108306-9494 ijcee-ijens

International Journal of Civil & Environmental Engineering IJCEE-IJENS Vol:10 No:06 37

108306-9494-IJCEE-IJENS © December 2010 IJENS I J E N S

Abstract— Modern construction requires advanced concretes,

which have high strength, cost effectiveness and durability. The

influence of sodium sulfate solutions on the physical properties of

neat and blended cement has been investigated. The ordinary

Portland cement was replaced by 35% ground granulated blast

furnace slag (GGBS). The cement pastes were mixed using

water/cement ratios 0.25. The hardened cement paste samples

were immersed in sodium sulfate solutions having different

concentrations between 0 and 1.5 %, for 3, 14 and 28 days. The

results showed that, the increase of sulfate ions leads to increase

in the porosity, the relative mass loss and the relative expansion.

Also show a decrease in the bulk density and the chemical

combined water. The critical variations in these parameters

occur at a concentration ≈ 0.5%.

Index Term— Blended cement paste, combined water,

Relative mass loss, sodium sulfate, Relative expansion.

I. INTRODUCTION

Cement is an economical and durable material that is one of

the world’s most utilized man-made building products.

Cement is utilized for its good compressive strength and

stiffness but the major drawback is that it lacks tensile

strength.

External sulfate attack is one of the durability problems

associated with concrete. Since its identification it has been

the subject of numerous studies and still not totally understood

[1]. The extent to which concrete is affected by sulfates

depends on several factors including its permeability, water to

cement (w/c) ratio, type of cement, exposure conditions and

the environment [2]. Assuming the same environmental

conditions, two factors will tend to control the resistance of a

given concrete to sulfate attack: the chemistry of the cement

and the permeability of the concrete. To control the cement

chemistry, American Standards [3] suggest a limit on the

(C3A) and (2C3A + C4AF) contents of sulfate resistant, type V,

cements as 5% and 25%, respectively. On the other hand,

cements with low C3A and C4AF compounds generally tend to

have a higher C3S/C2S ratio, and an increase in the C3S

content of cement generates a significantly higher amount of

calcium hydroxide, as the hydration of C3S produces nearly

2.2 times more calcium hydroxide (CH) than the hydration of

C2S. Calcium hydroxide is known to be responsible for the

formation of gypsum, and gypsum is known to be the first step

E. AL-Salami, is the head of physics department, Faculty of Science King

Kaled University, Abha, SA (e-mail: [email protected]).

A. Salem

of the formation of ettringite, which can be considered as the

principal cause of deterioration [4]. Low C3A and low C3S

cement showed 10 times less expansion than those with a low

C3A and a high C3S content [5]. Furthermore, to control the

permeability of concrete, lower w/c ratio and/or pozzolans are

recommended [6,7]. Effect of various pozzolans on the

resistance of cements to external sulfate attack has also been

studied by other researchers [8–10]. Pozzolans reduce not only

the permeability but also the C3A amount if they are a partial

replacement of cement. Moreover, use of pozzolans or use of

blended cements, in general, reduces the quantity of CH due to

the pozzolanic reactions which would otherwise react with

sulfates to form gypsum.

The present work examines the effects of slag and curing time

on corrosion resistance (sulfate attack) and durability of OPC

mortars, through the improvement of the physical properties of

cement–based materials due to their pozzolanic properties.

II. EXPERIMENTAL WORK

The starting materials used in this study were ground

granulated blast furnace slag (GGBS) of Blaine surface area ≈

3000 cm2/g and ordinary Portland cement of Blaine surface

area ≈ 3100 cm2/g, The oxide composition of OPC and GGBS

were shown in Table I.

The blended cement pastes were prepared using ordinary

Portland cement that was partially substituted by 35% ground

granulated blast furnace slag (GGBS) by weight of cement.

The ingredients were homogenized on a roller in a porcelain

ball mill with four balls for 1 hour to assure complete

homogeneity. The pastes were prepared using the

water/cement ratios 0.25. The pastes were molded into 2 cm

cubes. The moulds were vibrated for one minute to remove

any air bubbles. The samples were kept in moulds at 100%

relative humidity for 24 hours, and then it cured under water

for 3, 14 and 28 days. The hardened cement pastes were dried

at a temperature of 105oC for 24 hours in an oven. Then, they

were kept in sodium sulfate solution of concentration 0, 5000,

10000 and 15000 ppm for 3, 14 and 28 days. Each

measurement of the physical properties was carried out for

three identical cubes of the same mix composition.

