You are on page 1of 14

Magazine of Concrete Research Magazine of Concrete Research, 2012, 64(6), 513–525

Volume 64 Issue 6 http://dx.doi.org/10.1680/macr.11.00026


Paper 1100026
Use of recycled and secondary aggregates in Received 10/02/2011; revised 22/06/2011; accepted 03/08/2011
foamed concretes Published online ahead of print 29/03/2012

Jones, Zheng, Yerramala and Rao Thomas Telford Ltd & 2012

Use of recycled and secondary


aggregates in foamed concretes
Roderick Jones Amarnath Yerramala
University of Dundee, Dundee, UK University of Dundee, Dundee, UK
Li Zheng Kharidu Srinivasa Rao
University of Dundee, Dundee, UK University of Dundee, Dundee, UK

This paper describes research into improved sustainable construction through the development of foamed concrete
incorporating recycled and secondary aggregate fines (RSA). These were selected on the basis of being broadly
available and/or significant local production but being otherwise difficult to utilise in structural fill or conventional
normal weight concrete. The results indicate that the fine RSA can be successfully used to replace natural sand, either
partially or fully. Data presented cover mix design, consistence and strength development. With respect to the
results, significant advantage was obtained with the pozzolanic contribution of the conditioned fly ash. In addition,
the performance of RSA foamed concretes in terms of drying shrinkage, water permeation, sulfate attack and
thermal conductivity values are given. The data show that behaviour is mainly dependent on particle density and
given RSA heterogeneity. It is necessary to trial mix unknown materials to enumerate foamed concrete performance.

Introduction and precast concretes. As a result, the uptake of such materials


In recent years, the use of recycled and secondary aggregates has been limited. In response to this, the current paper reports a
(RSA) has become increasingly important to sustainable construc- laboratory study that examined the fresh, engineering and
tion by reducing demand for primary aggregates (WRAP, 2006). durability properties of the foamed concrete incorporating a
RSA covers a wide range and there are at least 18 materials that broad range of RSA, where natural sand was replaced with RSA
can be included in the family of RSA in the UK (WRAP, 2010). in two levels at 50% and 100%. The work covered foamed
Although coarse aggregate-sized RSA are successfully used to concretes with a wider range of plastic densities from 600 to
replace primary aggregates to produce normal weight concrete 1600 kg/m3 with the main study focused on that of 1000 kg/m3
(e.g. Andrade et al., 2007; Dhir et al., 2004; Filipponi et al., with free water/cement ratio (w/c) of 0.50, which is a common
2003; Khatib, 2005; Poon et al., 2004), the silt/sand-sized phase mix used by industry.
is problematic owing to its high water absorption and particle
shape (Jones et al., 2009). Foamed concrete has become an
Experimental details
increasingly popular material for diverse construction applica-
tions, making important contributions to sustainable construction Materials
by improving the thermal insulating properties of the building From the 18 RSA listed by WRAP (2010), seven different
elements (Jones and Giannakou, 2002). Some notable examples materials and sources were selected based on broad geographic
of the UK foamed concrete such as the 600 000 m3 Combe and quantity availability: that is, demolition fines (DF), incinera-
Downs stone mine stabilisation project, which used a natural sand tor bottom ash aggregate (IBA), glass fines (GF), foundry sand
filler (Dhir et al., 2005), bridge infills, pipework protection and (FS), china clay sand (CCS), conditioned fly ash (CFA) and
airport applications and so on (Propump, 2011). During the early crumb rubber (CR). All of these were obtained from sources
studies, Jones et al. (2005) identified that fine RSA materials across the UK, as summarised in Table 1 and Table 2. The
could effectively be used in foamed concrete to replace primary control mix used a natural sand conforming to BS EN 12620
aggregates owing to their fresh properties being dominated by the (BSI, 2002) Table 2. The cement used in this study was Portland
foam/paste phases, rather than the aggregate. As noted in the cement (PC), conforming to BS EN 197-1 (BSI, 2011), and a
Magcon project (Nixon, 2004), this makes foamed concrete an commercially available synthetic surfactant was used for foam
environmentally friendly fill material compared to quarry-sourced production.
products.
Particle density and water absorption tests were carried out in
Fine RSA is a highly heterogenous material group (Jones et al., accordance with BS EN 1097-6 (BSI, 2000) to enable foamed
2009), both physically and compositionally, and moreover, there concrete mix proportioning (Jones et al., 2005). It can be seen
are within-source variabilities that make its use in conventional that the densities of the inorganic RSA materials ranged from
concrete inefficient for large-scale producers, such as ready-mix 1.95 g/cm3 to 2.61 g/cm3 , which were less than that of reference

513
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

RSA material Mean Denoted Source and general nature


particle
size: m

Recycled aggregate Demolition fines/silt 0.8 DF Processed from construction, demolition and excavation
waste materials.
Secondary aggregate Incinerator bottom 2.5 IBA By-product from the combustion process in municipal solid
ash* waste incineration facilities.
Glass fines 0.5 GF Recycled from building flat glass.
Foundry sand 0.18 FS Sand which was used as a moulding material in the metal
casting industry.
China Clay sand 0.6 CCS Fines recovered from the processing of china clay waste.
Conditioned fly ash 0.02 CFA Fly ash is a by-product from thermal power station. CFA is
fly ash ‘conditioned’ by adding a controlled amount of
water to prevent dust blow problems in landfill.
Crumb rubber 1.0 CR Recycled from shredded truck tyres.

