REVIEW OF EXPANSIVE ALKALI-AGGREGATE REACTIONS IN CONCRETE

By J. E. GillottI

ABSTRACT: Expansive alkali-aggregate reactions (AAR) are a frequent and worldwide cause of poor durability
in Portland cement concrete. In plain concrete, deleterious AAR generally results in expansion and map-
cracking, but in reinforced or prestressed concrete cracks may show a linear relationship to reinforcement.
Expansion may lead to buckling, extrusion of joint filler, and displacement of equipment set in concrete.
White deposits of silica gel, rims surrounding aggregate particles, pop outs, loss of strength, and decrease in
modulus of elasticity may occur. Damage may be so severe that concrete has to be replaced, but less drastic
Downloaded from ascelibrary.org by UNIV OF STELLENBOSCH-PERIOD on 08/25/15. Copyright ASCE. For personal use only; all rights reserved.

remedial measures have been tried when the damage is less severe. Aggregates likely to cause expansive AAR
may be identified by petrographic methods, length-change tests, and examination of deteriorated concrete
made with similar materials and exposed to similar conditions. Potential expansion due to AAR may be
minimized by reduction of the available supply of alkalies in concrete, use of mineral and chemical admixtures,
design to limit wetting, and the beneficiation of aggregates. Classification of the types of reaction and mech-
anisms of expansion are reviewed.

INTRODUCTION toms often make it difficult to recognize the initial cause of

damage, so detailed laboratory work is required.
Durability failure of Portland cement concrete may result
from the expansion and cracking caused by reactions between SYMPTOMS LABORATORY
some types of aggregates and strong alkalies derived mainly
from the cement. Opal, chert, chalcedony, glassy volcanic Laboratory work general1y involves the examination of af-
rocks, and other natural, manufactured, or waste materials fected concrete on the binocular and petrographic micro-
containing poorly crystallized silica are liable to be alkali- scopes. Known reactive rocks and minerals and deposits of
expansive, but other rocks including some argillaceous do- silica gel may be identified and other useful information on
lomitic limestones and greywackes, phyllites, and argillites aggregates, binder, cracking, and the nature and distribution
are also involved. of alteration products (French 1991) may be obtained. Flu-
Expansion and cracking may develop within a few weeks orescence in short-wave ultraviolet light after treatment with
or months or may not appear for a number of years after uranyl acetate has been introduced as a means to supplement
concrete has been placed. The time frame required for the older optical techniques to recognize alkali-silica gel (Nate-
expansion to become excessive depends on the type of re- saiyer and Hover 1988). X-ray diffraction (XRD), scanning
active aggregate, concentration of alkalies in the pore solu- electron microscopy (SEM), energy dispersive x-ray analysis
tions, availability of water, mixture design including the pres- (EDXA), electron-probe microanalysis (EPMA), infrared
ence of chemical and mineral admixtures, curing conditions, spectroscopy (IR), image analysis, and other methods are also
temperature, and other factors related to the conditions of used.
exposure. Concrete deterioration from this cause has been Cores from an affected structure may be stored at high
found worldwide. Damage in the form of cracking and buck- humidity and at a control1ed temperature or immersed in
ling occurs in structures such as pavements and sidewalks and alkali or NaCl solution. Measurement of the length change
is of additional serious concern in mass concrete structures may indicate the amount of expansive reactivity remaining in
such as dams. the material but, as cores are free from restraint, measured
expansion may not duplicate the expansion of concrete in the
structure.
SYMPTOMS FIELD In concrete affected by AAR, strength may drop below
In plain concrete, expansive alkali-aggregate reactions ultimate; but no correlation has been found between strength
(AAR) generally result in map cracking caused by differential and expansion. However, AAR has a greater effect on tensile
stress due to variation in the amount of expansion within the than on compressive strength (Clayton 1989). The modulus
concrete element. In reinforced or prestressed concrete, cracks of elasticity generally decreases and change in this parameter
may show a linear relationship to reinforcement. Expansion may be more sensitive to expansive AAR than are strength
may cause buckling, extrusion of joint filler, and displacement changes (Ono 1989; Pleau et al. 1989). Severely damaged
of equipment set in concrete. White deposits of silica gel often concrete may show a loss of prestress and corrosion of steel;
fill voids and cracks in the concrete, rims may surround ag- but where deterioration is less severe, the bond is largely
gregate particles, and pop outs may occur. Some of these unaffected and so concrete expansion may lead to substantial
symptoms appear when concrete deterioration results from increases in the tensile stress of steel. This effect has been
causes other than AAR. Anything that impairs the properties termed "chemical stress" (Ohno et al. 1989; Inoue et aI.
of concrete reduces its ability to resist attack by other agents 1989).
of destruction and, by the time of failure, overlapping symp-
REMEDIAL MEASURES IN EXISTING CONCRETE

