EFFECT OF MIX DESIGN ON CONSISTENCE AND SETTING

TIME OF ALKALI ACTIVATED CONCRETE

Kofi Abora1, Keith Quillin2, Kevin A Paine3, Andrew M Dunster4

1,3
BRE Centre for Innovative Construction Materials, University of Bath
2,4
Building Research Establishment, UK

Abstract: In order for alkali activated systems to become technically and economically
viable construction materials, a number of technical and economic barriers need to be
addressed. As with Portland cement concretes, issues related to the mix design, curing
and use of alkali activated concrete have to be clearly understood; inappropriate mix
designs will lead to products that are not fit for purpose. This paper reviews two form of
sodium silicate based solution (water glass) at different concentration to assess its effects
on consistence, one major barrier in the use of alkali activated system. Achieving better
control of these parameters whilst retaining appropriate compressive strength
development will help move these promising low carbon dioxide (CO2) binders towards
mainstream use. The results obtained have demonstrated that the change in
concentration of sodium oxide (Na2O) within the activator has an effect on the volume of
activator required for a consistence mix. Additionally the results give indication of the
optimum activator binder ratio and the compressive strength, when compositional design
of the activator is varied. Although the concretes made with the lower concentrated
activator had the highest level of consistence, it could only achieve less than 25% of the
strength value of the concretes made with the higher concentrated activator. Thus the
type of activator to be used will depend on the application of the concrete and the level of
strength required.

Keywords: alkali activated concrete, strength, consistence, fly ash, ggbs

1. Introduction
In a bid to reduce CO2 emissions associated with the use of concrete, the cement
industry is providing incremental process improvements and the use of additions such as
fly ash and ground granulated blastfurnace slag (ggbs) in blended cements and
combinations is having some effect. However, significant savings from decarbonation can
only be made by changing the composition of the cement. This is due to the fact that, in
addition to emissions arising from intensive energy use, CO2 is generated from the
calcinations of the calcium carbonate; one of the main components of Portland cement
manufacture. Global data from the European Cement Association (Cembureau) on CO2
emissions from cement manufacture estimates an average of 0.8 tonnes of CO2 per tonne
of Portland cement produced.
Current research on alkali activated technologies indicates a more enhanced
environmental impact than Portland cement. This is due to the fact that an alkali activated

1
Researcher, k.abora@bath.ac.uk
2
Associate Director, quillink@bre.co.uk
3
Senior Lecturer, k.paine@bath.ac.uk
4
Principal Consultant, dunstera@bre.co.uk

1
system utilises industrial by-products, providing massive savings on energy consumption
and CO2 emissions.
In order for alkali activated systems to become technically and economically viable
construction materials, a number of technical and economic barriers need to be
addressed. As with Portland cement concretes, the key performance requirements from
any binders used in concrete are that they provide good consistence of the wet concrete
and an appropriate rate of strength development.
Actual formulations, hardening characteristics, strength and durability are dependent
on the composition of the materials used, the relative proportions, the formulation of the
alkali activator and the curing conditions; however the exact reaction mechanism leading
to the setting of the alkali activated binders is still being investigated.

1.1. Alkali activated technology

Alkali activated technology involves the use of various raw materials and industrial by-
products activated with a high pH alkali metal solution (generally sodium silicate-based)
and cured at room or higher temperatures. The source materials predominantly used are
fly ash, ggbs, metakaolin, and sewage treatment and paper processing residues. The high
alkali environment breaks down Si-O chemical bonds in the glassy phase of the material.
These phases then become available for reaction with the alkaline species in the liquid.
Reaction products are zeolitic in nature and the activator is consumed in the reaction. The
chemistry is dependent on the composition of the source material and the type of activator
used.
Basic research on alkali activated materials has been carried out at a number of
institutions, since the 1940’s (Roy 1998). Most work has been done in Eastern Europe,
Australia, New Zealand, Ukraine, France, Finland and the former Soviet Union. However,
comprehensive work was carried out at the Building Research Establishment, UK in the
1970’s. Smith and Osborne (1977) reported that by activating a combination of fly ash with
either ggbs or specially prepared synthetic granules with sodium hydroxide, there is the
potential of producing cement. However, at that time, the use of fly ash and ggbs as
partial Portland cement replacements was seen as being of more immediate benefit, in
terms of reducing the CO2 emissions associated with Portland cement production.
However, building construction using alkali activated slag cements in Russia, Ukraine and
Poland has been reported (Shi et al, 2006).

