

ings of the 7th International Congress on the Chemistry of Cements, I

V. I, pp. vii/2/1+ii/2/48, Septima, Paris.Chaiken, B., and Halstead, W. J., 1960. “Correlation Between

Chemical and Mortar Bar Test for Potential Alkali Reactivity ofConcrete Aggregate,” BuNetin No. 239, Highway (Transportation)Research Board, pp. 24-40.

1

“Bridge Deck Deterioration and Crack Control,” Proceedings, I

ASCE, V.

201.2R-34 MANUAL OF CONCRETE PRACTICE

Chamberlain, William P.; Irwin, Richard J.; and Am&r, DuaneE., 1977. “Waterproofing Membranes for Bridge Deck Rehabilita-tion.” Reseurch Report No. 52. New York State Department ofTransportation.

Chatterji. S., 1978. “Accelerated Method for the Detection of AI-kali-Aggregate Reactivities of Aggregates,” C&em and ConcreteResearch. V. 8, No. 5. Sept., pp. 647-650.

Chung, H. W., 1975. “Epoxy-Repaired Reinforced ConcreteBeams.” AC1 JOURNAL, proceedings V. 72, No. 5. May, pp. 233-234.

Clear, K. C., 1974a. “Evaluation of Portland Cement Concrete forPermanent Bridge Deck Repair,” Reporf No. FHWA-RD-74-5, Fed-eral Highway Administration, Washington, D.C.. Feb.

Clear, K. C., 1974b, “Permanent Bridge Deck Repair,” PublicRoads, V. 39, No., 2, Sept. 1975. pp. 53-62. Also, Report No.FHWA-RD-74-5, Federal Highway Administration, Washington,D.C.

Clear, K. C., 1976. “Time-to-Corrosion of Reinforcing Steel inConcrete Slabs, V. 3: Performance After 830 Daily Salt Applica-tions,” Report NO. FHWA-RD-7670, Federal Highway Administra-tion, Washington, D.C.

Clear, K. C.. and Harrigan E. T., 1977. “Sampling and Testing forChloride Ion in Concrete,” Report No. FHWA-RD-77-85, FederalHighway Administration, Washington, D.C.

Clear, K. C.. and Hay, R. E.. 1973. “Time-to-Corrosion of Rein-forcing Steel in Concrete Slabs, V. 1: Effect of Mix Design and Con-struct ion Parameters ,” In ter im Report No. FHWA-RD-73-32, Fed-eral Highway Administration, Washington, D.C.

Clear, K. C., and Ormsby, W. C. 1975. “Concept of InternallySealed Concrete.” Interim Report No. FHWA-RD-75-21, FederalHighway Administration, Washington, D.C.

Clemens, Gerard0 G.; Reynolds, Johon V.; and McCormickRandy, 1976. “Comparative Study of Procedures for the Analysis o;Chloride in Hardened Concrete,” Report No. VHTR-77-R7, Vir-ginia Highway and Transportation Research Council, Appendix 3.

Clifton, J. R.; Beeghly, H. F.; and Mathey, R. G., 1974. “Non-metallic Coatings for Concrete Reinforcing Bars.” Final Report No.FHWA-RD-74-18, National Bureau of Standards for Federal High-way Administration, Washington, D.C.

Collins, A. R., 1944. “Destruction of Concrete by Frost,” Jour-nu/, Institute of Civil Engineers (London), Paper No. 5412, pp. 29-41.

Coombes, L. H.; Cole, R. D.; and Clarke, R.,M., 1975. “Reme-dial Measures to Val-de-la-Mare Dam, Jersey, Channel Islands,”BNCOLD Symposium, Newcastle-upon-Tyne, England.

Coombes, L. H. 1976. “ValdelaMare Dam, Jersey, Channel Is-lands,” The wfect of Alkalies on the Properties of Concrete, A. B.Poole, Editor, Cement and Concrete Association, Wexham Springs,pp. 357-370.

Cordon, William A., 1966. “Freezing and Thawing of Concrete-Mechanisms and Control,” Monograph No. 3, American ConcreteInstitute/Iowa State University Press, Detroit, 99 pp.

Corps of Engineers, 1985. Standard Practice for Concrete, EM1110-2-2000, U.S. Army Corps of Engineers, Office, Chief of Engi-neers, Washington, DC.

Dahir, S. H., 1981. “Relative Resistance of Rained-On ConcretePavements to Abrasion, Skidding, and Scaling,” Cement, Concrete,

undAggregates, ASTM, V. 3, No. 1, Summer, pp. 13-20.Diamond, Sidney, 1975. “Review of Alkali-Silica Reaction and

Expansion Mechanisms: 1. Alkalies in Cements and in Concrete PoreSolutions,” Cement und Concrete Reseurch, V. 5. No. 4, July, pp.329-346.

Diamond, Sidney, 1976. “Review of Alkali-Silica Reaction andExpansion Mechanisms: 2. Reactive Aggregates,” Cement und Con-crete Research, V. 6, No. 4, July, pp. 549-560.

Diamond, Sidney. Editor, 1978a. Proceedings, 4th InternationalConference on Effects of Alkalies in Cement and Concrete, fublicu-tion No. CE-MAT-I-78, School of Civil Engineering, Purdue Uni-versity, West Lafayette, 376 pp.

Diamond, Sidney, 1978b. “Chemical Reactions Other than Car-bonate Reactions,” Significance of Tests and Properties of Concreteand Concrete Making Materials, STP-169B, Chapter 40. ASTM,

Philadelphia, pp. 708-821.Dikem J. T., 1975, “Flyash Itqreases Resistance of Concrete to

Sulfate Attack,” Reseatvh Report No. 23, U.S. Bureau of Recfama-tion, Denver, If pp.

Dikmu, J. T., 1976. “Review of Worldwide Developments and UseOf Polymers in Concrete,” Polymers in Concrete, Proceedings of the1st International Congress, Concrete Construction Publications, Ad-dison, pp. 2-8.

DNIBR, 19561%5. Committee on Alkali Reactions in ConcreteReports Al, 1957; Bl, 1958; B2. 1958; B3, 1958; Dl, 1957; D2, 1958;El, 1959; F1.2.3, 1958; HI, 1958; 11, 1958; 12. 1966; 13. 1967; Kl,1% K2, 1958; Ll, 1957; Ml. 1958; Nl, 1956; N2, 1961; NJ, I%l;N4, 1964; N5. 1964; N6, 1964. Danish National Institute of BuildingResearch and the Academy of Technical Sciences, Copehagen.

Dolar-Mantuani. L., 1964. “Expansion of Gull River CarbonateRocks in Sodium Hydroxide,” Highway Rescerch Record No. 45,Highway (Transportation) Research Board, pp. 178-195.

Dolar-Mantuani, L., 1969. “Alkali-Silica Reactive Rocks in theCanadian Shield.” Highway Research Record No. 268, Highway(Transportation) Research Board, pp. 99-l 17.

Dolar-Mantuani, Ludmila, 1971. “Late Expansion of Alkali-Re-active Carbonate Rocks,” Highwuy Research Record No. 353. High-way (Transportation) Research Board. pp. l-14.

Dolar-Mantuani, Ludmila, 1983. Hundbook of Concrete Aggegutes, Noyes Publications, Park Ridge, Chapter 7, pp. 79-125.

Duncan. M. A. G.: Swenson, E. G.; Gillott, J. I!.; and ForanM . R., 1973. “Alkali-Aggregate Reaction in Nova Scotia: 1. Sum:mary of a Five Year Study,” Cement und Concrete Revearrch, V. 3,No. 1, Jan., pp. 55-69.

Duncan, M. A. G.; Swenson, E. G.; and Gillott, J. E., 1973. “Al-kali-Aggregate Reaction in Nova Scotia: III. Laboratory Studies ofVolume Change,”pp. 233-245.

Cement and Concrete Reseamh, V. 3, No. 3, May,

Dunn, J. R., and Hudec, P. P., 1965. “Influence of Clays on Wa-ter and Ice in Rock Pores,” Report No. RR65-5. New York StateDepartment of Public Works.

Dunstan, E. R., Jr., 1976. “Performance of Lignite and Subbitu-minous Flyash in Concrete-A Progress Report,” Report No. REC-ERC-76-l. U.S. Bureau of Reclamation, Denver, 23 pp.

