Performance of Portland/Silica Fume Cement Concrete Produced With Recycled Concrete Aggregate
Performance of Portland/Silica Fume Cement Concrete Produced With Recycled Concrete Aggregate
Performance of Portland/Silica Fume Cement Concrete Produced With Recycled Concrete Aggregate
The use of recycled aggregate in concrete industry has a great new construction may play an important role in promoting
potential to reduce demand for natural aggregate and the amount the use of recycled aggregates in the concrete industry.
of solid waste dumped at landfill sites. The main objective of this Sustainable construction has become one of the key
study is to design a concrete made with different proportions of requirements of today’s concrete. A wide range of recycled
coarse recycled concrete aggregate (RCA) having a similar 28-day aggregate has been steadily introduced in a range of civil
design strength to corresponding natural aggregate concrete. engineering and construction applications as partial replace-
Recycled coarse aggregates, obtained by crushing concrete debris ment of natural aggregates in concrete.1-5 While various
from various sources, were used in three proportions of 30, 50, and
types of recycled aggregates, such as bricks,3 tiles,4,5 and
100% (by weight) to produce concrete with various water-cement
ratios (w/c) and different compressive strength grades.
glass might be incorporated in concrete, RCA is still consid-
The key mechanical properties and durability performance of ered to be the most available for use as secondary aggregates.
concrete produced with portland silica fume (PSF) and RCA were It is well recognized that concrete is the second-most-
investigated. The RCA used showed inferior mechanical properties consumed material, with an estimated worldwide consump-
(crushing and impact values) than the natural aggregates (NA) tion6 in 2006 of approximately 31 billion tonnes (6.8 ×
and, hence, RCA concrete exhibited slightly lower performance 104 billion lb), accounting for 20 to 50% of all resources
than NA concrete. The results showed that up to 30% coarse RCA explored. Using different types of industrial by-products and
had no major effect on the compressive strength of concrete and, recycled materials in concrete industry would significantly
thereafter, a gradual reduction in strength with an increase in RCA contribute to achieving sustainability in construction, as it
content was observed. Reducing the w/c of concrete treated with has a potential to reduce landfill charges and the exploration
the RCA has led to an enhanced compressive strength, higher and extraction of nonrenewable materials.
resistance to carbonation, and chloride ion ingress. It was also Previous work on the use of recycled concrete aggre-
found that, when properly designed, portland cement silica fume gate (RCA) in concrete production concluded that:
(PC-SF) concrete made with different proportions of coarse 1) the source7 and strength8-13 of the parent concrete used
RCA as substitute of NA may contribute to enhance the durability to produce RCA has no effect on the grading of the RCA
performance of concrete. but may have great influence on the strength characteristics
Keywords: carbonation; chloride ingress; drying shrinkage; durability; of the new concrete14; 2) for the same workability, RCA
mechanical properties; recycled concrete aggregate; silica fume; strength; requires more mixing water than NA concrete; 3) a reduc-
sulfate attack; sustainable concrete tion in the compressive strength of concrete made with RCA
as compared to concrete with NA has been reported8,15-18;
INTRODUCTION and 4) the replacement ratio strongly affects the mechanical
It is now widely acknowledged and acceptable that the properties and durability performance of the new concrete
use of recycled aggregates in concrete production offers made with RCA.16 On the other hand, one of the major diffi-
an environmentally responsible and economically viable culties with recycled aggregates is the variability in their
sustainable route. In recent years, concrete made with properties due to composition, contents, and proportions
recycled aggregate has started to become a practical reality, largely linked to the original source of debris, which conse-
gained general acceptance, and is considered to be one quently results in the variability of concrete produced.
of the most promising solutions to reduce the amount of Although significant research has been done on concrete
construction and demolition waste (CDW) that may end up using different types of recycled aggregates including RCA,
in landfills. With an estimated amount of more than 50% of a lack of information can still be observed regarding RCA
the total generated wastes, CDW combined to an excessive concrete durability performance and its use in blended cement
concrete. This paper presents various properties of concretes
extraction of nonrenewable mineral resources is largely
made with portland cement (PC) and portland cement silica
contributing to a permanent environmental degradation.