1.2. Alkali activated technology and CO2

The manufacturing of cement produces a more significant source of CO2 than any
industrial process other than electrical power generation and can be associated with two or
three contributions:

Č{total} → Č{raw materials} + Č{fuels} + Č{electric power} (1)

The most contributing factor of the CO2 emissions during manufacturing of Portland
cement is as a result of the electricity and energy used in firing the kiln to high
temperatures in the region of 1450°C, and the calcination of limestone and silica. The
chemical equation representing the above scenario can be expressed as:

5CaCO3 + 2SiO2 → (3CaO.SiO2 )(2CaO.SiO2) + 5CO2 (2)

Where energy consumption equates to about 0.4 tonnes of CO2 per tonne of clinker
(Quillin, 2007).

2
 When industrial by-products such as fly ash and ggbs are used as the binders for alkali
activated technologies, the major source of the CO2 arises from the production of the
sodium silicate-based chemical activator through thermal processing and decarbonation of
sodium carbonate. Information from silicate manufacturers indicates that the CO2 emission
should be less than 0.2 tonnes per tonne of activator.
Whilst fly ash may be activated alone, its use in combination with ggbs allows for the
curing to be done at room temperature, and therefore there is no need for elevated
temperatures which would increase the level of CO2 emissions further.

2. Experimental Investigation
The consistence and compressive strength characteristics of alkali activated concrete
using two different concentrations of alkali activators were investigated in this project. The
specimens for the compressive strength test were 100 mm cubes and these were seal
cured in air at room temperature. A flow table test was carried out on the alkali activated
concrete made from different activators at different volumes.
The same ggbs, fly ash, alkali activators, fine and coarse aggregates were used
throughout the investigation.

2.1. Cementitious materials

The chemical and physical properties of UK sourced fly ash and ggbs, conforming to
BS EN 450 and BS 6699 respectively, are given in Tables 1 and 2 below.

2.2. Alkali activators

Two forms of alkali activators with the same molecular ratio SiO2:Na2O (Ms) = 1.54
were used. The first was a concentrated sodium silicate solution (A1) with 20% SiO2 and
13% Na2O and a density of 1420 kg/m3. The second activator was a less concentrated
sodium silicate solution (A2) with 12.3% SiO2 and 8% Na2O and a density of 1320 kg/m3.

2.3. Mixture proportions

A series of alkali activated concrete using fly ash and ggbs were made and tested for
consistence and compressive strength. The compositions given in Table 3 were studied:
The mix design used in each case was as follows:
Total binder content: 440 kg/m3
Total activator volume: 170 – 205 litres per m3 of concrete
Total aggregate: binder ratio: 3.5 : 1
10/20 mm Thames valley aggregate: 35%
4/10 mm Thames valley aggregate: 25%
0/4 mm Thames valley aggregate: 40%

Each mix of about 7 litres was made in a pan mixer to produce six cubes to allow for a
7, 28 and 91 days compressive strength testing. The aggregates and the binders were
mixed for a few seconds to ensure all the materials were mixed thoroughly. The activator
was then added and mixing continued for about 3.5 minutes. The concrete was then
subjected to a flow table test before casting into a 100 mm cube moulds that using a
vibration table.
The samples were subject to a standard curing regime involving 24 hours in moulds
under damped Hessian, prior to demoulding, being wrapped in polythene and stored in a
controlled room maintained at room temperature until the test date.

