Fracture Mechanics of Concrete and Concrete Structures -

Assessment, Durability, Monitoring and Retrofitting of Concrete Structures- B. H. Oh, et al. (eds)
ⓒ 2010 Korea Concrete Institute, Seoul, ISBN 978-89-5708-181-5

Chloride penetration resistance properties of concrete containing mineral

admixtures
Jaekang Yoo, Bohwan Oh, Sangjoon Park & Seongwoon Kim
Daewoo Institute of Construction Technology, Suwon-si, Gyeonggi-do, South Korea

ABSTRACT: This paper describes the chloride penetration resistance properties of concrete containing min-
eral admixtures including fly ash, ground granulated blast-furnace slag powder, silica fume, and meta-kaolin.
An electrically accelerated chloride migration test and immersion test were performed for 91days and
330days, respectively. Total charge passed and chloride migration coefficient were derived from the acceler-
ated chloride migration test, and chloride penetration depth, water and acid soluble chloride content, and ap-
parent chloride diffusion coefficient were measured from the immersion test. The initial current has a linear
relationship with the total charge passed, and it was found that the binding capacity increased as the distance
from the concrete surface exposed to chloride solution increased. The effect of types of mineral admixture
and their replacement ratios on enhancing the concrete’s resistance against chloride ion penetration could be
clarified through the electrically accelerated chloride migration test and immersion test.

ing a reinforced concrete structure that is meant to

1 INTRODUCTION
One of the main causes of premature deterioration of
reinforced concrete structures exposed to salt
laden environments is chloride ingress and the re-
sulting corrosion of reinforced steel. The chloride
permeability of concrete is one of the important
properties that need to be assessed in the design of
structures to be built in coastal areas. The penetra-
tion capacity of chloride ions depends on the mi-
cro-structure of hardened concrete, such as the
continuity of pore paths and micro cracks. There-
fore, durable structures are made from dense con-
crete by increasing the resistance against chloride
invasion and restricting the formation of harmful
cracks through proper measures regarding the ma-
terial and construction processes (Boulfiza et al.
2003).
The BS 8500 (2006), ACI 318 (2008), and
KCSS (2003) codes suggest guidelines for mix
conditions, including a cementitious material sys-
tem for concrete exposed to chloride attack envi-
ronments. Moreover, the ACI 201 committee
(2000) reports that the use of several pozzolans
positively influences the durability of structures by
decreasing the permeability of concrete to as low
as 1/10th or 1/100th of that of comparable concrete
of equal strength made without pozzolans. It is
clear that choosing the optimum composition of
cementitious material is a key issue when design-
last for a long time in a chloride environment.
Many researchers have also developed test
methods to measure the chloride diffusion coeffi-
cient under steady state and non-steady state condi-
tions (Page 1981; Dhir et al. 1990; Torii et al.
1997; Tang & Nilsson 1992). The test method un-
der steady state is considerably time consuming,
often requiring months or years to obtain results.
As such, it is not suitable for a large variety of
concrete mixtures. A rapid test method for the
chloride permeability of concrete, developed by
Whiting (1981), is available and has been adopted
as a standard test method of AASHTO T 277
(2005) and ASTM C 1202 (2005). Whereas these
methods provide qualitative comparisons such as
total charge passed (coulomb), Tang’s method
(1996) provides chloride diffusivity based on the
theoretical relationship between diffusion and mi-
gration in the calculation of the chloride diffusion
coefficient from the accelerated measurement.
However, it has been pointed out that the pore
solution chemistry may affect the conductivity of
concrete containing mineral admixtures in an elec-
trically accelerated chloride migration test (Shi et
al. 1998). Also, compared to the natural immersion
regime, the binding effect is not taken into account
in the electrically accelerated test (Jiang et al.
2004; Tang 1996).
This paper investigates the effects of mineral
admixtures on the resistance against chloride ion
penetration of concrete containing fly ash, ground
granulated blast-furnace slag powder, silica fume, 2.2 Materials J = − D ( h , T ) ∇h
and meta-kaolin, respectively.
The physical properties and chemical compositions
of normal Portland The cementproportionality coefficient
(NPC) and mineral ad- D(h,T)
moisture permeability
mixtures used in this study are listed in Table 3.and it is a nonlinea
2 EXPERIMENTAL PROGRAM of the relative humidity
NPC corresponds to the standard composition h and temperature
requirements of & NajjarC1972).
ASTM 150 asThe Typemoisture mass balanc
I cement.
2.1 Experimental plan that the variation
The FA was ASTM Class F type with a SiO in time of the water
2
mas
As shown in Table 1, in the experimental plan, the volume of concrete (water
+ Al 2 O 3 + Fe 2 O 3 content of 89.1%; the specific content w ) be eq
water to cementitious materials ratio (W/Cm) of gravity was 2.15. divergence of the moisture flux J
concrete mixed with and without mineral admixture The GGBS corresponds to the physical require-
(plain) ranged from 0.40 to 0.55. The content of fly ments of ASTM C− 989. ∂w = ∇The• J SO3 content was 0.98%
ash (FA) was 10, 20, and 30% by mass of total cemen- and specific gravity ∂was
t 2.85.
titious materials as cement replacement. Ground The SF, dried and un-densified product, corre-
granulated blast-furnace slag powder (GGBS) was re- sponds to the ASTM TheCwater
1240.content w can be
It contained expressed a
91.2%
placed with ratios of 30, 50, and 70%. The replace- of the evaporable
SiO2, and had a specific gravity of 2.20. water w e (capillary wa
ment ratio of silica fume (SF) and meta-kaolin (MK) The chemical vapor,
component and adsorbed
of the MK water) and the non-e
was similar
ranged from 5 to 20% in four steps with equal grade. to that of the FA,(chemically
except the Al2bound)
O3 contents,waterand thewn (Mil
Table 2 shows the mix proportions and consis- Pantazopoulo
specific gravity was 2.59. & Mills 1995). It is reas
tency results of plain concrete. assume (Blaine)
The specific surface that the of evaporable
NPC, FA,water and is a fu
GGBS ranged from relative humidity,
300~400m 2
, degree
/kg, hand that of of SFhydration
degree of silica fume reaction, αs, i.e. we=w
Table 1. Experimental plan.
= age-dependent sorption/desorption
W/Cm Remark
Mineral admixture and replacement ratio (Norling Mjonell 1997). Under this assum
Type by substituting Equation
Replacement ratio (%)1 into Equati
NPC - obtains -
0.40
FA Fly ash 10, 20, 30
0.45
GGBS Ground granulated blast-furnace slag powder ∂w 30, 50, 70
∂h 5, 10, 15, 20 ∂w ∂w
0.50
SF Silica fume − e + ∇ • ( D ∇h ) = eα&c + e α&s + w
0.55
MK Meta-kaolin ∂h ∂t 5, 10, 15,h 20 ∂α ∂α
c s
0.40FA10: water to cementitious materials ratio, types of mineral admixture, replacement ratio
where ∂we/∂h is the slope of the sorption/
Table 2. Mix proportions of plain mixture. isotherm (also called moisture capac
W/Cm S/a (%)
Unit weight (kg/m3) governing equationSlump(Equation
Air 3) must be
Water Cement Sand Gravel by appropriate
SP *) boundary
(mm) and
content initial
(%) conditi
0.40 45.6 158 395 793 954 The relation180between3.1the amount of e
2.0
0.45 46.7 158 351 829 954 water
1.8 and relative
180 humidity
5.8 is called ‘‘
0.50 47.7 158 316 861 951 isotherm”
1.6 if measured
170 4.2with increasing
0.55 47.8 164 298 863 949 humidity
1.5 and ‘‘desorption
140 4.8 isotherm” in th
*) SP: super-plasticizer case. Neglecting their difference (Xi et al.
the following, ‘‘sorption isotherm” will be
Table 3. Physical properties and chemical compositions of cementitious materials. reference to both sorption and desorption c
Mark
Specific Blaine Ig. loss Chemical composition (%)By the way, if the hysteresis of the
gravity (m2/kg) (%) SiO2 Al2O3 isotherm
Fe2O3 would CaObe taken MgOinto account, two
NPC 3.14 329 0.99 22.0 4.6 relation,
3.2 evaporable
63.1 water2.1vs relative humi
FA 2.15 316 3.57 56.4 23.7 be9.0used according
2.5 to the 1.3
sign of the varia
GGBS 2.85 386 0.27 33.7 14.4 relativity
0.4 humidity.
41.7 The6.4 shape of the
SF
MK
2.20
2.59
22,800
1085
2.10
2.13
91.2
52.0
1.3
40.0
isotherm
0.8 for HPC
0.7 is influenced
0.3 by many p
3.0 1.2 0.5
especially those that influence extent and
Table 4. Physical properties of fine and coarse aggregate.
chemical reactions and, in turn, determ
structure and pore size distribution (water-
Kind Specific gravity Absorption (%) ratio,
Fineness cement Maximum
modulus chemicalsizecomposition,
(mm) SF
Fine aggregate 2.60 1.20 2.60 curing time and 5 method, temperature, mix
Coarse aggregate 2.63 0.53 6.96 etc.). In the literature
25 various formulatio
found to describe the sorption isotherm
Table 5. Physical properties of super-plasticizer. concrete (Xi et al. 1994). However, in th
Kind Specific gravity Solid contents (%) paper the semi-empirical
Base material Color expression pro
Super-plasticizer 1.20±0.02 40±2.0 Norling Mjornell
Naphthalene (1997) is adopted b
Dark brown