Erlin, Bernard, 1966. “Methods Used in Petrographic Studies ofConcrete,” Analytical Techniques for Hydraulic Cement and Con-crete, STP-395, ASTM, Philadelphia, pp. 3-17.

Erlin, Bernard, and Hime. William G., 1976. “Role of CalciumChloride in Concrete,” Concrete Construction, V. 21, No. 2, Feb.,pp. 57-61.

Erlin, B., and Woods, H., 1978. “Corrosion of Embedded Mate-rials Other Than Reinforcing Steel,” Signficunce of Test and Properties of Concrete and Concrete Making Materials, STP-169B.ASTM, Philadelphia, Chapter 20.

Federal Highway Administration, 1975a. “Bridge Deck ProtectiveSystems, Membranes, Polymer Concrete and Dense Portland Ce-ment Concrete,” Interim Report NEEP No. 12. Notice N5080.23,Washington, D.C.

Federal Highway Administration, 1975b. “Coated ReinforcingSteel,” Interim Report No. 1. NEEP Project No. 18, FHWA NoticeN5080.33, Washington, D.C.

Federal Highway Administration, 1975~. “AASHTO-FHWA Spe-cial Products Evaluation List (SPEL),” Report No. FHWA-RD-7641, Washington, D.C.

Federal Highway Administration, 1976. “Use of Galvanized Re-bars in Bridge Decks,” Notice No. 5, 140.10. Washington, D.C.

Feldman, R. F.. and Sereda, P. J. 1961. “Characteristics of Sorp-tion and Expansion Isotherms of Reactive Limestone Aggregate,”AC1 JOURNAL, Procex!dings V. 58. No. 2. Aug., pp. 203-214.

Felt. Earl J., 1960. “Repair of Concrete Pavement,” AC1 JOUR-NAL, Ptvceedings V. 57, No. 2. Aug., pp. 139.153.

French, W. J., and Poole, A. B.. 1976. “Alkali-Aggressive Reac-tions and the Middle East.” Concrete, V. 10. No. 1, pp. 18-20.

Gaynor. R. D., 1967. “Laboratory Freezing and Thawing Tests-A Method of Evaluating Aggregates,” National Aggregates Associa-tion Circular No. 101, 32 pp.

GUIDE TO DURABLE CONCRETE 201.2R-35

Gaynor, Richard D., 1985. “Understanding Chloride Percent-ages,” Concrete International: Design & Construction, V. 7, No. 9,Sept. pp. 26-27.

Gewertz, M. W., 1958, “Causes and Repair of Deterioration to aCalifornia Bridge Due to Corrosion of Reinforcing Steel in a MarineEnvironment: Part l-Method of Repair,” Highway Research Bulle-tin No. 182, Highway (Transportation) Research Board, pp. I-17.

Gibson, W. E., 1938 “Study of Map Cracking in Sand-GravelConcrete Pavements,” Proceedings, Highway (Transpor ta t ion) Re-search Board, V. 18, Part 1, pp. 227-231.

Gillott, J. E.. 1%3a. “Cell Test Method for Study of Alkali-Car-bonate Rock Reactivity,” Proceedings, ASTM, V. 63, pp. 11951206.

Gillott, J. E., 1%3b. “Mechanism and Kinetics of the Alkali-Car-bonate Rock Reaction,” Canadian Journal of Earth Sciences (Ot-tawa), V. 1, pp. 121-145.

Gillott, J. E.; Duncan, M. A. G.; and Swenson, E. G., 1973. “Al-kali-Aggregate Reaction in Nova Scotia: IV. Character of the Reac-tion,” Cement and Concrete Research, V. 3, No. 5, Sept. pp. 521-536.

Gillott, J. E., and Swenson, E. G., 1%9. “Mechanism of the Al-kaliCarbonate Reaction,” Journal of Engineering Geology, V. 2, pp.7-23.

Gjgrv, Odd E., 1957. Concrete in the Oceans, Marine SciencePublications, pp. 51-74.

Gjgrv, Odd E., 1%8. “Durability of Reinforced Concrete Wharvesin Norwegian Harbours,” Norwegian Committee on Concrete in SeaWater, lgeniorforlaget A/S Oslo, 208 pp.

Gogte, B. S., 1973. “Evaluation of Some Common Indian Rockswith Special Reference to Alkali-Aggregate Reactions,” EngineeringGeology, No. 7, pp. 135-153.

Grattan-Bellew, Patrick E., Editor, 1987. 7th International Con-ference on Alkal i -Aggregate React ion, Ottawa, Noycs P u b l i c a t i o n s ,Park Ridge, 509 pp.

Griffin, Donald F., 1%9. “Effectiveness of Zinc Coating on Re-inforcing Steel in Concrete Exposed to a Marine Environment,”Technical Note No. N-1032, U.S. Naval Civil Engineering Labora-tory, Port Hueneme, July, 42 pp. Also 1st Supplement, June 1970,and 2nd Supplement, June 1971.

Hadley, David W., 1964. “Alkali Reactivity of Dolomitic Carbon-ate Rocks,” Highway Research Record No. 45, Highway (Transpor-tation) Research Board, pp. l-20.

Hadley, David W., 1968. “Field and Laboratory Studies on theReactivity of Sand-Gravel Aggregates,” Journal, PCA Research andDevelopment Laboratories, V. 10, No. 1, Jan., pp. 17-33.

Hagerman. T., and Roosaar, H., 1955. “Damage to ConcreteCaused by Sulfide Minerals,” Betong (Stockholm), V. 40, No. 2, pp.151-161.

Halldorsson, Ottar P., Editor, 1975. Symposium on Alkali-Aggre-gate Reaction, Preventive Measures (Reykjavik, Aug.), Building Re-search Institute, Keldnaholt, Reykjavik, 270 pp.

Halstead, S., and Woodworth, L. A., 1955. “Deterioration of Re-inforced Concrete Structures Under Coastal Conditions,” Transac-

tions, South African Institute of Civil Engineers, V. 5, No. 4, pp.115-134.

Hansen, W. C., 1944. “Studies Relating to the Mechanism byWhich the Alkali-Aggregate Reaction Produces Expansion in Con-crete,” AC1 JOURNAL , Proceedings V. 40, No. 3, Jan. , pp. 213-228.

Hansen, W. C., 1%3. “Crystal Growth as a Source of Expansionin Portland-Cement Concrete,” Proceedings, ASTM, V. 63, pp. 932-945.

Hansen, W. C., 1964. “Anhydrous Minerals and Organic Materi-a ls as Sources of Dis t ress in Concrete ,” Highway Research RecordNo. 43, Highway (Transportation) Research Board, pp. l-7.

Harman, John W., Jr.; Cady, Philip D.; and Boiling, Nanna B.,1970. “Slow-Cooling Test for Frost Susceptibility of PennsylvaniaAggregates ,” Highway Research Record No. 328, Highway (Trans-portation) Research Board, pp. 26-37.

Helmuth, R. A., 1961. “Dimensional Changes of Hardened Port-land Cement Pastes Caused by Temperature Changes,” Proceedings,Highway (Transportation) Research Board, V. 40, pp. 315-336.

Helmuth, R. A., 196Oa. “Capillary Size Restrictions on Ice For-mation in Hardened Portland Cement Pastes,” Proceedings, Fourth

International Symposium on the Chemistry on Cement, MonographNo. 43, National Bureau of Standards, Washington, DC., V. 2, pp.855-869.

Helmuth, R. A., 196Ob. Discussion of “Frost Action in Concrete”by Paul Nerenst, Proceedings, Fourth International Symposium onthe Chemistry of Cement, Monograph No. 43, National Bureau ofStandards, Washington, D.C., V. 2. pp. 829-833.

Hill, G. A.; Spellman, D. L.; and Stratfull, R. F., 1976. “Labo-ratory Corrosion Tests of Galvanized Steel in Concrete,” Transpor-tation Research Record No. 604, pp. 25-37.

Hiltrop, C. L., and Lemish, J., 1960. “Relationship of Pore-SizeDistribution and Other Rock Properties to Serviceability of SomeConcrete Aggregates,” Bulletin No. 239, Highway (Transportation)Research Board, pp. l-23.

Hinds, Julian, and Tuthill, Lewis H., 1941. Discussion of “Crack-ing in Concrete Due to Expansive Reaction Between Aggregate andHigh Alkali Cement as Evidenced in Parker Dam,” by H. S. Meis-sner, AC1 JOURNAL, Proceedings V. 37, No. 5, Apr., pp. 568-lthrough 568-3.

