You are on page 1of 7

Construction and Building Materials 25 (2011) 4088–4094

Contents lists available at ScienceDirect

Construction and Building Materials


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

Recycled aggregate concrete produced with red granite dust as a partial


cement replacement
S.A. Abukersh, C.A. Fairfield ⇑
School of Engineering and the Built Environment, Edinburgh Napier University, 10 Colinton Road, Edinburgh, EH10 5DT, UK

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

Article history: Sustainable development involves reducing natural resource consumption by appropriate recycling. To
Received 30 September 2010 reduce the amount of waste going to landfill, red granite dust arising from the rock-crushing process
Received in revised form 31 March 2011 was assessed for suitability as a replacement for up to 30% by mass of the cement content of structural
Accepted 14 April 2011
grade concretes made with recycled coarse aggregates. While the impact on worldwide cement demand
Available online 8 May 2011
will be minimal, the contribution to waste reduction may be valuable. Laboratory trials showed that this
red granite dust rendered it suitable in terms of workability, stiffness and strength compared to normal
Keywords:
aggregate and Portland cement concretes. The potential for thereby overcoming the problem of low early
Concrete
Recycling
strength in some pulverised fuel ash concretes was demonstrated.
Aggregate Ó 2011 Elsevier Ltd. All rights reserved.
Cement
Granite
Testing

1. Introduction and research hypothesis waste product (RGD) generated from such rock crushers amounts
to one tonne per nine tonnes of aggregate produced [4]. Structural
Concrete has been a leading construction material for over a quality RAC here inferred that made with 100% coarse RA, pro-
century: its annual global production is about 3.8 billion m3 – duced from CDW and with a 28 day compressive strength above
roughly 1.5 tonnes per capita – according to Portland Cement 50 N mm 2. The research hypothesis was that a comparable work-
Association data [18]. Global natural aggregate (NA) consumption ability, strength, durability and overall performance were achiev-
is expected to rise by 2.9% annually to a 2013 total of around 28 able when RGD was used as a partial cement substitute. The
billion tonnes [14]; approximately 1 billion tonnes of construction authors’ null hypothesis is that RGD use caused no statistically sig-
and demolition waste (CDW) are generated annually in the EU nificant improvement over comparable NA concretes.
alone [20]. Due to advances in the manufacture of crushing
machinery and developments in recycling processes, it is now pos- 2. Test methods and materials used
sible to produce recycled aggregate (RA) cheaply. Annually,
All materials were tested in accordance with current British or European Stan-
approximately 65 million tonnes of aggregate are derived from
dards. Recycled aggregates were tested in accordance with standards originally pre-
recycled sources representing 23.6% of the total of 275 million ton- pared for natural aggregates while noting their inclusion in BS EN 12620 [8]. The
nes used in the UK [12]. Unfortunately, the European standard [6] coarse aggregates used were: a nominal 20 mm down crushed granite natural
for the specification of concrete did not include provision for the aggregate (for the control mixes) and a similarly graded recycled, crushed, screened
use of recycled aggregate concretes (RAC), however the UK com- and washed aggregate with approximately 6.6% impurities (timber, metal, glass, pa-
per, rubber and mortar). The fine aggregate was a natural medium sand. Lafarge’s
plementary standard thereto [10] permits their use but imposes
Portland PC Cem I (Class 42.5 N mm 2) cement was used throughout. The red gran-
additional requirements on maximum masonry, fines, asphalt ite crusher dust came from Cloburn quarry, Lanarkshire, Scotland. The rock-crush-
and sulphate contents. ing produces smaller aggregates; it also produces dust (the RGD which is separated
This research examined the potential for using recycled red by vacuum filtration) amounting to approximately 2% by mass of the coarse aggre-
granite dust (RGD) (from a crusher) as a partial cement substitute gate. This accumulates and, on windy days, is readily dispersed unless the stock-
piles are sprayed with water. Wet sieve analysis data for the RGD are given in
in concrete. The successful production of RAC, with this cement- Table 1. Chemical analyses of cement, pulverised fuel ash (PFA) (for comparison)
replacing RGD, could lead to its large-scale structural use, rather and RGD are given in Table 2.
than the prevailing low-value uses for many such materials. The Additionally, the cement grains all passed a 38 lm square aperture cloth mesh
sieve: manufacturer’s data indicated a specific surface area of 300 to 450 m2 kg 1.
The RGD’s high silica and alumina contents suggested a potential pozzolanic
⇑ Corresponding author. Tel.: +44 (0)131 455 2232; fax: +44(0)131 455 2239. reaction (although data shown later refutes this and the principal instigator of
E-mail address: c.fairfield@napier.ac.uk (C.A. Fairfield). strength was shown to be practically all physical) and a quasi-cementitious nature.

