The effect of Persian Gulf tidal zone

exposure on durability of mixes

containing silica fume and blast
furnace slag
Ganjian, E. and Pouya, H.S.
Author post-print (accepted) deposited in CURVE November 2012

Original citation & hyperlink:

Ganjian, E. and Pouya, H.S. (2009) The effect of Persian Gulf tidal zone exposure on
durability of mixes containing silica fume and blast furnace slag. Construction and Building
Materials, volume 23 (2): 644-652.
http://dx.doi.org/10.1016/j.conbuildmat.2008.02.009

Copyright © and Moral Rights are retained by the author(s) and/ or other copyright
owners. A copy can be downloaded for personal non-commercial research or study,
without prior permission or charge. This item cannot be reproduced or quoted extensively
from without first obtaining permission in writing from the copyright holder(s). The
content must not be changed in any way or sold commercially in any format or medium
without the formal permission of the copyright holders.

This document is the author’s post-print version, incorporating any revisions agreed during
the peer-review process. Some differences between the published version and this version
may remain and you are advised to consult the published version if you wish to cite from
it.

CURVE is the Institutional Repository for Coventry University

http://curve.coventry.ac.uk/open
 The effect of Persian Gulf tidal zone exposure on durability of mixes

containing silica fume and blast furnace slag

Eshmaiel Ganjiana,*, Homayoon Sadeghi Pouyaa

a
Coventry University, Faculty of Engineering and Computing, Department of Built Environment,

Coventry CV1 5FB, UK

Abstract

In this research the performance of cement paste and concrete mixes incorporating 7

and 10 percent of silica fume (S.F.) as a cement replacement was investigated in three

exposure conditions. The results showed that plain type II portland cement performed

better than blended S.F. cement under cyclic wetting and drying conditions. Silica fume

specimens under cyclic wetting and drying conditions in simulated seawater exhibited

higher strength loss compared to plain type II portland cement where cured under

potable water. In addition, the greater the silica fume amount used in the mixes, the

more the capillary water absorption under tidal zone exposure or/and under wetting and

drying simulation. Further, the ternary blended GGBS (Ground Granulated Blast

furnace Slag) mix was the worst performing mix in all exposure conditions.

Keywords: concrete durability; concrete deterioration; cement replacement materials;

blast-furnace-slag; marine conditions.

_________
* Corresponding author. Tel.: +44-24-7688-7625; Fax: +44-24-7688-8296.
E-mail address: [email protected].
1. Introduction

In recent decades the use of silica fume (S.F.) as a supplementary cementitious material

has increased substantially in concretes used for marine structures in the Persian Gulf.

The reason for use of silica fume in these structures is claimed to be an increase in the

durability of coastal structures (including those within the tidal zone).

The performance of concrete in sea water has been extensively studied [1–9]. The

findings of these studies are at the same time inconclusive and sometimes controversial

[10 and 11]. This reflects the complex nature of the durability problem which is

attributed to one or more of the following factors [12].

• Characteristics of the ingredient materials and mix design characteristics such as

water-cement ratio, cement to aggregate and fine to coarse aggregate ratios,

degree of compaction, curing type and age, etc.

• Test methods adopted to assess the concretes.

• Duration and severity of the exposure period.

• Variability of the seawater concentration and the variation in its temperature.

• Specimen type (paste, mortar, concrete), shape and size.

• Use of small specimens to predict the behaviour of structural concrete.

• The dominant deterioration mechanism or mechanisms that can operate

simultaneously.

One of the lessons learned from the case histories of concrete deterioration in seawater

is that the individual processes of deterioration tend to limit themselves to different

parts of a concrete structure, depending on the tidal lines [4]. Thus, a structure can be

divided into three zones: the atmospheric zone (above the high-tide level), the tidal zone

(between the tide lines), and the submerged zone (below the low tide level). The most
severe deterioration usually occurs in the tidal zone, which is the focus of this

investigation.

The most aggressive chemicals that affect the durability of concrete structures in

seawater are chlorides and sulfates with their associated cations. The chloride dissolved

in water increases the rate of leaching of portlandite and thus increases the porosity and

capillary absorption of concrete, and leads to loss of stiffness and strength. Calcium,

sodium, magnesium, and ammonium sulfates are (in increasing order) harmful to

concrete as they react with hydrated cement paste leading to expansion, cracking,

spallings and loss of strength [13].

It is noteworthy that the hydrated cement phase that binds chloride ions is tricalcium

aluminate (C 3 A) forming an insoluble compound, Friedel’s salt (calcium chloro-

aluminate hydrate), thereby removing a portion of chlorides from corrosive attack. This is

the same phase that, in excess of a certain critical level, reacts with sulphate ions in the

hardened state to form ettringite and gypsum, thereby leading to degradation of concrete

into a non-cohesive granular mass and to disruptive expansion and cracking [14].

