- Email: [email protected]
Chloride ingress and steel corrosion in geopolymer concrete based on long term tests
Accepted Manuscript Chloride ingress and steel corrosion in geopolymer concrete based on long term tests
Chandani Tennakoon, Ahmad Shayan, Jay G Sanjayan, Aimin Xu PII: DOI: Reference:
S0264-1275(16)31547-7 doi: 10.1016/j.matdes.2016.12.030 JMADE 2575
To appear in:
Materials & Design
Received date: Revised date: Accepted date:
21 August 2016 11 November 2016 11 December 2016
Please cite this article as: Chandani Tennakoon, Ahmad Shayan, Jay G Sanjayan, Aimin Xu , Chloride ingress and steel corrosion in geopolymer concrete based on long term tests. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Jmade(2016), doi: 10.1016/j.matdes.2016.12.030
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Chloride ingress and steel corrosion in geopolymer concrete based on long term tests Chandani Tennakoona,1, Ahmad Shayanb, Jay G Sanjayana, Aimin Xub a
Centre of Sustainable Infrastructure, Faculty of Science, Engineering and Technology, Swinburne University of Technology, Hawthorn, Victoria, Australia. ARRB Group Ltd, Vermont South, Victoria, Australia.
PT
b
RI
Abstract
Alkali activated binders (geopolymer) is an emerging technology to produce concrete without
SC
the use of any Ordinary Portland Cement (OPC) by alkali activation of alumino-silicate source materials such as fly ash and/or slag. Limited knowledge on durability issues like corrosion
NU
behaviour of reinforced geopolymer concrete impedes the usage of this technology in structural applications.
MA
This study explored the chloride permeability and initiation of chloride induced corrosion of geopolymer concrete in accelerated chloride environment using longer test period. Corrosion
D
initiation was also monitored in embedded rebar in 2% chloride contaminated concrete.
reference electrode.
PT E
Corrosion state of the rebar was monitored using non-destructive test method using Cu/CuSO4
The results showed that the apparent chloride diffusion coefficient of blended fly ash and slag
CE
geopolymer concrete is lower than that of OPC concrete. The diffusion coefficient also decreased with the increase of slag content in the binder. Blended fly ash and slag geopolymer
AC
concrete also exhibited higher aging factor than OPC concrete indicating improved resistance to chloride ingress with time. The study also showed that the embedded rebar in fly ash and slag based geopolymer concrete has higher protection against corrosion than a rebar in OPC concrete even when the concrete is contaminated with significant levels of chloride.
1
Corresponding author; email: [email protected] 1
ACCEPTED MANUSCRIPT
Highlights Chloride ingress rate in geopolymers containing slag is low
Age factor of geopolymer is significantly higher than OPC
Geopolymer can hinder corrosion in steel rebar
SC
RI
PT
AC
CE
PT E
D
MA
NU
Keyword: Age factor, Chloride ingress, Corrosion, Fly ash, Geopolymers, Slag
2
ACCEPTED MANUSCRIPT
1. Introduction The term geopolymer describes a broad class of synthetic zeolite materials that are produced by a series of reactions between alumina-silicate source materials and alkali activators. Due to the reduced carbon foot print and low embodied energy, geopolymer concrete technology has gained
PT
significant attention in recent years as an alternative to Ordinary Portland cement (OPC) concrete for production of structural and non-structural elements. To date, geopolymer concrete has had
RI
relatively minor application in field concrete due to the limited knowledge base on its durability [1]. Its long-term durability properties as well as its protection against reinforcements corrosion
SC
are yet to be fully investigated [2]. Some studies of marine-exposed concrete specimens have shown that the performance of geopolymer concrete specimens is comparable to that of OPC
NU
concrete with respects to reinforced corrosion [3-5].
The main cause of corrosion in concrete structures, particularly in concrete structures in marine
MA
environment is the chloride ingress [6, 7]. Ganesan et al. [8] studied the chloride diffusion of 100% fly ash based geopolymer and reported that chloride diffusion of geopolymer and OPC
D
concrete are similar. In their study authors employed AgNO3 spraying to identify the depth of chloride penetration, as suggested by ASTM C1556 [9]. In contrast, the study by Olivia et al.
PT E
[10] reported that the rate of chloride penetration in 100% fly ash based geopolymer concrete is higher than that of OPC concrete. Shayan et al. [11] characterised the properties of a 100% slag-based geopolymer concrete after having been in service for five years in the retaining walls
CE
of a bridge structure on the Yarra river in Melbourne, Australia. The rapid chloride permeability test ASTM 1202 [12] showed very low chloride permeability. Based on NT Build 443 [13]
AC
chloride diffusion coefficient was an order of magnitude lower than that of blended slag cement concrete used in the retaining walls of the same bridge. More recently, Ma et al. [14] followed NT build 443 test method [13] to evaluate the chloride diffusion coefficient of 100% slag-based geopolymer and found that chloride diffusion rate of slag based geopolymer concrete is lower than OPC concrete. However, in the same study, Ma et al. [14] found that the corrosion rate of rebar inside prism specimens is similar to that of the embedded rebar inside OPC concrete. The above studies were mainly focused on corrosion due to chloride ingress. There are also a few studies on chloride induced corrosion in contaminated concrete. Monticelli et al. [15] added
3
ACCEPTED MANUSCRIPT
2% chloride by mass of binder to 100% class F fly ash based geopolymer during the mixing process. The authors reported that the availability of silicate ions in the pore solution of contaminated geopolymer concrete can provide additional protection for the passivity of rebar. Miranda et al. [16] observed that addition of 2% chloride by mass of binder to fly ash based geopolymer mix increase the corrosion rate by 100 times compared to uncontaminated geopolymer concrete. They also found that the steel embedded in contaminated geopolymer did
PT
not have the protective passive layer and show similar corrosion behaviour to OPC. The result of Bastidas et al. [17] is in agreement with the previous observations. More recently, an innovative
RI
technique for repair and strengthening of reinforced concrete members with deteriorated cover
SC
concrete and/or corroded steel rebars was proposed and named inhibiting-repairing-strengthening (IRS) technique [18, 19]. It consists in the installation of a stainless steel fabric in the cover concrete, restoring this latter with a geopolymer-based matrix which also act as a corrosion
NU
protection.
MA
Blending of class F fly ash and slag enables geopolymers modifies the reaction phases and the microstructure of geopolymer compared to 100% fly ash or 100% slag based geopolymer leading to different level of chloride binding and electro chemical mechanisms. Slag leaches sulphate
D
ions into pore solution leading to high negative half-cell redox potential in the embedded rebar
PT E
[20]. Corrosion mechanism of embedded rebar in slag blended geopolymer concrete has not been discussed in the litreture. In the present study chloride ingress of slag blended geopolymer concrete is evaluated using potentiometric titration method [14, 21]. This study also examines
CE
the change in chloride diffusion rate with time in blended fly ash and slag geopolymer which can be used to quantify the age factor of concrete. The corrosion state of embedded rebar using electrochemical method (half-cell potential and corrosion current) was evaluated. These results
AC
are compared with the actual corrosion state of the rebar. Finally, the study evaluates the corrosion state of the embedded rebar in blended fly ash and slag geopolymer concrete contaminated with 2% (by mass of binder) The purpose is to accelerate the corrosion so that the investigation can be completed in realistic time period.
4
ACCEPTED MANUSCRIPT
2.
Materials and methods
2.1.Materials Precursor materials A class F fly ash (FA) which was a 50/50 mixture of Gladstone and Callide fly ashes (FA) and granulated ground blast furnace slag (Slag), with the chemical compositions shown in Table 1
PT
were used as the alumino-silicate precursor materials. The bulk chemical compositions were determined by X-ray florescence (XRF) analysis. The main component of class F fly ash is SiO2
RI
(over 51%) whereas slag contains high amount of CaO (41.5%). It is important to note that fly ash contains twice as much Al2O3 as slag does. The particle size distributions of fly ash and slag,
SC
obtained using a Cilas laser diffraction particle analyser are shown in Figure 1. The particle size analysis shows that both fly ash and slag have small particle sizes which are desirable for the
MA
while slag has about 2.5% of finer than 1µm.
NU
geopolymer reaction. It also shows that class F fly ash has 10% of very fine particles (< 1µm),
Table 1: Bulk chemical composition of materials (wt. %)
AC
CE
GP Cement 19.30 0.30 5.50 2.90 0.20 1.40 62.00 0.06 0.40 0.11 2.46 0.00 5.3
D
PT E
Oxides SiO2 TiO2 Al2O3 Fe2O3 Mn3O4 MgO CaO Na2O K2O P2O5 SO3 S2LOI
Slag 32.00 0.51 13.00 0.40 0.30 4.90 41.50 0.20 0.33 0.02 2.10 0.60 4.9
FA 51.50 1.50 27.60 11.80 0.15 1.30 2.20 0.40 0.60 0.73 0.2 0.00 1.8
5
PT
ACCEPTED MANUSCRIPT
(b)
RI
(a)
SC
Figure 1: Particle size distribution of (a) Slag (b) Fly ash. 1-10 g of source material was dispersed in water in order to obtain the particle size.
NU
Alkali activator
Granular sodium metasilicate penta-hydrate (28.7 wt% Na2O, 28.5 wt% SiO2 and 42.8 wt% H2O) was used as solid activator to activate the precursors. Usage of solid alkali activators in
MA
geopolymer systems is less hazardous than commonly used liquid activators and more convenient for field applications. The amount of activator used was calculated to provide an
D
alkali content of 4% Na2O by mass of precursor materials.
PT E
Aggregates
A coarse basalt aggregate of 14 mm nominal particle size and river sand were used in the geopolymer concrete mix. The particle density of coarse aggregates and the sand in saturated
CE
surface dry condition (SSD) was 2830 and 2620 kg/m3, respectively. The moisture absorption of
AC
coarse aggregates and the sand were 2.2 and 0.3% respectively. 2.2.Preparation of geopolymer concrete The mix designs for the geopolymer concretes employed in this work are shown in Table 2. The concrete mixing procedures consistent of: Sand, coarse aggregate and precursor materials were mixed thoroughly for 1 minute. Water was added to the mix and mixing was continued for another 2 minutes. Alkali activator (solid form) was added into wet mix and mixed for around 7 minutes to obtain workable geopolymer concrete. The workability of geopolymer concrete mix was measured using the standard slump test according to AS1012.3.1 [22]. The target slump was 80-100 mm. 6
ACCEPTED MANUSCRIPT
Two identical concrete prisms (95 × 95 × 300 mm) were cast from each, mix embedding a 20 mm diameter carbon steel rebar. A sample prism is shown in Figure 2. The concrete mixes were also used to cast two cylindrical specimens (100×200 mm) for NT build 443 testings. The cast specimens were transferred into a standard fog room (23°C, 100% RH). After 24 hours, concrete specimens were de-moulded and stored inside fog room until 28 days. During the storing period in the fog room, the samples were covered with plastic sheets to avoid the water dropping on the
PT
sample.
RI
To identify the corrosion behaviour in contaminated concrete, 50/50 FA/Slag geopolymer concrete specimens were cast with added 2% NaCl by mass of binder and partially immersed in
SC
water. The temperature during the immersion test was maintained at 23°C. An OPC concrete
NU
prism was also cast with added 2% NaCl by mass of binder as a control. Table 2: Mix design for concrete (Alkali content = 4% for geopolymers) Slag (kg/m3)
60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
240 200 180 0
180 200 240 0
OPC (kg/m3) 0 0 0 400
Aggregate (kg/m3)
Sand 14mm 680 680 680 680
630 630 630 630
10mm 450 450 450 450
Water/ Solid ratio 0.390 0.425 0.449 0.464
AC
CE
PT E
D
MA
Mix
FA (kg/m3)
(a) (b) Figure 2: (a) Specimens used for chloride induced corrosion testing (b) Gecor8 instrument
7
ACCEPTED MANUSCRIPT
2.3.Testing procedure Chloride diffusion tests After 28 days of moist curing at 23°C each concrete cylinder was segmented to three parts with heights 51, 51 and 98 mm respectively. The curved surfaces of the cylindrical segments were then coated with quick setting epoxy which prevented chloride transportation into the concrete
PT
from the sides. Chloride diffusion coefficients of geopolymer and OPC concretes were measured using NT
RI
Build 443 [13] method. The 98 mm cylindrical section was immersed in 2.826 M sodium
SC
chloride solution vertically such that approximately 1/4 of the height was submerged inside the solution. Cylindrical specimens were mounted on PVC rings in order to maintain uniform
NU
chloride exposure through the bottom surface. The chloride solution was stirred frequently in order to have a uniform solution. After 5 weeks of immersion, the concrete specimens were drilled in 1 mm depth increments using a profile grinder to obtain fine concrete powder from
MA
each 1 mm layer (75 mm diameter circular area is drilled). The obtained concrete powder was dried at 105°C for 24 hours (powder was covered properly avoiding direct heat exposure). Approximately, 5 g mass of the dried concrete powder was mixed with distilled water (1:5 solid
D
to water ratio) and HNO3. The resulting solution was boiled and filtered to extract the chloride
PT E
ions from concrete powder. Potentiometric titration was conducted on the filtered solution using standard 1N AgNO3 solution to obtain the chloride content (acid soluble chloride) of each layer.
