You are on page 1of 8

ACI MATERIALS JOURNAL TECHNICAL PAPER

Title no. 107-M27

Shrinkage of Precast, Prestressed


Self-Consolidating Concrete
by Kamal H. Khayat and Wu Jian Long

The proper evaluation of shrinkage is critical for the design of fillers employed in SCC. Therefore, both drying and
prestressed structural members. An experimental program was autogenous shrinkage deformations have to be accounted for
undertaken to evaluate autogenous and drying shrinkage of precast, in the structural detailing of reinforced concrete and
prestressed self-consolidating concrete (SCC). Sixteen SCCs with a prestressed concrete members.
slump flow of 680 ± 20 mm (26.7 ± 0.8 in.) were evaluated. These Some discrepancy exists regarding the shrinkage deformations
mixtures were made with 440 to 500 kg/m3 (742 to 843 lb/yd3) of
binder, Type MS cement or HE cement; and 20% Class F fly ash;
of SCC compared to HPC. For example, Persson,6,7 Vieira
0.34 to 0.40 water-cementitious material ratio (w/cm); viscosity- and Bettencourt,8 Pons et al.,9 Attiogbe et al.,10 and
modifying admixture content of 0 to 100 mL/100 kg (0 to 1.53 fl oz/cwt) Bouzoubaâ and Lachemi11 reported that the shrinkage of
of binder, and 0.46 to 0.54 sand-to-total aggregate volume ratio. SCC can be similar to that of normal concrete of similar
Two high-performance concretes (HPCs) with 0.34 and 0.38 w/cm compressive strength. On the other hand, according to
and slump of 150 mm (5.9 in.) were also investigated. The SCC Heirman and Vanderwalle,12 Song et al.,13 Hu and
developed 5 to 30% higher drying shrinkage at 300 days than the Barcelo,14 Hammer,15 Turcry and Loukili,16 Klug and
HPC but with similar autogenous shrinkage. The shrinkage was Holschemacher,17 and Turcry et al.,18 the shrinkage of SCC
compared to prediction models proposed by AASHTO 2004 and was found to be higher than that of normal concrete. Holt and
2007, CEB-FIP 90, GL 2000, and ACI 209R. The CEB-FIP 90 and a Schodet19 found early-age shrinkage of SCC to be lower
modified AASHTO 2004 model was found to provide an adequate
prediction of shrinkage for precast, prestressed SCC.
than that of normal concrete. Several factors could explain
this discrepancy in the shrinkage deformation of SCC
compared to concrete of a normal consistency, in particular,
Keywords: autogenous shrinkage; drying shrinkage; prestressed concrete;
self-consolidating concrete; shrinkage models. variations in mixture design including w/cm, binder content,
coarse aggregate volume, and type of chemical admixtures
in use. The discrepancy also lies in the type of comparison.
INTRODUCTION
Some studies evaluate the visco-elastic properties of SCC
Self-consolidating concrete (SCC) is highly flowable, mixtures proportioned differently from normal vibrated
nonsegregating concrete that can spread into place, fill the concrete, whereas other studies compare the performance
formwork, and encapsulate the reinforcement without any of SCC and normal concrete having the same mixture
mechanical consolidation.1-3 The workability of SCC is designs and compressive strength levels. Moreover, the
higher than the highest class of workability associated with shrinkage results are, in some cases, taking into consideration
high-performance concrete (HPC) of normal consistency only drying shrinkage determined after a certain time of curing
that is typically used in precast, prestressed concrete (or immediately after demolding), whereas others consider
applications. Changes in mixture design and the fluidity of shrinkage that is determined starting from the setting time
SCC can lead to changes in mechanical and visco-elastic where autogenous shrinkage is added to drying shrinkage.
properties of the SCC compared to what is commonly expected
Compared to existing prediction models, studies have
from HPC. Typically, SCC has a higher binder content and a
shown that the scatter between measured and predicted
lower coarse aggregate volume than HPC. These factors could
shrinkage values is greater for SCC than for normal
increase the risk of thermal, autogenous, and drying shrinkage,
concrete.20 Shrinkage values for SCC were found to be
leading to cracking and reduction in serviceability.
larger than those estimated by various prediction drying
The overall shrinkage of concrete corresponds to a combi- shrinkage models. Work is therefore required to compare the
nation of several shrinkages, that is, plastic shrinkage, shrinkage data of SCC mixtures made with representative
autogenous shrinkage, drying shrinkage, thermal shrinkage, mixture designs and material constituents to determine the
and carbonation shrinkage.4 Unless specifically designed, applicability of various shrinkage prediction models for SCC
for conventional concrete, shrinkage is taken as drying used in precast applications, such as those employed for
shrinkage, which is the strain associated with the loss of prestressed girders.
moisture from the concrete under drying conditions.
Conventional concrete with a relatively high water- RESEARCH SIGNIFICANCE
cementitious material ratio (w/cm) (higher than 0.40) Autogenous shrinkage and drying shrinkage significantly
exhibits a relatively low autogenous shrinkage, with values influence the design and performance of precast, prestressed
less than 100 microstrain.5 In contrast, the SCC used in
precast, prestressed applications is typically proportioned
with a low w/cm of 0.32 to 0.40. Relatively low w/cm values, ACI Materials Journal, V. 107, No. 3, May-June 2010.
MS No. M-2009-065 received February 27, 2009, and reviewed under Institute publication
coupled with a high content of binder, leads to greater autog- policies. Copyright © 2010, American Concrete Institute. All rights reserved, including the
enous shrinkage. Such shrinkage increases with the use of making of copies unless permission is obtained from the copyright proprietors. Pertinent
discussion including authors’ closure, if any, will be published in the March-April 2011
finely ground supplementary cementitious materials and ACI Materials Journal if the discussion is received by December 1, 2010.