Proceedings of FraMCoS-7, May 23-28, 2010

 D (h,was
J = −MK
and T )∇h22,800m2/kg and 1085m2/kg as deter- (1) explicitly accounts
After curing for 28fordays,
the epoxy
evolution of hydration
was coated on all
mined by the nitrogen adsorption method. reaction
of and SF
the surfaces content.
except This (100×400mm)
two sides sorption isotherm
at a
The proportionality
Table coefficient
4 shows the physical D(h,T)ofiswashed
properties called reads angle to the surface concrete poured. The
right
moisture permeability and it is a nonlinear
sand and crushed gravel, and Table function
5 shows the specimens were then immersed in 3.6% NaCl solu-
of the relative
physical humidity
properties and temperature
of ah naphthalene T (Bažant
based super- tion for 330 days. After ⎡ a specified amount ⎤ of
& Najjar 1972). The moisture mass balance requires
plasticizer used in this study. elapsed time, the chloride
we (h, α c , αby G1 (colorimetric⎢ penetration front
⎥ was
s ) =the α c , α s )⎢1 − +
1

that the variation in time of the water mass per unit measured method
10(g α
∞ and powdered
⎥
volume of concrete (water content w) be equal to the samples were extracted from ⎢
⎣ c − α c )hat⎥⎦ regular
e the1concrete (4)
2.3 Measurements
divergence of the moisture flux J 5mm intervals from the end exposed to
⎡ 10(g α ∞ − α )h
the chloride
⎤
The specimens for a compressive strength test and solution. Water soluble ⎢and acid soluble
1 c c chloride
an∂welectrically accelerated chloride migration test content was measured
K1 (α c , α sin) eaccordance with− ASTM1⎥
C
⎢ ⎥
− =cast
were ∇ • J in a cylinder of ø100×200mm. Prism (2) 1218 and 1152, respectively. ⎣ Based on the ⎦chloride
∂t
shaped specimens (100×100×400mm) were also cast profile (total chloride content), the apparent chloride
for The
an water
immersion
contenttestw can (nonbesteady-state
expressed asdiffusionthe sum where thecoefficient
diffusion first termwas (gelcalculated
isotherm) byrepresents
Fick’s Sec- the
test). After casting, the specimens
of the evaporable water we (capillary were sealed
water, with
watera physically
ond Law. bound (adsorbed) water and the second
poly-ethylene sheet and cured in a
vapor, and adsorbed water) and the non-evaporable constant tempera- term
The(capillary
test flow isisotherm)
shown in represents
Figure 1. the capillary
ture room withbound)
(chemically 23±2℃.water The specimens
wn (Millswere1966, de- water. This expression is valid only for low content
molded after 1 day at RH 95%
Pantazopoulo & Mills 1995). It is reasonable to and cured on satu- of SF. The coefficient G1 represents the amount of
rated
assume lime water
that the ofevaporable
23±2℃. water is a function of 3water per unitAND
RESULTS volume held in the gel pores at 100%
DISCUSSION
relative humidity, h, the
The test set-up for migration
degree test corresponds
of hydration, αc, and relative humidity, and it can be expressed (Norling
to ASTM
degree C 1202.
of silica fumeA reaction,
potentialαof, i.e.
s
30 wvolts DC was
e=we(h,αc,αs)
Mjornell
3.1 Initial1997) as and total charge passed
current
applied for the test tosorption/desorption
= age-dependent minimize changes inisotherm the test Figure 2 shows the total charge passed through the
regime such as the temperature of
(Norling Mjonell 1997). Under this assumption and the test solutions c
G1 (α c , α s according
concrete ) = k α c + to
s the
k vg α s mineral admixture type (5)
and current. Because
by substituting Equationthe applied
1 intopotential
Equationwas2 dif- one vg
and the replacement ratio.
c s
ferent
obtainsfrom that of the ASTM method (60V), the to- The total charge passed through the plain con-
tal charge passed derived from this study does not crete kcvg and kwith
wheredecreased s
vg areage material
in a rangeparameters. From the
of 2300~3300C
correspond
∂w ∂h
to the ASTM ∂w
reference.
∂w
The non-steady- maximum
at 7 days and amount of water at
1600~2400C per28unit volume
days. that can
Moreover, it
state
− e chloride migration
+ ∇ • ( D ∇h ) = e α&c +
coefficient e &s + w&calculated
αwas (3) decreased to 1200~2000C and 1000~1600C at one
fill all pores (both capillary pores and gel pores), 56
h t h
from the average temperature α α
of the n
can 91calculate K1 as one obtains
∂ ∂ ∂
c ∂
s anolyte solu- and days, respectively (data not shown in the fig-
tion, the penetration depth of the chloride ion ure).
(shown
whereand ∂wby /∂h
spraying
is the aslope
0.1Mofsilver nitrate acid solu-
the sorption/desorption ⎜ g α − α the ⎟h ⎥ con-
⎡ ⎛ ⎞ ⎤
e At 7days, the total charge passed 10 through
∞
c c
tion),
isotherm (also called moisture 1,
the test duration (Equation Tang 1996).
capacity). The crete replaced w − with α 30%s αofsFA
⎢
−G ⎢ −e
was
⎝
1
similar to
⎠
that of
0.188
c +0.22
s 1 ⎥

governingRTLequation
xd − a (Equation 3) must be completed (6)
0 1
⎢ ⎥
xd plain concrete
K (α c α s ) = while that of concrete
⎣ with FA10 ⎦ and
=
byDappropriate ⋅ boundary and initial conditions. (1) FA20
,
was slightly lower⎛⎜ gthan
zFU t α ∞ − αthat⎞ of plain concrete.
⎟h
1