fume (PCSF), designed with different proportions (30, 50,
Over the last decades, the amount of CDW has increased
and 100%) of the RCA obtained from different sources
considerably in line with increased construction activities of parent concrete. An appropriate experimental program
and due to the demolition and restoration of old buildings. was undertaken to explore the feasibility and potential of
Meanwhile, the substantial demand of new constructions
are recorded, which leads to an extensive use of natural
resources, especially natural aggregate (NA), as it represents ACI Materials Journal, V. 109, No. 1, January-February 2012.
approximately 70% of the total volume of concrete. It is MS No. M-2010-320.R2 received July 7, 2011, and reviewed under Institute
publication policies. Copyright © 2012, American Concrete Institute. All rights
believed that the taxes on landfilling and quarrying nonre- reserved, including the making of copies unless permission is obtained from the
newable materials, a shortage of raw materials, and a reduc- copyright proprietors. Pertinent discussion including author’s closure, if any, will be
published in the November-December 2012 ACI Materials Journal if the discussion is
tion in the dependency on primary materials and the cost of received by August 1, 2012.
attached mortar on the RCA. Whereas RCA has more fine Oven-dry 2.44 2.51 2.34
particles than NA, both aggregate types are within the limit
of grading for coarse aggregate set by BS 882 and EN 12620, Surface-dry 2.46 2.54 2.43
as shown in Fig. 1. Unit volume weight, kg/L — 1.48 1.28
Water absorption, % 0.77 1.4 5.3
Mixture proportions and concrete mixing
All concretes investigated were designed following Solid content — 59.1 56.8
BS 8500 requirements. For mixtures made with PC, four Shape — Round Spherical
concrete grades were tested: GEN 3, with a nominal 28-day Shape index, % — 20 to 22 14.4 to 16.2
design compressive strength of approximately 20 MPa
(2900.75 psi); RC 30; RC 35; and RC 40, with a nominal Flakiness index, % 13 to 16 7 to 9
28-day design compressive strength of 30, 35, and 40 MPa Fineness modulus 2.4 7.1 6.7
(4351, 5076.3, and 5801.5 psi), respectively. For mixtures
Surface texture — Smooth Rough and porous
containing SF, only two concrete grades were investigated:
RC 30 and RC 35. Table 4 provides the mixture proportions Mechanical properties
of all concretes investigated. For both PC and PC/SF concrete Crushing value, % — 12.4 23.4
types, NA was substituted by RCA at three replacement
Impact value, % — 6.3 to 7.3 18.3 to 23
ratios of 30, 50, and 100% (by mass). For practicability,
the RCAs have been used in surface-dry conditions and as 10% fines value, kN — 155 131
received from the recycling plant. Notes: 1 kN = 0.225 kips; 1 kg/m3 = 1.686 lb/yd3; 1 g/cm3 = 0.0361 lb/in.3; 1 mm =
0.0394 in.
As the densities of NAs and RCAs were different,
the actual amounts of fine and coarse aggregates in the
mixtures were slightly different, whereas the effective
water content was kept constant at 180 kg/m3 (303.5 lb/yd3)
for all the mixtures. The total mixing water was adjusted
to account for NA and RCA moisture content. To meet
BS 8500-2 requirements for the minimum compressive
strength of each concrete grade and to achieve a comparable
compressive strength for concrete made with RCA, the w/c
of 50 and 100% RCA concrete was reduced by an increase
in binder content.
Concrete were produced in a laboratory concrete mixer and
each mixture was appropriately labeled, as shown in Table 4.
Aggregates (NA, RCA, and fines) were first put in the mixer
and dry mixed for 1 minute, followed by the addition of
cement/SF, and mixed for another 1 minute. Water was then
added, followed by an additional mixing period of 3 minutes.
Immediately after mixing, fresh concrete properties were
determined using slump and compacting factor (CF) tests; Fig. 1—Grain size distribution of aggregates used. (Note:
the results are given in Table 4. Generally, the slump value 1 mm = 0.0394 in.)
varied between 20 and 120 mm (0.79 and 4.72 in.), and the
air content was limited to the entrapped air with values of
approximately 1.5 ± 0.5%. prior to exposure in specified conditions, unless otherwise
stated. The main curing regimes adopted for different tests
Specimen casting and curing are given in Table 5.