ACI Materials Journal/May-June 2010 231


Kamal H. Khayat, FACI, is a Professor of civil engineering at the Université de
Sherbrooke, Sherbrooke, QC, Canada. He chairs ACI Committee 237, Self-Consolidating
Concrete; and is a member of 234, Silica Fume in Concrete; 236, Material Science
of Concrete; 238, Workability of Fresh Concrete; and 552, Cementitious Grouting. His
research interests include self-consolidating concrete, high-performance concrete, rheology
of cement-based materials, and concrete repair.

Wu Jian Long is an Assistant Professor of civil engineering at the Shenzhen University,


Shenzhen Durability Center for Civil Engineering, Shenzhen, China. He received his
PhD in civil engineering from the Université de Sherbrooke. His research interests
include self-consolidating concrete, high-performance concrete, and durability of
concrete structures.

Table 1—Physical properties and chemical


compositions of cement and fly ash
Cement and supplementary Type MS Type HE Class F
cementitious materials cement cement fly ash
Physical properties
Specific gravity 3.14 3.15 2.53
Fig. 1—Grading of combined aggregate retained on various
Blaine specific surface area, m2/kg 390 530 410 sieve openings.
Passing No. 325 (45 μm), % 91 99 90
Chemical composition, %
SiO2 21.4 20.0 52.4 elements. Such early-age and long-term deformations for
Al2O3 4.6 5.4 27.2
SCC can differ from those of HPC with normal consistency,
given the differences in mixture design and the flow charac-
Fe2O3 2.9 2.3 8.3
teristics of these construction materials. The study presented
CaO 63.3 63.5 4.5 herein aims to determine the main factors affecting the
MgO 2.0 1.4 0.96 drying shrinkage of SCC, which is increasingly being used
SO3 3.4 4.4 0.05 in mainstream precast, prestressed concrete production. The
K2O 0.94 1.1 2.33 proper estimate of shrinkage of SCC designated for precast,
Na2O 0.07 0.15 0.20 prestressed applications, including prestressed bridge girders,
Na2O eq* 0.69 0.88 1.74
can provide engineers with the information necessary for
producing high-quality products manufactured with SCC.
LOI 0.98 0.80 2.73
Bogue composition, %
C3S
EXPERIMENTAL PROGRAM
50.0 54.9 —
Materials
C2S 23.7 15.8 —
Two types of portland cement, Type MS and Type HE
C3A 7.4 10.5 — (similar to ASTM C150 Type I/II cement and Type III
C4AF 8.8 6.9 — cement, respectively) and a Class F fly ash were used in this
*Na2O equivalent = Na2O + 0.64K2O. investigation. The physical properties and chemical compositions
of the cement and fly ash are given in Table 1. The Blaine
fineness values of the Type MS cement, Type HE cement,
and fly ash are 390, 530, and 410 m2/kg (212, 288, and
222 yd2/lb), respectively. The SCC mixtures were evaluated
Table 2—Grain-size distribution and physical using both Type MS cement and Type HE cement and 20%
properties of aggregate Class F fly ash replacement, by mass. The binder content
Siliceous sand Crushed coarse aggregate varied from 440 to 500 kg/m3 (742 to 843 lb/yd3).
3/4 in. to 1/2 in. to 3/8 in. to All concrete mixtures were prepared with a crushed coarse
0 to No. 4 No. 4 (19 No. 4 (12.5 No. 8 (9.5
Sieve opening (0 to 4.75 mm) to 4.75 mm) to 4.75 mm) to 2.36 mm) aggregate with 12.5 mm (1/2 in.) maximum size aggregate.
1 in. (25 mm) 100 100 100 100
Natural siliceous sand with a specific gravity of 2.66
3/4 in. (19 mm) 100 99 100 100
conforming to AASHTO T 27 specifications was used. The
grain-size distribution and physical properties of the various
1/2 in. (12.5 mm) 100 68 95 100
aggregates are summarized in Table 2. The particle-size
3/8 in. (9.5 mm) 100 40 69 100
distributions of the sand and coarse aggregate are within the
No. 4 (4.75 mm) 98 7 18 13
AASHTO recommended limits. The combined sand and
No. 8 (2.36 mm) 85 1 4 2
coarse aggregate particle-size distribution is plotted in Fig. 1.
No. 16 (1.18 mm) 72 1 3 2
A polycarboxylate-based high-range water-reducing admixture
No. 30 (600 μm) 55 — — —
(HRWRA) complying with ASTM C494C/C494M (Type F)
No. 50 (300 μm) 32 — — —
and an organic, thickening-type viscosity-modifying admixture
No. 100 (150 μm) 9 — — — (VMA) that are widely used in the precast industry in North
Pan 2 0 0 0 America were employed. The specific gravities of these
Specific gravity 2.66 2.72 2.71 2.73 chemical admixtures were 1.047 and 1.0, respectively, and their
Absorption, % 1.12 0.31 0.44 0.38 solid contents were 20.3% and 6%, respectively.

