This article was downloaded by: [American Public University System]

On: 26 February 2014, At: 03:09

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Sustainable Cement-Based

Materials
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/tscm20

Drying shrinkage and creep

performance of geopolymer concrete
a b a a
Kwesi Sagoe-Crentsil , Trevor Brown & Alan Taylor
a
CSIRO Materials Science and Engineering , Highett , Australia
b
Faculty of Engineering & Industrial Sciences , Swinburne
University of Technology , Hawthorn , Australia
Published online: 20 Feb 2013.

To cite this article: Kwesi Sagoe-Crentsil , Trevor Brown & Alan Taylor (2013) Drying shrinkage and
creep performance of geopolymer concrete, Journal of Sustainable Cement-Based Materials, 2:1,
35-42, DOI: 10.1080/21650373.2013.764963

To link to this article: http://dx.doi.org/10.1080/21650373.2013.764963

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
 Journal of Sustainable Cement-Based Materials, 2013
Vol. 2, No. 1, 35–42, http://dx.doi.org/10.1080/21650373.2013.764963

Drying shrinkage and creep performance of geopolymer concrete

Kwesi Sagoe-Crentsila,b*, Trevor Browna and Alan Taylora
a
CSIRO Materials Science and Engineering, Highett, Australia; bFaculty of Engineering &
Industrial Sciences, Swinburne University of Technology, Hawthorn, Australia
Downloaded by [American Public University System] at 03:09 26 February 2014

The creep behavior and drying shrinkage performance of fly ash geopolymer (GP)
concrete mixtures have been investigated using equivalent grade 40-MPa ordinary
Portland cement (OPC) concrete as the reference system. Drying shrinkage values,
measured in accordance with AS 1012.13-1992, up to one year fell well below the
nominal 700 μstrain limit, with GP concrete values typically less than 400 μstrain at
1 year. Variations in basic mean creep coefficient of concrete (Фcc.b), measured as
the ratio of the creep strain to elastic strain for a specimen loaded at 28 days under
constant stress of 0.4fc (fc – 28 day compressive strength) was monitored for up to
52 weeks. Values of Фcc.b obtained for steam-cured GP concrete was found to be of
the order of 45% lower than corresponding OPC concrete. Based on these results,
the effects of GP concrete mixtures on load-dependent (creep) deformations appear
to be negligible. The paper further discusses factors that contribute to observed GP
concrete creep deformation and drying shrinkage performance.
Keywords: geopolymer; fly ash; drying shrinkage; creep; compressive strength

1. Introduction are available concerning the effects of

The prediction of concrete creep and dry-
ing shrinkage remains a critical parameter
in concrete specification and is of crucial
importance for durability and long-time
serviceability of concrete structures.[1]
While creep of concrete is associated
with deformation of hardened concrete
due to applied constant load, drying
shrinkage refers to internal moisture loss
of the hardened concrete. In ordinary
Portland cement (OPC) systems, both
creep and drying shrinkage are known to
be affected by variable factors including
cement type, w/b ratio, aggregate type
and content, age, temperature relative
humidity of the environment, and age
and size of member.[2] In the literature, a
considerable amount of experimental data
each of these variables on creep and
shrinkage behavior of concrete made with
OPC and blended cements.[3–4] Drying
shrinkage, generally leads to cracking
and, even though it may not affect the
structural integrity, durability problems
are generally increased and is perhaps
one of the most deleterious properties of
Portland cement concrete.[5] Correspond-
ing studies investigating various aspects
of mix composition and engineering per-
formance of geopolymer (GP) systems
including shrinkage and creep are now
only beginning to emerge [5–9].
Hardjito et al. [10], in their study of
creep of medium strength GP concrete,
observed lower creep coefficient values
compared to equivalent grade OPC