Although type V cement can increase the chemical resistance to sulphate attack, it also

tends to reduce the corrosion resistance of reinforcement in presence of chloride ions.

This is due to the lower content of C 3 A in type V cement compared to type I and II that

reduces the chloride binding capacity of cement and therefore promotes steel corrosion

[15-17] in concrete. Type I cement with higher C 3 A content is however more vulnerable

to sulphate attack and performs worse than the other types of cement [18]. Other studies

[11 and 19] show that type II cement performs better than types I and V cements in

marine environments; for this reason type II cement was chosen for this investigation.

Moreover, use of supplementary cementitious materials such as silica fume and blast

furnace slag with portland cement has increased significantly in the Persian Gulf region.
The rate at which the hardened cement paste is deteriorated due to exposure to harmful

chemicals mainly depends on the concentration of the chemicals in water, the time of

exposure, and the chemical resistance of the concrete. Extensive investigations have been

carried out on the use of silica fume and blast furnace slag in concrete during the past two

decades and have led to their widespread application in the construction industry [20-24].

Many national standards also determine the degree of attack, and this is primarily on the

basis of the concentration of the aggressive substances. However, the chemical resistance

of high-performance concrete using silica fume or combined silica fume and blast furnace

slag is an issue that has not yet received sufficient attention from the research community

[24]. Hence, this investigation was undertaken to study the effect of silica fume and

GGBS blended with type II cement on progressive deterioration of pastes and concretes in

the marine tidal zone of the Persian Gulf, and in laboratory tanks simulating severe

conditions.

2. Experimental program

An experimental program was designed to produce a high-performance paste and

concrete for the best protection in marine tidal zone by optimising silica fume [25] and

GGBS replacement for the type II cement. The materials used and the experimental

procedures are described in the following sections.

2.1. Materials

Locally available Type II portland cement (ASTM C 150 Type II) complying with

Iranian specification 389 was used. The chemical composition, compound composition

(calculated complying Bogue), specific gravity and the specific surface of the cement

used are shown in Table 1.

Locally available silica fume and ground granulated blast furnace slag complying with

ASTM C 1240 and ASTM C 989-93 (except the lower fineness of GGBS) were used

with naphthalene sulphonate based high-range water reducing admixture (HRWRA) to

achieve constant workability in all paste and concrete mixes. Table 1 details the

chemical composition and the nitrogen BET fineness of the SF and the Blaine fineness

of the GGBS used in this investigation. The slag activity index of the GGBS used was

calculated as 67 percent based on ASTM C989 definitions.

Commercially available crushed siliceous aggregates complying with ASTM C33-78

were employed for both coarse and fine fractions to make concrete specimens. The

relative density, absorption and fineness modulus of the fine aggregate were 2.65, 4.2%

and 3.72, respectively. The specific gravity and absorption for coarse aggregate were

2.64 and 1.8%, respectively.

Reagent grade MgSO 4 , Na 2 SO 4 , NaCl and MgCl 2 were used for sulphate and chloride

exposure in wetting and drying simulation ponds. Potable water was used throughout

for mixing, initial 7 days curing of all specimens, and curing in the control laboratory

water tank. The chemical analysis of the water used is shown in Table 2.

2.2. Mix proportions

There were four basic mixes in which silica fume and GGBS were incorporated to make

paste and concrete mixes. The first mix was the control Type II cement mix. The second

mix was the optimum amount of silica fume replacement, i.e. 10 percent, which was

adopted from the preliminary studies on physical and mechanical properties reported by

Ganjian et al. [25]. The experiments [25] indicated that 10% replacement by silica fume

achieved the highest compressive strength at the age of 28 days. In the third mix, 7

percent silica fume replacement was used in order to evaluate the effect of the amount
of silica fume on the strength loss. The fourth mix was based on the optimum

replacement amounts of blended silica fume and GGBS with type II cement, which

were 10 and 50 percent by weight of cement respectively as obtained by several trial

mixes. The amount of silica fume in such trial mixes was kept constant at 10% of total

mass of cementitious materials, and the amount of GGBS was varied to reach the

highest relative compressive strength. The mix designs and characteristics of all pastes

and concretes are given in Tables 3 and 4, respectively.

The amount of water in mixes was chosen based upon the total content of cementitious

materials used in each mix. The water-cementitious materials ratio and per cent flow

were kept constant for all mixtures and HRWRA was added to give constant

workability. The required HRWRA per cent by mass of cementitious materials was

obtained by trial and error from preliminary mixes.

2.3. Mixing and casting

There was particular interest in selecting a mixing procedure that would minimize

dusting, especially during the initial stage of mixing and subsequent clustering of the SF

particles after contact with water. For this reason the dry and wet methods of mixing SF

and cement were examined. The dry mixing procedure of the portland cement, silica

fume and GGBS proved to be the most effective in the dispersion of the SF particles

and the most reliable method to obtain a homogenous mixture [25].