232 ACI Materials Journal/May-June 2010


Table 3—Mixture compositions
3
Mixture Binder kg/m VMA, mL/100 kg Binder
3
no. (lb/yd ) w/cm CM* (fl oz/cwt) type S/A, %
1 440 (742) 0.34 0 MS 0.54
2 440 (742) 0.34 0 HE† 0.46
Non-AEA SCC (slump flow 680 ± 20 mm)

3 440 (742) 0.34 100 (1.53) MS 0.46


4 440 (742) 0.34 100 (1.53) HE 0.54
5 440 (742) 0.40 0 MS 0.46
6 440 (742) 0.40 0 HE 0.54
7 440 (742) 0.40 100 (1.53) MS 0.54
8 440 (742) 0.40 100 (1.53) HE 0.46
Fig. 2—Steam curing regime specified by AASHTO and CSA.
9 500 (843) 0.34 0 MS 0.46
10 500 (843) 0.34 0 HE 0.54 followed by the addition of the binder. Once the aggregate
11 500 (843) 0.34 100 (1.53) MS 0.54 particles were coated by a layer of cement paste, the
12 500 (843) 0.34 100 (1.53) HE 0.46 HRWRA and the VMA diluted with the remaining mixing
13 500 (843) 0.40 0 MS 0.54
water were introduced over 30 seconds, and the concrete was
mixed for 2.5 minutes. The concrete remained at rest in the
14 500 (843) 0.40 0 HE 0.46
mixer for 2 minutes for fluidity adjustment and to enable any
15 500 (843) 0.40 100 (1.53) MS 0.46
large air bubbles entrapped during mixing to rise to the
16 500 (843) 0.40 100 (1.53) HE 0.54 surface. The concrete was then remixed for 3 minutes. The
w/cm = 0.34, Type MS cement, 12.5 mm (1/2 in.) fresh properties of SCC were measured at 10 and 40 minutes
17 crushed aggregate.
Normal consistency mixtures with 150 mm (5.9 in.) slump. after cement and water contact. During that period, the
HPC

w/cm = 0.38, Type HE + 20% Class F fly ash, 12.5 mm concrete was agitated at 6 rpm; the drum mixer was covered
18 (1/2 in.) crushed aggregate. during that time to prevent any water evaporation.
Normal consistency mixtures with 150 mm (5.9 in.) slump.
*

CM = cementitious materials. Curing
Type HE cement + 20% Class F fly ash. Concrete cylinders measuring 100 x 200 mm (3.9 x 7.9 in.)
Note: 1 mm = 0.0394 in.
and 150 x 300 mm (5.9 x 11.8 in.) were sampled 10 minutes
after the end of mixing to evaluate mechanical properties and
Table 4—Concrete curing conditions drying shrinkage under three different curing conditions
indicated in Table 4. All sampled specimens were cast in one
Curing
method Stage Details lift without any mechanical consolidation. Some of the
I Ambient temperature for 2 hours after water-cement contact samples were covered and remained in the laboratory at 23 ±
II Temperature raised for 2 hours
2°C (73.4 ± 3.6°F) to air cure until the time of testing,
Steam- whereas others were steam-cured according to the regime
III Concrete temperature maintained for 10 hours
cured described in Fig. 2.
IV Temperature decreases over 2 hours to ambient temperature
Based on AASHTO, CSA, and PCI specifications, the
V Air-curing until age of testing at 18 hours maximum curing temperature in the concrete should not exceed
Moist- I 18 hours in molds with wet burlap at 23 ± 2°C (73.4 ± 3.6°F) 70 or 71°C (158 or 160°F) to prevent the risk of delayed
cured II Moist-cured at 23 ± 2°C (73.4 ± 3.6°F) until testing age ettringite formation. The standards also stipulated that
Air- I 18 hours in molds with wet burlap at 23 ± 2°C (73.4 ± 3.6°F) temperature rise during steam curing should not exceed 20
cured II Air-dried at 23 ± 2°C (73.4 ± 3.6°F) until testing age and 22°C/h (68 and 72°F/h) (depending on the standard), and
the rate of cooling should not exceed 15 or 22°C/h (59 or
72°F/h) for the CSA and AASHTO standards, respectively.
Mixture composition
In this investigation, the targeted release compressive
The 56-day design compressive strength of the concrete strength after 18 hours was 34.5 MPa (5000 psi). In some
targeted for precast, prestressed applications varied between cases, the curing regime was slightly modified by increasing
55 and 70 MPa (7980 to 10,150 psi). One raw material the maximum curing temperature to achieve the targeted
parameter and four mixture proportioning parameters were release compressive strength. In addition to compressive
considered in this investigation, including the binder type, strength, all of the samples used for shrinkage testing were
binder content, w/cm, sand-to-total aggregate volume ratio subjected to steam curing to simulate field conditions in
(S/A), and dosage of VMA. The mixture proportioning of prestressed applications.
the 16 SCCs and two HPC mixtures of normal consistency
used in this investigation are summarized in Table 3. The Test methods
initial slump flow of the SCC mixtures was set at 660 to Autogenous shrinkage was determined on prismatic
700 mm (26 to 27.6 in.). samples measuring 75 x 75 x 300 mm (3 x 3 x 11.8 in.) using
The SCC mixtures were prepared in 110 L (3.88 ft3) embedded vibrating wire strain gauges. Immediately after
batches using a drum mixer. The mixer was modified to demolding at 18 hours, the prisms were sealed using two
promote a greater shearing action of the concrete and was layers of self-adhesive aluminum sheet to prevent any mass
equipped with a speed gear to enable the simulation of loss due to drying. The samples were stored at 23 ± 2°C
concrete agitation at a low rotational speed after the end of (73.4 ± 3.6°F) until the end of testing. Autogenous shrinkage
the mixing cycle. The mixing sequence consisted of wetting was obtained by subtracting the total shrinkage from thermal
the sand and coarse aggregate with half of the mixing water, deformation. The thermal expansion coefficient of the