Concrete was cast in metal molds (cubes, cylinders, and
prisms) in three layers and compacted with a plate vibrator, Mechanical properties
as specified by BS 1881:Part 108:1983. Afterwards, all Cubes measuring 100 x 100 x 100 mm (3.94 x 3.94 x
concrete specimens were kept in their molds in a laboratory 3.94 in.) were used for compressive strength and tested at 3,
environment at 20 ± 2°C (68 ± 6.8°F) and 55 ± 5% relative 7, 14, 28, 56, 91, and 365 days, as specified in EN 12390-3.
humidity (RH) for the first 24 hours. Specimens were then Cylindrical specimens measuring 150 x 300 mm (5.9 x
demolded and cured under the specific curing conditions 11.81 in.) were used to measure the modulus of elasticity
adopted (Table 5) prior to testing. at 28 days in accordance with BS EN 1352:1997. Flexural
strength was measured under four-point loading at 28 days
Testing procedures in accordance with EN 12390-5 using 100 x 100 x 500 mm
Three specimens from each concrete type were tested (3.94 x 3.94 x 19.7 in.) prisms. For compressive and flexural
at different ages and the average value was recorded. All testing, the specimens were loaded during the testing at a
specimens were cured underwater at 20 ± 2°C (68 ± 6.8°F) constant rate until failure.
Expansion and drying shrinkage were measured on 75 x the CU3 regime for 14 days. Afterwards, all sides of the
75 x 300 mm (2.95 x 2.95 x 11.8 in.) prisms with stainless specimens were sealed using a bituminous coating paint,
steel points fixed on two sides of the specimen. The samples except the top side of the specimen, which was exposed (in a
were stored under the CU2 and CU3 curing regimes, respec- carbonation tank) to a CO2-enriched atmosphere contfaining
tively. For both tests, the specimens were continuously cured 3.5% of CO2 at 20 ± 2°C (68 ± 6.8°F) and 60 ± 5% RH. The
for 90 days and a length change of the different sides of the depth of carbonation in the tested concrete was measured by
specimens was measured using a digital strain gauge.
applying a phenolphthalein color indicator spray on a freshly
Durability performance broken piece of the specimens after 2, 4, 8, 12, and 20 weeks
Carbonation—An accelerated carbonation test was carried of exposure. This turns noncarbonated concrete pink and
out on cubes measuring 100 x 100 x 100 mm (3.94 x 3.94 x remains colorless in carbonated concrete. The depth of the
3.94 in.) that were cured during the first 28 days under the uncolored zone of the concrete (the carbonated layer) from
CU2 regime. The specimens were then conditioned under the edges of the broken piece was measured at five points,
from 290 to 810 me and 150 to 195 me for the PC and PC-SF
mixtures, respectively. Drying-shrinkage strains increased Table 7—Total magnitude of shrinkage after
with the increase of RCA contents in PC mixtures, whereas 90 days of dry curing
no significant changes were observed in the PC-SF concrete
mixtures. Moreover, for the PC mixtures, the general trend Strains (× 10–6)
suggests that up to 30% RCA has no major effect on the
Mixture code RCA, % PC PC/SF PC PC/SF
drying shrinkage, and shrinkage significantly increased
0 290 125
with the increase in RCA content. As could be expected,
the highest shrinkage magnitude was exhibited by the PC 30 320 110
GEN3 — —
mixtures made with 100% RCA and the lowest shrinkage 50 450 100
was exhibited by the control mixture. Therefore, the greater 100 650 60
the amount of RCA added to the PC mixture, the greater the 0 340 195 100 220
magnitude of drying shrinkage. Such a large magnitude of
30 340 160 120 340
drying shrinkage in RCA concrete can be attributed to the RC30
50 520 170 80 430
high porosity and water absorption capacity of RCA, as well
as to the low modulus of elasticity of the treated concretes. 100 630 155 80 760
Additionally, RCA concretes have a larger volume of paste 0 280 190 120 265
and a lower content of aggregate skeleton to counteract the 30 320 180 130 235
development of shrinkage strains when considering the RC35
50 425 150 130 435
additional old mortar that is adhered to the RCA compared
100 810 160 140 730
to the PC concrete with NA. As a result, shrinkage of the PC
concrete made with RCA could be considered as the sum
of shrinkage strains of the old mortar attached to the RCA, PC-SF mixtures. As carbonation is a function of concrete
plus shrinkage of the new hydrated PC paste, leading to a quality, in particular the cement content, the w/c, and the
substantial ultimate magnitude of shrinkage. porosity and moisture content—the higher the pore volume
(low strength), the higher the carbonation rate of concrete.