'Prof. Emeritus, Dept. of Civ. Engrg., Univ. of Calgary, Calgary, Remedial measures have varied depending on the degree
Alberta, Canada TIN 1N4. of distress and the function of the concrete. Damage may be
Note. Discussion open until April 1, 1996. To extend the closing date so severe that complete replacement of the concrete is the
one month, a written request must be filed with the ASCE Manager of only solution, but, when it appears that early structural failure
Journals. The manuscript for this technical note was submitted for review is less likely to occur, the future progress of the reaction
and possible publication on November 28, 1994. This technical note is
part of the Journal of Materials in Civil Engineering, Vol. 7, No.4, becomes significant. This may be estimated from the expan-
November, 1995. ©ASCE, ISSN 0899-1561195/0004-0278-0282/$2.00 + sion rate and form of the curve of length change versus time
$.25 per page. Technical Note No. 9799. obtained from strain gauges or other instrumentation at-
278/ JOURNAL OF MATERIALS IN CIVIL ENGINEERING / NOVEMBER 1995

J. Mater. Civ. Eng. 1995.7:278-282.

 tached to concrete members. If the expansion rate is declining effect may be offset because some aggregates themselves re-
and the structural effects appear small repair may be at- lease alkalies (Kawamura et al. 1989; Gillott and Rogers 1994),
tempted. Techniques that have been tried include the partial and there are external sources from which alkalies may be
replacement of concrete; installation of improved reinforce- derived.
ment including steel plate adhesion (Fujii 1992); reduction in Admixtures such as CaCl2-based accelerators and super-
the ingress of water by drainage to prevent ponding, injection plasticizers aggravate the alkali-silica reaction (ASR) whereas
of resin, or cementitious material to decrease permeability air-entraining agents, some retarders, and Li and Ba salts
and the use of waterproof surface coatings (Kojima et al. decrease expansion (Wang and Gillott 1989; Stark 1992). A
1992; Swamy and Tanikawa 1992; Ono 1989); and the in- combination of admixtures, which individually decrease ex-
stallation of expansion joints (Silveira et al. 1989) and cutting pansion due to ASR may supplement one another and cause
of transverse slots by means of a diamond wire (Thompson a further decrease in expansion, but the additive effect is only
1992). significant when the decreased expansion is due to a different
Downloaded from ascelibrary.org by UNIV OF STELLENBOSCH-PERIOD on 08/25/15. Copyright ASCE. For personal use only; all rights reserved.

