Construction and Building Materials 73 (2014) 326–331

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Stress–strain behaviour of confined Geopolymer concrete

N. Ganesan a,1, Ruby Abraham b, S. Deepa Raj b,⇑, Divya Sasi b
a
Department of Civil Engineering, National Institute of Technology Calicut (NITC), Kerala, India
b
Department of Civil Engineering, College of Engineering Trivandrum (CET), Kerala, India

h i g h l i g h t s

 Studied the behaviour of PCC and GPC specimens.

 Stress–strain behaviour in uniaxial compression was studied.
 Effect of confinement on the stress–strain behaviour of GPC and PCC were studied.
 Proposed an analytical model for the stress–strain behaviour of confined GPC.

a r t i c l e i n f o a b s t r a c t

Article history: Fly ash based Geopolymer concrete (GPC) is an environment friendly alternative to conventional concrete
Received 20 April 2014 made from alkali activated aluminosilicate and aggregate. This study intends to examine the effects of
Received in revised form 29 August 2014 confinement on the behaviour of GPC and conventional Portland Cement Concrete (PCC). Out of the 36
Accepted 24 September 2014
cylinders tested under monotonic loading 24 cylinders were made with GPC and the remaining with
Available online 17 October 2014
PCC. The variable considered in this study is the volumetric ratio of confinement. An analytical model
is proposed for the stress–strain behaviour of confined GPC. The results showed that confinement
Keywords:
reinforcement greatly improved the strength and ductility of GPC than PCC.
Geopolymer
Fly ash
Ó 2014 Elsevier Ltd. All rights reserved.
Alkali-activated cement
Confinement

1. Introduction sodium hydroxide and sodium silicate solution, which is a poly-

merization process that differs widely from Portland cement
An important ingredient in conventional Portland Cement Con- hydration [3]. Thus the source material (fly ash), alkaline solution
crete (PCC) is Ordinary Portland Cement (OPC). The production of and aggregates constitute the main components of Geopolymer
cement is increasing by about 3% annually. The process of cement concrete (GPC). The properties of GPC include high early strength,
production is highly energy intensive and releases large volume of low shrinkage, excellent freeze–thaw resistance, sulphate resis-
greenhouse gases like CO2 [1]. Thus the cement industry is respon- tance and corrosion resistance [2,4]. Confinement is an important
sible for some of the greenhouse gas emissions into the atmo- factor which affects the behaviour of concrete. Since sectional
sphere. In this respect, geopolymer technology introduced by strength and ductility depend on the stress–strain characteristics
Davidovits provides an alternative low emission binding agent to of concrete they are also influenced by the confinement of the
PCC [2]. Geopolymers are inorganic aluminosilicates produced by members. The properties of confined concrete have been exten-
alkali activation of materials of geological origin such as kaolin or sively studied in the past [5,6]. Confinement increases the com-
bentonite or byproduct materials such as fly ash or rice husk ash. pressive strength and the capacity of concrete to sustain large
Thus Geopolymer concrete (GPC) is obtained by alkali activation deformations without substantial loss of strength. From the litera-
of industrial waste materials such as fly ash in the presence of ture review, it is found that the stress–strain behaviour varies con-
siderably depending on the type of concrete and the confinement.
Although information on the confinement of normal concrete is
⇑ Corresponding author. Tel.: +91 04712539280.
available in literature, the effect of confinement on GPC has been
E-mail addresses: [email protected], [email protected] (N. Ganesan),
only scantly investigated. This study has attempted to obtain
[email protected] (R. Abraham), [email protected] (S. Deepa Raj),
[email protected] (D. Sasi). experimentally the stress–strain curve of confined GPC and
1
Tel.: +91 04952286204. develop analytical models for the same.

http://dx.doi.org/10.1016/j.conbuildmat.2014.09.092
0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.
 N. Ganesan et al. / Construction and Building Materials 73 (2014) 326–331 327