*Corresponding author. Email: Kwesi.Sagoe-Crentsil@csiro.au

Ó 2013 Taylor & Francis

 36 K. Sagoe-Crentsil et al.

concretes. These researchers furthermore 2. Experimental

showed that the specific creep of GP con- 2.1. Raw materials
crete decreased as the compressive
Table 1 shows chemical composition of
strength increased. This observation cor-
ASTM Type I OPC used as reference for
relates well with conventional OPC con-
concrete mixtures and Class F fly ash
crete.[11–14] Typical literature values of
used for the GP mixtures. Commercial
specific creep for conventional 50–60-
“N” grade sodium silicate solution with
MPa Portland cement concrete after one
SiO2/Na2O molar ratio of 3.22 and
year are of the order of 50–60 micro-
28.7% SiO2 content with total solids con-
strain,[12] with this value decreasing for
tent of a 38% supplied by PQ Australia
higher strength concretes. In the case of
was used in mixtures.
Downloaded by [American Public University System] at 03:09 26 February 2014

high-performance high-volume fly ash

River sand and local 9-mm Hornfels
concrete, for instance, Malhotra and
coarse aggregate was used. The coarse
Mehta [15] reported specific creep was in
aggregate fraction was received in a satu-
the region of 30 microstrain/MPa after
rated surface-dry condition.
one year.
While the roles of curing on
mechanical properties of GP concretes
2.2. Specimen preparation and curing
are yet to be fully established, it has
been observed that the drying shrinkage Table 2 provides concrete mix propor-
strains of heat-cured GP concrete are tions of the concrete formulations. Both
generally lower compared to equivalent the reference OPC and GP concrete
values observed for concretes cured mixes were designed to achieve target
under ambient conditions,[16] This 28-day strength of 40 MPa.
effect was attributed to water which is The reference OPC concrete samples
released during the chemical reaction were prepared in a similar fashion as GP
process of ambient-cured GPs which concrete and molded according to stan-
subsequently evaporate over a period of dard requirements for compressive
time causing significantly large drying strength measurements. After casting, all
shrinkage strains particularly within the samples were placed in the steam curing
initial two weeks. chamber at 65 °C for 6hrs. Thereafter, all
Other material-related parameters that samples, were stored at 23 °C and 100%
can influence engineering performance of RH until required for testing. Standard
GP binders, such as the pore network concrete cylinders of 100
200 mm were
distribution, have also been investigated.
[17,18] Kriven et al. [17] examined the
pore network distribution of GP binders Table 1. Chemical analysis of OPC and fly
and observed that there exist several ash (mass %).
clusters of pore sizes and found them
not too dissimilar to those characterizing Oxide OPC Fly ash
OPC systems. In this paper, the early- SiO2 20.2 47.19
age basic creep behavior of GP concrete Al2O3 4.16 29.79
is investigated using equivalent grade Fe2O3 5.30 13.93
Portland cement concrete as the refer- CaO 64.8 3.29
MgO 1.29 1.38
ence system. The experiments assess the Na2O 0.22 0.24
influence of age and loading on the K2O 0.42 0.49
actual creep in early age, while measur- TiO2 – 1.77
ing drying shrinkage response of con- SO3 2.67 0.13
crete specimens. LOI 1.34 1.3
 Journal of Sustainable Cement-Based Materials 37

Table 2. Concrete mix proportions. accordance with a modified procedure of

AS 1012.1 -1993. The cast specimens
OPC were stored in the fog room (set at 23 °C
concrete (kg/ GP concrete
Material m3 ) (kg/m3) and 100% RH) for 6 days and then stored
in a standard temperature and humidity
Fly ash – 269 controlled room (at 23 °C and 50% RH)
OPC 346 –
Coarse 1112 1163 and measurement taken at regular inter-
aggregate vals.
(9 mm)
Silicate – 148
solution/ 2.5. Creep
NaOH
Downloaded by [American Public University System] at 03:09 26 February 2014

Creep measurements were carried out on

Dry sand 753 784
Water 186 27
the prisms cut from panels after 28 days
of casting, measuring 150
150