The relation between the amount of evaporable As the replacement ratio

10

e ⎝ 1 c c
of FA ⎠ was − increased, the
where, 1

water and relative humidity is called ‘‘adsorption total charge passed was influenced more by the den-
isotherm”
a=2
RTL if measured
⋅ erf −1 (1 − d ) with increasing relativity
2C sityThe material parameters
of hardened concrete, which kcvg and wasksvg and g1ascana
measured
humidityzFUand ‘‘desorption C0 isotherm” in the opposite be calibrated by fitting
compressive strength. However, the totalexperimental data relevant
charge to
case.
D Neglecting
: Diffusiontheir difference
coefficient (Xi et al. 1994), in
(m2/sec) free (evaporable)
passed decreased with water content FA
an increased in replacement
concrete at
the zfollowing,
: Absolute ‘‘sorption
value ofisotherm”
ion valence will(for
be Cl,
usedz with
=1) various
ratio at ages (Di Luzio
28 days, and the & Cusatis
effect of 2009b).
FA replacement
reference to bothconstant
F : Faraday sorption(96,481.04
and desorption
J/Vmol)conditions. increased as increased age.
ByU the: Potential
way, ifdifference
the hysteresis
(V) of the moisture 2.2The total charge passed through the GGBS con-
Temperature evolution
isotherm
R : Gas would be taken
constant into account, two different
(8.314J/Kmol) crete significantly decreased as the replacement ratio
relation, evaporable
T : Solution water vs (K)
temperature relative humidity, must Note
of GGBSthat, wasat early age, since
increased, eventhe at chemical
just 3 days. reactions
Com-
be used accordingthickness
L : Specimen to the sign (m)of the variation of the associated
pared to plain withconcrete,
cement replacement
hydration and of 30%SF reaction
GGBS
relativity humidity.depth
xd : Penetration Theof shape
chlorideofionthe (m) sorption are exothermic,
reduced the totalthe chargetemperature
passed by field is not uniform
30~40% and re-
isotherm for HPC duration
t : Diffusion is influenced by many parameters, for non-adiabatic
placement of 50~70% systemsGGBS even if reduced
the environmental
the total
especially those
erf : Error that influence extent and rate of the
function temperature
charge passedisbyconstant.
60~80%.Heat It is conduction
clear that the cantotal
be
chemical reactions
C d : Chloride ion and, in turn, ofdetermine
concentration pore
color changed described
charge in concrete,
passed through attheleast for temperature
concrete replaced with not
structure
area and pore size distribution (water-to-cement exceeding
GGBS 100°Cinfluenced
was more (Bažant by&theKaplan 1996), and
use of GGBS by
ratio,
C 0 cement
: Chloride chemical composition,
ion concentration SF content,
of cathode cell Fourier’s law,
accordingly its which
replacement reads ratio, than the W/B ratio.
curing time and method, temperature, mix additives,
etc.). In the literature various formulations can be q = − λ ∇T (7)
Measuring age (days)
found to describe the sorption isotherm of normal
concrete
Migration (Xi
test et Casting
al. 1994). However,
1 Water curing 6in the present
28 56
where
91
q is the heat flux, T is the absolute
paper the test
Immersion semi-empirical
Casting 1 Water expression
curing 27proposedImmersed
by temperature,
period 28 and λ is the
91 heat conductivity;
182 in330
this
Norling Mjornell
Figure 1. Test schedule.
(1997) is adopted because it

Proceedings of FraMCoS-7, May 23-28, 2010

 3500
= − D ( h , T ) ∇h
J 7days

Total charge passed(C)

7days 7days
3000
2500
The proportionality coefficient D(h,T)
2000
7days
moisture permeability and it is a nonlinea
1500 of the relative humidity h and temperature
1000
W/B 0.40
W/B 0.45
& Najjar 1972). The moisture mass balanc
500 W/B 0.50 that the variation in time of the water mas
W/B 0.55 volume of concrete (water content w) be eq
0 divergence of the moisture flux J
0 10 20 30 0 30 50 70 0 5 10 15 20 0 5 10 15 20
3500
28days ∂ = ∇•J
− 28days
w
Total charge passed(C)

28days 28days
3000 ∂t

2500
2000 The water content w can be expressed a
1500
of the evaporable water we (capillary wa
vapor, and adsorbed water) and the non-e
1000 (chemically bound) water wn (Mil
500 Pantazopoulo & Mills 1995). It is reas
0 assume that the evaporable water is a fu
0 10 20 30 0 30 50 70 0 5 10 relative
15 20humidity,0 5 h,10degree
15 20 of hydration
Replacement ratio(%) Replacement ratio(%)
degree
Replacement ratio(%) of silicaReplacement ratio(%) αs, i.e. we=w
fume reaction,
= age-dependent sorption/desorption
(a) FA (b) GGBS (c) SF (Norling Mjonell(d)1997). MK Under this assum
by substituting Equation 1 into Equati
Figure 2. Total charge passed for concrete according to the types of mineral admixture and replacement ratio.
The total charge passed through the SF concrete regime such as the obtains
temperature of the test solution
measured at 7 days gradually decreased as the re- and the current, this paper applied 30V for 6 hours
placement ratio of SF was increased and the W/B ra- and the temperature ∂w increases of the test
∂w solution ∂w
e ∂h + ∇10℃ e αtest.
& e α& + w
• ( D during
∇h) = the c +
tio decreased. The concrete with 15% SF showed a were controlled to−be below
∂h ∂t h ∂α ∂ α s
similar total charge passed with the concrete having c
Based on the results of the electrically accelerated s
30% GGBS at 7 days, while the concrete with 5% migration test, the total charge passed could be de-
SF showed lower total charge passed compared to termined with highwhere
accuracy
∂we/∂h by ismeasuring
the slopetheof the sorption/
initial
the concrete with 30% GGBS at 28 days. Based on isotherm (also called moisture
current before finishing the accelerated chloride mi- capac
these results, the resistance of concrete replaced with gration test. governing equation (Equation 3) must be
SF against charge passed was significantly improved by appropriate boundary and initial conditi
4000
during 7~28 days compared to other mixtures. TheY2=relation
20.487Xbetween the amount of e
At 7 days, the total charge passed for the concrete water and0.9969
R = relative humidity is called ‘‘
replaced with MK decreased gradually within a re- 3000 isotherm” if measured with increasing
Total charge passed(C)