ACI Materials Journal/May-June 2010 233


Table 5—Drying shrinkage models
Prediction models
εsh = –kvs khskf ktd 0.48 × 10–3
AASHTO 2007

in which: khs= (2.00 – 0.14H); kvs = 1.45 – 0.13(V/S) ≥ 1.0; k f = ---------------- ; k td = ⎛ -------------------------------------⎞
35 t
7 + f ct′ ⎝ 61 – 0.58f ct′ + t ⎠
where khs is humidity factor; H is relative humidity, %; kvs is volume-to-surface ratio factor; kf is concrete strength factor; ktd is time development factor;
t is maturity of concrete, days; ti is age of concrete at drying, days; V/S is volume-to-surface ratio; and fci′ is specified compressive strength of concrete
at time of prestressing.
ACI 209R-92 AASHTO 2004

t
------------------------------------
0.0142 ( V/S )
⎛ t ⎞ 0.56 10 –3 26e +t 1064 – 3.70 ( V/S )
ε sh = – k s k h -------------- × × (steam-cured); k s = ------------------------------------- -------------------------------------------
⎝ 55 + t⎠ t 923
--------------
45 + t
where t is drying time, days; ks is size factor; and kh is humidity factor.
t – t sh, 0 ( t – t sh, 0 )
- ε sh, ∞ (moist-cured); ε sh ( t, t sh, 0 ) = ---------------------------------
ε sh ( t, t sh, 0 ) = --------------------------------- - ε sh, ∞ (steam-cured)
35 + ( t – t sh, 0 ) 55 + ( t – t sh, 0 )
where εsh(t, tsh,0) is shrinkage strain (mm/mm); t is time, days; tsh,0 is time at initiation of drying, days; and εsh, ∝ is ultimate shrinkage strain (mm/mm).

( t – tc )
- ; εs(fcm) = [160 + 10βsc(9 – 0.1fcm)] × 10–6; βRH = –1.55βARH; βARH = 1 – (RH/100)3
ε cso = ε s ( f cm ) ( β RH ) ---------------------------------------------------------
CEB-FIP 90

2
2A
⎛ 350 ⎛ ------------ c ⎞
+ ( t – t c )⎞
⎝ ⎝ 100μ⎠ ⎠
where, εcso is drying shrinkage (mm/mm); εs is drying shrinkage obtained from RH-shrinkage chart; βsc is cement type factor; βRH is relative humidity
factor; fcm is mean 28-day compressive strength, MPa; Ac is cross-sectional area, mm2; and μ is perimeter, mm.
εsh = εshuβ(h)β(t)
1/2 ( t – tc ) 0.5
× 10 ; β(h) = 1 – 1.18h4; β(t) = ⎛ --------------------------------------------
εshu = 1000k ⎛ ------------⎞ ⎞
GL 2000

30 –6
⎝ f cm28
′ ⎠ ⎝ t – t + 0.12 ( V/S ) 2⎠
c
where εsh is shrinkage strain (mm/mm); εshu is ultimate shrinkage strain (mm/mm); β(h) is correction term for humidity; β(t) is correction term for time;
h is humidity; t is age of concrete, days; tc is age of drying initiation, days; k is cement type factor, 1, 0.75, and 1.15 for Type I, II, and III cement, respectively.
Note: 1 mm = 3.94 × 10–2 in.; 1 MPa = 145 psi.