300 mm, in accordance with Australian
Standard as 1012.16-1996 (methods of
used for compressive strength measure-
testing concrete – determination of creep
ments. Table 3 provides data on mea-
of concrete cylinders in compression).
sured fresh concrete properties.
The load for determining creep was
applied after 2 days and readjusted after 7
2.3. Compressive strength and 28 days. The loads in all creep tests
The compressive strength was measured corresponded to 40% of the compressive
in accordance with AS 1012.9-1999 [19] strength of the specific concrete mixtures
at a loading rate of 20 MPa/min – Stan- at the time of loading. Total strain was
dard Test Method for Compressive measured on specimens in the creep test
Strength of Cylindrical Concrete Speci- rig, while drying shrinkage strain was
men. Specimens were capped using con- obtained from parallel unloaded speci-
ventional sulfur capping method in mens in a separate complementary creep
accordance with the Australian Standard test rig under same conditions as for the
AS 1012.9 (1999) for concrete compres- creep measurements. The calculated load-
sive strength tests at 1, 3, 7, and 28 days. induced creep strain at any time is the
Tests were performed in a 2500 kN ELE difference between the average strain val-
Universal testing machine. ues of the loaded and control specimens.
The creep loads used in the current test
were calculated based on 40% of 28-day
2.4. Drying shrinkage strength.
The standard drying shrinkage measure-
ment of the aerated GP products was per-
formed on 75
75
300-mm prisms in 3. Results and discussion
3.1. Compressive strength
The mean compressive strength develop-
Table 3. Properties of fresh concrete. ment of steam-cured OPC and GP con-
cretes is shown in Table 4. The results in
Property OPC concrete GP concrete Table 4 show that GP and OPC samples
Mix temp. (°C) 15.7 25.5 varied slightly from the target strengths
Flow (mm) – 260 of 40 MPa at 28 days. It is also observa-
Slump (mm) 120 160 ble that different strength development
Wet density 2380 2430
(kg/m3)
trends occur for the two heat-cured con-
crete types up to 28 days.
 38 K. Sagoe-Crentsil et al.
Table 4. Compressive strength development
of ambient cured OPC and GP concretes.
Compressive strength (MPa)
Age (days) OPC steam-cure GP steam-cure
1 22.5 42.3
7 28.0 42.3
28 38.1 44.0
The heat-cured OPC specimens typi-
cally show greater strength development,
Downloaded by [American Public University System] at 03:09 26 February 2014
achieving nearly 40% increase between
1 day and 28 days during fog-room stor-
age. However, only a corresponding mar-
ginal increase in compressive strength of
less than 5% is observed between 1 day
and 28 days for the heat-cured GP sam-
ples. This strength increase is a result of
residual polymerization reactions that
occur beyond the initial heat exposure.
Furthermore, only a relative modest
Figure 1. (a) Uniaxially loaded specimens under constant temperature and humidity conditions.
(b) Close-up of loaded creep specimens.
strength increase occurs between 7 days
and 28 days for GP concretes, which is
typical of this class of materials, suggest-
ing that significant polymerization reac-
tion occurs during the initial heat curing
regime compared to equivalent hydration
reactions characterizing OPC systems
(see Figure 1).
3.2. Drying shrinkage
The drying shrinkages of the mixes are
shown in Figure 2, where it can be seen
that the OPC mixtures showed excellent
drying shrinkage properties both at early
ages and at ages up to one year. It is evi-
dent from the trends shown in the plots
that the drying shrinkage steam-cured
samples of both GP and OPC concrete
fell well below the 700 μstrain limit
specified in AS 3600-2001.[20]
 Journal of Sustainable Cement-Based Materials
0
0 50
-100
-200
MICROSTRAIN
-300
-400
-500
Downloaded by [American Public University System] at 03:09 26 February 2014
-600
Figure 2. Drying shrinkage for 40-MPa GP and OPC concrete.
The drying shrinkage trends of the
OPC and GP concrete at early ages (i.e.
<7 days) were very similar. At this early
age, as expected, both binders register