* Note: Material can be problematic if it contains metallic aluminium in which case it should not be used indoors (HSE, 2009).

Table 1. Summary of RSA test materials

Natural DF IBA GF FS CCS CFA CR


sand*

Particle density in SSD state: g/cm3 2.63 2.18 1.95 2.36 2.08 2.61 2.04 1.07
Water absorption, % by mass 0.50 8.83 10.97 0.01 6.03 0.58 –y 0.75

Particle size distribution, % passing by mass

8.0 mm 100 100 100 100 100 100 –y 100


4.0 mm 96 100 100 93 100 97 100
2.0 mm 85 92 75 73.7 90 77 100
1.0 mm 74 67 50 55.5 76 57 44
0.5 mm 58 39 34 43.5 67 34 3
0.25 mm 24 18 28 18 47 12 0
0.063 mm 2 5 2 4.5 4 1 0

* Conforming to BS EN 12620 (BSI, 2002)


† Not applicable

Table 2. Physical properties of natural sand and RSA materials

natural sand; therefore, increased volumes and less foam were have to be ‘wetted’ if in the dry form, but this was not a problem
required to achieve the specified plastic density, as summarised in with the CFA, as its moist content was around 20% by mass, as
Table 3. The measured density of crumb rubber was 1.07 g/cm3 , received.
around half that of the inorganic RSA materials. Three RSA
materials – DF, IBA and FS – had high water absorption, ranging The particle size distributions of the fine aggregates were meas-
from 6.0% to 11.0%, which had to be allowed for in the total ured by sieve analysis, to BS EN 933-1 (BSI, 1997), and are
water added to the base mix. The water absorption of the given in Table 2. It can be seen that most of the RSA fines have a
remaining RSAs was less than 1% and close to that of the similar particle size to natural sand and are within the 0/4 (MP)
reference natural sand. The high surface area of fly ash would limits, that is medium sand range to BS EN 12620 (BSI, 2002),

514
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

Mix Mix constituent proportions: kg/m3 Air: % Vol.

Sand RSA Water

D ¼ 600 kg/m3 , w/c ¼ 0.50, PC ¼ 300 kg/m3

Reference sand 150 0 150 70


100% RSA* 0 150 150 68–70y
50%/100% CFA{ 60/0 60/100 180/200 67/66

PC ¼ 400 kg/m3

‘‘Air’’ 0 0 200 67

D ¼ 1000 kg/m3 , w/c ¼ 0.50 (except CR which w/c ¼ 0.60), PC ¼ 300 kg/m3

Reference sand 550 0 150 55


50%/100% RSA* 275/0 275/550 150 47–55y
50%/100% CFA{ 220/0 220/367 260/333 45/39
50%/100% CR 260/0 260/520 180/180 38/24

PC ¼ 400 kg/m3

Reference sand 400 0 200 52


50%/100% RSA* 200/0 200/400 200 47–52y
50%/100% CFA{ 160/0 160/267 280/333 45/41
50%/100% CR 180/0 180/360 240 40/30

D ¼ 1400 kg/m3 , w/c ¼ 0.50, PC ¼ 300 kg/m3

Reference sand 950 0 150 39


100% RSA* 0 950 150 27–39y
50%/100% CFA{ 380/0 380/633 340/467 23/13

PC ¼ 400 kg/m3

Reference sand 800 0 200 37


100% RSA* 0 800 200 26–37y
50%/100% CFA{ 320/0 320/533 360/467 23/14

D ¼ 1600 kg/m3 , w/c ¼ 0.50, PC ¼ 400 kg/m3

Reference sand 1000 0 200 29


100% RSA* 0 1000 200 16–29y
50%/100 CFA{ 400/0 400/677 400/533 12/1

* Excluding CFA and crumb rubber mixes.


† Air percent varied depending on relative density of RSA used in the mix.
‡ CFA is also considered in w/c calculations.

Table 3. Mix proportions of natural sand and RSA foamed


concretes

515
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

while FS fell in finest side and CCS coarsest side. CR had a very (a) consistence, as measured by slump flow
narrow range of particle size distribution with more than 95% of (b) seal-cured cube strength
particles between 0.5–2.0 mm. (c) drying shrinkage
(d ) water permeability
Experimental programme (e) resistant to sulfate attack
The mix proportioning method used in the study was developed ( f ) thermal conductivity.
at the University of Dundee (Giannakou and Jones, 2002). Details
of the mix proportions of the foamed concrete studied are given These were considered to be the key characteristics for the most
in Table 3. The two ‘extreme’ mixes were common uses of foamed concrete.

(a) ‘air’ mix with target density ¼ 600 kg/m3 and PC ¼ 400 kg/m3 , Results and discussions
to which no fine aggregate was required
Consistence
(b) 100% CFA mix with target density ¼ 1600 kg/m3 and
Effect of RSA type and PC content
PC ¼ 400 kg/m3 , to which no additional foam was required,
The slump flow values of the 1000 kg/m3 sand and RSA foamed
that is, it was just the base mix.
concrete are shown in Figure 1. As can be seen in Figure 1
except CR mixes, all foamed concretes had slump flow larger
The tests were carried out at two levels of RSA replacement, that than 500 mm, which indicates a good self-flowing ability (Jones
is, 50% and 100%, with two levels of PC content, that is, et al., 2005).
300kg/m3 and 400 kg/m3 , and four levels of plastic density, that
is, 600 kg/m3 , 1000 kg/m3 , 1400 kg/m3 and 1600 kg/m3 : The w/c Comparing the slump flow values of the different RSA foamed
ratio for all of the mixes was maintained at 0.50, except that for concretes, when PC content was 300 kg/m3 , the CCS mixes and
CR mixes, which was increased to 0.60 to maintain flowing and 50% GF mix exhibited the similar slump flow to the reference
self-compacting behaviour. sand mix. CFA mixes, in which CFA is taken into account in w/c
calculation, had a higher slump flow than the sand mix, while DF,
The effects of using RSA on the properties of foamed concretes IBA, 100% GF and FS mixes had slightly lower slump flow but
were evaluated in terms of their still in the designed range from 500 to 700 mm. The reduced