concrete was determined from the slope of the total deformation- under standard conditions with a relative humidity of 40%
temperature curve of concrete prisms subjected to control for both moist and steam curing made with Type GU and HE
temperature changes. Two prisms were immersed in water at cements. The CEB-FIP code is applicable to ordinary struc-
an approximate temperature of 50°C (122°F). Once the tural concrete with 28-day compressive strength ranging
temperature of the samples was stabilized, the water was between 12 and 80 MPa (1740 and 11,600 psi), relative
allowed to cool down to approximately 20°C (68°F). The humidity of 40 to 100%, and temperature of 5 to 30°C (41 to
resulting deformation was used to estimate the coefficient of 86°F). The GL 2000 model can be used for concrete propor-
the thermal expansion/contraction of the concrete. tioned with 0.40 to 0.60 w/cm and compressive strength up
Three 150 x 300 mm (6 x 12 in.) cylindrical test specimens to 70 MPa (10,150 psi).
were cast to monitor drying shrinkage. The specimens were
steam cured for 16 hours and were then demolded to grind TEST RESULTS AND DISCUSSION
their ends and fix external studs for deformation measure- Autogenous shrinkage
ments. A digital-type extensometer was used to determine A wide range of autogenous shrinkage was obtained
drying shrinkage. Shrinkage testing started at 18 hours. depending on mixture proportioning and materials in use.
Shrinkage specimens were monitored for 11 months at 23 ± Autogenous shrinkage strains of the 18 tested mixtures are
2°C (73.4 ± 3.6°F) and 50% ± 4% relative humidity. presented in Table 6 and Fig. 3. The autogenous shrinkage
It is also important to note that the structural performance strains reached stable values after approximately 2 months
of full-scale precast, prestressed bridge girders constructed of age and ranged from 95 to 245 microstrain for mixtures
using two SCC and two HPC mixtures was evaluated to made with Type MS cement and 170 to 345 microstrain for
verify the applicability of existing design provisions those with Type HE cement and 20% fly ash.
(AASHTO and PCI). The following aspects were studied: The highest 56-day autogenous shrinkage was obtained
constructibility, temperature variations, transfer lengths, for SCC 10 made with 0.34 w/cm, 500 kg/m3 (843 lb/yd3) of
cambers, flexural cracking, shear cracking, and shear binder, and Type HE cement with 20% fly ash. Compared to
strengths. Further information on the performance of the the SCC made with the same binder type, the two HPC
full-scale AASHTO girders can be found in NCHRP Report mixtures exhibited similar autogenous shrinkage at 56 days.
No. 628. The lowest autogenous shrinkage at 28 and 56 days was
measured for SCC 5 made with w/cm of 0.40, Type MS
Shrinkage prediction models cement, 440 kg/m3 (742 lb/yd3) of binder, and 0.46 S/A.
Shrinkage strains were compared to values that can be As expected, the w/cm has a significant influence on
predicted using the AASHTO 2007, AASHTO 2004, ACI autogenous shrinkage. The autogenous shrinkage of
209, CEB-FIP 1990, and GL 2000 models. Detailed model Mixtures 1 to 4 and Mixtures 5 to 8 made with a binder
descriptions are summarized in Table 5. The ACI 209 model content of 440 kg/m3 (742 lb/yd3) are presented in Fig. 3(a).
is applicable to normalweight and lightweight concrete The SCC made with 0.34 w/cm exhibited an autogenous

234 ACI Materials Journal/May-June 2010


shrinkage of 115 to 330 microstrain compared to 95 to
275 microstrain for SCC Mixtures 5 to 8 proportioned with
a higher w/cm of 0.40. Similarly, Mixtures 9 to 12 made with
a lower w/cm of 0.34 and binder content of 500 kg/m 3
(843 lb/yd 3) exhibited a higher autogenous shrinkage of
165 to 345 microstrain compared to 105 to 210 microstrain
for Mixtures 13 to 16 made with a w/cm of 0.40 and the same
binder content, as illustrated in Fig. 3(b).
The binder type had a considerable influence on autogenous
shrinkage. The SCC made with Type HE cement and 20%
fly ash exhibited higher autogenous shrinkage compared to
similar mixtures proportioned with Type MS cement,
regardless of the binder content, w/cm, S/A, and the use of
thickening-type VMA. For a given binder content of 440 kg/m3
(742 lb/yd3), SCC Mixtures 2, 4, 6, and 8 made with Type HE
cement and 20% fly ash exhibited autogenous shrinkage
values of 330, 280, 210, and 275 microstrain after 56 days of
measurement, respectively. The autogenous shrinkage
values of SCC Mixtures 1, 3, 5, and 7 were 115, 135, 95, and
100 microstrain, respectively. Similar results were found for
SCC proportioned with 500 kg/m3 (843 lb/yd3) of binder.

