Construction and Building Materials 71 (2014) 246–253

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Behavior of ternary blended cements containing limestone filler and flyash in magnesium sulfate solution at low temperature

http://dx.doi.org/10.1016/j.conbuildmat.2014.08.0370950-0618/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +40 2421208; fax: +40 2420781.E-mail address: nastasiasaca@gmail.com (N. Saca).

Nastasia Saca a,⇑, Maria Georgescu b

a Technical University of Civil Engineering, RO-020396 Bucharest, Romaniab University ‘‘Politehnica’’ of Bucharest, RO-011061 Bucharest, Romania

h i g h l i g h t s

� Studied the behavior of some ternary blended cements into 5% MgSO4 solution, at 5 �C.� The presence of both limestone and fly ash does not decrease the vulnerability of cements.� After 90 days of exposure, compressive strengths decrease for all samples.� The deterioration products were gypsum, ettringite and thaumasite–ettringite solid solution.

a r t i c l e i n f o

Article history:Received 13 November 2013Received in revised form 28 July 2014Accepted 23 August 2014

Keywords:Ternary blended cementsLimestone fillerFly ashMortarsMagnesium sulfate solution

a b s t r a c t

The behavior of ternary blended cements with limestone filler and fly ash additions into 5% magnesiumsulfate solution was evaluated. The presence of both limestone and fly ash in cement does not seem todecrease the vulnerability of cements to magnesium sulfate attack at 5 �C in comparison with limestonefiller or fly ash cements. After 90 days, the compressive strengths decrease for all specimens (from33.5 MPa to 16.4 MPa for Portland cement mortar and from 35 MPa to 13.8 MPa for 10% fly ash cementmortar) as a consequence of higher amount of deterioration products which exerts pressure in the cap-illare pores.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Cement production is responsible for about 5% of the globalman made CO2 emission. For each tone of cement being produced,an average of 0.87 tons of CO2 is being emitted [1,2]. A reduction ofthe CO2 emission during cement production can be possible byusing alternative fuels, by optimizing the heat transfer and usingsupplementary cementitious materials (fly ash, granulated blastfurnance, silica fume etc.) and fillers, without compromising thequality of the cement and concrete properties. The use of limestonefillers has physical, chemical and environmental effects oncements. Limestone filler can accelerate early age hydration ofPortland cement by interacting with calcium aluminate hydrateprovided by Portland cement hydration. This leads to calcium car-boaluminate hydrate formation instead of calcium monosulfatealuminate hydrate [3–6]. In these conditions, higher quantity ofettringite can slightly improve mechanical strengths. Fly ash can

provide additional calcium aluminate hydrate which increasesthe effect of limestone filler.

During this research, investigations of ternary blended cementscontaining limestone filler and fly ash were performed in order toobtain supplementary information regarding the influence ofcement replacement ratio by fly ash and/or limestone powder[1,7–11]. According to [1], after 120 days of hardening, the com-pressive strengths of cements with 5–10% limestone filler and25–30% fly ash were close to those of Portland cement.

The carboaluminate formation, the ettringite stabilization andthe supplementary CSH (formed by pozzolanic reaction) increasethe solid volume of hydrates and decrease the permeability ofthe ternary blended cements [9,12,13]. As a result of this a higherresistance of blended cements against sulfate attack is expected.

The behavior of blended cements in sulfate solution is an actualresearch topic.

The sulfate attack of limestone Portland cement mortars/concretes involves the thaumasite formation, particularly at lowtemperatures (<5 �C). Thaumasite formation requires a calcium sil-icate source, carbonate and sulfate anions, excess humidity and

N. Saca, M. Georgescu / Construction and Building Materials 71 (2014) 246–253 247

low temperature [14–21]. Carbonate ions source can be limestonepowder or calcium carbonate formation as a result of atmosphericcarbonation [22]. The thaumasite formation mechanism is stillcontroversial.

According to Bensted [23,24], the following routes could formthaumasite:

� Direct route (reaction (1)) between CSH phases with SO42�

and CO32� ions or atmospheric CO2, Ca2+ ions and excess of

water;� Woodfordite route (reaction (2)), take place below 15 �C,

from ettringite by substitution in its structure of Al3+ ionswith Si4+ ions in the presence of CO3

2� ions [23–26].