0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.04.047
S.A. Abukersh, C.A. Fairfield / Construction and Building Materials 25 (2011) 4088–4094 4089

Table 1 Additionally, the RGD’s gravimetric water absorption was approximately 1%.
Wet sieve analysis of RGD. Fig. 1 shows typical scanning electron micrographs of the particles in both RGD
and PFA samples. The PFA used here was Category N to BS EN 450-1:2005 [9] with
Sieve size (lm) Percentage passing (%) a fineness of 30% (based on the amount passing a 45 lm square aperture mesh
150 99.4 sieve).
75 81.5 Evaluation of the particles’ shape, texture and morphology showed the RGD’s
63 77.3 grains to be slightly larger, more angular, more porous and of greater specific sur-
45 58.6 face and roughness than a typical PFA sample. Energy dispersive X-ray analysis con-
38 52.2 firmed the chemical compositions outlined in Table 2.

3. Mixes determining optimum RGD content


Table 2
Chemical compositions: the cement, a typical PFA (for comparison) and RGD.
3.1. Mix designs: NAC, RAC and RGD-for-cement replacement
Property Average content (%)
Cement PFA RGD Increasing the fines content, particularly for medium to high
Silica (SiO2) content 21.8 45.0 61.4
range cement contents, results in increased water demand. In
Alumina (Al2O3) content 4.20 30.0 16.3 the absence of sufficient water, the available free water and ce-
Iron oxide (Fe2O3) content 2.50 11.0 3.66 ment tend to work less effectively. This leads to agglomeration
Quicklime (CaO) content 65.1 5.90 3.69 of cement particles and a stiffer mix. The additional water for
Magnesia (MgO) content - 2.25 1.70
wetting the finer PFA causes reduced strength but it may be pos-
Potassium oxide (K2O) content 0.72 3.40 3.75
Sodium oxide (Na2O) content 0.13 1.30 3.62 sible to avoid the lack of workability and drop in strength by
Sulphate (as SO3) content – 1.50 0.05 increasing the cement content. An alternative to this is to reduce
Na2Oeq (Na2O + 0.658  K2O) content 0.60 3.51 6.09 the water content and use a superplasticiser (SP) [13]. The
Other oxides 3.15 – 1.10 authors chose to reduce the water content and use an SP admix-
Loss on ignition – 5.0–7.0 5.01
ture to maintain workability while avoiding the extra cost of an
increased cement content (the SP may add cost at a potentially
higher rate per m3 in certain markets). The authors used Fosroc’s
Structuro 11180 superplasticising admixture (to EN 934-2: T3.1/
3.2). This polycarboxylate polymer contained a maximum chlo-
ride content of less than 0.1% by mass and a maximum alkali con-
tent below 3.0% by mass.
For concretes with compressive strengths under 40 N mm 2, the
coarse aggregate rarely becomes strength-limiting as a relatively
high water:cement (w/c) ratio is used (i.e. within 0.5 to 0.7). Thus,
the hardened paste and interfacial zone between it and the
aggregate is the weakest link [3]. Similarly for such concrete, the
mineralogy of the coarse aggregate is not a major factor in its mix
design unless it is proven to contain sufficient deleterious minerals
[17]. Conversely, higher strength concrete, for which a lower w/c ra-
tio is often used (i.e. below 0.45), the hardened paste and interfacial
zone are strong enough and are non-strength limiting: it is the
coarse aggregate’s strength and mineralogy that control the
concrete’s strength [16]. All RA concretes herein were made with
100% RA: RGD was then used as a partial cement substitute.
The characteristic, 28 day, compressive strength of the authors’
natural aggregate control concrete was 50 N mm 2 (target mean
strength 63 N mm 2) and the UK’s most common mix design
method [21] yielded the following proportions (per m3 of fresh
concrete): a free water:cement ratio of 0.4, a cement content of
425 kg, a water content of 170 kg, a fine aggregate content of
545 kg and a coarse aggregate content of 1280 kg. The RAC, based
on the natural aggregate control concrete was modified to account
for the properties of the recycled aggregate, differed slightly in that
it had a free water:cement ratio of 0.4, a cement content of 425 kg,
a water content of 170 kg, a fine aggregate content of 530 kg and a
coarse aggregate content of 1235 kg.

