TF6912_CH06 alkali silica reaction on concrete chapter 06

6 Alkali-silica reaction—Danish experienceS.CHATTERJEE, Z.FÖRDÖS and N.THAULOW

Abstract

In Denmark research on alkali-silica reactivity and preventive measures hasbeen carried out in two distinct phases. In the first phase, flint and opalinelimestone were identified as the main reactive components in Danish aggregates.Suggested preventive measures were the use of low-alkali cement and/or theuse of aggregates containing less than 2% reactive components. At this stagethe basic assumptions were that the alkali content of concrete is determined bythat of its cement and the alkalis are evenly distributed.

The second phase of research, on the other hand, assumed that a real-life structuremay acquire extra alkalis from its environment and there is always a concentrationdistribution of alkalis in it. An accelerated test method of storing mortar prisms in asaturated NaCl solution at 50°C was devised. New and detailed mechanisms havebeen proposed for alkali-silica reaction and the associated expansion.

As regards the preventive measures, the following innovations have beenmade:

(1) The assumptions of a constant and uniform distribution of alkali in aconcrete structure have been dropped.

(2) The environment to which a structure may be exposed has been classifiedas regards its severity.

(3) The acceptance criteria of aggregates, both fine and coarse, have beenmade very strict, and have been collated with the environmental classes.

(4) The permissible total alkali content of a concrete mix has been collatedwith the environment to which the concrete would be exposed.

(5) A fly ash Portland cement has been introduced to the Danish market.

6.1 Historical introduction

As far as alkali-silica reaction (ASR) is concerned, Denmark is in anunfortunate situation; most of the aggregate sources contain reactive

Copyright 1992 Blackie and Son Ltd

components in one or another size fraction, i.e. there is always a risk of ASR. Itis hardly surprising that the first description of the effects which could beattributed to ASR was reported by Poulsen in 19141, long before Stanton’swork in the 1940s. However, systematic research on ASR did not start untilNerenst’s report in 19512. The research on ASR in Denmark can be broadlydivided into two phases-that carried out before the 1970s and that afterwards-with fairly clear differences in their research philosophies.

6.1.1 First phase

The earlier research was carried out under the auspices of the Danish Committee onAlkali Reaction in Concrete. The results of this extensive programme were publishedin 23 reports3–25. In this programme basically the American methodologies were usedfor the detection of alkali-silica reactivity of aggregates26,27. Subsequently, petrographicand X-ray diffraction techniques were employed both to characterise the reactiveaggregate types and to detect ASR in concrete structures. Flint and opaline limestonewere identified as the most prevalent reactive aggregate types in Denmark. The reactivecomponents were found in both sand and gravel fractions of many aggregate sources.One of the important results of this programme was verification of the concept of‘pessimum proportioning’. This concept was based on the assumption of a fixed alkalicontent of a mix composition and the observation that in mortar bar tests expansionmaxima occur within a rather narrow range of alkali to reactive silica ratio. It wasconsidered desirable to avoid the condition of maximum expansion. Anothercontribution was the classification of flint types, relating their porosity and reactivity.It was also inferred that, if the total flint content of a sand type is less than 2%, it is lessliable to cause alkali-silica expansion. The practical implication of this investigationwas that the risk of alkali-silica expansion could be avoided by using either a flint-freeaggregate or a low-alkali Portland cement or a pozzolan. The research was summarisedby Bredsdorff et al.28. The culmination of the research was Idorn’s dissertation in196729 in which a systematic approach to the assessment of field concrete deteriorationwas presented.

Implicit assumptions of this phase of research were that the alkali content ofa concrete structure is determined by the alkali content of the cement andremains constant and uniformly distributed throughout its life. Moreover,Powers and Steinour’s hypothesis of ASR, which is consistent with the aboveassumptions, formed the theoretical basis of this phase30.