placement ratio of 5~10% while that for the con- humidity and ‘‘desorption isotherm” in th
crete replaced with MK exceeding 10% decreased case. Neglecting their difference (Xi et al.
rapidly as the replacement ratio was increased. How-
2000 the following, ‘‘sorption isotherm” will be
ever, at 28 days, the total charge passed decreased reference to both sorption Plainand desorption c
gradually as the replacement ratio was increased. By the way, if the ACA hysteresis of the
Figure 3 shows the relationship between the ini-
1000
isotherm would be taken into account, two
FA
tial current and the total charge passed measured relation, evaporable water GGBS vs relative humi
through the accelerated chloride migration test. The be used according to the SF sign of the varia
initial current showed a linear relationship with the relativity humidity. The MK shape of the
total charge passed in a range from 1.85 to 3500C 0 isotherm for HPC is influenced by many p
and the correlation coefficient was found to be 0 especially
50 those
100 that150 influence200extent and
0.9969, indicating a strong regression. These results chemical reactions and, in turn, determ
Initial current(mA)
agree well with the findings of Feldman et al. (1994) Figure 3. Relationship between initialpore
structure and sizeanddistribution
current total charge (water-
and Zhao et al. (1998). However, Jiang et al. (2004) passed. ratio, cement chemical composition, SF
reported that the relationship between initial current curing time and method, temperature, mix
and total charge passed was linear up to only ap- etc.). Incoefficient
3.2 Chloride migration the literature various formulatio
proximately 2000C. The discrepancy between the found to describe the sorption isotherm
findings in this paper and those of Jiang et al. is at- Figure 4 shows the concrete
chloride (Ximigration
et al. 1994). However,
coefficient of in th
tributed to the difference in the applied voltage and concrete specimens paper the semi-empirical
according to the type of expression
mineral pro
test duration. To minimize the changes of the test admixture and theNorling
replacementMjornell
ratios.(1997) is adopted b

Proceedings of FraMCoS-7, May 23-28, 2010

J = − D (h, T50)∇h explicitly accounts for the evolution of hydration
(1)
7days reaction and SF 7days
content. This sorption7days isotherm
Migration coeff.(×10-12m2/s)
7days
The proportionality coefficient D(h,T) is called
40 reads
moisture permeability
30
and it is a nonlinear function
of the relative humidity h and temperature T (Bažant ⎡ ⎤
& Najjar 1972).
20 The moisture mass balance requires w (h α α ) = G (α α ) − ⎢ ⎥
+ 1
W/B 0.40
that the variation in time ofW/B
the0.45
water mass per unit e c s c s ⎢
,

(g α
∞
,
⎥
, 1

c − α c )h ⎥⎦ (4)
1
10

volume of10concrete (water content

W/B 0.50w) be equal to the
⎢
⎣ e 1

divergence0of the moisture W/B

flux 0.55
J ⎡ (g α
∞ − α )h ⎤
10K (α15 , α 20)⎢e
10

01 c5 c −15 ⎥
0 10 20 30 0 30 50 70 0 5
c s⎢ 10 1 20
− ∂ = ∇ •50J (2)28days ⎥
1
w

∂ 28days 28days ⎣ ⎦
28days
Migration coeff.(×10-12m2/s)

t
40
The water
30
content w can be expressed as the sum where physically
the first term (gel isotherm) represents the
bound (adsorbed) water and the second
of the evaporable water we (capillary water, water term (capillary isotherm) represents the capillary
vapor, and20 adsorbed water) and the non-evaporable water. This expression
(chemically bound) water wn (Mills 1966, of SF. The coefficient isG valid only for low content
1 represents the amount of
Pantazopoulo
10 & Mills 1995). It is reasonable to water per unit volume held
assume that 0
the evaporable water is a function of relative humidity, and it canin be the gel pores at 100%
expressed (Norling
relative humidity, h, degree of hydration, αc, and Mjornell
0 1997)
5 10as 15 20 0 5 10 15 20
degree of silica fume
0
reaction,
10 20
α s, i.e. we=we(h,αc,αs)
30 0 30 50 70
= age-dependent sorption/desorptionReplacement
Replacement ratio(%)
isotherm
ratio(%) Replacement ratio(%) Replacement ratio(%)

(Norling Mjonell 1997). Under this assumption and G (α c ofα smineral

(a) FA (b) GGBS (c) SFc
)= k α c + k α s
s (d) MK (5)
by substituting Equation 1 into Equation 2 one
Figure 4. Chloride migration coefficient of concrete according to the types 1 vg ,
c vg
admixture s
and replacement ratio.
obtains
The migration coefficient of plain concrete de- where
at kcvg and
28 days, are material
ksvg2~7
or was times lowerparameters.
than thatFrom the
of plain
creased
∂w ∂hwith age and was ∂w 20~36×10∂w
-12 2
m /s at 7 days maximumAlso,
concrete. amountthe of water percoefficient
migration unit volume that can
of concrete
− e16~27×10 e α&days.
+ e Itα&decreased
+ w& fill all pores (both capillary pores and gel pores), one
-12 2 -12 2
and + ∇ • (D m ∇h/s
) = at 28 (3)
to with MK was 4.6~17.9×10 m /s at 28 days, which
∂h ∂t -12m2/sh at 91days ∂α (data
c
∂αnot sshown n
can calculate K as one obtains
10~21×10 c s in the is roughly 1.4~3.3
1 times better than that of plain
figure). concrete in terms of resistance against chloride inva-
where
The ∂w e/∂h is the slope of the sorption/desorption
migration coefficient of concrete replaced sion. ⎡
⎢ ⎜ g α − α ⎟h ⎥
⎛ ∞
c c
⎞ ⎤
10

isotherm (also called

with FA increased as the moisture
replacement capacity).
ratio was The in- w −
The chloride c s + α s s −coefficient
αmigration
0.188 G ⎢ −e ⎝
0.22 of concrete
⎠
⎥ 1
1

governing equation (Equation 3) must be completed ⎦ (6)

0 1
creased at 7 days; these results deviated from those generally decreased with the elapse ⎢ of time. Thus,
⎥
K (α α ) =
⎣
by total
of appropriate
charge boundary
passed. Itand initialthat
is clear conditions.
the migration the migration
c s
1
,
or diffusion ⎛ coefficient
⎜ g α − α ⎟h
∞ ⎞ is a function of
The relation between the amount of penetration
coefficient derived from the chloride evaporable e ⎝ c property
time and this time dependent c ⎠ − can be deline-
10
1
1

water differs
depth and relativefrom humidity
the total ischarge called passed
‘‘adsorption
even ated by Equation 2.
isotherm” if measured with increasing relativity
though both results were obtained via the same test The material parameters kcvg and ksvg and g1 can
humidityHowever,
method. and ‘‘desorption isotherm”
the migration in the opposite
coefficient of FA be calibrated
⎛ t0 ⎞ by fitting experimental data relevant to
a