CE
The rapid chloride permeability of concretes was measured using ASTM 1202 standard [12]. In this test each ends of a 51 mm (2 inch) thick concrete disk is embedded between two cells each
AC
fitted with a stainless steel mesh. One cell is filled with a 0.6 M NaCl solution and connected to the negative terminal and the other filled with 0.3 M NaOH solution. A DC potential of 60 Volts is applied across the concrete specimen for 6 hours during which the current is recorded every 5 minutes, and the total charge after 6 hours is reported by the testing device. Chloride permeability of geopolymer concrete was measured at three ages, namely 28 days, 180 days and 500 days. Measurements of corrosion state of embedded rebar using electrochemical measurements After 28 days of ambient curing at 23°C, the concrete prisms were partially immersed in 2.826 and 0.6 M NaCl solutions up to 500 days. Temperature of the storing room was 23°C. The 8
ACCEPTED MANUSCRIPT
corrosion rate meter was used to measure the corrosion current. It measures the corrosion current over a confined length of steel bar, the half-cell potential (copper-copper sulphate) of the steel and the electrical resistance of concrete. This instrument uses the “linear polarization” technique and calculates the polarization resistance and obtains the corrosion current according to Stern and Geary relationship [23].
PT
Gecor6 instrument was used to measure the half-cell potential of embedded rebars with reference copper-copper sulphate electrode. As prescribed in ASTM C876 standard for Cu/CuSO4
RI
reference electrode [24], the corrosion state of rebar can be assessed as follows: the half-cell potential < -200 mV (less negative) is an indication of 90% chance of having no corrosion while
SC
if the potential<-350 mV (more negative) it is an indication of 90% probability of having corrosion (half-cell potential limits are marked in the figure). Previously it was also reported
NU
that, if the corrosion rate is higher than 1 µA/cm2 it is an indication of a higher corrosion activity (high risk of corrosion). Instead, if the corrosion rate is in between 0.1 and 0.5 µA/cm2 it is a
MA
sign for low to moderate corrosion activity [25].
After 500 days, immersed concrete prisms were split to observe the corrosion state of embedded
PT E
potentiometric titration [13].
D
rebar. Chloride penetration depth and chloride content at the rebar were calculated using
3. Results and Discussion
3.1.Chloride diffusion rate of concrete in unsaturated condition
CE
Corrosion due to chloride ingress is initiated when sufficient amount of chloride ions penetrate to the depth of reinforcement steel [7]. The resistance against chloride transportation into concrete
AC
is quantified using the chloride diffusion coefficient [26]. To determine the rate of chloride diffusion into concrete NT build 443 test as described in section 2.3 was conducted. The apparent diffusion coefficient of a concrete (Da mm2/year) can be calculated using the Fick’s second Law of Diffusion. According to this law the relationship between the calculated chloride content of concrete (C(t,x) ), and the depth from the exposed surface (𝑥) is as follows: 𝐶(𝑡,𝑥) − 𝐶0 = (𝐶𝑆 − 𝐶0 )erfc(
𝑥 t √∫t 𝐷(𝑡) 𝑑𝑡 s
)]
(1)
9
ACCEPTED MANUSCRIPT
Where CS is the chloride concentration at the surface (x=0), C0 is the chloride concentration in concrete before immersion and t is time in years and t s age of concrete when exposure started. The chloride profile of concrete specimens obtained using the above procedure was then used to calculate the chloride coefficient of each concrete type. Figure 3 shows the chloride profiles obtained for the geopolymer and OPC concretes while Table
PT
3 shows the calculated chloride diffusion coefficients. The chloride profile and the chloride diffusion coefficient results show that the chloride diffusion rate decreases significantly when the
RI
slag content of the geopolymer binder is increased.
SC
Geopolymer concrete containing significant level of slag has the following main reaction products: calcium silicate hydrate (C-S-H), calcium aluminate silicate hydrate (C-A-S-H),
NU
sodium aluminate silicate hydrate (N-A-S-H) and hydrotalcite phases [27-29]. When the slag content of the geopolymer blends increases, the resulting binder consists more C-S-H, C-A-S-H than N-A-S-H and higher amount of hydrotalcite phases [27, 30]. It is known that C-A-S-H
to
N-A-S-H
which
can
limit
MA
products have higher a binding capacity for chloride ions (and accompanied Na ions) compared the
ingress
of
those
ions
[27].
Hydrotalcite
(Mg6Al2(OH)16CO34H2O) has a layered double hydroxide structure. This layered crystal
D
structure consists of an outer layer with hydroxide and an inner layer with anions and water
PT E
molecules [30-32]. Therefore, hydrotalcite has large specific area which increase its capability to exchange ions with the concrete pore solutions and absorb anions to its structure [32]. Previous studies have shown that the addition of hydrotalcite into a chloride contaminated system
CE
externally result in the reduction of chloride content in the concrete without changing the
AC
structure of hydrotalcite [30, 33]. The X-ray diffraction patterns on slag blended geopolymer show the indication of hydrotalcite type phases with significant magnesium level in fly ash and slag based geopolymer which is also confirmed in literature (Figure 13) [30]. Geopolymer containing lower slag content mainly produces N-A-S-H type phases which have high pore volume [27, 28]. Provis et al.[34] and Lloyd et al. [35] showed that increasing fly ash content, the porosity of the geopolymer is increased significantly. They stated that geopolymer containing high slag content have high bound water (induced by calcium) which leads to more pore-filling capacity than geopolymer containing high fly ash [34]. This is in agreement with the studies by Yang et al. [36] and Li et
10
ACCEPTED MANUSCRIPT
al. [37]. Since fly ash based geopolymer has lower density than slag blended geopolymer, it can absorbs more external chloride ions and deposit in the binder pore solution. Previously it has been reported that geopolymers containing higher fly ash content (>50%) has halite (NaCl) in their pore solution than that of high slag based geopolymers [27]. The chloride diffusion coefficients results in Table 3 also show that slag blended geopolymer has
PT
higher resistance to chloride ingress compared to OPC based concrete. Geopolymer concrete containing significant level of calcium produces cross-linked tobermorite-like C-A-S-H which
RI
are high density structures compared to non-cross linked tobermorite-like C-S-H phases formed in OPC concrete [38]. These cross linked structures reduce the chloride ingress in slag blended
SC
geopolymer concrete compared to that of OPC concrete. OPC binder also contains calcium aluminates and calcium aluminoferrite phases that can absorb chloride ions forming
NU
3CaO·Al2O3·CaCl2·10H2O (Friedel’s salt). However, the study by Florea and Brouwers [39] stated that Friedel’s salt formation in OPC binder is significant only in chloride contaminated
MA
concrete during the initial hydration process. The Friedel’s salt formation after cement hydration is very slow hence this process would not be effective in absorbing chloride coming from the external sources in the long term [40, 41]. This explains the higher chloride diffusion in OPC
AC
CE
PT E
D
concrete.
Figure 3: Chloride profiles of concrete cylinders after 5 weeks of immersion test
Table 3: Chloride diffusion coefficient and surface chloride content of concrete Geopolymer
Surface Cl-. %
Apparent chloride 11
ACCEPTED MANUSCRIPT
formulation
(Cs ) (%wt. by concrete mass) 0.80 0.89 0.81 0.80
60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
diffusion coefficient (Da ) 10-12 m2/s 5.17 2.38 1.01 6.70
Table 4 shows the chloride content (Cl-% by mass of concrete) at the different depth of
PT
geopolymer and OPC concrete after 500 days immersion in 2.83 M NaCl solution. The results show that the chloride content of OPC concrete is higher than geopolymer concrete at all depths.
RI
At the rebar (40 mm depth), OPC concrete has 0.070% chloride (by mass of concrete) which is above the threshold chloride value that initiates corrosion [42]. On the other hand, geopolymer
SC
concrete shows 10 times lower chloride content at the rebar after 500 days which is not sufficient
NU
to initiate chloride induced corrosion.
The resistance to chloride ingress rate into concrete is reduced upon aging of the geopolymer concrete. It is well known that chloride diffusion rate of OPC concrete decreases with the age of
MA
concrete as a result of chloride binding and refinement in the pore structure [43]. It was reported that a 1 day old OPC binder has pore size range 8 nm to 700 nm while more than 365 days old
D
OPC binder has pores in the range of 16 nm to 64 nm [43] which significantly affects the chloride diffusion coefficient of concrete. Ignoring this effect can result in overestimation of the
PT E
chloride penetration rate [44]. Resistance to chloride ingress into concrete can vary due to modified microstructure and additional reaction products formed during extended curing. Geopolymerisation is a continual reaction process that can have different microstructure and new
CE
reaction phases in the later stages compared to premature age of concrete. Puertas et al. [29], showed that when blended fly ash and slag geopolymer matures, FTIR (T-O-T T: tetrahedral Si
AC
or Al) band shifted towards a higher value indicating higher degree of cross-linking in geopolymer than in early age concrete. This is attributed to the continual development of N-A-SH and C-A-S-H cross-linking and increase of number of reaction products. Their results are in agreement with several other studies [45, 46].
12
ACCEPTED MANUSCRIPT
Provis et al. [34] reported that longer curing of blended fly ash and slag geopolymer increases the pore-network tortuosity and decreases the porosity due to space filling nature of C-(N)-S-H phases. Xu et al. [45] reported that in the presence of slag, geopolymerisation is continual as slag provides hydroxyl buffering to react with the precursors in the later age. Moreover, the authors found that channels in the pore structure are disconnected due to maturity of geopolymer concrete. These studies show that resistance to chloride ingress into concrete is expected to be
RI
concrete will be quantified and analysed the factors affecting it.
PT
lower in the later age of geopolymer concrete. In the next section, the age factor of geopolymer
Average depth of the concrete (mm)
NU
Concrete type
SC
Table 4: Chloride percentage in the different depth of concrete after 500 days of immersion in 2.83 M NaCl solution
MA
50/50 FA/Slag
1 15 35 40 1 15 35 40
PT E
D
OPC
Measured chloride (%wt. by concrete mass) 0.880 0.300 0.020 0.007 0.989 0.364 0.299 0.077
3.2.1.
CE
3.2.Properties of geopolymer concrete Compressive strength
AC
Figure 4 (a) and (b) show the compressive strength development of 60/40 FA/Slag, 50/50 FA/Slag and 40/60 FA/Slag geopolymers. 40/60 and 50/50 FA/Slag based geopolymers showed the highest strength during first year. 60/40 FA/Slag based geopolymer had the maximum strength at 100 days and thereafter compressive strength is slightly reduced. Ismail et al. [47] found that the main reaction phase in geopolymer with slag percentage in the range of 50-100%, is calcium silicate hydrated gel (C-S-H) and calcium aluminate silicate hydrate (C-A-S-H). Moreover, in these binders, the number of reaction product formations and degree of crosslinking increased with extended curing time. The geopolymer with lower than 50% slag has a combination of C-A-S-H and N-A-S-H reaction products, and the reaction product formation and 13
ACCEPTED MANUSCRIPT
development of additional cross-linking of the geopolymer is slower compared to geopolymer containing higher slag (>50%). This reveals that geopolymer concrete synthesised using fly ash and slag under ambient curing has stable compressive strength within the test period of two years. Higher compressive strength indicates the denser microstructure that has lower gas
AC
CE
PT E
D
MA
NU
SC
RI
PT
permeability. Therefore, oxygen cannot be penetrated into concrete and the level of rebar.
Figure 4: Compressive strength development of fly ash and slag blended geopolymers
14
ACCEPTED MANUSCRIPT
3.3.Quantifying the age factor for geopolymer concrete The influence of the age on the chloride diffusion coefficient is an important parameter for service-life modelling of any concrete structure [48]. The chloride diffusion of concrete D(t) varies with the time as shown in the equation (2) [48]: ts D(t) = Dref [ ]m t
PT
(2)
Where, 𝒎(> 𝟎) is the age factor, 𝐃𝐫𝐞𝐟 is the chloride diffusion rate obtained using NTBuild 443
RI
test (Table 3) and𝐭 is exposure time in years and 𝐭 𝐬 age of concrete when exposure started. Rapid permeability test results at several ages of concrete can be used to obtain the value of
SC
𝐦(> 𝟎) for concrete. The rapid chloride permeability test (RCPT) ASTM C1202 [12] provides information on the instantaneous chloride diffusion of concrete as it is conducted within a short
NU
period of time (6 hours), whereas NTBuild 443 [12] test involves 35 days immersion period. Previous studies have shown that the changes in RCPT results upon aging would reflect the
MA
changes in the chloride diffusion change [49, 50]. In order to calculate the ′𝐦′factor of concrete, rapid chloride permeability of concrete was conducted at three ages of concrete (i.e. 28, 180 and 500 days) and total charges passed through the concrete is calculated according to the procedure
D
described in section 2.3 [26, 51]. The charges passed through concrete obtained from ASTM
PT E
C1202 [12] test method is used to calculate the 𝐦factor as shown in the equation (8) [52, 53]. G(t) = G0 t −m
(3)
CE
Where G(t)= charge passed through concrete at age t of concrete, GO = charge passed through concrete at aget o .
AC
Figure 5 shows the charges passed through concrete at different ages of concrete. The results show that the charges passed through 28 days old geopolymer concrete is in the range of 10002000 Coulumb which is classified as low permeable concrete [12]. Even though, 28 days old OPC concrete can be also classified as low permeable concrete, the number of charged passed through OPC concrete is slightly higher than the corresponding value for geopolymer concrete. At 180 days, the charges passed through geopolymer concrete are significantly lower than in 28 days. This changes the permeability classification of geopolymer concrete to “very low” permeable concrete range [12]. The charges passed through geopolymer concrete after one year is only slightly lower than that at 180 days. All the geopolymer concretes reached to the same 15
ACCEPTED MANUSCRIPT
level of permeability level after one year, with charges passed through concrete being 300-330 Columb. On the other hand, the 180 days old OPC concrete had the same permeability category (low) as the 28 days old OPC concrete. When the OPC concrete was older than one year the change in the permeability level was not significant. The calculated age factors are presented in Table 5. The results show that the age factor of
PT
geopolymer concrete is in the range of 0.40-0.60. On the other hand age factor of OPC concrete is only 0.192 which is half of the age factor of geopolymer concrete. The value obtained for
RI
100% OPC concrete is similar to the values reported in previous studies [53, 54]. This shows aging improves the resistance of geopolymer concrete for chloride diffusion more significantly
SC
than for OPC concrete. As mentioned previously, when fly ash and slag geopolymer concrete matures, a higher degree of cross-linked structure is achieved than in early age, more reaction
NU
products (like hydrotalcite) are formed from unreacted fly ash and slag particles and pore structure of the binder is refined [27, 30, 34]. The combination of these parameters reduced the
MA
chloride diffusion of the concrete upon aging at a higher rate than for OPC concrete. Even though the pore structure is refined upon aging of OPC binder, not as much additional products are formed for binding chloride [40, 41, 43] in turn a lower age factor is obtained.