Helm, J., 1987. “Comparison of the Corrosion Potential of Cal-cium Chloride and a Calcium Nitrite Based Non-Chloride Accelera-tor-A Macro-Cel l Corrosion Approach,” Corrosion, Concrete , and

Chlorides, SP-102, American Concrete Institute, Detroit, pp. 35-48.Honig, A., 1984. “Radiometric Determination of the Density of

Fresh Shielding Concrete In Situ,” In Situ/Nondestructive Testing ofConcrete, SP-82, American Concrete Institute, Detroit, pp. 603-618.

ldorn, G. M., 1958. “Concrete on the West Coast of Jutland-Parts 1 and II,” Danish Institute of Building Research and Academyof Technical Sciences, Committee on Alkali Reactions in Concrete,Copenhagen, 56 pp. and 54 pp.

ldorn, G. M., 1967. Durability of Concrete Structures in Den-

mark, Technical University of Denmark, Copenhagen, 208 pp.ldorn, G. M., 1980. “Interface Reactions Between Cement and

Aggregate in Concrete and Mortar-Bond Strength and Durability,”General Report, Theme VII, Proceedings, 7th International Congresson the Chemistry of Cement, Paris, V. IV, pp. 129-147.

ldorn, G. M., and Rostam, Steen, Editors, 1983. Proceedings, 6thInternational Conference on Alkalis in Concrete, Research and Prac-tice, (Copenhagen, June 22-25), Dansk Betonforening, Vester Fari-magsgade 31, DK-1606 Kobenhaven V. Denmark, 532 pp.

ldorn, G. M., and Roy, Della M., 1986. “Opportunities with Al-kalies in Concrete Testing, Research, and Engineering Practice,” Al-kalies in Concrete, STP-930, ASTM, Philadelphia, pp. 5-15.

Jackson, F. H., 1946. “Durability of Concrete in Service,” ACI-JOURNAL , Proceedings V. 43, No. 2, Oct. , pp. 165-180.

Jenkins, G. H., and Butler, J. M., 1975. “Internally Sealed Con-crete,” Report No. FHWA-RD-75-20, Monsanto Research Corpora-tion for the Federal Highway Administration, Washington, D.C., Fi-nal Report.

Kalousek, G. L.; Porter, L. C.; and Benton, E. J., 1972. “Con-crete for Long-Time Service in Sulfate Environment,” Cement and

Concrete Research, V. 2, No. I, Jan. pp. 79-89.Kammer, H. A., and Carlson, R. W., 1941. “Investigation of

Causes of Delayed Expansion of Concrete in Buck HydroelectricPlant,” AC1 JOURNAL, Proceedings V. 37, No. 6, June, pp. 665-671.

Kauer, J. A., and Freeman, R. L., 1955. “Effect of Carbon Di-oxide on Fresh Concrete,” AC1 JOURNAL, Proceedings V. 52, No. 4,Dec., pp. 447-454. See also Discussion, pp. 1299-1304.

Kettle, R., and Sadegzadeh, M., 1987. “Influence of ConstructionProcedures on Abrasion Resistance,” Concrete Durability-Proceed-

ings of the Katharine and Bryant Mather International Conference,SP-100, V. 2, American Concrete Institute, Detroit, pp. 1385-1410.

Keyser, J. Hode, 1971. “Resistance of Various Types of Bitumi-nous Concrete and Cement Concrete to Wear by Studded Tires,”Highway Research Record No. 352, Highway (Transportation) Re-search Board, pp. 16-31.

Klieger, P., 1956. “Curing Requirements for Scale Resistance ofConcrete,” Bulletin No. 150, Highway (Transportation) ResearchBoard, Jan.

Klieger, Paul, Editor, 1982. George Verbeck Symposium on SW-fate Resistance of Concrete, SP-77, American Concrete Institute,Detroit, 94 pp.

Klieger, Paul, and Hanson, J. A., 1%1. “Freezing and Thawing

201.2R-36 MANUAL OF CONCRETE PRACTICE

Tests of Lightweight Aggregate Concrete,*’ AC1 J OURNAL’ proceedings V. 57, No. I, Jan., pp. 7797%.

Krukar, Milan, and Cook, John C., 1973. “Effect of Studded Tireson Various Pavements and Surfaces,” Highway Reseamh Record No.

477, Highway (Transportation) Research Board’ pp. 4-8.Kukacka, L. E.. et al., 1974. “Concrete-Polymer Materials for

Highway Applications: Progress Report No. 3,” Report No. FHWA-RD-74-17’ Federal Highway Administration, Washington, D.C.

Kummer, H. W., and Meyer, W. E., 1967. “Tentative Skid-Resis-tance Requirements for Main Rural Highways,” NCHRP Report No.37’ Highway (Transportation) Research Board, 80 pp.

Lea, F. M., 1971. The Chemistry of Cement and Concrete, 1stAmerican Edition, Chemical Publishing Company’ New York, 727PP.

Ledbetter, W. B., 1973, “Synthetic Aggregates from Clay andShale: A Recommended Criteria for Evaluation”’ Highway ResearchRecord No. 430’ Highway (Transportation) Research Board, pp. 9-15.

Lemish, J.. and Moore, W. J., 1964. “Carbonate Aggregate Rcactions: Recent Studies and an Approach to the Problem,” HighwayReswrch Record No. 45, Highway (Transportation) Research Board’pp. 57-71.

Lenzer. Dennis, 1981. “Untersochonzen Zur Alkali-ZuschlagReaktion mit Opalsandstein aus Schleswig-Holsten,” dissertationRWTH, Aachen. 125 pp.

Lerch. William, 1959. “Cement-Aggregate Reaction that Occurswith Certain Sand-Gravel Aggregates,” Journal, PCA Research andDevelopment Laboratories, V. 1. No. 3, Sept., pp. 42-50.

Lewis’ D. A., 1962. “Some Aspects of the Corrosion of Steel inConcrete,” Proceedings, First International Congress on MetallicCorrosion, London, pp. 547-555.

Lewis’ D. W.’ 1956. “Deterioration of Structural Concrete in In-diana,” Extension Series No. 88, Engineering Reprint No. 88’ Engi-neering Experiment Station, Purdue University, 97 pp.

Litvan. G. G.’ 1972. “Phase Transitions of Adsorbates: IV,Mechanism of Frost Action in Hardened Cement Paste”’ Journ&,

American Ceramic Society’ V. 55. NO. 1, pp. 38-42.Litvan, 0.0.’ 1975. “Phase Transitions of Adsorbates: VI, Effect

of Deicing Agents on the Freezing of Cement Paste,” Journal,

American Ceramic Society, V. 58, No. 1-2, pp. 2630.Litvan. 0. G.. 1976. “Frost Action in Cement in the Presence of

De-icers.” Cement and Concrete Research, V. 6, No, 3. pp. 351-356.Lossing, Fay A., 1966. “Sulfate Attack on Concrete Pavements in

Mississippi,” Highwuy Reseorefr Record No. 113. Highway (Trans-portation) Research Board’ pp. 88-102.

Lovell. C. E., 1928. “Heavy Duty Concrete Floors,” AC1 JOUR-NAL. Proceedings V. 24, pp. 454465.

Ludwig, Udo, 1980. “Durability of Cement Mortars and Con-cretes,” Durability of Building Materials and Components, STP-691,ASTM, Philadelphia, pp. 269-281.

Luke, W. 1.. 1964. “Alkali-Carbonate Reaction in Concrete fromChickamauga Dam Powerhouse,” Proceedings, ASTM, V. 64, pp.887902.

Manns, W.. and Eichler, W. R., 1982. “Corrosion-Promoting Ac-tion of Concrete Admixtures Containing Thiocyanate,” Betonwerk &Fertigteil-Technik (Weisbaden). V. 48. No. 3, pp. 154-162.

Marusin, S. L., and Pfeifer, D. W., 1985. “Chloride Ion Penetra-tion into Concrete Made with Various Admixtures,” RILEM Sym-posium 85-Technology of Concrete when Pozzolans, Slags andChemical Admixtures Are Used, Monterey, N.L., Mexico.