3
 Figure 1: Fresh alkali activated concrete mix

Table 1: Typical chemical and physical properties of ggbs

Oxide % Physical properties
SiO2 35.3 Specific surface: 450 m2/kg
Al2O3 11.3 Specific gravity: 2.91
Fe2O3 0.2 Colour: off white
CaO 40.0
MgO 10.6
SO3 4.0
Na2O 0.3
K2O 0.4

Table 2: Typical chemical properties of fly ash

Oxide %
SiO2 44.0
Al2O3 24.6
Fe2O3 11.7
CaO 5.5
MgO 2.1
SO3 0.2
Na2O 1.1
K2O 2.4

Table 3: Concrete mix proportions

Volume of activator Activator : binder
Activator
MIX ID Fly ash Ggbs per m3 of concrete ratio
type
(litres)
MIX 1 90% 10% 205 0.66
MIX 2 90% 10% A1 186 0.60
MIX 3 90% 10% 170 0.55
MIX 4 90% 10% 205 0.61
MIX 5 90% 10% A2 186 0.55
MIX 6 90% 10% 170 0.50

4
2.4. Flow table test
The consistence of the concrete was investigated using a standard flow table in
accordance with the method specified for determining the flow of fresh concrete in BS EN
12350-5:1999. The measurement of spread was as shown in Figure 2.

Figure 2: Spread measurement of a fresh alkali activated concrete

2.5. Compressive strength

100 mm cubes specimens (Figure 3) were prepared to determine the effect of varying
the composition of the activator as well as the change in volume of activator on the
compressive strength. Tests in accordance with BS EN 12390-3:2009 were carried out at
7, 28 and 91 days.

Figure 3: 100 mm cube specimen of alkali activated concrete

3. Results and Discussion

3.1. Consistence of alkali activated concrete

Initial and final setting times of alkali activated concrete are difficult to determine,
because unlike Portland cement the consistence changes rapidly over a short period of
time (Atiş et al., 2009), making the fresh concrete difficult to place. Initial trials using the
slump test was used to measure the consistence of alkali activated concrete. However,
this was quickly identified as unsuitable because most of the mixes flowed freely under
their own weight and therefore there was no correlation between slump and the degree of
consistence. To ensure comparison of the consistence between the various mixes, a flow
table test was therefore used.

5
 It was observed that the concrete made with the less concentrated activator solution
exhibited the highest consistence value, largest flow value, when comparing the volume of
activator used per mix (Figure 4). It can also be observed that the higher the volume of
activator, the higher the consistence value.
The ability to control and increase the setting time of alkali activated concrete needs a
more detailed study and therefore lies outwith the scope of this paper.

Table 4 : Measurement of spread from the flow table test

Volume of activator per m3 of Measurement of spread (mm)
concrete (litres) Activator A1 MIX ID Activator A2 MIX ID
205 565 1 630 4
186 485 2 510 5
3
170 415 415 6

activator A1 activator A2

650

600
Measurement of spread (mm)

550

500

450

400
165.00 175.00 185.00 195.00 205.00
Volum e of
Figure 4: Relationship between theactivator/m of concrete
activator type (litres)
and consistence
3

3.2. Compressive strength development

The compressive strengths of the concrete are shown in Table 5. The compressive
strength of the alkali activated concrete was greater with the more concentrated activator
solution, A1. This is attributed to the high concentration of the Na2O in the activator, which
has been reported by previous authors using various ranges of activator concentration
(Atiş et al.). The results presented in Figure 5 indicate that the analysis is valid for each of
the curing ages of the mixes made with the A1 solution (MIX 1, 2 and 3).
Furthermore, the increase in compressive strength correlated to the decrease in
volume of activator used as shown in Figure 6 and Figure 7. Analysis of the concretes
made with solution A1 and A2 indicated that overall the highest compressive strength was
obtained when the lowest volume of activator was used. For the alkali activated concretes
at a constant concentration, as shown in Figure 5, the highest compressive strengths were