D
Previous studies [55, 56] have reported that, when used on geopolymer binders, the outcome of
PT E
the RCPT test is greatly influenced by the ion mobility of the pore solution rather than the chloride permeability. However, current study found that RCPT results correlates well with the
AC
CE
chloride induced corrosion.
16
SC
RI
PT
ACCEPTED MANUSCRIPT
Table 5: Age factor of concrete
Age factor of concrete (m) 0.407 0.604 0.509 0.192
D
MA
Mix proportions 60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
NU
Figure 5: The change of charged passed through concrete for concrete specimens of different ages
PT E
3.4.Service life prediction of concrete
Evaluation of the time to initiate corrosion on the embedded rebar in concrete is essential in modelling the service life of concrete structures. The results in Table 4 and the state of rebar
CE
(section 3.3) show that in OPC corrosion has initiated at a chloride level of 0.077% by mass of concrete. This chloride level has also been identified as the threshold level of the chloride to
AC
initiate corrosion process in rebar embedded in OPC concrete [42]. Assuming the same threshold level for geopolymer concrete, the time taken to initiate the corrosion can be calculated using the equations Fick’s Law [57]. The following assumptions are used in the calculation of the corrosion initiation time. 1. The threshold chloride level for the initiation of the steel corrosion is 0.077% by weight of concrete. 2. The diffusion rate only changes during the initial 10 years. There after it is assumed to be constant [53]. 17
ACCEPTED MANUSCRIPT
The results presented in Table 6 shows the time taken to initiate chloride induced corrosion when the cover thickness of concrete is 25 and 40 mm and the environment chloride concentration is 2.826 M. According to the calculations, OPC concrete initiates corrosion in 0.69 years which is consistent with the observations. OPC concrete takes 1.95 years to initiate corrosion when the cover thickness is 40 mm.
PT
On the other hand geopolymer concrete has very high corrosion initiation time, particularly the geopolymer concrete containing higher slag (>50%). The time taken to initiate chloride induced
RI
corrosion is increased with the slag content in the geopolymer concrete. This results show that geopolymer concrete structures would have higher service life in chloride environment compared
SC
to OPC concrete structures. It should be noted that a typical marine environment has a chloride concentration of around 0.6% which is approximately a quarter of the concentration used for the
NU
calculations (2.8%). Therefore, the initiation time will be much longer. The corrosion state of the rebar is consistent with the chloride calculations.
MA
Table 6: Number of years to reach chloride level of 0.077% in 2.826 M NaCl solutions. (Chloride (%) = mass of chloride/ concrete mass)
60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
CE
Number of years to reached 0.077% chloride in different concrete depths 25 mm 40 mm 1.28 5.26 7.87 37.28 13.19 41.39 0.69 1.95
D
PT E
Mix proportions
Results presented in the previous section showed that chloride diffusion in blended fly ash and
AC
slag geopolymer concrete is lower than in OPC concrete. Therefore, the initiation time for corrosion is expected to be longer in geopolymer concrete than in OPC. To determine the corrosion state of the rebar, the specimens which had been immersed in the 2.826 M NaCl solution for 500 days were split, and immediately 0.1M AgNO3 solution was sprayed to measure the chloride penetrated depth, based on the depth of white colour observed due to formation of AgCl. In general, when the chloride content of concrete is in the range 0.02-0.23, the colour change is observed [58, 59]. Figure 6 shows the concrete samples after spraying AgNO3 solution and Table 7 shows the measured chloride penetrated depth of concrete obtained from AgNO3. The depth obtained using AgNO3 colorimetric method is based on the apparent chloride content 18
ACCEPTED MANUSCRIPT
of the concrete. Geopolymer concrete immersed in 2.826 M NaCl solution has very low chloride penetrate depth (lower than 5 mm). Moreover, when the geopolymer concrete prisms were immersed in 0.600 M NaCl solution, apparent chloride penetration depth is further reduced to 2 mm. On the other hand, OPC concrete immersed in 2.826 M and 0.600 M NaCl solutions showed 35.0 mm and 18.0 mm chloride penetration depth respectively which is significantly higher than that of geopolymer concrete samples. Since all the geopolymer concrete types had
PT
similar corrosion and chloride penetrated depth, only the 50/50 FA/Slag geopolymer concrete
MA
NU
SC
RI
was samples are shown in the Figure 6.
(b)
CE
PT E
D
(a)
AC
(c) (d) Figure 6: Chloride penetration depth of concrete (a) 50/50 FA/S geopolymer immersed in 2.832 M NaCl solution (b) 50/50 FA/S geopolymer immersed in 0.6 M NaCl solution (c) OPC concrete immersed in 2.826 M NaCl solution (d) OPC concrete immersed in 0.6 M NaCl solution Table 7: Depth of chloride obtained from AgNO3 colorimetric method
Sample ID 60/40 FA/S 50/50 FA/S 40/60 FA/S OPC
Depth in 2.826 M NaCl solution (mm) 4.2 4.0 4.1 35.0
Depth in 0.6 M NaCl solution (mm) 1.2 2.5 1.8 18.0
3.5.Corrosion state of rebar 19
ACCEPTED MANUSCRIPT
Figure 8(a) and Figure 8(b) respectively show half-cell potential and corrosion current of embedded rebar. The electrochemical indicate that embedded rebar in geopolymer concrete has 90% probability of having corrosion which is not consistent with the actual corrosion state of the rebar. 50/50 FA/S geopolymer has half-cell potential around -700 mV consistently in entire measuring period. Furthermore, rebar in 60/40 FA/S shows the half-cell potential around -600 Mv. Rebar in OPC concrete shows the half-cell potential -200 mV which is an indication of
PT
uncertainty in corrosion. 40/60 FA/S geopolymer also shows high negative value that indicates corrosion activity. In the previous section, it has been observed that chloride was not reached to
RI
the level of the rebar in geopolymer concrete while chloride was reached to rebar level in OPC
SC
concrete when they are in 2.83 M NaCl solution.
Half-cell potential and corrosion current of embedded rebar in geopolymer concrete when they
NU
are immerged in 0.6 M NaCl solution shows scattered behaviour (Figure 8). Rebar in OPC concrete shows lower negative half-cell potential than the samples in 2.83 M NaCl solution.
MA
Despite the fact the chloride is not reached to the level of rebar for initiating corrosion, the corresponding half-cell potential for embedded rebars in geopolymer concrete pushed towards high negative values (>-300 mV). Slag blended geopolymer release sulphide ions into pore
D
solution of concrete leading significant reduction in redox potential. Slag-rich blends can reduce
PT E
100 mV of rebar in OPC concrete to minus 200-225 mV [Glaser, 1988 #58]. The similar observations were made by Yeau and Kim [60] for the half-cell potential in slag blended OPC concrete. They stated that sulphur species in slag-rich binders (S2-, SO3- and S2O32-) resulted
CE
higher negative half-cell potential in embedded rebar. The half-cell potential of slag blended geopolymer concrete is also high negative. The actual corrosion state of the embedded rebar in slag blended geopolymer concrete is not consistent with the half-cell potential. Half-cell
AC
potential of rebar embedded in OPC when immerge in 2.83 M NaCl solution, indicates uncertainty in corrosion. It has been observed that the embedded rebar is initiated to corrode. Corrosion state can be seen in the Figure 9 and Figure 10.
20
RI
PT
ACCEPTED MANUSCRIPT
(b)
PT E
D
MA
NU
SC
(a)
AC
CE
Figure 7: Corrosion state of embedded rebar when concrete is immersed in 2.826 M NaCl solution (a) half-cell potential (b) corrosion rate
(a)
21
PT
ACCEPTED MANUSCRIPT
(b)
SC
RI
Figure 8:Corrosion state of embedded rebar when concrete is immersed in 0.6 M NaCl solution (a) half-cell potential (b) corrosion rate Figure 9 shows the corrosion state of rebar inside 50/50 FA/Slag geopolymer and OPC concrete
NU
after immersion in 2.826 M NaCl solution. Rebar inside geopolymer concrete retained its passivity after 500 days (Figure 9(b)), while chloride induced corrosion was initiated in rebar embedded in OPC and rust products are evident in Figure 9(d). No rust products formed in either
MA
geopolymer concrete or OPC concrete for immersion in 0.6 M NaCl solution for 500 days (Figure 10). This shows that the corrosion state based on the electrochemical measurements and
D
the corrosion activity observed in the rebar do not have direct relationship. Therefore, the electrochemical measurements can over-estimate the corrosion of embedded rebar in geopolymer
AC
CE
PT E
concrete.
(a)
(b)
22
PT
ACCEPTED MANUSCRIPT
MA
NU
SC
RI
(c) (d) Figure 9: (a) geopolymer concrete with rebar (b) embedded rebar in geopolymer concrete (c) OPC concrete with rebar (d) embedded rebar in OPC concrete immersed in 2.826 M NaCl
(b)
CE
PT E
D
(a)
AC
(c) (d) Figure 10: (a) geopolymer concrete with rebar (b) embedded rebar in geopolymer concrete (c) OPC concrete with rebar (d) embedded rebar in OPC concrete immersed in 0.60 M NaCl 3.6.Corrosion rate of rebar embedded in chloride contaminated geopolymer concrete As observed in the previous section, the chloride ions take a long time to reach the embedded rebar in geopolymer concrete. Therefore, to accelerate the chloride induced corrosion the authors prepared specimens with added chloride. The half-cell potential and corrosion current of embedded rebar was measured using Cu/CuSO4 (CSE) reference electrodes [61].
23
ACCEPTED MANUSCRIPT
Again half-cell potential of embedded rebar in geopolymer concrete is higher negative than that of embedded rebar in OPC concrete (Figure 11(a) and (b). This is an indication of severe corrosion activity in embedded rebar in geopolymer concrete. The corresponding corrosion current for the rebar inside 50/50 FA/Slag geopolymer concrete is lower than the current obtained for rebar in OPC concrete [25]. According to electrochemical measurements, embedded
PT
rebars in both types of concrete expected to have corrosion. Figure 12 shows the corrosion products of the rebar embedded in contaminated concrete. Figure
RI
12(a) and (c) show the corrosion products in rebar embedded in OPC concrete. The rebar embedded in OPC shows significant corrosion activity after 150 days while the rebar embedded
SC
in geopolymer concrete does not show any pitting corrosion products. As observed in previous section, electrochemical measurements obtained for the rebar in contaminated geopolymer
NU
concrete misinterpret the corrosion state.
Slag rich concrete has high content of sulphide species (S2- and S2O32-) in the concrete pore
MA
solution [14, 20, 62]. These sulphide ions can oxidize into sulphur consuming locally available oxygen [14, 62]. The absence of oxygen in the rebar inhibits the cathode reaction in turn the corrosion is controlled. Ma et al. [14] reported higher sulphide ion concentration in the slag
D
based geopolymer concrete pore solution than OPC concrete which leads to reduction in the
PT E
embedded rebar. Hollway et al.[62] concluded that oxidation of sulphide ions inot sulphur
AC
CE
prevents potential rising into the pitting regimes which inhibits the oxygen cathode reaction.
(a) (b) Figure 11: (a) Half-cell potential (b) Corrosion current of rebar in chloride contaminated geopolymer and OPC concrete
24
PT
ACCEPTED MANUSCRIPT
(b)
MA
NU
SC
RI
(a)
(c)
(d)
PT E
D
Figure 12: Corrosion state of rebar embedded in chloride contaminated concrete (a) rebar in OPC closer image (b) rebar in 50/50 FA/Slag geopolymer closer image (c) rebar in OPC (d) rebar in 50/50 FA/Slag geopolymer after 150 days 3.7 XRD patterns of geopolymer
CE
The XRD patterns of the blended fly ash and slag geopolymer binders at two ages i.e. 28 and 270 days are shown in Figure 13. The results show that geopolymer binders containing higher than 50% slag have C-S-H phases indicated by the peak at 29°2θ. C-S-H has a disordered structure
AC
similar to riversideite and do not have 3-dimensional cross linking. When the Fly ash percentage is increased beyond 50% the resulting binders show a peak corresponding to N-(C)-A-S-H type garronite phases and N-A-S-H type zeolite crystalline phases. The high fly ash binders also showed a hump between 25-30°2𝜃 corresponds to the amorphous geopolymer gel. The XRD patterns at 28 days also show the presence of crystalline phases corresponding to unreacted particles from the original fly ash (mullite, quartz, and hematite). However, at 270 days, the height of these peaks decreased indicating continual geopolymerisation reaction. 25
ACCEPTED MANUSCRIPT
Above XRD results shows that geopolymers containing higher slag content produce higher content of calcium bearing phases including disordered calcium silicate hydrate phases. These phases do not have 3-dimensional cross linking like N-A-S-H phases which are produced in high fly ash binders. The disorderly structure in high slag binders can be easily attacked by
NU
SC
RI
PT
chemically.
(b)
AC
CE
PT E
D
MA
(a)
26
ACCEPTED MANUSCRIPT
(c) Figure 13: XRD patterns of geopolymer binders (a) 70/30 FA/S (b) 50/50 FA/S (c) 30/70 FA/S
4. Conclusions On the basis of obtained results, the following concluding remarks can be drawn: Apparent chloride diffusion coefficient of blended fly ash and slag geopolymer concrete is
PT
1.
lower than that of OPC concrete. The diffusion coefficient decreased with the slag content
2.