Mather, Bryant , 1948 . “Petrographic Identification of ReactiveConstituents in Concrete Aggregates”’ Proceedings, ASTM’ V. 48.pp. 1120-I 125.

Mather, Bryant, 1966. “Effects of Seawater on Concrete,” High-wuy Research Record No. 113, Highway (Transportation) ResearchBoard, pp. 33-42.

Mather. Bryant, 1968. “Field and Laboratory Studies of the Sul-phate Resistance of Concrete”’ Performance of Concrete-Resis-tance of Concrete to Sutphate and Other Environmental Conditions,Thorvaldson Symposium, University of Toronto Press’ Toronto’ pp.66-76.

Mather, Bryant, 1969. “Sulfate Soundness, Sulfate Attack, and

Expansive Cement in Concrete,” Proceedings, RILEM InternationalSymposium on the Durability of Concrete, Prague, 12 pp.

Mather, Bryant, 1975. “New Concern Over Alkali-Aggregate Re-action,” Proceedings, Symposium on Alkali-Aggregate Reaction,(Reykjavik, Aug.), Building Research Institute, Keldnoholt. Reykja-vik. pp. 17-20.

Mather, Bryant, 1980. “Mhmral Aggregates for Concrete-NeededResearch,” Progm in Concrete Technology, Canada Centre forMineral and Energy Technology, Ottawa, pp. 57-110 (with 127 Ref-erences and 4 Annexes).

Mather, Bryant, 1981. “Blends of Cementitious Materials forConcrete to be Exposed to Sea Water”’ Proceedings, 5th Interna-tional Symposium on Concrete Technology, Universidad Autonomade Nuevo Leon, Monterry, N. L., Mexico, pp. 289-310.

Mather’ Katharine, 1973. “Examination of Cores from FourHighway Bridges in Georgia,” MisceNanmus Paper No. C-73-11.U.S. Army Engineer Waterways Experiment Station, Vicksburg.

Mather, Katharine, 1978a. “Petrographic Examination”’ Signfl-

qmce of Test and Properties of Concrete and Concrete-Making MO-teriu&, STP-169B. ASTM, Philadelphia, pp. 132-145.

Mather, Katharine’ 1978b. “Tests and Evaluation of Portland andBlended Cements for Resistance to Sulfate Attack,“ Cement Ston-da&: Evolution and Trends, STP-663, ASTM, Philadelphia, pp. 74.86.

Mather, Katharine, 1981. “Factors Affecting Sulfate Resistance ofMortars,“ Proceedings, 7th International Congress on the Chemistryof Cement, Paris, V. IV, pp. 580-585.

Mather, Katharine, 1982. “Current Research in Sulfate Resistanceat the Waterways Experiment Station,“ George Verbeck Symposiumon Sulfate Resistonce of Concrete, SP-77, American Concrete Insti-tute, Detroit, pp. 63-74.

Mehta, P. K., 1976. Discussion of “Combating Sulfate Attack inCorps of Engineers Concrete Construction,“ by Thomas J. Reading,AC1 JOURNAL, Proceedings V. 73, No. 4, Apr., pp. 237-238.

Mehta, P. K.. 1981. “Sulfate Resistance of Blended Portland Ccments Containing Poxxolans and Granulated Blast-Furnace Slag,”Proceedings, 5th International Symposium on Concrete Technology,Univers idad Autonoma de Nuevo Leon, Monterry, N. L. , Mexico,pp. 35-50.

Meissner, H. S., 1941. “Cracking in Concrete Due to ExpansiveReaction Between Aggregate and High Alkali Cement as Evidencedin Parker Dam,” AC1 JOURNAL’ Proceedings V. 37. No. 5, Apr., pp.549-568.

Mielenz, Richard C., 1954. “Petrographic Examination of Con-crete Aggregate,“ Proceedings, ASTM, V. 54, pp. 1188-1218 (p.1214).

Mielent, R. C., 1958. “Petrographic Examination of ConcreteAggregate to Determine Potential Alkali-Reactivity,” Reseurch Pu-per No. 18-C, Highway (Transportation) Research Board, pp. 29-38.

Mielenx, R. C., 1964. “Reactions of Aggregates Involving Solubil-ity Oxidation, Sulfates or Sulfides,“ Highway Reseamh Record No.43, Highway (Transportation) Research Board, pp. 8-18.

Mielenr. Richard C., 1978. “Petrographic Examination” (Con-crete Aggregates), Signifcance of Tests and Properties of Concrete

Muking Muter& STP-169B, ASTM, Philadelphia, Chapter 33, pp.536-572.

Miesenhelder, P. D., 1960. “Effect of Design and Details on Con-crete Deterioration,” AC1 J OURNAL, Proceedings V. 56, No. 7, Jan.,pp. 581-590.

Missouri Highway Department, 1%7a. “Effects of Reactive Car-bonate Aggregates on the Durability of Laboratory Concrete Speci-mens,“ Report No. 67-5, Missouri Cooperative Highway ResearchProgram.

Missouri Highway Department, 1%7b. “Alkali Reactivity of Car-bonate Rocks in Missouri,“ Report No. 76-6, Missouri CooperativeHighway Research Program.

Monfore’ G. E., and Verbeck, 0. J., 1960. “Corrosion of Pre-stressed Wire in Concrete”’ AC1 JOURNAL. Wings V. 57, No.5, Nov., pp. 491-515.

Mourn, Johan. and Rosenquist, I. Th., 1959. “Sulfate Attack onConcrete in the Oslo Region,“ AC1 ;ICWRNAL. Proceedings V. 56, No.

3, Sept., pp. 257-264.Murphy, Wiliam E., 1975. “Skidding Resistance of Concrete

GUIDE TO DURABLE CONCRETE 201.2R-37

Pavements: A Review of Research, Development and Practice in theUnited Kingdom,” Roudwuys ond ‘Airport Pavements, SP-5 1, Amer-ican Concrete Institute, Detroit, pp. 231-256.

Myers, John J., Editor, 1969. Handbook of Oceun und Underwu-ter Engineering, McGraw-Hill Book Company, New York.

National Bureau of Standards, 1939. Reseurch Puper RP1252, Nov.Newlon, Howard, Jr., 1978. “Resistance to Weathering,” Signif-

cunce of Tests and Properties of Concrete and Concrete-Muking Mu-teriuls, STP-169B, ASTM, Philadelphia, pp. 351-368.

Newlon, Howard H., Jr.; 0201, Michael A.; and Sherwood, W.Cullen, 1972. “Potentially Reactive Carbonate Rocks, Progress Re-port No. 5, An Evaluation of Several Methods for Detecting AlkaliCarbonate Reaction,” Virginia Highway Research Council.

Newlon, Howard H., Jr., and Sherwood, W. Cullen, 1964.“Methods for Reducing Expansion of Concrete Caused by Alkali-Carbonate Rock Reactions,” Highwuy Reseurch Record No. 45,Highway (Transportation) Research Board, pp. 134-150.

Newlon, Howard H., Jr.; Sherwood, W. Cullen; and 0201, Mi-chael A., 1972. “Potentially Reactive Carbonate Rocks, ProgressReport No. 8, A Strategy for Use and Control of Potentially Reac-tive Carbonate Rocks Including an Annotated Bibliography of Vir-ginia Research,” Virginia Highway Research Council.

New Zealand Portland Cement Association, 1975. “Vacuum Con-crete Dewatering,” IBOOl, Wellington.

Nmai, Charles K., and Corbo, Jack M.. 1989. “Sodium Thiocya-nate and the Corrosion Potential of Steel in Concrete and Mortar,”Concrete Internotional: Design & Construction, V. 1 I, No. 11, Nov.,pp. 59-67.

Oberholster, R. E., Editor, 1981. Proceedings, Fifth InternationalConference on Alkali-Aggregate Reaction in Concrete, (Cape Town,Mar.-Apr.), National Building Research Institute, Pretoria.

Oberholster, R. E., and Davies, G., 1986. “Accelerated Method forTesting the Potential Reactivity of Siliceous Aggregates,” Cementund Concrete Research, V. 16, No. 2, Mar., pp. 181-189.

Orbon, B., 1969. “Effect on Concrete Pavement of Studded Win-ter Tires,” First European Symposium on Concrete Pavements, Paris.

0~01, M. A., and Newlon, H. H., Jr., 1974. “Bridge Deck Dete-rioration Promoted by Alkali-Carbonate Reaction: A DocumentedExample,” Trunsportation Research Record No. 525, Transporta-tion Research Board, pp. 55-63.