Table 1Chemic

Chem

SiO2

Al2OFe2OCaOMgOSO3

PC

a %Na

C� S�Hþ CaCO3 þ CaSO4 þ xH2O

! CaSiO3 � CaSO4 � CaCO3 � 15H2O ð1Þ

3CaO � Al2O3 � 3CaSO4 � 32H2Oþ 3CaO � 2SiO2 � 3H2Oþ 2CaCO3 þ 4H2O! CaSiO3 � CaSO4 � CaCO3 � 15H2Oþ CaSO4 � 2H2Oþ 2AlðOHÞ3 þ 4CaðOHÞ2 ð2Þ

According to Crammond [27], SO42� ions react with Ca2+ ions,

Al3+ ions, CO32� or HCO3

� ions and Si4+ ions to form ettringite, gyp-sum or thaumasite depending on ions concentration, stability ofprecipitates and relative solubility of competing species.

Kohler et al. [28] proposes a heterogeneous nucleation mecha-nism of thaumasite on the surface of ettringite including the disin-tegration of CSH takes place in cement paste.

The aim of this study is to investigate the behavior of ternarycomposite cements (Portland cement-limestone filler-fly ash sys-tem) as pastes and mortars immersed into magnesium sulfatesolution at 5 �C. The time evolution of compressive strengths ofsamples cured in water/immersed in 5% sulfate solution weredetermined. In addition to analyzing the visual appearance of sam-ples, SEM and EDX analysis were used in order to asses the deteri-oration of the samples and to identify the deterioration products.Thermal analysis (DTA) was also applied.

2. Materials and methods

The materials used in this research work are: Portland cement – CEM I 52.5(CEM I) according to SR EN 197-1 [29], limestone filler (L), and fly ash (FA) (seeTable 1).

Mineralogical composition of Portland cement was: 72.63% C3S, 1.02% C2S,9.76% C3A and 10.34% C4AF. Its specific surface area (Blaine) was 4190 cm2/g.

The limestone powder contained 85% CaCO3 and had a Blaine specific surfacearea of 5200 cm2/g.

The type F fly ash [30] had a Blaine specific surface area of 2108 cm2/g and thepozzolanic activity index was 87%. Pozzolanic activity index was determinedaccording to Romanian standard SR 13298 [31]. It was calculated as a ratio betweencompressive strength of blended cement (75% Portland cement + 25%FA) mortarand Portland cement mortar. The mortars had been left in the mold for 24 h, thencured for 4 days in water at 20 �C, for 46 h in water at 50 �C, for 2 h in water at20 �C and then tested.

al composition of CEM I and FA.

ical composition (%w) CEM I FAa

18.46 53.10–53.403 5.85 26.50–27.873 3.40 8.34–8.84

63.16 2.82–3.500.41 1.51–1.601.65 0.25–0.386.48 1.46–2.29

2O = 0.72–0.75; %K2O = 2.22–2.78.

Blended cements were prepared by homogenization of Portland cement with flyash and limestone into a rolling ball mill. For comparison, only fly ash and limestonewere considered as addition in blended cements (10–30% FA and 10–20% L, respec-tively –Table 2).

The mortars prepared with such binders had a water/cement ratio of 0.5 andbinder/siliceous sand ratio of 1:3. The prepared samples for compressive strengthsdeterminations having sizes of 20 mm � 20 mm � 20 mm had been preserved forone day in the mold and up to 28 days in water at 20 �C. At this age, some sampleswere immersed in 5% MgSO4 solution at 5 �C and the others had been continuouslycured in water at 20 �C until testing time (from 2 to 360 days).

The compressive strength was determined using a WPM machine. The compres-sive strengths were considered as relative strengths, (CSrel (%)), according torelation:

CSrel ¼CS1

CS2� 100

where: CS1 is compressive strength of blended cements immersed in sulfate solutionor cured in water for t days;

CS2 – compressive strength of blended cements or CEM I cured in water for thesame period of time.

All values presented in the paper are the average value of three determinations.Paste prisms with sizes 20 mm � 20 mm � 120 mm of some selected binders

(water/binder ratio = 0.5) were prepared in order to study the processes and prod-ucts of their interaction with magnesium sulfate solution at low temperature. After28 days of curing in water, the specimens were immersed in 5% MgSO4 solution at5 �C. The magnesium sulfate solution was replaced every three months.