mechanism in each case. For example, a greater decrease in

PREVENTING AND MINIMIZING DAMAGE the expansion of mortar bars caused by ASR was achieved
BEFORE PLACING when an air-entraining agent was used along with a pozzolan
than when either material was used separately. It was con-
Prevention is undoubtedly the best solution and AAR is cluded that the decrease caused by the pozzolan was added
best avoided by the use of nonexpansive aggregates. This to the decrease caused by the admixture, so the total decrease
requires the recognition of aggregates likely to cause a prob- was greater because two different mechanisms were involved
lem. The performance of existing concrete may provide in- (Gillott and Wang 1993). Concrete test specimens containing
sight into the behavior of new aggregates if they are the same
alkali-carbonate reactive aggregates often respond to admix-
as those used in existing structures. Care is needed, however, tures very differently from specimens containing alkali-silica
in the interpretation of field evidence. If behavior is to be reactive aggregates (Wang and Gillott 1995).
comparable to that of old concrete, the new material must Expansion may be restrained parallel to the reinforcement
be similar to that used in the existing structure. Not only must
so, in principle, cracking may be controlled. Research showed
the aggregates be the same, but other factors must be similar
that very careful detailing of secondary reinforcement or use
including the cement, particularly in regard to its alkali con-
of spiral reinforcement is needed if both axial and horizontal
tent, the mixture design, including cement content and the
expansion are to be restricted (Swamy 1989; Takemura et al.
use of pozzolans and other admixtures, and conditions of 1989).
exposure. The structures examined must be old enough for
effects resulting from AAR to become evident. This kind of
information may be unavailable due to the lack of adequate FACTORS INFLUENCING AAR
records or because construction is planned at a new or remote Certain factors influence the behavior of concrete affected
site where there has been little or no previous construction. by AAR regardless of the type of reactive rock or mineral.
For example, such sites are often selected for the development In all cases increases in expansion result from increases in
of new hydraulic structures. available alkali, water is a requirement, and higher temper-
There is now a wide range of rocks and minerals that are atures of exposure or accelerated curing increase expansion.
known to be potentially alkali-expansive and may be readily However, there are significant differences in behavior be-
identified by petrographic methods (ASTM C295). However, tween different types of materials. For example, concrete
concrete behavior often depends very much on the afore- containing opal, cristobalite, etc., often expands relatively
mentioned factors, in addition to the petrographic nature of rapidly whereas the rate of expansion of concrete made with
the aggregate and the prediction of performance often re- greywacke and related rocks is often very slow. However,
quires length change and other data. Measurements are made expansion may continue for a very long time even under
on mortar bars or concrete prisms held at a constant tem- conditions that strongly favor the active leaching of alkalies.
perature and high relative humidity or on rock cylinders im- Concrete affected by ACR shows no significant quanities of
mersed in NaOH solution [ASTM C22?, C586; Canadian silica gel. Thus, the identification of aggregate types by pet-
(1994), A23.2-14A]. Traditional tests of that sort require many rographic examination is essential but, for proper interpre-
months to a year or more. Such long delays are often unac- tation, it must be combined with a classification based on
ceptable in construction practice and much effort has been meaningful engineering performance.
made to develop more rapid tests. Mortar or concrete spec-
imens have been exposed to high humidities and high tem- CLASSIFICATION
peratures, or to increased alkali concentrations, and auto-
clave techniques have been used and chemical procedures In Canada two types of alkali-aggregate reaction are rec-
tried. These and other methods have been described and ognized: the alkali-silica reaction and the alkali-carbonate
reviewed in the literature [ASTM C1260, C289; Canadian reaction (ACR). The alkali-silica reaction is divided into two
(1994), A23.2-25A, A23.2-26A; Tang et al. 1983; Nishibay- categories depending on the types of expansive material [Ca-
ashi et al. 1986; Alasali et al. 1991; Shayan 1992; Hooton and nadian (1994a), A23.1-94, p. 112-135]: (1) Includes opal,
Rogers 1989; Grattan-Bellew 1981]. tridymite, cristobalite and natural and man-made silicate
If a new aggregate is found to contain potentially expansive glasses; and (2) includes chalcedony, microcrystalline quartz,
rocks and minerals it is best to use aggregates from a different greywackes, and related rocks. The alkali carbonate reaction
source, which are free of that type of material. If that is not includes certain varieties of argillaceous dolomitic limestone.
possible, for economic or other reasons, concrete mixture One advantage of separating aggregates into different cat-
design may be changed to minimize the risk of expansive egories is that it facilitates the recognition of potentially del-
behavior. Reduction in the overall alkali content or the use eterious behavior. Different classes of materials may be sub-
of mineral and/or chemical admixtures have been employed, jected to the same standard tests but it may be possible to
and other methods include design to limit wetting, selective use different limits in the different categories. Lack of a clas-
quarrying, and the beneficiation of gravel. The use of low sification, which recognized differences in behavior, frus-
alkali cement or mixture design to limit ,the alkali content of trated early attempts to characterize the nature of durability
the concrete is generally good practice, but the beneficial problems found in concrete made with greywacke aggregates.
JOURNAL OF MATERIALS IN CIVIL ENGINEERING / NOVEMBER 1995/279

J. Mater. Civ. Eng. 1995.7:278-282.

 Standard tests/were unable to show that such aggregates were
deleteriously alkali-expansive because limits were used, which
work well for aggregates containing opal and related minerals
but are not applicable to slowly expanding aggregates. Une-
quivocal laboratory evidence demonstrating the alkali-expan-
sivity of greywackes was obtained only when accelerated test
methods and different limits were developed. Differences in
behavior between the different types of aggregates may also
reflect differences in the mechanisms of expansion.