Porosity and bulk density of each cube sample are measured

according to ASTM C140-01 [11]. The procedures are as

follows:

Effects of mix composition on the sulfate

resistance of blended cements A. E. AL-Salami and, A. Salem

mailto:[email protected]



Table I

Chemical analysis of the starting materials

Constituents

OPC

(%)

GGBS

(%)

CaO 62.56 33.76

SiO2 20.85 38.60

Al2O3 4.70 13.20

Fe2O3 3.86 3.53

MgO 1.23 5.36

SO3 2.79 0.55

Na2O 0.49 1.33

K2O 0.12 0.66

Cl- --- 0.07

I.L. 2.82 2.94

Immerse the specimens in water at room temperature (22 oC)

for 24 h. Weigh the specimens while suspended by a thin wire

and completely submerged in water and record Wi (immersed

weight). Remove the specimens from water and allow water to

drain for 1 min by placing them on a wire mesh, removing

visible surface water with a damp cloth; weigh and record as

Ws (saturated weight). Then, dry all specimens in a ventilated

oven at 105 oC for not less than 24 h and two successive

weightings at intervals of 2 h show an increment of loss not

greater than 0.2% of the last previously determined weight of

the specimen. Record the weight of dried specimens as Wd

(oven-dry weight).

The absorption and oven-dry density are calculated as

Absorption% = [(Ws - Wd)/Wd] x 100;

Density (kg/m3) = [ (Wd/(Ws - Wi) x 1000

Where Wi, Ws, Wd are in kg

loss of weight (Lw )

Lw=(Wd-Ig)/Wd x100

Where Ig is the ignited weight at any temperature level (g).

Combined water (Wn)

Wn= [Wd-W'/W'] - L

Where W' : is the ignited weight of specimen (g) for one hour

at 10000c, then caold in a dissicator then weighted.

L ; is the ignition loss of un hydrated specimen, an is equal to

(W'-W10000C).

III. RESULTS AND DISCUSSION

The physical properties of cement pastes subjected to different

concentration of Na2SO4 were studied to clarify its physical

performance affected during sodium sulfate expunger.

The results of bulk density of the hardened cement pastes,

which exposed to a different concentration of Na2SO4 in

absence and presence of slag are graphically illustrated in fig

(1-a,b). In general, the bulk density increases gradually with

increasing, the hydration age of the cement pastes. This is due

to the decrease in the total porosity with increasing hydration

age as a result of the filling of pores by the hydration products.

The results of figure (1-a) indicate that the bulk density of the

hardened cement pastes increases at concentration ≈ 5000 ppm

of Na2 SO4, then decreases gradually with increasing of Na2

SO4. The increase of bulk density at 5000 ppm is due to

formation of ettringite and gypsum. Therefore the samples

with higher C3S content that decreases bulk density, the

sulfate attack may have involved a complex mechanism of

ettringite expansion, gypsum formation, and decalcification of

the C-S-H. Obviously, the bulk density of blended cement

pastes containing 35% slag increases as the concentration of

sodium sulfate increases. This is mainly due to decrease of the

amount of clinker and then the C3A (tray calcium aluminates)

which produce sulphoaluminate as a results of C3A and

Na2SO4 reaction. This is leading to increase the bulk density

with the concentration of Na2SO4 in the samples as illustrated

in figure (1-b). The increase of the bulk density is mainly due

to the activation of blended cement with sulfate ions in the

samples forming hydrated products which fill some open

pores in the samples [12-14].

Fig. (1-a,b). Variation of the bulk density of the hardened cement pastes

versus Na2 SO4 concentration f at different curing ages; (a) without slag; (b)

with 35% slag

Figures (2-a,b) shows the variation in the mass loss verses the

concentration of Na2 SO4 at different curing ages . It is clear

that the rate of mass loss increases with increasing the curing

age due to the precipitation of hydrated phases in to the

available pores of cement pastes. But in fig.(2-a) the rate of

the of mass loss increases with the concentration of Na2SO4 in

the samples, and that is related to the increase and formation

of gypsum and ettringite which are produced by the chemical



reaction of sulfate salt with CH and C3A. The rate of

formation of gypsum and ettringite plays an important role in

the increase of mass loss, where the gypsum results makes the

samples softening and brittle as well as the hydration product

of the (C-S-H) phase converts to gypsum ,and this leads to

mass loss. In fig (2-b) the rate of the mass loss decreases with

increasing of Na2SO4 in the samples, that is due to the slag

reacts with the liberated lime from the hydration of clinker,

hence, the free lime is nearly consumed. So there is no chance

to reaction of lime with sulfate group, and that is leading to

decreasing in the formation of gypsum, which is responsible to

increasing the mass of the samples [15,16].