rapid increase in drying shrinkage. Given
the 7-day moist curing of the specimens
prior to shrinkage measurements, the initial
high shrinkage of the GP samples could
reflect the time taken for them to equili-
brate with the surrounding environment.
The average change in shrinkage values
between 7 and 56 days of the GP samples
was of the order of 250 μstrain. This value
is much lower compared to the value of
400–450 μstrain experienced by Portland
cement concrete in agreement with obser-
vations of Hardjito et al. [21] for heat-
cured fly ash-based GP concrete after one
year.
Similar to OPC concrete, it may be
expected that the main mechanism for
drying shrinkage in GP concrete is the
build-up of negative pressure within the
capillary network of the bulk cement
paste as meniscii form. The resulting
stresses lead to contraction of the con-
crete unless restrained.[22] The observed
results suggest that increased condensa-
tion processes occurring in steam-cured
GP concretes may have implications on
the shrinkage process. Thus, when the
condensation reaction of GP concretes
39
100 150 200 250 300 350 400
GP
OPC
TIME (DAYS)
are accelerated by heat curing, the associ-
ated initial drying shrinkage was signifi-
cantly reduced, signifying the completion
of initial condensation reactions and
moisture loss from capillaries, as also
confirmed by the associated rapid
strength development.
The lower drying shrinkage observed
for GP concretes may also be partly due
to a finer disconnected capillary network
structure. Correspondingly, contributions
of chemical or autogenous shrinkage fac-
tors of GP concrete shrinkage also remain
unknown. Chemical shrinkage has a pro-
found effect on the macroscopic deforma-
tion of the concrete volume. For instance,
in OPC concrete, the dimensional
changes caused by chemical shrinkage
are a result of the hydration reactions that
occur when the anhydrous cement phases
react with available water. Thus, the vol-
ume change caused by chemical shrink-
age occurs when the concrete is in a
plastic state. Overall, early age concrete
shrinkage is of increasing concern, as it
can be responsible for cracking when the
concrete has not gained significant
strength to withstand internal stresses.
3.3. Creep
Figure 3 shows variation of the basic
creep coefficient with time up to
 40 K. Sagoe-Crentsil et al.
Downloaded by [American Public University System] at 03:09 26 February 2014
Figure 3. Average basic creep coefficient for steam cured 40-MPa grade OPC and GP concrete.
52 weeks. As shown in the figure, basic
creep coefficient for 40-MPa grade GP
concrete is 40–60% lower than the corre-
sponding OPC concrete. Creep is the
time dependant increase in strain under
sustained stress for a given material. At
the application of a load, the instanta-
neous strain includes elastic strain and
some creep strain.
Basic creep is taken as the increase in
strain ½ðtðtÞ above the initial elastic
strain under sustained loading, minus
shrinkage strains over the same time
period. Therefore,
cbðtÞ ¼ TðtÞ  EðtÞ  sðtÞ ð1Þ
where at any time t after loading:
cbðtÞ = basic creep strain
E = elastic strain upon loading
sðtÞ = shrinkage
Australian Standard AS3600-2009
(Concrete structures) Section 3.1.8.2 [16]
gives the basic creep coefficient of con-
crete (Фcc.b), as the ratio of the creep strain
to elastic strain for a specimen loaded
at 28 days under constant stress of 0.4fc
(fc – 28 day compressive strength). Thus,
basic creep coefficient is an indicator of
concrete deformation due to loading; thus,
lower creep coefficient represents a higher
resistance to deformation under load. The
basic creep coefficient, determined from a
standard test, can where necessary, be
extrapolated to a final test value by any
acceptable mathematical model for creep
behavior, calibrated such that Фcc.b is also
predicted by the chosen model.
From the instantaneous creep, the
basic creep factor has been calculated for
test specimens, as illustrated in Figure 3.
After adding trend line of best fit, the
resulting mathematical formula or model,
for Basic Creep coefficient is Фcc.b =
