Cement and Concrete Research 32 (2002) 1301 – 1311

Porosity of recycled concrete with substitution

of recycled concrete aggregate
An experimental study
José M.V. Gómez-Soberón*
3 Sur #503 Cd. Serdán, Puebla, CP 75520, Mexico
Received 8 June 2001; accepted 5 March 2002

Abstract

In this paper, we present the experimental analysis of samples of recycled concrete (RC) with replacement of natural aggregate (NA) by
recycled aggregate originating from concrete (RCA). The results of the tests of mechanical properties of RC were used for comparison with
tests of mercury intrusion porosimetry (MIP), in which the distribution of the theoretical pore radius, critical pore ratio, the surface area of the
concrete, threshold ratio and average pore radius were studied at ages of 7, 28 and 90 days. The results showed some variation in the
properties of the RC with respect to ordinary concrete. Porosity increases considerably when NA is replaced by RCA. Additionally, a
reduction in the mechanical properties of the RC is seen compared with ordinary concrete when porosity increases. D 2002 Elsevier Science
Ltd. All rights reserved.

Keywords: Compressive strength; Porosity; Recycled concrete; Recycled concrete aggregate; Young’s modulus

1. Introduction Tests of mercury intrusion porosimetry (MIP) in RC

show increases in the total volume of pores, especially in
Current studies of recycled concrete (RC), with partial larger pores ( > 100 nm). This suggests that this property is a
substitution of natural aggregate (NA) by recycled aggregate function of both the age of the concrete and the amount of
originating from concrete (RAC), promise a feasible path mortar that the RCA contains [12].
for its application, thus satisfying a need, saving energy, Five types of concrete are presented in this work with
improving environmental conditions and providing a solu- different RCA contents, together with the behavior of the
tion for the 200 million tons/year of construction waste concrete that provided the RCA, and were prepared for
generated in the EU [1 – 7]. the study of the properties of both the aggregates and
RCs can be regarded as porous concretes, having per- the concretes.
meability values that are twice those of ordinary concretes.
Their general behavior shows a decrease in their mechanical
and physical properties as the percentage of replacement of 2. Experimental
NAs by RCA increases [8]. It is also widely accepted that
there is an inverse correlation between the volume of pores 2.1. Original concrete
and the levels of stress to which the concrete can be
subjected, and this relationship should also take into account In this study, 4 m3 of original concrete (OC) was used.
the distribution of the sizes of the pores and the intercon- The concrete was made in a mixer and was poured into
nection between them [9 –11]. wooden formwork frames measuring 0.40
0.20
0.10 m.
Fifty cylinders measuring ; 0.15
0.30 m, eight measuring
* L’Hospitalet de Llobregat, Av. Carrılet #338 5-3, Barcelona 08901
; 0.15
0.45 m and four cubes measuring 0.10
0.10 m
Spain. Tel.: +34-933-37-05-69; fax: +34-934-01-72-62. were also used to study the porosity, mechanical behavior,
E-mail address: [email protected] (J.M.V. Gómez-Soberón). shrinkage and creep.

0008-8846/02/$ – see front matter D 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 8 - 8 8 4 6 ( 0 2 ) 0 0 7 9 5 - 0
1302 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311