900

800 Slump flow spread of reference sand mix

700
Slump flow spread: mm

600

500

400
Considering self-flowing and levelling

300
Cement content: Plastic density ⫽ 1000 kg/m3
3
200 PC ⫽ 300 kg/m w/c ⫽ 0·50*
3
PC ⫽ 400 kg/m *w/c ⫽ 0·60 for CR aggregate mixes

100
Sand 50% 100% 50% 100% 50% 100% 50% 100% 50% 100% 50% 100% 50% 100%
DF DF IBA IBA GF GF FS FS CCS CCS CFA CFA CR CR
Aggregate type: % mass

Figure 1. Slump spread of foamed concretes with different types


and levels of RSA

516
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

workability of these mixes may be attributed to the more angular concrete mixes with increasing foamed concrete density from
particle shapes of the RSAs. In contrast, when PC content was 600 kg/m3 to 1000 kg/m3 : The increase rates were different for
increased to 400 kg/m3 , slump flow was generally increased, different RSA, varied from the highest with 100% CFA to similar
owing to the reduced RSA content. with DF. Further increase in density exhibited two different
trends: while there was a continued increase in slump flow with
Owing to its very low density and hence high content, it was CFA mixes, it was decreased with the DF, IBA, CCS and sand
found that for the CR mixes, there was insufficient free water at mixes.
0.5 w/c ratio. To overcome this, the w/c ratio was increased to
0.6. For 50% CR mix with 300 kg/m3 PC content, although there When PC content was increased to 400 kg/m3 (Figure 2(b)), no
was no slump flow for its base-mix, normal foamed concrete aggregate was required in 600 kg/m3 density mix, that is, the ‘air’
workability was observed with the addition of foam. At 100%CR mix (Table 3), and the slump flow measured was around 600 mm.
mix and 300 kg/m3 PC content there was insufficient foam to When the density was increased to 1000 kg/m3 , slump flow
make it flow and no further testing was carried out. Although increased with all foamed concrete mixes, with the highest
increased workability was obtained with the increase in w/c for observed for 100% CFA mix. However, increasing the density
all CR mixes, it was still lower (slump flow , 500 mm) compar- further resulted in a decrease in slump flow for all mixes.
ing to the other foamed concrete mixes owing to the low foam
content. However, self-flowing ability was also observed from The results suggest that, for a particular aggregate, foamed
these mixes and the slump flow observed was within the self- concrete has an optimal density with respect to workability. In
flowing range (.400 mm) given by Jones et al. (2005), except case of 400 kg/m3 PC content and 0.5 w/c ratio, a density of
the 50% CR mix with 300 kg/m3 PC. around 1000 kg/m3 gave the highest flow while for 300 kg/m3 PC
content; however, the results suggested that the 1000 kg/m3
Effect of plastic density and air content density was still appropriate for DF, IBA, CCS and sand mixes
The effects of plastic density and air content of foamed concrete but not for CFA mixes.
on consistence are shown in Figure 2. For the PC 300 kg/m3
mixes (Figure 2(a)), the slump flow increased for all foamed Since the density of foamed concrete was controlled by its

900

800

700
Slump flow spread: mm

600

500

400

Sand Sand
300 100%DF 100%DF
100%IBA 100%IBA
100%CCS 100%CCS
200 PC ⫽ 300 kg/m3
50%CFA 50%CFA
PC ⫽ 400 kg/m3 w/c ⫽ 0·50
100%CFA 100%CFA
100
600 1000 1400 600 1000 1400 1600 10 20 30 40 50 60 70
Plastic density: kg/m3 Plastic density: kg/m3 Air content: % vol.
(a) (b) (c)

Figure 2. Effect of density and air content of foamed concretes


with different RSA on slump spread: (a) effect of density
(PC ¼ 300 kg/m3 ); (b) effect of density (PC ¼ 400 kg/m3 );
(c) effect of air content