The following mixing procedure was used for paste mixes:

1- Dry mix the portland cement and SF (and GGBS if required) in a Hobart ASTM

mixer at low speed for 60 seconds. It is imperative to use a plastic cover to prevent

escape of fine particles.

2- Mix the required water with superplasticizer in a measuring vessel.

3- Without stopping the mixer, gradually add all of the solution (water and HRWRA) in

next 60 seconds.

4- Stop mixer and change to medium speed. Resume mixing for further 60 seconds.

5- Stop mixer. Hand mix with a spatula by scraping the sides and dispersing the large

visible clusters for 60 seconds.

6- Leave the mixture to stand for 30 seconds.

7- Resume mixing at high speed for 2 minutes.

Total time required is therefore six and a half minutes. The following mixing procedure

was used for concrete mixes using a tilting drum mixer of 0.15 cubic meter capacity.

The interior of the drum was initially wetted with water to prevent absorption. The

coarse aggregate fractions were mixed first, followed by the cement, part of the required

amount of sand, and the water containing half of the required amount of

superplasticizer. The final mixing stage involved the addition of the SF (and GGBS if

required), and the remaining sand. One-half of the super-plasticiser was always retained

with one litre of water to be added during the last three minutes of the mixing period.

50 mm and 100 mm cube moulds were used for paste and concrete specimens

respectively, and a vibrating table was used to achieve full compaction. After casting,

all paste and concrete specimens were covered with wet burlap in the laboratory at 20±1

°C and 65 % relative humidity for 24 hours.

2.4. Initial curing

After de-moulding, all specimens were initially cured in a potable water-curing tank at

laboratory temperature for 7 days and then transferred to the three specific exposure

conditions. Chemical analysis of potable water used for making mixes and used for

initial curing is given in Table 2.

2.5. Long term simulation ponds and tidal zone site exposures

Specimens were kept in three different conditions for longer ages as follows:

A) In fresh potable water curing tank, control curing regime (for both paste and concrete

specimens) at 20±1 °C.

B) In a simulation pond under cyclic wetting and drying in synthetic seawater solution

(for both paste and concrete specimens). The concentration of major ions in the solution

used for the simulation pond is given in Table 2. Simulation pond temperature was kept

at about 37 to 42 oC according to dominant temperature and time intervals of wetting

and drying cycles within tidal zone of south coast of Iran by using electric heater rods

for drying and wetting period respectively. An ultraviolet lamp was mounted on the top

of each pond to bring sunlight effect to the specimens. Owing to the fact that wetting

and drying cycles were set to occur at 6 hourly intervals and temperature was kept

higher than real site environment, the simulation conditions were harsher and more

severe than site exposure tidal zone conditions. The solution in the simulation ponds

was refreshed every two weeks.

C) In a coastal environment on the north coast of the Persian Gulf, in the south of Iran.

The paste specimens with the same characteristics of those cured in water tank

(condition ‘A’) and in the simulation ponds (condition ‘B’), were placed in the seashore

of Kish Island in the south of Iran. The concentration of major ions in the Kish Island’s

seawater is given in Table 2. In this exposure site the specimens were kept in tidal zone

for 6 months, protected from waves by surrounding rock barriers. The specimens were

on average subjected to 18 hours of wetting and 6 hours of drying per 24 hours. The

average seawater temperature was approximately 33 °C.

2.6. Testing program

The various paste and concrete mixes were tested for compressive strength

development, at ages of 3, 7, 28, 90, and 180 days. These tests were performed in

accordance with BS 1881 (Part 116 1983) [26].

The water absorption test was carried out on pastes and concrete cubes in accordance

with the method specified by RILEM-CPC-11.2 (RILEM 1994) [27] at 180 days of age.

The Linear correlation between initial water absorption and square root of time was

obtained. The tangent of the line i.e. the absorption factor was calculated to compare the

effect of the admixtures on the pore structures of concrete and paste samples.

3. Experimental results and discussion

The results of the compressive strength tests for paste mixes are shown in Figs. 1 to 3

and for concrete mixes in Figs. 4 and 5. Reduction in compressive strength of paste and

concrete specimens exposed to laboratory simulation ponds and site tidal zone

compared with corresponding control strengths are plotted in Figs. 6 to 8. Water

absorption measurements were taken at 3, 6, 24 and 72 hours and the absorption

equations for each mix are given in Tables 5 and 6 for pastes and concretes respectively.

The calculated absorption factors are shown in Figs. 9 and 10 for paste and concrete

mixes respectively.

All of the paste and concrete mixes cured in fresh potable water (control curing) show

an increase in compressive strength with age as expected (Figs. 1 and 4). The strength

increases sharply up to 28 days and rate of increase slows at later ages. In both paste and

concrete mixes, the optimum S.F. mix exhibited the highest strength development as

expected due to the superior pozzolanic activity and pore refinement process of S.F.