Pepper, Leonard, and Mather, Bryant, 19.59. “Effectiveness ofMineral Admixtures in Preventing Excessive Expansion of ConcreteDue to Alkali-Aggregate Reaction,” Proceedings, ASTM, V. 59, pp.1178-1202.

Pfeifer, Donald W.; Landgren, J. Robert; and Zoob, Alexander,1987. “Protective Systems for New Prestressed and SubstructureConcrete,” FHWA Report No. RD-86-193, Federal Highway Ad-ministration, Washington, D.C.

Philleo, R. E., 1986. “Freezing and Thawing Resistance of High-Strength Concrete,” NCHRP Synthesis No. 129, Transportation Re-search Board, Washington, D.C., 31 pp.

Pickett, Gerald, 1953. “Flow of Moisture in Hardened CementDuring Freezing,” Proceedings, Highway Research Board, V. 32, pp.276-284.

Poole, Alan B., Editor, 1976. “Effect of Alkalies on the Proper-ties of Concrete,” Cement and Concrete Association, England, 374pp., available as WES MP C-77-10 from NTIS as AD A044 565(CTIAC Report No. 26).

Portiand Cement Association (PCA), 1968. “Effect of VariousSubstances on Concrete and Protective Treatments, Where Re-quired,” Publication No. ISOOIT, Skokie.

Powers, T. C., 1945. “Working Hypothesis for Further Studies ofFrost Resistance of Concrete,” AC1 JOURNAL, Proceedings V. 41,No. 4, Feb., pp. 245-272.

Powers, T. C., 1954. “Void Spacing as a Basis for Producing Air-Entrained Concrete,” AC1 JOURNAL, Proceedings V. 50, No. 9. May,pp. 741-760.

Powers, T. C., 1955. “Basic Considerations Pertaining to Freezingand Thawing Tests,” Proceedings, ASTM, V. 55, pp. 1132-1155.

Powers, Treval C., 1956. “Resistance of Concrete to Frost at EarlyAges,” Proceedings, RILEM Symposium on Winter Concreting (Co-penhagen, 1956). Danish National Institute of Building Research,

Copenhagen, Session C, pp. l-50. Also, Research Bulletin No. 71,Portland Cement Association.

Powers, T. C., 1965. “Mechanisms of Frost Action in Concrete,”Stanton Walker Lecture No. 3, National Sand and Gravel Associa-tion/National Ready Mixed Concrete Association, Silver Spring, 35PP.

Powers, T. C., 1975. “Freezing Effects in Concrete,” Durubility ofConcrete, SP-47, American Concrete Institute, Detroit, pp. l-1 1.

Powers, T. C., and Steinour, H. H., 1955. “Interpretation of SomePublished Researches on the Alkali-Aggregate Reaction: Part 1, TheChemical Reactions and Mechanism of Expansion,” AC1 JOURNAL,Proceedings V. 5 1, No. 6, Feb., pp. 497-5 16.

Preus, C. K., 1971. Discussion of “Resistance of Various Types ofBituminous Concrete and Cement Concrete to Wear By StuddedTires,” by J. Hode Keyser, Highway Research Record No. 352.Highway (Transportation) Research Board, pp. 31-38.

Preus, C. K., 1973. “After Studs in Minnesota,” Highwuy Re-search Record No. 477, Highway (Transportation) Research Board,pp. 11-15.

Price, Waiter H., 1947. “Erosion of Concrete by Cavitation andSolids in Flowing Water,” AC1 JOURNAL, Proceedings V. 43, No. 9,May, pp. 1009-1023. See also Discussion, pp. 1024-l and 1024-2.

Prior, M. E., 1966. “Abrasion Resistance-Hardened Concrete,”Significunce of Test and Properties of Concrete and Concrete-Muk-ing Muteriuls, STP-169A, ASTM, Philadelphia, pp. 246-260.

Reading, Thomas J., 1975. “Combating Sulfate Attack in Corps ofEngineers Concrete Construction,” Durability of Concrete, SP-47,American Concrete Institute, Detroit, pp. 343-366

Regourd, M.; Hornain, H.; Levy, P.; and Morturaux, B., 1980.“Resistance of Concrete to Physico-Chemical Attack,” Proceedings,Seventh International Conference on the Chemistry of Cement, Paris.

RILEM, 1962. “Durability of Concrete,” RILEM Symposium,General Reports, RILEM Bulletin (Paris), No. 14, Mar., 149 pp.

Rogers, Christopher A., 1986. “Evaluation of the Potential forExpansion and Cracking of Concrete Caused by the Alkali-Carbon-ate Reaction,” Cement, Concrete und Aggregates, ASTM, V. 8, No.I, Summer, pp. 13-23.

Rose, J. G., and Ledbetter, W. B., 1971. “Summer of SurfaceFactors Influencing the Friction Properties of Concrete Pavements,”Highway Research Record No. 357, Highway (Transportation) Re-search Board, pp. 53-63.

Roy, Della M., and Idorn, Gunnar M., 1982. “Hydration, Struc-ture, and Properties of Blast-Furnace Slag Cements, Mortars, andConcrete,” AC1 JOURNAL, Proceedings V. 79, No. 6, Nov.-Dec., pp.444-457.

Scholer, C. H., and Smith, G. M., 1954. “Rapid Accelerated Testfor Cement-Aggregate Reaction,” Proceedings, ASTM, V. 54, pp.1165-1177.

Schuman, Louis, and Tucker, John, Jr., 1939. “Portable Appara-tus for Determining the Relative Wear Resistance of ConcreteFloors,” Journul of Research, National Bureau of Standards, V. 23,No. 5, Research Paper RP-1262, pp. 549-570.

Scripture, E. W., Jr.; Benedict, S. W.; and Bryant, D. E., 1953.“Floor Aggregates,” AC1 JOURNAL, Proceedings V. 50, No. 4, pp.305316. See also Discussion, pp. 316-1 through 316-5.

Sellevold. E. J., and Nilson, T., 1987. “Condensed Silica Fume inConcrete-A World View,” Supplementary Cementing Muterials forConcrete, SP-86-8E, CANMET, Energy Mines and Resources, Ot-tawa.

Sherwood, W. Cullen, and Newlon, Howard H., Jr., 1964. “Stud-ies on the Mechanisms of Alkali-Carbonate Reaction, Part I. Chem-ical Reactions,” Highway Reseurch Record No. 45, Highway (Trans-portation) Research Board, pp. 41-56.

Smith, F. L., 1956. “Effect of Various Surface Treatments UsingMagnesium and Zinc Fluosilicate Crystals on Abrasion Resistance ofConcrete Surface,” Concrete Loboratory Report No. C-819, U.S.Bureau of Reclamation, Denver.

Smith, F. L., 1958. “Effect of Aggregate Quality on Resistance ofConcrete to Abrasion,” Cement und Concrete, STP-205, ASTM,Philadelphia, pp. 91-106.

Smith, Peter, 1964. “Learning to Live with a Reactive CarbonateRock,” Highwuy Research Record No. 45, Highway (Transporta-

201.2R-38 MANUAL OF CONCRETE PRACTICE

tion) Research Board, pp. 126133.Smith, Peter, 1974. “Fifteen Years of Living at Kingston with a

Reactive Carbonate Rock,” Transportat ion Research Record, No.525. Transportation Research Board, pp. 23-27.

Smith, P., and Schonfeld. R.. 1970. “Pavement Wear Due toStudded Tires and the Economic Consequences in Ontario,” High-way Research Record No. 331, Highway (Transportation) ResearchBoard, pp. 54-79.

Smith, P., and Schonfeld, R., 1971. “Studies of Studded-TireDamage and Performance in Ontario During the Winter of 1969-70,”Highway Research Record No. 352, Highway (Transpor ta t ion) Re-search Board, pp. l-15.

Smock, J. G. 1975. “Polymer Impregnation of New ConcreteBridge Deck Surface, Interim User’s Manual of Procedures andSpecifications,” Report No. FHWA-RD-75-72, U.S. Bureau of Rec-lamation for the Federal Highway Administration, Washington,D.C., June.

Soles, J. A., 1982. “Thermally Destructive Particles in Sound Do-lostone Aggregate from an Ontario Quarry,” Cement, Concrcre andAggregates, ASTM, V. 4, No. 2. Winter, pp. w-102.

