ACI 212.3R-91

Chemical Admixtures for ConcreteReported by ACI Committee 212

(Reapproved 1999)

Edwin A. Decker Joseph P. FlemingChairman Secretary

John M. Albinger Bayard M. CallFloyd Best David A. HuntReid H. Brown Kenneth R. LauerW. Barry Butler Bryant Mather*

Members of the committee voting on the 1991 revisions:

Richard C. MielenzWilliam F. Perenchio*William S. Phelan*Michael F. Pistilli

Dale P. RechDonald L SchlegelRaymond J. Schutz*Billy M. Scott

Paul R. Stodola*David A. Whiting*Arthur T. WintersJ. Francis Young

William F. Perenchio Chairman

Joseph P. FlemingSecretary

Greg BobrowskiReid H. BrownW. Barry ButlerBayard M. CallEdwin A. Decker

Guy DetwilerGunnar M. IdornBryant MatherRichard C. Mielenz

William S. PhelanMichael F. PistilliJohn H. ReberDale P. Rech

M. Roger RixomDonald L SchlegelRaymond J. SchutzBilly M. Scott

Paul R. StodolaDavid A. WhitingArthur T. WintersJ. Francis Young

This sixth report of ACI Committee 212, now named “Chemical Ad-mixtures for Concrete, ” updates the previous reports of 1944, 1954,1963, 1971, and 1981. Admixtures discussed herein are those knownas chemical admixtures; finely divided mineral admixtures have beentransferred to ACI Committee 226. Admixtures are classified into fivegroups: (I) air-entraining; (2) accelerating; (3) water-reducing and set-controlling; (4) admixtures for flowing concrete; and (5) miscella-neous.

Preparation and batching, which had a separate chapter in the 1981report, are included here in Chapter 1. Chapter 5, “Admixtures forNo wing Concrete, ”is new, representing technology that has ma-tured since 1981. Any of those admixtures possessing propertiesidentifiable with more than one group are discussed with the groupthat describes its most important effect on concrete.

Keywords: accelerating agents; adhesives; admixtures: air-entraining agents; al-kali-aggregate reactions; bactericides; batching; calcium chlorides; colors (materials);concretes; corrosion inhibitors; expanding agents; flocculating; foaming agents;fungicides; gas-forming agents; insecticides; permeability-reducing admixtures; pig-ments; plasticizers; pumped concrete; retardants; waterproofing admixtures;water-reducing agents.

CONTENTSChapter 1-General information, p. 212.3R-2

1 .1 -Introduction1.2-Reasons for using admixtures1 .3-Specifications for admixtures1.4-Sampling1.5-Testing1.6-Cost effectiveness1.7-Considerations in the use of admixtures1.8-Decision to use1.9-Classification of admixtures1.10 -Preparation and batching

ACI Committee Reports, Guides, Standard Practices, andCommentaries are intended for guidance in designing, plan-ning, executing, or inspecting construction and in preparingspecifications. Reference to these documents shall not be madein the Project Documents. If items found in these documentsare desired to be part of the Project Documents they shouldbe phrased in mandatory language and incorporated into theProject Documents.

212.3R-

Chapter 2-Air-entraining admixtures, p. 212.3R-72.1-Introduction2.2-Entrained-air-void system2.3-Effect on concrete properties2.4-Materials for air entrainment2.5-Applications2.6-Evaluation, selection, and control of purchase2.7-Batching, use, and storage2.8-Proportioning of concrete2.9-Factors influencing amount of entrained air2.10-Control of air content of concrete

Chapter 3-Accelerating admixtures, p. 212.3R-103.1-Introduction3.2-Types of accelerating admixtures3.3-Use with special cements3.4-Consideration of use3.5-Effect on freshly mixed and hardened concrete3.6-Wet- and dry-process shotcrete3.7-Control of purchase3.8-Batching and use3.9-Proportions of concrete3.10-Control of concrete

Chapter 4-Water-reducing and set-controllingadmixtures, p. 212.3R-14

4.1-General4.2-Classification and composition4.3-Application4.4-Typical usage4.5-Effects on fresh concrete4.6-Effects on hardened concrete4.7-Preparationand batching4.8-Proportioning4.9-Qualitycontrol4.10-Precautions

* Chairman of the task groups that prepared this report. The following former members ofcommittee 212 contributed to the preparatioo of the document: Sanford L. Bauman, Jr.; RogerW. Black; Edward J. Hyland (chairman); Rolland L. Johns; and Herman G. Protze, III.

The 1991 revisions became effective July 1, 1991. A number of minor editorial revisions weremade to the report, including Section 5.8.7. The year designations of the recommended refer-ences of standards-producing organizations have been removed so that the current editionsbecome the referenced version.

Copyright 0 1991, American Concrete Institute.All rights reserved including the rights of reproduction and use in any form or by any means.

including the making of copies by any photo process, or by any electronic or mechanical deviceprinted or written or oral, or recording for sound or visual reproduction or for use in anyknowledge or retrieval system or device, unless permission in writing is obtained from the copy-right proprietors.

1

212.3R-2 MANUAL OF CONCRETE PRACTICE

Chapter 5-Admixtures for flowing concrete,p. 212.3R-19

5.1-General5.2-Materials5.3-Evaluation and selection5.4-Application5.5-Performance criteria5.6-Proportioning of concrete5.7-Effect on fresh concrete5.8-Effect on hardened concrete5.9-Quality assurance5.10-Control of concrete

Chapter 6-Miscellaneous admixtures, p. 212.3R-226.1-Gas-forming admixtures6.2-Grouting admixtures6.3-Expansion-producing admixtures6.4-Bonding admixtures6.5-Pumping aids6.6-Coloring admixtures6.7-Flocculating admixtures6.8-Fungicidal, germicidal, and insecticidal admixtures6.9-Dampproofing admixtures6.10-Permeability-reducing admixtures6.11-Chemical admixtures to reduce alkali-aggregate expansion6.12-Corrosion-inhibiting admixtures

Chapter 7-References, p. 212.3R-287.1-Recommended references7.2-Cited references

CHAPTER 1-GENERAL INFORMATION1.1-Introduction

An admixture is defined in ACI 116R and in ASTMC 125 as: “a material other than water, aggregates, hy-draulic cement, and fiber reinforcement, used as an in-gredient of concrete or mortar, and added to the batchimmediately before or during its mixing.” This reportdeals with commonly used admixtures other than poz-zolans. Admixtures whose use results in special types ofconcrete are assigned to other ACI committees, such as:expansive-cement concrete (ACI Committee 223), insu-lating and cellular concretes (ACI Committee 523), andpolymers in concrete (ACI Committee 548). Pozzolansused as admixtures are assigned to ACI Committee 226,which also deals with ground granulated iron blast-fur-nace slag (a latent hydraulic cement) added at themixer.

Admixtures are used to modify the properties ofconcrete or mortar to make them more suitable for thework at hand, or for economy, or for such other pur-poses as saving energy. In many instances, (e.g., veryhigh strength, resistance to freezing and thawing, re-tarding, and accelerating), an admixture may be theonly feasible means of achieving the desired result. Inother instances, certain desired objectives may be bestachieved by changes in composition or proportions ofthe concrete mixture if so doing results in greater econ-omy than by using an admixture.

1.2-Reasons for using admixturesSome of the more important purposes for which ad-

mixtures are used are:To modify properties of fresh concrete, mortar, and

grout so as to:

Increase workability without increasing water con-tent or decrease the water content at the same work-abilityRetard or accelerate time of initial settingReduce or prevent settlement or create slight expan-sionModify the rate and/or capacity for bleedingReduce segregationImprove pumpabilityReduce the rate of slump lossTo modify properties of hardened concrete, mortar,

and grout so as to:Retard or reduce heat evolution during early hard-eningAccelerate the rate of strength development at earlyagesIncrease strength (compressive, tensile, or flexural)Increase durability or resistance to severe conditionsof exposure, including application of deicing saltsDecrease permeability of concreteControl expansion caused by the reaction of alkalieswith certain aggregate constituentsIncrease bond of concrete-to-steel reinforcementIncrease bond between existing and new concreteImprove impact resistance and abrasion resistanceInhibit corrosion of embedded metalProduce colored concrete or mortar

1.3-Specifications for admixturesThe following specifications cover the types or classes

that make up the bulk of current products:Air-entraining admixtures . . . . . . . . . . . . ASTM:

AASHTO:Water-reducing and set-controlling

admixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ASTM:AASHTO:

Calcium chloride.. . . . . . . . . . . . . . . . . . . . . . . ASTM:AASHTO:

Admixtures for use in producingflowing concrete . . . . . . . . . . . . . . . . . . . . . ASTM:

C 260M 154

C 494M 194D 98M 144

C 1017

1.4-SamplingSamples for testing and inspection should be ob-

tained by procedures prescribed for the respective typesof materials in the applicable specifications. Such sam-ples should be obtained by random sampling fromplant production, from previously unopened packagesor containers, or from fresh bulk shipments.

1.5-TestingAdmixtures are tested for one or more of three rea-

sons: (a) to determine compliance with specifications;(b) to evaluate the effect of the admixture on the prop-erties of the concrete to be made with job materials un-der the anticipated ambient conditions and construc-tion procedures; and (c) to determine uniformity ofproduct.

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-3

The manufacturer of the admixture should be re-quired to certify that individual lots meet the require-ments of applicable standards or specifications.

It is important that quality control procedures beused by producers of admixtures to insure productcompliance with uniformity and other provisions ofASTM specifications and with the producer’s own fin-ished-product specifications. Since such test methodsmay be developed around a particular proprietaryproduct, they may not be applicable to general use oruse by consumers.

Although ASTM tests afford a valuable screeningprocedure for selection of admixtures, continuing useof admixtures in production of concrete should be pre-ceded by testing that allows observation and measure-ment of the performance of the chemical admixtureunder concrete plant operating conditions in combina-tion with concrete-making materials then in use. Uni-formity of results is as important as the average resultwith respect to each significant property of the admix-ture or the concrete.

1.6-Cost effectivenessEconomic evaluation of any given admixture should

be based on the results obtained with the particularconcrete in question under conditions simulating thoseexpected on the job. This is highly desirable since theresults obtained are influenced to an important degreeby the characteristics of the cement and aggregate andtheir relative proportions, as well as by temperature,humidity, and curing conditions.

In evaluating an admixture, its effect on the volumeof a given batch should be taken into account. If add-ing the admixture changes the yield, as often is the case,the change in the properties of the concrete will be duenot only to direct effects of the admixture, but also tochanges in the yield of the original ingredients. If theuse of the admixture increases the volume of the batch,the admixture must be regarded as effecting a displace-ment either of part of the original mixture or of one oranother of the basic ingredients-cement, aggregate, orwater. All such changes in the composition of a unitvolume of concrete must be taken into account whentesting the direct effect of the admixture and in esti-mating the benefits resulting from its use.

The increase in cost due to handling an additionalingredient should be taken into account, as well as theeconomic effect the use of the admixture may have onthe cost of transporting, placing, and finishing the con-crete. Any effect on rate of strength gain and speed ofconstruction should be considered. An admixture maypermit use of less expensive construction methods orstructural designs to more than offset any added costdue to its use. For example, novel and economical de-signs of structural units have resulted from the use ofwater-reducing and set-retarding admixtures (Schutz1959).

In addition, placing economies, ability to pump atgreater heads, and economies of concrete cost versuscompetitive building materials have been realized. Wa-

ter-reducing and set-retarding admixtures permit place-ment Of large volumes of concrete over extendedperiods, thereby minimizing the need for forming,placing, and joining separate units. Accelerating ad-mixtures reduce finishing and forming costs. Requiredphysical properties of lightweight concrete can beachieved at lower densities (unit weight) by using air-entraining and water-reducing admixtures.

1.7-Considerations in the use of admixturesAdmixtures should conform to ASTM or other ap-

plicable specifications. Careful attention should begiven to the instructions provided by the manufacturerof the admixture. The effects of an admixture shouldbe evaluated whenever possible by use with the partic-ular materials and conditions of use intended. Such anevaluation is particularly important when (1) the ad-mixture has not been used previously with the particu-lar combination of materials; (2) special types of ce-ment are specified; (3) more than one admixture is to beused; and (4) mixing and placing is done at tempera-tures well outside generally recommended concretingtemperature ranges.

Furthermore, it should be noted that: (1) a change intype or source of cement or amount of cement, or amodification of aggregate grading or mixture propor-tions, may be desirable; (2) many admixtures affectmore than one property of concrete, sometimes ad-versely affecting desirable properties; (3) the effects ofsome admixtures are significantly modified by suchfactors as water content and cement content of themixture, by aggregate type and grading, and by typeand length of mixing.

Admixtures that modify the properties of fresh con-crete may cause problems through early stiffening orundesirable retardation, i.e., prolonging the time ofsetting. The cause of abnormal setting behavior shouldbe determined through studies of how such admixturesaffect the cement to be used: Early stiffening often iscaused by changes in the rate of reaction between tri-calcium aluminate and sulfate. Retardation can becaused by an overdose of admixture or by a lowering ofambient temperature, both of which delay the hydra-tion of the calcium silicates.

Another important consideration in the use of ad-mixtures relates to those cases where there is a limit onthe amount of chloride ion that is permitted in concreteas manufactured. Such limits exist in the ACI 318Building Code, the recommendations of ACI Commit-tees 201, 222, 226, and others. Usually these limits areexpressed as maximum percent of chloride ion byweight (mass)* of cement. Sometimes, however, it ischloride ion per unit weight (mass) of concrete, andsometimes it is “water-soluble” chloride ion per unitweight (mass) of cement or concrete.

Regardless of how the limit is given, it is obvious thatto evaluate the likelihood that using a given admixture

*In this report, when reference is made to mass it is called “weight” becausethe committee believed this would be better understood; however, the correctterm “mass” is given in parentheses.

212.3R-4 MANUAL OF CONCRETE PRACTICE

will jeopardize conformance of concrete with a specifi-cation containing such a limit, one needs to know thechloride-ion content of the admixture that is being con-sidered for use expressed in terms relevant to those inwhich the specification limit is given. If in using theavailable information on the admixture and the pro-posed dosage rate it is calculated that the specificationrequirement will be exceeded, alternate admixtures orprocedures should be considered for achieving the re-sults that were sought through the use of the admixturethat cannot be used in the originally intended amount.

The user should be aware that in spite of such termsas “chloride-free,” no truly chloride-free admixtureexists since admixtures often are made with water thatcontains small but measurable amounts of chloride ion.

1.8-Decision to useAlthough specifications deal primarily with the in-

fluence of admixtures on standard properties of freshand hardened concrete, the concrete supplier, contrac-tor, and owner of the construction project are inter-ested in other features of concrete construction. Of pri-mary concern may be workability, pumping qualities,placing and finishing qualities, early strength develop-ment, reuse of forms or molds, appearance of formedsurfaces, etc. These additional features often are ofgreat importance when the selection and dosage rate ofan admixture are determined.

Specific guidance for use of accelerating admixtures,air-entraining admixtures, water-reducing and set-con-trolling admixtures, admixtures for flowing concrete,and admixtures for other purposes is given in the rele-vant chapters of this report. Those responsible for con-struction of concrete structures should bear in mindthat increasing material costs and continuing develop-ment of new and improved admixtures warrant reeval-uation concerning the benefits of admixture use.

1.9-Classification of admixturesIn this report, admixtures are classified generically or

with respect to the characteristic effects of their use.Information to characterize each class is presentedalong with brief statements of the general purposes andexpected effects of the use of materials of each group.The wide scope of the admixture field, the continualentrance of new or modified materials into this field,and the variations of effects with different concretingmaterials and conditions preclude a complete listing ofall admixtures and their effects on concrete.

Commercial admixtures may contain materials thatseparately would belong in two or more groups, orwould be covered by two or more ASTM standards orACI committees. For example, a water-reducing ad-mixture may be combined with an air-entraining ad-mixture, or a pozzolan may be combined with a water-reducing admixture. Those admixtures possessingproperties identifiable with more than one group or onecommittee are considered to be in the group or com-mittee that is concerned with their most important ef-fect.

1.10-Preparation and batching1.10.1 Introduction-The successful use of admix-

tures depends upon the use of appropriate methods ofpreparation and batching. Neglect in these areas mayaffect properties, performance, and uniformity of theconcrete significantly.

Certain admixtures such as pigments, expansiveagents, pumping aids, and the like are used in ex-tremely small dosages and most often are batched byhand from premeasured containers. Other hand-addedadmixtures may include accelerators, permeability re-ducers, and bonding aids, which often are packaged inamounts sufficient for proper dosage per unit volumeof concrete.

Most admixtures usually are furnished in ready-to-use liquid form. These admixtures are introduced intothe concrete mixture at the concrete plant or into atruck-mounted admixture tank for introduction into theconcrete mixture at the jobsitc. Although measurementand addition of the admixture to the concrete batch orinto the truck-mounted tank often is by means of a so-phisticated mechanical or electromechanical dispensingsystem, a calibrated holding tank should be part of thesystem so that the plant operator can verify that theproper amount of admixture has been batched into theconcrete mixer or into the truck-mounted tank.

1.10.2 Conversion of admixture solids to liq-uids-Most admixtures are furnished in liquid formand do not require dilution or continuous agitation tomaintain their solution stability.

The preparation of admixtures may involve makingdilutions of the various concentrations to facilitate ac-curate batching or dispensing. As a result, the recom-mendations of the manufacturer should be followed ifthere is any doubt about procedures being used.

Some chemical admixtures are supplied as water-sol-uble solids requiring job mixing at the point of use.Such job mixing may require that low-concentrationsolutions be made due to difficulty in mixing. In somecases, it is convenient to prepare standard solutions ofuniform strength for easier use. Since many low-con-centration solutions contain significant amounts offinely divided insoluble materials or active ingredients,which may or may not be readily soluble, it is impor-tant that precautions be taken to insure that these bekept in uniform suspension before actual batching.

1.10.3 Storage and protection-Because admixturesfurnished as dry powders sometimes are difficult todissolve, admixtures supplied as ready-to-use liquidsmay be of much higher concentrations than job-mixedsolutions. As a result, any finely divided insoluble mat-ter, if present, will tend to stay in suspension, and con-tinuous agitation usually is not required. Admixturemanufacturers ordinarily can furnish either completestorage and dispensing systems or at least informationregarding the degree of agitation or recirculation re-quired with their admixtures. Timing devices com-monly are used to control recirculation of the contentsof storage tanks to avoid settlement or, with someproducts, polymerization.

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-5

In climates subject to freezing, the storage tank andits contents must be either heated or placed in a heatedenvironment. The latter is preferred for the followingreasons:

1. If the storage tank contains pipe coils for heatingthe contents by means of hot water or steam, care mustbe taken to avoid overheating the admixture since hightemperatures can reduce the effectiveness of certain ad-mixture formulations.

2. Some heating probes can overheat the admixturelocally, pyrolize certain constituents, and produce ex-plosive gases.

3. ‘Electrical connections to heating probes, bands, ortapes can be disconnected, allowing the admixture tofreeze and damage equipment.

4. The cost of operating electric probes, bands, tapes,etc. is normally higher than the cost of maintainingabove-freezing temperatures in a heated storage room.

5. A heated admixture storage room protects notonly storage tanks, but pumps, meters, valves, and ad-mixture hoses from freezing and from other problemssuch as dust, rain, ice, and vandalism. Further, sincethe storage temperature is subject to less widespreadvariation throughout the year, admixture viscosity ismore constant and dispensers require less frequent cal-ibration.

6. If plastic storage tanks or hoses are used, caremust be taken to avoid heating these materials to thepoint of softening and rupture.