6.1.2 Second phase

The second phase of alkali-silica research started with an investigation ofconcrete roads which had deteriorated within 4 years of their inauguration.During this investigation, it was observed that extensive ASR had occurred in


the concrete roads. Since non-reactive granite was used as coarse aggregates andthe cement contained about 0.6% Na2O eq. alkali, it was inferred that theextensive reaction was a result of an interaction between the de-icing salt NaCland reactive silica present in the sand, i.e. NaCl acted as an accelerator. Thisinference was then verified by laboratory experiments31. The above roadinvestigation also revealed that the cement-stabilised sub-base of the roads(containing about 6% cement) though it used the same reactive sand as theroads, suffered no ASR. A subsidiary investigation indicated that this absence ofASR in the sub-base may be because of absence of free Ca(OH)2 in the subbasematerial32. At about the same time, the electron probe microanalytical techniquewas applied to determine the chemical composition of alkali-silica gel33. A highcontent of CaO in the gels analysed, up to 20%, was found. The gel wasanalysed in situ within the cracks at known and increasing distances from theperimeter of the original reacted particle. The CaO content of the gel increasedwith increasing distances. This paper, which remained somewhat dormant for atime, played a very important role in the elucidation of the ASR mechanism.

The explicit assumptions of this second phase of research were that the real-life structures often receive alkalis from outside sources, e.g. from thegroundwater, from de-icing salt, etc., and that there is nearly always aconcentration gradient of alkalis in any structure.

Most of the subsequent research in Denmark could be treated as a followupto the above-mentioned three papers. The subsequent research could bedivided into the following subgroups: (1) That directed to the elucidation of mechanisms of ASR and attendant

expansion.(2) That directed to the development of methods for the measurement of the

expansion capacity of reactive aggregates.(3) That directed towards the development of preventive methods. Beside these systematic studies, there were some isolated studies of the effectsof ASR on reinforced concrete beams, life expectancy of a structure sufferingfrom ASR reaction, occurrence of ASR in swimming pools, etc.

6.2 Fundamental studies

6.2.1 Mechanisms of ASR and attendant expansion

The first investigation to be carried out was on the role of Ca(OH)2 in theexpansive ASR34. From Ref. 32, one may reason that expansive ASR may beavoided, even when cement content is high, by using a cement which does notform crystalline Ca(OH)2 during hydration, e.g. high-slag Portland cement. Theresults of this investigation confirmed the above expectation of no expansion


in prisms made with slag Portland cement. The X-ray diffraction diagram of theprism made with slag Portland cement and stored in an NaCl bath showed thatprisms contained no Ca(OH)2 but had substantial amounts of Friedel’s salt,indicating that NaOH had formed in those prisms by the following reaction:

4CaO+Al2O3+2NaCl+13H2O=3CaO·Al2O3·CaCl2·12H2O+2NaOH A point to note is that no expansion occurred in spite of this extra NaOHformation. This confirmed the importance of crystalline Ca(OH)2 in theexpansive ASR.

The next investigation addressed the mechanisms by which NaCl andCa(OH)2 accelerate the alkali-silica expansion35. In that investigation, prismsmade from reactive sand and Portland cement containing varying amounts ofdiatomites were exposed to saturated NaCl solution at 50°C. The expansion ofthe prisms was measured up until about 1 year, and at the end of that period theprisms were further examined by petrographic, X-ray diffraction and electronprobe microanalytical techniques. Figure 6.1 shows that the alkali-silicaexpansion decreases with the increase in the diatomite content in the cement.The X-ray diffraction diagrams showed that crystalline Ca(OH)2 contentdecreased with the increase in diatomite in cement.

The most interesting information came from the electron probemicroanalyses of reactive grains and their surrounding pastes. Normalisedmajor oxide analyses of reacted grains and their surrounding areas are shownin Figures 6.2 to 6.5. The figures show that (i) the entry of CaO and Na2O intoreacted grains decreases with the increase in diatomite content in cement, (ii)more CaO and Na2O enter the grains when the prisms are stored in NaClsolution than when stored in water, (iii) CaO enters reactive grains irrespectiveof the nature of the bath, (iv) in samples free of diatomite, SiO2 concentration

Figure 6.1 Expansion characteristics of mortar prisms. Figures indicate moler contents.


Figure 6.2 The distributions of oxides in and around a reactive grain. Mortar prism stored inwater bath. No moler in cement.

Figure 6.3 The distribution of oxides in and around a reactive grain. Mortar prism stored in anNaCl bath. No moler in cement.


Figure 6.4 The distribution of oxides in and around a reactive grain. Mortar prism stored inwater bath. Cement contains 25% moler.