The blended GGBS/SF mix gave the lowest strength development. This was due to the

lower portland cement contents in these mixes and the slower rate of reaction of GGBS.
In general, as expected, the S.F mixes show higher strength development than type II

cement in control curing condition. This is in agreement with the absorption test results

in Figs. 9 and 10, which show lower absorption factors for the control curing specimens.

However, this strength development trend changes in other exposure conditions for

paste and concrete. In simulation pond exposure, portland cement type II exhibits

greater compressive strength development and less loss (Figs. 2 and 5). In general after

initial 7 days curing the strength gain continues up to 28 days for all paste and concrete

mixes in both exposure conditions but the harmful effect of environmental condition

and seawater chemicals causes strength loss in pastes (Fig. 2) and slower rates of

strength gain in concretes (Fig. 5). It can be postulated that the reason for further

strength gain in concrete specimens (compare to loss of strength of paste specimens) is

attributed to the bigger size of concrete specimens as the volume-related ion-

intrusion/attack to the specimens would be less than for smaller paste specimens that are

more sensitive to ion attack and induced cracks. In other words, the intrusion of ions

throughout the bigger concrete specimens would not be deep enough to have a major

influence on the compressive strength of samples when compared to paste specimens.

Thus, further strength loss should be expected at later ages for concrete specimens in

these exposure conditions.

Portland cement type II mixes exposed to laboratory simulation ponds or site exposure

tidal zone exhibit lower percentage strength loss (see Figs. 6 to 8). Figs. 2, 5 and Figs. 6

to 8 show that S.F. paste and concrete mixes have greater strength loss compared to type

II portland cement mixes in tidal zone and simulation pond exposure conditions, in

contrast to their control curing condition in which S.F. mixes exhibit highest strength

development. The greater intensity of the attack on portland cement plus silica fume

specimens as compared to the portland cement specimens may be due to the absence of
magnesium hydroxide (Brucite) in the portland cement plus silica fume mixes.

Magnesium hydroxide, formed as a result of calcium hydroxide reacting with

magnesium sulphate, is relatively insoluble in water and therefore blocks the pores and

protects the C-S-H gel from further attack. Its absence in portland cement plus silica

fume specimens, therefore, makes the C-S-H gel more prone to magnesium sulphate

attack. In addition, cyclic wetting and drying exposure will affect the crystallization of

salts and results in degradation of the protective layer, facilitating the ingress of ions.

The greater intensity of the attack may also be the result of the formation of an

additional amount of C-S-H gel due to the pozzolanic reaction (‘pozzolanic C-S-H gel’),

which is different in composition from the C-S-H gel produced by the hydration of alite

and belite of ‘portland cement C-S-H gel’ with respect to lower density and fewer

hydrated molecules of water as explained by Cohen et al. [28] and Kjellsen et al. [29].

As shown in Table 5, replacement of cement with silica fume in pastes exposed to

simulation ponds and site tidal zone has made the gradient coefficient of capillary

equation to be increased by square root of time elapsed. This increase depends on the

amount of silica fume replaced in the mix. The higher the amount of silica fume, the

higher the gradient coefficient of absorption equations (see Figs. 9 and 10). This finding

is in accordance with the research results and mechanism developed by Kjellsen [29]. It

should be mentioned that the effect of amount of silica fume on absorption factor of

concrete samples was not as distinctive as of paste mixes. This may be due to the size

effect and less cement matrix in concrete specimens. Moreover, longer exposure of

these specimens will lead to greater absorption factor in these specimens. In addition,

the small accessible pore volume of these concrete specimens due to salt crystallization

can be another explanation to the above observation. These figures also show that the

absorption factors of specimens in simulation ponds are lower than corresponding

specimens in potable water. This is due to partial blockage of pores as a result of the salt

crystallization in the pores network. Furthermore, using GGBS in the mix causes the

absorption factor to be profoundly increased in simulation ponds and tidal zone

exposure conditions. Therefore it can be concluded that the pore structures of GGBS

plus SF mixes under wetting and drying condition in deleterious solutions are more

porous than others. GGBS is considered to be effective in refining of the pore structure

and improving the durability of concrete against sulphate attack; however, the higher

percentage of GGBS, which is a slow-reacting cementitious material (especially the

GGBS used in this study with lower fineness value than ASTM specification), together

with silica fume replacement and insufficient initial curing could make GGBS mixes

more vulnerable to sulphate attack. It is considered that the hydration product at 7 days

when placed in site tidal zone or simulation pond exposure is not dense enough to

prevent the ingress of ions [30]. Therefore, it is imperative to cure the concrete for an

extended period of time when using supplementary cementitious material in wetting and

drying exposures. Moreover, the deterioration of binder matrix due to magnesium

sulphate and decomposition of C-S-H gel to M-S-H, which is non-cementitious, can be

another probable reason for this phenomenon.