Storage tanks should be vented properly so that for-eign materials cannot enter the tank through the open-ing. Likewise, fill nozzles and any other tank openingsshould be capped when not in use to avoid contamina-tion.

1.10.4 Batching-Batching of liquid admixtures anddischarging into the batch, mixer, or truck-mountedtank generally is accomplished by a system of pumps,meters, timers, calibration tubes, valves, etc., generallycalled the admixture-dispensing system or dispenser.

Dispensing of admixtures into a concrete batch in-volves not only accurately measuring the quantity ofadmixture and controlling the rate of discharge but alsothe timing in the batching sequence. In some instances,changing the time at which the admixture is added dur-ing mixing can vary the degree of effectiveness of theadmixture. For example, Bruere (1963) and Dodson,Farkas, and Rosenberg (1964) reported that the retard-ing effect of water-reducing retarders depends on thetime at which the retarder is added to the mixture. Thewater requirement of the admixture also may be af-fected significantly.

For any given condition or project, a procedure forcontrolling the time and rate of the admixture additionto the concrete batch should be established and ad-hered to closely. To insure uniform distribution of theadmixture throughout the concrete mixture during thecharging cycle, the rate of admixture discharge shouldbe adjustable.

Foster (1966) noted that two or more admixturesoften are not compatible in the same solution. For ex-

ample, a vinsol resin-based air-entraining admixtureand a water-reducing admixture containing a lignosul-fonate should never come in contact prior to actualmixing into the concrete because of their instant floc-culation and loss of efficiency of both admixtures. It isimportant, therefore, to avoid intermixing of admix-tures prior to introduction into the concrete unless testsindicate there will be no adverse effects or the manu-facturer’s advice permits it. It generally is better to in-troduce the various admixtures into the batch at differ-ent times or locations during charging or mixing.

It is important that batching and dispensing equip-ment meet and maintain tolerance standards to mini-mize variations in concrete properties and, con-sequently, better performance of the concrete.Tolerances of admixture batching equipment should bechecked carefully. ASTM C 94 requires that volumetricmeasurement of admixtures shall be accurate, to +3percent of the total volume required or plus or minusthe volume of dose required for 94 lb (43 kg, or onebag) of cement, whichever is greater. ASTM C 94 re-quires that powdered admixtures be measured byweight (mass), but permits liquid admixtures to bemeasured by weight (mass) or volume. Accuracy ofweighed admixtures is required to be within + 3 per-cent of the required weight (mass).

1.10.5 Batching equipment1.10.5.1 General-In terms of batching systems,

admixtures may be grouped in two categories: (a) thosematerials introduced into the batch in liquid form,which may be batched by weight (mass) or volume; and(b) powdered admixtures that normally are batched byweight (mass). The latter case includes such specialtymaterials as pumping aids and others that are added inextremely small amounts and, thus, often are intro-duced by hand in premeasured packages. When high-volume usage of these admixtures is contemplated, themanufacturer of the admixture normally supplies asuitable bulk dispensing system.

1.10.5.2 Liquid batching systems-Liquid admix-ture dispensing systems are available for manual, sem-iautomatic, and automatic batching plants. Simplemanual dispensing systems designed for low-volumeconcrete plants depend solely on the care of the con-crete plant operator in batching the proper amount ofadmixture into a calibration tube and discharging itinto the batch.

More sophisticated systems intended for automatedhigh-volume plants provide automatic fill and dis-charge of the sight or calibration tube. It is necessary tointerlock the discharge valve so that it will not openduring the filling operation or when the fill valve is notclosed fully. Usually, the fill valve is interlocked withthe discharge valve so that it will not open unless thedischarge valve is closed fully. A low-level indicator inthe calibration tube often is used to prevent the dis-charge valve from being closed before all the admixtureis dispensed into the batch.

Several methods of batching liquid admixtures are incommon use. All require a visual volumetric container,

212.3R-6 MANUAL OF CONCRETE PRACTICE

called a calibration tube, to enable the plant operator toverify the accuracy of the admixture dosage. The sim-plest consist of a visual volumetric container, whileothers include positive volumetric displacement, and avery limited number use weigh-batching systems. Someof these can be used readily with manual, semiauto-matic, and automatic systems.

Positive volumetric displacement devices are wellsuited for use with automatic and semiautomaticbatchers because they may be operated easily by re-mote control with appropriate interlocking in thebatching sequence. They include flow meters and mea-suring containers equipped with floats or probes. Mostmeters are calibrated for liquids of a given viscosity.Errors caused by viscosity changes due to variations intemperature can be avoided by recalibration and ad-justment made by observation of the visual volumetriccontainer or calibration tube.

Flow meters and calibration tubes equipped withfloats or probes often are combined with pulse-emit-ting transmitters that give readouts on electromechani-cal or electronic counters. Often they are set by input-ting the dosage per unit of cementitious material. Theamount of cementitious material input to the panelcombined with the dosage rate sets the dispensing sys-tem to batch the proper amount of admixture.

Timer-controlled systems involve the timing of flowthrough an orifice. There are a number of variables as-sociated with these systems that can introduce consid-erable error. These variables include changes in powersupply, partial restrictions of the measuring orifice, andchanges in viscosity of the solution due to temperature.

Timer-controlled systems must be recalibrated fre-quently, and the plant operator must be alert to verifythe proper admixture dose by observation of the cali-bration tube. Although timer-controlled systems havebeen used successfully, because of these inherent dis-advantages, their use, in general, is not recommended,except perhaps for dispensing calcium chloride.

A number of different methods are used by admix-ture manufacturers to fill and discharge calibrationtubes. A major objective is to insure that the fill valvewill not open until the discharge valve is completelyclosed and to provide that, in the event of electrical ormechanical malfunction, the admixture cannot be ov-erbatched.

Power-operated valves are used frequently; a vac-uum release also may be provided to prevent venturiaction from the concrete plant’s water line, causing anoverbatching. Prior to installation of the dispenser, thesystem should be analyzed carefully to determine whatpossible batching errors could occur and, with the helpof the admixture supplier, they should be eliminated.

Discharge of the admixture from the calibration tubeto the concrete batch should be to the point where theadmixture achieves the greatest dispersion throughoutthe concrete. Thus, for example, the discharge end ofthe water line leading to the mixer is a preferred loca-tion, as is the fine-aggregate weigh hopper or the beltconveyor carrying fine aggregate.

Often, the calibration tube is emptied either by grav-ity or by air pressure and the admixture may have aconsiderable distance to flow through a discharge hoseor pipe before it reaches its ultimate destination.Therefore, the dispenser control panel should beequipped with a timer-relay device to insure that all ad-mixture has been discharged from the conveying hosesor pipes. If the admixture dispenser system is operatedmanually, the plant operator should be furnished avalve with a detente discharge side to prolong the dis-charge cycle until it is ascertained that all admixture isin the concrete batch.

When more than one admixture is intended for thesame concrete batch, the dispenser must be designed sothat: (1) an appropriate delay is built into the system toprevent the admixtures from becoming intermingled; or(2) each is batched separately so as to be properlymaintained apart before entering into the mixer. Like-wise, in a manual system, the operator must be in-structed in methods to prevent such comingling of ad-mixtures.

Weigh batching (batching by mass) of liquid admix-tures ordinarily is not used because the weigh batchingdevices are more expensive than volumetric dispensers.In some cases, it is necessary to dilute admixture solu-tions to obtain a sufficient quantity for accurate weigh-ing (determination of mass).

Because of the high rate of slump loss associated withcertain high-range water-reducing admixtures (super-plasticizers), jobsite introduction of such admixtureshas become common. Such addition may be fromtruck-mounted admixture tanks or jobsite tanks ordrums. When using drums, the dispensing system oftenis similar to that used in concrete plants, e.g., pumps,meters, pulse transmitters, and counters to dispense theproper admixture volume to the truck mixer at the job-site.

If truck-mounted tanks are used, the proper dosageof admixture for the concrete in the truck is measuredat the batch plant and discharged to the truck-mountedtank at a special filling station. At such a station, a se-ries of lights or other signals tells the driver when theadmixture batching is complete and when his tank con-tains the proper amount. At the jobsite, the driver setsthe mixer at mixing speed and discharges the entireamount of admixture from the truck-mounted tank intothe concrete.

Care should be taken that the mixer remain in themixing mode until the admixture has been thoroughlydistributed throughout the concrete. The condition ofthe mixer and its blades influences the distribution. Toinsure that all the admixture is introduced, air pressureshould be used to force the admixture into all parts ofthe mixer drum. To shorten the mixing time, the truckmixer should operate at maximum speed, preferablyover 19 rpm.

1.10.5.3 Maintenance-Batching systems requireroutine periodic maintenance to prevent inaccuraciesdeveloping from such causes as sticky valves, buildupof foreign matter in meters or in storage and mixing

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-7

tanks, or worn pumps. It is important to protect com-ponents from dust and temperature extremes, and theyshould be readily accessible for visual observation andmaintenance.

Although admixture batching systems usually are in-stalled and maintained by the admixture producer,plant operators should thoroughly understand the sys-tem and be able to adjust it and perform simple main-tenance. For example, plant operators should recali-brate the system on a regular basis, not to exceed 90days, noting any trends that indicate worn parts need-ing replacement.

Tanks, conveying lines, and ancillary equipmentshould be drained and flushed on a regular basis, andcalibration tubes should have a water fitting installed toallow the plant operator to water flush the tube so thatdivisions or markings may be clearly seen at all times.Because of the marked effect of admixtures on con-crete performance, care and attention to the timing andaccuracy of batching admixtures is necessary to avoidserious problems.

CHAPTER 2-AIR-ENTRAINING ADMIXTURES2.1-Introduction

ACI 116R defines an air-entraining agent as “anaddition for hydraulic cement or an admixture for con-crete or mortar which causes entrained air to be incor-porated in the concrete or mortar during mixing, usu-ally to increase its workability and frost resistance.”This chapter is concerned with those air-entrainingagents that are added to the concrete batch immedi-ately before or during its mixing, and are referred to asair-entraining admixtures.

Extensive laboratory testing and long-term field ex-perience have demonstrated conclusively that concretemust be properly air entrained if it is to resist the ac-tion of freezing and thawing (Cordon et al. 1946;Blanks and Cordon 1949). Air entrainment should al-ways be required when concrete must withstand manycycles of freezing and thawing, particularly where theuse of such chemical deicing agents as sodium or cal-cium chlorides is anticipated. Highway pavements, ga-rage floors, and sidewalks placed in cold climatesprobably will be exposed to such conditions.

The mechanism of air entrainment in concrete hasbeen discussed in the literature (Powers 1968) but is be-yond the scope of this report. The resistance of con-crete to freezing and thawing also is affected by plac-ing, finishing, and curing procedures; therefore,acceptable practice in these respects must be followed(ACI 304R-85, ACI 308-81).

2.2-Entrained-air-void systemImprovements in frost resistance are brought about

by the presence of minute air bubbles dispersed uni-formly through the cement-paste portion of the con-crete. Because of their size and great number, there areliterally billions of such bubbles in each cubic yard ofair-entrained concrete. The void size must be small toprovide adequate protection with a relatively low totalvolume of void space.

The cement paste in concrete normally is protectedagainst the effects of freezing and thawing if the spac-ing factor (Powers 1949) is 0.008 in. (0.20 mm) or lessas determined in accordance with ASTM C 457. Addi-tional requirements are that the surface of the air voidsbe greater than 600 in?/in? (23.6 mm2/mmJ) of air-voidvolume, and that the number of air voids per 1 in. (25mm) of traverse be significantly greater than the nu-merical value of the percentage of air in the concrete.

The air content and the size distribution of air voidsproduced in air-entrained concrete are influenced bymany factors (Mielenz et al. 1958), the more importantof which are the (1) nature and quantity of the air-en-training admixture; (2) nature and quantity of the con-stituents of the concrete admixture; (3) type and dura-tion of mixing employed; (4) slump; and (5) kind anddegree of consolidation applied in placing the concrete.The factors are discussed in more detail in Section2.9.1. Vibration applied to air-entrained concrete re-moves air as long as the vibration is continued (Mielenzet al. 1958); however, laboratory tests have shown thatthe resistance of concrete to freezing and thawing is notreduced by moderate amounts of vibration.

Most investigators (Tynes 1977; Mather 1979; Schutz1978; Whiting 1979; Litvan 1983) have found in labo-ratory tests that the addition of high-range water re-ducers to air-entrained concrete may increase the spac-ing factor and decrease the specific surface area of theair-void systems. However, early reports of a reductionin frost resistance of such concretes (Tynes 1977;Mather 1979) have not been substantiated by later re-search. Nevertheless, it would be prudent to evaluatethe effect a specific high-range water reducer on thefrost resistance of a concrete mixture if this is a signif-icant factor and if the manufacturer cannot supply suchan evaluation.

For a discussion of the mechanism of protection byair entrainment, other sources should be consulted(Cordon 1966; Litvan 1972; MacInnis and Beaudoin1974; Powers 1975).

2.3-Effect on concrete properties2.3.1 Fresh concrete -Air entrainment alters the

properties of fresh concrete. These changes should beconsidered in proportioning a mixture (ACI 211.1 and211.2; Powers 1964). At equal slump, air-entrainedconcrete is considerably more workable and cohesivethan similar non-air-entrained concrete except at highercement contents. Segregation and bleeding are reduced.The reduction in bleeding, in turn, helps to prevent theformation of pockets of water beneath coarse-aggre-gate particles and embedded items such as reinforcingsteel, and also to prevent the accumulation of laitanceor weak material at the surface of a lift. At high ce-ment contents, air-entrained concrete becomes stickyand difficult to finish.

2.3.2 Hardened concrete-Air-entrainment usuallyreduces strength, particularly in concretes with moder-ate to high cement contents, in spite of the decreasedwater requirements. The reduction is generally propor-

212.3R-8 MANUAL OF CONCRETE PRACTICE

tional to the amount of air entrained, but the rate ofreduction increases with higher amounts. Therefore,while a proper air-void system must be provided, ex-cessive amounts of air must be avoided. A detailed dis-cussion of air requirements is included in ACI 211.1.

When the cement content and slump are maintainedconstant, the reduction in strength is partially or en-tirely offset by the resulting reduction in water-cementratio (w/c) and fine-aggregate content. This is particu-larly true of lean mass concretes or those containing alarge maximum-size aggregate. Such concretes may not’have their strength reduced; strengths even may be in-creased by the use of air entrainment.

2.4-Materials for air entrainmentMany materials are capable of functioning as air-en-

training admixtures. Some materials, such as hydrogenperoxide and powdered aluminum metal, can be used toentrain gas bubbles in cementitious mixtures but are notconsidered to be acceptable air-entraining admixtures,since they do not necessarily produce an air-void sys-tem that will provide adequate resistance to freezingand thawing.

2.4.1 Liquid or water-soluble powdered air-entrain-ing agents-These agents are composed of salts ofwood resins, synthetic detergents, salts of sulfonatedlignin, salts of petroleum acids, salts of proteinaceousmaterials, fatty and resinous acids and their salts, andorganic salts of sulfonated hydrocarbons. Not everymaterial that fits the preceding description will producea desirable air-void system.

Any material proposed for use as an air-entrainingadmixture should be tested for conformance withASTM C 260. This specification is written to assurethat the admixture functions as an air-entraining ad-mixture, that it causes a substantial improvement in theresistance of concrete to freezing and thawing, and thatnone of the essential properties of the concrete are se-riously impaired. Air-entrained concrete also can bemade by using an air-entraining portland cement meet-ing ASTM C 150, Type IA, IIA, or IIIA.

2.4.2 Particulate air-entraining admixtures-Solidparticles having a large internal porosity and suitablepore size have been added to concrete and seem to actin a manner similar to that of air voids. These particu-late materials may be composed of hollow plasticspheres or certain crushed bricks, expanded clay orshale, or spheres of certain diatomaceous earths. Thesematerials currently are not being used extensively.

Research has indicated that when using inorganicparticulate materials, the optimum particle size shouldrange between 290 and 850 pm, total porosity of theparticles should be at least 30 percent by volume, and apore-size distribution should be in the range of 0.05 to3 pm (Gibbons 1978; Sommer 1978). Inclusion of suchparticulates in the proper proportion has producedconcrete with excellent resistance to freezing and thaw-ing in laboratory tests using ASTM C 666 (Litvan andSereda 1978; Litvan 1985).

Particulate air-entraining admixtures have the ad-vantage of complete stability of the air-void system.Once added to the fresh concrete, changes in mixingprocedure or time; changes in temperature, workabil-ity, or finishing procedures; or the addition of otheradmixtures such as fly ash, or other cements such asground slag, will not change the air content, as may bethe case with conventional air-entraining admixtures.

2.5-ApplicationsThe use of entrained air in concrete is recommended

for several reasons. Because of its greatly improved re-sistance to frost action, air-entrained concrete must beused wherever water-saturated concrete is exposed tofreezing and thawing, especially when salts are used fordeicing. Its use also is desirable wherever there is a needfor watertightness.

Since air-entrainment improves the workability ofconcrete, it is particularly effective in lean mixtures thatotherwise may be harsh and difficult to work. It iscommon practice to provide air-entrainment in variouskinds of lightweight aggregate concrete, including notonly insulating and fill concrete (ACI 523.1R-67) butalso in structural lightweight concrete. However, ad-mixtures for cellular concrete are not covered in thisreport since ACI Committee 523 covers that subject.

There is no general agreement on benefits resultingfrom the use of air-entraining admixture in the manu-facture of concrete block (Farmer 1945; Kennedy andBrickett 1986; Keunning and Carlson 1956). However,satisfactory results using air-entraining admixtures havebeen reported in the manufacture of cast stone andconcrete pipe.

2.6-Evaluation, selection, and control ofpurchase

To achieve the desired improvement in frost resis-tance, intentionally entrained air must have certaincharacteristics. Not only is the total volume of air sig-nificant but, more importantly, the size and distribu-tion of the air voids must be such as to provide effi-cient protection to the cement paste.

To assure that an air-entraining admixture producesa desirable air-void system, it should meet the require-ments of ASTM C 260. This specification sets limits onthe effects that any given air-entraining admixture un-der test may exert on bleeding, time of setting, com-pressive and flexural strength, resistance to freezing andthawing, and length change on drying of a hardenedconcrete mixture in comparison with a similar concretemixture containing a standard-reference air-entrainingadmixture such as neutralized vinsol resin. The methodby which these effects may be determined is given inASTM C 233.

Extensive testing and experience have shown thatconcrete having total air contents in the range of thoserecommended in ACI 211.1 generally will have theproper size and distribution of air voids when the air-entraining admixture used meets the requirements ofASTM C 260. Use of ASTM C 457 to determine the

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-9

actual characteristics of the air-void system in hard-ened concrete in investigations of concrete proportion-ing provides greater assurance that concrete of sat-isfactory resistance to freezing and thawing will beobtained.

Most commercial air-entraining admixtures are inliquid form, although a few are powders, flakes, orsemisolids. The proprietary name and the net quantityin pounds (kilograms) or gallons (liters) should beindicated plainly on the containers in which the admix-ture is delivered. The admixture should meet require-ments on allowable variability within each lot andbetween shipments (see ASTM C 260). Acceptancetesting should be as stated in ASTM C 233.

2.7-Batching, use, and storageTO achieve the greatest uniformity in a concrete mix-

ture and in successive batches, it is recommended thatwater-soluble air-entraining admixtures be added to themixture in the form of solutions rather than solids.Generally, only small quantities of air-entraining ad-mixtures (about 0.05 percent of active ingredients byweight [mass] of cement) are required to entrain thedesired amount of air. If the admixture is in the formof powder, flakes, or semisolids, a solution must beprepared prior to use, following the recommendationsof the manufacturer.