MECHANISMS OF EXPANSION
The mechanism of expansive ASR has been debated for
Downloaded from ascelibrary.org by UNIV OF STELLENBOSCH-PERIOD on 08/25/15. Copyright ASCE. For personal use only; all rights reserved.

more than 30 years (Powers and Steinour 1955; Diamond (a)

1989; Chatterji 1989; Wang and Gillott 1991). Attack by alkali
involves diffusion into the reactive silica of Ca, alkali, and
OH ions accompanied by proton and cation exchange, leading
to the formation of a hygroscopic alkali-silica gel. It was sug-
gested that expansion implies that composition, solution con-
centration gradient, and temperature conditions favor mois-
ture uptake because the cement paste or gel acts much like
a semipermeable membrane. More probably, swelling simply
indicates that the forces of attraction between the polar water
molecules and the alkali-silicate ions composing the gel ex-
ceed the attractive forces between the silicate groups them-
selves. Therefore, the porous and disordered three-dimen-
sional network swells and disaggregates into complex alkali-
silica and lime-alkali-silica groups surrounded by a loosely (b)
associated hydrous envelope. The expansive response of a FIG. 1. (a) Clay Inclusions in Dolomite Crystals; (b) Acid-Etched
siliceous material to alkaline attack evidently depends on its Dolomitic Limestone (Pittsburg Quarry, Kingston, Ontario, Canada)
permeability and particle size as well as on the degree of
crystallographic disorder (Pettersson 1992; Wang and Gillott exposed by the dedolomitization reaction (Gillott 1964; Swen-
1992; Brouxel 1993). son and Gillott .1967). Clay minerals occur both in the matrix
In greywackes and related rocks the silica minerals are and as inclusions in dolomite crystals (Fig. 1). Dry clay has
commonly microcrystalline, strained, and crystallographically been shown to cause large moisture expansions in sulphur
disordered and these minerals undoubtedly contribute to the concrete, which has engineering properties resembling those
mechanism of expansion (Grattan-Bellew 1992). Although of Portland cement concrete (Gillott 1986). In sulphur con-
the difference in behavior of these aggregates is no doubt crete any clay present is dry because the aggregates and fillers
largely due to the mineralogy of the silica minerals, other are heated before being cast with molten sulphur. In my opin-
factors may be involved. When the silica minerals are en- ion the evidence is convincing that the dedolomitization re-
closed in the tight fabric of a rock they may respond differ- action, the fabric of the rock, and moisture uptake are all
ently from similar minerals that are exposed as isolated frag- involved in the mechanism of expansion. A better under-
ments. In addition, phyllosilicates and other minerals were standing of the role of dedolomitization in the mechanism of
shown to react readily with alkaline solutions (Gillott 1970, ACR expansion may require the application of a method to
1986; Gillott and Swenson 1973). There is now a wealth of trace moisture movements in materials (Raichle 1994) or mea-
recent data confirming that high pH solutions at low tem- surement of the source of expansive pressure on the micro-
peratures react with silicates, the frequent products being scale.
zeolites and clay minerals (Chermak 1992), but the signifi-
cance of these reactions to expansive AAR still remains un- CONCLUSIONS
known.
The ACR differs from the other two varieties of AAR in The poor durability of Portland cement concrete due to
that the affected concrete does not contain significant quan- excessive expansion and cracking caused by AAR is now
tities of silica gel and known alkali-expansive silica minerals recognized as a frequent and worldwide problem. Deterio-
have not been found in the aggregates. There seems to be ration may set in within a few months to a number of years
no pessimum (a ratio of expansive aggregate to alkali by mass after casting.
at which maximum expansion occurs), and the response to Affected concrete often develops map-cracking, cracks with
mineral and chemical admixtures is very different (Wang and a linear relationship to reinforcement, evidence of expansion,
Gillott 1995). Details of the expansive mechanism are still white deposits of silica gel, rims surrounding particles of ag-
debated. In one proposal supported by excellent electron gregate, loss of strength, and decrease in the modulus of
micrographs, expansion is attributed to the growth and rear- elasticity.
rangement of brucite crystals in the restricted space in which Remedial measures tried included complete or partial re-
dedolomitization takes place (Tang et al. 1986; Tong 1994). placement of damaged concrete, use of additional reinforce-
Another suggestion is that expansion is due to cryptocrys- ment, protection from water, and cutting of slots or expansion
talline quartz (Katayama 1992); alkali-reactive silica is not joints.
uncommon in limestones and is a well-known cause of ex- AAR is best avoided by the use of nonexpansive aggre-
pansive AAR but gel, in significant quantities, is normally gates. Problem aggregates may be identified by petrographic
present in affected concrete. An early explanation is that methods, length-change tests, and examination of sufficiently
expansion results from the swelling pressure generated by the old exposed concrete made with similar materials and mixture
pickup of moisture by fine constituents, particularly dry clay designs.
280/ JOURNAL OF MATERIALS IN CIVIL ENGINEERING / NOVEMBER 1995