Figure 6.5 The distribution of oxides in and around a reactive grain. Mortar prism stored in anNaCl bath. Cement contains 25% moler.


drops sharply at the grain boundary, whereas in samples containing diatomiteSiO2 concentration decreases slowly. A detailed analysis of the microanalyticaldata indicated that some SiO2 diffused out of the reactive grains, and theamount of SiO2 that diffused out increased with decreasing Ca(OH)2 content inthe paste. It is obvious that calcium and sodium must have entered the grainsas hydrated ions and have been accompanied by OH– ions so as to maintainelectroneutrality.

From the results, it can be seen that expansion of mortar prisms, extent ofreaction and penetration of hydrated Na+, Ca2+ and OH– decreased with thedecrease in crystalline Ca(OH)2 in the paste. However, the diffusion of SiO2 outof a reactive grain increased with the decreasing Ca(OH)2 content.

The observed high Na2O content of reacted grains of prisms stored in anNaCl bath indicates that a reaction of the type shown in Figure 6.6 occursduring ASR. Availability of both NaCl and Ca(OH)2 near a reactive grain is thecontrolling factor in the penetration of hydrated Na+ and OH–. The abovereaction explains how NaCl and Ca(OH)2 accelerate ASR.

In ASR, Ca(OH)2 has at least four roles: it accelerates the penetrations ofhydrated Ca2+, Na+, K+ and OH– in a reactive grain; it promotes the reactionshown in Figure 6.6; it hampers the diffusion of SiO2 out of a reactive grain;and the presence of solid Ca(OH)2 acts as a buffer to maintain a high OH–

concentration.In a cement paste environment, expansion will occur if the rate of

penetration of matter in a reactive grain exceeds that of the SiO2 diffusingout; the process is shown schematically in Figure 6.7. If the alkali content islow, only a limited amount of hydrated Ca2+ and OH– can penetrate a grain;this is because of the large size of hydrated Ca2+. If the alkali concentrationis high, smaller hydrated Na+, K+ and OH– will be able to penetrate a reactivegrain unhindered. This penetration of Na+, K+ and OH– will cause a break-down of Si–O–Si bonds, thereby opening the grain for further penetration ofCa2+, Na+, K+ and OH–. In the presence of excess Ca(OH)2 and high alkali

Figure 6.6 A hypothetical reaction.


concentration, only a limited amount of SiO2 can diffuse out, but morematerials are pumped in. This generates the pressure necessary for expansion.Note that according to this hypothesis, the chemical reaction andaccompanied expansion are not directly related.

The above roles of Ca(OH)2 are quite different from that assigned to it in thehypothesis of Powers and Steinour30. According to the Powers and Steinourhypothesis the expanding grain should be low in CaO, whereas according tothe newly proposed hypothesis the expanding grain should containsubstantial amounts of CaO. The near universally high content of CaO in gelscan be seen in Figure 6.836.

Figure 6.7 A model for the alkali-silica reaction. If the amounts of water, alkali hydroxide andcalcium hydroxide entering the reactive particle are larger than the amount of alkali-silica gel

seeping out, the particle expands and cracks the surrounding cement paste.

Figure 6.8 Composition of alkali-silica gel in concrete. Data published in Ref. 20 andmicroanalyses in Ref. 36.


6.2.2 Testing of reaction mechanisms

From the newly proposed hypothesis, it follows that all alkali metal salts arecapable of accelerating alkali-silica expansion, and for a given alkali metalsalt and reactive aggregate at a given temperature the expansion will decreasewith decreasing salt concentration. A recent investigation shows that the aboveexpectations are valid37. Figures 6.9 and 6.10 show the results of thisinvestigation. Figure 6.9 shows that all alkali salts accelerate expansion. Alkalihydroxide, which depresses the solubility of Ca(OH)2, causes less expansionthan other salts. Figure 6.10 shows that expansion decreases with decreasingKCl concentration. Figure 6.10 shows that a 0.5 N KCl solution, i.e. a ca 3%solution, does not cause any expansion; this is consistent with the result of a10-year submersion of concrete prisms in seawater38.

The reaction shown in Figure 6.6 constitutes another test for the newlyproposed hypothesis. The points to note are that the above reaction shouldoccur in a reactive sand–Ca(OH)2–KCl solution system and during reaction theliquid phase is enriched with CaCl2. This newly formed CaCl2 will in turndepress the solubility of Ca(OH)2 in the liquid phase. High reactivity of a sandcreates a high concentration of CaCl2 and a low concentration of Ca(OH)2 inthe liquid phase. This inference has been tested recently39. Table 6.1 shows theresults of this investigation. From Table 6.1, it can be seen that the suspensionscontaining more reactive sand types have lower OH– concentrations.