If the manufacturer’s recommended amounts of air-entraining admixture do not result in the desired aircontent, it is necessary to adjust the amount of admix-ture added. For any given set of conditions and mate-rials, the amount of air entrained is roughly propor-tional to the quantity of agent used. However, in somecases, a ceiling may be reached. The ceiling may occurin low-slump, high cement-content mixtures made inhot weather with finely ground cements and containingfine aggregate with large amounts of material passingthe 75 pm (No. 200) sieve. A change in the fundamen-tal type of material used to make the air-entraining ad-mixture or a change in the cement or fine aggregate oran increase in slump may be necessary to obtain the re-quired air content.

Attention should be given to proper storage of air-entraining admixtures. The manufacturer’s storage rec-ommendations should be obtained and followed. Air-entraining admixtures usually are not damaged byfreezing, but the manufacturer’s instructions should befollowed regarding the effects of freezing on the prod-uct. An admixture that is stored at the point of manu-facture for more than six months after completion oftests prior to shipment, or an admixture in local stor-age in the hands of a vendor or contractor for morethan six months, should be retested before use and re-jected if it fails to conform to any of the requirementsof ASTM C 260.

2.8-Proportioning of concreteThe proportioning of air-entrained concrete is simi-

lar to that of non-air-entrained concrete. Methods ofproportioning air-entrained concrete should follow the

procedures of ACI Committee 211. These proceduresincorporate the reduction in water and fine aggregatepermitted by the improved workability of air-entrainedconcrete.

2.9-Factors influencing amount of entrained air2.9.1 Effects of materials and proportions-There

are numerous factors that can influence the amount ofair entrained in concrete. The amount of air-entrainingadmixture required to obtain a given air content willvary widely depending on the particle shape and grad-ing of the aggregate used. Organic impurities in the ag-gregate usually decrease the air-entraining admixturerequirements, while an increase in the hardness of wa-ter generally will increase the air-entraining admixturerequirements.

As the cement content or the fineness of a cement in-creases, the air-entraining potential of a given amountof an admixture will tend to diminish. Thus, largeramounts of air-entraining admixture generally are re-quired in concrete containing high early strength (TypeIII, ASTM C 150) or Portland-pozzolan cement (TypeIP, ASTM C 595). High-alkali cements generally re-quire a smaller amount of air-entraining admixture toobtain a given air content than do low-alkali cements.

Increasing the amount of finely divided materials inconcrete by the use of fly ash or other pozzolans, car-bon black or other finely divided pigments, or benton-ite usually decreases the amount of air entrained by anadmixture. As concrete temperature increases, higherdosages of air-entraining admixtures will be required tomaintain proper air content. A given amount of an air-entraining admixture generally produces slightly moreair where calcium chloride is used as an accelerator.

Similarly, the amount of air-entraining admixture re-quired to produce a given air content may be reducedone-third or more when used with certain water-reduc-ing admixtures. Various types of admixtures can influ-ence the air content and quality of the air-void system;therefore, special care should be taken when such ad-mixtures are used in conjunction with air-entrainingadmixtures to assure that there is compatibility.

Increasing the air content of concrete generallyincreases the slump. However, relatively high-slumpmixtures may have a larger spacing factor and aretherefore less desirable than low-slump mixtures. Anincrease in w/c is likely to result in an increase in aircontent and in larger air voids. As the temperature ofthe concrete increases, less air is entrained.

2.9.2 Effect of mixing, transporting, and consolidat-ing-The amount of air entrained varies with the typeand condition of the mixer, the amount of concretebeing mixed, and the mixing speed and time. The effi-ciency of a given mixer will decrease appreciably as theblades become worn or when mortar is allowed to ac-cumulate in the drum and on blades.

There also may be changes in air content if there is asignificant variation in batch size for a given mixer, es-pecially if the batch size is markedly different from therated capacity of the mixer. Adams and Kennedy (1950)

212.3R-10 MANUAL OF CONCRETE PRACTICE

found in the laboratory that, for various mixers andmixtures, air content increased from a level of about 4percent to as much as 8 percent, as the batch size wasincreased from slightly under 40 percent to slightly over100 percent of rated mixer capacity.

The amount of entrained air increases with mixingtime up to a point beyond which it slowly decreases.However, the air-void system, as characterized by spe-cific surface and spacing factors, generally is notharmed by prolonged agitation. If more water is addedto develop the desired slump, the air content should bechecked since some adjustment may be required; addi-tion of water without thorough or complete mixing mayresult in nonuniform distribution of air and waterwithin the batch. See ACI 304R for further details.

The methods used to transport concrete after mixingcan reduce the air content. Pumping the concrete gen-erally will reduce the air content.

The type and degree of consolidation used in placingconcrete can reduce the air content. Fortunately, air-void volume lost by these manipulations primarily con-sists of the larger bubbles of entrapped air that con-tribute little to the beneficial effects of entrained air.

2.10-Control of air content of concreteTo achieve the benefits of entrained air in a consis-

tent manner requires close control of the air content.For control purposes, samples for determination of aircontent should be obtained at the point of placement.Tests for air content of freshly mixed concrete shouldbe made at regular intervals for control purposes. Testsalso should be made when there is reason to suspect achange in air content.

The air content of importance is that present in con-crete after consolidation. Losses of air that occur dueto handling, transportation, and consolidation will notbe reflected by tests for air content of concrete taken atthe mixer (see ACI 309). This is why air content in thesample should be checked at the point of discharge intothe forms.

There are three standard ASTM methods for mea-suring the air content of fresh concrete: (1) the gravi-metric method, ASTM C 138; (2) the volumetricmethod, ASTM C 173; and (3) the pressure method,ASTM C 231, which, however, may not be applicableto lightweight concretes. An adaptation of the volu-metric method using the so-called Chace Air Indicator(Grieb 1958), in which a small sample of mortar fromthe concrete is used, has not been standardized andshould not be used to determine compliance with spec-ification limits.

These methods measure only air volume and not theair-void characteristics. The spacing factor and othersignificant parameters of the air-void system in hard-ened concrete can be determined only by microscopicalmethods such as those described in ASTM C 457. Theuse of these methods in coordination with investiga-tions of proportioning of concrete for new projectsprovides greater assurance that concrete of satisfactoryresistance to freezing and thawing will be obtained. It

has been shown, however, that the air content of aconcrete mixture generally is indicative of the adequacyof the air-void system when the air-entraining admix-ture used meets the requirements of ASTM C 260.

The properties of the concrete-making materials, theproportioning of the concrete mixture, and all aspectsof mixing, handling, and placing should be maintainedas constant as possible so that the air content will beuniform and within the range specified for the work.This is important because too much air may reducestrength without a commensurate improvement in du-rability, whereas too little air will fail to provide de-sired workability and durability.

Proper inspection should insure that air-entrainingadmixtures conform to the appropriate specifications,that they are stored without contamination or deterio-ration, and that they are accurately batched and intro-duced into the concrete mixture as specified. The aircontent of the concrete should be checked and con-trolled during the course of the work in accordancewith the recommendations of ACI Committee 311 asreported in the ACI Manual of Concrete Inspection(ACI SP-2). Practices causing excessive air loss shouldbe corrected or additional compensating air should beentrained initially.

CHAPTER 3-ACCELERATING ADMIXTURES3.1-Introduction

An accelerating admixture is a material added toconcrete for the purpose of reducing the time of settingand accelerating early strength development.

Accelerators should not be used as antifreeze agentsfor concrete; in the quantities normally used, accelera-tors lower the freezing point of concrete only a negligi-ble amount, less than 2 C (3.6 F). No commonly usedaccelerators will substantially lower the freezing pointof water in concrete without being harmful to the con-crete in other respects.

The best-known accelerator is calcium chloride, butit is not recommended for use in prestressed concrete,in concrete containing embedded dissimilar metals, orin reinforced concrete in a moist environment becauseof its tendency to promote corrosion of steel. Proprie-tary nonchloride noncorrosive accelerating admixtures,certain nitrates, formates, and nitrites afford users al-ternatives, although they may be less effective and aremore expensive than calcium chloride. Other chemicalsthat accelerate the rate of hardening of concrete in-clude triethanolamine and a variety of soluble salts suchas other chlorides, bromides, fluorides, carbonates, sil-icates, and thiocyanates.

3.2-Types of accelerating admixturesFor convenience, admixtures that accelerate the

hardening of concrete mixtures can be divided into fourgroups: (1) soluble inorganic salts, (2) soluble organiccompounds, (3) quick-setting admixtures, and (4) mis-cellaneous solid admixtures.

Accelerators purchased for use in concrete shouldmeet the requirements for Type C or E in ASTMC 494. Calcium chloride also should meet the require-

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-11

ments of ASTM D 98. Forms of calcium chloride areshown in Table 3.2.

3.2.1 Soluble inorganic salts-Studies (Edwards andAngstadt 1966; Rosskopf, Linton, and Peppler 1975)have shown that a variety of soluble inorganic salts,such as chlorides, bromides, fluorides, carbonates, thi-ocyanates, nitrites, nitrates, thiosulfates, silicates, alu-minates, and alkali hydroxides, will accelerate the set-ting of portland cement.

Research by numerous investigators over recent yearshas shown that inorganic accelerators act primarily byaccelerating the hydration of tricalcium silicate; com-prehensive calorimetric data illustrating this point havebeen reported.

Calcium chloride is the most widely used acceleratorsince it is the most cost effective.

It has been postulated (Tenoutasse 1969; Ramachan-dran 1972) that in portland cement concrete mixturescontaining calcium chloride (C&l,), gypsum combineswith the calcium aluminate to form ettringite (calciumtrisulfoaluminate [3CaO - A&O, l 3CaS0, l 32H,O]) andthe calcium chloride combines with the calciumaluminate to form calcium chloroaluminate(3CaO.CaCl,* lOH,O).

Table 3.2 - Calcium chloride - Amounts ofchloride ion introduced per 100 lb cement

Amount ofchloride

Liquid form, ionPercent calcium

chloride by Solid form, lb29 percent added tosolution* concrete,

*Commercial flake products generally have an assay of 77 to 80 percent cal-cium chloride, which is close to the dihydrate.

‘Commercial anhydrous calcium chloride generally has an assay of 94 to 97percent calcium chloride. The remaining solids usually are chlorides of magne-sium, sodium, or potassium, or combinations thereof. Thus, with regard to thechloride content, assuming that the material is 100 percent, calcium chlorideintroduces very little error.

‘A 29 percent solution often is the concentration of commercially used liquidforms of calcium chloride, and is made by dissolving 1 lb dihydrate to make 11qt of solution.

3.2.2 Soluble organic compounds-The most com-mon accelerators in this class are triethanolamine andcalcium formate, which are used commonly to offsetthe retarding effects of water-reducing admixtures or toprovide noncorrosive accelerators. Accelerating prop-erties have been reported for calcium acetate (Washaand Withey 1953), calcium propionate (Arber and Vi-vian 1961), and calcium butyrate (RILEM 1968), butsalts of the higher carboxylic acid homologs are retard-ers (RILEM 1968).

A number of organic compounds are found (Bashand Rakimbaev 1969) to accelerate the setting of port-land cement when low water-cement ratios are used.Organic compounds reported as accelerators includeurea (RILEM 1968), oxalic acid (Bash and Rakimbaev1969; Djabarov 1970), lactic acid (Bash and Rakim-baev 1969; Lieber and Richartz 1972), various ringcompounds (Lieber and Richartz 1972; Wilson 1927),and condensation compounds of amines and formal-dehyde (Rosskopf, Linton, and Peppler 1975; Kossivas1971). However, severe retardation can be experiencedwhen the amounts of these compounds used in a mix-ture are excessive.

Recent reports (Ramachandran 1973, 1976) indicatethat triethanolamine accelerates the hydration of trical-cium aluminate but retards tricalcium silicate. Thus,triethanolamine can act as a retarder of cement hydra-tion as well as an accelerator. Other organic accelera-tors may behave in a similar fashion.

Studies have shown that production of ettringite isgreater in mixtures containing calcium formate(Bensted 1978). Also, other data (Gebler 1983) haveshown that the effectiveness of formates is dependenton the sulfate content of the cement and the tricalciumaluminate-to-sulfate ratio (CJA/SOj). Cements that areundersulfated provide the best potential for calcium

formate to accelerate the early-age strength of con-crete. If the value for C,A/SO, is greater than 4.0, cal-cium formate has a good potential for accelerating thestrength of concrete.

3.2.3 Miscellaneous solid admixtures-In certain in-stances, hydraulic cements have been used in place ofaccelerating admixtures. For example, calcium-alumi-nate cement can shorten the time of setting of portlandcement concrete (Robson 1952).

The “seeding” of portland cement concrete with 2percent by weight (mass) of the cement with finelyground hydrated cement has been reported (Baslazs,Kelmen, and Kilian 1959; Duriex and Lezy 1956) to beequivalent to the use of 2 percent calcium chloride. Theeffects of seeding, in addition to calcium chloride, aresaid to be supplementary.

Various silicate minerals have been found (Angstadt and Hurley 1967; Kroone 1968) to act as accelerators.

Finely divided silica gels and soluble quaternary am-monium silicates have been found (Nelson and Young1977) to accelerate strength development, presumablythrough the acceleration of tricalcium-silicate hydra-tion (Stein and Stevels 1974). Very finely divided mag-nesium carbonate has been proposed (Ulfstedt andWatesson 1961) for accelerating time of setting of hy-draulic binders. Finely ground calcium carbonate tendsto accelerate time of setting (RILEM 1968).

3.3-Use with special cementsIt has been reported (USBR 1975) that the effective-

ness of calcium chloride in producing acceleratedstrength of concrete containing pozzolans is propor-tional to the amount of cement in the mixture. Variouseffects may be produced when calcium chloride is usedas an admixture in concrete containing shrinkage-com-pensating cement (ACI 223). The limited and conflict-ing data available on the effect of acceleration on theexpansion of concrete containing shrinkage-compen-sating or self-stressing cements suggest that the con-crete proposed for use should be evaluated with the ac-celerating admixture to determine its effect.

212.3R-12 MANUAL OF CONCRETE PRACTICE

Calcium chloride should not be used with calcium-aluminate cement since it retards the hydration of thealuminates. Similarly, calcium chloride and potassiumcarbonate increase the time of setting and decrease theearly strength development of rapid-hardening cementsbased on calcium fluoroaluminate (C11A7

. CaF2). How-ever, strengths after one day are improved by theseadditions. The effects of calcium chloride on blendedcements are similar to those for portland cements, theeffects being greater for cements using ground granu-lated blast-furnace slag than for those using pozzolanicadditions (Collepardi, Marcialis, and Solinas 1973).

The usual tests should be made for the control ofconcrete, such as slump, unit weight, and air content.If the concrete stiffens rapidly and difficulty is encoun-tered in achieving proper consolidation or finishing ofthe concrete, the accelerator used should be investi-gated

3.4-Consideration of useAccelerating admixtures are useful for modifying the

properties of concrete, particularly in cold weather, to:(a) expedite the start of finishing operations and, wherenecessary, the application of insulation for protection;(b) reduce the time required for proper curing and pro-tection; (c) increase the rate of early strength develop-ment to permit earlier removal of forms and earlieropening of construction for service; (d) permit moreefficient plugging of leaks against hydrostatic pressure;and (e) accelerate time of setting of concrete placed byshotcreting.

The use of accelerators in cold-weather concrete usu-ally is not sufficient in itself to counteract effects of lowtemperature. Recommendations for cold-weather con-creting usually include such practices as heating the in-gredients, providing insulation, and applying externalheat (see ACI 306R). Accelerators should not be usedas antifreeze agents for concrete.

Accelerators should be used with care in hot weather.Some of the detrimental effects that may result are veryrapid evolution of heat due to hydration, rapid setting,and increased shrinkage cracking.

3.5-Effect on freshly mixed and hardenedconcrete

The effects of accelerators on some properties ofconcrete include the following:

3.5.1 Time of setting-Initial and final times of set-ting are reduced. The amount of reduction varies withthe amount of accelerator used, the temperature of theconcrete, the ambient temperature, and characteristicsof other materials used in the concrete. Excessiveamounts of some accelerators may cause very rapidsetting; also, excessive dosage rates of certain accelera-tors may cause retardation.

Times of setting as short as 15 to 30 sec can be at-tained. There also are ready-to-use mixtures of cement,sand, and accelerator that have an initial set of 1 to 4min and a final set of 3 to 10 min. Mortars thus pre-pared are employed to seal leaks in below-grade struc-

tures, for patching, and for emergency repair, The ul-timate strength of such mortars will be much lowerthan if no accelerator had been added.

The concentration of an admixture may determine itsbehavior. For example, at high rates of addition (6 per-cent by weight [mass] of cement), calcium nitrate be-gins to show retarding properties (Murakami and Tan-aka 1969). Ferric chloride is a retarder at additions of 2to 3 percent by weight (mass) but is an accelerator at 5percent (Rosskopf, Linton, and Peppler 1975). The useof calcium-aluminate cement as an admixture maycause flash set depending on dosage rate.

Temperature also may be an important parametersince calcium chloride is stated (RILEM 1968) to havea greater accelerating effect at 0 to 5 C (32 to 41 F) thanat 25 C (77 F).

3.5.2 Air entrainment-Less air-entraining admix-ture may be required to produce the required air con-tent when an accelerator is used. However, in somecases, large bubble sizes and higher spacing factors areobtained, possibly reducing the beneficial effects ofpurposely entrained air. Evaluation of concrete con-taining the specific admixture(s) may be performed toascertain air-void parameters or actual resistance tofreezing and thawing using tests such as ASTM C 457and C 666, respectively.

3.5.3 Heat of hydration-Earlier heat release is ob-tained, but there is no appreciable effect on the totalheat of hydration.

3.5.4 Strength -- When calcium chloride is used,compressive strength may be increased substantially atearly ages; later strength may be reduced slightly. Thepercentage increase in flexural strength usually is lessthan that of the compressive strength.

The effects of other accelerating admixtures onstrength development are not completely known, al-though a number of salts that accelerate setting maydecrease concrete strengths even as early as one day.Some carbonates, silicates, and aluminates are in thiscategory. Organic accelerators, such as triethanolamineand calcium formate, appear to be sensitive in their ac-celerating action to the particular concrete mixture towhich they are added.

The addition of 2 percent calcium chloride by weight(mass) of cement, the 77-percent dihydrate type, in-creases strength at one day in the range of 100 to 200percent depending on the cement used.

The compressive strength at one day of neat cementpaste, mortar, or concrete prepared with mixtures ofportland and calcium-aluminate cements generally willbe materially lower than those obtained with either ofthe two cements alone.

Seeding of portland cement with 2 percent by weight(mass) of cement with finely ground hydrated cementhas been reported to increase 90-day compressivestrengths by 20 to 25 percent (Baslazs, Kelmen, andKilian 1959, Duriex and Lezy 1956).

3.5.5 Durability3.5.5.1 Volume change-Accelerators have been

reported to increase the volume changes that occur un-

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-13

der both moist curing and drying conditions. Calciumchloride is reported to increase creep and dryingshrinkage of concrete (Shideler 1942). A discussion ofliterature relating to the presumed association of the useof calcium chloride with increased drying shrinkagewith an alternative hypothesis has been advanced(Mather 1964).

More recent work (Bruere, Newbegin, and Wilson1971) has indicated that such changes depend on thelength of curing prior to beginning measurements, thelength of the drying or loading periods, and the com-position of the cement used. Also, changes in the rateof deformation are greater than changes in the totalamount of deformation. It has been suggested (Berger,Kung, and Young 1967) that the influence of calciumchloride in drying shrinkage may be the result ofchanges in the size distribution of capillary pores due tothe effect of calcium chloride on hydration of the ce-ment.