RI
in the binder.
Blended fly ash and slag geopolymer concrete has higher age factor than in OPC concrete
SC
indicating improved resistance to chloride ingress with time. The lower permeability is attributed to the higher binding capacity in the hydrolacite phases and higher degree of
3.
NU
geopolymerisation in blended fly ash and slag which have been reported in the literature. The time taken to initiate the corrosion in the rebar in blended fly ash and slag geopolymer
MA
concrete is significantly higher than that of OPC concrete. The time is longer when the slag in the geopolymer is higher. 4.
Chloride induced corrosion started in the rebar in OPC concrete exposed to 2.826 M NaCl
D
solution for 500 days while the blended fly ash and slag geopolymer concrete bar did not
PT E
corrode. The chloride content at the rebar in geopolymer is ten times lower than chloride content in OPC concrete. The results are consistent with the calculated initiation time. 5.
Geopolymer concrete containing slag provides greater protection against reinforced
CE
corrosion than that of OPC concrete when concrete is contaminated with chloride. This is due to reduction in oxygen around the rebar in the presence of slag in the binder (which
6.
AC
has been confirmed by the previously reported work). Electrochemical measurements (i.e. Half-cell potential and corrosion current) of the embedded rebar in geopolymer concrete do not correlate well with the corrosion activity.
Acknowledgements The authors gratefully acknowledge the financial support provided by Austroads.
27
ACCEPTED MANUSCRIPT
References [1] Mobili A, Belli A, Giosuè C, Bellezze T, Tittarelli F. Metakaolin and fly ash alkali-activated mortars compared with cementitious mortars at the same strength class. Cement and Concrete
PT
Research. 2016;88:198-210. [2] Chindaprasirt P, Chalee W. Effect of sodium hydroxide concentration on chloride penetration
RI
and steel corrosion of fly ash-based geopolymer concrete under marine site. Construction and
SC
Building Materials. 2014;63:303-10.
NU
[3] Reddy DV, Edouard J-B, Sobhan K. Durability of Fly Ash–Based Geopolymer Structural Concrete in the Marine Environment. Journal of Materials in Civil Engineering. 2013;25(6):781-
MA
7.
[4] Fernando Pacheco-Torgal JL, Leonelli C, Palomo A, Chindaprasit P. Handbook of Alkali-
PT E
D
Activated Cements, Mortars and Concretes. 2015. [5] Provis JL, Van Deventer JSJ. Alkali-activated materials: state of the art report. 2013.
CE
[6] Ghafoori N, Najimi M, Sobhani J, Aqel MA. Predicting rapid chloride permeability of self-
AC
consolidating concrete: A comparative study on statistical and neural network models. Construction and Building Materials. 2013;44:381-90. [7] Silva N. Chloride Induced Corrosion of Reinforcement Steel in Concrete Threshold Values and Ion Distributions at the Concrete-Steel Interface [PhD]: Chalmers University Of Technology, Sweden.; 2013. [8] Ganesan N, Abraham R, Deepa Raj S. Durability characteristics of steel fibre reinforced geopolymer concrete. Construction and Building Materials. 2015;93:471-6. 28
ACCEPTED MANUSCRIPT
[9] ASTM International. ASTM C1556: Standard test method for determining the apparent chloride diffusion coefficient of cementitious mixtures by bulk diffusion. 2011. [10] Olivia M, Nikraz HR. Durability of Fly Ash Geopolymer Concrete in a Seawater
PT
Environment. Concrete conference. Perth, Australia2011. [11] Shayan A, Xu, A., Andrews-Phaedonos, F. Field Performance of Geopolymer Concrete,
SC
Concrete Institute of Australia. Sydney, Australia2013.
RI
Used as a Measure Towards Reducing Carbon Dioxide Emission. 26thBiennial Conference
NU
[12] ASTM International. ASTM C1202: Standard test method for electrical indication of concrete's ability to resist chloride ion penetration. 2012.
MA
[13] Nordtest Method. NT BUILD 443: Concrete, Hardened: Accelerated Chloride Penetration.
D
1995.
PT E
[14] Ma Q, Nanukuttan SV, Basheer PAM, Bai Y, Yang C. Chloride transport and the resulting corrosion of steel bars in alkali activated slag concretes. Materials and Structures. 2015:1-15.
CE
[15] Monticelli C, Natali ME, Balbo A, Chiavari C, Zanotto F, Manzi S, et al. Corrosion behavior of steel in alkali-activated fly ash mortars in the light of their microstructural,
AC
mechanical and chemical characterization. Cement and Concrete Research. 2016;80:60-8. [16] Miranda JM, Fernández-Jiménez A, González JA, Palomo A. Corrosion resistance in activated fly ash mortars. Cement and Concrete Research. 2005;35(6):1210-7. [17] Bastidas DM, Fernández-Jiménez A, Palomo A, González JA. A study on the passive state stability of steel embedded in activated fly ash mortars. Corrosion Science. 2008;50(4):1058-65.
29
ACCEPTED MANUSCRIPT
[18] Bencardino F, Condello A. Innovative solution to retrofit RC members: InhibitingRepairing-Strengthening (IRS). Construction and Building Materials. 2016;117:171-81. [19] Bencardino F, Condello A. Eco-friendly external strengthening system for existing
PT
reinforced concrete beams. Composites Part B: Engineering. 2016;93:163-73. [20] Glasser FP, Luke K, Angus MJ. Modification of cement pore fluid compositions by
RI
pozzolanic additives. Cement and Concrete Research. 1988;18(2):165-78.
SC
[21] RILEM TC 178-TMC. Testing and modelling chloride penetration in concrete analysis of
NU
total chloride content in concrete recommendation. Materials and Structures. 2002;35:583-5. [22] Australian Standards. AS 1012.3.1:Determination of properties related to the consistency of
MA
concrete—Slump test. 2014.
D
[23] Aimin X, Ahmad S. Relationship between Reinforcing Bar Corrosion and Concrete
PT E
Cracking. Materials Journal.113(01).
[24] ASTM International. ASTM C896: Standard Test Method for Half-Cell Potentials of
CE
Uncoated Reinforcing Steel in Concrete. 1999.
AC
[25] Kupwade-Patil K, Allouche EN. Examination of Chloride-Induced Corrosion in Reinforced Geopolymer Concretes. Journal of Materials in Civil Engineering. 2013;25(10):1465-76. [26] Lizarazo-Marriaga J, Claisse,Peter., . Determination of the concrete chloride diffusion coefficient based on an electrochemical test and an optimization model. Materials, chemistry and physics. 2009;113(3):536-43.
30
ACCEPTED MANUSCRIPT
[27] Ismail I, Bernal SA, Provis JL, San Nicolas R, Brice DG, Kilcullen AR, et al. Influence of fly ash on the water and chloride permeability of alkali-activated slag mortars and concretes. Construction and Building Materials. 2013;48:1187-201. [28] Kumar S, Kumar R, Mehrotra SP. Influence of granulated blast furnace slag on the reaction,
PT
structure and properties of fly ash based geopolymer. Journal of Materials Science.
RI
2009;45(3):607-15.
SC
[29] Puertas F, Fernández-Jiménez A. Mineralogical and microstructural characterisation of
NU
alkali-activated fly ash/slag pastes. Cement and Concrete Composites. 2003;25(3):287-92. [30] Kayali O, Khan MSH, Sharfuddin Ahmed M. The role of hydrotalcite in chloride binding
MA
and corrosion protection in concretes with ground granulated blast furnace slag. Cement and Concrete Composites. 2012;34(8):936-45.
D
[31] Chen W, Brouwers HJH. A method for predicting the alkali concentrations in pore solution
PT E
of hydrated slag cement paste. Journal of Materials Science. 2011;46(10):3622-31.
CE
[32] Evans DG, Slade RCT. Structural aspects of layered double hydroxides. Layered double hydroxides2006. p. 1-87.
AC
[33] Narayanan S, Krishna K. Structure activity relationship in Pd/hydrotalcite: effect of calcination of hydrotalcite on palladium dispersion and phenol hydrogenation. Catalysis Today. 1999;49(1–3):57-63. [34] Provis JL, Myers RJ, White CE, Rose V, Van Deventer JSJ. X-ray microtomography shows pore structure and tortuosity in alkali-activated binders. Cement and Concrete Research. 2012;42(6):855-64.
31
ACCEPTED MANUSCRIPT
[35] Lloyd RR, Provis JL, Smeaton KJ, Van Deventer JSJ. Spatial distribution of pores in fly ash-based inorganic polymer gels visualised by Wood’s metal intrusion. Microporous and Mesoporous Materials. 2009;126(1–2):32-9. [36] Yang T, Yao X, Zhang Z. Quantification of chloride diffusion in fly ash–slag-based
PT
geopolymers by X-ray fluorescence (XRF). Construction and Building Materials. 2014;69:109-
RI
15.
SC
[37] Li Z, Liu, Sifeng. Influence of Slag as Additive on Compressive Strength of Fly Ash-Based
NU
Geopolymer. Journal of Materials in Civil Engineering. 2007;19(6):470-4. [38] Bernal SA, Provis JL, Fernández-Jiménez A, Krivenko PV, Kavalerova E, Palacios M, et
MA
al. Binder Chemistry – High-CalciumAlkali- Activated Materials Alkali activated materials2013. p. 59-77.
D
[39] Florea MVA, Brouwers HJH. Chloride binding related to hydration products: Part I:
PT E
Ordinary Portland Cement. Cement and Concrete Research. 2012;42(2):282-90.
CE
[40] Odler I. Hydration, Setting and Hardening of Portland Cement. Lea's Chemistry of Cement and Concrete1998. p. 241-91.
450.
AC
[41] Neville AM. Durability of concrete. Properties of Concrete. 5th edition ed1996. p. 1350-
[42] Gjorv OE. Durability analysis.
Durability design of concrete structure in severe
enviournments2014. [43] Roy DM, Idorn GM. Concrete Microstructure. National Research Council, Washington, DC: Strategic Highway Research Program,; 1993. p. 50-115. 32
ACCEPTED MANUSCRIPT
[44] Attari A, McNally, C., Richardson, M.G.,. Probability-based modeling of corrosion initiation times using mix-dependent age factors.
Safety, Reliability, Risk and Life-Cycle
Performance of Structures and Infrastructures2013. p. 2935-345. [45] Xu H, Provis, John L., Van Deventer, J.S.J., Krivenko, P.V. Characterization of aged slag
PT
concretes. ACI Materials Journal. 2008;105(2):131-9.
RI
[46] Ismail I, Bernal SA, Provis JL, San Nicolas R, Hamdan S, Van Deventer JSJ. Modification
SC
of phase evolution in alkali-activated blast furnace slag by the incorporation of fly ash. Cement
NU
and Concrete Composites. 2014;45:125-35.
[47] Ismail I, Bernal SA, Provis JL, San Nicolas R, Brice DG, Kilcullen AR, et al. Influence of
MA
fly ash on the wate and chloride permeability of alkali-activated slag mortars and concretes. Construction and Building Materials. 2013;48:1187-201.
D
[48] Song L, Sun W, Gao J. Time dependent chloride diffusion coefficient in concrete. Journal of
PT E
Wuhan University of Technology-Mater Sci Ed. 2013;28(2):314-9.
CE
[49] Cement concrete and aggregates Australia. Chloride resistance of concrete. 2009. p. 30-5. [50] Buenfeld NR, Newman JB. Examination of three methods for studying ion diffusion in
AC
cement pastes, mortars and concrete. Materials and Structures.20(1):3-10. [51] Luping T, Nilsson, L-O. Chloride diffusivity in high strength concrete as different ages. Nordic concrete research. 1992;10:162-71. [52] Costa A, Appleton J. Chloride penetration into concrete in marine environment-Part II: Prediction of long term chloride penetration. Materials and Structures.32(5):354-9.
33
ACCEPTED MANUSCRIPT
[53] Xu A, Shayan, A. Determination of Chloride Diffusion Coefficient of Concrete: Comparison of Bulk Diffusion and Electrical Field Method. 27th Biennial National Conference of the Concrete Institute of Australia in conjunction with the 69th RILEM Week. Melbourne, Australia2015.
PT
[54] Hou H, Tian L. The time-dependence of chloride penetration resistance of concrete by the
RI
rapid chloride permeability test Applied mechanics and materials. 2013;357:651-4.
SC
[55] Roy DM, Jiang, W. and Silsbee, M.R. Chloride diffusion in ordinary, blended, and alkaliactivated cement pastes and its relation to other properties. Cement and Concrete Research.
NU
2000;30(12):1879-84.
MA
[56] Shi C. Strength, pore structure and permeability of alkali-activated slag mortars. Cement and Concrete Research. 1996;26(12):1789-99.
D
[57] Luping T, Gulikers J. On the mathematics of time-dependent apparent chloride diffusion
PT E
coefficient in concrete. Cement and Concrete Research. 2007;37(4):589-95.
CE
[58] Elfmarkova V, Spiesz P, Brouwers HJH. Determination of the chloride diffusion coefficient in blended cement mortars. Cement and Concrete Research. 2015;78, Part B:190-9.
AC
[59] Meck E, Sirivivatnanon V. Field indicator of chloride penetration depth. Cement and Concrete Research. 2003;33(8):1113-7. [60] Yeau KY, Kim EK. An experimental study on corrosion resistance of concrete with ground granulate blast-furnace slag. Cement and Concrete Research. 2005;35(7):1391-9. [61] Bushman JB. Calculation of corrosion rate from corrosion current (Faraday's Law). 2000. p. 1-4. 34
ACCEPTED MANUSCRIPT
[62] Holloway M, Sykes JM. Studies of the corrosion of mild steel in alkali-activated slag cement mortars with sodium chloride admixtures by a galvanostatic pulse method. Corrosion
AC
CE
PT E
D
MA
NU
SC
RI
PT
Science. 2005;47(12):3097-110.