The visul examinations on selected mortars imersed into sulfate solution at reg-ular intervals were performed in order to record surface deteriorations.

DTA and SEM analysis using a Shimadzu DTG-TA-50H apparatus and a HITACHIS2600N Scanning Electron Microscope equipped with Energy Dispersive X-ray spec-trometer, were used in order to obtain information regarding the interaction pro-cesses between specimens and sulfate solution and the morphology of reactionproducts. The electron microscopy analyses were performed on samples taken fromeither hard-core or surface zone (corroded zone). A thin conductive coating (gold)was added on the samples prior to imaging.

3. Results and discussion

3.1. Visual examination

The visual examination of samples cured in sulfate solution wascarried out after 60, 90 and 196 days of exposure into sulfatesolution.

After 60 days, the samples presented some deteriorations ofcorners and edges; samples with blended cements containing10% and 20% limestone (C-L10 and C-L20) were more affected thanreference sample (C) and samples with blended cements contain-ing fly ash-single or associated with limestone filler.

After 90 days of exposure in sulfate solution, samples C-L10 andC-L20 were characterized by some white efflorescence (substance)and surface spalls. These are present on the surface of referencesample (C) too.

Increasing the time exposure to 196 days results in serious dam-ages of specimens C, C-L10 and C-L20 and small pieces falling off theedges and surfaces of the samples (Fig. 1). The presence of fly ashonly does not seems to retard the sulfate attack considering theappearance of C-FA10 sample in comparison with C sample. Thepresence of both limestone and fly ash seem to decrease the vulner-

Table 2Codes and compositions of studied binders.

Code CEM I (%w) L (%w) FA (%w)

C 100 – –C-L10 90 10 –C-L20 80 20 –C-FA10 90 – 10C-FA20 80 – 20C-FA30 70 – 30C-FA10-L10 90 10 10C-FA20-L10 70 10 20C-FA10-L20 70 20 10

Fig. 1. Aspect of the samples immersed for 196 days in 5% MgSO4 solution at 5�C.

50

60

70

80

90

100

110

2 28 60 90 360

CSr

el. (

%)

Time (days)

C C-FA10-L10 C-FA20-L10 C-FA10-L20

Fig. 3. Relative compressive strengths of ternary blended cement mortars (C-FA10-L10, C-FA10-L20, C-FA20-L10), cured in water.

248 N. Saca, M. Georgescu / Construction and Building Materials 71 (2014) 246–253

ability of cements to magnesium sulfate attack at 5 �C, in compari-son with samples containing limestone or fly ash only. According tosome literature data [32,33], blended cements with fly ash seem tohave a good behavior in magnesium sulfate solution at low temper-ature only if the fly ash content is higher than 50%.

0

5

10

15

20

25

30

35

40

60 90 196

Com

pres

sive

stre

ngth

(MPa

)

Time (days)

C C-L10 C-FA10 C-FA10-L10

Fig. 4. The compressive strength of some blended cement mortars immersed into5% MgSO4 solution at 5 �C.

3.2. Compressive strengths

The blended cements which developed the best compressivestrengths by hardening in water at 20 �C [8] were selected andimmersed in 5% MgSO4 solution at 5 �C, from 2 to 369 days.

The compressive strength values of cement mortars with 10%,20% and 30% fly ash addition, cured into water, showed that flyash presence generally causes decreases in mechanical strengthsin comparison with the reference (Fig. 2). The diminishes are sig-nificant during the initial 28 days of hardening and are attenuatedfor longer periods of time. This is due to a negative effect of activecomponent dilution in blended binder (Portland cement) – espe-cially for a short hardening time. The very slow rate of pozzolanicreactions is due to a relative coarse character of fly ash.

The compressive strengths of mortars with 10%, 20% and 30% flyash content after 360 days are 96%, 92% and, respectively, 78% from

405060708090

100110120130140150

2 28 60 90 360

CS r

el. (

%)

Time (days)

C C-FA10 C-FA20

C-FA30 C-L10 C-L20

Fig. 2. Relative compressive strengths of cement Portland mortar (C), limestonefiller cement mortars (C-L10, C-L20) and fly ash cement mortars (C-FA10, C-FA20, C-FA30), cured in water.

reference sample value. This proves the contribution of pozzolanicreaction to strength development for long period of time.