Drying shrinkage and swelling in water are higher formixtures containing both portland and calcium-alumi-nate cements, and their durability may be affected ad-versely by use of an accelerating admixture (Feret andVenuat 1957).

3.5.5.2 Frost damage-The resistance to deterio-ration due to cycles of freezing and thawing and toscaling caused by the use of deicing salts may be in-creased at early ages by accelerators but may be de-creased at later ages (see comments in previous sectionon air entrainment).

3.5.5.3 Sulfate resistance-The resistance to sul-fate attack is decreased when portland cement concretemixtures contain calcium chloride (USBR 1975).

3.5.5.4 Alkali-silica reaction-The expansion pro-duced by alkali-silica reaction is greater when calciumchloride is used (USBR 1975). This can be controlled bythe use of nonreactive aggregates, low-alkali cement, orcertain pozzolans.

3.5.5.5 Corrosion of metals-One of the majordisadvantages of calcium chloride is its tendency tosupport corrosion of metals in contact with concretedue to the presence of chloride ions moisture, andoxygen. In accordance with ACI 222, the maximumacid-soluble chloride contents of 0.08 percent for pre-stressed concrete and 0.20 percent for reinforced con-crete, measured by ASTM C 114 and expressed byweight (mass) of the cement, are suggested to minimizethe risk of chloride-induced corrosion.

Values for water-soluble chloride ion are given asmaxima in ACI 318-83, Table 4.5.4: prestressed con-crete-0.06; reinforced concrete exposed to chloride inservice-0.15; reinforced concrete that will be dry orprotected from moisture in service-1.00; other rein-forced concrete-0.30.

The user should exercise good judgment in applyingthese limits, keeping in mind that other factors (mois-ture and oxygen) always are necessary for electrochem-ical corrosion.

The use of calcium chloride as an accelerator will ag-gravate the effects of poor-quality concrete construc-

tion, particularly when the concrete is exposed to chlo-rides during service. Adherence to the limits just men-tioned does not guarantee absence of corrosion if goodconstruction practices are not followed.

Thus, admixtures have been sought that emulate theaccelerating properties of calcium chloride withouthaving its corrosive potential. Formulations based oncalcium formate with a corrosion inhibitor have beenpatented (Dodson, Farkas, and Rosenberg 1965). Theuse of stannous chloride, ferric chloride, and sodiumthiosulfate (Arber and Vivian 1961), calcium thiosul-fate (Murakami and Tanaka 1969), ferric nitrite (RI- LEM 1968), and calcium nitrite (Bruere 1971) are re-ported to inhibit the corrosion of steel while still accel-erating setting and hardening.

However, all accelerators that do not contain chlo-ride are not necessarily noncorrosive. Manns and Ei-chler (1982) reported that thiocyanates may promotecorrosion. Until additional published data becomeavailable, the Committee recommends that users re-quest suppliers of admixtures containing thiocyanatesto provide test data regarding the corrosion of steel inconcrete made with these admixtures. The test dataprovided should include corrosion results associatedwith the dosage range.

3.5.5.6 Discoloration of flatwork-Discolorationof concrete flatwork has been associated with the use ofcalcium chloride (Greening and Landgren 1966). Twomajor types of mottling discoloration can result fromthe interaction between cement alkalies and calciumchloride. The first type has light spots on a dark back-ground and is characteristic of mixtures in which theratio of cement alkalies to calcium chloride is relativelylow. The second consists of dark spots on a light back-ground and is characteristic of mixtures in which theratio of cement alkalies to chlorides is relatively high.

Available evidence indicates that the magnitude andpermanence of discoloration increases as the calciumchloride concentration increases from 0 to 2 percent byweight (mass) of cement. This type of discoloration canbe aggravated by high rates of evaporation during cur-ing and improper placement of vapor barriers. Use ofcontinuous fog spray or curing compounds can help al-leviate this problem.

3.5.6 Quick-setting admixtures-Some of the admix-tures in this category are used to produce quick-settingmortars or concretes suitable for shotcreting opera-tions, sealing leaks, or other purposes. Quick-settingadmixtures are believed to act by promoting the flashsetting of tricalcium aluminate (Schutz 1977). Amongthose admixtures used (Mahar, Parker, and Wuellner1975) or purported to produce quick set are ferric salts,sodium fluoride, aluminum chloride, sodium alumi-nate, and potassium carbonate. However, many pro-prietary formulations are mixtures of accelerators.These proprietary compounds are available in liquid orpowder form to be mixed with cement.

3.5.7 Rapid accelerators for shotcrete-Rapid accel-erators for shotcrete are employed extensively in bothdry- and wet-process shotcrete (ACI 506-66).

212.3R-14 MANUAL OF CONCRETE PRACTICE

Rapid shotcrete accelerators traditionally are basedon soluble aluminates, carbonates, and silicates. Thesematerials are highly caustic and are hazardous to work-ers. Newer neutral-pH chloride-free proprietary com-pounds are penetrating the market slowly.

3.6-Wet- and dry-process shotcreteSince the wet-process shotcrete mixture is mixed with

water as in conventional concrete, the rapid-setting ac-celerator is added at the nozzle during shooting. Gen-erally, the shotcrete mixture quickly stiffens andreaches an initial set, with a final set occurring muchlater than would occur with the dry process. However,the early stiffening imparted by the accelerator aids invertical and overhead placement.

Accelerated shotcrete is used for providing early rocksupport in tunneling, applying thick sections in verticalor overhead positions, sealing flowing water, and ap-plying of shotcrete between tides. The rate of strengthgain can be greatly accelerated using rapid acceleratorsin dry-process shotcrete. Strength in excess of 3000 psi(21 MPa) in 8 hr would be typical for a noncaustic ac-celerator and 2000 psi (14 MPa) with a conventionalcaustic accelerator.

Using dry-process shotcrete and a compatible cementand accelerator, an initial set of less than 1 min and afinal set of less than 4 min can be attained.

3.7-Control of purchaseAccelerators should meet the requirements of ASTM

C 494 for Type C or E. Calcium chloride also shouldmeet the requirements of ASTM D 98, solid or liquid.

3.8-Batching and useThe amount of accelerator needed to obtain the de-

sired acceleration of the time of setting and strengthdevelopment depends on local conditions and specificmaterials used; for calcium chloride, generally 1 to 2percent of the dihydrate form (77 to 80 percent) basedon the weight (mass) of cement, is added.

Practice in the industry has been to equate 1 lb of thedihydrate form to represent one percent of cement byweight (mass) (Calcium Chloride Institute 1959). Var-ious researchers (Abrams 1924; Ramachandran 1976)have studied the effects of calcium chloride on concreteusing this dosage basis. For convenience and means ofreference to various research data, this practice contin-ues and prevails.

It is recognized, however, that this practice does notresult in 1 percent anhydrous calcium chloride goinginto the mixture (1 lb dihydrate x 77 percent minimumassay/100 lb cement = 0.8 percent CaCl,). Multiples ofthis one percent (1 lb) of dihydrate are then used de-pending on temperatures (see Table 3.1).

In many locations, anhydrous (94 to 97 percent) solidforms or solutions of calcium chloride are more eco-nomical. Table 3.1 lists the common dosage rates ofeach form. The total chloride contributed to the mix-ture is shown in Table 3.1, and this includes chlorides

contributed by normal impurities (NaCl, KCl, MgCl,)in technical-grade products.

Calcium chloride should be introduced into the con-crete mixture in solution form. The dihydrate and an-hydrous solid forms should be dissolved in water priorto use. Preparation of a standard solution from drycalcium chloride requires that the user be aware of thepercent calcium chloride printed on the container, Indissolving the dry product, it should be added slowly tothe water, rather than the water to the calcium chlorideas a coating may form that is difficult to dissolve. Theconcentration of the solution may be verified by check-ing the density, which should be approximately 1.28

g/ml (0.17 oz/gal.) at 73 F (23 C), for a 29 percent so-lution. The correct density should be obtained from thesupplier.

All forms of calcium chloride should conform toASTM D 98. Accelerating admixtures based on cal-cium chloride should meet the requirements of ASTMC 494. The amount of water in the solution should bededucted from the water required for the desired w/c.Batching systems are available and are recommended toassure accurate and uniform addition of calcium chlo-ride in liquid form.

3.9-Proportions of concreteThe mix proportions for concrete containing an ac-

celerator generally are the same as for those without theaccelerator. The maximum recommended chloride-iondosage should not exceed those mentioned in the sec-tion of this chapter dealing with corrosion of metals.

3.10-Control of concretePerformance tests should be made if adequate infor-

mation is not available to evaluate the effect of a par-ticular admixture on properties of job concrete usingjob materials with expected job temperatures and con-struction procedures. Since some accelerators containsubstantial amounts of chlorides, the user should de-termine whether or not the admixture under considera-tion contains a significant amount of chlorides and, ifso, the percent by weight (mass) of the cement that itsuse will introduce into the concrete. The in-service po-tential for corrosion should then be evaluated accord-ingly.

CHAPTER 4-WATER-REDUCING AND SET-CONTROLLING ADMIXTURES

4.1 -GeneralCertain organic compounds or mixtures of organic

and inorganic compounds are used as admixtures inboth air-entrained and non-air-entrained concrete toreduce the water requirement of the mixture for a givenslump or to modify the time of setting, or both. Re-duction in water demand may result in either a reduc-tion in w/c for a given slump and cement content or anincreased slump for the same w/c and cement content.

When the w/c is reduced, the effect on the hardenedconcrete is increased compressive strength and reduc-tion in permeability and, in combination with adequateair entrainment, improved resistance to freezing and

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-15

thawing. The gain in compressive strength is frequentlygreater than is indicated by the decrease in w/c alone.This may be due to improved efficiency of hydration ofthe cement. Such admixtures also may modify the timeof setting of concrete or grouts.

A common side effect of many water-reducing ad-mixtures is a tendency to retard the time of setting ofthe concrete. Water-reducing admixtures that do notretard frequently are obtained by combining water-re-ducing and retarding materials with accelerators toproduce admixtures that still retain the water-reducingproperty but are less retarding, nonretarding (some-times called normal setting), or even somewhat accel-erating. The degree of effect depends upon the relativeamounts of each ingredient used in the formulation.Such formulations may contain other materials to pro-duce or modify certain other effects such as the inclu-sion of an air-entraining admixture to produce air-en-trained concrete, or an air-detraining admixture to re-duce or eliminate air-entrainment produced by certainingredients in the formulation when air entrainment iseither not desired or the amount of air produced is ex-cessive.

High-range water-reducing admixtures, also referredto as superplasticizers, behave much like conventionalwater-reducing admixtures in that they reduce the in-terparticle forces that exist between cement grains in thefresh paste, thereby increasing the paste fluidity. How-ever, they differ from conventional admixtures in thatthey do not affect the surface tension of water signifi-cantly; therefore, they can be used at higher dosageswithout excessive air entrainment.

The specific effects of water-reducing and set-con-trolling admixtures vary with different cements, addi-tion sequence, changes in w/c, mixing temperature,ambient temperature, and other job conditions.

4.2-Classification and compositionWater-reducing and set-controlling admixtures

should meet the applicable requirements of ASTMC 494, .which classifies them into the following seventypes:

1. Water-reducing2. Retarding3. Accelerating4. Water-reducing and retarding5. Water-reducing and accelerating6. Water-reducing, high-range*7. Water-reducing, high-range, and retarding?This ASTM specification gives detailed requirements

with respect to water requirement, time of setting,strength (compressive and flexural), drying shrinkage,and resistance to freezing and thawing.

The materials that generally are available for use aswater-reducing and set-controlling admixtures fall intoone of eight general classes:

1. Lignosulfonic acids and their salts

*Also covered by ASTM C 1017 as Type I.t Also covered by ASTM C 1017 as Type I I.

2. Modifications and derivatives of lignosulfonicacids and their salts

3. Hydroxylated carboxylic acids and their salts4. Modifications and derivatives of hydroxylated

carboxylic acids and their salts5. Salts of the sulfonated melamine polycondensa-

tion products6. Salts of the high molecular weight condensation

product of naphthalene sulfonic acid7. Blends of naphthalene or melamine condensates

with other water-reducing or set-controlling materials,or both

8. Other materials, which include: (a) inorganic ma-terials, such as zinc salts, borates, phosphates, chlo-rides; (b) amines and their derivatives; (c) carbohy-drates, polysaccharides, and sugar acids; and (d) cer-tain polymeric compounds, such as cellulose-ethers,melamine derivatives, naphthalene derivatives, sili-cones, and sulfonated hydrocarbons.

These materials may be used singly or in combina-tion with other organic or inorganic, active, or essen-tially inert substances.

4.3-Application Water-reducing admixtures are used to produce con-

crete of higher strength, obtain specified strength atlower cement content, or increase the slump of a givenmixture without an increase in water content. They alsomay improve the properties of concrete containing ag-gregates that are harsh or poorly graded, or both, ormay be used in concrete that may be placed under dif-ficult conditions. They are useful when placing con-crete by means of a pump or tremie.

Set-retarding admixtures are used primarily to offsetthe accelerating effect of high ambient temperature (hotweather) and to keep concrete workable during the en-tire placing period, thereby eliminating form-deflectioncracks (Schutz 1959). This method is particularly valu-able to prevent cracking of concrete beams, bridgedecks, or composite construction caused by form de-flections. Set retarders also are used to keep concreteworkable long enough so that succeeding lifts can beplaced without development of cold joints or discon-tinuities in the structural unit. Their effects on rate ofslump loss vary with the particular combinations ofmaterials used.

High-range water-reducing admixtures can be used toreduce the water content of concrete. Concrete of avery low w/c can be made to have high strength whilemaintaining a higher slump [over 3 in. (75 mm)] thanotherwise obtainable using a w/c as low as 0.28 byweight (mass). Water reduction up to 30 percent hasbeen achieved. Moderate water reductions (10 to 15 per-cent) also have been obtained at somewhat higherslumps (6 to 7 in).

With no reduction in water content, achievement offlowing concrete with slumps in excess of 8 in. is typi-cal (see Chapter 5). High-range water reducers alsohave been employed to reduce cement content. Sincethe w/c controls the strength of concrete, the cement

2212.3R-16 MANUAL OF CONCRETE PRACTICE

content may be reduced with a proportional reductionof the water content for equivalent strength concretes.

4.4-Typical usageExpected performance of a given brand, class, or

type of admixture may be projected from one or moreof the following sources:

1. Results from jobs where the admixture has beenused under good field control, preferably using thesame material and under conditions similar to thoseanticipated.

2. Laboratory tests made to evaluate the admixture.3. Technical literature and information from the

manufacturer of the admixture.The addition rate (dosage) of the admixture should

be determined from information provided by one ormore of the sources just mentioned. Informationshould be available on past performance of the pro-posed admixture substantiating the desired perfor-mance.

Various results can be expected with a given admix-ture due to differences in dosage, cements, aggregates,other materials, and weather conditions. Water-reduc-ing and set-controlling admixtures usually are found tobe more effective with respect to water reduction andstrength increase when used with portland cements oflower tricalcium aluminate (CA) and alkali content.Differences in setting times also can be expected withdifferent types and sources of cement as well as con-crete mixtures and ambient temperatures.

These admixtures generally are used to take advan-tage of water reduction to increase the strength of con-crete. If it is desired to provide a given level of strength,the cement content generally can be reduced, resultingin cost savings. In mass concrete, low cement content isparticularly desirable since it lowers the temperaturerise of the concrete. Water-reducing and set-controllingadmixtures do not lower heat of hydration of concreteexcept as they permit a reduction in cement content.The early temperature characteristics may be modifiedsomewhat due to the modification of the setting prop-erties of the concrete.

In the production of high-strength concrete [above6000 psi (41 MPa)], it has been found beneficial to in-crease the dosage of the admixture. This usually pro-vides extra water reduction as well as, typically, a re-tarded time of setting and slower early strength gain.Concrete having slow early strength gain characteristicsgenerally exhibits higher later strengths.

Concretes containing high-range water reducers oftenhave shown rapid slump loss. To overcome this, a sec-ond dosage of high-range water reducer may be used torestore the slump without any apparent ill effects. Gen-erally, more than two dosages are less effective andconcrete may lose its workability faster than with a sin-gle dosage. It has been found that redosage may resultin an increase or decrease in air content on the order of1 to 2 percent for each redose. When redosages areused, the concrete may experience a greater potential

for bleeding, segregation, and possible set retardation.Therefore, trial mixtures should be conducted to deter-mine the effects of redosing.

ASTM C 494 includes specifications for admixturesof the water-reducing and set-controlling types, It pro-vides for evaluation of the admixture for specificationcompliance under controlled conditions such as tem-perature, fixed cement content, slump, and air contentusing aggregates graded within stipulated limits. Thisstandard requires certain minimum differences in waterrequirement and strength of the concrete, range in timeof setting, and requirements in other properties such asshrinkage and resistance to freezing and thawing.

Most water-reducing admixtures perform consider-ably better than the minimum requirements of ASTMC 494. Good quality water-reducing admixtures reducethe water requirement of the concrete as much as 8 to10 percent or more and substantially increase thestrength at the same cement content. High-range waterreducers may reduce the water requirement by morethan 30 percent in some instances.

4.5-Effects on fresh concrete4.5.1 Water reduction-Water-reducing admixtures,

ASTM C 494, Type A, reduce the water required forthe same slump concrete by at least 5 percent, and insome cases up to 12 percent. Concrete containing lig-nosulfonate or hydroxylated carboxylic acid salts gen-erally reduce the water content 5 to 10 percent for agiven slump and cement content. High-range water re-ducers must reduce the water requirement at least 12percent but may reduce the water by more than 30 per-cent at a given slump. They also can be used to signifi-cantly increase slump without increasing water content.They’ also may be employed to achieve a combinationof these two objectives- a slump increase with a water-content reduction.

As the cement content of a concrete mixture in-creases, the required dosage of a high-range water re-ducer, as a percentage by weight (mass) of cement, isreduced. The effects of these admixtures also are de-pendent on the calculated CA, C,S, and alkali con-tents of the cement. Concretes made with cementsmeeting requirements for Type II and Type V cementsrequire lower admixture dosages than concretes con-taining Type I or Type III cements. In some cases, ithas been found that higher SO, content may be desir-able when using high-range water reducers.

4.5.2 Time of setting-Lignosulfonates and hydrox-ylated carboxylic acids retard times of setting by 1 to 3hr when used at temperatures of 65 to 100 F (18 to 38C). Sugar acids, carbohydrates, zinc salts, borates, andphosphates in unmodified form retard the setting ofportland cement in varying degrees. Most other mate-rials, including high-range water reducers, do not pro-duce appreciable retardation.

Accelerators may be incorporated in the formulationto produce acceleration or decrease or eliminate retar-dation. Retarders generally are not recommended for

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-17

controlling false set; water-reducing retarders have beenknown to contribute to premature stiffening.

4.5.3 Air entrainment-Lignosulfonates are air-en-training agents to various degrees. The amount of airentrainment generally is in the range of 2 to 6 percent,although higher amounts have been reported. This air-entrainment may consist of large unstable bubbles thatcontribute little to resistance to freezing and thawing.The air-entraining properties may be controlled bymodifying formulations. Materials in Classes 3 through8 (see Section 4.2) generally do not entrain air, but ma-terials in all eight classes may affect the air-entrainingcapability of both air-entraining cements and air-en-training admixtures.