517
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

aggregate/air (foam) content when its cement content and w/c These differences are probably due to the interaction of internal
ratio were fixed, the optimal density can also be expressed by the rheological properties of the foamed concrete and its self-weight.
optimal aggregate content or air content. Indeed, as shown in In foamed concrete, plastic density is controlled by the water,
Figure 2(c), there was an optimum air content in terms of flow: aggregate and cement content. Increased density and hence self-
around 50% in volume obtained for all foamed concretes, weight overcomes its yield stress to initiate flowing behaviour. On
excepting the CFA and CR mixes. the other hand, the increased aggregate content increases plastic
viscosity and yield stress value, as noted previously by Dhir et al.
In general, the consistency of foamed concrete was found to (1999) and adversely affects flow. The optimum density seems to
increase with w/c ratio, water content and superplasticiser dosages balance these two opposite effects.
and is reduced at lighter plastic densities due to lower self-weight
(Karl and Wörner, 1993; Zain et al., 1999). However, Dhir et al. Compressive strength
(1999) examined foamed concretes of 1400 to 1800 kg/m3 plastic Cube strength was measured in accordance with BS EN 12390-3
density with 500 kg/m3 PC content and found that both plastic (BSI, 2009) on the 100 mm sealed–cured cube specimens at
viscosity and yield value were directly proportional to the density between 28 and 360 days.
of foamed concrete, and the fly ash mixes generally possessed a
lower plastic viscosity and yield value compared with the sand
ones of equal flow time. Jones and McCarthy (2005), examined Effect of RSA type and PC content
foamed concretes of 1000 to 1400 kg/m3 plastic density with The overall effect of different RSAs on cube strength of 1000
300 kg/m3 PC content and found that the slump flow of the sand kg/m3 foamed concretes are summarised in Figure 3. An increase
mixes decreased with plastic density while that of CFA/sand and in strength was observed in all mixes up to 1 year. With CFA,
CFA mixes increased with plastic density. The results presented in compressive strength was greatly improved (175%) compared to
Figure 2 agreed with that of Zain et al. (1999) in the plastic density 28 days. It appears the high PC content got more out of the FA.
range from 600 kg/m3 to 1000 kg/m3 (1400 kg/m3 in case of For the inert aggregates, the average strength gain was relatively
PC ¼ 300 kg/m3 with CFA) and they agreed with that of Jones and lower and close to that of the sand reference mix: around 19% to
McCarthy (2005) in the plastic density range from 1000 kg/m3 to 29% in one year compared to 28 days. It is not clear why this
1400 kg/m3 : improvement happens, as it is higher than would be expected for

5·0

28 d, PC ⫽ 300 kg/m3 Plastic density ⫽ 1000 kg/m3


4·5
28 d, PC ⫽ 400 kg/m3 w/c ⫽ 0·50*
4·0 360 d, PC ⫽ 300 kg/m 3
*w/c ⫽ 0·60 for CR aggregate mixes

3·5
S–C cube strength: N/mm2

3·0

2·5

2·0

1·5

1·0

0·5

0
Sand 50% 100% 50% 100% 50% 100% 50% 100% 50% 100% 50% 100% 50% 100%
DF DF IBA IBA GF GF FS FS CCS CCS CFA CFA CR CR
Aggregate type

Figure 3. Sealed–cured (S–C) cube strength of foamed concretes


with different types and levels of RSA

518
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

a normal weight concrete. One possibility is self-desiccation or increasing density from 600 kg/m3 to 1600 kg/m3 : High cement
loss of internal moisture to the atmosphere. content (PC ¼ 400 kg/m3 , Figure 4(b)) gave a more significant
increase than low cement content (PC ¼ 300 kg/m3 , Figure 4(a)).
The strength difference between different RSAs can be explained The results generally agree with those reported in the literature
by their different characteristics, such as particle density (mainly (De Rose and Morris, 1999; Jones and Giannakou, 2005;
affect air content as discussed below), water absorption, particle Kearsley, 1999; Ramamurthy and Narayanan, 2000).
distributions, shapes, softness and pozzolanic reactivity. In addi-
tion, the very low strength of the IBA mixes may be attributable Since the density of foamed concrete is controlled by its aggre-
to a particularly weak particle nature. Similar observations have gate and air contents, as might be expected, an exponential
been reported in normal weight concrete with IBA (Andrade et correlation between air content and compressive strength can be
al., 2007; Muller and Rubner, 2006). established as shown in Figure 4(c), which agrees with those
reported by Kearsley and Wainwright (2002). Figure 4(c) also
When PC content was increased from 300 kg/m3 to 400 kg/m3 , indicates clearly that at the same air content, higher strength is
strength was generally increased, but this was particularly clear for obtained with a higher PC content in the mixes.
the 100% CFA mix. This result is quite different from normal
weight concrete, and suggests that the increased amount of calcium For given cement content, w/c, and target plastic density, more
hydroxide available results in a larger amount of the fly ash foam was required with higher density RSAs, and, thereby,
reacting. It was not, however, possible to confirm or quantify this in higher voids content and lower strength. However, other RSA
this work but it is presumed that the bubble ‘walls’ are improved, characteristics, such as water absorption, particle distribution
although there are also changes in the bubble size and shape. and shape complicate this picture, as they influence the inter-
facial transition zone. In addition, owing to the pozzolanic
Effect of plastic density and air content reactivity of CFA and ‘softness’ of CR, foamed concretes with
The relationship between plastic density and air content and 28 these two aggregates were found to not follow the same
day cube strength is given in Figure 4. As would be predicted by strength–air content relationships and, therefore, are not in-
the void ratio effect, cube strength increased exponentially with cluded in Figure 4(c).