Fig. (2-a,b). Variation of the mass loss of the hardened cement pastes with the

Na2SO4 concentration at different curing ages: (a) without slag;(b) with 35%

slag.

The total porosity of the blended cement pastes cured for 3,14

and 28 days are graphically represented as a function of the

concentration of Na2SO4 in the samples in fig(3-a,b). It is clear

that the total porosity decreases with curing time for all

hardened cement pastes. This is due to the filling up of a part

of the available pore volume with the hydration products, as

the hydration proceeds. In fig (3-a) the total porosity

decreasing up ≈ 5000 ppm , then increasing with the

concentration of Na2SO4 in the samples. This variation mainly

depending on the rate of the geometrical random arrangement

of the material hydrated phases during the hydration

mechanism inside the matrix of the sample [17].

Fig. (3a,b). The variation of the total porosity as a function of the different

concentration of Na2SO4 in the cement pastes at different curing age (a)

without slag, (b) with 35% slag.

The increase in the total porosity values of hardened

blended cement pastes after the concentration ≈ 5000 ppm is

related to the increase of gypsum and other products which

increase the total porosity with the concentration of Na2SO4 in

the samples. The variation of the total porosity in fig(3-b) is

decreasing with increasing of Na2SO4 in the specimens. That

is due to the formation of mush amounts of (C-S-H) [reaction

of slag with CH, also the content of slag in the opc matrix to

decrease the interaction of the liberated lime (C-H) with the

sulfate group SO42- to form gypsum, which is increase the

total porosity.

The relative expansion of the blended cement pastes as a

function in the concentration of the Na2SO4 and at different

curing time were represent in fig(4-a,b). The general behavior

of the relative expansion is decreasing with increasing the

curing time duo to the precipitation of hydrated phases into the

pores of the system, and that is leading to decreasing in the

relative expansion of the samples.

In fig (4-a) the relative expansion increasing with increase the

concentration of Na2SO4 in the sample, and that is attributed to

the chemical interaction of the sulfate group with the C3A,

producing the ettringite formation, and this reaction depending

on the concentration of Na2SO4 in cement paste.

The ettringite formation results in expansion of the material

and that are related to the water absorption of crystalline

ettringite, the morphology and size of the ettringite crystal.

But in fig(4-b) the relative expansion decreasing with

increasing Na2SO4 in the samples after the concentration ≈

5000( ppm), and that is due to the random arrangement of the



hydrated phases in the cement matrix. The decreasing in the

relative expansion with Na2SO4 is mainly duo to the increase

of the slag hydration, in the cement matrix, and that is leading

to decrease the contents of C3A in the cement clinker, and

decrease the rate of formation the ettringite interaction of C3A

with Na2SO4.

Fig. (4-a,b). The variation of the relative expansion of the hardened cement

pastes versus the concentration of Na2SO4 at different curing age (a) without slag (b) with 35% slag

The results of the chemically combined water contents (water

represents that part of water retained by the cement paste after

the free water has been removed it has been determined as a

measure of the degree of hydration for the blended cement

pastes hydrated for 3,7 and 28 days, are graphically

represented as a function in the concentration of Na2SO4 in

fig(5-a,b). In general in this figure (5-a,b) the chemical

combined water increases with the curing time, and that is

related to the increase of formation of the hydration phases

with curing time, that is leading to increase the combined

water inside the samples [18-20].

In fig(5-a) the combined water decreasing with increasing the

concentration of Na2SO4 in the samples. that is attributed to

the interaction of sulfate group with hydrated phases CH and

(C-S-H) and producing un hydrated phases [ gypsum and

ettringite ], that is leading to decrease in the combined water

inside the matrix. But in fig (5-b) it is clear that the combined

water content of the blended cement pastes increase with the

Na2SO4. This is duo to high pozzolanic activity of slag with

the limestone (C-H) during the hydration mechanism, forming

additional amounts of calcium silicate hydrates (C-S-H).

There fore, the combined water tends to increase [21]. Finally

the reactions between the sulphate group SO42- and the

hydrated lime (CH) and traycalcium aluminate (C3A) in the

cement paste, result in the formation of gypsum and ettringite

(has a large specific volume), which causes undesirable effects

on the physico-mechanical properties, that may produce

failure of the cement paste. The present of the pozzolanic

material (slag) in the cement matrix can be improved the

physico-mechanical properties by reducing the (CH) phase

C3A content in ordinary Portland cement, and producing the

(C-S-H) hydrate phase.