35
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
NU
SC
RI
PT
Graphical Abstract
36
Chandani Tennakoon, Ahmad Shayan, Jay G Sanjayan, Aimin Xu PII: DOI: Reference:
S0264-1275(16)31547-7 doi: 10.1016/j.matdes.2016.12.030 JMADE 2575
To appear in:
Materials & Design
Received date: Revised date: Accepted date:
21 August 2016 11 November 2016 11 December 2016
Please cite this article as: Chandani Tennakoon, Ahmad Shayan, Jay G Sanjayan, Aimin Xu , Chloride ingress and steel corrosion in geopolymer concrete based on long term tests. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Jmade(2016), doi: 10.1016/j.matdes.2016.12.030
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Chloride ingress and steel corrosion in geopolymer concrete based on long term tests Chandani Tennakoona,1, Ahmad Shayanb, Jay G Sanjayana, Aimin Xub a
Centre of Sustainable Infrastructure, Faculty of Science, Engineering and Technology, Swinburne University of Technology, Hawthorn, Victoria, Australia. ARRB Group Ltd, Vermont South, Victoria, Australia.
PT
b
RI
Abstract
Alkali activated binders (geopolymer) is an emerging technology to produce concrete without
SC
the use of any Ordinary Portland Cement (OPC) by alkali activation of alumino-silicate source materials such as fly ash and/or slag. Limited knowledge on durability issues like corrosion
NU
behaviour of reinforced geopolymer concrete impedes the usage of this technology in structural applications.
MA
This study explored the chloride permeability and initiation of chloride induced corrosion of geopolymer concrete in accelerated chloride environment using longer test period. Corrosion
D
initiation was also monitored in embedded rebar in 2% chloride contaminated concrete.
reference electrode.
PT E
Corrosion state of the rebar was monitored using non-destructive test method using Cu/CuSO4
The results showed that the apparent chloride diffusion coefficient of blended fly ash and slag
CE
geopolymer concrete is lower than that of OPC concrete. The diffusion coefficient also decreased with the increase of slag content in the binder. Blended fly ash and slag geopolymer
AC
concrete also exhibited higher aging factor than OPC concrete indicating improved resistance to chloride ingress with time. The study also showed that the embedded rebar in fly ash and slag based geopolymer concrete has higher protection against corrosion than a rebar in OPC concrete even when the concrete is contaminated with significant levels of chloride.
1
Corresponding author; email: [email protected] 1
ACCEPTED MANUSCRIPT
Highlights Chloride ingress rate in geopolymers containing slag is low
Age factor of geopolymer is significantly higher than OPC
Geopolymer can hinder corrosion in steel rebar
SC
RI
PT
AC
CE
PT E
D
MA
NU
Keyword: Age factor, Chloride ingress, Corrosion, Fly ash, Geopolymers, Slag
2
ACCEPTED MANUSCRIPT
1. Introduction The term geopolymer describes a broad class of synthetic zeolite materials that are produced by a series of reactions between alumina-silicate source materials and alkali activators. Due to the reduced carbon foot print and low embodied energy, geopolymer concrete technology has gained
PT
significant attention in recent years as an alternative to Ordinary Portland cement (OPC) concrete for production of structural and non-structural elements. To date, geopolymer concrete has had
RI
relatively minor application in field concrete due to the limited knowledge base on its durability [1]. Its long-term durability properties as well as its protection against reinforcements corrosion
SC
are yet to be fully investigated [2]. Some studies of marine-exposed concrete specimens have shown that the performance of geopolymer concrete specimens is comparable to that of OPC
NU
concrete with respects to reinforced corrosion [3-5].
The main cause of corrosion in concrete structures, particularly in concrete structures in marine
MA
environment is the chloride ingress [6, 7]. Ganesan et al. [8] studied the chloride diffusion of 100% fly ash based geopolymer and reported that chloride diffusion of geopolymer and OPC
D
concrete are similar. In their study authors employed AgNO3 spraying to identify the depth of chloride penetration, as suggested by ASTM C1556 [9]. In contrast, the study by Olivia et al.
PT E
[10] reported that the rate of chloride penetration in 100% fly ash based geopolymer concrete is higher than that of OPC concrete. Shayan et al. [11] characterised the properties of a 100% slag-based geopolymer concrete after having been in service for five years in the retaining walls
CE
of a bridge structure on the Yarra river in Melbourne, Australia. The rapid chloride permeability test ASTM 1202 [12] showed very low chloride permeability. Based on NT Build 443 [13]
AC
chloride diffusion coefficient was an order of magnitude lower than that of blended slag cement concrete used in the retaining walls of the same bridge. More recently, Ma et al. [14] followed NT build 443 test method [13] to evaluate the chloride diffusion coefficient of 100% slag-based geopolymer and found that chloride diffusion rate of slag based geopolymer concrete is lower than OPC concrete. However, in the same study, Ma et al. [14] found that the corrosion rate of rebar inside prism specimens is similar to that of the embedded rebar inside OPC concrete. The above studies were mainly focused on corrosion due to chloride ingress. There are also a few studies on chloride induced corrosion in contaminated concrete. Monticelli et al. [15] added
3
ACCEPTED MANUSCRIPT
2% chloride by mass of binder to 100% class F fly ash based geopolymer during the mixing process. The authors reported that the availability of silicate ions in the pore solution of contaminated geopolymer concrete can provide additional protection for the passivity of rebar. Miranda et al. [16] observed that addition of 2% chloride by mass of binder to fly ash based geopolymer mix increase the corrosion rate by 100 times compared to uncontaminated geopolymer concrete. They also found that the steel embedded in contaminated geopolymer did
PT
not have the protective passive layer and show similar corrosion behaviour to OPC. The result of Bastidas et al. [17] is in agreement with the previous observations. More recently, an innovative
RI
technique for repair and strengthening of reinforced concrete members with deteriorated cover
SC
concrete and/or corroded steel rebars was proposed and named inhibiting-repairing-strengthening (IRS) technique [18, 19]. It consists in the installation of a stainless steel fabric in the cover concrete, restoring this latter with a geopolymer-based matrix which also act as a corrosion
NU
protection.
MA
Blending of class F fly ash and slag enables geopolymers modifies the reaction phases and the microstructure of geopolymer compared to 100% fly ash or 100% slag based geopolymer leading to different level of chloride binding and electro chemical mechanisms. Slag leaches sulphate
D
ions into pore solution leading to high negative half-cell redox potential in the embedded rebar
PT E
[20]. Corrosion mechanism of embedded rebar in slag blended geopolymer concrete has not been discussed in the litreture. In the present study chloride ingress of slag blended geopolymer concrete is evaluated using potentiometric titration method [14, 21]. This study also examines
CE
the change in chloride diffusion rate with time in blended fly ash and slag geopolymer which can be used to quantify the age factor of concrete. The corrosion state of embedded rebar using electrochemical method (half-cell potential and corrosion current) was evaluated. These results
AC
are compared with the actual corrosion state of the rebar. Finally, the study evaluates the corrosion state of the embedded rebar in blended fly ash and slag geopolymer concrete contaminated with 2% (by mass of binder) The purpose is to accelerate the corrosion so that the investigation can be completed in realistic time period.
4
ACCEPTED MANUSCRIPT
2.
Materials and methods
2.1.Materials Precursor materials A class F fly ash (FA) which was a 50/50 mixture of Gladstone and Callide fly ashes (FA) and granulated ground blast furnace slag (Slag), with the chemical compositions shown in Table 1
PT
were used as the alumino-silicate precursor materials. The bulk chemical compositions were determined by X-ray florescence (XRF) analysis. The main component of class F fly ash is SiO2
RI
(over 51%) whereas slag contains high amount of CaO (41.5%). It is important to note that fly ash contains twice as much Al2O3 as slag does. The particle size distributions of fly ash and slag,
SC
obtained using a Cilas laser diffraction particle analyser are shown in Figure 1. The particle size analysis shows that both fly ash and slag have small particle sizes which are desirable for the
MA
while slag has about 2.5% of finer than 1µm.
NU
geopolymer reaction. It also shows that class F fly ash has 10% of very fine particles (< 1µm),
Table 1: Bulk chemical composition of materials (wt. %)
AC
CE
GP Cement 19.30 0.30 5.50 2.90 0.20 1.40 62.00 0.06 0.40 0.11 2.46 0.00 5.3
D
PT E
Oxides SiO2 TiO2 Al2O3 Fe2O3 Mn3O4 MgO CaO Na2O K2O P2O5 SO3 S2LOI
Slag 32.00 0.51 13.00 0.40 0.30 4.90 41.50 0.20 0.33 0.02 2.10 0.60 4.9
FA 51.50 1.50 27.60 11.80 0.15 1.30 2.20 0.40 0.60 0.73 0.2 0.00 1.8
5
PT
ACCEPTED MANUSCRIPT
(b)
RI
(a)
SC
Figure 1: Particle size distribution of (a) Slag (b) Fly ash. 1-10 g of source material was dispersed in water in order to obtain the particle size.
NU
Alkali activator
Granular sodium metasilicate penta-hydrate (28.7 wt% Na2O, 28.5 wt% SiO2 and 42.8 wt% H2O) was used as solid activator to activate the precursors. Usage of solid alkali activators in
MA
geopolymer systems is less hazardous than commonly used liquid activators and more convenient for field applications. The amount of activator used was calculated to provide an
D
alkali content of 4% Na2O by mass of precursor materials.
PT E
Aggregates
A coarse basalt aggregate of 14 mm nominal particle size and river sand were used in the geopolymer concrete mix. The particle density of coarse aggregates and the sand in saturated
CE
surface dry condition (SSD) was 2830 and 2620 kg/m3, respectively. The moisture absorption of
AC
coarse aggregates and the sand were 2.2 and 0.3% respectively. 2.2.Preparation of geopolymer concrete The mix designs for the geopolymer concretes employed in this work are shown in Table 2. The concrete mixing procedures consistent of: Sand, coarse aggregate and precursor materials were mixed thoroughly for 1 minute. Water was added to the mix and mixing was continued for another 2 minutes. Alkali activator (solid form) was added into wet mix and mixed for around 7 minutes to obtain workable geopolymer concrete. The workability of geopolymer concrete mix was measured using the standard slump test according to AS1012.3.1 [22]. The target slump was 80-100 mm. 6
ACCEPTED MANUSCRIPT
Two identical concrete prisms (95 × 95 × 300 mm) were cast from each, mix embedding a 20 mm diameter carbon steel rebar. A sample prism is shown in Figure 2. The concrete mixes were also used to cast two cylindrical specimens (100×200 mm) for NT build 443 testings. The cast specimens were transferred into a standard fog room (23°C, 100% RH). After 24 hours, concrete specimens were de-moulded and stored inside fog room until 28 days. During the storing period in the fog room, the samples were covered with plastic sheets to avoid the water dropping on the
PT
sample.
RI
To identify the corrosion behaviour in contaminated concrete, 50/50 FA/Slag geopolymer concrete specimens were cast with added 2% NaCl by mass of binder and partially immersed in
SC
water. The temperature during the immersion test was maintained at 23°C. An OPC concrete
NU
prism was also cast with added 2% NaCl by mass of binder as a control. Table 2: Mix design for concrete (Alkali content = 4% for geopolymers) Slag (kg/m3)
60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
240 200 180 0
180 200 240 0
OPC (kg/m3) 0 0 0 400
Aggregate (kg/m3)
Sand 14mm 680 680 680 680
630 630 630 630
10mm 450 450 450 450
Water/ Solid ratio 0.390 0.425 0.449 0.464
AC
CE
PT E
D
MA
Mix
FA (kg/m3)
(a) (b) Figure 2: (a) Specimens used for chloride induced corrosion testing (b) Gecor8 instrument
7
ACCEPTED MANUSCRIPT
2.3.Testing procedure Chloride diffusion tests After 28 days of moist curing at 23°C each concrete cylinder was segmented to three parts with heights 51, 51 and 98 mm respectively. The curved surfaces of the cylindrical segments were then coated with quick setting epoxy which prevented chloride transportation into the concrete
PT
from the sides. Chloride diffusion coefficients of geopolymer and OPC concretes were measured using NT
RI
Build 443 [13] method. The 98 mm cylindrical section was immersed in 2.826 M sodium
SC
chloride solution vertically such that approximately 1/4 of the height was submerged inside the solution. Cylindrical specimens were mounted on PVC rings in order to maintain uniform
NU
chloride exposure through the bottom surface. The chloride solution was stirred frequently in order to have a uniform solution. After 5 weeks of immersion, the concrete specimens were drilled in 1 mm depth increments using a profile grinder to obtain fine concrete powder from
MA
each 1 mm layer (75 mm diameter circular area is drilled). The obtained concrete powder was dried at 105°C for 24 hours (powder was covered properly avoiding direct heat exposure). Approximately, 5 g mass of the dried concrete powder was mixed with distilled water (1:5 solid
D
to water ratio) and HNO3. The resulting solution was boiled and filtered to extract the chloride
PT E
ions from concrete powder. Potentiometric titration was conducted on the filtered solution using standard 1N AgNO3 solution to obtain the chloride content (acid soluble chloride) of each layer.
CE
The rapid chloride permeability of concretes was measured using ASTM 1202 standard [12]. In this test each ends of a 51 mm (2 inch) thick concrete disk is embedded between two cells each
AC
fitted with a stainless steel mesh. One cell is filled with a 0.6 M NaCl solution and connected to the negative terminal and the other filled with 0.3 M NaOH solution. A DC potential of 60 Volts is applied across the concrete specimen for 6 hours during which the current is recorded every 5 minutes, and the total charge after 6 hours is reported by the testing device. Chloride permeability of geopolymer concrete was measured at three ages, namely 28 days, 180 days and 500 days. Measurements of corrosion state of embedded rebar using electrochemical measurements After 28 days of ambient curing at 23°C, the concrete prisms were partially immersed in 2.826 and 0.6 M NaCl solutions up to 500 days. Temperature of the storing room was 23°C. The 8
ACCEPTED MANUSCRIPT
corrosion rate meter was used to measure the corrosion current. It measures the corrosion current over a confined length of steel bar, the half-cell potential (copper-copper sulphate) of the steel and the electrical resistance of concrete. This instrument uses the “linear polarization” technique and calculates the polarization resistance and obtains the corrosion current according to Stern and Geary relationship [23].