3.2. RAC with partial RGD-for-cement replacement


Fig. 1. Scanning electron micrograph of RGD particles (PFA shown for comparison).

BS EN 12620 defines crushed rock dust, or simply dust, as aggre-


Fly ash meeting BS EN 450 should contain at least 25% reactive silica, far below that
found in this RGD (61.4%), albeit inevitably at a differing reactivity: X-ray diffraction gate particles passing a 75 lm aperture sieve. This RGD would thus
revealed variable amorphous/crystalline silica ratios and this in itself warrants fur- be classified as a very fine powder. In contrast to alternative PFA
ther investigation outwith the scope of this paper. Another limitation is that the mixes, the ideal level of cement replacement for concrete mixes
sum of ferric oxide (Fe2O3), alumina (Al2O3) and silica (SiO2) must be at least 70% with RGD was unknown. In this case, selection of the optimum
in PFA for possible use in concrete: the sum of these oxides is over 80% in the chosen
RGD, leading to the belief that it may be useful for concrete mixes. Although seem-
RGD content was based on the workability and strength of trial
ingly high, the RGD’s alkali content (as Na2O equivalent) was not freely available mixes. Mix proportions trialled here at different RGD contents
and therefore not able to contribute to any potential alkali-aggregate reaction. are given in Table 3.
4090 S.A. Abukersh, C.A. Fairfield / Construction and Building Materials 25 (2011) 4088–4094

Table 3
Mix proportions with differing RGD contents.
100

RGD content (%) Mass (kg per m3 of fresh concrete)


90
Cement RGD Water Coarse Fine aggregate
aggregate
0 425 0 170 1280 545 80
20 372 93 160 1270 545
30 345 145 155 1265 540
40 312 208 150 1250 535 70
50 280 280 145 1230 525

60

Strength (Nmm-2 )
Table 4
Compressive strength at 28 days
Workability of mixes with RGD as a partial cement substitute.
Compressive strength at 56 days
50
RGD (%) WR SP w/c Slump Vebe Notes
Compressive strength at 90 days
(%) (%) ratio (mm) time (s)
0 0 0 0.40 15 12 Uniform mix 40 Tensile splitting strength at 28 days
0 16 0.8 0.34 20 10 Uniform mix
20 16 0.8 0.34 25 9 Uniform mix
Tensile splitting strength at 56 days
30 16 0.8 0.34 20 11 Slightly cohesive 30
Tensile splitting strength at 90 days
40 16 0.8 0.34 25 8 Less cohesive
50 16 0.8 0.34 30 6 Segregating slightly Flexural strength at 28 days
20
Flexural strength at 56 days
For selection purposes, a total of five NA concrete trials, with
RGD partially replacing cement, were produced with 0.8% SP and 10
16% concomitant water reduction (WR). It must be stressed that
these mixes do not represent an exhaustive list – see Abukersh 0
[1] for more details – and are subject to the limitation that two 0 10 20 30 40 50
parameters (total cement content and water content) are varied RGD content (%)
simultaneously. Fuller treatment of the optimisation of the RGD
content remains a subject of ongoing research. Fig. 2. Mechanical properties of mixes with RGD as a partial cement substitute.