case. Neglecting their difference (Xi et al. 1994),

concrete was similar to that of plain concrete in
at 28 D = D 0⎜ ⎟
free (evaporable) water content in concrete (2)at
the following, ‘‘sorption isotherm” will be used with ⎝t ⎠
days, and decreased with an increase of the replace- various ages
where, D (Di : Luzio & Cusatis
Diffusion 2009b).
coefficient at t (m2/sec)
reference
ment ratioto both28sorption
after days. and desorption conditions.
ByThe themigration
way, if coefficient
the hysteresis of GGBSof the moisture
concrete de- D0 : Diffusion coefficient at t0 (m2/sec)
isotherm would be taken into account, two different
creased when the replacement ratio was increased to 2.2 Temperature
t evolution
: Age(day)
relation, evaporable water vs relative humidity, must
50% at 7 days while that of concrete with 70% Note that,t0at early age, since
: Datum the chemical reactions
age(day)
be usedincreased
GGBS accordingslightly
to the sign of the variation
in comparison. of the
However, associateda with: cement hydration and SF reaction
Age factor
relativity
there was no humidity.
difference Thein theshape of thecoefficient
migration sorption are exothermic, the temperature field is not uniform
isotherm
of concreteforaccording
HPC is influenced by many parameters,
to the replacement ratio at 28 forFigure
non-adiabatic
5 showssystems even ifofthe
the age factor environmental
concrete accord-
especially those that influence extent and rate ofcon-
days and the migration coefficient of GGBS the temperature is constant. Heat conduction
ing to the type of mineral admixture and the can re-
be
chemical
crete reactionsofand,
was 40~56% that of in plain
turn, concrete.
determineThepore to- described in concrete,
placement ratios. at least for temperature not
structure
tal chargeand pore through
passed size distribution
the concrete(water-to-cement
with GGBS exceeding 100°Cof (Bažant
The age factor & Kaplan
plain concrete 1996), and
was 0.18~0.23 by
ratio, cement chemical composition, SFincreased
decreased as the replacement ratio was content, Fourier’s
that law, which
of concrete reads
replaced with a mineral admixture in-
curingthe
while timemigration
and method, temperature,
coefficient showedmix no additives,
variation creased as the replacement ratio was increased.
etc.).the
with In replacement
the literatureratio various
after 28formulations
days. can be q =The
− λ∇age
T factor of FA concrete increased as (7) the
found
The to describecoefficient
migration the sorption of isotherm of normal
concrete specimens replacement of FA was increased and that of FA30
concrete
with SF or(XiMK et al.replacement
1994). However, decreasedin the present
as the re-
where
was q is the
0.39~0.49, heatis twofold
which flux, Thigher
is thethanabsolute
that of
paper the semi-empirical expression proposed
placement ratio was increased. The migration
-12 2
by
coeffi-
temperature,
plain concrete.and
The is the heat
λ GGBS conductivity;
concrete showed an in age
this
Norling
cient Mjornellwith
of concrete (1997)
SF was is 2.4~12.5×10
adopted becausem /sit factor in a range of 0.28~0.52. The age factor devi-

Proceedings of FraMCoS-7, May 23-28, 2010

 1.2 J = − D ( h , T ) ∇h
W/B0.40
1.0 W/B0.45
W/B0.50 The proportionality coefficient D(h,T)
0.8
W/B0.55 moisture permeability and it is a nonlinea
Age factor a

0.6 of the relative humidity h and temperature

& Najjar 1972). The moisture mass balanc
0.4 that the variation in time of the water mas
0.2
volume of concrete (water content w) be eq
divergence of the moisture flux J
0.0
0 30 50 70 0 5 10 15∂w 20 0 5 10 15 20
0 10 20 30 − = ∇•J
Replacement ratio(%) Replacement ratio(%) ∂t
Replacement ratio(%) Replacement ratio(%)
(a) FA (b) GGBS (c) SF (d) MK
The water content w can be expressed a
of the evaporable water we (capillary wa
Figure 5. Age factor of concrete according to the types of mineral admixture and replacement ratio.

0.6 0.7 vapor, and adsorbed water) and the non-e

(chemically bound) water wn (Mil
Chloride cont.(%, wt of conc.)

0.5 0.6 Pantazopoulo & Mills 1995). It is reas

Fixed/total chloride ratio

0.4 0.5 assume that the evaporable water is a fu
0.4
relative humidity, h, degree of hydration
0.3
Free chloride
degree of silica fume reaction, αs, i.e. we=w
0.3 = age-dependent sorption/desorption
Average
0.2 Fixed chloride
0.2 (Norling Mjonell 1997). MKUnder this assum
0.1 by substituting Equation 1 into Equati
0.1
obtains
0.0 0.0
0-5 ∂w 15-20
5-10 10-15 15-20 20-25 25-30 30-35 35-40 ∂w 35-40 ∂w
5-10 10-15
Distance from concrete surface(mm)
− from
Distance e ∂hconcrete
20-25 25-30 30-35
+ ∇ • ( Dsurface(mm)
∇h ) = eα e 0-5
&+ α&s + w
∂h ∂t h ∂α c
∂α
(a) Chloride content c
(b) Ratio of fixed to total chloride content s
Figure 6. Average chloride content (a) and fixed / total chloride content ratio (b) versus distance from the concrete surface.
where ∂we/∂h is the slope of the sorption/
ated from the W/B ratios while it was not signifi- isotherm
ride was considered as free (also
chloride called
and the moisture
residue capac
cantly affected by the replacement ratio of GGBS. of water soluble governing equation (Equation 3)acid
chloride subtracted from the must be
The concrete with SF showed a noticeable in- by appropriate
soluble chloride was considered boundary and initial conditi
as fixed chloride.
crease in age factor with an increase of the replace- Thechloride
The free and fixed relationcontent
between the amount
decreased as of e
ment ratio. The age factor of the concrete with 5% the distance fromwater and relative humidity is called ‘‘
the surface increased. In a range of
SF was similar to that of the concrete with 30% FA, isotherm”
0~5mm, in the vicinity of theif surface,
measured with chlo-
the total increasing
and the concrete with W/B 0.40 and 15% SF showed humidity
ride content was 0.48% andconcrete
of the ‘‘desorption
weight isotherm”
and it in th
the highest age factor, at 1.1. consisted of 0.33%case.ofNeglecting
free chloride theiranddifference
0.15% of (Xi et al.
The concrete replaced with MK also showed an the following, ‘‘sorption isotherm”
fixed chloride. The free chloride holds up to 68% ofwill be
increasing slope in its age factor curve as the re- the total chloridereference
and is 2.1 to both
timessorption
higher and thandesorption
the c
placement ratio was increased. By theHowever,
fixed chloride content. way, ifthetheratiohysteresis
of free toof the
The age factor indicates the improvement capac- isotherm
total chloride content wouldasbethetaken
decreased into account,
distance from two
ity of concrete in terms of resistance against chloride relation, evaporable water vs relative humi
the surface increased and approached the fixed chlo-
invasion with the elapse of time. More specifically, ride content. The be used
high according toofthe
concentration freesign of the varia
chloride
the higher the age factor is, the greater the increase relativity
in the surface layer humidity.
of the concrete The toshape
appears be at- of the
in resistance with time. From an analysis of the age isothermadsorption
tributable to the physical for HPC isofinfluenced
concrete by hy-many p
factor, it is clear that the resistance of concrete could drates (Sakurata &especially those that influence extent and
Maruya 1999).
be improved through replacement with a mineral chemical
Figure 6(b) shows the ratioreactions
of fixedand, in chlo-
to total turn, determ
admixture but the effect varies according to the type structure
ride content (binding ratio)and forpore size distribution
the average of all of (water-
of mineral admixture. ratio, and
the concrete specimens cementfor thechemical
concretecomposition,
replaced SF
curing time and method, temperature,
with MK. The average binding ratio was 31% be- mix
3.3 Chloride profile
tween 0~5mm, and etc.). In the with
increased literature variousfrom
the distance formulatio
found to
the surface, converging to 50%
describe
beyondthe20mm.
sorption isotherm
Figure 6(a) shows the average chloride content ac- concrete
The binding ratio of the (Xiconcrete
et al. 1994).
replaced However,
with in th
cording to the distance from the concrete surface ex- MK was slightly paper
higher the
thansemi-empirical
the average value. expression
The pro
posed to chloride solution. The water soluble chlo- Norling
binding ratio between 0~5mm Mjornellwas 35%(1997) (4% ishigher
adopted b