2. Experimental investigation

2.1. Materials and mix proportions

Fly ash (ASTM Class F) was used as the main source material for synthesizing
the geopolymer binder. Coarse aggregate of 20 mm nominal size was used for mak-
ing GPC and PCC. Locally available river sand conforming to zone II as per IS: 383-
1970 was used as fine aggregate. A combination of sodium silicate solution and
sodium hydroxide (NaOH) solution was chosen as the alkaline liquid to activate
the source material. Commercially available sodium silicate solution with SiO2–
to-Na2O ratio by mass of 2 (Na2O = 14.7%, SiO2 = 29.4%) and water = 55.9% by mass
were used for the study. A naphthalene based superplasticiser was also used to
improve the workability of GPC. Ordinary Portland cement of 53 grade was used
for preparing PCC. HYSD bars of 6 mm diameter and 360 N/mm2 yield strength
were used for making spiral reinforcement cages of 90 mm diameter. The pitch of
the spirals used were 75 mm, 50 mm and 25 mm (volumetric ratios 1.36%, 2.05%
and 4.1%). The GPC and PCC specimens were designated as GPCP1, GPCP2, GPCP3
and PCCP1, PCCP2, PCCP3 corresponding to pitches of 75 mm, 50 mm, 25 mm
respectively. GPCP0 and PCCP0 represents unconfined GPC and PCC specimens.
Since there are no codal recommendations available for the mix design of GPC,
Fig. 1. Test set-up.
different trial mixtures of GPC were prepared as per the guidelines given in the lit-
erature [7]. For the trial mixes, the alkaline activator to fly ash ratio, amount of extra
water, superplasticiser content, fine aggregate to total aggregate ratio, molarity of
NaOH, mixing time and curing temperature were considered as variables. The final
ment in GPC is faster than in PCC. This may be attributed to the fast
mix proportion for M30 grade GPC was selected based on the 28th day compressive
strength and a workability giving compacting factor of 0.9. PCC mix of the same polymerization process due to heat curing. In the case of GPC the
grade was also prepared as per IS: 10262 [8] and the details are given in Table 1. splitting tensile strength increased by approximately 13%, whilst
the flexural strength increased by 12%. This is probably due to
2.2. Preparation of test specimens the geopolymer paste present in GPC which provides better bond-
ing between the fine and coarse aggregate than that of cement
Coarse aggregates and sand in the saturated surface dry condition were first
paste in PCC. The studies conducted by Frantisek et al. [10] have
mixed in laboratory mixer with fly ash for about three minutes. Then alkaline solu-
tion, super plasticizer and extra water were added to the dry materials and were
shown that the interfacial transition zone which is considered as
mixed for four minutes. The GPC resembles PCC in its appearance. Immediately after the weakest part in ordinary concrete is not found between geo-
mixing, the slump and compacting factor of fresh concrete were measured to observe polymer and aggregate and the absence of such a layer contributes
the consistency of the mixture. Cubes of 150 mm, prisms of 100 mm  100 mm  to the superior mechanical properties of GPC. The modulus of elas-
500 mm, cylinders of 150 mm diameter and 300 mm height were prepared for deter-
ticity increased by 50% for GPC compared to PCC.
mining the compressive strength, modulus of rupture and splitting tensile strength
respectively. For finding the stress–strain behaviour cast-iron moulds of 150 mm
diameter and 300 mm height were fabricated. Special provision was provided in 3.2. Stress–strain behaviour
the mould to insert the plates for fixing the LVDTs so that the core strain could be
measured accurately [9]. The test set-up is shown in Fig. 1. The plates were inserted
in such a way that the gauge points were symmetrical about the centre of the spec-
The stress–strain curves of GPC and PCC specimens with various
imen and the gauge length was 100 mm. After casting, all GPC specimens were kept percentages of spiral confinement are shown in Fig. 3. From figure,
at room temperature for one day. The GPC specimens were then placed inside the it can be seen that the stress–strain behaviour is almost similar for
oven along with moulds and cured at 60 °C for 24 h. After curing, the specimens were both GPC and PCC. However GPC mixes have shown improved
removed from the chamber and left to air-dry at room temperature for another 24 h
stress values for the same strain levels compared to that of PCC
before demoulding. The test specimens were then left in the laboratory ambient con-
ditions till the day of testing. PCC specimens were also prepared and kept immersed in the unconfined state. At the initial stage, the deformation of
in water for 28 days after one day of casting. Six GPC and three PCC specimens for GPC specimens increases at a slower rate than that of PCC.
each volumetric ratio of confinement were cast. This trend continued up to about 80% of the peak stress. The
2.3. Testing
increase in deformation was faster in GPC. This may be due to
the development of large number of micro cracks in the geopoly-
After 28 days of casting, the specimens were tested for cube compressive mer paste near the peak stress point as noted by other researchers
strength, flexural tensile strength, splitting tensile strength and modulus of elastic- [4,10]. Since the plain GPC specimens (GPCP0) had a brittle failure,
ity. The stress–strain behaviour was determined by carrying out tests on cylindrical
the descending branch of stress–strain curve could not be deter-
specimens. The 5 mm LVDTs having least count of 0.001 mm were used. The speci-
mens were placed in a compression testing machine of 3000 kN capacity and tested mined in any of these specimens. But the behaviour of confined
under uni-axial compression. The loading arrangement is shown in Fig. 2. The LVDTs specimens was comparatively ductile than unconfined specimens.
were attached to the plates on opposite sides of the specimen and parallel to the lon- The behaviour of all unconfined specimens in the ascending
gitudinal axis. The LVDT readings were taken at equal increments of 250 N load. branch up to the peak stress is similar. This is due to the fact that
in the case of confined concrete at low levels of stress the trans-
3. Results and discussions verse reinforcement is hardly stressed; hence the concrete is
unconfined. The concrete becomes confined at stresses approach-
3.1. Fresh and hardened properties ing the uniaxial strength [11]. The confinement considerably
improved the stress–strain characteristics of GPC at higher strain
The fresh and hardened properties of all the mixes are shown in levels. The stress–strain curves were analyzed to obtain the effect
Table 2. From the table it can be seen that the strength develop- of confinement on the strength and ductility of GPC.