Fig. (5-a,b). The variation of the combined water contents in the blended

cement pastes with the concentration of Na2SO4 at different curing age (a) with out slag (b) with 35% slag

IV. CONCLUSIONS

The results are summarized as follows.

1- When cement paste phases are exposed to sodium

sulfate gypsum and ettringite are produced depending

on the concentration of sodium sulfate in the sample,

by chemical reaction of sulfate with Ca(OH)2 and

C3A.

2- The formation of gypsum plays an important role in

the decrease in the balk density and mass loss and

combined water, where the gypsum results makes the

specimen material brittle and softening.

3- Ettringite formation results in cracking and expansion

of the material and this expansion leads to a increase

in the total porosity and the relative expansion of the

cement matrix, and that is related to the water

absorption of crystalline ettringite , the morphology

and size of the ettringite crystal .



4- The physico- mechanical are affected by the present

of sulfate solt in cement paste, and that is leads to

decrease the durability of cement paste.

5- The variation in the physico–mechanical properties

occurs at the concentration≈ 5000 ppm can be

regarded as a geometrical random arrangement of

material phases in the samples.

6- The pozzolanic addition (slag) to the blended cement

past improved the durability of the physico-

mechanical properties and corrosion inhibitor. This is

because the interaction the active pozzolanic material

(slag) with the free lime (CH) during the hydration

mechanism leading to form the (C-S-H) phases, and

consequently decreasing the formation rate of

gypsum and ettringite [interaction of So42- with CH].

REFERENCES

[1] A. Neville The confused world of sulfate attack on concrete. Cem Concr Res 2004(34):1275–96.

[2] J. Skalny Sulfate attack issues: an overview. In: Marchand

Jacques, Skalny Jan, editors. Materials science of concrete – sulfate attack mechanisms. The American Ceramic Society; 1999.

p. 49–63.

[3] ASTM C 150-02, Standard specification for portland cement, West Conshohocken (PA): American Society for Testing and

Materials, vol. 04.01, 2002.

[4] PK. Mehta. Mechanism of expansion associated with ettringite formation. Cem Concr Res 1973(3):1–6.

[5] Irassar EF, Gonzalez M, Rahhal V. Sulfate resistance of type V

cements with limestone filler and natural pozzolana. Cem Concr Comp 2000(22):361–8.

[6] A. Neville Properties of concrete. 4th ed. Prentice Hall; 1995.

[7] PK. Mehta, PJM. Monteiro Concrete – microstructure,

properties, and materials, Chennai, India, 1997.

[8] RE. Rodriguez-Camacho Uribe-Afif R. Importance of using

natural pozzolans on concrete durability. Cem Concr Res 2002(32):1851–8.

[9] Al-Amoudi OSB. Attack on plain and blended cements exposed

to aggressive sulfate environments. Cem Concr Comp 2002(24):305–16.

[10] MD. Cohen, A. Bentur Durability of portland cement–silica fume

pastes in magnesium sulfate and sodium sulfate solutions. ACI Mater J 1988(85):148–57.

[11] ASTM C140-01, Standard test methods for sampling and testing

concrete masonry units and related units. Annual book of American Society for Testing and Materials; 2001.

[12] A. E. Al-Salami, A. Al-Hajry,. M. A. Ahmed and S. Taha accepter for publication in Silicate Indsutrial(2005).

[13] H. El-Didamony, M. Y. Haggag and. S. A. Abo.El-Enein Cem.

Concr. Res. 8,351(1,78). [14] W. H. Gutte and. A. D. Russel. Meg . Concr.Res.41(146)9(1989).

[15] Hansan Biricik , Fevziye AKoz, Fikret Turker, Ilhan Berktay.

Cem.Concr. Res.30,1189(2000). [16] M. D. Cohen and A. Bentur. ACI Mater.J-85, 148(1998).

[17] G. E. Monfor, J.of the PCA Research and development

laboratories. 10 (21). 35,May (1968). [18] S. Brunauer, M. Yudenfreund, M. I.older. and R. Mikhail Sh.,

Final report clarskon college of technology postdam, New York

(1971). [19] S. Wild, J. M. Katib and M. O. Farrell ,Cem.Concr.Res.27(5),

697(1997).

[20] T. Ramlochan, M. D. A. Thomas and R.D. Hooton.. Cem.Concr.Res.34,1341 (2004).

[21] P. C. Hewlett., leas chemistry of cement and concrete 4th Edward

Arnod, London (1998).

108306-9494 ijcee-ijens

Documents