PT
Gecor6 instrument was used to measure the half-cell potential of embedded rebars with reference copper-copper sulphate electrode. As prescribed in ASTM C876 standard for Cu/CuSO4
RI
reference electrode [24], the corrosion state of rebar can be assessed as follows: the half-cell potential < -200 mV (less negative) is an indication of 90% chance of having no corrosion while
SC
if the potential<-350 mV (more negative) it is an indication of 90% probability of having corrosion (half-cell potential limits are marked in the figure). Previously it was also reported
NU
that, if the corrosion rate is higher than 1 µA/cm2 it is an indication of a higher corrosion activity (high risk of corrosion). Instead, if the corrosion rate is in between 0.1 and 0.5 µA/cm2 it is a
MA
sign for low to moderate corrosion activity [25].
After 500 days, immersed concrete prisms were split to observe the corrosion state of embedded
PT E
potentiometric titration [13].
D
rebar. Chloride penetration depth and chloride content at the rebar were calculated using
3. Results and Discussion
3.1.Chloride diffusion rate of concrete in unsaturated condition
CE
Corrosion due to chloride ingress is initiated when sufficient amount of chloride ions penetrate to the depth of reinforcement steel [7]. The resistance against chloride transportation into concrete
AC
is quantified using the chloride diffusion coefficient [26]. To determine the rate of chloride diffusion into concrete NT build 443 test as described in section 2.3 was conducted. The apparent diffusion coefficient of a concrete (Da mm2/year) can be calculated using the Fick’s second Law of Diffusion. According to this law the relationship between the calculated chloride content of concrete (C(t,x) ), and the depth from the exposed surface (𝑥) is as follows: 𝐶(𝑡,𝑥) − 𝐶0 = (𝐶𝑆 − 𝐶0 )erfc(
𝑥 t √∫t 𝐷(𝑡) 𝑑𝑡 s
)]
(1)
9
ACCEPTED MANUSCRIPT
Where CS is the chloride concentration at the surface (x=0), C0 is the chloride concentration in concrete before immersion and t is time in years and t s age of concrete when exposure started. The chloride profile of concrete specimens obtained using the above procedure was then used to calculate the chloride coefficient of each concrete type. Figure 3 shows the chloride profiles obtained for the geopolymer and OPC concretes while Table
PT
3 shows the calculated chloride diffusion coefficients. The chloride profile and the chloride diffusion coefficient results show that the chloride diffusion rate decreases significantly when the
RI
slag content of the geopolymer binder is increased.
SC
Geopolymer concrete containing significant level of slag has the following main reaction products: calcium silicate hydrate (C-S-H), calcium aluminate silicate hydrate (C-A-S-H),
NU
sodium aluminate silicate hydrate (N-A-S-H) and hydrotalcite phases [27-29]. When the slag content of the geopolymer blends increases, the resulting binder consists more C-S-H, C-A-S-H than N-A-S-H and higher amount of hydrotalcite phases [27, 30]. It is known that C-A-S-H
to
N-A-S-H
which
can
limit
MA
products have higher a binding capacity for chloride ions (and accompanied Na ions) compared the
ingress
of
those
ions
[27].
Hydrotalcite
(Mg6Al2(OH)16CO34H2O) has a layered double hydroxide structure. This layered crystal
D
structure consists of an outer layer with hydroxide and an inner layer with anions and water
PT E
molecules [30-32]. Therefore, hydrotalcite has large specific area which increase its capability to exchange ions with the concrete pore solutions and absorb anions to its structure [32]. Previous studies have shown that the addition of hydrotalcite into a chloride contaminated system
CE
externally result in the reduction of chloride content in the concrete without changing the
AC
structure of hydrotalcite [30, 33]. The X-ray diffraction patterns on slag blended geopolymer show the indication of hydrotalcite type phases with significant magnesium level in fly ash and slag based geopolymer which is also confirmed in literature (Figure 13) [30]. Geopolymer containing lower slag content mainly produces N-A-S-H type phases which have high pore volume [27, 28]. Provis et al.[34] and Lloyd et al. [35] showed that increasing fly ash content, the porosity of the geopolymer is increased significantly. They stated that geopolymer containing high slag content have high bound water (induced by calcium) which leads to more pore-filling capacity than geopolymer containing high fly ash [34]. This is in agreement with the studies by Yang et al. [36] and Li et
10
ACCEPTED MANUSCRIPT
al. [37]. Since fly ash based geopolymer has lower density than slag blended geopolymer, it can absorbs more external chloride ions and deposit in the binder pore solution. Previously it has been reported that geopolymers containing higher fly ash content (>50%) has halite (NaCl) in their pore solution than that of high slag based geopolymers [27]. The chloride diffusion coefficients results in Table 3 also show that slag blended geopolymer has
PT
higher resistance to chloride ingress compared to OPC based concrete. Geopolymer concrete containing significant level of calcium produces cross-linked tobermorite-like C-A-S-H which
RI
are high density structures compared to non-cross linked tobermorite-like C-S-H phases formed in OPC concrete [38]. These cross linked structures reduce the chloride ingress in slag blended
SC
geopolymer concrete compared to that of OPC concrete. OPC binder also contains calcium aluminates and calcium aluminoferrite phases that can absorb chloride ions forming
NU
3CaO·Al2O3·CaCl2·10H2O (Friedel’s salt). However, the study by Florea and Brouwers [39] stated that Friedel’s salt formation in OPC binder is significant only in chloride contaminated
MA
concrete during the initial hydration process. The Friedel’s salt formation after cement hydration is very slow hence this process would not be effective in absorbing chloride coming from the external sources in the long term [40, 41]. This explains the higher chloride diffusion in OPC
AC
CE
PT E
D
concrete.
Figure 3: Chloride profiles of concrete cylinders after 5 weeks of immersion test
Table 3: Chloride diffusion coefficient and surface chloride content of concrete Geopolymer
Surface Cl-. %
Apparent chloride 11
ACCEPTED MANUSCRIPT
formulation
(Cs ) (%wt. by concrete mass) 0.80 0.89 0.81 0.80
60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
diffusion coefficient (Da ) 10-12 m2/s 5.17 2.38 1.01 6.70
Table 4 shows the chloride content (Cl-% by mass of concrete) at the different depth of
PT
geopolymer and OPC concrete after 500 days immersion in 2.83 M NaCl solution. The results show that the chloride content of OPC concrete is higher than geopolymer concrete at all depths.
RI
At the rebar (40 mm depth), OPC concrete has 0.070% chloride (by mass of concrete) which is above the threshold chloride value that initiates corrosion [42]. On the other hand, geopolymer
SC
concrete shows 10 times lower chloride content at the rebar after 500 days which is not sufficient
NU
to initiate chloride induced corrosion.
The resistance to chloride ingress rate into concrete is reduced upon aging of the geopolymer concrete. It is well known that chloride diffusion rate of OPC concrete decreases with the age of
MA
concrete as a result of chloride binding and refinement in the pore structure [43]. It was reported that a 1 day old OPC binder has pore size range 8 nm to 700 nm while more than 365 days old
D
OPC binder has pores in the range of 16 nm to 64 nm [43] which significantly affects the chloride diffusion coefficient of concrete. Ignoring this effect can result in overestimation of the
PT E
chloride penetration rate [44]. Resistance to chloride ingress into concrete can vary due to modified microstructure and additional reaction products formed during extended curing. Geopolymerisation is a continual reaction process that can have different microstructure and new
CE
reaction phases in the later stages compared to premature age of concrete. Puertas et al. [29], showed that when blended fly ash and slag geopolymer matures, FTIR (T-O-T T: tetrahedral Si
AC
or Al) band shifted towards a higher value indicating higher degree of cross-linking in geopolymer than in early age concrete. This is attributed to the continual development of N-A-SH and C-A-S-H cross-linking and increase of number of reaction products. Their results are in agreement with several other studies [45, 46].
12
ACCEPTED MANUSCRIPT
Provis et al. [34] reported that longer curing of blended fly ash and slag geopolymer increases the pore-network tortuosity and decreases the porosity due to space filling nature of C-(N)-S-H phases. Xu et al. [45] reported that in the presence of slag, geopolymerisation is continual as slag provides hydroxyl buffering to react with the precursors in the later age. Moreover, the authors found that channels in the pore structure are disconnected due to maturity of geopolymer concrete. These studies show that resistance to chloride ingress into concrete is expected to be
RI
concrete will be quantified and analysed the factors affecting it.
PT
lower in the later age of geopolymer concrete. In the next section, the age factor of geopolymer
Average depth of the concrete (mm)
NU
Concrete type
SC
Table 4: Chloride percentage in the different depth of concrete after 500 days of immersion in 2.83 M NaCl solution
MA
50/50 FA/Slag
1 15 35 40 1 15 35 40
PT E
D
OPC
Measured chloride (%wt. by concrete mass) 0.880 0.300 0.020 0.007 0.989 0.364 0.299 0.077
3.2.1.
CE
3.2.Properties of geopolymer concrete Compressive strength
AC
Figure 4 (a) and (b) show the compressive strength development of 60/40 FA/Slag, 50/50 FA/Slag and 40/60 FA/Slag geopolymers. 40/60 and 50/50 FA/Slag based geopolymers showed the highest strength during first year. 60/40 FA/Slag based geopolymer had the maximum strength at 100 days and thereafter compressive strength is slightly reduced. Ismail et al. [47] found that the main reaction phase in geopolymer with slag percentage in the range of 50-100%, is calcium silicate hydrated gel (C-S-H) and calcium aluminate silicate hydrate (C-A-S-H). Moreover, in these binders, the number of reaction product formations and degree of crosslinking increased with extended curing time. The geopolymer with lower than 50% slag has a combination of C-A-S-H and N-A-S-H reaction products, and the reaction product formation and 13
ACCEPTED MANUSCRIPT
development of additional cross-linking of the geopolymer is slower compared to geopolymer containing higher slag (>50%). This reveals that geopolymer concrete synthesised using fly ash and slag under ambient curing has stable compressive strength within the test period of two years. Higher compressive strength indicates the denser microstructure that has lower gas
AC
CE
PT E
D
MA
NU
SC
RI
PT
permeability. Therefore, oxygen cannot be penetrated into concrete and the level of rebar.
Figure 4: Compressive strength development of fly ash and slag blended geopolymers
14
ACCEPTED MANUSCRIPT
3.3.Quantifying the age factor for geopolymer concrete The influence of the age on the chloride diffusion coefficient is an important parameter for service-life modelling of any concrete structure [48]. The chloride diffusion of concrete D(t) varies with the time as shown in the equation (2) [48]: ts D(t) = Dref [ ]m t
PT
(2)
Where, 𝒎(> 𝟎) is the age factor, 𝐃𝐫𝐞𝐟 is the chloride diffusion rate obtained using NTBuild 443
RI
test (Table 3) and𝐭 is exposure time in years and 𝐭 𝐬 age of concrete when exposure started. Rapid permeability test results at several ages of concrete can be used to obtain the value of
SC
𝐦(> 𝟎) for concrete. The rapid chloride permeability test (RCPT) ASTM C1202 [12] provides information on the instantaneous chloride diffusion of concrete as it is conducted within a short
NU
period of time (6 hours), whereas NTBuild 443 [12] test involves 35 days immersion period. Previous studies have shown that the changes in RCPT results upon aging would reflect the
MA
changes in the chloride diffusion change [49, 50]. In order to calculate the ′𝐦′factor of concrete, rapid chloride permeability of concrete was conducted at three ages of concrete (i.e. 28, 180 and 500 days) and total charges passed through the concrete is calculated according to the procedure
D
described in section 2.3 [26, 51]. The charges passed through concrete obtained from ASTM
PT E
C1202 [12] test method is used to calculate the 𝐦factor as shown in the equation (8) [52, 53]. G(t) = G0 t −m
(3)
CE
Where G(t)= charge passed through concrete at age t of concrete, GO = charge passed through concrete at aget o .
AC
Figure 5 shows the charges passed through concrete at different ages of concrete. The results show that the charges passed through 28 days old geopolymer concrete is in the range of 10002000 Coulumb which is classified as low permeable concrete [12]. Even though, 28 days old OPC concrete can be also classified as low permeable concrete, the number of charged passed through OPC concrete is slightly higher than the corresponding value for geopolymer concrete. At 180 days, the charges passed through geopolymer concrete are significantly lower than in 28 days. This changes the permeability classification of geopolymer concrete to “very low” permeable concrete range [12]. The charges passed through geopolymer concrete after one year is only slightly lower than that at 180 days. All the geopolymer concretes reached to the same 15
ACCEPTED MANUSCRIPT
level of permeability level after one year, with charges passed through concrete being 300-330 Columb. On the other hand, the 180 days old OPC concrete had the same permeability category (low) as the 28 days old OPC concrete. When the OPC concrete was older than one year the change in the permeability level was not significant. The calculated age factors are presented in Table 5. The results show that the age factor of
PT
geopolymer concrete is in the range of 0.40-0.60. On the other hand age factor of OPC concrete is only 0.192 which is half of the age factor of geopolymer concrete. The value obtained for
RI
100% OPC concrete is similar to the values reported in previous studies [53, 54]. This shows aging improves the resistance of geopolymer concrete for chloride diffusion more significantly
SC
than for OPC concrete. As mentioned previously, when fly ash and slag geopolymer concrete matures, a higher degree of cross-linked structure is achieved than in early age, more reaction
NU
products (like hydrotalcite) are formed from unreacted fly ash and slag particles and pore structure of the binder is refined [27, 30, 34]. The combination of these parameters reduced the
MA
chloride diffusion of the concrete upon aging at a higher rate than for OPC concrete. Even though the pore structure is refined upon aging of OPC binder, not as much additional products are formed for binding chloride [40, 41, 43] in turn a lower age factor is obtained.