Table 1
Mix Proportions.

Mix Cement Fly ash Sodium silicate Sodium hydroxide CA FA Water SP

(kg/m3) (kg/m3) solution (kg/m3) solution (kg/m3) (kg/m3) (kg/m3) (kg/m3) (kg/m3)
GPC – 408 103 41 1294 554 14.5 10.2
PCC 426 – – – 1266 598 192 0
328 N. Ganesan et al. / Construction and Building Materials 73 (2014) 326–331

3.2.2. Behaviour of specimens

In the case of GPC specimens with low values of volumetric
ratio of confinement both cover and core concrete were found to
have spallen (GPCP1). On the other hand as the volumetric ratio
of confinement increases the core concrete was found to be intact
even after spalling of cover concrete (GPCP2, GPCP3). Fig. 4 shows
the tested specimens.

3.3. Stress–strain model

In order to develop an analytical model for the stress–strain

behaviour of confined GPC, the test results were compared with
the available models for confined PCC. Most of the models were
able to estimate correctly the ascending part of the stress–strain
curve. But there were wide variations in the descending portion
of the stress–strain curve. The model proposed by Mander et al.
Fig. 2. Loading arrangement. [12] gives a better representation of the stress–strain behaviour
of confined GPC. It adopted the 3-parameter equation proposed
by Popovics to describe the entire stress–strain curve of confined
3.2.1. Strength enhancement concrete and is given below.
Strength enhancement, which is the ratio of the peak stress of
 
confined concrete to the peak stress of unconfined concrete is ec
f cc
ecc r
shown in Table 3. The strength increased by 22%, 60% and 110% fc ¼  r ð1Þ
when the volumetric ratio of transverse steel was 1.36%, 2.05% r  1 þ eeccc
and 4.10% respectively for both GPC and PCC. Strain ductility ratio,
which is the ratio of axial strain of confined concrete at 85% of the where, fc , fcc, ec, ecc are the stress at any point, peak stress, strain at
peak stress on the descending branch to the strain of unconfined any point and strain corresponding to peak stress respectively and r
concrete corresponding to peak stress [9], indicates the deformabi- is the curve fitting factor.
lity of confined concrete. The strain ductility ratio was determined In Mander’s model, the expression for ‘r’ is given as,
and is given in Table 3. The ratio of the strain at peak stress of con-
fined concrete to the strain at peak stress of unconfined concrete EC
r¼ ð2Þ
known as strain enhancement, which is also a measure of the duc- EC  Esec
tility was calculated and is given in Table 3. It can be observed that
where Ec is the modulus of elasticity of concrete and the parameter
the strain ductility ratio varied as 3.09, 7.14 and 11.43 and strain
Esec is given by Eq. (3).
enhancement varied as 1.47, 3.28 and 4.71 with increase in the vol-
umetric ratio of confinement indicating an increase in the ductility f cc
Esec ¼ ð3Þ
of GPC with increase in the transverse confinement. ecc

Table 2
Fresh and hardened properties of GPC and PCC mixes.