3.3. Results: workability as a determinant of optimum RGD content


The superplasticised NAC with up to 30% RGD content reached a
Table 4 displays mixes used for the selection of the best RGD ce- compressive strength of 72 N mm 2, (above the 63 N mm 2 target
ment replacement level and their resulting workability. mean strength of the control mix): data at greater ages exhibited
Table 4 indicates that as the RGD content increased, the slump similar trends. Visual inspection of the failure surfaces revealed
initially increased – possibly attributable to the SP’s dispersive shearing of aggregate grains at lower RGD contents indicating a
influence – however beyond 20% cement replacement, the slump higher matrix strength: higher RGD (40 and 50%) contents under-
decreased to reach a minimum at 30% RGD content (still above that went bond failure and had lower strengths. In addition to the ce-
of the control mix). The physical reason for reduced slump within ment content reduction, the decrease in tensile strength may be
the range 20–30% RGD content was possibly due to local grain rear- attributed to the increased amount of free water due to the lower
rangement (around this RGD content the density of the wet concrete absorption of the RGD – this despite the constant w/c ratio used for
peaked). Beyond this range this closest packed arrangement re- each mix – thus the higher RGD content mixes became wet and
collapsed and the slump increased again possibly due to the lean resulting in lower strengths. However, the superplasticised
increased free water content in the voids due to the lower absorp- concrete without RGD was stronger than that with it but mixes
tion of RGD grains when compared with cement. Cement also reacts with 20% and 30% RGD were stronger than the control mix without
chemically with water and when there is more cement (or less RGD) SP and RGD. Of all mixes tested, that with 30% RGD attained the
there will be less free water for grain lubrication. In contrast, no highest 28 day compressive and tensile strengths, with a flexural
reaction of RGD with water takes place; some water is absorbed strength of 93% of the control mix’s. The small compressive
by the grains and the fluidity is increased when there is less cement strength changes between 28 and 90 days suggested physical
in the mix. However, further increases in RGD content were pre- rather than chemical (pozzolanic) action: in fact the greatest
dicted to result in a lack of mix cohesion. Below 50% substitution, re- change over time occurred in the 0% RGD mixes.
sults showed that it was possible to maintain adequate concrete Workability data showed the 30% RGD mix to have a uniform
workability, with the help of a superplasticising admixture. The 20 mm slump, which, although the lowest, nevertheless repre-
authors also acknowledge the fact that test-to-test, a 5 mm change sented a mid-range value compared to that designed (10–
in slump is negligible and within the range of inherent variability for 30 mm). This mix’s Vebe time of 11 s was also within the original
the slump test itself. Over the range 0% RGD to 30% RGD the trend is design limits (6–12 s).
at least established despite these concretes appearing to be slightly
less workable than comparable mixes used on construction sites. 3.5. Results: density and ultrasonic pulse velocity as determinants of
optimum RGD content
3.4. Results: mechanical properties as determinants of optimum RGD
content Table 5 summarises the time dependency of the density and
ultrasonic pulse velocity of concrete mixes with RGD contents
Fig. 2 shows the time dependency of the three key mechanical ranging from zero (control mix) to 50% cement replacement.
properties of concrete mixes with RGD contents ranging from zero The data show that both density and pulse velocity generally
(control mix) to 50% cement replacement. decreased with the addition of RGD, however, the density at 30%
S.A. Abukersh, C.A. Fairfield / Construction and Building Materials 25 (2011) 4088–4094 4091

RGD content was slightly higher at both 56 and 90 days (yet not at replaced by RGD were produced. Two mixes (one NAC, the other
28 days). This was possibly caused by the gradual development of RAC, both 30% RGD-for-cement replacement, no superplasticiser,
an improved pore structure arising from the continued cement no water reduction, 0.4 w/c ratio and 0.32 w/b ratio) were de-
hydration and void-filling nature of the RGD grains. Despite the signed to reach 50 N mm 2 compressive strength at 28 days
density improvement at 30% RGD, the density was marginally (slump, 10–30 mm; Vebe time, 6–12 s) for NAC and RAC mixes,
(2.2% and 2.4% at 56 and 90 days respectively) below that of the in accordance with the Building Research Establishment mix de-
control mix. The pulse velocity followed a similar trend with a sign method [21]. In the other two mixes (as above but with a
maximum at 30% RGD and a range from 4.65 to 5.00 km s 1; there- 0.34 w/c ratio and a 0.27 w/b ratio) the water content was re-
fore, all RGD mixes could be classified as excellent against the cri- duced by 16% in combination with 0.8% SP addition (as proven
teria of BS 1881-209:1990 [5]. The density differences measured optimal for the materials at hand). Section 4 continues by pre-
largely corresponded to the lighter material replacing the cement senting test data relating to these mixes’ workability and
and the water content differences between mixes; with this in mechanical properties.
mind the small differences observed may well have been the result
of the test’s inherent variability. On the basis of the measured
4.2. Results: workability of concretes with 30% RGD-for-cement
properties (Sections 3.2–3.5 inclusive), the 30% RGD mix was
replacement
selected for further testing.
Table 6 shows the workability measurements and the visual
observations made on these latest mixes.
4. Mixes with 30% RGD content Data presented in Table 6 showed that it was possible to pro-
duce concrete with 30% RGD cement replacement to an acceptable
standard of workability; a little extra compaction was needed for
4.1. Concrete mixes with 30% RGD-for-cement replacement non-SP mixes. Comparing 30% RGD with the control mixes, a mar-
ginal improvement was seen; however, it could be merely the ef-
To investigate further the possibility of using RGD in NA and fect of the reduction of absorption capacity of the newly formed
RA concretes, four other mixes in which 30% of their cement was binary binder (70% cement, 30% RGD). This implied that RGD was
Table 5 less absorbent than cement. The workability was much improved
Density and ultrasonic pulse velocity of mixes with RGD as a partial cement in superplasticised mixes despite the water reduction. Another
substitute. noteworthy point was the reddish external appearance of the
RGD (%) Density (kg m 3
) and (ultrasonic pulse velocity (km s 1
)) 30% RGD which may be beneficial in some circumstances, undesir-
28 days 56 days 90 days
able in others.