Stanton, Thomas E., 1940. “Expansion of Concrete Through Re-action Between Cement and Aggregate,” Proceedings, ASCE. V. 66,Dec., pp. 1781-1812.

Stark, David, 1978. “Alkali-Silica Reactivity in the Rocky Moun-tain Region,” Proceedings, 4th International Conference on Effectsof Alkalies in Cement and Concrete, Purdue University, pp. 235-243.

Stark, David, and Bhatty, Muhammed S. Y., 1986. “Alkali-SilicaReactivity: Effect of Alkali in Aggregate on Expansion,” A/ka/ies in

Concrete, STP-930. ASTM, Philadelphia, pp. 16-30.Stark. David, and Perenchio. William, 1975. “Performance of

Galvanized Reinforcement in Concrete Bridge Decks,” Final Report,

Portland Cement Association, Skokie.Stratfull, R. F., 1973. “Half Cell Potentials and the Corrosion of

Steel in Concrete,” presented at the 52nd Annual Meeting of theHighway Research Board.

Stratfull, R. F., 1974. “Experimental Cathodic Protection of aBridge Deck,” Interim Report No. FHWA-RD7431, California De-partment of Transportation.

Stratfull, R. F.; Jurkovich, W. J.; and Spelfman, D. L., 1975.“Corrosion Testing of Bridge Decks,” Transportation Reseaffh Ret-ord No. 539, Transportation Research Board. pp. 50-59.

Stratton, F. W., and McCollum, B. F.. 1974. f’lepair of Hollowor Soft Areas in Bridge Deck by Rebonding with Injected EpoxyResin or Other Polymer,” Report No. TF 72-5. State Highway Com-mission of Kansas.

Sturrup, V.; Hooton, R., Mukherjee, P.; and Carmichael, T.,1987. “Evaluation and Prediction of Concrete Durability-OntarioHydro’s Experience,” Concrerc Durability-Katharine and Bryant

Mather International Conference, SP-100, American Concrete Insti-tute, Detroit, V. 2, pp, 1121-I 154.

Swenson, E. G., 1957. “Reactive Aggregate Undetected by ASTMTests,” ASTM Bullet in No. 226, Dec. , pp . 48 -50 .

Swenson, E. G., Editor, 1968. Performance of Concrete-Resb-tame of Concrete to Sulphate and Other Environmental Conditions,Thorvaldson Symposium, University of Toronto Press, Toronto, 243PP.

Swenson, E. G., and Giilott, J. E.. 1960. “Characteristics ofKingston Carbonate Rock Reaction,” Bulletin No. 275, Highway(Transportation) Research Board. pp. 18-3 1.

Swenson, E. G., and Gillott, J. E., 1964. “Alkali-Carbonate RockReaction.” Highway Research Record No. 45. Highway (Transpor-tation) Research Board, pp. 21-40.

Swenson, E. G., and Gillott, J. E., 1967. “Alkali Reactivity ofDolomit ic Limestone Aggregate ,” Magazine o f Concrete Research

(London), V. 19, No. 59, June, pp. 95-104.Terzaghi, Ruth D., 1948. “Concrete Deterioration in a Shipway,”

AC1 JOURNAL, Proceedings V. 44, No. 10, June, pp. 977-1005.Terzaghi, Ruth D., 1949. “Concrete Deterioration Due to Car-

bonic Acid.” Journa/, Boston Society of Civil Engineers, V. 36.Thornton, Henry T., Jr., 1978. “Acid Attack of Concrete Caused

by Sulfur Bacter ia Action,” AC1 JOURNAL , Proceedings V. 75. No .

Il. Nov., pp. 577-584.

Thurmann. Moe T., 1969. “Pavement Wear Caused bv StuddedTires” (Piggdekkenes Slitasge pa Vegdekker), Norwegian StateHighway Laboratory, C&lo. ’

Tran&rtation Research Board, 1958. “Chemical Reactions ofAggregates in Concrete,” Special Report No. 31. Highway (Trans-portation) Research Board, pp. 1-12.

Transportation Research Board. 1959. “Report on CooperativeFreezing and Thawing Tests of Concrete,” Special Report No. 47,Highway (Transportation) Research Board, 67 pp.

Transportation Research Board, 1975. “Effects of Studded Tires,”NCHRP Synthesis of Highway Pract ice No. 32, Transporta t ion Re-search Board, 1975.46 pp.

Tremper, Bailey; Beaton, John L.; and Stratfull, R. F., 1958.“Causes and Repair of Deterioration to a California Bridge Due toCorrosion of Reinforcing Steel in a Marine Environment: Part II-Fundamental Factors Causing Corrosion,” Bulletin No. 182, High-way (Transportation) Research Board, pp. 18-41.

Tremper, Bailey, and Spellman, D. L., 1961. “Tests for Freeze-Thaw Durability of Concrete Aggregates,” Bulletin No. 305, High-way (Transportation) Research Board, pp. 28-50.

Tuthill. Lewis H., 1960. “ Conventional Methods of Repair ingConcrete,” AC1 JOURNAL . Proceedings V. 57, No. 1, July, pp. 129-138.

Tuthill, Lewis H., 1978. “Resistance to Chemical Attack,” Signif-icance of Tests and Properties of Concrete and Concrete-Making

Materials, STP-169B, ASTM. Philadelphia, pp. 369-387.Tuthill, Lewis H., 1982. “Alkali-Silica Reaction-40 Year Later,”

Concrete International: Design & Construction, V. 4, No. 4, Apr.,pp. 32-36.

Tutti. Kyosti, 1982. “Corrosion of Steel in Concrete,” SwedishCement and Concrete Research Institute, S-100 44 Stockholm.

Tyler, I. L., 1960. “Long-Time Study of Cement Performance inConcrete, Chapter It-Concrete Exposed to Sea Water and FreshWater,” AC! JOURNAL, Proceedings V. 56, No. 9, Mar., pp. 825-836.

USBR, 1975. Concrete Manual. 8th Edit ion, U.S. Bureau of Rec-lamation, Denver, 627 pp.

Van Daveer, J. R., and Sheret, G. D., 1975. “Concrete CoverStudy,” Finat Report No. FHWA-DP-15, Federal Highway Admin-istration, Washington, D.C.

Vanden Bosch, V. D., 1980. “Performance of Mortar Specimensin Chemical and Accelerated Marine Exposure,” Performance ofConcrete in Marine Environment, SP-65, American Concrete. Insti-tute. Detroit, pp. 487-507.

Van Til, C. J.; Carr, B. J.; and Vallerga, B. A., 1976. “Water-proof Membranes for Protection of Concrete Bridge Deck-Labora-tory Phase,”Board, 70 pp.

IVCHRP Report No. 165, Transportation Research

Verbeck, George J., 1958. “Carbonation of Hydrated PortlandCement.” cement and Concrete, STP-205, ASTM, Philadelphia, pp.17-36. Also. Research and Development Laboratories, Portland Ce-ment Association, Bulletin 37.

Verbeck. G. J., 1%8. “Field and Laboratory Studies of the Sul-phate Resistance of Concrete.” Performance of Concrete-Resistanceof Concrete to Sulphate and Other Environmental Conditions, Thor-valdson Symposium, University of Toronto Press, Toronto, pp. 113-124. Also reprinted as PCA Research Department Bulletin 227, Port-land Cement Association. Skokie, 1969.

Verbeck, G. J., 1978. “Pore Structure-Hardened Concrete,”Significance of Test and Propetiies of Concrete and Concrete-Mak-ing Materio& STP-169B. ASTM, Philadelphia, pp. 262-274.

Verbeck, G. J., and Klieger, P., 1957. “Studies of ‘Salt’ Scaling ofConcrete,” Bulletin No. 150, Highway (Transportation) ResearchBoard, pp. l-13.

Verbeck, G. J., and Landgren, R., 1960. “Influence of PhysicalCharacteristics of Aggregates on the Frost Resistance of Concrete.”Proceedings, ASTM. V. 30, pp. 1063-1079.

Verein Deutscher Zementwerke, 1973. “Vorbeugende Massnah-men gegen Alklaireaktion im Beton,” Schriftenreihe der Zementin-dustrie, V. 40. 101 pp.