J. Mater. Civ. Eng. 1995.7:278-282.

 The use of nonexpansive aggregate may not be possible, Katayama, T (1992). "A critical review of carbonate rock reactions-
is their reactivity useful or harmful?" Proc., 9th Int. Con! on Alkali-
but expansive behavior may be minimized by reduction of Aggregate Reaction in Concrete, Concrete Soc. Pub!. CS 104, Slough,
the available alkali in concrete, use of mineral and chemical England,508-518.
admixtures, design to limit wetting, and the beneficiation of Kawamura, M., Koike, M., and Nakano, K. (1989). "Release of alkalies
aggregates. from reactive andesitic aggregates and fly ashes into pore solution in
Classification aids the recognition of expansive aggregates mortars." Proc., 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada,
and, when coupled with an understanding of the expansive S. Nishibayashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto,
mechanism, provides a reasonable framework within which Japan, 271-278.
Kojima, T, Miyagawa, T., Nakano, K., Yamaguchi, Y., and Kobayashi,
poor performance may be anticipated and effective measures S. (1992). "Effect of coating to inhibit AAR in concrete structures."
selected to minimize damage to concrete from AAR. Proc., 9th Int. Con! on Alkali-Aggregate Reaction in Concrete, Con-
crete Soc. Pub!. CS 104, Slough, England, 550-555.
APPENDIX. REFERENCES Natesaiyer, K. C., and Hover, K. C (1988). "Insitu identification of
Downloaded from ascelibrary.org by UNIV OF STELLENBOSCH-PERIOD on 08/25/15. Copyright ASCE. For personal use only; all rights reserved.