Carbonation The results presented in Fig. 5 to 7 showed that the depth of
The measured carbonation depths through the top side of carbonation values decrease when the compressive strength
the tested specimens exposed to CO2-enriched environment of concrete increases. The largest depth of carbonation
in the carbonation chamber for a period of 20 weeks are ranged between 20 and 25 mm (0.79 and 0.98 in.) recorded
shown in Fig. 5 to 7. For both PC and PC-SF, the depth of for the GEN3 PC concrete (Fig. 5), whereas the lowest
carbonation seems to be proportional to its compressive carbonation depth ranged between 4 and 8 mm (0.16 and
strength. The higher the compressive strength of concrete, 0.31 in.), exhibited by the RC40 PC concretes (Fig. 7).
the lower the values of the depth of carbonation. Although no The carbonation depth is also intimately linked to the pore
consistent trend between carbonation depth and RCA content system characteristics. A large pore size and open porosity
was observed, it could be seen, generally, that the addition led to a greater carbonation depth, as could be expected for
of RCA has increased the carbonation rate in both PC and the GEN3 PC mixtures.
Fig. 6—Rate of carbonation versus exposure time (RC30 PC Fig. 7—Rate of carbonation versus exposure time (RC40 PC
concrete grade). (Note: 1 mm = 0.0394 in.) and RC35 SF concrete grade). (Note: 1 mm = 0.0394 in.)
As moisture is needed for the carbonation process to be trend suggests that the chloride ion concentration is still at
processed, the increased rate of carbonation in the RCA moderate level with the use of 30% RCA and quite similar
mixtures might be attributed to the high porosity and to the to that found in the control NA concrete, whereas beyond
high moisture content of RCA due to the water that might 30%, a substantial increase of the chloride concentration is
be stored in its pore system. On the other hand, it seems that recorded with an increase in RCA content. The results have
the use of SF in RC35 PC-SF mixtures did not improve the shown that chloride concentration is proportional to concrete
resistance to carbonation of concrete because the rate and strength. Generally, the increase in concrete strength has
depth of carbonation of these mixtures is somewhat higher led to a reduction in chloride concentration (Fig. 9 and 10).
than that of corresponding PC mixtures. The variety of RCA, Whereas the results of chloride ions for the PC-SF concretes
in terms of sources and properties, could be the main reason presented in Fig. 8 showed some anomalies because the
for these anomalies. It should also be noted that for the same PC mixtures have exhibited relatively higher resistance to
concrete types (PC), the carbonation rate develops quickly chloride penetration than the PC-SF mixtures, the general
over time in GEN3 PC mixtures compared to RC30 and trend suggests that using SF may improve the resistance to
RC40 PC mixtures. Thus, the rate of carbonation decreases chloride ingress of RCA concrete (Fig. 9), as has also been
as the compressive strength increases. reported previously.23 The higher chloride concentration
found in RC30 PC-SF mixtures is mainly due to the high
Chloride ion penetration w/c of these mixtures compared to their corresponding PC
The chloride concentration of the specimens measured mixtures (absence of HRWRA), as well as the large variety
at different depths and immersed for 42 days in the in RCA sources and properties.
chloride solution are shown in Fig. 8 to 10. For both PC As pointed out previously, pore characteristics and
and PC-SF at a given depth from the exposed surface, concrete strength are the two major parameters governing
chloride concentration seems to increase with an increase carbonation, as well as the resistance to chloride ingress.
in RCA content, especially at a depth of 5 to 10 mm Enhancing concrete strength by lowering the w/c and using
(0.196 to 0.39 in.) near the exposed surface. The general SF could significantly contribute to decreasing the porosity
Fig. 9—Rate of chloride ingress versus depth (RC35 PC and RC35 SF). (Note: 1 mm =
0.0394 in.)
Sulfate attack
A sulfate attack test was carried out on a single concrete
type (GEN3); the results obtained are presented in Fig. 11. It
could be seen that the expansion recorded in test specimens
increased with an increase in exposure duration. The
expansion induced by the sulfate solution was also found to
increase with an increase in the RCA concrete mixture. The
full replacement of NA by RCA has exhibited the highest
expansion of approximately 80 me after 2 months of exposure
time. In fact, the expansion induced by the penetration of
dissolved sulfate into the concrete via the pore network
induces changes in the composition and microstructure of
concrete. The transformation of the monosulfate phase into Fig. 11—Expansion due to sulfate attacks for GEN3 PC
ettringite led to the expansion observed in Fig. 11. Such an concrete made with RCA.