0 2393 (4.85) 2288 (4.76) 2278 (4.86)


0 + 0.8% SP 2316 (4.85) 2314 (5.04) 2322 (5.10) 4.3. Results: mechanical properties of mixes with 30% RGD-for-cement
20 + 0.8% SP 2216 (4.88) 2203 (4.92) 2207 (4.91) replacement
30 + 0.8% SP 2193 (5.00) 2237 (4.94) 2223 (4.97)
40 + 0.8% SP 2189 (4.73) 2185 (4.90) 2190 (4.96)
50 + 0.8% SP 2176 (4.65) 2184 (4.82) 2181 (4.87) RGD concrete mixes were tested and the results shown in Table
7 with data from the control mixes.

Table 6
Workability data: control mixes versus those at 30% RGD-for-cement replacement.

Mix Description w/c w/b* SP (%) WR (%) Slump (mm) Vebe time (s) Notes
Mixes without RGD
1 NAC (control mix) 0.40 0.40 0 0 15 12 Uniform mix
2 RAC (control mix) 0.40 0.40 0 0 10 14 Uniform mix
3 NAC + SP 0.34 0.34 0.8 16 15 10 Uniform & cohesive
4 RAC + SP 0.34 0.34 0.8 16 25 7 Uniform & cohesive
Mixes with 30% RGD-for-cement replacement
5 NAC 0.40 0.32 0 0 20 9 Uniform mix, needs 2 min compaction
6 RAC 0.40 0.32 0 0 15 13 Uniform mix, needs 3 min compaction
7 NAC + SP 0.34 0.27 0.8 16 55 5 Uniform & cohesive
8 RAC + SP 0.34 0.27 0.8 16 45 7 Uniform & cohesive
*
Water to binder ratio.

Table 7
Mechanical properties: control mixes versus 30% RGD-for-cement replacement.
2 2 2
Mix RGD % Compressive strength/Nmm Tensile splitting strength/Nmm Flexural strength/N mm
28 days 56 days 90 days 28 days 56 days 90 days 28 days 56 days
1 0 69.0 (1) 71.0 76.0 (1) 4.7 3.8 4.6 8.1 9.8
2 0 52.0 (0.75) 58.0 58.5 (0.78) 3.6 4.1 3.6 7.8 7.9
3 0 74.0 (1.07) 89.5 91.5 (1.21) 5.8 7.2 7.1 11 11
4 0 64.0 (0.93) 72.0 73.0 (0.96) 4.8 4.4 4.9 11 8.2
5 30 64.0 (0.93) 62.0 72.5 (0.96) 4.8 3.6 4.7 6.2 6.8
6 30 46.0 (0.67) 53.0 53.5 (0.70) 3.5 3.5 3.7 7.5 5.5
7 30 72.0 (1.04) 76.5 74.0 (0.97) 4.6 4.5 5.4 7.6 6.7
8 30 53.5 (0.78) 54.0 54.5 (0.72) 3.5 4.6 5.2 7.3 7.0

Values in brackets are ratios of the mix strength at the specified age to the corresponding strength of the NAC control mix.
4092 S.A. Abukersh, C.A. Fairfield / Construction and Building Materials 25 (2011) 4088–4094

Fig. 3. Stress–strain plots: NAC (left) & RAC (right) mixes at 30% RGD content.