6
associated with the activator with the highest sodium concentration, A1. Among the six
concrete mixes, MIX 3 exhibited the highest compressive strength at all testing ages.
Table 5: Compressive strength test results at 7, 28 & 91 days
Volume of activator Compressive strength (N/mm2)
per m3 of concrete A1 A2
(litres) 7
7 days 28 days 91 days 28 days 91 days
days
205 14.0 27.5 46.5 3.3 4.4 5.8
186 15.5 30.5 49.5 2.8 4.5 5.8
170 17.5 33.8 50.3 4.0 5.3 6.8

activator A1 activator A2

60

50
)
2
compressive strength (N/mm

40

30

20

10

0
7 28 91
Test age (days)
Figure 5: Relationship between strength development and type of activator used

7 days 28 days activator:binder ratio

35 70%
)
2

30
Compressive strength (N/mm

65%
activator : binder ratio

25

20
60%
15

10
55%

0 50%
170.00 186.00 205.00

Figure 6: RelationshipVolum e of strength

activator/m 3
between of concrete
development (litres)
and volume of activator (A1)

7
 7 days 28 days activator:binder ratio

65%

2)
compressive strength (N/mm

activator : binder ratio

4
60%

2
55%

0 50%
170.00 186.00 205.00
Figure 7: RelationshipVolum
between
e of strength development
activator/m and
3 of concrete volume of activator (A2)
(litres)

4. Conclusion
The overall analysis has demonstrated that when the compositional design of the
activator is varied, in this instance the change in concentration of NaOH within the
activator, there is an effect on the volume of activator required for a mix of a given
consistence and compressive strength. The properties of the alkali activated concrete are
dependent on the type of activator used. The following observations were made during
the investigations:

• The results identified from the research indicates that fly ash in the presence of the
ggbs, which provides calcium to the binders leads to higher strength and quicker
hardening of the concrete within 24 hrs.
• The more concentrated the NaOH within the activator solution, the higher
compressive strength. This indicates that the strength of the concrete is influenced
by the concentration of NaOH.
• The analysis indicates that the lower the volume of activator for the mix, the higher
the compressive strength.
• The consistence value of the alkali activated concrete increased with the use of
lower concentrated activator solutions when compared to more concentrated
activators based on the same volume of activator used.
• Observation of the concrete over time indicates that the consistence reduces much
more quickly with an increase in the sodium concentration of the activator.
• A slump test could not be used for determining the consistence of the concrete.
This is due to the fact that alkali activated concrete fell freely under its own weight
during the slump test.

In summary although the concretes made with the lower concentrated activator had the
highest level of consistence, it could only achieve 25% of the strength value of the
concretes made with the higher concentrated activator. Thus the type of activator to be
used will depend on the application of the concrete and the level of strength required.

8
5. Acknowledgements
The authors wish to express their gratitude and sincere appreciation to the laboratory
technicians at the Building Research Establishment especially Les White and Clive Tipple,
whom without their assistance this paper would not have been possible.

6. References
ATIŞ C. D., BILIM C., ÇELIK Ö AND KARAHAN O., 2009. Influence of activator on the
strength and dry shrinkage of alkali-activated slag mortar. Construction and building
materials, 23(2009), pp.548 – 555.
EUROPEAN CEMENT ASSOCIATION (Cembureau), Energy is a sensitive factor in
cement manufacture, www.cembureau.be, 2000.
QUILLIN K, 2007. Calcium sulfoaluminate cements. CO2 reduction, concrete properties
and applications. Bracknell: IHS BRE Press.
ROY D. M., 1998. Alkali-activated cements, opportunities and challenges. Cement and
concrete research, 29(1999), pp.249-254.
SHI C., KRIVENKO P. V AND ROY D., 2006. Alkali-activated cements and concrete.
Oxon: Taylor & Francis Ltd, pp.298-319.
SMITH M.A. AND OSBORNE G. J., 1977. Slag/fly ash cements. World Cement
Technology, Nov\Dec, pp.223-233.