ASR products in concrete." Cement and Concrete Res., 18(3), 455-

Alasali, M. M., Malhotra, V. M., and Soles, J. A. (1991). "Performance 463.
of various test methods for assessing the potential alkali reactivity of Nishibayashi, S., Yamura, K., and Matsushita, H. (1986). "Rapid method
some Canadian aggregates." ACI Mat. J., 88(6), 613-619. of determining the alkali-aggregate reaction in concrete by autoclave."
Annual book of ASTM standards; Vol. 04.02, concrete and aggregates, Proc., 7th Int. Con! on Concrete Alkali-Aggregate Reactions, P. E.
Philadelphia, Pa. Grattan-Bellew, ed., Noyes Publications, Park Ridge, N.J., 299-303.
Brouxel, M. (1993). "The alkali-aggregate reaction rim: Na 2 0, Si0 2 , Ohno, S., Yoshioka, Y., Shinozaki, Y., and Morikawa, T (1989). "The
K20 and CaO chemical distribution." Cement and Concrete Res., 23(2), mechanical behaviour of beams coated after alkali silica reaction dam-
309-320. age." Proc., 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada,
Canadian Standards Association. (l994a). "Concrete materials and S. Nishibayashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto,
methods of concrete construction." A23.1-94, Rexdale, Ont., Canada, Japan, 697-702.
1-148. Ono, K. (1989). "Assessment and repair of damaged concrete structure. "
Canadian Standards Association. (1994b). "Methods of test for con- Proc., 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada, S. Nish-
crete." A23.2-94, Rexdale, Ont., Canada, 149-350. ibayashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto, Japan,
Chatterji, S. (1989). "Mechanisms of alkali-silica reaction and expan- 647-658.
sion." Proc., 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada, Pettersson, K. (1992). "Effects of silica fume on alkali-silica expansion
S. Nishibayashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto, in mortar specimens." Cement and Concrete Res., 22(1), 15-22.
Japan, 101-105. Pleau, R., Berube, M. A., Pigeon, M., Fournier, B., and Raphael, S.
Chermak, J. A. (1992). "Low temperature experimental investigation (1989). "Mechanical behaviour of concrete affected by ASR." Proc.,
of the effect of high pH NaOH solutions of the Opalinus shale, Switz- 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada, S. Nishibay-
erland." Clays and Clay Minerals, 40(6), 650-658. ashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto, Japan, 721-
Clayton, N. (1989). "Structural performance of ASR affected concrete." 726.
Proc., 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada, S. Nish- Powers, T. C, and Steinour, H. H. (1995a). "An interpretation of some
ibayashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto, Japan, published researches on the alkali-aggregate reaction. Part I: The
671-676. chemical reaction and mechanism of expansion." J. Am. Concrete
Diamond, S. (1989). "ASR-another look at mechanisms." Proc., 8th Inst., 26(6), 497-516.
Int. Con! on Alkali-Aggregate Reaction, K. Okada, S. Nishibayashi, Powers, T C, and Steinour, H. H. (1995b). "An interpretation of some
and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto, Japan, 83-94. published researches on the alkali-aggregate reaction. Part 2: A hy-
French, W. J. (1991). "Concrete petrography: a review." Quarterly J. pothesis concerning safe and unsafe reactions with reactive silica in
Engrg. Geology, 24(1), 17-48. concrete." J. Am. Concrete Inst., 26(8), 785-811.
Fujii, M. (1992). "Reinforcement of AAR damaged concrete beam." Raichle, M. (1994). "Visualizing the mind." Scientific Am., 270(4), 58-
Proc., 9th Int. Con! on Alkali-Aggregate Reaction in Concrete, Con- 64.
crete Soc. Publ. CS 104, Slough, England, 353-360. Shayan, A. (1992). "Prediction of alkali reactivity potential of some
Gillott, J. E. (1964). "Mechanism and kinetics of expansion in the alkali- Australian aggregates and correlation with service performance." ACI
carbonate rock reaction." Can. J. Earth Sci., Ottawa, Ont., Canada, Mat. J., 89(1), 13-23.
1(2), 121-145. Silveira, J. F. A., Degaspare, J. C, and Cavalcanti, A. J. C T (1989).
Gillott, J. E. (1970). "Examination of rock surfaces with the scanning "The opening of expansion joints at the Moxoto powerhouse to coun-
electron microscope." J. Microscopy, 91(1970),203-205. teract the alkali-silica reaction." Proc., 8th Int. Con! on Alkali-Ag-
Gillott, J. E. (1986a). "Some clay-related problems in engineering ge- gregate Reaction, K. Okada, S. Nishibayashi, and M. Kawamura, eds.,
ology in North America." Clay Minerals, 21(3), 261-278. Soc. of Mat. Sci., Kyoto, Japan, 747-751.
Gillott, J. E. (l986b). "Alkali-reactivity problems with emphasis on Ca- Stark, D. C (1992). "Lithium salt admixtures-an alternative method
nadian aggregates." Engrg. Geology, 23(1), 29-43. to prevent expansive alkali-silica reactivity." Proc., 9th Int. Con! on
Gillott, J. E., and Rogers, C A. (1994). "Alkali-aggregate reaction and Alkali-Aggregate Reaction in Concrete, Concrete Soc., Slough, En-
internal release of alkalies." Mag. Concrete Res., 46(167), 99-112. gland, 1017 -1025.
Gillott, J. E., and Swenson, E. G. (1973). "Some unusual alkali-expan- Swamy, R. N. (1989). "Structural implications of alkali silica reaction."
sive aggregates." Engrg. Geology, Vol. 7,181-195. Proc., 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada, S. Nish-
Gillott, J. E., and Wang, H. (1993). "Improved control of alkali-silica ibayashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto, Japan,
reaction by combined use of admixtures." Cement and Concrete Res., 683-689.
23(4), 973-980. Swamy, R. N., and Tanikawa, S. (1992). "Acrylic rubber coating to
Grattan-Bellew, P. E. (1981). "A review of test methods for alkali ex- control alkali silica reactivity." Proc., 9th Int. Con! on Alkali-Aggre-
pansivity of concrete aggregates." Proc., 5th Int. Con! on Alkali- gate Reaction in Concrete, Concrete Soc. Pub!. CS 104, Slough, En-
Aggregate Reaction in Concrete; Paper S252/9, Nat. Build. Res. Inst., gland, 1026-1034.
Pretoria, South Africa, 303-314. Swenson, E. G., and Gillott, J. E. (1967). "Alkali reactivity of dolomitic
Grattan-Bellew, P. E. (1992). "Microcrystalline quartz, undulatory ex- limestone aggregate." Mag. Concrete Res., 19(59),95-104.
tinction and the alkali-silica reaction." Proc., 9th Int. Con! on Alkali- Takemura, K., Tazawa, E.-I., Yonekura, A., and Abe, Y. (1989). "Me-
Aggregate Reaction in Concrete, Concrete Soc. Publ. CS 104, Slough, chanical characteristics of reinforced concrete column affected by alkali
England, 383-394. aggregate reaction." Proc., 8th Int. Con! on Alkali-Aggregate Reac-
Hooton, R. D., and Rogers, C A. (1989). "Evaluation of rapid test tion, K. Okada, S. Nishibayashi, and M. Kawamura, eds., Soc. of
methods for detecting alkali-reactive aggregates." Proc., 8th Int. Con! Mat. Sci., Kyoto, Japan, 665-670.
on Alkali-Aggregate Reaction, K. Okada, S. Nishibayashi, and M. Tang, M., Han, S.-F., and Zhen, S.-H. (1983). "A rapid method for
Kawamura, eds., Soc. of Mat. Sci., Kyoto, Japan, 439-444. identification of alkali reactivity of aggregate." Cement and Concrete
Inoue, S., Fujii, M., Kobayashi, K., and Nakano, K.-I. (1989). "Struc- Res., 13(3),417-422.
tural behaviours of reinforced concrete beams affected by alkali-silica Tang, M.-S., Liu, Z., and Han, S.-F. (1986). "Mechanism of alkali
reaction." Proc., 8th Int. Con! on Alkali-Aggregate Reaction, K. Okada, carbonate reaction." Proc., 7th Int. Con! on Concrete Alkali-Aggre-
S. Nishibayashi, and M. Kawamura, eds., Soc. of Mat. Sci., Kyoto, gate Reactions, P. E. Grant-Bellew, ed., Noyes Publication, Park Ridge,
Japan, 727- 732. N.J., 275-279.