10·0

9·0
Sand Sand
8·0 PC ⫽ 300 kg/m3
100%DF 100%DF
100%IBA
28 day S–C cube strength: N/mm2

100%IBA PC ⫽ 400 kg/m3


7·0
100%CCS 100%CCS
50%CFA 50%CFA w/c ⫽ 0·50
6·0
100%CFA
100%CFA
5·0

4·0

3·0

2·0

1·0

0
600 1000 1400 600 1000 1400 1600 10 20 30 40 50 60 70
Plastic density: kg/m3 Plastic density: kg/m3 Air content: % vol.
(a) (b) (c)

Figure 4. Effect of density and air content of foamed concretes


with different RSA on cube strength: (a) effect of density
(PC ¼ 300 kg/m3 ); (b) effect of density (PC ¼ 400 kg/m3 );
(c) effect of air content

519
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

Drying shrinkage water contents, and thereby greater quantities of evaporable water,
Foamed concrete will also have relatively high drying shrinkage subsequently causing higher shrinkage strains. In addition, in-
strain compared to normal weight concrete owing to its high creased volume of hydrated cement paste, which is liable to
porosity, lack of coarse aggregate and high paste-to-fine aggregate shrink and reduced volume of fine aggregate to provide restraint,
content ratio (Dransfield, 2000). The amount of shrinkage is also increased shrinkage.
dependent upon various factors such as type of curing, size and
content of fine aggregates, type and amount of cement, w/c ratio The greater magnitude of shrinkage of the DF, IBA and FS mixes
and density of foamed concrete. The effects of RSA on drying can be attributed to their higher water absorption. The 100% CFA
shrinkage were examined in accordance with BS 6073-1 (BSI, mixes had no sand-sized particles to provide restraint to the
1981) . The test was carried out on 50 3 50 3 200 mm specimens shrinkage potential seen previously by Jones and Giannakou
after 1 week of sealed curing. The specimens were placed at the (2005). In comparison, the 50% CFA mix had lower or similar
environment of 20  28C with 55  5% RH and the length change drying shrinkage strains compared to reference sand mixes. This
were measured at intervals from 2 to 20 weeks. Around 75% of suggests that with this particular mix, there was sufficient
the total 20 week shrinkage occurred in the first 2 to 3 weeks and optimum particle packing of cement and fine aggregate to
in most cases reached a maximum at 4 to 6 weeks’ exposure. minimise shrinkage.
Similar observations have been reported by McGovern (2000).
Overall there was a poor relationship between shrinkage and air
Figure 5 shows drying shrinkage of sand and RSA concretes after content, and no equation to predict its magnitude could be easily
8 weeks of exposure. There were clear differences between RSAs derived. The phenomenon of drying shrinkage in foamed concrete
at the higher density, but at 600 kg/m3 shrinkage strain was requires further study to connect the characteristics of the fine
independent of aggregate type. In addition, high strength may aggregate, bubbles, air content and the resultant strain magnitude.
provide some restraint and then reduce drying shrinkage. When
cement content increased from 300 kg/m3 to 400 kg/m3 , drying Water permeation
shrinkage was generally increased. This is expected as greater Conventional methods for measuring permeation characteristics
cement contents (at a given w/c ratio) would result in higher free of normal concrete (e.g. ISAT as described in BS 1881-208:1996

5000

4500

4000
56 day drying shrinkage: μ strain

3500

3000

2500

2000

Sand Sand
1500
100%DF 100%DF
1000 100%IBA 100%IBA
100%CCS w/c ⫽ 0·50
100%CCS
500 50%CFA 50%CFA PC ⫽ 300 kg/m3
100%CFA 100%CFA PC ⫽ 400 kg/m3
0
600 1000 1400 600 1000 1400 1600 10 20 30 40 50 60 70
Plastic density: kg/m3 Plastic density: kg/m3 Air content: % vol.
(a) (b) (c)

Figure 5. Effect of density and air content of foamed concretes


with different RSA on drying shrinkage: (a) effect of density
(PC ¼ 300 kg/m3 ); (b) effect of density (PC ¼ 400 kg/m3 );
(c) effect of air content

520
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

(BSI, 1996) are not suitable for foamed concrete interconnected gypsum (CaSO4.2H2 O) in combination with epsomite
bubble structure (Dhir et al., 1999). The permeation test method (MgSO4.7H2 O) at 30% and 70%, respectively. Epsomite exhibits
applied in this study was based on the soil permeability test a chemical reaction of more aggressive nature (owing to its
method described by Barnes (2000). The test was carried out on a greater solubility) and is more commonly present in soils with
28 day sealed–cured 150 mm cube specimen with Ö15 mm tube high sulfate content. In addition, reference specimens were placed
cast 50 mm into the as-cast surface. Only specimens with in de-ionised water to determine swelling strain owing to
1000 kg/m3 plastic density and 300kg/m3 cement content were immersion. In addition to length changes, X-ray diffraction
measured. (XRD) analyses were carried out to determine whether expansive
reaction products, such as ettringite, were formed even though
The relationship between water permeability coefficients and air there was no expansion, which would suggest they form inside
content of RSA foamed concretes is shown in Figure 6. The bubbles.
water permeability coefficients of the RSA foamed concretes
ranged from around 1011 to 106 m/s. A reduction in the air The 200 day net expansions of the sand and RSA mixes owing to
content from 55% to 45% resulted in a significant reduction of the sulfate attack in relation to their air content are shown in
water permeability, and further reduction only marginally affected Figure 7. In general, expansions due to sulfate attack were similar
it, which suggests that there is a critical air content below which between the RSA and reference sand foamed concrete mixes, and
the bubbles are not interconnected. With the materials tested in they increased with reducing air content. The maximum expan-
this project, it seems that an air content of around 50% defines sion observed was on the 50% IBA specimen (3380 ì strain),
this point. Since all RSA foamed concretes in this study had whereas the minimum was on the 100% GF specimen (530 ì
lower air content than reference sand mix at the same density and strain). Visual observations found some physical defects on the
cement content level (Table 3) owing to the lower densities of the specimens which exhibited higher expansions above the trend line
RSA, their permeability coefficients were lower than that of sand and XRD analysis found significant gypsum and ettringite peaks
mix. from these specimens as discussed below.