Drying shrinkage
The results of drying shrinkage for the 18 tested mixtures
are summarized in Table 7. The drying shrinkage of the 16
SCCs and the two HPC mixtures after 300 days ranged from
495 to 975 microstrain. In general, SCC exhibited 5 to 30%
higher drying shrinkage at 300 days, compared to HPC made
with a similar w/cm. Figures 4(a) and (b) show the variations
of drying shrinkage with time for Mixtures 1 to 4 and 9 to 12
proportioned with a w/cm of 0.34 and different binder contents
and Mixtures 5 to 8 and 13 to 16 made with 0.40 w/cm.
As compared in Fig. 4(a) and (b), the SCC made with
500 kg/m3 (843 lb/yd3) of binder exhibited a higher drying Fig. 3—Variations of autogenous shrinkage with time for SCC
shrinkage compared to similar mixtures proportioned with and HPC mixtures (refer to Table 6 for mixture codification).
440 kg/m3 (742 lb/yd3) of binder. For a given w/cm of 0.34,
Mixtures 9 to 12 made with a higher binder content of
500 kg/m3 (843 lb/yd3) exhibited greater drying shrinkage
values ranging between 630 and 965 microstrain after Table 6—Autogenous shrinkage strains at various
250 days compared to 510 and 815 microstrain for a similar ages (microstrain)
SCC prepared with a lower binder content of 440 kg/m 3 Time, days
(742 lb/yd 3). In particular, SCC Mixture 9 proportioned Mixture
no. Codification 14 28 56
with 0.34 w/cm, 500 kg/m3 (843 lb/yd3) of Type MS
1 34-440- MS-S/A54 95 105 115
cement, and an S/A of 0.54 exhibited the highest drying
2 34-440-HE20%FA-S/A46 280 310 330
shrinkage of 965 microstrain after 250 days of drying.
Similar results were found in the case of SCC proportioned 3 34-440-MS-S/A46-VMA 100 115 135
with 0.40 w/cm. It is important to note that, in this investigation, 4 34-440-HE20%FA-S/A54-VMA* 235 260 280
drying shrinkage includes both the intrinsic drying shrinkage 5 40-440-MS-S/A46 90 90 95
and the autogenous shrinkage. 6 40-440-HE20%FA-S/A54 165 190 210
7 40-440-MS-S/A54-VMA 95 100 100
Comparisons of drying shrinkage prediction models 8 40-440-HE20%FA-S/A46-VMA 200 245 275
Drying shrinkage data between 1 and 252 days are SCC
9 34-500- MS-S/A46 155 155 165
compared in Fig. 5 to values predicted by the AASHTO 10 34-500-HE20%FA-S/A54 285 315 345
2007, AASHTO 2004, ACI 209R, CEB-FIP 90, and GL 2000
11 34-500-MS-S/A54-VMA 215 230 245
models. The AASHTO 2007 model appeared to underestimate
12 34-500-HE20%FA-S/A46-VMA 270 305 330
shrinkage and resulted in larger scattering compared to the
AASHTO 2004 model. The AASHTO 2004 model provided 13 40-500-MS-S/A54 95 110 115
a better prediction of shrinkage data with a coefficient of 14 40-500-HE20%FA-S/A46 105 140 170
correlation factor (R2) of 0.75 compared to 0.71 for the 15 40-500-MS-S/A46-VMA 90 100 105
AASHTO 2007 model. The AASHTO 2004 model had a 16 40-500-HE20%FA-S/A54-VMA 125 170 210
mean predicted-to-measured shrinkage ratio of 0.87 17 34-MS 110 115 120
HPC
compared to 0.69 for the 2007 model. 18 38-HE20%FA 140 180 210
The AASHTO 2007, ACI 209R, and GL 2000 shrinkage *
34(0.34 w/cm)-440(binder content)-HE20%FA(binder type)-S/A54(0.54 S/A)-VMA
models were found to underestimate shrinkage for SCC with (VMA incorporation).

ACI Materials Journal/May-June 2010 235


Fig. 4—Variations of drying shrinkage with time for SCC
and HPC mixtures (refer to Table 7 for mixture codification). Fig. 5—Comparison of measured and predicted drying
shrinkage values.

Table 7—Drying shrinkage strains shrinkage values greater than 600 microstrain. In general,
at various ages (microstrain) these mixtures were proportioned with the relatively high w/cm
Time, days of 0.40 and a binder content of 500 kg/m3 (843 lb/yd3). As in
Mixture
no. Codification 28 252 300 the case of the CEB-FIP 90 model, the GL 2000 model uses
1 34-440- MS-S/A54 465 815 830 cement input parameters, including Type GU, MS, and HE
2 34-440-HE20%FA-S/A46 265 520 535
cements. The cement factor is not considered as an input
parameter in the AASHTO 2004 and 2007 models. The ACI
3 34-440-MS-S/A46-VMA 270 550 565
209 model is applicable for concrete made with Types GU
4 34-440-HE20%FA-S/A54-VMA 365 725 735
and HE cements. For the ACI 209R model, the ultimate
5 40-440-MS-S/A46 130 470 495 shrinkage was taken as that measured at 11 months.
6 40-440-HE20%FA-S/A54 230 530 545 The GL 2000 and ACI 209R models had an adequate
7 40-440-MS-S/A54-VMA 195 570 585 prediction for shrinkage values in the range of 200 to
SCC
8 40-440-HE20%FA-S/A46-VMA 270 665 670 500 microstrain. The CEB-FIP 90 model provided the best
9 34-500-MS-S/A46 550 965 975 fit of the measured drying shrinkage values for SCC with a
10 34-500-HE20%FA-S/A54 540 910 930 maximum shrinkage of 600 microstrain after 9 months. The
11 34-500-MS-S/A54-VMA 255 660 680 mean predicted-to-measured shrinkage ratio at 9 months was
12 34-500-HE20%FA-S/A46-VMA 330 630 635 0.90 compared to 0.69 to 0.87 for the other models. This can
13 40-500-MS-S/A54 225 670 690 be due to the fact that the CEB-FIP 90 model takes into
14 40-500-HE20%FA-S/A46 280 630 640
consideration the cement type that was changed for the
tested SCC mixtures (Type MS cement and Type HE cement
15 40-500-MS-S/A46-VMA 190 525 545
with 20% fly ash). A coefficient for cement Type A was
16 40-500-HE20%FA-S/A54-VMA 380 710 720
introduced for the AASHTO 2004 drying shrinkage model to
17 34-MS 230 560 570 provide a better prediction for the drying shrinkage. The
HPC
18 38-HE20%FA 380 625 630 coefficient is taken as the mean of eight coefficients derived