These findings are in agreement with the results of other investigators [28–33]. Cohen

and Bentur [28] showed that adding 15% silica fume to cement paste (type I and V)

exposed to 5% sodium sulphate solution resulted in improving its durability, but when

exposed to magnesium sulphate solution a major deterioration resulted due to softening

of C-S-H and converting it to hydrated magnesium silicate with loss of strength and

mass. With the presence of silica fume, the decrease in strength observed was 5 to 10

times larger than in pastes without this pozzolan [28]. Likewise, Hekal et al. [31]

reported that partial replacement of portland cement with 10 to 15% of silica fume did
not show a significant improvement in sulphate resistance of hardened cement pastes.

The decreased amount of magnesium hydroxide, due to its reaction with hydrated

silicate, and also the formation of M 4 SH 8.5 which is non-hydraulic, increase the

exposure of cementitious C-S-H to aggressive components. In the same way, the

durability study on mortar specimens carried out by Moon et al. [32] and Lee et al. [33]

confirmed that the conversion of secondary C-S-H gel into M-S-H (or M-C-S-H) gel,

which results in the softening of cement matrix, eventually led to the pronounced

deterioration of silica fume blended mixes exposed to magnesium sulphate

environment. Magnesium silicate formation in concrete exposed to seawater was

reported by Cole [34] and by Roy et al. [35] who illustrated that the formation of Mg-

rich phases is normally associated with the near-surface regions of concrete exposed to

Mg-containing solutions such as seawater. Furthermore, investigation carried out by

Nehdi et al. [36] indicated that decalcification of CSH was significant in presence of

magnesium sulphate and silica fume did not show any additional benefits in resisting

surface deterioration and even led to worse performance than OPC specimens. Also

Brown and Doerr [37] established the mechanism by which Mg is transported

throughout the pore structures of site concretes even though magnesium hydroxide is

known to be highly insoluble in pore solution of concrete. Further studies are needed to

determine the behaviour and mechanism of deterioration of ternary blended mixes.

In summary, the results discussed in this section clearly demonstrate the complex effect

of silica fume on the durability of cement mixes in seawater and under wetting and

drying conditions. The presence of S.F. is beneficial in normal water, or other solutions

where the deterioration mechanism is the result of expansion and believed to be

associated with ettringite formation. In seawater solutions with high concentration of

magnesium sulphate, a major cause for attack is decomposition of C-S-H; in this case a

silica fume addition (and GGBS) becomes detrimental.

4. Conclusions

The results of an investigation into potential mixes for use in tidal zone marine

structures in Persian Gulf waters revealed the detrimental effect of silica fume. The

specific conclusions that can be drawn from this study are as follows:

1. In potable water, the S.F. mix exhibits the highest strength development, while the

blended ternary GGBS - SF mix exhibits the lowest strength development.

2. In laboratory simulation ponds and site tidal zone exposures, type II portland cement

exhibits better compressive strength development. Type II portland cement mixes

exhibit lower percentage of strength loss than S.F. mixes or blended ternary S.F. -

GGBS mixes.

3. The strength loss and the capillary absorption factors show that in general the higher

the amount of silica fume, the higher the strength loss and the higher the absorption

factor in simulation and tidal zone exposure conditions.

Furthermore, using GGBS in the mix causes the absorption factor to be profoundly

increased in simulation ponds and tidal zone exposure conditions. However, for

concrete samples stored in simulation pond, SF and GGBS had less influence in the

absorption factor than the paste mixes.

4. This study indicates that the effect of silica fume in various seawater environments

is not clear-cut. There is a need for a thorough understanding of its influence on the

different deterioration mechanisms to predict its potential for improvement in

durability. For a long-term durable performance of concrete in marine and offshore

tidal zone structures, S.F. or GGBS cement can be used in low magnesium sulphate
 concentration waters. To further enhance the resistance against deterioration by

sulphate ions and / or salt crystallisation, additional protective measures, such as the

application of a water-resistant epoxy based coating, may be needed.

5. It is imperative to cure the concrete for an extended period of time when using

supplementary cementitious material in wetting and drying exposures.

Acknowledgments

The present research described has been carried out in the concrete technology

laboratory in department of Civil Engineering, K.N.T. University of Technology,

Tehran, Iran. The authors express gratitude to the technicians and the university for the

support rendered throughout the research program. The support and site exposure

security provided by Kish University of Iran is appreciated.

References

[1] Mather, B. Effects of seawater on concrete. Highway Research records, PT 113,

1966, 33-43.

[2] Wood, H. Durability of concrete construction. ACI Monograph No. 4, ACI/Iowa State

Univ. Press, Detroit, 1968.