Mix Slump (mm) Compacting factor Compressive strength (N/mm2) Splitting tensile Flexural strength Modulus of
strength (N/mm2) (N/mm2) elasticity (N/mm2)
7th day 28th day
GPC 123 0.90 32.0 38.55 3.56 4.46 39,992
PCC 93 0.89 23.3 39.00 3.15 3.79 26,678

60

50

40
Stress (N/mm2)

GPCP0
30 GPCP1
GPCP2
20 GPCP3
PCCP0
PCCP1
10 PCCP2
PCCP3
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045
Strain

Fig. 3. Stress–strain curve for GPC specimens.

 N. Ganesan et al. / Construction and Building Materials 73 (2014) 326–331 329

Table 3
Effect of confinement on strength and ductility.

Specimen ID Peak stress (N/mm2) Strength enhancement Strain at peak stress Strain enhancement Strain ductility ratio
GPCP0 25.48 1.0 0.0021 1.00 1.00
PCCP0 24.06 1.0 0.0020 1.00 1.00
GPCP1 31.14 1.2 0.0031 1.47 3.09
PCCP1 29.72 1.2 0.0027 1.35 2.25
GPCP2 41.05 1.6 0.0069 3.28 7.14
PCCP2 39.63 1.6 0.0066 3.05 5.00
GPCP3 53.79 2.1 0.0099 4.71 11.43
PCCP3 51.96 2.2 0.0102 4.55 8.50

Table 4
Curve fitting factors.

Pitch of Volumetric ratio of r values

confinement (mm) confinement (%)
Mander’s Modified
model values
0 0.00 1.00 1.62
75 1.36 1.35 1.52
50 2.05 1.18 1.46
25 4.10 1.17 1.34

3.3.1. Equation for the modified parameter ‘r’

An attempt has been made to derive an equation for the param-
(a) GPCP1 (b) GPCP2 (c) GPCP3 eter ‘r’ in the case of GPC as follows. The normalized lateral confin-
Fig. 4. Tested specimens. ing pressure developed as a result of confinement gives the
confinement index [13]:

Fig. 5 shows the stress–strain curves obtained from the experi-

2f l
Confinement index ðkÞ ¼ 0 ð4Þ
ment for confined GPC and the curves obtained by using Mander’s fc
model. From figure it can be observed that Mander’s equation did
where fl is the lateral pressure and f0 c is the compressive strength of
not match satisfactorily with the test results of GPC. Mander’s
unconfined specimen.
model required a modification when applied to GPC. An attempt
From the free body diagram of spirally reinforced concrete as
has been made to modify the curve fitting factor ‘r’ present in
shown in Fig. 7, the lateral pressure
the Mander’s model by regression analysis using MAT LAB. Table 4
shows the values ‘r’ thus modified for various values of pitch of 2As f y
confinement. Figs. 5 and 6 show the comparison of experimental fl ffi ð5Þ
dS
stress–strain curves with Mander’s model and modified Mander’s
model for GPC specimens. From these figures it may be noted that where As, fy, S are the section area, yield strength and pitch of the
the modified Mander’s model matches satisfactorily with the confining reinforcement respectively and d is the diameter of the
experimental results. confined core.
The volumetric ratio of confining reinforcement

Fig. 5. Comparison of experimental values with Mander’s model.

330 N. Ganesan et al. / Construction and Building Materials 73 (2014) 326–331

Fig. 6. Comparison of experimental values with modified Mander’s model.

4. Conclusions

From the present study the following conclusions were

obtained.