From the results of Figure 6.10, it is possible to get an order of magnitudeestimation of the maximum tolerable alkali content. Figure 6.10 shows that inthe case of a highly expansive sand a 0.5 N alkali salt solution does not giverise to any expansion, but a 1 N solution does. One may consider that the

Figure 6.9 Expansion characteristics of mortar bars stored in different sodium salt solutions.


maximum tolerable alkali salt solution will be around 0.75 N. After some time,the concentration in the pore liquid and in the salt bath will be the same. Anestimate of the amount of alkali needed to trigger expansive reaction may beestimated from the following considerations. Consider a concrete mixcontaining 350 kg cement per cubic metre and a water-cement (w/c) ratio of0.5, i.e. an initial water content of 175 litres. Further assume that, afterhardening, the free water content is 85 litres. The maximum tolerable alkalicontent can be calculated as:

For a less expansive sand, more Na2O could be tolerated. In the above

Figure 6.10 Expansion characteristics of mortar bars stored in KCl solutions of varyingnormality.

Table 6. 1 The OH– concentration of the filtrates in milligrams per litreof solution for sand tested by a simple chemical method39.

* Have some difficulties during filtration.


estimation, it has been assumed that no alkali salts may migrate from anoutside source and that alkali salt is uniformly distributed throughout the massof the concrete mix. If alkali salts migrate from an outside source or localevaporation takes place at certain parts, then the local concentration of alkalisalt may go above the critical value and give rise to alkali-silica expansioneven though the overall alkali content may be lower than the critical value.

Various practical implications of this newly proposed hypothesis have beendiscussed in a paper40.

6.3 Prevention of ASR

The development of ASR and cracking of concrete depends primarily on thefollowing parameters:

The alkali content of the concreteThe type and content of the reactive silica in the aggregateThe quality of the concrete (strength, density, air content)The environmental condition (temperature, humidity, salt).

6.3.1 Alkalis in Danish cement

The alkalis in the concrete normally originate from the cement, but alkalis canalso be transported from external sources, for example from de-icing salts.However, the total alkali content of the concrete depends, on most occasions,on the alkali content of the cement and the cement content of the concrete.

The alkali content of the most commonly used Danish types of cement variesbetween 0.6% and 0.8% by weight of cement, and that is generally expressed asequivalent acid-soluble sodium oxide (Na2O+0.658 K2O) (Table 6.2).

From Table 6.2 it can be seen that the acid-soluble alkali content of thePortland cement is close to 0.6% Na2O eq. According to the American StandardASTM C 150-72, a Portland cement with an alkali content of less than 0.6% isdefined as a low-alkali cement. In Denmark two special cements with lowalkali contents are produced:

Low-alkali sulphate-resisting Portland cement (less than 0.4% alkali)White Portland cement (less than 0.2% alkali).

6.3.2 Danish Code of Practice for the prevention of ASR

The Danish Code of Practice for the structural use of conctrete41 requires thatthe cement should be Portland cement or Portland fly ash cement complying


with the requirement of DS 427, Portland cement and Portland fly ash cement,2nd Edn, April 1983. In DS 427 there is no regulation of the alkali content ofthe cement. However, in May 1986, a new Code of Practice was issued. This isthe so-called BBB (In Danish: Basis Betonbeskrivelsen forBygningskonstruktioner)42. This is a fundamental concrete specification to beused in all public building construction works. A system is set out to preventharmful alkali-silica reaction.

The parameters used in this system are as follows:

Alkali content of the concreteReactivity of the aggregateEnvironmental conditions.

6.3.2.1 Cement and its requirements. The cement falls into one of four groupsof alkali content:

EA extra low-alkali � 0.4% eq. Na2OLA low-alkali � 0.6% eq. Na2OMA medium-alkali � 0.8% eq. Na2OHA high-alkali > 0.8% eq. Na2O.

The alkali contributions of the cement and all other alkali sources in the mixare summed up as eq. Na2O kg/m3 of concrete.

Table 6.2 Danish types of cement: chemical composition.

1, Standard cement (Portland fly ash cement); 2, Rapid cement; 3,Lowalkali sulphate-resisting Portland cement; 4, White cement.