The additions of the limestone filler in cement cause significantdecrease of the compressive strengths, in comparison with CEM I,especially after 28 days when the dilution effect of the active com-ponent in binder is the highest. The decrease of mechanicalstrength due to the presence of 20% limestone in cement is almostthe same with those corresponding to cement with 30% fly ashaddition, for 360 days.

Fig. 5. DTA curves of some blended cement pastes cured in water for 28 days orimmersed into 5% MgSO4 solution, at 5 �C, for 196 days.

N. Saca, M. Georgescu / Construction and Building Materials 71 (2014) 246–253 249

The compressive strengths developed by the ternary blendedcements were smaller in comparison with those of the cementswith similar contents of fly ash or limestone (Fig. 3). The binderC-FA10-L10 developed the best compressive strengths. The com-pressive strength of the cements with 30% addition (both lime-stone filler and fly ash) was very similar, regardless of thepreponderance of one or another addition.

The compressive strengths of specimens immersed into 5%MgSO4 solution at 5 �C showed a similar evolution, for binders C-L10, C-FA10-L10 and C, C-FA10, respectively. The compressivestrengths of C and C-FA10 binders increase after 60 days of expo-sure into sulfate solution (Fig. 4). Such a behavior can be attributedto continuos hydration with ettringite formation in conditions ofbrucite layer formation as a barrier of damage process. After90 days the compressive strengths decrease for all type of binders(for example, from 33.5 MPa to 16.4 MPa for C, from 35 MPa to

(a) x 500

(c) x 2500

(e)

Fig. 6. SEM images (a–c) of inner zone for C-FA10 immersed into 5% MgSO4 solutioninterpretation of the references to color in this figure legend, the reader is referred to th

13.8 MPa for C-FA10). It is assumed that formation of higheramount of ettringite exerts a pressure in the capillare pores whichcreates damage and strength loss.

The samples containing limestone – C-L10 and C-FA10-L10,immersed into sulfate solution showed a continuous decrease ofthe compressive strengths. This negative effect of limestone filleron the compressive strengths of specimens exposed to sulfate solu-tion could be explained by certain microstructural changes as aresult of the interaction between sulfate solution and binder speci-mens. This interaction is described by the chemical reaction (3)–(5).

3.3. DTA analysis

DTA curves of C, C-L10, C-FA10, and C-L10-F10 samples cured inwater for 28 days and of samples immersed into magnesium sul-fate solution for up to 196 days are presented in Fig. 5.

(b) x 1000

(d)

at 5 �C; EDX spectra of area in b figure (red mark)-(d) and all inner area (e). (Fore web version of this article.)

(a) x 500 (b) x 2500

(c) (d)

Fig. 7. SEM images (a, b) of a pore inside of C-FA10 sample; EDX spectra of all area (c) and of area in b figure (black mark)-(d).

(a) x 500 (b) x 2500

(c)

Ettringit-thaumasitesolid solution

Fig. 8. SEM images (a, b) and EDX spectra (c) of C-FA 10 sample’s surface.

250 N. Saca, M. Georgescu / Construction and Building Materials 71 (2014) 246–253

Pore Fly ash grain

(a) x 500 (b) x 1000

(c) x 1000 (d) x 2500

(e)

Fig. 9. SEM images (a–d) and EDX spectrum (e-all inner area) of C-FA10-L10 binder specimens immersed into 5% MgSO4 solution at 5 �C, for 360 days.

N. Saca, M. Georgescu / Construction and Building Materials 71 (2014) 246–253 251

DTA curves for specimens cured for 28 days in water show thefollowing effects:

� 98–107 �C and 150–162 �C, corresponding to dehydrationprocesses of calcium silicate hydrates(gel) and sulfate alumi-nate hydrates (AFm, AFt possible carbonated);

� An endothermic peak at 470–480 �C, associated with the lossof mass, corresponding to Ca(OH)2 dehydration;

� An endothermic peak associated with the loss of mass, at670–750 �C, caused by calcium carbonate decomposition.The source of CaCO3 could be unreacted limestone or calciumcarbonate accidentally formed during the preparation of thesamples (by reaction between Ca(OH)2 and CO2 from the air).