Table 1 than the NA (with an average of 14% less in Ds and 9% less

Mixtures used for OC
in Dsss) The RCA shows an increase in density, which is
Component OC directly proportional to the greater particle size. Finally, the
3 a
Cement (kg/m ) 380 differences between dry and dry surface saturated conditions
Water (kg/m3) 168 are greater for the RCA than for the NA.
Fine gravel (5 – 12) (kg/cm3)b 252
The RCA used in this study can be considered as being
Gravel (12 – 20) (kg/cm3)b 773
Sand (0 – 5) (kg/m3)b 784 within the RILEM recommendation for Type II RCA
W/C 0.44 (absorption  10% and Ds  2000 kg/m3). For the Belgian
Coarse A/Fine A (vol) 1.3 recommendation, they are GBSBII (absorption < 9% and
Additives (plastifier) 2.69 Ds > 2100 kg/m3) and in the Japanese case they comply with
a
CEM I 42.5 R. the absorption requirement ( 7% and Ds  2200 kg/m3) in
b
Limestone aggregate, Garraf quarry, Barcelona. the fractions used [13 – 16]. Consequently, the RCA
employed in this study may be used in both plain and
Twenty-four hours after pouring, the samples were reinforced concrete if its application and factors of behavior
removed from the formwork and submitted to a curing are taken into account.
process for 150 days (see Table 1, in which the specific
characteristics of this concrete are given). The specimens 2.3. Mix of RCs
were then passed once through a semifixed roller grinder
with an inlet width of 0.45 m and a maximum outlet size of Due to the difficulty in determining the real water/
0.025 m. Finally, the resulting material was classified into cement (W/C) ration because of the high variation of
sizes (mm in all cases): 0 – 5, 5 – 10, 10 – 20, 20 –25. The 5 – absorption in the RCA, it was decided to use basic ACI
10 and 10 – 20 fractions were used as RCA in this work. 211.1 and ACI 211.2 mix concepts in accordance with the
following criteria.
2.2. Recycled aggregate and NA (1) The substitution of NA by RCA was done using equal
volume fractions with the following condition:
The designation used by sizes was: for RCA, gravel 10 –
20 and fine gravel 5 –10; and for the NA, gravel 12– 20 and r ¼ RCAcoarse =ðRCAcoarse þ NAcoarse Þ
fine gravel 5 – 12.
The criterion used for this fit was the compacted max- ð0:00  r  1:00Þ
imum density (which reduced the possible influences of
different particle size). These were: where: r =percentage of NA replaced by RCA, by volume;
RCAcoarse = 55% recycled gravel + 45% recycled fine
 For RCA, the combination was 55% gravel and 45% gravel; and NAcoarse =70% natural gravel +30% natural
fine gravel. fine gravel.
 For NA, the combination was 70% gravel and 30% The percentages of the five studied samples of RC were:
fine gravel. r = 0.00, 0.15, 0.30, 0.60 and 1.00. As fine aggregate, 100%
crushed natural limestone sand from the Garraf quarry
Table 2 shows the properties of the aggregate used. The (Barcelona) was used.
total porosity to water is the variable that shows the greatest (2) The RCA showed an increase in absorption propor-
difference between the RCA and NA and, in the worst case, tional to the time spent in water. The time allowed for the
reaches 2.82% for the NA and 14.86% for the fine gravel mixture was 20 min of immersion, with up to 97% fine gravel
fraction of the RCA. As regards density, the RCA is lighter and 77% gravel, in all cases with comparison after 24 h.

Table 2
Properties of recycled aggregate and NA
RCA NAa
Property 10 – 20 5 – 10 0–5 12 – 20 5 – 12 0–5
Dry specific gravity (kg/m3) 2280 2260 2170 2570 2640 2570
Specific gravity (surface dry) (kg/m3) 2410 2420 2350 2590 2670 2600
Water absorption (%) 5.828 6.806 8.160 0.876 1.134 1.49
Total porosity (%) 13.42 14.86 – 2.70 2.82 –
Shape coefficient 0.363 0.466 – 0.364 0.576 –
Longs indices 6 15 – 8 19 –
Modulus of fineness 7.2 6.2 3.8 6.9 5.0 3.3
Sand equivalent (%) – – 93.6 – – 93.8
Particles < 200 mm (%) 0.06 0.29 9.85 0.50 2.46 9.24
a
Limestone aggregate, Garraf quarry, Barcelona.
 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311 1303