Mixes 7 and 8 were also tested at 7 days to check the early age Table 8
strength of RGD concrete giving compressive strengths of 65.5 and Compressive strength and modulus of elasticity of 30% RGD mixes.
41.5 N mm 2 respectively. The 28 day tensile strengths were con- Mix RGD (%) Compressive strength Modulus of elasticity Poisson’s
sistently between 6% and 8% of the compressive strengths at the (N mm 2) 28 days (GPa) 28 days ratio
same age. 1 0 69.0 (1) 36.3 (1) 0.10
This testing showed that NAC concrete made with 30% RGD-for- 2 0 52.0 (0.75) 28.3 (0.78) 0.28
cement replacement achieved the target mean strength of 3 0 74.0 (1.07) 51.4 (1.42) 0.39
63 N mm 2. This concrete, which can be regarded as a recycled 4 0 64.0 (0.93) 40.0 (1.10) 0.30
5 30 64.0 (0.93) 41.4 (1.14) 0.25
concrete, attained better compressive and tensile (and comparable
6 30 46.0 (0.67) 28.3 (0.78) 0.25
flexural) strengths than the control mix. RAC mixes with 30% RGD 7 30 72.0 (1.04) 49.7 (1.37) 0.27
were able to achieve the characteristic strength only after 56 days 8 30 53.5 (0.78) 39.0 (1.08) 0.25
and were unable to attain the target mean strength at any age un-
Values in brackets are the ratios of the compressive strength and modulus of
less an appropriate SP was used (with concomitant water reduc- elasticity at 28 days to the corresponding values for the NAC control mix.
tion). The SP did not markedly improve the strength of RGD
concrete mixes: a comparison of mixes with and without SP
showed small increases which decreased with time. Although
the middle cube in a cruciform pattern. Stress–strain plots are pre-
RAC mixes with 30% RGD achieved an acceptable strength
sented in Fig. 3 with derived properties in Table 8.
(53.0 N mm 2) after 56 days, it is clear that when two recycled
Results presented in Table 8 show that Es tended to improve
materials (recycled coarse aggregate and RGD) were combined,
with the addition of RGD and SP, particularly with NAC mixtures;
the resulting concrete was unlikely to achieve this high a target
NAC modulus increased by 14–37%, most likely due to the greater
mean strength even with admixtures. It is notable that NAC and
stiffness of NA and a strong matrix. NAC with 30% RGD and SP
RAC mixes made with 30% RGD-for-cement replacement achieved
achieved the highest Es. In contrast, superplasticised RAC improved
practically equal strengths irrespective of the presence of the
by 8% due to the decreased strength of RAC without SP. The non-
added SP. The superior rate of strength increase of concrete mixes
superplasticised RAC (Mix 6 with only RGD) achieved an elastic
with added SP was more pronounced until 90 days for NAC mixes
modulus equal to that of the RAC control mix; while the RAC mix
and 56 days for RAC mixes: thereafter, this difference vanished.
with SP attained the highest modulus of 39 GPa (better than the
The data indicated that the improvement of compressive strength
NAC control mix). However, although the moduli achieved by
of mixes without RGD varied from 10% to 24% and 2% to 16% for
RAC were lower than those for NAC, they remained acceptably stiff.
mixes incorporating 30% RGD; this difference was not deemed
Slender sections need high strength and stiffness; such sections
significant.
are preferable for cost- and space-saving. In this context, it has
Tests on three 100 mm (side length) cubes from Mix 7 (NAC,
been found that the density of 30% RGD concrete enables manufac-
0.8% SP content, 16% water reduction and 30% RGD) yielded an
ture of lighter sections. Combined with increased strength and
early compressive strength of 67 N mm 2 at 7 days. A similar mix
stiffness at the same level of RGD-for-cement replacement (30%),
produced with PFA instead of RGD achieved 50 N mm 2 at the
this indicates potential widespread acceptability and applicability
same age. This indicated improved early strength of this RGD
for this novel recycled concrete.
mix when compared to a similar PFA mix. This is potentially valu-
able in situations where high early strength is needed for rapid
construction or removal of formwork. 4.4. Comparisons to other strength and stiffness estimates
The static modulus of elasticity Es for RGD concrete mixes was
determined from strain-gauged compression tests. A vertical stack The elastic modulus Es is ideally measured directly on a con-
of three dental plaster-jointed 100 mm cubes was subjected to two crete specimen. From an experimental point of view, this is not al-
partial load cycles before crushing at a rate of 180 kN per minute ways straightforward. A concrete’s elastic modulus Es, flexural
and a temperature of 20 °C. The quarter-bridge, 120 X circuit resis- strength rf and tensile strength rt are often estimated using either
tance, 2.10 gauge factor, strain gauges were mounted on the face of theoretical or empirical approaches. Various sources suggest
S.A. Abukersh, C.A. Fairfield / Construction and Building Materials 25 (2011) 4088–4094 4093

Table 9
Models used to estimate concrete properties from the compressive strength.