Materials in Classes 5, 6, and 7 (high-range water re-ducers) have an effect on air entrainment. The perfor-mance and effectiveness of an air-entraining admixtureis strongly dependent on the nature of the high-rangewater-reducing admixtures with which it may be used.Certain air-entraining admixtures have been reported tobe more effective with certain high-range water-reduc-ing admixtures in the production of adequate air en-trainment.

Since the key factor in producing frost-resistant con-crete is the air-void system-that is, the total volume ofair, spacing factor, voids per inch, and specific sur-face-these factors should be quantified to achieve thedesired durability and that reliance is not placed whollyon the air content of the freshly mixed concrete. There-fore, different combinations of air-entraining admix-tures and high-range water-reducing admixtures shouldbe evaluated to achieve concrete that is resistant tofreezing and thawing, with determination of air-voidcontent and parameters of the air-void system in hard-ened concrete specimens being a desirable addition tothe test program.

It is prudent to include testing for resistance to freez-ing and thawing in the evaluation, as in some instancesspacing factors may exceed generally accepted limits[i.e., 0.20 mm (0.008 in.)] yet concrete may still be frostresistant when subjected to freezing and thawing.

4.5.4 Workability-When otherwise comparableconcretes with and without a water-reducing admixturehaving the same slump and air content are compared,differences in workability are difficult to detect sincethere is no standard test for workability. However,concrete containing a water-reducing admixture gener-ally is less likely to segregate. When vibrated, someworkers detect better flowability for the concrete con-taining the admixture.

Concretes proportioned for high strength using high-range water reducers usually have a sufficiently highcement content to supply the fines required. Repropor-tioning such concrete can be accomplished by makingup the volume of water reduced by increasing the vol-ume of coarse and fine aggregate equally. If trial mix-tures are sticky, the volume of coarse aggregate shouldbe increased and that of the fine aggregate reduced.This usually results in a mixture that is easier to placeand finish.

4.5.5 Bleeding-Admixtures affect bleeding capacityin varying degrees. For example, Class 3 admixtures(see Section 4.2) tend to increase bleeding, while cer-tain Class 4 admixtures, which are derivatives of theClass 3 admixtures, do not. Class 1 and 2 admixturesreduce bleeding and segregation in freshly mixed con-crete, in part due to the air entrainment. Class 5through 7 admixtures, when used as high-range waterreducers, generally decrease bleeding, except when at avery high slump.

4.5.6 Heat of hydration and temperaturerise-Within normal w/c ranges, adiabatic temperaturerise and heat of hydration are not reduced at equal ce-ment contents with the use of set-controlling admix-tures. Acceleration or retardation may alter the rate ofheat generation characteristics, which may change theearly rate of temperature rise under job conditions. Ifthe use of the admixture permits a reduction in cementcontent, heat generated is proportionally reduced.

4.5.7 Raze of slump loss-Rate of slump loss maynot be reduced and often is increased. With the low w/cattainable with high-range water reducers, the concretemay show a greater-than-normal rate of slump loss.Because of this, high-range water reducers often areadded at the jobsite. Working time can be extendedwith the careful use of an ASTM C 494, Type B re-tarding admixture or a Type D water-reducing and re-tarding admixture. The working time depends on manyfactors, including the high-range water-reducer dosage,the use of other chemical admixtures, cement charac-teristics, concrete temperature, slump, and the age ofthe concrete when the high-range water reducer is in-traduced.

4.5.8 Finishing-The finishing characteristics of con-crete containing Class 3 and 4 admixtures generally areimproved.

At reduced water contents achieved with high-rangewater reducers, finishing may become more difficultdue to the decrease in bleeding, and surfaces may havea tendency to crust and promote plastic-shrinkagecracking. The surface may be kept from drying by fog-ging, use of evaporation retarder, or other procedures(see ACI 308). This should be done with caution so thatthe durability of the surface is not affected adversely.

4.6-Effects on hardened concrete4.6.1 Strength-Reduction in w/c causes an increase

in strength. There is a further increase in strength dueto the use of a water-reducing admixture, apart fromthat due to reduction of w/c, due to modification ofthe hydration reaction and microstructure. Unless usedat unusually high rates, retarding admixtures generallywill produce an increase in strength at 24 hr. The re-tarding types may decrease the very early strength whilethe normal setting and accelerating types increase thevery early strength.

Later strength may be increased 20 percent or moreat the same cement content. Cement contents, thus, canbe reduced without lowering 28-day strengths. Whenhigh-range water reducers are used to decrease the w / c

212.3R-18 MANUAL OF CONCRETE PRACTICE

28-day compressive strength may be increased by 25percent or more. Increases in flexural strength of con-crete containing a water-reducing admixture usually areattained, but they are not proportionally as great as in-creases in compressive strength.

4.6.2 Shrinkage and creep-Information on the ef-fects of these admixtures on shrinkage and creep isconflicting. Long-term shrinkage may be less, depend-ing on the degree to which the water content of theconcrete is reduced. Creep will be reduced proportionalto the increase in the strength of the concrete. The de-’gree to which the use of an admixture, in given dos-ages, affects shrinkage and creep may be different ifcements of different compositions are used in the con-crete.

4.6.3 Durability-The effect of these admixtures onresistance to freezing and thawing, including deicerscaling, is small since resistance to freezing and thaw-ing is almost wholly a function of the air-void systemin the hardened concrete. An improvement may resultfrom a decrease in w/c and increased strength.

Some high-range water reducers cause bubble-spac-ing factor& higher than typically deemed necessary toproduce concrete that will survive freezing and thawingif critically saturated. A spacing factor of 0.008 in.(0.20 mm) or less generally is needed to insure resis-tance to freezing and thawing. Concrete made withsome high-range water reducers having spacing factorsof 0.010 in. (0.25 mm) or higher was found upon test-ing to be highly resistant to freezing and thawing. Thismay be a function of increased strength and density andreduced permeability, which allowed the concrete to re-main less than critically saturated in the presence ofwater while being tested.

A small increase in resistance to freezing and thaw-ing and to aggressive waters and soils results from wa-ter reduction. This is due largely to decreased permea-bility and increased strength.

.

4.7-Preparation and batchingWater-reducing and set-controlling admixtures

should be batched and dispensed as liquids. When sup-plied as solids, they should be mixed to a suitable so-lution concentration following the manufacturer’s rec-ommended practices.

The density of admixtures mixed on the job, or thoseapplied as solutions, should be determined and com-pared with the manufacturer’s standards. Determina-tion of density can be made easily and quickly with ahydrometer or volumetric flask. The determinationsshould be made at a standard temperature and re-corded for future reference as part of the job qualitycontrol program. Storage tanks for solutions should beplainly identified and the solutions should be protectedfrom contamination, dilution, evaporation, and freez-ing.

Two or more admixtures of different types, such as awater-reducing and an air-entraining admixture, maynot be compatible when mixed together. Unless it isknown that admixtures can be satisfactorily mixed to-

gether, they should be added to the batch separately sothat they will be adequately diluted before coming incontact with each other. The manufacturer of the ad-mixtures should recommend proper procedures.

4.8-ProportioningA concrete mixture may need reproportioning when

an admixture is used if the water content, cement con-tent, or air content is changed. By definition, for ex-ample, the water requirement of a concrete mixture forgiven consistency is reduced 5 percent or more with theintroduction of a water-reducing admixture. Proce-dures for proportioning and adjusting concrete mix-tures are covered by ACI 211.1.

One fundamental rule to remember is that when aconcrete mixture that is considered satisfactory inworkability and finishing qualities is modified to incor-porate a chemical admixture, the ratio of volumetricproportions of mortar to coarse aggregate should re-main the same. Changes in water content, cement con-tent, and air content are compensated for by corre-sponding changes in the content of fine aggregate, allon a solid or absolute volume basis, so that the volumeof mortar remains the same.

Most chemical admixtures of the water-reducing typeare water solutions. The water they contain becomes apart of the mixing water in the concrete and usually isconsidered in the calculation of w/c ratio. The propor-tiona1 volume of the solids included in the admixture isso small in relation to the size of the batch that it canbe neglected.

4.9-Quality controlIt sometimes is necessary or desirable to determine

that an admixture is the same as that previously tested,or that successive lots or shipments are the same. Teststhat can be used to identify admixtures include solidscontent, density, infrared spectrophotometry for or-ganic materials, chloride content, pH, and others.Guidelines for determination of uniformity (variability)of chemical admixtures are given in ASTM C 494.

Job inspectors may be instructed to sample deliveriesof the admixture as part of the job quality control.Density can be determined on the job by a hydrometeror volumetric flask as mentioned previously.

Admixture users should become familiar with ad-mixtures’ appearances and odors. This knowledge hassometimes prevented errors and mixups.

4.10-PrecautionsIf adequate information is not available, tests should

be made to evaluate the effect of the admixture on theproperties of concrete made with job materials underthe anticipated ambient conditions and constructionprocedures. Tests of water-reducing admixtures and set-controlling admixtures should indicate their effect onthe following properties of concrete, insofar as they arepertinent to the job: (1) water requirement, (2) air con-tent, (3) slump, (4) bleeding and possible loss of airfrom the fresh concrete, (5) time of setting, (6) corn-

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-19

pressive and flexural strength, (7) resistance to freezingand thawing, (8) drying shrinkage, and (9) setting char-acteristics.

When admixtures are evaluated in laboratory trialbatches prior to job use, the series of mixtures shouldbe planned to provide necessary information. Theyneed not follow ASTM C 494 procedures, althoughthese may be a helpful guide. The trial mixtures shouldbe made with the same materials, particularly cement,that will be used on the job and as close to job condi-tions as possible. Temperature is particularly importantto time of setting and early strength development.

Trial mixtures can be made at midrange slump andair contents expected or specified for the job. The ce-ment content or w/c ratio should be that required forthe specified design strength and durability require-ments for the job. Trial mixtures also can be made witha range of cement contents or w/c or other propertiesto bracket the job requirements.

Air content and time of setting of job concrete candiffer considerably from laboratory concrete with thesame materials and mixture proportions. All partiesshould be alert to this possibility at the start of a joband be ready to make adjustments in the addition ratesof materials (particularly air-entraining admixtures) toachieve the specified properties of the concrete at theproject site.

Admixtures of all classes may be available in eitherpowder or liquid form. Since relatively small quantitiesare used, it is important that suitable and accuratelyadjusted dispensing equipment be employed. Refer toChapter 1 for information on dispensing admixtures.

CHAPTER 5-ADMIXTURES FOR FLOWINGCONCRETE

5.1 -GeneralASTM C 1017 defines flowing concrete as “concrete

that is characterized as having a slump greater than 7 V2in. (190 mm) while maintaining a cohesive nature. . .”Flowing concrete can be placed so as to be self-levelingyet remaining cohesive without excessive bleeding, seg-regation, or abnormal retardation.

Since production of flowing concrete by addition ofwater only would result in concrete of extremely lowquality, flowing concrete must be obtained through theuse of a plasticizing admixture, either normal (Type 1)or retarding (Type 2). These materials used as plasticiz-ing admixtures for production of flowing concrete gen-erally are identical to those used as high-range water-reducing admixtures (superplasticizers) and conform toASTM C 494, Types F and G (see Chapter 4).

As an example, concrete could be delivered to thejobsite at an initial slump of 2 to 3 in. (50 to 75 mm)and the plasticizing admixture then could be added toincrease the slump to 8 in. or more, or the plasticizingadmixture could be added at the plant to achieve thisslump level.

The plasticizing admixture, either a conventional wa-ter-reducing admixture or a high-range water-reducingadmixture, is adsorbed onto the hydrating cement par-

ticles and causes a repulsion among them. The concreteloses slump at a more rapid rate than the same concretewithout the plasticizing admixture.

The admixtures that are used to achieve flowing con-crete should meet the requirements of ASTM C 1017,Type 1 (plasticizing) or Type 2 (plasticizing and retard-ing). Commonly used materials are:

1. Sulfonated napthalene condensates, Types 1 or 22. Sulfonated melamine condensates, Types 1 or 23. Modified lignosulfonates4. A combination of these types plus a water-reduc-

ing admixture, Type A; or water-reducing retardingadmixture, Type D; or water-reducing accelerating ad-mixture, Type E

5. High dosages of a water-reducing admixture, TypeA, plus a water-reducing accelerating admixture, TypeE

This latter combination requires higher water con-tents than are required when using a high-range water-reducing admixture (superplasticizer).

5.3-Evaluation and selectionA decision to produce and use flowing concrete

should include selecting the type of admixture to use.Factors to be considered in the choice of admixture(s)include type of construction; restriction imposed on thechloride-ion content, time interval from introduction ofcement and water into the mixer; availability of accu-rate admixture dispensing equipment at the plant, job-site, or both; and ambient temperature.

If the decision is made to add the plasticizing admix-ture at the jobsite, an accurate means of introducingthe admixture into the concrete mixer must be assured.Truck mixers should be equipped with admixture tanksdesigned to introduce the admixture into the concretemixer so that it can be distributed evenly throughoutthe batch, and adequate mixing speed and revolutionsshould be maintained. The concrete plant must beequipped to accurately measure the admixture into thetruck-mounted tanks.

Admixtures must be handled and measured properly.An admixture batching system must include a means ofvisual verification of the dosage.

5.4-ApplicationFlowing concrete commonly is used in areas requir-

ing maximum volume placement (slabs, mats, pave-ments) in congested locations where the member is un-usually shaped or a great amount of reinforcement ispresent. Proper consolidation of high-strength concretefor columns is difficult. Flowing concrete can be usedin areas of limited access or where the maximum hori-zontal movement of the concrete is desirable.

Flowing concrete is useful for pumping because it re-duces pumping pressure and increases both the rate anddistance that the concrete can be pumped. Such con-crete is useful for projects requiring rapid form cyclingwith a maximum volume of concrete required per day,

2 MANUAL OF CONCRETE PRACTICE

coupled with a low w/c to achieve the early strengthsrequired for stripping or tensioning. A short time cycleoften can be used on such projects.

5.5-Performance criteriaExpected performance of a given brand, class, or

type of admixture may be estimated from one or moreof the following sources: (a) results from jobs where theadmixture has been used under good technical control,preferably using the same concreting materials and un-der conditions similar to those anticipated; (b) labora-tory tests made to evaluate the admixture; and (c) tech-nical literature and information from the manufacturerof the admixture.

The dosage required to increase the slump to flowingconsistency varies depending upon the cement, the ini-tial slump, w/c, temperature, time of addition, andconcrete mix proportions. The dosage required to in-crease slump from 1 to 8 in. may be 50 percent higherthan that required if the starting slump is 3 in. Theproposed flowing concrete mixture should be used ini-tially in noncritical work so that proportions and pro-cedures can be verified before the mixture is used in theareas requiring flowing concrete. The proportions ofthe various concrete ingredients can be adjusted and thedosage or the type of admixture varied to achieve anacceptable initial slump, rate of slump loss, and settingcharacteristics.

Results may vary with a given admixture due to dif-ferences in cement, aggregates, other material, andweather conditions from day to day. The use of admix-tures to increase slump from the 2 to 3 in. range mayalso allow a cement reduction with a resultant cost sav-ing. Since very little concrete is placed at that lowslump level, the additional water required to raise theslump of conventional concrete from 2 to 3 in. to 5 to6 in. would have to be matched with an increase in thecement content if the strength and, consequently, thew/c is kept constant.

Flowing concrete is desirable for use in mass place-ments. The cement content may be kept low, which willminimize heat development, and the lower water con-tent will reduce shrinkage. The plasticizing admixturedoes not lower the temperature rise in concrete exceptas a result of reducing cement content. The early tem-perature-rise characteristics also may be modified withthe use of the retarding version of the plasticizing ad-mixture (Type 2) or in combination with a conven-tional water-reducing retarding admixture (Type D).

Concrete intended to have a compressive strengthhigher than 6000 psi (41 MPa) may be produced asflowing concrete; since a low w/c is required, reducingthe mixing water is the best approach. Flowing con-crete, being easier to consolidate, also contributes toproper bond between reinforcing steel and concrete inareas where reinforcement is congested.

ASTM C 1017 is the specification for admixtures forflowing concrete. It provides for evaluation of the ad-mixture for specification compliance under controlledconditions of temperature, fixed cement content,

slump, and air content, using aggregates graded withinstipulated limits. This standard requires certain mini-mum differences in strength of concrete, range of timesof setting, and requirements regarding other aspects ofperformance such as shrinkage and resistance to freez-ing and thawing.

It is preferable to keep the cement content constant,allowing the significant slump increase to assist inplacement. When admixtures are evaluated in labora-tory trial mixtures prior to job use, the series of mix-tures should be planned to provide the necessary infor-mation. Assuming specification compliance has beenestablished, the tests need not follow ASTM C 1017procedures such as slump, air content, and cementcontent; however, consistency of procedures should bemaintained.

The trial mixtures should be made with the same ma-terials, particularly cement, that will be used on the job,and simulate the job conditions as closely as possible.Temperature is particularly important to times of set-ting and early strength development. Trial mixtures canbe made with a starting slump and air content in thespecified range. The dosage of the plasticizing admix-ture can be varied to achieve various slump increases.If allowed, the starting slumps also may be varied. Thespecified w/c should be maintained in each case, and arange of slump can be reviewed. In this manner, theoptimum mixture proportions can be selected and therequired results achieved.

Air content and time of setting of job concrete candiffer considerably from laboratory concrete with thesame materials and mixture proportions. Therefore,adjustment of the proposed mixture on the jobsite priorto its use in the required locations usually is beneficial.

5.6-Proportioning of concreteA concrete mixture usually needs reproportioning

when a plasticizing admixture is added to achieve flow-ing concrete. Procedures for proportioning and adjust-ing concrete mixtures are covered in 211.1. Thefine aggregate-coarse aggregate ratio may require ad-justment to assure that sufficient fines are present toallow a flowable consistency to be achieved without ex-cessive bleeding or segregation. It also may be neces-sary to increase the cement content or add other finematerials such as pozzolan or slag.

Since 0.5 gal. or larger volume of plasticizing admix-ture is customarily used per yd’ (m’) of concrete to pro-duce flowing concrete, the water in the admixture mustbe accounted for in calculating w/c and the effect onmixture volume.

5.7-Effect on fresh concrete5.7.1 Times of selling-ASTM C 1017 Type 1 ad-

mixtures do not have much effect on times of setting.Therefore, flowing concrete with a water content thatwould give a 2 to 4 in. slump if an admixture were notused will set as quickly as if the admixture had not beenused. Type 2 admixtures can reduce slump loss signifi-cantly and retard the initial times of setting of the con-

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-21

crete. At concrete temperatures below 60 F (15 C), thetime of setting of concrete containing the Type 1 ad-mixture may be increased.

5.7.2 Workability and finishing-When concretemixtures are properly proportioned, flowing concrete isextremely workable without bleeding and segregation.The fine-to-coarse aggregate ratio often has to be ad-justed by an increase in fine aggregate content to pre-vent segregation at high slump. Flowing concreteshould be vibrated to achieve proper consolidation.

The characteristics of flowing concrete at the time itis being machine floated or troweled will be similar tothose of conventional concrete with the same ingredi-ents. Properly proportioned flowing concrete shouldnot exhibit objectionable bleeding even at high slump.Proper timing is imperative in the finishing operation.

If a concrete is over-sanded or the air content is toohigh, or both, the surface of the concrete tends to drybefore it sets. This condition may cause the concrete tofeel rubbery or jelly-like and cause finishing problemsby its stickiness and rolling. The problem of excessiveair entrainment in concrete used in floor slabs is partic-ularly apparent when the initial machine-finishing op-erations begin.

5.7.3 Bleeding and segregation-Properly propor-tioned concrete mixtures should not bleed excessively orsegregate. The upper slump limit of cohesive yet flow-able concrete varies and it can be determined from test-ing the mixture prior to its use. Segregation and bleed-ing may be reduced by increasing the fine-to-coarse ag-gregate ratio, or by the addition of other fine materials.