D
Previous studies [55, 56] have reported that, when used on geopolymer binders, the outcome of
PT E
the RCPT test is greatly influenced by the ion mobility of the pore solution rather than the chloride permeability. However, current study found that RCPT results correlates well with the
AC
CE
chloride induced corrosion.
16
SC
RI
PT
ACCEPTED MANUSCRIPT
Table 5: Age factor of concrete
Age factor of concrete (m) 0.407 0.604 0.509 0.192
D
MA
Mix proportions 60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
NU
Figure 5: The change of charged passed through concrete for concrete specimens of different ages
PT E
3.4.Service life prediction of concrete
Evaluation of the time to initiate corrosion on the embedded rebar in concrete is essential in modelling the service life of concrete structures. The results in Table 4 and the state of rebar
CE
(section 3.3) show that in OPC corrosion has initiated at a chloride level of 0.077% by mass of concrete. This chloride level has also been identified as the threshold level of the chloride to
AC
initiate corrosion process in rebar embedded in OPC concrete [42]. Assuming the same threshold level for geopolymer concrete, the time taken to initiate the corrosion can be calculated using the equations Fick’s Law [57]. The following assumptions are used in the calculation of the corrosion initiation time. 1. The threshold chloride level for the initiation of the steel corrosion is 0.077% by weight of concrete. 2. The diffusion rate only changes during the initial 10 years. There after it is assumed to be constant [53]. 17
ACCEPTED MANUSCRIPT
The results presented in Table 6 shows the time taken to initiate chloride induced corrosion when the cover thickness of concrete is 25 and 40 mm and the environment chloride concentration is 2.826 M. According to the calculations, OPC concrete initiates corrosion in 0.69 years which is consistent with the observations. OPC concrete takes 1.95 years to initiate corrosion when the cover thickness is 40 mm.
PT
On the other hand geopolymer concrete has very high corrosion initiation time, particularly the geopolymer concrete containing higher slag (>50%). The time taken to initiate chloride induced
RI
corrosion is increased with the slag content in the geopolymer concrete. This results show that geopolymer concrete structures would have higher service life in chloride environment compared
SC
to OPC concrete structures. It should be noted that a typical marine environment has a chloride concentration of around 0.6% which is approximately a quarter of the concentration used for the
NU
calculations (2.8%). Therefore, the initiation time will be much longer. The corrosion state of the rebar is consistent with the chloride calculations.
MA
Table 6: Number of years to reach chloride level of 0.077% in 2.826 M NaCl solutions. (Chloride (%) = mass of chloride/ concrete mass)
60/40 FA/Slag 50/50 FA/Slag 40/60 FA/Slag OPC
CE
Number of years to reached 0.077% chloride in different concrete depths 25 mm 40 mm 1.28 5.26 7.87 37.28 13.19 41.39 0.69 1.95
D
PT E
Mix proportions
Results presented in the previous section showed that chloride diffusion in blended fly ash and
AC
slag geopolymer concrete is lower than in OPC concrete. Therefore, the initiation time for corrosion is expected to be longer in geopolymer concrete than in OPC. To determine the corrosion state of the rebar, the specimens which had been immersed in the 2.826 M NaCl solution for 500 days were split, and immediately 0.1M AgNO3 solution was sprayed to measure the chloride penetrated depth, based on the depth of white colour observed due to formation of AgCl. In general, when the chloride content of concrete is in the range 0.02-0.23, the colour change is observed [58, 59]. Figure 6 shows the concrete samples after spraying AgNO3 solution and Table 7 shows the measured chloride penetrated depth of concrete obtained from AgNO3. The depth obtained using AgNO3 colorimetric method is based on the apparent chloride content 18
ACCEPTED MANUSCRIPT
of the concrete. Geopolymer concrete immersed in 2.826 M NaCl solution has very low chloride penetrate depth (lower than 5 mm). Moreover, when the geopolymer concrete prisms were immersed in 0.600 M NaCl solution, apparent chloride penetration depth is further reduced to 2 mm. On the other hand, OPC concrete immersed in 2.826 M and 0.600 M NaCl solutions showed 35.0 mm and 18.0 mm chloride penetration depth respectively which is significantly higher than that of geopolymer concrete samples. Since all the geopolymer concrete types had
PT
similar corrosion and chloride penetrated depth, only the 50/50 FA/Slag geopolymer concrete
MA
NU
SC
RI
was samples are shown in the Figure 6.
(b)
CE
PT E
D
(a)
AC
(c) (d) Figure 6: Chloride penetration depth of concrete (a) 50/50 FA/S geopolymer immersed in 2.832 M NaCl solution (b) 50/50 FA/S geopolymer immersed in 0.6 M NaCl solution (c) OPC concrete immersed in 2.826 M NaCl solution (d) OPC concrete immersed in 0.6 M NaCl solution Table 7: Depth of chloride obtained from AgNO3 colorimetric method
Sample ID 60/40 FA/S 50/50 FA/S 40/60 FA/S OPC
Depth in 2.826 M NaCl solution (mm) 4.2 4.0 4.1 35.0
Depth in 0.6 M NaCl solution (mm) 1.2 2.5 1.8 18.0
3.5.Corrosion state of rebar 19
ACCEPTED MANUSCRIPT
Figure 8(a) and Figure 8(b) respectively show half-cell potential and corrosion current of embedded rebar. The electrochemical indicate that embedded rebar in geopolymer concrete has 90% probability of having corrosion which is not consistent with the actual corrosion state of the rebar. 50/50 FA/S geopolymer has half-cell potential around -700 mV consistently in entire measuring period. Furthermore, rebar in 60/40 FA/S shows the half-cell potential around -600 Mv. Rebar in OPC concrete shows the half-cell potential -200 mV which is an indication of
PT
uncertainty in corrosion. 40/60 FA/S geopolymer also shows high negative value that indicates corrosion activity. In the previous section, it has been observed that chloride was not reached to
RI
the level of the rebar in geopolymer concrete while chloride was reached to rebar level in OPC
SC
concrete when they are in 2.83 M NaCl solution.
Half-cell potential and corrosion current of embedded rebar in geopolymer concrete when they
NU
are immerged in 0.6 M NaCl solution shows scattered behaviour (Figure 8). Rebar in OPC concrete shows lower negative half-cell potential than the samples in 2.83 M NaCl solution.
MA
Despite the fact the chloride is not reached to the level of rebar for initiating corrosion, the corresponding half-cell potential for embedded rebars in geopolymer concrete pushed towards high negative values (>-300 mV). Slag blended geopolymer release sulphide ions into pore
D
solution of concrete leading significant reduction in redox potential. Slag-rich blends can reduce
PT E
100 mV of rebar in OPC concrete to minus 200-225 mV [Glaser, 1988 #58]. The similar observations were made by Yeau and Kim [60] for the half-cell potential in slag blended OPC concrete. They stated that sulphur species in slag-rich binders (S2-, SO3- and S2O32-) resulted
CE
higher negative half-cell potential in embedded rebar. The half-cell potential of slag blended geopolymer concrete is also high negative. The actual corrosion state of the embedded rebar in slag blended geopolymer concrete is not consistent with the half-cell potential. Half-cell
AC
potential of rebar embedded in OPC when immerge in 2.83 M NaCl solution, indicates uncertainty in corrosion. It has been observed that the embedded rebar is initiated to corrode. Corrosion state can be seen in the Figure 9 and Figure 10.
20
RI
PT
ACCEPTED MANUSCRIPT
(b)
PT E
D
MA
NU
SC
(a)
AC
CE
Figure 7: Corrosion state of embedded rebar when concrete is immersed in 2.826 M NaCl solution (a) half-cell potential (b) corrosion rate
(a)
21
PT
ACCEPTED MANUSCRIPT
(b)
SC
RI
Figure 8:Corrosion state of embedded rebar when concrete is immersed in 0.6 M NaCl solution (a) half-cell potential (b) corrosion rate Figure 9 shows the corrosion state of rebar inside 50/50 FA/Slag geopolymer and OPC concrete
NU
after immersion in 2.826 M NaCl solution. Rebar inside geopolymer concrete retained its passivity after 500 days (Figure 9(b)), while chloride induced corrosion was initiated in rebar embedded in OPC and rust products are evident in Figure 9(d). No rust products formed in either
MA
geopolymer concrete or OPC concrete for immersion in 0.6 M NaCl solution for 500 days (Figure 10). This shows that the corrosion state based on the electrochemical measurements and
D
the corrosion activity observed in the rebar do not have direct relationship. Therefore, the electrochemical measurements can over-estimate the corrosion of embedded rebar in geopolymer
AC
CE
PT E
concrete.
(a)
(b)
22
PT
ACCEPTED MANUSCRIPT
MA
NU
SC
RI
(c) (d) Figure 9: (a) geopolymer concrete with rebar (b) embedded rebar in geopolymer concrete (c) OPC concrete with rebar (d) embedded rebar in OPC concrete immersed in 2.826 M NaCl
(b)
CE
PT E
D
(a)
AC
(c) (d) Figure 10: (a) geopolymer concrete with rebar (b) embedded rebar in geopolymer concrete (c) OPC concrete with rebar (d) embedded rebar in OPC concrete immersed in 0.60 M NaCl 3.6.Corrosion rate of rebar embedded in chloride contaminated geopolymer concrete As observed in the previous section, the chloride ions take a long time to reach the embedded rebar in geopolymer concrete. Therefore, to accelerate the chloride induced corrosion the authors prepared specimens with added chloride. The half-cell potential and corrosion current of embedded rebar was measured using Cu/CuSO4 (CSE) reference electrodes [61].
23
ACCEPTED MANUSCRIPT
Again half-cell potential of embedded rebar in geopolymer concrete is higher negative than that of embedded rebar in OPC concrete (Figure 11(a) and (b). This is an indication of severe corrosion activity in embedded rebar in geopolymer concrete. The corresponding corrosion current for the rebar inside 50/50 FA/Slag geopolymer concrete is lower than the current obtained for rebar in OPC concrete [25]. According to electrochemical measurements, embedded
PT
rebars in both types of concrete expected to have corrosion. Figure 12 shows the corrosion products of the rebar embedded in contaminated concrete. Figure
RI
12(a) and (c) show the corrosion products in rebar embedded in OPC concrete. The rebar embedded in OPC shows significant corrosion activity after 150 days while the rebar embedded
SC
in geopolymer concrete does not show any pitting corrosion products. As observed in previous section, electrochemical measurements obtained for the rebar in contaminated geopolymer
NU
concrete misinterpret the corrosion state.
Slag rich concrete has high content of sulphide species (S2- and S2O32-) in the concrete pore
MA
solution [14, 20, 62]. These sulphide ions can oxidize into sulphur consuming locally available oxygen [14, 62]. The absence of oxygen in the rebar inhibits the cathode reaction in turn the corrosion is controlled. Ma et al. [14] reported higher sulphide ion concentration in the slag
D
based geopolymer concrete pore solution than OPC concrete which leads to reduction in the
PT E
embedded rebar. Hollway et al.[62] concluded that oxidation of sulphide ions inot sulphur
AC
CE
prevents potential rising into the pitting regimes which inhibits the oxygen cathode reaction.
(a) (b) Figure 11: (a) Half-cell potential (b) Corrosion current of rebar in chloride contaminated geopolymer and OPC concrete
24
PT
ACCEPTED MANUSCRIPT
(b)
MA
NU
SC
RI
(a)
(c)
(d)
PT E
D
Figure 12: Corrosion state of rebar embedded in chloride contaminated concrete (a) rebar in OPC closer image (b) rebar in 50/50 FA/Slag geopolymer closer image (c) rebar in OPC (d) rebar in 50/50 FA/Slag geopolymer after 150 days 3.7 XRD patterns of geopolymer
CE
The XRD patterns of the blended fly ash and slag geopolymer binders at two ages i.e. 28 and 270 days are shown in Figure 13. The results show that geopolymer binders containing higher than 50% slag have C-S-H phases indicated by the peak at 29°2θ. C-S-H has a disordered structure
AC
similar to riversideite and do not have 3-dimensional cross linking. When the Fly ash percentage is increased beyond 50% the resulting binders show a peak corresponding to N-(C)-A-S-H type garronite phases and N-A-S-H type zeolite crystalline phases. The high fly ash binders also showed a hump between 25-30°2𝜃 corresponds to the amorphous geopolymer gel. The XRD patterns at 28 days also show the presence of crystalline phases corresponding to unreacted particles from the original fly ash (mullite, quartz, and hematite). However, at 270 days, the height of these peaks decreased indicating continual geopolymerisation reaction. 25
ACCEPTED MANUSCRIPT
Above XRD results shows that geopolymers containing higher slag content produce higher content of calcium bearing phases including disordered calcium silicate hydrate phases. These phases do not have 3-dimensional cross linking like N-A-S-H phases which are produced in high fly ash binders. The disorderly structure in high slag binders can be easily attacked by
NU
SC
RI
PT
chemically.
(b)
AC
CE
PT E
D
MA
(a)
26
ACCEPTED MANUSCRIPT
(c) Figure 13: XRD patterns of geopolymer binders (a) 70/30 FA/S (b) 50/50 FA/S (c) 30/70 FA/S
4. Conclusions On the basis of obtained results, the following concluding remarks can be drawn: Apparent chloride diffusion coefficient of blended fly ash and slag geopolymer concrete is
PT
1.
lower than that of OPC concrete. The diffusion coefficient decreased with the slag content
2.
RI
in the binder.