[3] Gjorv, O. E. Long-term durability of concrete in seawater. In: ACI Journal

Proceedings 1971(68):60-67.

[4] Mehta, P.K. Durability of concrete in marine environment – a review, Performance of

concrete in marine environment. ACI SP-65, American Concrete Institute, Detroit, 1980,

1-20.

[5] Conjeaud, M. L. Mechanism of seawater attack on cement mortar. ACI SP-65

1980:39-61.

[6] Regourd, M. Physico-chemical studies of cement pastes, mortars and concrete

exposed to seawater. ACI SP-65 1980:63-82.

[7] Eglinton, M. S. Concrete and its chemical behaviour. Thomas Telford Ltd., London,

1987.

[8] Gerwick, B. C. International experience in the performance of marine concrete.

Concrete International 1990(5):47-53.

[9] Hoff, G. C. Durability of offshore and marine concrete structures. ACI SP-126

1991(1):33-64.

[10] Al-Amoudi, O. S. B. Durability of plain and blended cements in marine

environments. Adv Cem Res 2002;14(3):89-100.

[11] Al-Ghamdi, H. A. Effect of curing and mix design on durability of Portland cement

and Portland cement – silica fume cement mortars in a hot-marine environment. PhD

thesis, Dept of Chemistry, University of Aberdeen, UK, 1999.

[12] Ghoddousi, P., Ganjian, E., Parhizkar, T. and Ramezanianpour, A. Concrete

technology in the environmental conditions of the Persian Gulf. Building and Housing

research Centre Publication No. B 283, Tehran, Iran, 1998.

[13] Wee, T. H., Suryavanshi, A.K., Wong, S. F. and Rahman, A. K. Sulfate resistance

of concrete containing mineral admixtures. ACI Materials Jour 2000;97:536-49.

[14] Mehta, P.K. Mechanisms of expansion associated with ettringite formation. Cem

Concr Res 1973;3(1):1-6.

[15] Suryavanshi, A.K., Scantlebury, J.D., Lyon, S.B. Pore size distribution of OPC and

SRPC mortars in presence of chlorides. Cem Concr Res 1995;25:980-988.

[16] Neville, A.M. Properties of Concrete. England: Fourth Edition, Pearson Education

Limited Press, 2002, p.76, 569.

[17] Mehta, P.K., Monteiro, P.J.M. Concrete Structure, Properties, and Materials. USA:

New Jersey, Second Edition, Prentice Hall Inc., 1993, 160-170.

[18] Al-Dulajian, S.U., Maslehuddin, M., Al-Zahrani, M.M., Sharif, A.M., Shameem,

M., Ibrahim, M. Sulfate resistance of plain and blended cements exposed to varying

concentration of sodium sulfate. Cem Concr Compos 2003;25:429-437.

[19] Habibi, M. K. Performance of Silica Fume concrete in the Persian Gulf condition.

MSc. thesis, Dept. of Civil Eng., K.N.T. Univ. of Tech., Tehran , Iran, 1997.

[20] ACI committee 226. Silica fume in concrete. ACI Materials Jour 1987;84:158-66.

[21] ACI committee 232. Natural pozzolans in concrete. ACI Materials Jour

1994;91:410-426.

[22] Goldman, A. and Bentur, A. Bond effects in high strength silica fume concretes.

ACI Materials Jour 1989;86:440-7.

[23] Malhotra, V.M., Carette, G.G. and Sivasundaram, V. Role of silica fume in concrete:

a review. In: Proceedings International Conference on Advances in Concrete Technology,

CANMET, Ottawa, Canada, 1992.

[24] Tamimi, A. K. High-performance concrete mix for an optimum protection in acidic

conditions. Mater Struct 1997;30:188-91.

[25] Ganjian, E., Ghodosi, p. and Parhizkar, T. Mechanical properties of concrete

containing Iranian Silica Fume. In: 4th Int. Conf. on Civil Eng., Sharif Univ. of

Technology, Tehran, Iran 1997;3:111-123.

[26] British Standards Institution BS 1881, Part 116. Recommendations for the

measurement of compressive strength, 1983.

[27] RILEM, “CPC-11.2. Absorption of water by concrete by capillarity. TC14-CPC,

1982, RILEM Technical Recommendations for the Testing and Use of Construction

Materials, International Union of Testing and Research Laboratories for Materials and

Structures, E & FN Spon, 1994.

[28] Cohen, M.D. and Bentur, A. Durability of Portland cement–silica fume pastes in

magnesium sulfate and sodium sulfate solutions. ACI Materials Jour 1988;85(3):148–

157.