 The strength development of GPC is faster than that of PCC

in the initial stages.
 Confined GPC appears to be suitable for earthquake resis-
Fig. 7. The relation between lateral pressure and confining force. tant design since the spiral confinement improves the
strength, strain at peak stress, strain ductility ratio and
energy absorption capacity of GPC.
A pd 4A
q ¼ ps 2 ¼ s ð6Þ  The strength of GPC increased by 20% to 110% when the vol-
d S dS
4 umetric ratio of confinement was increased from 1.36% to
Substituting q in Eq. (5), the following equation for lateral 4.10%.
pressure is obtained:  The strain ductility ratio and energy absorption capacity of
GPC increased by 11.4 and 5.4 times respectively compared
qf y to unconfined GPC at 4.1% volume of transverse steel.
fl ¼ ð7Þ
2  The stress–strain model proposed by Mander et al. for PCC
Therefore, confinement index, with confinement can be used for GPC with a modification
in the curve fitting factor.
qf y
k¼ 0 ð8Þ
fc
To obtain a generalized equation for the curve fitting factor (r), a
regression analysis is performed among ‘r’, ‘k’ and ‘Ec’ and equation Acknowledgements
for the curve fitting factor thus obtained is given below.
The authors would like to thank Centre for Engineering
Ec
r¼ Ec
ð9Þ Research and Development (CERD) and Kerala State Council for Sci-
1:62
þ 8888k ence Technology and Environment (KSCSTE) for providing financial
where Ec is the modulus of elasticity of Geopolymer concrete given assistance to this work.
as follows [14].
qffiffiffiffiffiffi
Ec ¼ 6965 f ck ð10Þ References

where fck is the characteristic compressive strength of GPC. [1] Rajamane NP, Nataraja MC, Lakshmanan N. An introduction to geopolymer
concrete. Indian Concr J 2011:25–8.
Thus, the stress–strain behaviour of spirally confined GPC spec- [2] Bakharev T. Durability of geopolymer materials in sodium and magnesium
imens can be represented by the following model sulfate solutions. Cem Concr Res 2005;35:1233–46.
  [3] Ganesan N, Indira PV, Santhakumar Anjana. Prediction of ultimate strength of
ec
f cc
ecc r reinforced geopolymer concrete wall panels in one-way action. Constr Build
fc ¼  r ð11Þ Mater 2013;48:91–7.
r  1 þ eeccc [4] Duxson Peter, Provis John L, Lukey Grant C, van Deventer Jannie SJ. The role of
inorganic polymer technology in the development of green concrete. Cem
Concr Res 2007;37:1590–7.
as given by Mander et al. with modified equation of ‘r’ given [5] Ahmad SH, Shah SP. Stress–strain curves of concrete confined by spiral
by Eq. (9). reinforcement. ACI J 1982;79:484–90.
 N. Ganesan et al. / Construction and Building Materials 73 (2014) 326–331 331

[6] Braga Franco, Gigliotti Rosario, Laterza Michelangelo, Amato Michele D’. An [11] Park R, Paulay T. Reinforced concrete structures. 4th ed. New York: Wiley
analytical formulation of stress block parameters of confined concrete. Open Interscience Publication; 1974.
Constr Build Technol J 2008;2:156–65. [12] Mander BJ, Priestley JNM, Park R. Theoretical stress–strain model for confined
[7] Rangan BV. Studies on low-calcium fly ash based geopolymer concrete. Indian concrete. ASCE J Struct Eng 1988;144(8):1804–26.
Concr J 2006:9–17. [13] Moghaddam H, Samadi M, Mohebbi S. On the effect of external active
[8] Indian standard code of practice for recommended guidelines for concrete mix confinement on spirally reinforced concrete columns. In: The 14th world
design. IS: 10262-2009. Delhi: Bureau of Indian Standards. conference on earthquake engineering, Beijing, China; 2008.
[9] Ganesan N, Indira PV, Ruby Abraham. Compressive constitutive behaviour of [14] Sasi Divya, Deepa Raj S, Abraham Ruby. Evaluation of Mechanical Properties
SFRHPC subjected to cyclic loading. In: The tenth East Asia-Pacific conference of Geopolymer Concrete. In: National Conference on Technological
on structural engineering and construction, Bangkok, Thailand; 2006. Trends. Thiruvananthapuram (Kerala): College of Engineering Trivandrum;
[10] Frantisek S, Lubomir K, Jiri N, Zdenek B. Microstructure of geopolymer 2013.
materials based on fly ash. Ceram-Silik 2006;50:208–15.