Blended fly ash and slag geopolymer concrete has higher age factor than in OPC concrete
SC
indicating improved resistance to chloride ingress with time. The lower permeability is attributed to the higher binding capacity in the hydrolacite phases and higher degree of
3.
NU
geopolymerisation in blended fly ash and slag which have been reported in the literature. The time taken to initiate the corrosion in the rebar in blended fly ash and slag geopolymer
MA
concrete is significantly higher than that of OPC concrete. The time is longer when the slag in the geopolymer is higher. 4.
Chloride induced corrosion started in the rebar in OPC concrete exposed to 2.826 M NaCl
D
solution for 500 days while the blended fly ash and slag geopolymer concrete bar did not
PT E
corrode. The chloride content at the rebar in geopolymer is ten times lower than chloride content in OPC concrete. The results are consistent with the calculated initiation time. 5.
Geopolymer concrete containing slag provides greater protection against reinforced
CE
corrosion than that of OPC concrete when concrete is contaminated with chloride. This is due to reduction in oxygen around the rebar in the presence of slag in the binder (which
6.
AC
has been confirmed by the previously reported work). Electrochemical measurements (i.e. Half-cell potential and corrosion current) of the embedded rebar in geopolymer concrete do not correlate well with the corrosion activity.
Acknowledgements The authors gratefully acknowledge the financial support provided by Austroads.
27
ACCEPTED MANUSCRIPT
References [1] Mobili A, Belli A, Giosuè C, Bellezze T, Tittarelli F. Metakaolin and fly ash alkali-activated mortars compared with cementitious mortars at the same strength class. Cement and Concrete
PT
Research. 2016;88:198-210. [2] Chindaprasirt P, Chalee W. Effect of sodium hydroxide concentration on chloride penetration
RI
and steel corrosion of fly ash-based geopolymer concrete under marine site. Construction and
SC
Building Materials. 2014;63:303-10.
NU
[3] Reddy DV, Edouard J-B, Sobhan K. Durability of Fly Ash–Based Geopolymer Structural Concrete in the Marine Environment. Journal of Materials in Civil Engineering. 2013;25(6):781-
MA
7.
[4] Fernando Pacheco-Torgal JL, Leonelli C, Palomo A, Chindaprasit P. Handbook of Alkali-
PT E
D
Activated Cements, Mortars and Concretes. 2015. [5] Provis JL, Van Deventer JSJ. Alkali-activated materials: state of the art report. 2013.
CE
[6] Ghafoori N, Najimi M, Sobhani J, Aqel MA. Predicting rapid chloride permeability of self-
AC
consolidating concrete: A comparative study on statistical and neural network models. Construction and Building Materials. 2013;44:381-90. [7] Silva N. Chloride Induced Corrosion of Reinforcement Steel in Concrete Threshold Values and Ion Distributions at the Concrete-Steel Interface [PhD]: Chalmers University Of Technology, Sweden.; 2013. [8] Ganesan N, Abraham R, Deepa Raj S. Durability characteristics of steel fibre reinforced geopolymer concrete. Construction and Building Materials. 2015;93:471-6. 28
ACCEPTED MANUSCRIPT
[9] ASTM International. ASTM C1556: Standard test method for determining the apparent chloride diffusion coefficient of cementitious mixtures by bulk diffusion. 2011. [10] Olivia M, Nikraz HR. Durability of Fly Ash Geopolymer Concrete in a Seawater
PT
Environment. Concrete conference. Perth, Australia2011. [11] Shayan A, Xu, A., Andrews-Phaedonos, F. Field Performance of Geopolymer Concrete,
SC
Concrete Institute of Australia. Sydney, Australia2013.
RI
Used as a Measure Towards Reducing Carbon Dioxide Emission. 26thBiennial Conference
NU
[12] ASTM International. ASTM C1202: Standard test method for electrical indication of concrete's ability to resist chloride ion penetration. 2012.
MA
[13] Nordtest Method. NT BUILD 443: Concrete, Hardened: Accelerated Chloride Penetration.
D
1995.
PT E
[14] Ma Q, Nanukuttan SV, Basheer PAM, Bai Y, Yang C. Chloride transport and the resulting corrosion of steel bars in alkali activated slag concretes. Materials and Structures. 2015:1-15.
CE
[15] Monticelli C, Natali ME, Balbo A, Chiavari C, Zanotto F, Manzi S, et al. Corrosion behavior of steel in alkali-activated fly ash mortars in the light of their microstructural,
AC
mechanical and chemical characterization. Cement and Concrete Research. 2016;80:60-8. [16] Miranda JM, Fernández-Jiménez A, González JA, Palomo A. Corrosion resistance in activated fly ash mortars. Cement and Concrete Research. 2005;35(6):1210-7. [17] Bastidas DM, Fernández-Jiménez A, Palomo A, González JA. A study on the passive state stability of steel embedded in activated fly ash mortars. Corrosion Science. 2008;50(4):1058-65.
29
ACCEPTED MANUSCRIPT
[18] Bencardino F, Condello A. Innovative solution to retrofit RC members: InhibitingRepairing-Strengthening (IRS). Construction and Building Materials. 2016;117:171-81. [19] Bencardino F, Condello A. Eco-friendly external strengthening system for existing
PT
reinforced concrete beams. Composites Part B: Engineering. 2016;93:163-73. [20] Glasser FP, Luke K, Angus MJ. Modification of cement pore fluid compositions by
RI
pozzolanic additives. Cement and Concrete Research. 1988;18(2):165-78.
SC
[21] RILEM TC 178-TMC. Testing and modelling chloride penetration in concrete analysis of
NU
total chloride content in concrete recommendation. Materials and Structures. 2002;35:583-5. [22] Australian Standards. AS 1012.3.1:Determination of properties related to the consistency of
MA
concrete—Slump test. 2014.
D
[23] Aimin X, Ahmad S. Relationship between Reinforcing Bar Corrosion and Concrete
PT E
Cracking. Materials Journal.113(01).
[24] ASTM International. ASTM C896: Standard Test Method for Half-Cell Potentials of
CE
Uncoated Reinforcing Steel in Concrete. 1999.
AC
[25] Kupwade-Patil K, Allouche EN. Examination of Chloride-Induced Corrosion in Reinforced Geopolymer Concretes. Journal of Materials in Civil Engineering. 2013;25(10):1465-76. [26] Lizarazo-Marriaga J, Claisse,Peter., . Determination of the concrete chloride diffusion coefficient based on an electrochemical test and an optimization model. Materials, chemistry and physics. 2009;113(3):536-43.
30
ACCEPTED MANUSCRIPT
[27] Ismail I, Bernal SA, Provis JL, San Nicolas R, Brice DG, Kilcullen AR, et al. Influence of fly ash on the water and chloride permeability of alkali-activated slag mortars and concretes. Construction and Building Materials. 2013;48:1187-201. [28] Kumar S, Kumar R, Mehrotra SP. Influence of granulated blast furnace slag on the reaction,
PT
structure and properties of fly ash based geopolymer. Journal of Materials Science.
RI
2009;45(3):607-15.
SC
[29] Puertas F, Fernández-Jiménez A. Mineralogical and microstructural characterisation of
NU
alkali-activated fly ash/slag pastes. Cement and Concrete Composites. 2003;25(3):287-92. [30] Kayali O, Khan MSH, Sharfuddin Ahmed M. The role of hydrotalcite in chloride binding
MA
and corrosion protection in concretes with ground granulated blast furnace slag. Cement and Concrete Composites. 2012;34(8):936-45.
D
[31] Chen W, Brouwers HJH. A method for predicting the alkali concentrations in pore solution
PT E
of hydrated slag cement paste. Journal of Materials Science. 2011;46(10):3622-31.
CE
[32] Evans DG, Slade RCT. Structural aspects of layered double hydroxides. Layered double hydroxides2006. p. 1-87.
AC
[33] Narayanan S, Krishna K. Structure activity relationship in Pd/hydrotalcite: effect of calcination of hydrotalcite on palladium dispersion and phenol hydrogenation. Catalysis Today. 1999;49(1–3):57-63. [34] Provis JL, Myers RJ, White CE, Rose V, Van Deventer JSJ. X-ray microtomography shows pore structure and tortuosity in alkali-activated binders. Cement and Concrete Research. 2012;42(6):855-64.
31
ACCEPTED MANUSCRIPT
[35] Lloyd RR, Provis JL, Smeaton KJ, Van Deventer JSJ. Spatial distribution of pores in fly ash-based inorganic polymer gels visualised by Wood’s metal intrusion. Microporous and Mesoporous Materials. 2009;126(1–2):32-9. [36] Yang T, Yao X, Zhang Z. Quantification of chloride diffusion in fly ash–slag-based
PT
geopolymers by X-ray fluorescence (XRF). Construction and Building Materials. 2014;69:109-
RI
15.
SC
[37] Li Z, Liu, Sifeng. Influence of Slag as Additive on Compressive Strength of Fly Ash-Based
NU
Geopolymer. Journal of Materials in Civil Engineering. 2007;19(6):470-4. [38] Bernal SA, Provis JL, Fernández-Jiménez A, Krivenko PV, Kavalerova E, Palacios M, et
MA
al. Binder Chemistry – High-CalciumAlkali- Activated Materials Alkali activated materials2013. p. 59-77.
D
[39] Florea MVA, Brouwers HJH. Chloride binding related to hydration products: Part I:
PT E
Ordinary Portland Cement. Cement and Concrete Research. 2012;42(2):282-90.
CE
[40] Odler I. Hydration, Setting and Hardening of Portland Cement. Lea's Chemistry of Cement and Concrete1998. p. 241-91.
450.
AC
[41] Neville AM. Durability of concrete. Properties of Concrete. 5th edition ed1996. p. 1350-
[42] Gjorv OE. Durability analysis.
Durability design of concrete structure in severe
enviournments2014. [43] Roy DM, Idorn GM. Concrete Microstructure. National Research Council, Washington, DC: Strategic Highway Research Program,; 1993. p. 50-115. 32
ACCEPTED MANUSCRIPT
[44] Attari A, McNally, C., Richardson, M.G.,. Probability-based modeling of corrosion initiation times using mix-dependent age factors.
Safety, Reliability, Risk and Life-Cycle
Performance of Structures and Infrastructures2013. p. 2935-345. [45] Xu H, Provis, John L., Van Deventer, J.S.J., Krivenko, P.V. Characterization of aged slag
PT
concretes. ACI Materials Journal. 2008;105(2):131-9.
RI
[46] Ismail I, Bernal SA, Provis JL, San Nicolas R, Hamdan S, Van Deventer JSJ. Modification
SC
of phase evolution in alkali-activated blast furnace slag by the incorporation of fly ash. Cement
NU
and Concrete Composites. 2014;45:125-35.
[47] Ismail I, Bernal SA, Provis JL, San Nicolas R, Brice DG, Kilcullen AR, et al. Influence of
MA
fly ash on the wate and chloride permeability of alkali-activated slag mortars and concretes. Construction and Building Materials. 2013;48:1187-201.
D
[48] Song L, Sun W, Gao J. Time dependent chloride diffusion coefficient in concrete. Journal of
PT E
Wuhan University of Technology-Mater Sci Ed. 2013;28(2):314-9.
CE
[49] Cement concrete and aggregates Australia. Chloride resistance of concrete. 2009. p. 30-5. [50] Buenfeld NR, Newman JB. Examination of three methods for studying ion diffusion in
AC
cement pastes, mortars and concrete. Materials and Structures.20(1):3-10. [51] Luping T, Nilsson, L-O. Chloride diffusivity in high strength concrete as different ages. Nordic concrete research. 1992;10:162-71. [52] Costa A, Appleton J. Chloride penetration into concrete in marine environment-Part II: Prediction of long term chloride penetration. Materials and Structures.32(5):354-9.
33
ACCEPTED MANUSCRIPT
[53] Xu A, Shayan, A. Determination of Chloride Diffusion Coefficient of Concrete: Comparison of Bulk Diffusion and Electrical Field Method. 27th Biennial National Conference of the Concrete Institute of Australia in conjunction with the 69th RILEM Week. Melbourne, Australia2015.
PT
[54] Hou H, Tian L. The time-dependence of chloride penetration resistance of concrete by the
RI
rapid chloride permeability test Applied mechanics and materials. 2013;357:651-4.
SC
[55] Roy DM, Jiang, W. and Silsbee, M.R. Chloride diffusion in ordinary, blended, and alkaliactivated cement pastes and its relation to other properties. Cement and Concrete Research.
NU
2000;30(12):1879-84.
MA
[56] Shi C. Strength, pore structure and permeability of alkali-activated slag mortars. Cement and Concrete Research. 1996;26(12):1789-99.
D
[57] Luping T, Gulikers J. On the mathematics of time-dependent apparent chloride diffusion
PT E
coefficient in concrete. Cement and Concrete Research. 2007;37(4):589-95.
CE
[58] Elfmarkova V, Spiesz P, Brouwers HJH. Determination of the chloride diffusion coefficient in blended cement mortars. Cement and Concrete Research. 2015;78, Part B:190-9.
AC
[59] Meck E, Sirivivatnanon V. Field indicator of chloride penetration depth. Cement and Concrete Research. 2003;33(8):1113-7. [60] Yeau KY, Kim EK. An experimental study on corrosion resistance of concrete with ground granulate blast-furnace slag. Cement and Concrete Research. 2005;35(7):1391-9. [61] Bushman JB. Calculation of corrosion rate from corrosion current (Faraday's Law). 2000. p. 1-4. 34
ACCEPTED MANUSCRIPT
[62] Holloway M, Sykes JM. Studies of the corrosion of mild steel in alkali-activated slag cement mortars with sodium chloride admixtures by a galvanostatic pulse method. Corrosion
AC
CE
PT E
D
MA
NU
SC
RI
PT
Science. 2005;47(12):3097-110.
35
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
NU
SC
RI
PT
Graphical Abstract
36