5.7.4 Rate of slump loss- The rate of slump loss maybe altered by many factors such as concrete tempera-ture, type and amount of cement, water content, timeof admixture addition, and amount of admixture em-ployed. Therefore, an acceptable rate of slump loss canbe achieved by monitoring these conditions and bychanging the initial time of setting characteristics of theconcrete.

5.7.5 Additional dosages-Additional dosages ofplasticizing admixture should be used when delays oc-cur and the required slump has not been maintained.Two additional dosages have been used with success;more dosages generally are less effective. In general, thecompressive strength level is maintained or increasedand the air content decreased. Therefore, if air entrain-ment is of concern, it must be checked after the con-crete has been redosed and returned to its intendedslump.

5.8-Effect on hardened concrete5.8.1 Heat of hydration and temperature rise-Heat

of hydration is not reduced if the cement content is notreduced. If the use of flowing concrete involves the useof a lower cement content, the heat evolved will be re-duced. If the rate of hydration is not changed, the tem-perature rise will not be changed if the cement contentis not reduced.

5.8.2 Strength-Since concrete that is intended to beflowing often is batched with a water content thatwould result in a slump of 2 to 4 in. (51 to 102 mm),

the w/c is lower than that of conventional concrete ofSimilar cement content at a 5-in. (127-mm) slump, andstrength improvement therefore is realized. Flowingconcrete with no water reduction, as compared to con-ventional concrete, often shows strength increases,When concrete strengths above 6000 psi (41 MPa) at 28 days are desired, a high-range water-reducing admix-ture often is added to achieve a low w/c. It then maybe added ‘again in the field as a plasticizing admixtureto increase the slump in order to obtain the flowingconcrete required for the placing conditions.

When flowing concrete is used, the flexural strengthis not changed significantly from that of the initialconcrete of the same w/c at a lower slump.

5.8.3 Drying shrinkage and creep-When low-slumpconcrete and flowing concrete are compared, the dryingshrinkage will be approximately the same if the watercontents are virtually identical. If, in conjunction withproducing flowing concrete, the water content of themixture at equal cement content is lowered, then thedrying shrinkage may be reduced. There seems to belittle change in the creep characteristics of concrete withthe use of these admixtures when comparisons are madeon the basis of equal w/c concretes.

5.8.4 Air entrainment-Higher dosages of air-en-training admixture usually are required for flowingconcrete to maintain proper air content as compared toconventional concrete. As with any air-entrained con-crete, the air content in the field must be checked sothat the air-entraining admixture dosage can be modi-fied as required to keep the air content in the specifiedrange.

5.8.5 Resistance to freezing and thawing-Flowingconcrete exhibits degrees of resistance to freezing andthawing similar to conventional concrete with a similarw/c and air-void system. The air-void structure mayhave larger spacing factors and a decrease in the num-ber of voids per inch compared to the control concrete;however, satisfactory resistance to freezing and thaw-ing has been achieved in most cases. Lucas (1981) indi-cated that concrete made with a high-range water-re-ducing admixture (superplasticizer) has a smaller ten-dency to absorb chloride than do untreated concretes ofthe same w/c.

5.8.6 Permeability-Flowing concrete has resistanceto chloride penetration similar to, if not slightly greaterthan, that of conventional concrete with the same w/c.When the admixture is used to reduce the w/c, the re-sistance of the concrete to chloride penetration is evengreater. Flowing concrete tends to be of lower permea-bility because of better consolidation, reduced bleed-ing, and increased cement hydration.

Low permeability results from low w/c concretewhen it is properly placed and cured. Flowing concretewith a low w/c can be placed and consolidated easily.This allows concrete with a w/c below 0.40 to be placedeasily; therefore, the resultant concrete, if properlycured, can be of extremely low permeability and havegood resistance to the penetration of aggressive solu-tions.

212.3R-22 MANUAL OF CONCRETE PRACTICE

5.8.7 Bond-Flowing concrete can improve bondstrength to reinforcing steel when compared to similarconcrete with a 100-mm (4-in.) slump (Collepardi andCorradi 1979) conventional and flowing concrete. Brett-man, Darwin, and Donahey (1986) found that in reinforcedconcrete beams bond strength of concrete of equal w4c wasdecreased as the slump was increased and was decreased thelonger the concrete remained unhardened. Proper vibrationis required for both concretes.the reinforcing is more easilycrete.

Proper consolidation aroundachieved with flowing con-

5.9-Quality assuranceIt is desirable and sometimes necessary to determine

that an admixture is the same as that previously testedor that successive lots or shipments are the same. Teststhat can be used to identify admixtures include solidscontent, density, infrared spectrophotometry for or-ganic materials, chloride content, pH, and others. Ad-mixture manufacturers can recommend which tests aremost suitable for their admixtures and the results thatshould be expected. Guidelines for determining uni-formity of chemical admixtures are given in ASTMc 1017.

5.10-Control of concrete5.10.1 Concrete mixture proportioning-Concrete

should be proportioned with flowing characteristics inmind; therefore, sufficient fines must be present in themixture to allow the desired slump to be achieved with-out excessive bleeding and segregation. Trial batchesusually are prepared to confirm concrete characteris-tics. The mixture is adjusted in the field to verify flow-ing characteristics.

Necessary adjustments can be made to assure the userof the optimum mixture with regard to slump, work-ability, rate of slump loss, and setting characteristics.Verification of early strengths can be accomplished ifrequired. The rate of slump loss should be noted andadjusted as required. The mixture proportions alsoshould conform to the procedures indicated in ACI211.1 or ACI 211.2. Flowing concrete should be placedin accordance with ACI 304 and consolidated in ac-cordance with ACI 309.

5.10.2 Field control-Flowing concrete control re-quires checking the initial slump or water content priorto the addition of the admixture to assure that the wa-ter content and the w/c are as required. After the plas-ticizing admixture is added and thoroughly mixed intothe batch, the resulting slump should be in the speci-fied range. For air-entrained concrete, the air contentalso must be checked at the point of discharge into theforms.

Rate of slump loss, initial setting characteristics, andboth early and final strength results may require mix-ture adjustments. Slump loss and setting characteristicsmay be adjusted by changes in the plasticizing admix-ture dosage or by concurrent use of accelerating or re-tarding admixtures. When the concrete placement is

abnormally slow or the temperature is high, orthe USC of a Type 2 admixture may be desirable.

both,

Changes in cement composition or in aggregategrading, or both, can cause significant variations in theflowing concrete characteristics. Therefore, thesechanges should be minimized.

CHAPTER 6-MISCELLANEOUS ADMIXTURES6.1 -Gas-forming admixtures

6.1.1 Introduction-The gas-void content of con-crete can be increased by the use of admixtures thatgenerate or liberate gas bubbles in the fresh mixtureduring and immediately following placement and priorto setting of the cement paste. Such materials are addedto the concrete mixture to counteract settlement andbleeding, thus causing the concrete to retain morenearly the volume at which it was cast. They are notused for producing resistance to freezing and thawing;any such effect is incidental. Air-entraining admixturesare discussed in Chapter 2.

6.1.2 Materials-Admixtures that produce these ef-fects are hydrogen peroxide, which generates oxygen;metallic aluminum (Menzel 1943; Shideler 1942), whichgenerates hydrogen; and certain forms of activated car-bon from which adsorbed air is liberated.

Only aluminum powder is used extensively for gasformation. An unpolished powder usually is preferred,although polished powder may be used when a slowerreaction is desired. The rate and duration of gas evolu-tion of the cement (particularly alkali content), temper-ature, w/c, and fineness and particle shape of thealuminum powder.

The effectiveness of the treatment is controlled by theduration of mixing, handling, and placing operationsrelative to the speed of gas generation. The additionrate may vary from 0.006 to 0.02 percent by weight(mass) of cement under normal conditions, althoughlarger quantities may be used to produce low-strengthcellular concrete. Approximately twice as much alumi-num powder is required at 40 F (4 C) as at 70 F (21 C)to produce the same amount of expansion.

Because of the very small quantities of aluminumpowder generally used [a few grams per 100 lb (45 kg)of cement], and because aluminum powder has a ten-dency to float on the mixing water, it generally is pre-mixed with fine sand, cement, or pozzolan, or incor-porated in commercially available admixtures havingwater-reducing set-retarding effects.

In cold weather, it may be necessary ‘to speed up therate of gas generation by the addition of such causticmaterials as sodium hydroxide, hydrated lime, or tri-sodium phosphate. This may be done to insure suffi-cient gas generation before the mixture has set andhardened.

6.1.3 Effectiveness- The release of gas, when prop-erly controlled, causes a slight expansion of freshlymixed concrete. When such expansion is restrained,there will be an increase in bond to horizontal reinforc-ing steel without excessive reduction in compressive

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-23

strength. Too much gas-producing material may pro-duce large voids, seriously weakening the matrix. To aconsiderable extent, the effect on compressive strengthdepends on the degree to which the tendency of themixture to expand is restrained; therefore, it is impor-tant that confining forms be tight and adequatelyclosed. Gas-forming agents will not overcome shrink-age after hardening caused by drying or carbonation.

6.2-Grouting admixtures6.2.1 Introduction-Many of the admixtures used for

specific purposes in concrete are used as grouting ad-mixtures to impart special properties to the grout. Oil-well cementing grouts encounter high temperatures andpressures with considerable pumping distances in-volved. Grout for preplaced-aggregate concrete re-quires extreme fluidity and nonsettling of the heavierparticles. Nonshrink grout requires a material that willnot exhibit a reduction from its volume at placement.Installation of tile subjects bonding and joint-fillinggrout to very fast drying or loss of water through ab-sorption by the substrate and the tile. A wide variety ofspecial purpose admixtures are used to obtain the spe-cial properties required.

6.2.2 Materials-For oil-well cementing grouts, re-tarders, as described in Chapter 4, are useful in delay-ing setting time. Bentonite clays may be used to reduceslurry density, and materials such as barite and iron fil-ings may be used to increase the density (Hansen 1954).Tile grouts and certain other grouts use materials suchas gels, clays, pregelatinized starch, and methyl cellu-lose to prevent the rapid loss of water.

Grout fluidifiers for preplaced-aggregate concretegrouts usually contain water-reducing admixtures alongwith admixtures to prevent settlement of heavy constit-uents of the grout. Nonshrink grouts may contain gas-forming or expansion-producing admixtures, or both.

Special grout applications may require such admix-tures as accelerators and air-entraining materials as de-scribed in other sections. Tests should be conducted todetermine the compatibility of admixtures with the ce-ment to be used.

6.2.3 Effect-Retarders may be used to keep a groutfluid at temperatures up to 400 F (204 C) and pressuresas high as 18,000 psi (124 MPa) for 1 hr or more.Grouting is a highly specialized field, usually requiringmaterial properties not necessary for ordinary concret-ing operations. The admixture manufacturer’s sugges-tions on addition rates should be followed; however,tests must be performed on the grout to determine ifthe properties of the grout meet the project require-ments.

6.3-Expansion-producing admixtures6.3.1 Introduction-Admixtures that expand during

the hydration period of the concrete or react with otherconstituents of the concrete to cause expansion are usedto minimize the effects of drying shrinkage. They areused both in restrained and unrestrained concreteplacement.

6.3.2 Materials-The most common admixture forthis purpose is a combination of finely divided or gran-ulated iron and chemicals to promote oxidation of theiron. Expansion is greatest when the mixture is exposedalternately to wetting and drying. Expansive cementsare used on large projects where a predetermined uni-form degree of expansion is required (Klein and Trox-ell 1958). Materials are available (calcium sulfoalumi-nates) that can be added to portland cements toproduce useful amounts of expansion. These materialsare part of the cementitious component and, therefore,not admixtures. For additional information regardingthese cements refer to ACI 223.

6.3.3 Effect of expansion on concrete-The con-trolled expansion produced by these admixtures may beof about the same magnitude as the drying shrinkageexpected at later ages or it may be slightly greater. Fora given application, the extent of expansion and thetime interval during which it takes place are very im-portant and must be under control for the most satis-factory results.

For unrestrained concrete, the expansion must nottake place before the concrete gains sufficient tensilestrength, or else the concrete will be disrupted. For re-strained applications, the concrete must be strongenough to withstand the compressive stresses devel-oped. It is reported that restraint in only one direction(Klein, Karby, and Polivka 1961) creates some degreeof compression in the other two orthogonal directions.

6.4-Bonding admixtures6.4.1 Introduction-Admixtures specifically formu-

lated for use in portland cement mixtures to enhancebonding properties generally consist of an organicpolymer emulsion commonly known as latex (Goeke1958; Ohama 1984). In general, latex forms a filmthroughout the concrete.

6.4.2 Materials-A wide variety of types of latex isused in paints, paper coatings, textile backing, etc. La-tex for use as a concrete admixture is formulated to becompatible with the alkaline nature of the portland ce-ment paste and the various ions present. An unstableemulsion will coagulate in the mixture, rendering it un-suitable for use.

6.4.3 Function- When used as admixtures in thequantities normally recommended by the manufactur-ers (5 to 20 percent by weight [mass] of the cement),different latexes may affect the unhardened mixturedifferently. For example, a film-forming latex may feeltacky in contact with air.

Water still is necessary to hydrate the portland ce-ment of the cement-polymer system. The polymer com-ponent becomes effective only when the emulsion isbroken through a drying out process. The polymeremulsion carries a portion of the mixing water into themixture, the water being released to the cement duringthe hydration process.

At the same time, this release of water sets the emul-sion. Hence, after an initial 24 hr of moist curing toeliminate chances of cracking, additional moist curing

212.3R-24 MANUAL OF CONCRETE PRACTICE

is not necessary and actually is undesirable since theemulsion will not have an opportunity to dry and de-velop the desired strength. The only exception to this iswhen a very low w/c is used (less than 0.3 by weight[mass]).

Upon drying or setting, the polymer particles coa-lesce into a film, adhering to the cement particles andto the aggregate, thus improving the bond between thevarious phases. The polymer also fills microvoids andbridges microcracks that develop during the shrinkageassociated with curing (Isenburg et al. 1971). This sec-ondary bonding action preserves some of the potentialstrength normally lost due to microcracking.

Greater strength and durability are associated withthe lower w/c of latex mixtures. The polymer particlesact as a water replacement, resulting in more fluiditythan in mixtures without latex, but having a similar w/cratio.

The compressive strength of moist-cured grouts,mortars, and concrete made with these materials maybe greater or less than that of mixtures of the same ce-ment content without the admixture, depending on theadmixture used (Grenley 1967). However, the increasein bond, tensile, and flexural strengths far outweigh thepossible disadvantage of slight compressive-strength re-duction. Latex-modified concrete has better abrasionresistance, better resistance to freezing and thawing,and reduced permeability.

6.4.4 Limitations-Surfactants present in latex emul-sions can entrap air and may require that a foam-sup-pressing agent be used. Dosage rates for air-entrainingagents will be affected. Some types of polymers willsoften in the presence of water; therefore, these typesshould not be used in concrete that will be in contactwith water during service. The ultimate result obtainedwith a bonding admixture is only as good as the sur-face to which the mixture is applied. The surface mustbe clean, sound, and free from such foreign matter aspaint, grease, and dust.

6.5-Pumping aids6.5.1 Introduction-Pumping aids for concrete are

admixtures with the sole function of improving con-crete pumpability. They normally will not be used inconcrete that is not pumped or in concrete that can bepumped readily.

The primary purpose of using admixtures to enhancepumpability of concrete is to overcome difficulties thatcannot be overcome by changes in the concrete mixtureproportions. As in the use of many ingredients in con-crete, the objective is economic.

6.5.2 Materials-Many pumping aids are thickenersthat increase the cohesiveness of concrete. The Stan-dards Association of Australia* identified five cate-gories of thickening admixtures for concrete and mor-tar as follows:

* " I n f o r m a t i o n on Thickening Admixtures for Use in Concrete and Mor-tar, " Draft for Comment, Miscellaneous Publication No. DR 73146, Stan-dards Association of Australia, Committee BD/33, Oct. 1973, unpublished.

1. Water-soluble synthetic and natural organic poly-mers that increase the viscosity of water-cellulose de-rivatives (methyl, ethyl, hydroxyethyl, and other cellu-lose gums); polyethylene oxides; acrylic polymers;polyacrylamides; carboxyvinyl polymers; natural wa-ter-soluble gums; starches; and polyvinyl alcohol,

2. Organic flocculants-carboxyl-containing styrenecopolymers, other synthetic polyelectrolytes, and natu-ral water-soluble gums.

3. Emulsions of various organic materials-paraffin,coal tar, asphalt, and acrylic and other polymers.

4. High-surface-area inorganic materials-bentonitesand organic-modified bentonites and silica fume.

5. Finely divided inorganic materials that supplementcement in cement paste-fly ash and various raw orcalcined pozzolanic materials, hydrated lime, and nat-ural or precipitated calcium carbonates and variousrock dusts.

This list does not include all of the materials listed inMcCutcheon’s Functional Materials (McCutcheon Di-vision 1975). Classifications may be misleading sincethe performance of a given admixture can changedrastically with change of dosage rates, cement com-position, mixing temperature and time, and other fac-tors. An example is provided by the polyethylene ox-ides. When used in small amounts of 0.01 to 0.05percent of cement weight (mass), they improve pumpa-bility. Larger amounts produce thickening that may ormay not disappear upon prolonged mixing.

Other examples are provided by synthetic polyelec-trolytes, which act as flocculants or thickeners depend-ing upon dosage levels. It would appear to be highlyundesirable to induce flocculation and increase bleed-ing in pumped concrete. Nevertheless, these admixturesare considered effective in pumped concrete becausethey lower bleeding capacity or total bleeding despitecausing increases in initial rates of bleeding.

Other problems with the listing given previously oc-cur with the natural gums (algins, tragacanth, arabic).They can function as thickeners or flocculants depend-ing upon dosage levels and other factors. These agentsand some of the synthetic materials also can have dis-persing or water-reducing effects. Gum arabic is apowerful water reducer for calcium-sulfate plasters, butin portland cement pastes, it can produce a gluelikestickiness.

Factors to consider in the use of emulsions (paraf-fins, polymers) are whether they function in the desiredway in cement paste by remaining stable or by breakingof the emulsion. Both types of paraffin emulsion areconsidered to be useful in Australian concrete technol-ogy*

The listing given previously does not include such air-entraining agents or surface-active agents widely used inconcrete as hydroxylated carboxylic acid derivatives,lignosulfonates and their derivatives, formaldehyde-condensed naphthalene sulfonates, melamine poly-mers, and other set-retarding or water-reducing admix-tures, The omission here is deliberate because asubstantial proportion of concrete that is to be pumped

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-25

in North America will be specified as air-entrainedconcrete, and probably also will contain a water-reduc-ing or set-retarding admixture. Therefore, such admix-tures may be considered to be normal constituents ofconcrete.

Evaluation of and experience with these admixturesare well known. In this chapter, these types of admix-tures are not considered specifically as pumping aids.They will be present in many cases in combination withother agents introduced for the specific purpose of im-proving pumpability. In such cases, evaluation of ef-fects of such combinations on pumpability and otherproperties of concrete will be required to determinewhether or not adverse interactions occur between ad-mixtures.