6.3.2.2 Aggregate classification. The aggregates are classified into threegroups:

Class P for use in passive environmentsClass M for use in moderate environmentsClass A for use in aggressive environments.

The classification of the sand with regard to alkali-silica reactivecomponent (porous flint) is shown in Table 6.3. The methods used are thepetrographic thin-section point-count method TI-B 5243 and the mortar barexpansion test in saturated sodium chloride solution TI-B 5144. The TI-B51 method has recently been accepted as a common Nordic test method:Nordtest Build NT 29545.

For the coarse aggregate (Table 6.4), amount of reactive aggregate is limitedby the allowable amount of particles with a density below 2400 kg/m3.Furthermore, there is a limit on the absorption of the flint with density largerthan 2400 kg/m3. This value is determined on 10% of the flint with porous crust.

6.3.2.3 Environment classification. The environmental classes are set out in DS41141. The requirements refer to three environmental classes, characterised bydifferent degrees of aggressiveness commonly found in Denmark. (1) Aggressive environmental class. Comprises environment containing salt or

flue gases, seawater or brackish water.(2) Moderate environmental class. Comprises moist, unaggressive outdoor

and indoor environment, and flowing or standing fresh water.(3) Passive environmental class. Comprises dry, unaggressive environment,

i.e. particularly an indoor climate. 6.3.2.4 Requirements for the fresh concrete from BBB. The combination ofrequirements for the aggregate, the cement and the concrete as a function ofthe environmental conditions from BBB42 are given in Table 6.5.

Table 6.3 Classification of sand.


6.3.2.5 Requirements for hardened concrete. The requirements for thehardened concrete after BBB42 are found in Table 6.6.

6.3.3 Experimental results with fly ash and fly ash cement

The effect of fly ash on the expansion due to ASR of mortar or concretesamples can be evaluated either by the use of closed systems in which the totalalkali content of the systems does not change with time, which is the casewhen testing is carried out according to, for example, ASTM C441 or themodified ASTM method or by the use of open systems in which samples are incontact with an alkali salt solution, as is the case with the so-called NaCl bathmethod. The evaluation can also be carried out by an accelerated test methodor by a long-time storage method46–49. In Denmark both open and closedsystems as well as accelerated and long-time storage methods have been usedto evaluate fly ash or Portland fly ash cement systems.

In normal circumstances (closed systems), the use of a Danish Portlandcement, which has an alkali content of less than 0.8%, in conjunction with areactive aggregate does not give rise to any destructive ASR. However, in anopen system, in which alkali salts from an outside source can migrate into astructure, destructive ASR can develop very quickly.

It is common practice (in Denmark) to use either a low-alkali Portlandcement or a pozzolanic cement (Portland fly ash cement) as a safety factoragainst ASR.

In recent years papers50,51 have been published by the Nordic ConcreteResearch, from which it appears that: (1) Long-time storage tests show that addition of fly ash to concrete

mixes has no adverse effect even if its alkali content is as high as2.34% Na2O eq.

(2) Accelerated tests according to ASTM show that the addition of fly ash,even of high alkali content, to Portland cements reduces expansions dueto ASR.

(3) Accelerated tests with unlimited supplies of alkali salt show that theaddition of fly ash to Portland cement reduces, at least, the rate of

Table 6.4 Classification of coarse aggregate.


Table 6.5 BBB requirements for fresh concrete.

*Silica fume is included in calculation of the water-cement ratio by anactivity factor of 2.0, fly ash by 0.5.

†Class P sand is allowed in moderate environments if type EA or LAcement is used and the total equivalent Na2O is less than 1.8 kg/m3 concrete.

‡In moderate and aggressive environments HA cement (high-alkali) may beused and the requirement to a maximum Na2O content omitted provided thesand is class A with the further requirements:

Max. 1.0% reactive aggregateMax. 0.1 % expansion of mortar bars after 20 weeks.

and the coarse aggregate is type A with the further requirement:Max. 1.0% aggregate under 2500 kg/m3 density.§ MS should not be used together with Portland fly ash cement in

moderate and aggressive environments.** The project may specify less than 1.0% aggregate under 2200 kg/m3 in

order to reduce the number of pop-outs.†† It is not allowed to add chloride to reinforced concrete. For

unreinforced concrete the limit is 1.5%.‡‡ The kitmassevolumen is calculated as the total concrete volume minus

the volume of aggregates. The requirement may be omitted for structuralelements not exposed to frost.