Sulfate attack Some XRD analysis patterns to determine the reaction products
As foamed concrete is commonly used in void filling, there is formed during the two exposures are shown in Figure 8. When
clear interest in its ability to resist attack by sulfate-bearing soils. the foamed concrete samples were exposed in the sulfate environ-
To assess this, 1000 kg/m3 foamed concrete specimens ment, the sulfate reacted with calcium aluminate hydrate. The
(25 3 25 3 300 mm prisms) were selected to test for length product of the reaction, ettringite, has a considerably greater
expansion when exposed to a DS-4 m aggressive chemical volume than the compounds replaced, and then led to expansion
environment, in accordance with BRE special digest 1-1 (BRE, and disruption of the concrete. The magnesium sulfate also
2003) for up to 200 days. The exposure solution was made using attacks calcium silicate hydrates as well as calcium hydroxide

10⫺05
50%CCS
Sand mix
Water permeability coefficient: m/s (log scale)

10⫺06 50%GF
Cement content ⫽ 300 kg/m3 100%GF 100%CCS
Plastic density ⫽ 1000 kg/m3
10⫺07

10⫺08 50%DF
100%FS
50%FS
100%DF
10⫺09 100%CR
100%IBA
50%CR 50%IBA

10⫺10
100%CFA
50%CFA
10⫺11
20 25 30 35 40 45 50 55 60
Air content: %

Figure 6. Water permeability of RSA foamed concretes

521
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

4000
Cement content ⫽ 300 kg/m3
3500 Plastic density ⫽ 1000 kg/m3 50%IBA
200 days: neat length expansion: μ strain

3000
100%CR
2500 50%DF 100%CCS
50%CR

2000

100%CFA 100%IBA 50%CCS


1500

50%CFA 50%FS
1000
100%DF Sand mix
500 100%FS
100%GF
50%GF
0
20 25 30 35 40 45 50 55 60
Air content: %

Figure 7. The 200 days expansion of the sand and RSA mixes
owing to sulfate attack

(Ca(OH)2 ) and calcium aluminate hydrate and produces gypsum these mixes highlighted any differences between the RSAs. It can
and magnesium hydroxide (Mg(OH)2 ). Traces of ettringite (E, be seen that the ºind values, as expected, decreased with density
2Ł ¼ 9.1, 15.8 and 22.9) and gypsum (G, 2Ł ¼ 11.6, 20.7 and and increased air content. There was no significant difference
29.1) were detected in all the DS-4 m exposure specimens. In found between RSA and reference sand foamed concrete mixes,
those specimens with physical defects, the intensity in peaks of which is dominated by the air content.
gypsum was greater than that of ettringite, such as in 50% DF
(Figure 8(a)) and 100%CCS (Figure 8(f)) and both peaks were For foamed concrete with 1000 kg/m3 , ºind values were found
quite high. In others, the intensity in peaks of ettringite and slightly higher at high PC level; however, it was not apparent at
gypsum were relatively low, such as in 50% and 100%FS, and other density grades. The varying ranges of the ºind values for
100% DF specimens. The expansion of these samples was those with densities from 1000 to 1400 kg/m3 are in line with
generally on the trend line. Magnesium hydroxide may exist in that reported by Jones and Giannakou (2002). For the whole
the form of caminite (Mg3 (SO4 )2 (OH)2 ); however, its strongest range of the densities tested from 600 to 1600 kg/m3 , the ºind
peak (M, 2Ł ¼ 26.63) is close to that of quartz (Q, 2Ł ¼ 26.65) values ranged between 0.10 and 0.70 W/mK; this also conforms
and therefore cannot be clearly detected. Calcite (C, CaCO3 , with the range of foamed concrete thermal conductivities reported
strongest peak at 2Ł ¼ 38.0) was also detected in the samples, in the literature (Aldridge, 2000).
which indicated that the samples were more or less carbonated.
Only a few peaks of portlandite (P, Ca(OH)2 , 2Ł ¼ 18.1, 34.1, Conclusions
and 47.1) were detected, indicating that it was consumed in the The data show the RSA can be used in place of primary
reaction with magnesium sulfate, carbonation and the pozzolanic aggregates. However, there is considerable heterogeneity in these
reaction. materials, which gives rise to different foamed concrete charac-
teristics. The main points of this study are outlined below.
Thermal conductivity
One potentially useful application of the foamed concrete is for (a) The slump flow of the RSA foamed concretes was
thermal insulation (Cox and Van Dijk, 2003) but also to predict comparable with or, in some cases, that is, CFA, better than
temperature rises in large pours owing to the heat of hydration. that of reference sand foamed concrete. All RSA foamed
The indicative thermal conductivity (ºind ) of the RSA foamed concretes were self-flowing, except for 100% CR and 300 kg/
concrete was measured in accordance with the method described m3 cement content, most likely owing to low air (bubble)
in Jones and Giannakou (2002). phase required. Slump flow increased with cement content,
probably due to decreased yield stress with increasing volume
Figure 9 shows the indicative thermal conductivities of RSA and of cement sized particles. It also increased with plastic
sand mixes in relation to their plastic density and air content. In density, probably owing to increased self-weight. However,
this case, an additional mix series at 1600 kg/m3 was cast, as above 1000 kg/m3 slump flow reduced, most likely owing to