0.2687ln(x)  0.168. The creep coeffi-
cient, defined as the ratio of creep
strain-to-elastic strain after one year of
loading for the GP concrete with 40 MPa
compressive strength is of the order of
0.10. The values are well within the rec-
ommended range for conventional con-
crete as specified in the Australian
Standard AS3600 for Portland cement
concrete.[20]
Wallah [7] obtained creep coefficient
values between 0.6 and 0.7 for heat-
cured GP concrete after one year for
concrete with compressive strength of
40, 47, and 57 MPa and 0.4 for GP
concrete with compressive strength of
67 MPa.
While the dependence of creep strain
on factors such as binder loading, aggre-
gate type and ratio, air entrainment as
 Journal of Sustainable Cement-Based Materials
well as the age of the concrete at the time
of loading, are well established; the rela-
tive effects of these factors on GP con-
crete remain uncertain. Hence, in
comparing current results with published
work, such as that of Hardjito et al. [21],
it is vital to recognize contributions aris-
ing from differences in GP mix composi-
tions and cure conditions. However,
notwithstanding results obtained from the
current study are in good agreement with
Downloaded by [American Public University System] at 03:09 26 February 2014
the limited published data for GP con-
cretes which fall well within specification
range of OPC concretes as stipulated, for
example, in the Australian standard AS
3600.[20]
The scope of the current investiga-
tion did not allow for more comprehen-
sive testing to determine the influence
of stress levels or other materials and
process parameters. Hence, further test-
ing need to be undertaken to develop a
more comprehensive data-set for creep
behavior and also to benchmark against
the performance of conventional con-
cretes.
It is further worth noting that similar
to the current results, the measured creep
strains of fly ash-based GP concrete
obtained by Wallah [7] were significantly
smaller than predicted equivalent grade
Portland cement concrete. While the
exact cause for this trend remains
unclear, the observed smaller creep
strains of fly ash-based GP concrete may
be also due in-part to unreacted or
partially reacted fly-ash particle residues
acting as ‘micro-aggregates’ in the sys-
tem, as suggested by Davidovits.[9]
Hence, the additional restraining action
due to micro-aggregates could potentially
contribute to smaller creep compared to
traditional OPC concrete for which
cement hydration reaction is relatively
more complete. Particularly, given that
the aggregate fraction remains primarily
responsible for counteracting creep and
drying shrinkage deformations of the bin-
der phases.
41
4. Conclusions
The following observations have been
made based on laboratory drying shrink-
age and creep tests for nominal 40-MPa
GP and OPC concretes samples up to
52 weeks:
(1) Test results obtained from this
study disclose that primary struc-
tural properties of GP concretes,
namely, compressive strength, dry-
ing shrinkage, and creep behavior
broadly conforms to existing stan-
dards and guidelines for OPC sys-
tems.
(2) All mixes displayed excellent dry-
ing shrinkage properties, as mea-
sured in accordance with AS
1012.13-1992. Shrinkage values
obtained at both early and latter
ages up to one year fall well
below the nominal 700 μstrain
limit (Australian Standard for
Concrete Structures AS3600 -
2005), with GP concrete values
typically less than 400 μstrain at
one year.
(3) Similar to OPC concrete, the main
mechanism for drying shrinkage
in GP concrete appears similar to
the build-up of negative pressure
within the capillary network of the
bulk cement paste as meniscii
form which generate stresses
resulting in contraction of OPC
concrete. However, acceleration of
condensation reactions of heat-
cured GP concretes suggest
significant mitigation of early age
drying shrinkage arising from the
combined effects of nearly
completed condensation reactions
as confirmed by associated rapid
initial strength development.
(4) Variations in basic mean creep
coefficient of concrete (Фcc.b)
obtained for steam-cured GP con-