DTA curves of specimens cured in magnesium sulfate solutionfor up to 196 days show the following effects:

� 103–111 �C and 136–144 �C are related to dehydration ofCSH gel, ettringite, gypsum and thaumasite [16,34];

� Magnesium hydroxide formation is highlighted by the endo-thermic effect in the range of 396–407 �C;

� The small effects in range of 452–463 �C suggest the pres-ence of some remaining Ca(OH)2 quantities which did notreact with sulfate solution;

� An endothermic effect associated to calcium carbonatedecomposition at 723–750 �C.

During the deterioration process, the buffering calcium cationsare depleted and CSH decalcification could begin. Furthermorepotential protective brucite (Mg(OH)2) layer formation involvescalcium hydroxide consumption and pH decrease (for Mg(OH)2

pH = 10.5). The low solubility of brucite favors calcium hydroxideconsumption, according to reaction (3).

CaðOHÞ2 þMgSO4 þ 2H2O! CaSO4 � 2H2OþMgðOHÞ2 ð3ÞAs a consequence, CSH releases more calcium hydroxide into

intergranular solution, the pH rises so it remains stable. In theseconditions, CSH decalcification and noncementitious magnesium

(a) x 1500 (b)Fig. 10. SEM image (a) and EDX spectrum (b) of the surface area of C-FA10-L10 sample immersed into 5% MgSO4 solution at 5 �C, for 360 days.

252 N. Saca, M. Georgescu / Construction and Building Materials 71 (2014) 246–253

silicate hydrate (MSH) formation can occur, according to reaction(4) and (5) [35,36].

xMgSO4 þ xCaO � ySiO2 � zH2Oþ ð3xþ n� zÞH2O

! xCaSO4 � 2H2Oþ xMgðOHÞ2 þ ySiO2 � nH2O ð4Þ

4MgðOHÞ2 þ SiO2 � nH2O! 4MgO � SiO2 � 8:5H2Oþ ðn� 4:5ÞH2O

ð5Þ

These phenomena and the possible thaumasite formation jus-tify the significant decrease of compressive strengths of samplesimmersed into 5% MgSO4 solution at 5 �C.

3.4. SEM–EDX analysis

The figures from 6 to 10 represent SEM images and EDX spectraof C-FA10 and C-FA10-L10 samples, after 360 days of exposure intosulfate solution.

SEM images of inner zone of sample C-FA10 exposed to magne-sium sulfate solution at 5 �C show a weaker microcrystalline aspectof cement matrix (Fig. 6a and b) in comparison with the same sam-ple exposed into 5% MgSO4 solution at 20 �C [37]. This differencecould be a consequence of a lower rate of hydration – cristalisationprocesses due to both lower temperature and small surface reac-tion of FA (Blaine specific surface of 2108 cm2/g).

The main suggested hydrates are small-crystallized CSH andneedle like crystallized ettringite (5–10 lm in length) in pores(Fig. 7). The presence of solid solution ettringite–thaumasite couldbe supposed (Figs. 6c and 7b).

Elemental analyse of black selected zone in Fig. 7b, shows thepresence of Si (4.40%) by the side of Ca (73.88%), Fe (2.14%), S(12.38%), Al (6.71%) and Mg (0.19%). Red marked area in Fig. 6bcontains more Si (10.16), Ca (74.93%), Fe (2.76%), Mg (0.40%) andless Al (5.05%) and S (6.70) than black marked crystal.

SEM images of the surface specimen C-FA10, cured in MgSO4

solution for 360 days (Fig. 8) show a more deteriorated crystalsmorfology. The presence of ettringite–thaumasite clusters couldbe assumed. According to the elemental analyse data, on the spec-imen surface were identified Ca (76.29%), Si (9.91%), Fe (3.76%), S(7.65%), Al (1.97) and Mg (0.42%).

SEM images and EDX spectrum of C-FA10-L10 inner area sam-ple show an irregular morphology that could suggest the presenceof uncrystalised calcium silicate hydrates, small needle likeettringite crystals which are somewhat bigger in pores (Figs. 9aand d) and slabs like AFm phases.

On the sample‘s surface area, thaumasit small crystals togetherwith solid solution ettringite–thaumasite presence can be assumedaccording to SEM image (Fig. 10). The EDX spectrum reveals a com-position of type: Ca – 78.86%, Si – 9.81%, S – 6.92%, Fe – 3.72% andonly 2.23% Al. In the thaumasite structure, Al is replaced by Si.According to Barnett et al. [38], the thaumasite tolerates none orlittle Al in its structure. Ettringite can accept the replacement of½ of its Al by Si [39]. Mg was also found in EDX analyse (0.46%).Probably, magnesium ions are present as impurities into thauma-site crystals or adsorbed on their surface.