Proceedings of FraMCoS-7, May 23-28, 2010

J = − D ( h , T ) ∇h (1) explicitly accounts for the evolution of hydration
Diffusivity(×10-12m2/s) 20 reaction and SF content. This sorption isotherm
The
91days
proportionality coefficient D(h,T) is called
91days reads91days 91days
15
moisture permeability and it is a nonlinear function
of the10relative humidity h and temperature T (Bažant ⎡ ⎤
& Najjar 1972). The moisture mass balance requires we (h α c α s ) = G1 (α c , α s )⎢⎢1 − 1 ⎥
+
that the variation in time of the water mass per unit ∞
, ,

(g α
c 10 − α c )h ⎥⎥
volume 5
of concrete (water content w) be equal to the e ⎢
⎣ 1
⎦ (4)
divergence of the moisture flux J ⎡ (g α
∞ − α )h ⎤
0
K (α α s ) e 0 5 10c −15⎥
c 10
⎢ 1
10 15c 20
, 1
0 5 20
− ∂ 20
0 10 20 30 0 30 50 70 ⎢ ⎥
(2)
1
= ∇•J w
⎣ ⎦
∂ W/B 0.40
Diffusivity(×10-12m2/s)

t
330days 330days 330days 330days
15
The
W/B 0.45
water content w W/B
can be
0.50expressed as the sum
where the first term (gel isotherm) represents the
of the evaporable waterW/B (capillary water, water
we 0.55 physically bound (adsorbed) water and the second
vapor, and adsorbed water) and the non-evaporable
10 term (capillary isotherm) represents the capillary
(chemically bound) water wn (Mills 1966, water. This expression is valid only for low content
Pantazopoulo
5 & Mills 1995). It is reasonable to of SF. The coefficient G1 represents the amount of
assume that the evaporable water is a function of water per unit volume held in the gel pores at 100%
0 humidity, h, degree of hydration, αc, and
relative relative humidity, and it can be expressed (Norling
degree of silica
0 fume
10 reaction,
20 30αs, i.e.0 we=w30e(h,α50
c,αs) 70
Mjornell
0 5
1997) as
10 15 20 0 5 10 15 20
= age-dependent Replacementsorption/desorption
ratio(%) isotherm
Replacement ratio(%) Replacement ratio(%) Replacement ratio(%)
(Norling Mjonell(a)1997).
FA Under this assumption (b) GGBS and G (α c α s )(c) c α c+ ks α s
= kSF (d) MK (5)
byFigure
substituting Equation 1 into Equation 2 one 1 vg c,
vg s
obtains7. Chloride diffusion coefficient of concrete according to the types of cmineral admixture and replacement ratio.
than average) and that for 25~30mm was 58% (8% cient k vg and
wherecould vg are material
not kcalculated
s
parameters.
due to the lack ofFrom the
a suffi-
∂w ∂than
higher ∂w binding
average). The ∂w capacity of the maximum
cient amount
chloride of water
profile. per unitatvolume
However, 330 days,that can
the
− e
concrete
h with Dh ∇happears
+ ∇ • (MK ) =
e αto
& +
be e α& + w& to (3)
attributable the fill all
diffusionpores (both
coefficient capillary
decreasedpores
as and
the gel
W/B pores),
ratio one
de-
∂h ∂t component, Al ∂αO which
c
∂α holdss n
can calculate K1 as one obtainsratio of SF increased,
chemical 2 c3 s up to 40% creased and the replacement
of MK. falling in a range of 0.9~4.4×10-12m2/s.
where ∂we/∂h is the slope of the sorption/desorption The diffusion coefficient of the ⎡
⎢
concrete
⎜ g α − αwith
⎛ ∞
c c
⎞ ⎤ MK
⎟h ⎥
10

isotherm
3.4 Chloride (also calledcoefficient
diffusion moisture capacity). The decreased was −the replacement
α s + α s − Gratio
c
0.188
s 0.22 ⎢ − e was increased
⎝ 1
⎠
⎥ to
1

governing equation (Equation 3) must be completed (6)

0 1
⎢ ⎥
15% while that of the concrete⎣ with 20% MK⎦ was
K (α c α sto) =that of the concrete with 15% MK. Based on
by appropriate
Figure 7 showsboundary
the chlorideand initial conditions.
diffusion coefficient at similar
1
,

⎜ g α − α ⎟h
⎛ ∞ ⎞

The relation between the amount of evaporable

91 and 330 days for the immersion test. the results for the diffusion c c ⎠ − it appears that
10

e ⎝ coefficient,
1
1

water
The and relative
diffusion humidity ofis plain
coefficients calledconcrete
‘‘adsorption
with the effective replacement ratio of MK was 15% in
isotherm”
W/B ratios if measured
ranging with toincreasing
from 0.40 relativity
0.55 was 8.7~14.1 Theofmaterial
terms parameters
increasing kcvg and kagainst
the resistance s
vg and chloride
g1 can
humidity
×10 m /sand
-12 2
at 91‘‘desorption isotherm” in-12the
days and 4.7~12.0×10 m2/sopposite
at 330 be calibrated by fitting experimental data relevant to
penetration.
case. Neglecting
days. The diffusiontheir difference (Xi et al. 1994),was
coefficient of FA concrete in freeTable
(evaporable)
6 shows thewater content
regression in between
results concrete theat
the following, ‘‘sorption
6.8~15.2×10 -12 2
m /s, similarisotherm” will
to that of plainbe used with
concrete, various ages (Di Luzio & Cusatis 2009b). derived
total charge passed, the migration coefficient
reference
at 91 daystowhilebothitsorption
decreased andtodesorption
2.7~6.7×10 conditions.
-12 2
m /s at from the electrically accelerated migration test, and the
By the
330 days.way, if the the
Moreover, hysteresis
diffusionofcoefficient
the moisture de- diffusion coefficient derived from the immersion test.
2.2 relationship
Temperaturebetween evolution
isotherm would be taken into account, two different
creased as the replacement ratio was increased, and The the results of the migration
relation,
that evaporable
of concrete withwater
30% FA vs relative
was lesshumidity,
than halfmustthat Note that, at early age, since the
test and the diffusion coefficient chemical
showed reactions
a good corre-
be plain
of used according to the2.73~4.87×10
concrete, i.e., sign of the variation
m /s, atof330
-12 2
the associated
lation with cementmeasuring
at corresponding hydration stages.
and SFIt reaction
appears
relativity
days. The humidity.
concrete The shapewith
replaced of 30% the sorption
GGBS are exothermic,
that both the hydration the temperature
of cementfield and theis not uniform
pozzolanic
isothermdifferent
showed for HPCdiffusion
is influenced by manyaccording
coefficients parameters, to for non-adiabatic
reaction systems
of the mineral even ifcomplexly
admixture the environmental
influence
especially those that influence extent and rate of the
the W/B ratios while the 50% and 70% GGBS spe- temperature
the concrete’sisresistance
constant.against
Heat chloride
conduction can be
penetration
chemicalversus
cimens reactions and, in
W/B ratio. Theturn, determine
diffusion pore
coefficient described tointheconcrete,
according at least for temperature not
elapsed time.
structure
of and pore
the concrete size30%
with distribution
GGBS was (water-to-cement
40~50% of exceeding
Based on100°C (Bažant
the results of this&study, Kaplan 1996),
the total by
charge
ratio,ofcement
that comparable chemical
plain composition,
concrete at 91 SFdayscontent,
and Fourier’s law, which reads
passed and migration coefficient are useful variables to
curing time
25~35% and days.
at 330 method, temperature, mix additives, investigate the resistance of concrete against chloride
etc.).
TheIndiffusion
the literature various
coefficient formulations
of the concrete with can SFbe q = − λ ∇T within a short time. However, it is neces-
penetration (7)
foundlower
was to describe
than thatthe sorption
of plain isotherm
concrete and of normal
decreased sary to consider the effects of the mineral admixture
concrete
as the W/B (Xiratio
et al. 1994). However,
decreased. At 91 days, in some
the present
of the on the resistance of concrete according to the time
where q is the heat flux, T is the absolute
paper
SF the semi-empirical
concrete showed chloride expression
penetration proposed
depth be- by elapsed when selecting the test time for concrete
temperature, and λ is the heat conductivity; in this
Norling
low 5mmMjornell (1997) is the
and, consequently, adopted
diffusionbecause
coeffi-it containing mineral admixtures.