Term References Key formula Mix 5 Mix 6 Mix 7 Mix 8


rc = 64.0 N mm 2 rc = 46.0 N mm 2 rc = 72.0 N mm 2 rc = 53.5 N mm 2
Es = 41.4 GPa Es = 28.3 GPa Es = 49.7 GPa Es = 39.0 GPa
rf = 6.15 N mm 2 rf = 7.54 N mm 2 rf = 7.56 N mm 2 rf = 7.30 N mm 2
rt = 4.8 N mm 2 rt = 3.5 N mm 2 rt = 4.6 N mm 2 rt = 3.5 N mm 2
Es (GPa) ACI 318 [2] 4127rc 33.0 n/a 35.0 n/a
1=
CEB-FIP [11] 21.5(0.1rc) 3 39.9 n/a 41.5 n/a
GB 50010 [15] 105/(2.2 + 34.7/rc) 36.5 n/a 37.3 n/a
Dhir et al. [13] 370rc + 13100 n/a 30.1 n/a 32.9
1=
Ravindrarajah et al. [19] 7700(rc) 3 n/a 27.8 n/a 29.0
Xiao et al. [22] 105/(2.8 + 40.1/rc) n/a 27.2 n/a 28.2
Eurocode 2, BSI [7] 22(rc/10)0.3 38.4 34.8 39.8 36.4
rf (N mm 2) ACI 318 [2] 0.54rc 4.32 3.66 4.58 3.95
CEB-FIP [11] 0.81rc 6.49 5.49 6.87 5.92
Xiao et al. [22] 0.75rc n/a 5.10 n/a 5.50
rt (N mm 2) ACI 318 [2] 0.49rc 3.92 n/a 4.16 n/a
2=
CEB-FIP [11] 1.4(0.1rc) 3 4.83 n/a 5.22 n/a
3=
GB 50010 [15] 0.19(rc) 4 4.30 n/a 4.70 n/a
Xiao et al. [22] 0.24(rc)0.65 n/a 2.90 n/a 3.19
Eurocode 2, BSI [7] 2.12 ln(1 + (rc/10)) 4.24 3.24 4.46 3.92