[29] Kjellsen, K.O., Atlassi, E.H. Pore structure of cement silica fume systems Presence

of hollow-shell pores. Cem Concr Res 1999;29(1):133-142

[30] Toutanji, H., Delatte, N., Aggoun, S., Duval, R., Danson, A. Effect of

supplementary cementitious materials on the compressive strength and durability of

short-term cured concrete. Cem Concr Res 2004;24:311-319

[31] Hekal, E.E., Kishar, E., Mostafa, H. Magnesium sulphate attack on hardened

cement pastes under different circumstances. Cem Concr Compos 2002;32:1421-1427

[32] Moon, H.Y., Lee, S.T., Kim, S.S. Sulphate resistance of silica fume blended

mortars exposed to various sulphate solutions. Can Jour Civil Eng 2003;30(4):625-636.

[33] Lee, S.T., Moon, H.Y, Swamy, R.N. Sulfate attack and role of silica fume in

resisting strength loss. Cem Concr Compos 2005;27:65-76.

[34] Cole, W.F. A crystalline hydrated magnesium silicate formed in the breakdown of

concrete sea-wall. Nature 1953;171:354-355.

[35] Roy, D.M., Sonnenthal, E., Prave, R. Hydrotalcite observed in mortars exposed to

sulfate solutions. Cem Concr Res 1985;15:912-916.

[36] Nehdi, M., Hayek, M. Behavior of blended cement mortars exposed to sulfate

solutions cycling in relative humidity. Cem Concr Res 2005;35:731-742

[37] Brown, P.W., Doerr, A. Chemical changes in concrete due to the ingress of

aggressive species. Cem Concr Res 2000;(30):411-418.

Figure Captions:

Fig. 1- Compressive strength development with age for paste mixes cured in fresh

potable water (control curing).

Fig. 2 - Compressive strength development with age for paste mixes exposed to

simulation ponds.

Fig. 3 - Compressive strength development with age for paste mixes exposed to site

tidal zone.

Fig. 4 - Compressive strength development with age for concrete specimens cured in

fresh potable water (control curing).

Fig. 5 - Compressive strength development with age for concrete specimens exposed to

simulation ponds.

Fig. 6 - Reduction in compressive strength of pastes exposed to simulation ponds.

Fig. 7- Reduction in compressive strength of concretes exposed to simulation ponds.

Fig. 8 - Reduction in compressive strength of pastes exposed to site tidal zone.

Fig. 9 - Absorption factors for pastes exposed to different conditions.

Fig. 10 - Absorption factors for concretes exposed to different conditions.

 110

100

90

80
Strength (MPa)

70

60

50
P-type II
40 P-type II+7%SF
P-type II+10%SF
P-type II+10%SF+50%GGBS
30
0 20 40 60 80 100 120 140 160 180 200
Curing period (days)

Fig. 1- Compressive strength development with age for paste mixes cured in fresh

potable water (control curing).

110

100

90

80
Strength (MPa)

70

60

50

40 P-type II
P-type II+7%SF
30 P-type II+10%SF
P-type II+10%SF+50%GGBS
20
0 20 40 60 80 100 120 140 160 180 200
Exposure period (days)

Fig. 2 - Compressive strength development with age for paste mixes exposed to

simulation ponds.
 110

100

90

80
Strength (MPa)

70

60

50
C-type II
40
C-type II+7%SF
30 C-type II+10%SF
C-type II+10%SF+50%GGBS
20
0 20 40 60 80 100 120 140 160 180 200
Curing period (days)

Fig. 3 -Compressive strength development with age for paste mixes exposed to site

tidal zone.

110

100

90

80
Strength (MPa)

70

60

50
C-type II
40
C-type II+7%SF
30 C-type II+10%SF
C-type II+10%SF+50%GGBS
20
0 20 40 60 80 100 120 140 160 180 200
Curing period (days)

Fig. 4- Compressive strength development with age for concrete specimens cured

in fresh potable water (control curing).

 110

100

90

80
Strength (MPa)

70

60

50
C-type II
40 C-type II+7%SF
C-type II+10%SF
30
C-type II+10%SF+50%GGBS
20
0 20 40 60 80 100 120 140 160 180 200
Exposure period (days)

Fig. 5– Compressive strength development with age for concrete specimens

exposed to simulation ponds.

50

40

30
Strength loss (%)

20

10

-10 P-type II
P-type II+7%SF
-20 P-type II+10%SF
P-type II+10%SF+50%GGBS
-30
7 28 49 70 91 112 133 154 175
Exposure period (days)

Fig. 6- Reduction in compressive strength of pastes exposed to simulation ponds.

 25

20

15

10
Strength loss (%)

-5

C-type II
-10
C-type II+7%SF

-15 C-type II+10%SF

C-type II+10%SF+50%GGBS
-20
7 28 49 70 91 112 133 154 175
Exposure period (days)

Fig. 7- Reduction in compressive strength of concretes exposed to simulation

ponds.