Proceedings of FraMCoS-7, May 23-28, 2010

 J = − D ( h , T ) ∇h
Table 6. Regression results between the total charge passed, migration coefficient and apparent diffusion coefficient.
Apparent diffusion coefficient (Y*))
RCPT Measured age
91days**) 182days The proportionality
330days coefficient D(h,T)
2 2 moisture permeability 2and it is a nonlinea
28days Y=0.006C, R =0.8273 Y=0.0035C, R =0.8198 Y=0.0033C, R =0.7472
2 2 of the relative humidity h and temperature
Total charge passed(C) 56days Y=0.0077C, R =0.5591 Y=0.0047C, R =0.7867 Y=0.0045C, R2=0.8246
91days Y=0.0095C, R2=0.3623
& Najjar 1972).
Y=0.0058C, R2=0.6815
The moisture mass balanc
Y=0.0056C, R2=0.8040
2 2 that the variation in time of the water mas
28days Y=0.4905X, R =0.7937 Y=0.2825X, R =0.6905 Y=0.2632X, R2=0.5663
Migration
56days Y=0.6395X, R2=0.7146
volume
Y=0.3766X, R2=0.7709
of concrete (water content w) be eq
Y=0.3556X, R2=0.7215
coefficient(X*))
91days Y=0.7308X, R2=0.6950
divergence
Y=0.4310X, R2=0.7514
of the moisture flux J
Y=0.4087X, R2=0.7306
*) Unit: ×10-12m2/s
**) Immersed period on chloride solution − ∂ = ∇•J
w

∂ t

4 CONCLUSION REFERENCES
The water content w can be expressed a
The main conclusions that can be drawn from this AASHTO T 277 2005. of the evaporable
Standard method ofwater (capillary wa
weelectrical
test for
study are as follows: vapor, and adsorbed water) andpene-
indication of concrete’s ability to resist chloride
tration, American association of state highway and trans-
ion the non-e
(1) The total charge passed has a linear relation- (chemically
portation officials.
bound) water w n (Mil
ship with the initial current and the total charge Pantazopoulo & Mills
ACI Building Code 201 2000. ACI manual of concrete practice 1995). It is reas
passed could be determined with high accuracy by assume that the evaporable
2000, Materials and general properties of concrete, Chapter water is a fu
measuring the initial current before finishing the ac- relative humidity, h, degree of hydration
2-Aggressive chemical exposure, American concrete insti-
celerated chloride migration test. tute: 8-12. degree of silica fume reaction, αs, i.e. we=w
(2) The age factor of plain concrete was ACI Building Code = age-dependent
318 2008. sorption/desorption
Building code requirements for
structural concrete (ACI 318M-08) and commentary, Chapter
0.18~0.23 and that of concrete replaced with a min- (Norling
4-Durability requirement, Mjonell
American 1997).
concrete Under55-61.
institute: this assum
eral admixture increased as the replacement ratio ASTM C 1202 2005.byStandard substituting
test method Equation 1 into
for electrical indi- Equati
was increased. It is clear that the resistance of con- obtains
cation of concrete’s ability to resist chloride ion penetra-
crete with the elapse of time could be improved tion, ASTM international.
Boulfiza, M., Sakai, K., Banthia, N. & Yoshida, H. 2003. Predic-
through replacement with a mineral admixture. ∂w ∂h ∂w ∂w
− e e α con-
&+ eα w
&+
tion of chloride ions ingress in uncracked and cracked
(3) In a range of 0~5mm, in the vicinity of the + ∇ • ( D ∇h ) =
∂h ∂t January-February
crete, ACI materials journal h α 38-48.∂α s
∂2003: c
surface, the total chloride content was 0.48% of the BS 8500 2006. Concrete complementary British Standard c tos
concrete weight, and it consisted of 0.33% of free BS EN 206-1, Part 1: Method of specifying and guidance
chloride and 0.15% of fixed chloride. However, the wherestandard:
for the specifier, British ∂we/∂h 18-31.is the slope of the sorption/
ratio of free to total chloride content decreased as Dhir, R. K., Jones, M. isotherm
R., Ahmed, (also
H. E. H. called moistureA. capac
& Seneviratne,
M. G. 1990. Rapid estimation of chloride diffusion coeffi-
cient in concrete,governing
Magazine ofequation (Equation 3) must be
the distance from the surface increased and ap- concrete research, 42(152):
proached the value of the fixed chloride content. 177-185. by appropriate boundary and initial conditi
(4) The binding ratio of the concrete replaced Feldman, R. F., Chan, G. The W.,relation
Brousseau, between the amount of e
R. J. & Tumidajski,
with MK was slightly higher than that of average of water ofand
P. J. 1994, Investigation the relative
Rapid Chloridehumidity is called ‘‘
Permeability
all of the concrete specimens. The high binding ca- Test, ACI materialsisotherm”
journal May-Juneif measured with increasing
1994: 246-255.
pacity of the concrete with MK appears to be attrib- humidity and ‘‘desorption isotherm” in th
Jiang, L., Zhang, M. H., & Malhotra, V. M. 2004. Evaluation
of test method for determining the resistance of concrete to
utable to the chemical component Al2O3, which case. Neglecting
chloride-ion penetration, American concrete their difference
institute spe- (Xi et al.
holds up to 40% of MK. the following,
cial publication, SP221-1: 1-28. ‘‘sorption isotherm” will be
(5) The results of the migration test and diffusion KCSS 2003. Concrete reference
standard to specification,
both sorptionChapter and desorption
15- c
coefficient showed a good correlation at correspond- Coastal concrete,By the ofway,
Ministry if the and
construction hysteresis
transporta-of the
ing measuring stages. However, it is necessary to tion: 396-411 isotherm would be taken into account, two
Page, C. L., Short, N. R. & El Tarras, A. 1981. Diffusion of
consider the effect of the mineral admixture on the relation,cement
chloride ions in hardened evaporable
pastes, water
Cementvsand relative
con- humi
resistance of concrete with the elapse of time when be used according
crete research, Vol.11(3): 395-406. to the sign of the varia
selecting the test time for a concrete specimen con- Sakurata, R. & Maruya, relativity
K. 1999.humidity.
Relationship The shapeand of the
of porosity
taining mineral admixtures. isotherm
diffusion of chloride ion in for HPCmatrix,
concrete is influenced
Proceeding byofmany p
especially those that influence extent and
Japan concrete institute, Vol.21(2): 829-834.
chemical reactions and, in turn, determ
Shi, C., Stegemann, J. A., & Caldwell, R. J. 1998. Effect of
supplementary cementing materials on the specific conduc-
ACKNOWLEDGEMENTS structure
tivity of pore solution and and pore size distribution
its implications on the rapid (water-
ratio,testcement
chloride permeability (AASHTO chemical
T 277 andcomposition,
ASTM C SF
The work presented in this paper was funded by 1202) results, ACI curing time and
materials journalmethod, temperature,
July-August 1998: mix
Center for Concrete Corea (05-CCT-D11), sup- 389-394 etc.). In the literature various formulatio
ported by Korea Institute of Construction and Tang, L. & Nilsson, L. found
O. 1992.to Rapid
describe the sorption
determination
ride diffusivity in concrete by applying an electrical field,
isotherm
of the chlo-
Transportation Technology Evaluation and Planning concrete
ACI materials journal (Xi et al.1992:
January-February 1994).
49-53.However, in th
(KICTEP) under Ministry of Land, Transport and paper the semi-empirical expression pro
Maritime Affairs (MLTM). Norling Mjornell (1997) is adopted b