Visvervaraya, H. C.; Rajkumar,’ C.; and Mullick, A. K., 1987;“Analysis of Distress Due to Alkali-Aggregate Reaction in GalleryStructures of a Concrete Dam,” Proeedings, 7th International Con-

GUIDE TO DURABLE CONCRETE 2bl*2R-39

ference on Alkali-Aggregate Reaction, Noyes Publications, ParkRidge, Ottawa, pp. 188-193.

Walker, Hollis N., 1974. “Reaction Products in Expansion TestSpecimens of Carbonate Aggregate,” Transportation Research Rec-ord No. S25, Transportation Research Board, pp. 28-37.

Walker, Hollis N., 1978. “Chemical Reactions of Carbonate Ag-gregates in Cement Paste,” Significance of Tests and Properries ofConcrete and Concrete-Making Materials, STP-169B, ASTM, Phila-delphia, pp. 722-743.

Wehner, B., 1966. “Beanspruchung der Strassenooerflaech durchWinterreifen mit Spikes,” Technische Universitaet Berlin, Institutefur Strassen und Verkehrswesen.

Welp, Theodore L., and De Young, Clarence E., 1964. “Varia-

tions in Performance of Concrete with Carbonate Aggregates inIowa,” Highway Research Record No. 45, Highway (Transporta-tion) Research Board, pp. 159-177.

Wilk, W., 1978. “Consideration of the Question of Skid Resis-tance of Carriageway Surfaces, Particularly of Concrete,” Beton-strassen N. 117. Monograph.

Witte, L. P., and Backstrom, J. E., 1951. “Some Properties Af-fecting the Abrasion Resistance of Air-Entrained Concrete,” Pro-

ceedings, ASTM, V. 51, pp. 1141-1155.

Woods, Hubert, 1%8. Durability of Concrete Construction, Mon-ograph No. 4, American Concrete Institute/Iowa State UniversityPress, Detroit, 187 pp.

THE FOLLOWING DISCUSSIONS, WHICH WERE PUBUSHED IN THE MAY-JUNE 1992 ACI MaterialsJournal (PP. 311.313), ARE NOT PART OF THE REPORT ACI 201.2R-992, BUT ARE PROVIDED ASADDITIONAL INFORMATION TO THE READER.

Proposed revision of: Guide to Durable Concrete. Paper by ACI Committee 201

Discussion by Alexander M. Leshchinsky, Richard A. Mackow, and ACI Committee 201

by ALEXANDER M. LESHCHINSKY

Member American Concrete hiitute. Consulting Engineer, Concrete AdviceP/L, Croydon, Victoria, Austral ia

Of course it is difficult to classify guides in concretetechnology on a criterion of their importance. But, ifone could do it, the Guide to Durable Concrete wouldbe among the most important. Members of the com-mittee have done great work; therefore, the writerwants to discuss only some positions, mainly from thefirst two chapters of this guide.

From the chapter devoted to freezing and thawing,the effects of freezing and thawing of concrete are dif-ferent in different conditions. Therefore, in Clause 1.1,it is reasonable to give the typical conditions of freez-ing and thawing, for instance, according to the classi-fication of Alexeev et al. (1990): 1) freezing in air andthawing in water; concrete with different inital degreesof water saturation; 2) freezing and thawing in perma-nent capilliary suction; 3) freezing and thawing in wa-ter.

Describing four main processes (and the causes ofconcrete deterioration in each of them) which takeplace during freezing of concrete is also desirable(Moskvin and Golubyh 1975). In this case, mechanismsof frost action could be discussed with regard to spe-cific conditions of this action. In Clause 1.2, the phe-nomenological theory of concrete durability (Podvalny1976) also could be mentioned, since it has a practicalvalue.

To the requirements which must be satisfied for con-crete durability (Clause 1.4), it is necessary to add“minimal water content.” Shestoperov (1966) showed

that concretes with less water content had higher dura-bility (at a constant W/C ratio).

For durable concretes, cements with C,A < 8 per-cent should be used (Clause 1.4.4. l), since mineral3Ca0-AlZOj-6H20 loses strength in water (Shestope-rov 1966).

The writer assumes that it is necessary to accept andsay that concretes with blast furnace slag cements haveless freeze-thaw resistance than concretes with portlandcements (other things being equal), in spite of the factthat using air-entrained admixtures considerably in-creased the freeze-thaw resistance of the former. Thehigh durability of concretes with blast furnace slag ce-ments in marine structures of European countries (Bel-gium, the Netherlands, France, and Germany) takesplace mainly because of higher resistance of concretewith these cements to chemical and physical (e.g.,washing out calcium hydroxide) attack of seawater thanbecause of their high durability in freezing and thawingenvironments, which cannot be classified as “severe” inthis region.

Regarding the chapter on aggressive chemical expo-sure, there are opinions (about Clause 2.2.1) that theformation of ettringite is not the main cause of con-crete disruption in sulfate attack (Chaterji and Jeffrey1983). Mehta (1983) did not find the direct correlationbetween an amount of ettringite formed in cement andits expansion.

Alexeev et al. (1990) discussed works of Ludwig,Mehr, and Morales, who had found in old structures,as a result of sulfate attack and carbonization, a for-mation of thaumasite (CaCO,+CaSO,*CaSiO,* 14 H2Owhich loosens a cement structure.

201‘2R-40 MANUAL OF CONCRETE PRACTICE

It is also reasonable to describe attack mechanism inthe case of seawater exposure, e.g., according to Re-gourd, Hornain, and Mortureux (1980).

In general, the authors did not mention biocorrosionof concrete. A lot of space was given to alkali-silica andalkali-carbonate reactions, for instance, in comparisonwith sulfate attack, which is more dangerous and takesplace more often (Cohen and Mather 1991).

REFERENCESAkxeev. S. N.; Ivanov, F. M.; Modry, Sl.; and SchiesseJ. P., 1990.

‘Durability of Reinforced Concrete in Aggressive Media,” Strojiz-dat. Moscow, 320 pp.

Chaterji, S., and Jeffrey, I. M., 1983. “A New Hypothesis of Sul-phate Expansion,” hfogazine of Concrete Rcmwch, No. 44. pp. 83-86.

Cohen, M. D., and Mather. B., 1991. “Sulphate Attack on Con-crete-Research Needs,” V. 88, Jan.-Feb., pp. 62-69.

Mehta. K., 1983. “Mechanism of Sulphate Attack on PortlandCement,” Cement and Conctvte Restwmh, V. 13, No. 3, pp. 401406.

Moskvin, V. M., and Golubyh, N. D., 1975. “Deterioration ofConcrete under Freezing”, 2nd RILEM Symposium on Winter Con-creting, Strojizdat, Moscow, V. 1. pp. 1 M-125.

Podvahty, A. M., 1976 “Phenomenological Aspect of ConcreteDurability Theory.” Moteriols and Structures, No. 51, pp. 151-162.

Regourd. M.; Hornain, H.; and Mortureux, B., 1980. “Micro-structure of Concrete in Aggressive Environments,” Proceedings, 1stNational Conference on Durability of Buildings Materials and Com-ponents, ASTM, STP 691, pp. 253-268.

Shestoperov. S. V., 1966. “Durability of Concrete in TransportStructures,” Transport, Moscow, 500 pp.

By RICHARD A, MACKOWMembar Amwican Concrete Institute, P.E., District Manager, East/NorthuavtRegion. JTU Industries. Inc., AlIentown, PA

My comments will be written from the point of viewof a Class F fly ash supplier, while other comments aregeneral in nature.

Section 1.4.4.2--This section states that “most flyashes and natural pozxolans, when used as admixtures,have little effect on the durability of concrete providedthat the air content, strength and moisture of the con-crete are similar.” This is true, but it should be pointedout that equal strengths are a function of how fly ashmixes are proportioned. I believe that the method ofproportioning to achieve equal strength is somewhatesoteric. The reason why I say this is that many re-search projects that compare strength and durabilityseem destined to show Class F fly ash mixes at a dis-advantage because of 25 percent or 50 percent portlandcement replacements on very lean mixes. When pro-portioning Class F fly ash mixes, it is normal proce-dure to add a greater amount of fly ash than portlandcement replaced. The replacement range is normally inthe 10 to 20 percent range, while the final percentage offly ash to toal cementitious materials is usually in the 15to 25 percent range. Class C fly ashes are unlike ClassF ashes in that cement replacements can easily be 25 to35 percent, The key to durable FAC is to understandthat strength (assuming equal air-void systems) must besimilar to the ordinary portland cement concrete(OPCC).