522
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

25 000 25 000
Q/M Q/M
(a) 50%DF (b) 100%DF
C – Calcite
20 000 G – Gypsum 20 000
Q – Quartz
E – Ettringite
M – Caminite CSH
15 000 15 000
CSH G/C
G
CSH G/Q
10 000 G/Q 10 000
DS-4m M G/C DS-4m
E G
E G E M
E CSH E E G CSH
5000 5000
H2O H2O

0 0
25 000 25 000
Q/M Q/M
(c) 50%FS (d) 100%FS

20 000 20 000

CSH
15 000 G/C
15 000
G/Q
CSH
P M
10 000 DS-4m G/C DS-4m 10 000
G/Q E
P M E E CSH
E G E CSH E
E E G G G
5000 5000
H2O H2O

0 0
25 000 25 000
Q/M Q/M
(e) 50%CCS (f) 100%CCS
M G
20 000 20 000

15 000 15 000
DS-4m DS-4m
E M CSH
E CSH G/Q
10 000 P G/C G/C 10 000
G/Q G
P
E E CSH E E E CSH
G E G
5000 5000
H2O H2O

0 0
5 10 15 20 25 30 35 5 10 15 20 25 30 35
2θ: Degrees 2θ: Degrees

Figure 8. XRD patterns of the RSA foamed concretes subjected


to 200 days DS-4 m exposure and reference water solution

the reduced foam/paste phase offsetting the increase of self- concretes was also comparable with that of reference sand
weight effect. Therefore, in most cases, 1000 kg/m3 was an foamed concrete. However, the pozzolanic effect of the CFA
optimum plastic density in terms of slump flow. resulted in a significant increase in strength.
(b) Seal-cured cube strength development of the RSA foamed (c) Drying shrinkage strain of the RSA foamed concretes was

523
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

0·8

0·7
Indicative thermal conductivity, λind: W/mk

0·6

Sand
0·5
100%DF
100%IBA
0·4 100%CCS
50%CFA
100%CFA
0·3
Sand
100%DF
0·2
100%IBA
100%CCS PC ⫽ 300 kg/m3
0·1 50%CFA
PC ⫽ 400 kg/m3
100%CFA
0
600 1000 1400 600 1000 1400 1600 10 20 30 40 50 60 70
Plastic density: kg/m3 Plastic density: kg/m3 Air content: % vol.
(a) (b) (c)

Figure 9. Effect of density and air content of foamed concretes


with different RSA on indicative thermal conductivity: (a) effect of
density (PC ¼ 300 kg/m3 ); (b) effect of density (PC ¼ 400 kg/m3 );
(c) effect of air content

higher than that of the reference sand mix and varied with the Acknowledgements
RSA characteristics and sand replacement level. Shrinkage The Waste and Resources Action Programme is thanked for
increased with cement content but decreased with increasing commissioning and funding this research. The support and
density. technical advice is gratefully acknowledged from WRAP, Ballast
(d ) Water permeation had a varying relationship with air content. Phoenix Ltd, BASF Construction Chemicals (UK) Ltd, British
Up to 45% air, water permeation did not change significantly. Airport Authority plc, MPA Cement (formerly British Cement
However, from an air content of 45% water permeability Association), John Doyle Construction Ltd, Propump Engineering
increased by an order of magnitude to a 55% air content. Ltd, Mineral Products Association (formerly Quarry Products
This suggests that there is a critical air content above which Association), Southern Water Asset Investment, Tarmac Group
the bubble space phase becomes connected. With the and the United Kingdom Quality Ash Association.
materials tested in this project, it seems that an air content of
around 45% defines this point. REFERENCES
(e) Resistance to sulfate attack was generally related to air Aldridge D (2000) Foamed concrete. Concrete 34(4): 20–22.
content, which reduced with RSA densities. XRD analyses Andrade LB, Rocha JC and Cheriaf M (2007) Evaluation of
showed traces of both ettringite and gypsum but rather less concrete incorporating bottom ash as a natural aggregates
than would be expected for the strains measured. Clearly, replacement. Waste Management 27(9): 1190–1199.
additional work is required. Barnes GE (2000) Soil Mechanics: Principles and Practice, 2nd
( f ) There was no significant difference found between RSA and edn. Palgrave.
sand foamed concretes in their thermal conductivity BRE (Building Research Establishment) (2003) BRE Special
property. The ºind values increased with density and reduced Digest 1-1, Concrete in Aggressive Ground, Part 1: Assessing
with air content. For the range of the densities tested from the aggressive chemical environment. Watford, UK.
600 to 1600 kg/m3 , the ºind values ranged between 0.10 and BSI (1981) BS 6073-1: Precast concrete masonry units.
0.70 W/mK; which conformed with the range reported in the Specification for precast concrete masonry units. BSI,
literature. London, UK.