236 ACI Materials Journal/May-June 2010


shrinkage at 56 days varied between 100 and 350 microstrain
depending on the mixture composition. The majority of
autogenous shrinkage occurs in the first 28 days and reaches
approximately 85 to 95% of the ultimate value at 56 days.
The 14-day autogenous shrinkage is approximately 80% of
the 56-day value.
3. SCC exhibits 5 to 30% higher drying shrinkage compared
with HPC made with a similar w/cm at the age of 300 days.
4. For a given w/cm, SCC made with higher binder content
can exhibit higher drying shrinkage varying between 500
and 1000 microstrain after 300 days, depending on the
mixture composition.
5. The AASHTO 2004 model provides better prediction
compared with the AASHTO 2007 model. Moreover, the
AASHTO 2007, ACI 209, and GL 2000 shrinkage models
Fig. 6—Comparison of measured drying shrinkage and underestimate shrinkage values greater than 600 microstrain.
predicted values from modified models. 6. The CEB-FIP 90 Code can provide good prediction of
shrinkage values of SCC used in precast, prestressed
for SCC made with Type MS cement and eight others for applications. In addition, the AASHTO 2004 model with a
SCC made with Type HE cement with 20% fly ash. For each suggested modification for a cement coefficient can also be
mixture, a cement coefficient was derived to best fit the used to estimate drying shrinkage.
drying shrinkage data for that particular mixture with time
versus predicted shrinkage values for the AASHTO 2004 ACKNOWLEDGMENTS
model modified with a cement coefficient value. Finally, the The authors acknowledge the financial support provided by the Transportation
cement coefficients of 0.918 and 1.065 were suggested for Research Board (TRB) of the National Academies (NAS-NRC) of the
SCC made with Type MS cement and Type HE with 20% fly United States of America for NCHRP Project 18-12 (NCHRP Report 628).
The assistance of S.-D. Hwang and G. Lemieux during this project is especially
ash, respectively. A coefficient for cement type was added to acknowledged.
the AASHTO 2004 drying shrinkage model as an input
parameter to enhance the prediction of drying shrinkage. The REFERENCES
modified AASHTO 2004 model can then provide a mean 1. ACI Committee 237, “Self-Consolidating Concrete (237R-07),”
predicted-to-measured shrinkage of 0.94 with a relatively American Concrete Institute, Farmington Hills, MI, 2007, 30 pp.
low scattering of the data from the 1:1 equality line (R2 = 2. Khayat, K. H., “Workability, Testing, and Performance of Self-
0.89). This model, along with the CEB-FIP 90 model, could Consolidating Concrete,” ACI Materials Journal, V. 96, No. 3, May-June
1999, pp. 346-353.
then be used to predict drying shrinkage of SCC used for 3. Khayat, K. H.; Assaad, J.; and Daczko, J., “Comparison of Field-Oriented
precast, prestressed applications, as illustrated in Fig. 6. Test Methods to Assess Dynamic Stability of Self-Consolidating Concrete,”
The modified AASHTO 2004 model for SCC can then be ACI Materials Journal, V. 101, No. 2, Mar-Apr. 2004, pp. 168-176.
expressed as 4. Aïtcin, P. -C.; Neville, A. M.; and Acker, P., “Integrated View of
Shrinkage Deformation,” Concrete International, V. 19, No. 9, Sept. 1997,
pp. 35-41.
εsh = –kskh ⎛ --------------⎞ × 0.56 × 10 × A (steam-cured) (1)
1 –3 5. Davis, H. E., “Autogenous Volume Change of Concrete,” Proceedings,
⎝ 55 + t⎠ 43th Annual American Society for Testing Materials Meeting, Atlantic
City, NJ, June 1940, pp. 1103-1113.
6. Persson, B., “Creep, Shrinkage and Elastic Modulus of Self-
Compacting Concrete,” Proceedings of the 1st International RILEM
t
--------------------------------------
0.0142 ( V/S )
- Symposium on Self-Compacting Concrete, Stockholm, Sweden, 1999,
26e +t 1064 – 3.70 ( V/S ) pp. 239-250.
k s = --------------------------------------- ------------------------------------------ 7. Persson, B., “A Comparison between Mechanical Properties of Self-
t -
------------- 923 Compacting Concrete and the Corresponding Properties of Normal
45 + t Concrete,” Cement and Concrete Research, V. 31, No. 2, Feb. 2001, pp. 193-198.
8. Vieira, M., and Bettencourt, A., “Deformability of Hardened SCC,”
Proceedings of the 3rd International Symposium on Self-Compacting
where t is drying time, in days; ks is the size factor; kh is the Concrete, Reykjavik, Iceland, 2003, pp. 606-618.
humidity factor; V/S is the volume-to-surface ratio (mm3/mm2); 9. Pons, G.; Proust, E.; and Assié, S., “Creep and Shrinkage of Self-
and A is the cement factor: 0.918 for Type MS cement and Compacting Concrete: A Different Behaviour Compared with Vibrated
1.065 for Type HE + 20% fly ash. Concrete?” Proceedings of the 3rd International RILEM Symposium on
Self-Compacting Concrete, Reykjavik, Iceland, 2003, pp. 645-654.
10. Attiogbe, E. K.; See, H. T.; and Daczko, J. A., “Engineering Properties
CONCLUSIONS of Self-Consolidating Concrete,” Proceedings of the 1st North American
Based on the test results of the experimental program and Conference on the Design and Use of SCC, Chicago, IL, 2002, pp. 331-336.
the detailed comparisons among five shrinkage prediction 11. Bouzoubaâ, N., and Lachemi, M., “Self-Compacting Concrete
models, the following conclusions can be drawn for SCC Incorporating High Volumes of Class F Fly Ash: Preliminary Results,”
Cement and Concrete Research, V. 31, No. 3, Mar. 2001, pp. 413-420.
proportioned for precast, prestressed structural applications: 12. Heirman, G., and Vanderwalle, L., “The Influence of Fillers on the
1. Autogenous shrinkage is highly affected by w/cm and Properties of Self-Compacting Concrete in Fresh and Hardened State,”
binder content. SCC made with Type HE cement with 20% Proceedings of the 3rd International RILEM Symposium on Self-
Class F fly ash can exhibit higher autogenous shrinkage Compacting Concrete, Reykjavik, Iceland, 2003, pp. 606-618.
compared to SCC with Type MS cement. 13. Song, H. W.; Byun, K. J.; Kim, S. H.; and Choi, D. H., “Early-Age
Creep and Shrinkage in Self-Compacting Concrete Incorporating GGBFS,”
2. Compared to HPC made with a given binder type, SCC Proceedings of the 2nd International Symposium on Self-Compacting
is shown to develop similar autogenous shrinkage. Autogenous Concrete, Tokyo, Japan, 2001, pp. 413-422.