§§ For exposed aggregate finish, the requirement of air content may be omitted.C, cement; PFA, fly ash; MS, silica fume, microsilica.


expansion and perhaps also the ultimate expansion. Therefore, the use ofthe Portland fly ash cement reduces the risk of expansion due to ASR evenin an open system in which alkali salts from an outside source maypenetrate into a concrete structure. For the case of a moderately reactivesand, this reduction in expansion by using Portland fly ash cement maymake the sand acceptable for concrete-making.

(4) In an open system the use of a low-alkali sulphate-resistant Portlandcement has very little advantage over an ordinary Portland cement. Themost promising results are obtained with the use of Portland fly ashcement (Figures 6.11 to 6.13).

Figure 6.11 Expansion characteristics of sand-cement mortar bars stored in a saturated NaClbath at 50°C (ordinary Portland cement compared with low-alkali sulphate-resistant cement

and with Portland fly ash cement (Danish standard cement).

Table 6.6 Requirements for hardened concrete.


Figure 6.12 Alkali-silica test. Expansion characteristics of mortar bars of Nymølle sand togetherwith low-alkali cement and Portland fly ash cement.

Figure 6.13 Alkali-silica test. Expansion characteristics of mortar bars of Kallerup sand togetherwith low-alkali cement and Portland fly ash cement.


6.4. Conclusions

Denmark is a small country with many of its aggregate sources containingalkali-reactive components. This has led to extensive research programmes onASR, on both its fundamental aspect and its avoidance in concrete structures.As regards the preventive measures, the following innovations have beenmade: (1) The assumptions of a constant and uniform distribution of alkali in a

concrete structure have been dropped.(2) The environment to which a structure may be exposed has been classified

as regards its severity.(3) The acceptance criteria of aggregates, both fine and coarse, have been

made very strict, and have been correlated with the environmental classes.(4) The permissible total alkali content of a concrete mix has been correlated

with the environment to which the concrete would be exposed.(5) A fly ash Portland cement has been introduced to the Danish market. It is hoped that as a result of the above innovations the incidence of ASR willbe very much restricted in the future.

References

1 . Poulsen, A. (1914) Betons Holdbarhed (lecture in the Institution of Danish Civil Engineers,21 January 1914). Ingeniøren, 31, 293–300 (discussion pp. 300–304, in Danish).

2 . Nerenst, P. (1952) Betonteknologiske Studier i USA—Rapport over ECA Studierejse,14 November 1950 to 15 February 1951 (in Danish). The Danish National Institute ofBuilding Research, Copenhagen, Series SBI Studie No. 7.

3 . Nerenst, P. (1957) Alment om Alkali Reaktioner i Beton (with an English summary). TheDanish National Instutute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete, Progress Report Al, Copenhagen.

4 . Idorn, G.M. (1958) Concrete on the West Coast of Jutland, Part I. The DanishNational Institute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete, Progress Report B1, Copenhagen.

5 . Idorn, G.M. (1958) Concrete on the West Coast of Jutland, Part II. The DanishNational Institute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete, Progress Report B2, Copenhagen.

6 . Jeppesen, A. (1958) Durability and Maintenance of Concrete Structures on Danish Railways.The Danish National Institute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete, Progress Report B3, Copenhagen.

7 . Tovborg Jensen, A., Wøhlk, C.J., Drenck, K. and Krogh Andersen, E. (1957) AClassification of Danish Flints etc. Based on X-Ray Diffractometry. The DanishNational Institute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete, Progress Report D1, Copenhagen.

8 . Grey, H. and Søndergaard, B. (1958) Flintforekomster i Danmark (The Occurring ofFlint in Denmark. With an English Summary). The Danish National Institute ofBuilding Research and the Academy of Technical Sciences, Committee on AlkaliReactions in Concrete, Progress Report D2, Copenhagen.


9. Søndergaard, B. (1959) Petrografisk Undersøgelse af Daske Kvartœre Grusaflejringer.(Petrographic Investigation of Quaternary Danish Gravel Deposits. With an English Summary).The Danish National Institute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete. Progress Report El, Copenhagen.

10. Haulund Christensen, K.E. (1958) Evaluation of Alkali Reactions in Concrete by the ChemicalTest. The Danish National Institute of Building Research and the Academy of TechnicalSciences, Committee on Alkali Reactions in Concrete. Progress Report H1, Copenhagen.