524
Magazine of Concrete Research Use of recycled and secondary aggregates
Volume 64 Issue 6 in foamed concretes
Jones, Zheng, Yerramala and Rao

BSI (1996) BS 1881-208: Testing concrete. Recommendations for Jones MR, McCarthy A and Dhir RK (2005) Recycled and
the determination of the initial surface absorption of concrete. Secondary Aggregates in Foamed Concrete. WRAP, UK.
BSI, London, UK. Jones MR, Ansell T and Aldridge D (2009) Foamed concrete for
BSI (1997) BS EN 933-1: Tests for geometrical properties of sustainable construction. Concrete 43(5): 16–18.
aggregates. Determination of particle size distribution. Karl S and Wörner JD (1993) Foamed concrete – mixing and
Sieving method. BSI, London, UK. workability. Workability of Special Fresh Concretes (Bartos
BSI (2000) BS EN 1097-6: Tests for mechanical and physical PJM (ed.)). E & FN Spon, London, UK, pp. 217–224.
properties of aggregates. Determination of particle density Kearsley EP (1999) Just foamed concrete – an overview.
and water absorption. BSI, London, UK. Specialist Techniques and Materials for Concrete
BSI (2002) BS EN 12620: Aggregates for concrete. BSI, London, Construction (Dhir RK and Henderson NA (eds)). Thomas
UK. Telford, London, UK, pp. 227–237.
BSI (2009) BS EN 12390-3: Testing hardened concrete. Kearsley EP and Wainwright PJ (2002) The effect of porosity on
Compressive strength of test specimens. BSI, London, UK. the strength of foamed concrete. Cement and Concrete
BSI (2011) BS EN 197-1: Cement. Composition, specifications Research 32(2): 233–239.
and conformity criteria for common cements. BSI, London, Khatib ZM (2005) Properties of concrete incorporating fine
UK. recycled aggregate. Cement and Concrete Research 35(4): pp.
Cox L and Van Dijk S (2003) Foam concrete for roof slopes and 763–769.
floor levelling. Concrete 37(2): 37–39. McGovern G (2000) Manufacture and supply of ready-mix
De Rose L and Morris J (1999) The influence of the mix design foamed concrete. One Day Awareness Seminar on ‘Foamed
on the properties of micro-cellular concrete. Creating with Concrete: Properties, Applications and Potential’, University
Concrete (Dhir RK and Henderson NA (eds)). Thomas of Dundee, UK, pp. 12–25.
Telford, London, UK, pp. 185–197. Muller U and Rubner K (2006) The microstructure of concrete
Dhir RK, Jones MR and Nicol L (1999) Development of Structural made with municipal waste incinerator bottom ash as an
Grade Foamed Concrete Technology. Technical Report CTU/ aggregate component. Cement and Concrete Research 36(8):
898, Department of Environment Transport and the Regions. 1434-1443.
Dhir RK, Newlands MD and McCarthy A (2005) Use of Foamed Nixon P (2004) Concrete with Minimal or No Primary Aggregate
Concrete in Construction. Thomas Telford, London, UK. Concrete – the Magcon Pilot Study. Building Research
Dhir RK, Paine KA, Dyer TD, Tang MC (2004) Value-added Establishment, Watford, UK, IP 12/04.
recycling of domestic, industrial and construction arisings as Poon CS, Shui ZH, Lam L, Fok H and Kou SC (2004) Influence of
concrete aggregate. Concrete Engineering International 8(1): moisture states of natural and recycled aggregates on the
43–48. slump and compressive strength of concrete. Cement and
Dransfield JM (2000) Foamed concrete: Introduction to the Concrete Research 34(1): 31–36.
product and its properties. One day awareness seminar on Propump (2011) The foamed concrete specialists. See
‘Foamed Concrete: Properties, Applications and Potential’, www.foamedconcrete.co.uk, accessed (21/12/2011).
University of Dundee, UK, pp. 1–11. Ramamurthy K and Narayanan N (2000) Factors influencing the
Filipponi P, Polettini A, Pomi R and Sirini P (2003) Physical and density and compressive strength of aerated concrete.
mechanical properties of cement-based products containing Magazine of Concrete Research 52(3): 163–168.
incineration bottom ash. Waste Management 23(2): 145–156. WRAP (Waste and Resources Action Programme) (2006) The
Giannakou A and Jones MR (2002) Potential of foamed concrete Sustainable Use of Resources for the Production of
to enhance the thermal performance of low rise dwellings. Aggregates in England. The Waste and Resources Action
Innovations and Developments in Concrete Materials and Programme, Oxon, UK.
Construction (Dhir RK, Hewlett PC and Csetenyi LJ (eds)). WRAP (2010) http://aggregain.wrap.org.uk/more_information.
Thomas Telford, London, UK, pp. 533–544. html (accessed 21/12/2011).
HSE (Health and Safety Executive) (2009) Foamed concrete Zain MFM, Safiuddin M and Yusof KM (1999) A study on the
explosion – HSE interim position 03 December 2009. See properties of freshly mixed high performance concrete.
www.hse.gov.uk/construction/liveissues/foamedconcrete.htm, Cement and Concrete Research 29(9): 1427–1432.
accessed (09/12/10).
Jones MR and Giannakou A (2002) Foamed concrete for energy-
efficient foundations and ground slabs, Concrete 36(3): 14–17. WH AT DO YO U T HI NK?
Jones MR and Giannakou A (2005) Preliminary views on the To discuss this paper, please submit up to 500 words to
application of foamed concrete in structural sections using the editor at www.editorialmanager.com/macr by 1
pulverized fuel ash as cement or fine aggregate. Magazine of December 2012. Your contribution will be forwarded to
Concrete Research 57(1): 21–31. the author(s) for a reply and, if considered appropriate
Jones MR and McCarthy A (2005) Utilising unprocessed low-lime by the editorial panel, will be published as a discussion
coal fly ash in foamed concrete, Fuel 84(11): 1398–1409. in a future issue of the journal.

525
Copyright of Magazine of Concrete Research is the property of Thomas Telford Ltd and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.

You might also like