Proceedings of FraMCoS-7, May 23-28, 2010

 D (h1996.
J = − L.
Tang, , T ) ∇h
Electrically accelerated methods for determin- (1) explicitly
Zhao, accounts
T. J., Zhou, Z. H.,for
Zhu,the evolution
J. Q., & Feng, N.ofQ.hydration
1998. An
ing chloride diffusivity in concrete – current development, reaction and SF content. This sorption isotherm
alternating test method for concrete permeability, Cement
Magazine of concrete research, 48, No.176: 173-179.
The
Torii, K., proportionality coefficient
Sasatani, T. & Kawamura, D(h,T)
M. 1997. is called
Application of
reads
and concrete research, Vol.28(1): 7-12.

moisture permeability
rapid chloride andtestit tois evaluate
permeability a nonlinear function
the chloride-ion
of penetration
the relativeintohumidity and temperature
concrete,hAmerican T (Bažant
concrete institute spe- ⎡ ⎤
& Najjar 1972). The moisture mass balance requires
cial publication, SP170-21: 421-435.
we (h α c α s ) = G1 (α c , α s )⎢⎢1 − 1 ⎥
+
Whiting, D. 1981. Rapid determination of the chloride perme-
thatability
the variation in time of the water mass per unit ∞
, ,

of concrete, Report No. FHWA/RD-81/119, FHWA, (g α

c 10 − α c )h ⎥⎥
volume
Washington, D.C. (water content w) be equal to the
of concrete e
⎢
⎣ 1
⎦ (4)
divergence of the moisture flux J ⎡ (g α
∞ − α )h ⎤
K (α c α s ) e
,⎢ c c −
10
1
1⎥

− ∂ = ∇•J (2) ⎢ ⎥
1
w
⎣ ⎦
∂ t

The water content w can be expressed as the sum where the first term (gel isotherm) represents the
of the evaporable water we (capillary water, water physically bound (adsorbed) water and the second
vapor, and adsorbed water) and the non-evaporable term (capillary isotherm) represents the capillary
(chemically bound) water wn (Mills 1966, water. This expression is valid only for low content
Pantazopoulo & Mills 1995). It is reasonable to of SF. The coefficient G1 represents the amount of
assume that the evaporable water is a function of water per unit volume held in the gel pores at 100%
relative humidity, h, degree of hydration, αc, and relative humidity, and it can be expressed (Norling
degree of silica fume reaction, αs, i.e. we=we(h,αc,αs) Mjornell 1997) as
= age-dependent sorption/desorption isotherm
(Norling Mjonell 1997). Under this assumption and c α c+ ks α s
G (α c α s ) = k vg (5)
by substituting Equation 1 into Equation 2 one 1
,
c vg s
obtains
where kcvg and ksvg are material parameters. From the
∂w ∂h maximum amount of water per unit volume that can
− e + ∇ • ( D ∇h) = ∂we ∂w
α&c + e α&s + w&n (3) fill all pores (both capillary pores and gel pores), one
∂h ∂t h ∂α ∂α can calculate K1 as one obtains
c s
where ∂we/∂h is the slope of the sorption/desorption ⎡
⎢ 10⎜
⎛
g α c∞ − α c ⎞⎟h ⎤⎥
isotherm (also called moisture capacity). The w − 0.188 α s + 0.22α s G
c s − ⎢1 − e ⎝ 1 ⎠
⎥

governing equation (Equation 3) must be completed (6)

0 1
⎢ ⎥
K (α c α s ) = ⎣ ⎦
by appropriate boundary and initial conditions. 1
,
⎛
g αc − αc h
∞ ⎞

The relation between the amount of evaporable

10⎜ ⎟
e ⎝ 1
− ⎠ 1

water and relative humidity is called ‘‘adsorption

isotherm” if measured with increasing relativity
humidity and ‘‘desorption isotherm” in the opposite
case. Neglecting their difference (Xi et al. 1994), in
the following, ‘‘sorption isotherm” will be used with
reference to both sorption and desorption conditions.
By the way, if the hysteresis of the moisture
isotherm would be taken into account, two different
relation, evaporable water vs relative humidity, must
be used according to the sign of the variation of the
relativity humidity. The shape of the sorption
isotherm for HPC is influenced by many parameters,
especially those that influence extent and rate of the
chemical reactions and, in turn, determine pore
structure and pore size distribution (water-to-cement
ratio, cement chemical composition, SF content,
curing time and method, temperature, mix additives,
etc.). In the literature various formulations can be
found to describe the sorption isotherm of normal
concrete (Xi et al. 1994). However, in the present
paper the semi-empirical expression proposed by
Norling Mjornell (1997) is adopted because it
The material parameters kcvg and ksvg and g1 can
be calibrated by fitting experimental data relevant to
free (evaporable) water content in concrete at
various ages (Di Luzio & Cusatis 2009b).
2.2 Temperature evolution
Note that, at early age, since the chemical reactions
associated with cement hydration and SF reaction
are exothermic, the temperature field is not uniform
for non-adiabatic systems even if the environmental
temperature is constant. Heat conduction can be
described in concrete, at least for temperature not
exceeding 100°C (Bažant & Kaplan 1996), by
Fourier’s law, which reads
q = − λ ∇T (7)
where q is the heat flux, T is the absolute
temperature, and λ is the heat conductivity; in this

Proceedings of FraMCoS-7, May 23-28, 2010