formulae for normal and high strength concrete property predic- structural and civil engineering practice. Concrete with 30% RGD
tion and were used [1] on the NA and RA concretes described here- demonstrated better early age strength than similar PFA-based
in. Key results are given in Table 9. concrete. Any RGD addition beyond 30% caused a decrease in
The purpose of the analysis carried out here is to check the con- strength, although the mechanical properties remained acceptable
formity of the measured properties to those estimated. Results pre- at up to and including 50% RGD-for-cement replacement.
sented in Table 9 showed that the relationships between rc and Es,
rf and rt for normal and recycled aggregate concretes were not
necessarily valid for NAC and RAC mixes with chemical and/or References
mineral admixtures. Further research is recommended for the
[1] Abukersh SA. High quality recycled aggregate concrete. PhD thesis, Edinburgh:
development of new prediction models, incorporating the effects Edinburgh Napier University; 2009.
of admixtures. [2] American Concrete Institute. ACI 318 Building code requirements for
reinforced concrete. Committee for structural concrete, Detroit: Am Concrete
Meanwhile, in the absence of such models, the appropriate
Inst.; 2002.
inter-relationships of the mechanical properties of such concretes [3] Beshr H, Almusallam AA, Maslehuddin M. Effect of coarse aggregate quality on
remain best evaluated experimentally. This would thereby encom- the mechanical properties of high strength concrete. Constr Build Mater
pass the variety of aggregate combinations, regional geologies, 2003;17(2):97–103.
[4] British Geological Survey. Sustainable utilization of quarry by-products.
production methods, curing processes, test methods and specific Keyworth, UK: BGS; 2009. <http://www.sustainableaggregates.com/docs/
standards used. revs/t2c_suqbp.pdf>. [accessed 21.02.11].
[5] British Standards Institution. BS 1881-209:1990 Testing concrete.
Recommendations for the measurement of dynamic modulus of elasticity.
5. Conclusions London: BSI; 1990.
[6] British Standards Institution. BS EN 206-1:2000 Concrete. Specification,
performance, production and conformity. London: BSI; 2000.
Substantial quantities of RGD, estimated at between 1% and 2% [7] British Standards Institution. EN 1992-1-1:2004 Eurocode 2: Design of
by mass of the total aggregate crushed, are produced in quarries. concrete structures. London: BSI; 2004.
[8] British Standards Institution. BS EN 12620:2002 + A1:2008 Aggregates for
This RGD is currently disposed of as waste: its large-scale use as
concrete. London: BSI; 2002.
a cement replacement in concrete mixes could therefore be [9] British Standards Institution. BS EN 450-1:2005 + A1:2007 Fly ash for concrete.
beneficial. The RGD’s physical and chemical properties exhibited London: BSI; 2005.
similarities to PFA allowing mix design to proceed along [10] British Standards Institution. BS 8500-2:2006 Concrete. Complementary
standard to BS EN 206-1. Specification for constituent materials and
well-established lines. Despite the RGD’s higher silica content concrete. London: BSI; 2006.
(approximately 61% compared to PFA at approximately 51%), the [11] Comité Euro-International du Béton & Fédération Internationale de la
RGD was a less reactive material than typical PFAs. Précontrainte. CEB-FIP Model Code 1990. Design Code. London: Thomas
Telford Ltd; 1993.
The mechanical properties of RA concretes should be directly [12] Dhir RK, Payne KA. Performance related approach to use of recycled
measured on representative samples; the estimation by numerical aggregates. Waste & resources action programme Tech. Report AGG 0074,
formulae may lead to wide variations. The results showed that NA Banbury: WRAP; 2007.
[13] Dhir RK, Zhu WZ, McCarthy MJ. Use of portland pfa cement in combination
concretes produced with RGD at 30% cement replacement level with superplasticizing admixtures. Cem Concr Res 1998;28(9):1209–16.
had strengths either comparable to, or better than, equivalent [14] Freedonia Group Inc. World Construction Aggregates Industry. Rockville,
control mixes, although they were marginally weaker than similar Maryland, USA: Freedonia Group Inc.; 2009.
[15] Ministry of Construction of the People’s Republic of China. National Standard
PFA-based concretes. An equal, or improved, modulus of elasticity GB 50010-2002 Code for design of concrete structures. Beijing: China
ensued. In the same way, RA aggregate concretes with 30% RGD Architecture & Building Press; 2002.
showed strengths comparable to equivalent PFA-based concretes: [16] Illston JM, Domone PLJ. Construction materials: their nature and behaviour.
3rd ed. London: E & FN Spon; 2001.
in these cases a higher elastic modulus was achieved with the RA
[17] Neville AM. Neville on concrete. An examination of issues in concrete
concretes. practice. Farmington Hills, Michigan: Am. Concrete Inst. International; 2003.
Concrete mixes containing 30% RGD showed good fresh proper- [18] Portland Cement Association. Cement and concrete basics. PCA on-line. Skokie,
ties, better than expected mechanical properties and excellent (al- Illinois, USA: PCA <http://www.cement.org/basics/concretebasics_lessonone
c.asp>. [accessed 21.03.11].
beit reddish coloured) surface finish. The concretes produced [19] Ravindrarajah RS, Steward M, Greco D. Variability of recycled concrete
covered the range of the design strengths typically required in aggregate and its effect on concrete properties: a case study in Australia. In:
4094 S.A. Abukersh, C.A. Fairfield / Construction and Building Materials 25 (2011) 4088–4094

Proceedings of international workshop on recycled concrete. Niigata, Japan; [21] Teychenné DC, Franklin RE, Erntroy HC. Design of normal concrete mixes. 2nd
2000. p. 9–25. ed. Watford: Building Research Establishment; 1997.
[20] Sonigo P, Hestin M, Mimid S. Management of construction and demolition [22] Xiao J-Z, Li J-B, Zhang C. On relationships between the mechanical properties
waste in the EU. In: Proceedings of stakeholders workshop, Brussels, Belgium; of recycled aggregate concrete: an overview. Mater Struct 2006;39(6):
2010. 655–64.

You might also like