JOURNAL OF MATERIALS IN CIVIL ENGINEERING / NOVEMBER 1995/281

J. Mater. Civ. Eng. 1995.7:278-282.

 Thompson, G. A. (1992). "Alkali aggregate reactivity remedial measures Mat. Sci., Kyoto, Japan, 187-192.
Mactaquac intake structure." Proc., Workshop on Alkali-Aggregate Wang, H., and Gillott, J. E. (1991). "Mechanism of alkali-silica reaction
Reactions in Concrete: Occurrences, Testing and Control. J. A. Soles, and the significance of calcium hydroxide." Cement and Concrete Res.,
ed., Energy Mines and Resour. (EMR), Ottawa, Canada. 21(4),647-654.
Tong, L. (1994). "Alkali-carbonate rock reaction," PhD thesis, Dept. Wang, H., and Gillott, J. E. (1992). "Competitive nature of alkali-silica
of Mat. Sci. and Engrg., Nanjing Inst. of Chemical Techno!., Nanjing, fume and alkali-aggregate (silica) reaction." Mag. Concrete Res. , 44( 161 ),
ROC. 235-239.
Wang, H., and Gillott, J. E. (1989). 'The effect of superplasticizers on Wang, H., and Gillott, J. E. (1995). "Alkali-carbonate reaction-sig-
alkali-silica reactivity." Proc., 8th Int. Conf. on Alkali-Aggregate Re- nificance of chemical and mineral admixtures." Mag. Concrete Res.,
action, K. Okada, S. Nishibayashi, and M. Kawamura, eds., Soc. of 47(170),69-75.
Downloaded from ascelibrary.org by UNIV OF STELLENBOSCH-PERIOD on 08/25/15. Copyright ASCE. For personal use only; all rights reserved.

282/ JOURNAL OF MATERIALS IN CIVIL ENGINEERING / NOVEMBER 1995

J. Mater. Civ. Eng. 1995.7:278-282.