6.5.3 Effects on concrete-A side effect of a con-crete pumpability-enhancing admixture is any effect itmay have on fresh concrete other than that on pumpa-bility, and any effect that changes the characteristics ofthe hardened concrete. Since the main effect of a waterthickener is to increase viscosity, substantial thickeningcan increase water requirements with the usual conse-quences of reduced strength. By using a suitable dis-persant in combination with a thickening agent, no in-crease in water may be required. At certain dosagelevels, some thickeners act as dispersants of solid par-ticles. Many of the thickening agents cause entrainmentof air. To control air content, a defoamer (for exam-ple, tributyl phosphate) may be needed, especially whenhigher concentrations of pumping aid are used in mor-tars and concretes.

Many of the synthetic and natural organic thickeningagents retard the setting of portland cement pastes. Fordosages of methyl or hydroxyethyl cellulose of 0.1 per-cent or more by weight (mass) of portland cement, re-tardation may be substantial. In any case, the particu-lar concrete system in which a pumpability-enhancingadmixture is incorporated must be evaluated in terms ofside effects upon the fresh and hardened concrete inaddition to assessing the effectiveness of the admixturein performing its intended function.

6.6-Coloring admixtures6.6.1 Introduction-Pigments specifically prepared

for use in concrete and mortar are available both asnatural and synthetic materials. They are formulated toproduce adequate color without materially affecting thedesirable physical properties of the mixture. They arecovered by ASTM C 979.

6.6.2 Materials-The pigments listed in Table 6.6.2may be used to obtain a variety of colors.

6.6.3 Effects on concrete properties-The additionrate of any pigment to concrete normally should notexceed 10 percent by weight (mass) of the cement (Wil-son 1927); however, some pigments, such as carbonblack, should be used at lesser quantities. Natural pig-ments usually are not ground as finely as, and often arenot as pure as, synthetic materials and generally do notproduce as intense a color per unit of addition. Exceptfor carbon black, additions of less than 6 percent of

Table 6.6.2pigments

- Colors produced by various

Grays to black

BlueBright red to deep redBrown

Ivory,Green

White

cream, or buff

Black iron oxideMineral blackCarbon blackUltramarine blueRed iron oxideBrown iron oxideRaw and burnt umberYellow iron oxideChromium oxidePhthalocyanine greenTitanium dioxide

pigment generally have little or no effect on the physi-cal properties of the fresh or hardened concrete. Largerquantities may increase the water requirement of themixture to such an extent that the strength and otherproperties, such as abrasion resistance, may be ad-vcrscly affected.

The addition of an unmodified carbon black will in-crease considerably the amount of air-entraining ad-mixture needed to provide resistance of the concrete tofreezing and thawing (Taylor 1948). However, mostcarbon blacks available for coloring concrete do con-tain air-entraining materials in sufficient quantity tooffset the inhibiting effect of the carbon black.

Brilliant concrete colors are not possible with eithernatural or synthetic pigments due to their low allow-able addition rates and the masking effects of the ce-ment and aggregates. Stronger colors can be obtainedif white rather than grey cement is used. If bright colorsare required, a surface coating should be specified inlieu of pigment admixtures.

6.7-Flocculating admixturesSynthetic polyelectrolytes, such as vinyl acetate-mal-

eic anhydride copolymer, have been used as flocculat-ing admixtures. Published reports (Bruere and Mc-Gowan 1958; Vivian 1962) indicate that these materialsincrease the bleeding rate, decrease the bleeding capac-ity, reduce flow, increase cohesiveness, and increaseearly strength. Although the mechanism of this actionis not understood fully, it is believed that these com-pounds, containing highly charged groups in theirchains, are adsorbed on cement particles, linking themtogether. The net result is equivalent to an increase ininterparticle attraction, which increases the tendency ofthe paste to behave as one large floc.

6.8-Fungicidal, germicidal, and insecticidaladmixtures

6.8.1 Introduction- Certain materials have beensuggested as admixtures for concrete or mortar to im-part fungicidal, germicidal, and insecticidal properties.The primary purpose of these materials is to inhibit andcontrol the growth of bacteria and fungus on concretefloors and walls or joints. They may not always becompletely effective.

6.8.2 Materials-The materials that have been foundto be most effective are: polyhalogenated phenols (Le-

212.3R-26 MANUAL OF CONCRETE PRACTICE

vowitz 1952), dieldrin emulsion (Gay and Wetherly1959), and copper compounds (Robinson and Austin195 1; and Young and Talbot 1945).

6.8.3 Effectiveness-Addition rates vary from 0.1 to10 percent by weight (mass) of the cement, dependingon the concentration and composition of the chemical.The higher rates, above 3 percent, may have an adverseeffect on the strength of the concrete. The effectivenessof these materials, particularly the copper compounds,is reported to be of a temporary nature. This probablywill vary with the type of wear and cleaning methodsemployed.

6.9-Dampproofing admixtures6.9.1 Introduction-The term “dampproofing” im-

plies prevention of water penetration of dry concrete orstoppage of water transmission through unsaturatedconcrete. However, admixtures have not been found toproduce such effects and the term has come to mean areduction in rate of penetration of water into dry con-crete or in rate of transmission of water through unsat-urated’ concrete.

When some concrete dams, retaining walls, tanks,and other structures show evidence of leakage, it usu-ally is the result of faulty production and placement ofconcrete. When properly proportioned concrete mix-tures are used and placed with high-quality workman-ship under qualified inspection, the concrete in a struc-ture should be virtually impermeable, although leakagestill may occur through cracks. Dampproofing admix-tures cannot be expected to be as reliable or effective asapplying a moisture-barrier system to the concrete.

Dampproofing may reduce the rate of penetration ofaggressive chemicals found in water; however, it willnot stop them. Dampproofing admixtures also mayreduce the penetration of water into concrete, thus de-laying the effects of damage caused by freezing andthawing by reducing the amount or rate of moistureentering the concrete.

An admixture described as a dampproofer may havea secondary effect on the properties of fresh concretenot directly indicated by the name. For example, it maypromote entrainment of air; thus, it may more prop-erly be considered an air-entraining admixture.

This section deals with those effects directly impliedby the word “dampproofing” on the properties ofhardened concrete.

6.9.2 Materials-Admixtures for dampproofing areused to render the concrete hydrophobic and thereforecapable of repelling water that is not under hydrostaticpressure.

Admixtures for dampproofing include soaps, butylstearate, and certain petroleum products (Dunagan andErnst 1934; Uppal and Bahadur 1958).

1. The soaps are composed of salts of fatty acids,usually calcium or ammonium stearate or oleate. Thesoap content usually is 20 percent or less, the remain-der being calcium chloride or lime. Total soap addedshould not exceed 0.2 percent by weight (mass) of con-crete. Soaps cause entrainment of air during mixing.

2. Butyl stearate reportedly performs better thansoap as a water repellent. It does not entrain air andhas a negligible effect on strength. It is added as anemulsion with the stearate being 1 percent by weight(mass) of the cement.

3. Among petroleum products are mineral oils, as-phalt emulsions, and certain cutback asphalts. Heavymineral oil is effective as a water repellent for concreteand in reducing its permeability. The petroleum prod-uct must be a fluid and have a viscosity approximatelyequal to SAE60, with no fatty or vegetable oils.

4. In addition, there is a group of miscellaneous ma-terials sometimes advertised as dampproofing agents.All of these are usually detrimental to concretestrength, and none are truly dampproofers. These in-clude barium sulfate and calcium and magnesium sili-cates, finely divided silica and naphthalene, colloidalsilica and fluosilicate, petroleum jelly and lime, cellu-lose materials and wax, silica and aluminum, coal tarcut with benzene, and sodium silicate.

6.9.3 Effectiveness-Dampproofing admixtures mayaid in shedding or repelling water from the surface ofthe concrete. Some dampproofing admixtures also mayaid in reducing the ability of rain to penetrate the sur-face of the concrete or reducing the wicking or wettingproperties of concrete. These admixtures, by reducingpenetration of the visible pores, may retard penetrationof rain into concrete block made of nonplastic mix-tures. Test data show that they also reduce the rate ofpenetration of moisture into the micropores of dryconcrete, but there is no indication that there are com-parable effects on the transmission of moisture throughunsaturated concrete, except when the concrete con-tains paste with relatively high porosity.

A paste of high porosity results from low cementcontent and correspondingly high w/c, lack of curing,or from both factors. If the concrete has a sufficientlylow porosity, such as that obtained by producing awell-cured paste having w/c not over 0.6 by weight(mass), dampproofing agents give no appreciable im-provement.

The Building Research Advisory Board (1958) re-ported that in the opinion of the majority of 61 observ-ers, dampproofing admixtures are not “. . . effective oracceptable in controlling moisture migration throughslabs-on-ground.”

A special advisory committee to the Building Re-search Advisory Board reached the following conclu-sion on the basis of data from tests of moisture trans-mission through unsaturated concrete slabs: “TheCommittee does not find adequate data to demonstratethe effectiveness of any admixture to reduce the trans-mission of moisture through concrete slabs-on-groundin a manner sufficient to replace either a vapor barrieror granular base, or both, under conditions where suchprotection would be needed.”

6.10-Permeability-reducing admixturesPermeability refers to the rate at which water is

transmitted through a saturated specimen of concrete

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-27

under an externally maintained hydraulic gradient. Ad-mixtures of the kinds discussed in the previous sectiondo not reduce the permeability of saturated concrete.However, mineral powders (essentially fly ash, raw orcalcined natural pozzolans, and silica fume), properlyproportioned, reduce the permeability of mixtures inwhich the cement content of the paste is relatively low.This is due to the production of additional cementi-tious material, primarily calcium silicate hydrates andcalcium aluminum silicate hydrates, which form by thecombination of lime from the cement and silica andother components in the mineral powder.

The reduction of total water content by means of awater-reducing admixture should reduce the total po-rosity slightly, hut there are no adequate data to dem-onstrate that permeability is reduced materially. How-ever, decreased permeability with the use of high-rangewater reducers at equivalent w/c has been reported.

Polymer-emulsion admixtures have been used to re-duce permeability of concrete overlays for bridge decksand parking decks. The polymer particles coalesce intoa continuous film, which reduces permeability by seal-ing air voids and blocking microcracks (Isenburg et al.1971; Whiting 1981).*

Accelerating admixtures such as calcium chloride in-crease the rate of hydration (see Section 3.2.2), therebyreducing the length of time required for a concretemixture to attain a given fraction of its ultimate degreeof impermeability (see Section 3.2.1). However, anyadvantage attained this way is likely to be temporarysince, if conditions are such that water is being trans-mitted through the concrete, they also are conducive tocontinued hydration of cement.

6.11 -Chemical admixtures to reduce alkali-aggregate expansion

6.11.1 Introduction-The use of pozzolans to reduceexpansion caused by alkali-aggregate reaction has beenstudied widely and reported (Stanton 1950). As early as1950, reports began to appear on admixtures other thanpozzolans to reduce expansion caused by alkali-aggre-gate reaction. Since that time there has been little newinformation added in the form of meaningful researchof field practice (see ACI Committee 212’s 1963 re-port).

6.11.2 Materials-Soluble salts of lithium, barium,and certain air-entraining and some water-reducing set-retarding admixtures have been reported to produce re-duction in expansion of laboratory mortar specimens.Outstanding reductions have been obtained in suchspecimens using 1 percent additions of the lithium saltsand 2 to 7 percent additions by weight (mass) of ce-ment of certain barium salts. The lithium salts are veryexpensive.

Salts of proteinaceous materials and water-reducingset-retarding materials have shown moderate reduc-tions in expansion. Data on the protein air-entraining

*Also, Grenley, Dallas G., Michalyshin, J., and Molodovan, D., Mar. 1983,“Effect of Certain Concrete Admixtures on the Chloride Permeability ofConcrete,” Dow Chemical Company, Inc., presented at the Annual Meeting ofthe American Concrete Institute, Los Angeles.

admixtures are based on the use of 0.2 percent byweight (mass) of concrete. Some laboratories have re-ported on the use of calcium chloride with the bariumsalts to counteract strength loss.

Air entrainment, regardless of the admixture used,has been shown to lower the expansion slightly,

6.11.3 Effectiveness-Limited laboratory data on theuse of chemical admixtures to reduce expansion result-ing from the alkali-silica reaction are available; there-fore, no recommended practices are presented. Anyuser of these materials should test them thoroughly be-fore proceeding to field use.

6.12-Corrosion-inhibiting admixture6.12.1 Introduction-Many investigators have stud-

ied the corrosion of iron and steel with particular ref-erence to protective coatings. It has been found thatconcrete furnishes ample protection to the steel embed-ded in it, although a limited number of cases have beenreported in which infiltrating or percolating waters finda way through the concrete, removing or carbonatingthe calcium hydroxide (Cushman 1909; Cushman andGardener 1910). Reports indicate that the use of smallquantities of bentonite in the mixing water reduces cor-rosion by reducing the permeability of the concrete.

The major contributor to corrosion of reinforcingsteel is the presence of chlorides in the concrete. Thechlorides may come from circumstances such as expo-sure of the concrete to saline or brackish waters, expo-sure to saline soils from which chlorides can reach thesteel by diffusion through the concrete, by entrance ofdeicer solutions through cracks or pores in the con-crete, or use of calcium chlorides as an admixture con-stituent (see Chapter 3).

Corrosion caused by the inclusion or infiltration ofchlorides in concrete is difficult to control once it hasstarted. Numerous chemicals have been evaluated aspotential corrosion-inhibiting admixtures for concrete(Verbeck 1975; Clear and ‘Hay 1973; Griffin 1975).These include chromates, phosphates, hypophosphor-ites, alkalies, nitrites, and fluorides. Recently, calciumnitrite has been reported as an effective corrosion in-hibitor (Rosenberg et al. 1977). Studies of this materialare continuing (Virmani, Clear, and Pasko 1983).

6.12.2 Materials-The use of sodium benzoate at arate of 2 percent in the mixing water or a 10 percentbenzoate-cement slurry painted on reinforcement, orboth, have been described as effective (Lewis, Mason,and Brereton 1956). Analysis showed that the sodiumbenzoate remained in the concrete after five years’ ex-posure. It also accelerates compressive strength devel-opment.

Calcium lignosulfonate reduces the tendency for cor-rosion of steel in concrete containing calcium chloride(Kondo, Takeda, and Hideshima 1959).

Sodium nitrite has been investigated by Moskvin andAlekseyev (1958) as an inhibitor of corrosion of steel inautoclaved products. These authors suggest that thehigh alkalinity, which normally is present in concreteand which serves to passivate the steel, may be reduced

2 1 2 . 3 MANUAL OF CONCRETE PRACTICE

and Brereton 1956). Analysis showed that the sodiumbenzoate remained in the concrete after five years’ ex-posure. It also accelerates compressive strength devel-opment.

Calcium lignosulfonate reduces the tendency for cor-rosion of steel in concrete containing calcium chloride(Kondo, Takeda, and Hideshima 1959).

Sodium nitrite has been investigated by Moskvin andAlekseyev (1958) as an inhibitor of corrosion of steel inautoclaved products. These authors suggest that thehigh alkalinity, which normally is present in concreteand which serves to passivate the steel, may be reducedconsiderably by autoclave treatment, especially whensiliceous admixtures are present. Two to 3 percent so-dium nitrite by weight (mass) of cement was found tobe an efficient inhibitor under these conditions.

Sarapin (1958) found by storage tests that 2 percentsodium nitrite was effective in preventing corrosion ofsteel in concrete containing calcium chloride under cer-tain conditions of storage. Low-solubility salts such ascertain phosphates or fluosilicates and fluoaluminatesare beneficial, according to limited reports. Dosageshould be limited to 1 percent by weight (mass) of ce-ment.

6.12.3 Effect-Warnings have been sounded againstthe use of inhibitors, For example, the South AfricanNational Building Research Institute (1957) made thefollowing statement, “Integral additives: although cer-tain inert and reactive materials have shown promise,their use cannot be recommended at this stage, as in-sufficient evidence of their effectiveness or possible di-seffects from the inhibitor addition might reasonably beexpected, if the steel surface is clean and chlorides ab-sent; but under such conditions, there is unlikely to beserious trouble without added inhibitor.”

CHAPTER 7-REFERENCES

7.1-Recommended references

The documents of the various standards-producing organizations referred to in this documentfollow with their serial designation.

American Concrete Institute

116R201.2R211.2

212.1R

Cement and Concrete TerminologyGuide to Durable ConcreteStandard Practice for SelectingProportions for Normal, Heavyweight,and Mass ConcreteAdmixtures for Concrete

212.2R

222R223

304R

306R308309R311.1R311.4R318

506R523.lR

548R

Guide for Use of Admixtures inConcreteCorrosion of Metals in ConcreteStandard Practice for Use of Shrinkage-Compensating ConcreteGuide for Measuring, Mixing, Trans-porting, and Placing ConcreteCold Weather ConcretingStandard Practice for Curing ConcreteGuide for Consolidation of ConcreteManual of Concrete Inspection (SP-2)Guide for Concrete InspectionBuilding Code Requirements forReinforced ConcreteGuide to ShotcreteGuide for Cast-in-Place Low DensityConcretePolymers in Concrete

ASTM

C 94

C 114

C 125

C 138

C 150

C 173

C 231

C 233

C 260

C 457

C 494

C 595

C 666

C 979

Standard Specification for Ready-MixedConcreteStandard Methods for Chemical Analysisof Hydraulic CementTerminology Relating to Concrete andConcrete AggregatesStandard Test Method for Unit Weight,Yield, and Air Content (Gravimetric) ofConcreteStandard Specification for PortlandCementStandard Test Method for Air Contentof Freshly Mixed Concrete by theVolumetric MethodStandard Test Method for Air Contentof Freshly Mixed Concrete by thePressure MethodStandard Test M e t h o d f o r Air-Entraining Admixtures for ConcreteStandardSpecificationforAir-EntrainingAdmixtures for ConcreteStandard Practice for MicroscopicalDetermination of Air Void Content andParameters of the Air Void System inHardened ConcreteStandard Specification for Chemical Ad-mixtures for ConcreteStandard Specification for BlendedHydraulic CementsStandard Test Method for Resistance ofConcrete to Rapid Freezing andThawingStandard Specification for Pigments, forIntegrally Colored Concrete

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-29

C 1017

D 98

Standard Specification for Chemical Ad-mixtures for Use in Producing FlowingConcreteStandard Specification for CalciumChloride

These publications may be obtained from the follow-ing organizations:

American Concrete InstituteP. 0. Box 19150Detroit, MI 48219-0150

ASTM 1916 Race StreetPhiladelphia, PA 19103

7.2-Cited referencesAbrams, Duff A., 1924, “Calcium Chloride as an Admixture in

Concrete,” Proceedings, ASTM, V. 24, Part 2, pp. 781-834.Adams, R. F., and Kennedy. J. C., Dec. 1950, “Effect of Batch

Size and Different Mixers on the Properties of Air-Entrained Con-crete,” Laboratory Report No. C-532, U. S. Bureau of Reclamation,Denver.

Angstadt, R. L., and Hurley, F. R., July 18, 1967, “Spodumeneas an Accelerator for Hardening Portland Cement,” U. S. Patent No.3.331.695.

Arber, M. G., and Vivian, H. E., 1961, “Inhibition of the Corro-sion of Steel Imbedded in Mortars,” Australian Journal of AppliedScience (Melbourne), V. 12, No. 12, pp. 339-347.

Bash, S. M., and Rakimbaev, Sh. M., June 1969, “Quick-SettingCement Mortars Containing Organic Additives,” Beton i Zhelezobe-ton (Moscow), V. 15, No. 7, pp. 44-45. (in Russian)

Baslazs, B. Y.; Kelmen, J.; and Kilian, J., 1959, “New Method forAccelerating the Hardening of Concrete,” Epitoanyag (Budapest), V.10, No. 9, pp. 326-331.

Bensted, J., 1978, “Effect of Accelerator Additives on the EarlyHydration of Portland Cement,” I l Cemenro (Rome), V. 75, pp. 13-19.