4. Conclusion

The behavior of some blended cements containing fly ash and/or limestone filler into 5% MgSO4 solution, at 5 �C, was investigatedin this research.

The mechanical strengths of the blended ternary cements (ashardened mortars in water) were lower than the one of Portlandcement (mortar). The samples C, C-L10, C-FA10 and C-FA10-L10were selected for studying their behavior into 5% MgSO4 solutionat low temperature (5 �C) considering compressive strengthcriteria.

The compressive strengths of samples C and C-FA10 continue toincrease after exposure to sulfate solution until 90 days. This couldbe a result of the hydration process in conditions of a brucite pro-tective layer formation. This was highlighted by the endothermiceffect in the range of 396–407 �C on DTA curves.

The presence of both limestone and fly ash does not seem todecrease the vulnerability of cements to magnesium sulfate attackat 5 �C. After 90 days of exposure into sulfate solution, the com-pressive strengths decrease for all samples, probably as a conse-quence of the ettringite formation as needle like crystals andthaumasite–ettringite solid solution crystals.

The fly ash content does not seem to have a positive effect onthe magnesium sulfate resistance of studied cements at low tem-perature. This is a consequence of a low pozzolanic reactivity dueto a relative coarse character of fly ash (Blaine specific surface areaof 2108 cm2/g).

References

[1] Damtoft JS, Lukasik J, Herfort D, Sorrentino D, Gartner EM. Sustainabledevelopment and climate change initiatives. CemConcr Res2008;38(2):115–27.

[2] Georgescu M, Panait N. Influence of CaCO3 on the hydration and hardeningprocesses in C3S–H2O system. Rom J Mater 2004;34(1):27–35.

N. Saca, M. Georgescu / Construction and Building Materials 71 (2014) 246–253 253

[3] Georgescu M, Panait N. Influence of CaCO3 addition on the hydration andhardening processes in C3A–(CSH2)–H2O system. Rom J Chem 2004;55(12):971–4.

[4] Kuzel HJ, Pöllmann H. Hydration of C3A in the presence of Ca(OH)2,CaSO4�2H2O and CaCO3. Cem Concr Res 1991;21:885–95.

[5] Matschei T, Lothenbach B, Glasser FP. The role of calcium carbonate in cementhydration. Cem Concr Res 2007;37:551–8.

[6] Lothenbach B, Le Saout G, Gallucci E, Scrivener K. Influence of limestone on thehydration of Portland cements. Cem Concr Res 2008;38:848–60.

[7] De Weerdt K, Justnes H. Synergy in ternary cements — the interaction betweenlimestone powder and fly ash. In: Black L, Purnell P, editors. 29th cement andconcrete science conference. UK: Leeds; 2009. p. 153–6.

[8] Georgescu M, Saca N. Properties of blended cements with limestone filler andfly ash content. U.P.B. Sci Bull Series B 2009;71:11–22.

[9] Georgescu M, Saca N, Voicu G. Hydration–hydrolysis processes in blendedcements with limestone filler and fly ash content. Rom J Mater2008;38:260–70.

[10] De Weerdt K, Kjellsen KO, Sellevold E, Justnes EH. Synergy between fly ash andlimestone powder in ternary cements. CemConcr Compos 2011;33:30–8.

[11] De Weerdt K, Ben Haha M, Le Saout G, Kjellsen KO, Justnes EH, Lothenbach B.Hydration mechanism of ternary Portland cements containing limestonepowder and fly ash. CemConcr Res 2011;41:279–91.

[12] Githachuri K, Alexander M. Durability performance potential and strength ofblended Portland limestone cement concrete. CemConcr Compos 2013;39:115–21.

[13] Vance K, Aguayo M, Oey T, Sant G, Neithalath N. Hydration and strengthdevelopment in ternary portland cement blends containing limestone and flyash or metakaolin. CemConcr Compos 2013;39:93–103.

[14] Hartshorn SA. Thaumasite formation in Portland-limestone cement pastes.CemConcr Res 1999;29:1331–40.