11. Jones, F.E. (1958) Investigations of Danish Aggregates at Building Research Station.The Danish National Institute of Building Research and the Academy of TechnicalSciences, Committee on Alkali Reactions in Concrete, Progress Report II, Copenhagen.

12. Bredsdorff, P., Poulsen, E. and Spøhr, H. (1966) Experiments on Mortar BarsPrepared with Selected Danish Aggregates. The Danish National Institute of BuildingResearch and the Academy of Technical Sciences, Committee on Alkali Reactions inConcrete, Progress Report 12, Copenhagen.

13. Bredsdorff, P., Poulsen, E. and Spøhr, H. (1967) Experiments on Mortar BarsPrepared with a Representative Sample of Danish Aggregates. The Danish NationalInstitute of Building Research and the Academy of Technical Sciences, Committee onAlkali Reactions in Concrete, Progress Report 13, Copenhagen.

14. Efsen, A. and Glarbo, O. (1960) Experiments on Concrete Bars. Expansions DuringStorage in Climate Room. The Danish National Institute of Building Research and theAcademy of Technical Sciences, Committee on Alkali Reactions in Concrete, ProgressReport K1, Copenhagen.

15. Trudsø, E. (1958) Experiments on Concrete Bars. Freezing and Thawing Tests. TheDanish National Institute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete, Progress Report K2, Copenhagen.

16. Andreasen, A.H.M. and Haulund Christensen, K.E. (1957) Investigation of the Effectof Some Pozzolans on Alkali Reactions in Concrete. The Danish National Institute ofBuilding Research and the Academy of Technical Sciences, Committee on AlkaliReactions in Concrete, Progress Report L1, Copenhagen.

17. Poulsen, E. (1958) Preparation of Samples for Microscopic Investigation. The DanishNational Institute of Building Research and the Academy of Technical Sciences,Committee on Alkali Reactions in Concrete, Progress Report M1, Copenhagen.

18. Idorn, G.M. (1956) Disintegration of Field Concrete. The Danish National Institute ofBuilding Research and the Academy of Technical Sciences, Committee on AlkaliReactions in Concrete, Progress Report N1, Copenhagen.

19. Meyer, E.V. (1958) The Alkali Content of Danish Cements. Meyer, E.V. (1958) A NewDanish Alkali Resistant Cement. Andersen, J. and Ditlevsen, L. (1958) Methods for theDetermination of Alkalis in Aggregate and Concrete. The Danish National Institute ofBuilding Research and the Academy of Technical Sciences, Committee on AlkaliReactions in Concrete, Progress Reports F1, 2 and 3, Copenhagen.

20. Idorn, G.M. (1961) Studies of Disintegrated Concrete, Part I. The Danish NationalInstitute of Building Research and the Academy of Technical Sciences, Committee onAlkali Reactions in Concrete, Progress Report N2, Copenhagen.

21. Idorn, G.M. (1961) Studies of Disintegrated Concrete, Part II. The Danish NationalInstitute of Building Research and the Academy of Technical Sciences, Committee onAlkali Reactions in Concrete, Progress Report N3, Copenhagen.

22. Idorn, G.M. (1964) Studies of Disintegrated Concrete, Part III. The Danish NationalInstitute of Building Research and the Academy of Technical Sciences, Committee onAlkali Reactions in Concrete, Progress Report N4, Copenhagen.

23. Idorn, G.M. (1964) Studies of Disintegrated Concrete, Part IV. The Danish NationalInstitute of Building Research and the Academy of Technical Sciences, Committee onAlkali Reactions in Concrete, Progress Report N5, Copenhagen.

24. Idorn, G.M. (1964) Studies of Disintegrated Concrete, Part V. The Danish NationalInstitute of Building Research and the Academy of Technical Sciences, Committee onAlkali Reactions in Concrete, Progress Report N6, Copenhagen.

25. Plum, N.M. (1961) Alkaliudvalgets Vejledning. Foreløbig Vejledning i Forbyggelse afSkadelige Alkali-Kiselreaktioner i Beton (in Danish). The Danish National Institute ofBuilding Research, Copenhagen, Series SBI.