ACI Materials Journal/May-June 2010 237


14. Hu, C., and Barcelo, L., “Investigation on the Shrinkage of Self- of the 3rd International RILEM Symposium on Self-Compacting
Compacting Concrete for Building Construction,” Proceedings of the 1st Concrete, Reykjavik, Iceland, 2003, pp. 596-605.
Symposium International Conference on Self-Compacting Concrete, Kochi, 18. Turcry, P.; Loukili, A.; and Haidar, K., “Mechanical Properties,
Japan, 1998, pp. 228-242. Plastic Shrinkage and Free Deformations of Self-Consolidating Concrete,”
15. Hammer, T. A., “Cracking Susceptibility Due to Volume Change of Proceedings of the 1st North American Conference on the Design and Use
Self-Compacting Concrete,” Proceedings of the 3rd International RILEM of Self-Consolidating Concrete, Center for Advanced Cement-Based Materials,
Symposium on Self-Compacting Concrete, Reykjavik, Iceland, 2003,
Northwestern University, Evanston, IL, Nov. 2002, pp. 301-306.
pp. 553-557.
16. Turcry P., and Loukili, A., “A Study on Plastic Shrinkage of Self- 19. Holt, E., and Schodet, O., “Self-Consolidating Concrete: Early-Age
Compacting Concrete,” Proceedings of the 3rd International RILEM Shrinkage,” Internal Report (RTE40-IR-21/2002), Technical Research
Symposium on Self-Compacting Concrete, Reykjavik, Iceland, 2003, Center of Finland, VTT Building and Transport, 2002, 26 pp.
pp. 576-585. 20. Byun, K. J.; Kim, J. K.; and Song, H. W., “Self-Compacting Concrete in
17. Klug, Y., and Holschemacher, K., “Comparison of the Hardened Korea,” Proceedings of the International Workshop on Self-Compacting
Properties of Self-Compacting and Normal Vibrated Concrete,” Proceedings Concrete, Kochi, Japan, 1998, pp. 23-33.

238 ACI Materials Journal/May-June 2010

You might also like