Berger, R. L.; Kung, J. H.; and Young, J. F., 1967, “Influence ofCalcium Chloride in the Drying Shrinkage of Alite Paste,” ASTMJournal of Testing and Evaluation, V. 4, No. 1, pp. 85-93.

Blanks, R. F., and Cordon, W. A., Feb. 1949, “Practices, Expe-riences, and Tests with Air-Entraining Agents in Making DurableConcrete,” ACI JOURNAL, Proceedings V. 45, No. 6, pp. 469-488.

Brettman, Barie B.; Darwin, David; and Donahey, Rex C., Jan.-Feb. 1986, “Bond of Reinforcement to Superplasticized Concrete,”ACI JOURNAL, Proceedings V. 83, No. 1, pp. 98-107.

Bruere, G. M., 1963, “Importance of Mixing Sequence When Us-ing Set-Retarding Agents with Portland Cement,” Nature (London),V. 199, pp. 32-33.

Bruere, G. M., Mar. 1971, Journal of Applied Chemistry and Bio-technology (Oxford), V. 21, No. 3, pp. 61-64.

Bruere, G. M., and McGowan, J. K., 1958, “Synthetic Polyelec-

trolytes as Concrete Admixtures,” Australian Journal of AppliedScience (Melbourne), V. 9, No. 2, pp. 127-140.

Bruere, G. M.; Newbegin, J. D.; and Wilson, L. M., 1971, "Lab-oratory Investigation of the Drying Shrinkage of Concrete Contain-ing Various Types of Chemical Admixtures,” Technical Paper No. 1,Division of Applied Mineralogy, Commonwealth Scientific and ln-dustrial Research Organization, East Melbourne, 26 pp.

Building Research Advisory Board, July 1958, “Effectiveness ofConcrete Admixtures in Controlling the Transmission of MoistureThrough Slabs-on-Ground,” Publication No. 596, National Re-search Council, Washington, D.C.

Calcium Chloride Institute, 1959, Calcium Chloride in Concrete,3rd Edition, pp. 40-41.

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-struction Parameters,” Report No. FHWA-RD-73-32, Federal High-way Administration, Washington, D.C., 103 pp.

Collepardi, M., and Corradi, M., 1979, “Influence of Naphtha-lene-Sulfonated Polymer Based Superplasticizers on the Strength ofOrdinary and Lightweight Concrete,” ACI SP 62-16.

Collepardi, M.; Marcialis, A.; and Solinas, V., 1973, “The Influ-ence of CaCl, on the Properties of Cement Paste,” Il Cemenro, V.70, pp. 83-92.

Cordon, William A., 1966, Freezing and Thawing of Concrete,Mechanisms and Control, ACI Monograph No. 3, American Con-crete Institute/Iowa State University Press, Detroit, 99 pp.

Cordon, William, A., et al., June 1946, “Entrained Air in Con-crete: A Symposium,” ACI JOURNAL, Proceedings V. 42, No. 6, pp.601-699, and Discussion, pp. 700-1-700-12.

Cushman, A. S., 1909, “Preservation of Iron and Steel,” Engi-neering (London), V. 87, p. 742.

Cushman, A. S., and Gardener, H. A., 1910, Corrosion and Pres-ervation of Iron and Steel, McGraw-Hill Book Co., New York, p. 11.

Djabarov, N. D., Feb. 1970, “Oxalic Acid as an Additive to Ce-ment,” Zement-Ku/k-Gips (Wiesbaden), V. 23, No. 2, pp. 88-90. (inGerman)

Dodson, V. H.; Farkas, E.; and Rosenberg, A. M., Oct. 1965,“Non-Corrosive Accelerator for Setting of Cement,” U. S. PatentNo. 3,210,207.

Dunagan, W. M., and Ernst, G. C., 1934, “Study of the Permea-bility of a Few Integrally Water-Proofed Concretes,” Proceedings,ASTM, V. 34, Part 1, pp. 383-392.

Duriex, M., and Lezy, R., 1956, “New Possibilities for Ensuringthe Rapid Hardening of Cements, Mortars, and Concretes,” An-na/es, Institute Technique du Batiment et des Travaux Publics (Paris),No. 9, p. 138.

Edwards, G. C., and Angstadt, R. L., 1966, “Effect of Some Sol-uble Inorganic Admixtures on the Early Hydration of Portland Ce-ment,” Journal of Applied Chemistry and Biotechnology (Oxford),V. 16, No. 5, pp. 166-168.

Farmer, H. G., Jan. 1945, “Air-Entraining Portland Cement inConcrete Block,” Rock Products, V., 48, p. 209.

Feret, L., and Venuat, N., 1957, “Effect on Shrinkage and Swell-ing of Mixing Different Cements to Obtain Rapid Set,” Revue desMateriaux de Construction (Paris), No. 496, pp. l-10.

Foster, Bruce, 1966, “Chemical Admixtures,” Significance of Testsand Properties of Concrete and Concrete-Making Materials, STP-169A, ASTM, Philadelphia, pp. 556-665, 571.

Gay, F. J., and Wetherly, A. H., Sept. 1959 (Mar. 1960), “Ter-mite Proofing of Concrete,*’ Constructional Review (Sydney), V. 32,NO. 9, pp. 26-28. Also, Summary, ACI JOURNAL, Proceedings V. 56,No. 9, p. 904.

Gebler, Steven, Sept.-Oct. 1983, “Evaluation of Calcium Formateand Sodium Formate as Accelerating Admixtures for Portland Ce-ment Concrete,” A C I JOURNAL , Proceedings V. 80, No. 5, pp. 439-494.

Gibbons, C. S., 1978, “Porous Particulate Materials for Impart-ing Freeze-Thaw Resistance to Concrete,” Report No. 78-4, OntarioResearch Foundation.

Goeke, D. J., May 1958, “Bonding of Cementitious Materials,”Concrete Construction, V. 3, NO. 5, pp. 18-30.

Greening, N. R., and Landgren, R., Sept. 1966, “Surface Discol-

212.3R-30 MANUAL OF CONCRETE PRACTICE

Cement and Concrete Association, London, pp. 598-632.Isenburg, J. E.; Rapp, D. E.; Sutton, E. J.; and Vanderhoff, J.

W., 1971, “Microstructure and Strength of the Bond Between Con-crete and Sytrene-Butadiene Latex-Modified Mortar,” Highway Re-search Record No. 370, Htghway Research Board, pp. 75-89.

Kennedy, H., and Brickett, E. M., Mar. 1986, “Application of Air-Entraining Agents in Concrete and Products,” Pit and Quarry, V. 38,p. 144.

Keunning, William H., and Carlson, C. C., Dec. 1956, “Effect ofVariations in Curing and Drying on the Physical Properties of Con-crete Masonry Units,” Development Department Bulletin No. D13,Portland Cement Association, Skokie, 129 pp.

Klein, Alexander; Karby, Tsevi; and Polivka, Milos, July 1961,“Properties of an Expansive Cement for Chemical Prestressing,”ACI JOURNAL , Proceedings V. 58, No. , pp. 59-82.

Klein, Alexander, and Troxell, G. E, 1958, “Studies of CalciumSulfoaluminate Admixtures for Expansive Cement,” Proceedings,ASTM, V. 58, pp. 986-1008.

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

Kossivas, T. G, Oct. 14, 1971, “Setting Accelerators for PortlandCement,” German Patent No. 2,114,081.

Kroone, B., Mar. 2, 1968, “Reaction Between Hydrating PortlandCement and Ultramarine Blue,” Chemistry and Industry (London),pp. 287-288.

Levowitz, L. D., June 1952, “Anti-Bacterial Cement Gives LongerLasting Floors,” Food Engineering, V. 24, pp. 57-60 and 134-135.

Lewis, J. I. M.; Mason, E. E.; and Brereton, D., Aug. 1956, “So-dium Benzoate in Concrete,” Civil Engineering and Public WorksReview (London), V. 51, No. 602, pp. 881-882.

Lieber, W., and Richartz, W., 1972, “Effect of Triethanolamine,Sugar, and Boric Acid on the Setting and Hardening of Cements,”Zement-Kalk-Gips (Wiesbaden), V. 25, No. 9, pp. 403-409. (in Ger-man)

Litvan, G. G., Jan. 1972, “Phase Transitions of Adsorbates: IV.Mechanism of Frost Action in Hardened Cement Paste,” Journal ofthe American Ceramic Society, V. 55, No. 1, pp. 38-42.

Litvan, G. G., July-Aug. 1983, “Air Entrainment in the Presenceof Superplasticizers,” ACI JOURNAL, Proceedings V. 80, No. 4,pp. 326-331.

Litvan, G. G., Sept.-Oct. 1985, “Further Study of Particulate Ad-mixtures for Enhanced Freeze-Thaw Resistance of Concrete,” ACIJOURNAL, Proceedings V. 82, No. 5, pp. 724-730.

Litvan, G. G., and Sereda, P. J., Jan. 1978, “Particulate Admix-tures for Enhanced Freeze-Thaw Resistance of Concrete,” Cementand Concrete Research, V. 8, No. 1, pp. 53-60.

Lucas, Walter, 198 1, “Chloride Penetration in Standard Concrete,Water-Reduced Concrete, and Superplasticized Concrete,” Develop-ments in the Use of Superplasticizers, SP-68, American Concrete In-stitute, Detroit, pp. 253-257.

Maclnnis, C., and Beaudoin, J. J., Mar. 1974, “Mechanism ofFrost Damage in Hardened Cement Paste,” Cement and ConcreteResearch, V. 4, No. 2, pp. 139- 148.

Mahar, J. W.; Parker, H. W.; and Wuellner, W. W., 1975,“Shotcrete Practice in Underground Construction,” Report No.FRA-OR&D 75-90, Department of Transportation, Federal RailroadAdministration, Washington, D.C., 505 pp.

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

Mather, Bryant, 1964, “Drying Shrinkage-Second Report ,”Highway Research News, No. 15, pp. 34-38.

Mather, Bryant, 1979, “Tests of High-Range Water-Reducing Ad-mixtures” Superplasticizers in Concrete, SP-62, American ConcreteInstitute, Detroit, pp. 157-166.

McCutcheon Division, 1975, “Functional Materials,” Mc-Cutcheon Division MC Publishing Co., Ridgewood.

Menzel, Carl A., Jan. 1943, “Some Factors Influencing theStrength of Concrete Containing Admixtures of Powdered Alumi-num, ” ACI JOURNAL , Proceedings V. 39, No. 3, pp. 165-184.

Mielenz, Richard C.; Wolkodoff, Vladimir E.; Backstrom, JamesE.; Burrows, Richard W.; and Flack, Harry L., 1958, “Origin, Evo-lution and Effects of the Air-Void System in Concrete,” ACI J O U R -NAL, Proceedings V. 55, No. 1-4;: “Entrained Air in UnhardenedConcrete” July, pp. 95-122; “Influence of Type and Amount ot Au-Entraining Agent,” Aug., pp. 261-272; “Influence of Water-CementRatio and Compaction,” Sept., pp. 359-376; and “Air-Void Systemin Job Concrete,” Oct., pp. 507-5 18.

Moskvin, V. M., and Alekseyev, S. N., Dec. 1958, “Methods forIncreasing the Resistance IO Corrosion of Reinforcement in Rein-forced Concrete Structural Members,” Beton i Zhelezobeton (Mos-cow), V. 36, No. 12, pp. 21-23. (in Russian)

Murakami, J., and Tanaka, G., 1969, “Construction of CalciumThiosulfate to the Acceleration of the Hydration of Porrland Cementand Comparison with Other Soluble Inorganic Salts,” Proceedings,5th International Symposium on the Chemistry of Cement, CementAssociation of Japan, Tokyo, V. 2, pp. 422-436.

Nelson, J. A., and Young, J. F., May 1977, “Addition of Colloi-dal Silicas and Silicates to Portland Cement Pastes,” Cement andConcrete Research, V. 7, No. 3, pp. 277-282,

Ohama. Y., 1984, “Polymer Modified Mortars and Concretes,”Concrete Admixtures Handbook, Noyes Publications, Park Ridge,pp. 337-429.

Powers, T. C., 1949, “Air Requirement of Frost Resistant Con-crete,” Proceedings, Highway Research Board, V. 29, pp. 184-202.

Powers, T. C., Sept. 1964, “Topics in Concrete Technology; Part3, Mixtures Containing Intentionally Entrained Air,” Journal, PCAResearch and Development Laboratories, V. 6, No. 3, pp. 19-42.

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

Powers, Treval C., 1968, Properties of Fresh Concrete, John Wiley& Sons, New York, 664 pp.

Ramachandran, V. S., 1972, “Elucidation of the Role of ChemicalAdmixtures in Hydrating Cements by DTA Technique,” Thermo-chimica Acta (Amsterdam), V. 3, pp. 343-366.

Ramachandran, V. S. , Jan. 1973, “Action of Triethanolamine onthe Hydration of Tricalcium Aluminate,” Cement and Concrete Re-search, V. 3, No. 1, pp. 41-54.

Ramachandran, V. S., Sept. 1976, “Hydration of Cement-Roleof Triethanolamine,” Cement and Concrete Research, V. 6, No. 5,pp. 623-632.

Ramachandran, V. S., 1976a, Calcium Chloride in Concrete: Sci-ence and Technology, Applied Science Publishers, London, 216 pp.

RILEM, 1968, Proceedings, RILEM-ABEM International Sympo-sium on Admixtures for Mortar and Concrete (Brussels, 1967), RI-LEM, Paris, 6 volumes. Also, General Reports in Materials andStructures, Research and Testing (RILEM, Paris), V. 1, No . 2, pp.75-149.

Robinson, R. F., and Austin C. R., 1951, “Effect of Copper-Bearing Concrete on Molds,” Industrial and Engineering’ Chemistry,V. 43, pp. 2077-2082.

Robson, J. D., 1952, “Characteristics and Applications of Mix-tures of Portland Cement and High-Alumina Centers,” Chemistryand Industry (London), No. 1, pp. 2-7.

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

Rosskopf, P. A.; Linton, F. J.; and Peppler, R. B., 1975, “Effectof Various Accelerating Chemical Admixtures on Setting andStrength Development of Concrete,” Journal of Testing and Evalu-ation, V. 3, NO. 4, pp. 322-330.

Sarapin, E. G., 1958, “Corrosion of Wire Reinforcement in Con-crete Containing Calcium Chloride,” Promyshlennoye Stroitel’stvo(Moscow), V. 36, No. 12, pp. 21-23. (in Russian)

Schutz, Raymond J., Jan. 1959, ‘Setting Time of Concrete Con-trolled By the Use of Admixtures,” ACI JOURNAL, Proceedings,V. 55, No. 7, pp. 769-781.

Schutz, R. J., 1977, “Properties of Shotcrete Admixtures,” Shot-crete for Ground Support, SP-54, American Concrete Insti-tute/American Society of Civil Engineers, Detroit, pp. 45-58.

Schutz, R. J., May 1978, “Durability of Superplasticized Con-

CHEMICAL ADMIXTURES FOR CONCRETE 212.3R-31

crete," Proceedings, International Symposium on Superplasticizers inConcrete, Ottawa.

Shideler, J. J., Sept. 1942, “Use of Aluminum Powder to ProduceNon-Settling Concrete,” Report No. C-192, Engineering Laborato-ries, U. S. Bureau of Reclamation, Washington, D.C.

Sommer, D. R., Sept. 1978, “New Method of Making ConcreteResistant to Frost and De-icing Salts,” Betonwerk and Fertigteil-Technik (Wiesbaden), V. 44.

South African National Building Research Institute, Aug. 1957,“Interim Recommendations by the National Building Research Insti-tute to Reduce the Corrosion of Reinforcing Steel in Concrete,”Transactions, South African Institute of Civil Engineering (Johan-nesburg), V. 7, No. 8, pp. 248-250.

Stanton, T. E., 1950, “Studies of Use of Pozzolans for Counter-acting Excessive Concrete Expansion Resulting from Reaction Be-tween Aggregates and the Alkalies in Cement,” Symposium on Useof Pozzolanic Materials in Mortars and Concretes, STP-99, ASTM,Philadelphia, pp. 178-201.

Stein, H. M., and Stevels, J. M., 1974, “Influence of Silica on theHydration of Tricalcium Silicate,” Journal of Applied Chemistry, V.14, No. 8, pp. 338-346.

Taylor, Thomas G., Apr. 1948, “Effect of Carbon Black andBlack Iron Oxide on Air Content and Durability of Concrete,” ACIJOURNAL , Proceedings V. 44, No. 8, pp. 613-624.

Tenoutasse, N., 1969, “Hydration Mechanism of C3A and C3S inthe Presence of Calcium Chloride and Calcium Sulfate,” Proceed-ings, 5th International Symposium on the Chemistry of Cement, Ce-ment Association of Japan, Tokyo, V. 2, pp. 372-378.

Tynes, William O., 1977, “Investigation of Proprietary Admix-tures,” Technical Report No. C-77-1, U.S. Army Engineer Water-ways Experiment Station, Vicksburg, 38 pp.

Ulfstedt, L.; Wijard, E.; and Watesson, A. G.; 1961, “Accelerat-ing the Setting of Hydraulic Binders,” U.S. Patent No. 2,987.407.

Uppal, 1. S., and Bahadur, S. R., 1958, “Water-Proofing Cement-Sand Mortar, Part 3-Effect of Pretreatment of Sand on Permeabil-ity of Cement, Mortar and Concrete,” Indian Concrete Journal(Bombay), V. 32, p. 55.

. .

USBR, 1975, Concrete Manual, 8th Edition, U.S. Bureau of Rec-lamation, Denver, p. 627.

Verbeck, George J., 1975, “Mechanisms of Corrosion of Steel inConcrete,” Corrosion of Metals in Concrete. SP-49, American Con-crete Institute, Detroit, pp. 21-38.

Vivian, H. E., 1962, “Some Chemical Additions and Admixturesin Cement Paste and Concrete,” Proceedings, 4th InternationalSymposium on the Chemistry of Cement, (Washington, D.C., 1960),Monograph No. 43, National Bureau of Standards, Washington,D.C., pp. 909-923.

Virmani, Y. P.; Clear, K. C.; and Pasko, T. J., Jr., 1983, “Time-to-Corrosion of Reinforcing Steel in Concrete Slabs, V.5: CalciumNitrite Admixture or Epoxy-Coated Reinforcing Bars as CorrosionProtection Systems,” Report No. FHWA-RD-83-012, Federal High-way Administration, Washington, D.C., 71 pp.

Washa, G. W., and Withey, N. H., Apr. 1953, “Strength and Du-rability of Concrete Containing Chicago Fly Ash,” ACI JOUR-NAL, Proceedings V. 49, No. 8, pp. 701-712.

Whiting, D., Aug. 1981, “Rapid Determination of the ChloridePermeability of Concrete,” Report No. FHWA/RD-81/119, FederalHighway Administration, Washington, D.C., 174 pp.

Whiting, David, 1979, “Effects of High-Range Water Reducers onSome Properties of Fresh and Hardened Concrete,” Research andDevelopment Bulletin No. RD-61-01T, Portland Cement Associa-tion, Skokie, 15 pp.

Wilson, Raymond, 1927, “Tests of Colors for Portland CementMortars,” Proceedings, American Concrete Institute, V. 23, pp. 226-252.

Young, R. S., and Talbot, H. L., 1945, “Copper-Containing Ce-ments Inhibiting Footborne Diseases,” South African Mining Engi-neering Journal (Johannesburg), V. 56, pp. 475-477.

This report was submitted to letter ballot of the committee and was ap-proved in accordance with ACI balloting procedures.