[15] Torres SM, Sharp JH, Swamy RN, Lynsdale CJ, Huntley SA. Long term durabilityof Portland–limestone cement mortars exposed to magnesium sulfate attack.CemConcr Compos 2003;25:947–54.

[16] Hartshorn SA. The thaumasite form of sulfate attack in Portland-limestonecement mortars stored in magnesium sulfate solution. CemConcr Compos2002;24(3–4):351–9.

[17] Barker AP, Hobbs DW. Performance of Portland limestone cements in mortarprisms immersed in sulfate solutions at 5 �C. CemConcr Compos1999;21:129–37.

[18] Zhou Q, Hill J, Byars EA, Cripps JC, Lynsdale CJ, Sharp JH. The role of pH inthaumasite sulfate attack. CemConcr Res 2006;36:160–70.

[19] Crammond NJ, Halliwell MA. The thaumasite form of sulfate attack in concretecontaining a source of carbonate ions. In: 2nd Symposium AdvConcr Tech ACISP vol. 154, 1995. p. 357–80.

[20] Torres SM, Lynsdale CJ, Swamy RN, Sharp JH. Microstructure of 5-year-oldmortars containing limestone filler damaged by thaumasite. CemConcr Res2006;36:384–94.

[21] Pipilikaki P, Papageorgiou D, Teas Ch, Chaniotakis E, Katsioti M. The effect oftemperature on the thaumasite formation. CemConcr Compos 2008;30:964–9.

[22] Collett G, Crammond NJ, Swamy RN, Sharp JH. The role of carbon dioxide in theformation of thaumasite. Cem Concr Res 2004;34:1599–612.

[23] Bensted J. Thaumasite – background and nature in deterioration of cementmortars and concretes. CemConcr Compos 1998;21:117–21.

[24] Bensted J. Thaumasite – direct, woodfordite and other possible formationroutes. CemConcr Compos 2003;25:873–7.

[25] Heinz D, Urbonas L. About thaumasite formation in Portland-limestonecementpastes and mortars-effect of heat treatment at 95 �C and storage at5 �C. CemConcr Compos 2003;25:961–7.

[26] Schmidt T. Sulfate attack and the role of internal carbonate on the formation ofthaumasite. PhD Dissertation, École Polytechnique Fédérale de Lausanne 2007.

[27] Crammond NL. The thaumasite form of sulfate attack in the UK. CemConcrCompos 2003;25:809–18.

[28] Kohler S, Heinz D, Urbonas L. Effect of ettringite on thaumasite formation.CemConcr Res 2006;36:697–706.

[29] SR EN 197-1:2011 – Cement – Part 1: Composition, specifications andconformity criteria for common cements.

[30] SR EN 450-1:2006 – Fly ash for concrete. Part 1: Definitions, specifications andconformity criteria.

[31] SR 13298:1995 – Natural and artificial pozzolanic materials. Determination ofpozzolanic activity indication.

[32] Mulenga DM, Stark J, Nobst P. Thaumasite formation in concrete and mortarscontaining fly ash. CemConcr Compos 2003;25:907–12.

[33] Nobst P, Stark J. Investigation on the influence of cement type on thaumasiteformation. CemConcr Compos 2003;25:899–906.

[34] Tsivilis S, Kakali G, Skarropoulou A, Sharp JH, Swamy RN. Use of mineraladmixtures to prevent thaumasite formation in limestone cement mortar.CemConcr Compos 2003;25:969–76.

[35] Baghabra Al-Amoudi OS. Attack on plain and blended cements exposed toaggressive sulfat environments. CemConcr Compos 2002;3–4:305–16.

[36] Badanoiu A, Voicu G. Sulfate resistance of binary and ternary blended cements.Rom J Mater 2008;38:271–83.

[37] Saca N. Implication of calcium carbonate into the hardening and properties ofcomposite materials with mineral binder. PhD Thesis, University Politehnica ofBucharest; 2009.

[38] Barnett SJ, Macphee DE, Crammond NJ. Extent of immiscibility in theettringite-thaumasite system. CemConcr Compos 2003;25:851–5.

[39] Tosun K, Felekoglu B, Baradan B, Altun A. Effect of limestone replacement ratioon the sulfate resistance of Portland limestone cement mortars exposed toextraordinary high sulfate concentration. Constr Build Mater2009;23:2534–44.