30

20

10
Strength loss (%)

-10
P-type II
-20 P-type II+7%SF

P-type II+10%SF
-30
7 28 49 70 91 112 133 154 175
Exposure period (days)

Fig. 8- Reduction in compressive strength of pastes exposed to site tidal zone.

 0.3
A=P-type II

B=P-type II+7%SF

0.25 C=P-type II+10%SF

D=P-type II+10%SF+50%GGBS
Absorption Factor

0.2

0.15

0.1

0.05

0
A B C D A B C D A B C D

POTABLE WATER SIMULATION PONDS SITE TIDAL ZONE

Fig. 9- Absorption factors for pastes exposed to different conditions.

Fig. 10- Absorption factors for concretes exposed to different conditions.

Table 1 - Chemical analysis and physical properties of cement, silica fume (SF) and
ground granulated blast furnace slag (GGBS) used.

Chemical composition (%) Cement SF GGBS

SiO 2 21.7 91.7 33.03

Al 2 O 3 4.9 1.27 10.87
Fe 2 O 3 3.2 1.45 0.25
CaO 62.4 1.69 38.45
MgO 3.58 0.6 10.26
SO 3 1.84 0.45 <0.1
Na 2 O 0.35 --- 0.65
K2O 0.95 --- 0.78
Cl 0.017 0.08 0.007
Loss On Ignition 1 2 0.25
Physical properties
Relative Density 3.15 2.2 2.5
Specific surface (m2/kg) 290.8 14 3103 200
Compound composition (%)
C3S 46.3 --- ---
C2S 27.1 --- ---
C3A 7.6 --- ---
C 4 AF 9.7 --- ---
Table 2 - Chemical analysis of the water used for mixing, initial curing, control
curing, Kish island site exposure seawater and simulation ponds exposure

Tap water Kish island seawater

Constituent
ppm ppm
Na 173 11400
K 6 397
Ca 24 450
Mg 16 1600
Cl 372 22330
SO 4 23 3070
 Table 3 - Characteristics of the pastes made.

Superplasticizer Flow (%)

Paste mix W/B
(%)
P-type II 0.2 1.5 63 ± 5

P-type II+7%SF 0.2 1.8 64 ± 5

P-type II+10%SF 0.2 2 65 ± 5

P-type II+10%SF+50%GGBS 0.2 2 65 ± 5

 Table 4 - Concrete mix proportions made (1 m3)
Super-
Cement SF GGBS Sand Gravel Water W/B Approximate
Concrete mix plasticizer
Slump
(kg) (kg) (kg) (kg) (kg) (kg) (ratio) (kg) (mm)
C-type II 400 - - 736 1104 160 0.4 4.8 20

C-type II+7%SF 372 28 - 728 1104 160 0.4 5.2 20

C-type II+10%SF 360 40 - 723 1104 160 0.4 6 20

C-type II+10%SF+50%GGBS 160 40 200 680 1104 160 0.4 6 20

 Table 5 - Results of absorption by capillary test for pastes

(I=absorption percent, x=√t).

Capillary Correlati Exposure

Paste mix absorption on condition

equation coefficie

nt

P-type II I=0.3003x+0.881 0.92 Potable water

P-type II+7%SF I=0.1607x+0.492 0.90 Potable water

P-type II+10%SF I=0.131x+0.484 0.93 Potable water

P-type II+10%SF+50%GGBS I=0.1849x+0.0581 0.98 Potable water

P-type II I=0.0701x+0.0824 0.99 Simulation ponds

P-type II+7%SF I=0.1301x+0.1261 0.99 Simulation ponds

P-type II+10%SF I=0.1428x+0.228 0.99 Simulation ponds

P-type II+10%SF+50%GGBS I=0.1945x+0.2133 0.98 Simulation ponds

P-type II I=0.1231x+0.4368 0.94 Site tidal zone

P-type II+7%SF I=0.1275x+0.4791 0.94 Site tidal zone

P-type II+10%SF I=0.1539x+0.4792 0.96 Site tidal zone

P-type II+10%SF+50%GGBS I=0.2077x+0.4982 0.96 Site tidal zone

Table 6 - Results of absorption by capillary test for concretes

(I=absorption percent, x=√t).

Capillary Correlation Exposure

Concrete mix absorption coefficient condition

equation

C-type II I=0.1762x+0.908 0.97 Potable water

C-type II+7%SF I=0.1136x+1.1037 0.96 Potable water

C-type II+10%SF I=0.1303x+0.9177 0.96 Potable water

C-type II+10%SF+50%GGBS I=0.1507+0.5141 0.97 Potable water

C-type II I=0.0399x+0.2396 0.91 Simulation ponds

C-type II+7%SF I=0.0292x+0.197 0.97 Simulation ponds

C-type II+10%SF I=0.0489x+0.2802 0.99 Simulation ponds

C-type II+10%SF+50%GGBS I=0.0521x+0.2942 0.98 Simulation ponds