When cement factors are high and 20 to 30 percentfly ash is used, permeability can drop dramatically (Ar-maghani, Larsen, and Roman0 1991).

If durability is viewed from a slightly different pro-spective, the CANMET study of Malhotra, Carette,and Bremner (1987) indicates that at about 900 F-T cy-cles, a 25 percent mix with 1: 1 replacement of cementwith fly ash performs well at w/(c + p) = 0.40 to 0.50,but declined slightly at w(c + p) = 0.60 (see Table 12,SP-100). The 28day strengths of the FAC were typi-cally 25 percent lower than the OPCC mixes at w(c +P) = 0.40 and 0.50, and were 20 to 35 percent lowerthan the OPCC mixes at w(c + p) = 0.60. The infer-ence here is that a FAC mix will outperform OPCCwhen proportioned for equal strength because of re-duced permeability.

If minimum cement factors versus visual rating arecompared in the CANMET study, the following may bethe minimum suggested cement content and strengths toproduce F-T durability results comparable to OPCC:

WC + P) Minimum portland cement0.40 281 kg/m3 (474#/CY)0.50, 23 1 kg/m’ (39O#/CY)0.60 182 kg/m’ (31OUKY)

As far as freeze-thaw durability with deicing chemicals,minimum cement factors may have to be raised to in-sure adequate protection, since deicers can increase wa-ter penetration into concrete.

Section 3.3 - With regard to abrasion, the state-ment is made on p. 556 that “compressive strengths atthe surface can be improved by . . . 2. Eliminatingbleeding.” Since all concretes bleed, except perhapsthose incorporating silica fume, and nom&-entrainedconcrete will have more bleed water than air-entrainedmixes, I wonder if the committee means “reduce bleed-ing” rather than “eliminate bleeding.”

Section 5.2.5b. - Since the ASTM C 441 mortar barexpansion test limit of 0.020 percent is unrealisticallylow for today’s cement and fly ash combinations, analternative would be to compare a low-alkali cementexpansion with a higher alkali cement plus fly ash. AC1Committee 201 may wish to contact ASTM CommitteeC 618 to update itself on the progress of the proposedASTM C 618 revision.

With regard to the comment on drying shrinkage inthe same section, good quality ASTM C 618 Class F flyash does not increase drying shrinkage or water de-mand. Higher water demand can result if nonspecifi-cation grade fly ash which is coarse (high percent re-tained on No. 325 sieve) and has a very high carboncontent. With a finer fly ash, a greater water reductioncan be realized. Class F fly ash mixes can be propor-tioned for equal, early, or later strength, dependingupon the amount of cement replaced by fly ash, theamount of fly ash used, and the initial cement factor.

Perhaps future research can be geared around FACthat is proportioned for strength equal to OPCC, so thetrue potential of fly ash can be realized.

GUIDE TO DURABLE CONCRETE 201.2R-41

REFERENCESArmaghani; Larsen; and Romano, 1991. “Strength and Durability

of Concrete in Florida,” Florida DOT, FL/DOT/SM0/91-389, July.Malhotra, V.; Carette, G.; and Bremner, T., “Durability of Con-

crete Containing Supplementary Cementing Materials in Marine En-vironment,” Concrete Durability - Katharine and Bryant Matherinternational Conference, SP-100, American Concrete Institute, De-troit, V. 2, pp. 1227-1258.

COMMITTEE CLOSUREThe Committee appreciates very much the interest in

its report manifested by Messrs. Leshchinsky and Mac-kow, and finds their detailed comments to be ofconsiderable interest. The Committee especially appre-ciates Mr. Leshchinsky’s references to the Russian lit-erature since, as has already been called to its attentionby one of its members, one weakness of the report is itsfailure comprehensively to take account of considera-ble literature arising in parts of the world other thanNorth America, Western Europe, and the Pacific Rim.

Mr. Leshchinsky notes our failure to mention “bio-corrosion of concrete” and he references Cohen andMather (1991). This reference also fails to mention“biocorrosion.” In fact, the Committee does not havea proper definition of the term nor a reference to theoccurrence of the phenomenon. Kleinlogel (1950) dis-cusses many influences on concrete, many of which areorganic or produced by plants or animals, but theCommittee is not aware of any reason for groupingthese as “biocorrosion.” They include acetic acid, al-cohol, animals, beer, buttermilk, dung, fatty oils, gly-cerine, etc.

Mr. Leshchinsky is entirely correct - the effects offreezing and thawing are different in different condi-tions. The Committee believed that this point was,madein the report and in the literature it cited. We did notreference the classification proposed by Ivanov (1990)because the idea of doing so was not suggested to us.Had it been, we might have chosen not to do so be-cause it seems unlikely that the completely continuousgradation change from conditions where an unlimitednumber of cycles of any degree of severity have no ef-fect (concrete with no freezable water) to those where afew cycles can cause significant damage seems not wellindicated by his three-part classification. We invite Mr.Leshchinsky’s attention to Mather (1990), which ap-peared too late for us to cite.

The Committee does not find the evidence in favor ofan 8 percent limit on C9A in cement as a requirementfor frost resistance to be convincing. AC1 225R, whichwe probably should have referenced, remarks “. . . theproperties of cement are important as they influence thestrength and permeability of the concrete at the time ofexposure to freezing and thawing,” but makes no ref-erence to chemical or phase composition of cement.

The Committee is aware that some authors havequestioned that the formation of ettringite is the majorfactor in damage to concrete by sulfate attack. TheCommitte cited Lea (1971) and Mehta (1976) for therebeing two basic reactions involving formation of gyp-sum and formation of ettringite. We find convincingthe concept that the cause of most of the expansioncaused by sulfate attack to be the result of ettringiteformation, partly by analogy to the extensive workdone on expansive cements. (See AC1 223, another AC1report we could have cited.)

Mr. Mackow writes from the standpoint of a Class Ffly ash supplier. The Committee hopes he will make hisexperience available to AC1 Committee 226 on Fly Ash.It was our assumption that, normally, project concretehas a required design strength that is not changed as afunction of whether or not fly ash is an ingredient ofthe mixture. Hence, it seemed reasonable to discuss theeffect of fly ash on durability of concrete “providedthat the air content, strength, and moisture content ofthe concrete are similar” (Section 1.4.4.1). It may wellbe, as Mr. Mackow infers from the studies reported byMalhotra et al. (1987). that when mixtures containingfly ash are proportioned for equal strength to compa-rable mixtures without fly ash, there will be improveddurability due to reduced permeability. The Committeedid report (Section 2.3.3) that the “permeability ofconcrete made with appropriate amounts of suitableground blast-furnace slag or pozzolan can be as low as1110th or l/lOOth that of comparable concrete of equalstrength made without slag or pozzolan (Bakker1980).” In Section 5.2.5 the Committee listed as a“recommended procedure to be used with alkali-reac-tive aggregates,” use “with appropriate amounts of aneffective pozzolan or slag, or both” (Pepper andMather 1959).

REFERENCESAC1 223, Practice for the Use of Shrinkage-Compensating ConcreteAC1 225R, Guide to the Selection and Use of Hydraulic Cements

Cohen, Menashi D., and Mather, Bryant. 1991. “Sulfate Attack onConcrete - Research Needs,” ACI Material Journal, V. 88, No. 1,Jan.-Feb., pp. 62-69.

Ivanov, F. M., in Alexeev, S. N.; Ivanov, F. M.; Modry, Sl.; andSchiessel, P., 1990. “Durability of Reinforced Concrete in Aggres-sive Media,” Strojizdat, Moscow, 320 p p .

Kleinlogel, A., 1950. Influences on Concrete, translated from theGerman (1941 edition) by F. S. Morgenroth, Frederick Ungar Pub-lishing Co., New York, 281 pp.

Mather, Bryant, 1990. “How to Make Concrete That Will be Im-mune to the Effects of Freezing and Thawing,” Paul Klieger Sym-posium on Performance of Concrete, AC.1 SP-122, American Con-crete Institute, Detroit, 1990, pp 1-18.

Mahta, P. K., 1983. “Mechanism of Sulfate Attack on PortlandCement Concrete - Another Look,” Cement and Concrete Re-

search, V. 13, pp. 401-406.