Table 3 2.4. Properties of the concretes

Mixtures used for RC
Component r = 1.00 r = 0.60 r = 0.30 r = 0.15 r = 0.00 The tests on the different concretes comprised the study
Cement 400 of the physical properties such as absorption, density,
(kg/m3)a porosity and water permeability; and mechanical properties
Water 207.6
such as compression, tensile strength, Young’s modulus,
(kg/m3)
RCA Fine gravel 406 258 134 69 0 shrinkage and creep.
(kg/cm3) (5 – 10) Tests of the physical properties of the concrete were car-
Gravel 497 315 164 84 0 ried out on 0.10
0.10 m cubic samples, while the mecha-
(10 – 20) nical tests were done on ; 0.15
0.30 m (compression,
NA (kg/cm3) Fine gravel 0 268 488 604 710
tensile strength and Young’s modulus) and ; 0.15
0.45 m
(5 – 12)b
Gravel 0 115 209 259 304 cylindrical (shrinkage and creep) samples with ages of 7, 28
(12 – 20)b and 90 days, and 179, 200 and 262 for RC. Tables 4 and 5
Sand (0 – 5) 662 show the results of the tests, in which the Spanish UNE
(kg/m3)b [21 –33] and ASTM standards were used. Each value in the
W/C 0.52
table is the average of two tests for physical properties and
Coarse A/Fine A (vol) 1.53
a
an average of three for the mechanical tests (two for
CEM I 52.5R UNE 80 301 96 RC/97.
b shrinkage and creep).
Limestone aggregate, Garraf quarry, Barcelona.
The sections below provide a summary of the behavior of
the physical and mechanical properties.

(3) The amount of water absorbed by the aggregate was 2.4.1. Physical properties
taken into account separately, in addition to its wetness The absorption of the RC increases proportionally with
before mixing and the free water that formed part of the RCA content, while the density decreases slightly. Water
mixture. The above aspect is justified by criteria that were porosity, like absorption, increases proportionally with RCA
emphasized in a previous publication of the authors [17 – 20]. content. The above comments are presented in Table 4.
With the established mixing time and the required
amount of water, the order of mixing the materials guar- 2.4.2. Mechanical properties
anteed (as far as possible) the immobility of the water and Simple compression decreases as the r factor increases
an improvement in the transition zone. The following for the studied ages: when r 0.30, fc0 is appreciably the
sequence was adopted: (a) all of the coarse aggregates and same as the reference concrete, and if the evolution of fc0 is
water were introduced into the mixer; (b) these were mixed compared with the age of the RC, it is seen that its behavior
for 2 min; (c) the mixer was switched off for 3 min; (d) is the same as the reference concrete, although the stress
stages (b) and (c) were repeated twice; (e) the cement was levels are of course lower. Indirect stress shows a similar
introduced and mixed for 3 min; and (f) the sand was added evolution with age as with ordinary concretes. However,
and mixed for another 3 min. samples with r = 0.60 and r = 1.00 show tensile strength
The mixes obtained using the above criteria are given in values that are appreciably lower than the rest of the
Table 3. As can be seen, the variation in consistency and studied concretes for the age range under study. Finally,
volumetric weight for the different percentages of aggregate comparing results without reference to the age of the test,
replaced is within tolerable limits (slump 0.1 ± 0.03 m and Young’s modulus shows its minimum value for r= 0.60,
concrete with volumetric weight normal). closely followed by the value for r= 1.00. This property,

Table 4
Mechanical and physical properties of RC

Agea Tensile strength (MPa) Compressive strength (MPa) Young’s modulus (GPa) Absorption Water porosity Ds Dsss
factor 7 28 90 7 28 90 7 28 90 (%) (%) (kg/m3) (kg/m3)
r = 0.00 3.6 3.7 3.9 33.3 39.0 42.1 27.6 29.7 32.4 8.40 18.0 2130 2310
r = 0.15 3.3 3.7 3.9 33.9 38.1 41.6 27.2 291 30.1 8.60 18.5 2140 2360
r = 0.30 3.3 3.6 3.9 34.8 37.0 39.5 26.5 27.8 29.4 8.60 18.5 2150 2330
r = 0.60 3.2 3.4 3.7 30.6 35.8 38.3 25.5 26.6 27.6 9.00 19.2 2120 2320
r = 1.00 3.5 3.3 3.6 30.7 34.5 37.5 26.9 26.7 26.4 9.60 20.1 2090 2290
OC 3.2 3.8 – 35.2 38.4 – 33.0 33.7 – 5.90 13.4 2270 2410
OCb 4.1 4.1 4.2 45.1 45.4 47.0 35.2 34.5 34.6
a
Days.
b
At 172, 179, and 262 days of age.
1304 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311