crete was found to be of the order
 42 K. Sagoe-Crentsil et al.
of 45% lower than corresponding
OPC concrete. The combined
effects of greater shrinkage occur-
ring within OPC material, on the
one hand, and possible restraining
effects due to unreacted fly ash
residue particles acting as micro-
aggregates in GP concretes appear
to be major contributors to
observed differences in creep per-
formance of the two concrete
Downloaded by [American Public University System] at 03:09 26 February 2014
types.
References
[1] Gilbert RI. Creep and shrinkage models
for high strength concrete - proposal for
inclusion in AS3600. Austr. J. Struct.
Eng. 2002;4:95–106.
[2] Hanson W. Early age creep and stress
relaxation tests. In: Bentur A, Kovler K,
editors. Early age cracking in cementi-
tious systems. Cachan Cedex: RILEM
Publications S.A.R.L; 2002. 257–65.
[3] Standards-Australia, methods of testing
concrete - Determination of creep of con-
crete cylinders in compression, AS
1012.16-1996.
[4] Nawy Edward G, editor. Chapter 26 in
concrete construction engineering hand-
book. 2nd ed. New York (NY): CRC
Press; 2007.
[5] Collins F, Sanjayan JG. Cracking
Tendency of Alkali-Activated Slag Con-
crete Subjected to Restrained Shrinkage,
Cement and Concrete Research, Vol. 30,
No. 5, 2000, pp. 791–798.
[6] Krivenko PV. Alkaline cements: termi-
nology, classification aspects of durabil-
ity. In: Proceedings of 10th International
Congress on the Chemistry of Cement;
1997; Goteborg (Sweden). p. 4iv046–
4iv050.
[7] Wallah SE. Creep behaviour of fly ash-
based geopolymer concrete. Civil Eng.
Dimen. 2010;12:73–78.
[8] Davidovits J. Properties of geopolymer
cements. In: Krivenko PV, editor. Alkaline
cements and concretes, Vol. 1. Vipol
Stock. Ukraine: Kiev; 1994. 131–49.
[9] Davidovits J. Soft mineralurgy and geo-
polymers. In: Proceedings of Geopoly-
mer ‘88, First European Conference on
Soft Mineralurgy; Compiegne, France:
The Geopolymer Institute; 1988.
[10] Hardjito D, Wallah SE, Sumajouw DMJ,
Rangan BV. On the development of fly
ash-based geopolymer concrete. ACI
Mater. J. 2004;101:467–472.
[11] Neville AM. Properties of concrete
(Fourth and Final ed.). Essex: Pearson
Education, Longman Group; 2000.
[12] Warner RF, Rangan BV, Hall AS, Faul-
kes KA. Concrete structures. Melbourne:
Longman; 1998.
[13] Neville AM, Dilger WH, Brooks JJ.
Creep of plain and structural concrete.
London: Construction Press, Longman
Group; 1983.
[14] Gilbert RI. Time effects in concrete
structures. Amsterdam: Elsevier; 1988.
[15] Malhotra VM, Mehta PK. High-perfor-
mance, high-volume fly ash concrete: mate-
rials, mixture proportioning, properties,
construction practice, and case histories.
Ottawa: Supplementary Cementing Materi-
als for Sustainable Development; 2002.
[16] Rangan VB. Developments in Porous, bio-
logical and geopolymer ceramics: ceramic
engineering and science. In: Manuel Brito,
Eldon Case and Waltraud M. Kriven edi-
tors. Proceedings, vol. 28, issue 9, Devel-
opments in porous, biological and
geopolymer ceramics. The American Cera-
mic Society; Daytona Beach, Fl 2008.
[17] Kriven WM, Bell JL, Mallicoat SW,
Gordon M. Intrinsic microstructure and
properties of metakaolin-based geopoly-
mers’. In: International workshop on
geopolymer binders, interdependence of
composition, structure and properties;
Weimar, Germany. Edited by Bauhaus-
Universitat; 2006, p. 71–86.
[18] Steveson M, Sagoe-Crentsil K. Relation-
ships between composition, structure and
strength of inorganic polymers. J. Mater.
Sci. 2005;40:2023–2036.
[19] Australian Standard AS 1012.9-1999
Methods of testing concrete - determina-
tion of the compressive strength of con-
crete specimens. Standards Australia.
19999.
[20] Standards-Australia, concrete structures,
Draft for Public Comment Standards Aus-
tralia (Revision of AS 3600–2001); 2005.
[21] Hardjito D, Wallah SE, Sumajouw DMJ,
Rangan BV. Fly ash-based geopolymer
concrete. Aust. J. Struct. Eng.
2005;6:77–86.
[22] Bissonnette B, Pierre P, Pigeon M. Influ-
ence of key parameters on drying shrink-
age of cementitious materials. Cement
Concr. Res. 1999;29:1655–1662.