26. ASTM C-227: Test Method for Potential Alkali Reactivity of Cement-AggregateCombinations (Mortar Bar Method).

27. ASTM C-289 Test Method for Potential Reactivity of Aggregates (Chemical Method).28. Bredsdorff, P., Idorn, G.M., Kjær, A., Plum, N.M. and Poulsen, E. (1962) Chemical

Reactions Involving Aggregate. Chemistry of Cement. Proceedings of the FourthInternational Symposium, Washington 1960, Vol. 2. National Bureau of Standards,Washington, pp. 749–806 (with bibl.). Series: National Bureau of StandardsMonograph 43, Vol. 2.

29. Idorn, G.M. (1967) Durability of Concrete Structures in Denmark. DABI, Holte,Denmark.

30. Powers, T.C. and Steinour, H.H. (1955) An interpretation of some publishedresearches on the alkali-aggregate reaction”. ACI Journal Proc. V 51, 497–516.

31. Chatterji, S. (1978) An accelerated method for the detection of alkali silica reactivitiesof aggregates. Cement Conor. Res. 8, 647–650.

32. Chatterji, S. (1979) The role of Ca(OH)2 in the breakdown of Portland cementconcrete due to alkali silica reaction. Cement Concr. Res. 9, 185–188.

33. Knudsen, T. and Thaulow, N. (1975) Quantitative microanalyses of alkali-silica gel inconcrete. Cement Conc. Res. 5, 443–454.

34. Chatterji, S. and Clauson-Kaas, N. (1984) Prevention of alkali silica expansion byusing slag Portland cement. Cement Concr. Res., 14, 816–818.

35. Chatterji, S., Jensen, A.D., Thaulow, N. and Christensen, P. (1986) Studies of alkalisilica reaction. Mechanisms by which NaCl and Ca(OH)2 affect the reaction. CementConcr. Res. 16, 246–254.

36. Thaulow, N. and Thordal Andersen, K. (1988) Ny viden om alkalireaktioner (inDanish). Dansk Beton 1, 14–19.

37. Chatterji, S., Thaulow, N. and Jensen, A.D. (1981) Studies of alkali silica reaction. 4.Effects of different alkali salt solutions on expansion. Cement Concr. Res. 17, 777–783.

38. Bonzel, J., Krell, J. and Siebel, E. (1986) Alkalireaktion im Beton. Beton 345–348,385–389.

39. Chatterji, S. and Jensen, A.D. (1978) A simple chemical method for the detection ofalkali silica reaction. Cement Concr. Res. 18, 654–656.

40. Chatterji, S., Thaulow, N. and Jensen, A.D. (1988) Studies of alkali silica reaction. 6.Practical implications of a proposed reaction mechanism. Cement Concr. Res. 18, 363–366.

41. Danish Code of Practice for the Structural Use of Concrete, 3rd Edn, March 1984.Danish Standard DS 411.

42. Basisbetonbeskrivelsen for Bygningskonstruktioner (in Danish). Byggestyrelsen,Copenhagen, 1987.

43. Anon (1985) Test Method TI-B52 Petrographical Investigation of Sand. TechnologicalInstitute, Copenhagen.

44. Anon (1985) Test Method TI-B51 Alkali Silica Reactivity of Sand (in Danish).Technological Instutute, Copenhagen.

45. NORDTEST NT Build 295 (1985) Sand-Alkali-Silica Reactivity Accelerated Test.Nordtest, Finland.

46. Brotschi, J. and Mehta, P.K. (1978) Test methods for determining potential alkali-silicareactivity in cements. Cement Concr. Res. 8, 191–200.

47. Fördös, Z. (1981) Undersøgelse of Alkalikiselreaktioner i Beton Tilsat flyveaske (inDanish). Aalborg Portland, CBL Internal Report No. 29.

48. Chatterji, S. (1978) An accelerated method for the detection of alkali-aggregatereactivities of aggregates. Cement Concr. Res. 8, 647–650.

49. Justesen, C.F. and Nepper-Christensen, P. (1985) Portlandflyveaske cement (in Danish).Aalborg Portland, CBL Internal Report No. 36.

50. Chatterji, S. and Fördös, Z. (1985) Effect of Flyash Addition on Alkali SilicaExpansion. Nordic Concrete Research Publication No. 4.

51. Chatterji, S. and Nepper-Christensen, P. (1987) Evaluation of Portland Flyash cement asa preventive against alkali-silica reaction. Nordic Concrete Research Publication No. 6.


Documents

TF6912_CH06 alkali silica reaction on concrete chapter 06