Table 5
Shrinkage and creep of RC
Component properties r = 1.00 r = 0.60 r = 0.30 r = 0.15 r = 0.00 OC
fc0(28 days) (MPa) 34.5 35.8 37.0 38.1 38.8 38.4a
Stress level for creep (MPa)b 12.08 12.53 12.95 13.34 13.58 13.44
esh basic (mm/m) 0.0138 0.0310 0.0040 0.0800 0.0220 0.0260c
esh drying (mm/m) 0.4029 0.4104 0.3524 0.3763 0.3740 0.1940c
ec basic Instantaneous (mm/m) 0.1370 0.1470 0.1645 0.1350 0.1430 0.1180c
j (90days, t0) 0.85 0.72 0.55 0.81 0.44 0.34c
ec drying Instantaneous (mm/m) 0.1580 0.1530 0.1380 0.1350 0.1600 0.1265c
j (90days, t0) 4.04 3.85 3.65 3.55 2.90 1.67c
a
The age of the test was 200 days.
b
0.35fc0.
c
Shrinkage (262 and 172 days) and j (262 and 172 days).

therefore, shows a similar behavior to ordinary concretes which covers this spectrum to a great extent. It was applied
(see Table 4). using the ASTM D 4404 standard [34] and the concepts
For the shrinkage and creep tests, the samples, after expressed below.
28 days in a curing chamber (T = 20 ± 2 C and RH =
90 ± 5%), were submitted to a climatic chamber (T= 20 C 2.5.1. Procedure
and RH = 50%) for 90 days. The specimens used for basic The tests were carried out at 7, 28 (samples of concrete
shrinkage and creep measurement were sealed with paraffin submitted to curing chamber conditions to correlate them
(± 0.03 m thick) and wrapped in three layers of aluminum with physical and mechanical properties of RCs) and 90 days
foil. The details of these tests have been published by the (samples of concrete submitted to climatic chamber condi-
authors [17,18]. tions to correlate them with shrinkage and creep). The
Table 5 presents the results of this experimental cam- samples for the MIP tests were extracted from the center of
paign. This table shows the increase in the strain due to the samples (0.10
0.10 m) using a core catcher with a
shrinkage when the RCA increases its RC content. Sim- diamond bit. The resulting cylinder (; = 0.02 m
h= 0.10 m)
ilarly, from these results, it is concluded that the creep was then cut with a fine saw to obtain two central cylinders
coefficients report a direct correlation with the increase of with dimensions of ;= 0.02 m
h= 0.03 m.
the factor r (principally in shrinkage and creep for drying). The samples for tests at 7 and 28 days were dried by
stages: firstly, the samples were dehydrated by submerging
2.5. Porosimetry by mercury intrusion them in alcohol for 8 days. The alcohol was changed every
24 h. After this stage, the samples were dried by putting
Due to the broad pore size spectrum of the concretes them in an oven at T=70 C for 24 h. Finally, the samples
(1 nm – 1 cm), it was decided to use the MIP technique, were put into a drier until the tests.

Fig. 1. Distribution in the black test.

 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311 1305

Fig. 2. General distribution (age 7 days).

The samples for the tests at 90 days, after 28 days in the The following constants and hypotheses were applied
curing chamber, were put in the climatic chamber until the in this study (for all the samples): mercury (Hg) contact
age of 90 days. Upon arriving at this age, the samples were angle with the concrete q = 130, Hg surface tension
extracted and dried by putting them in an oven at T= 120 C s = 0.480 N/m 2 , pore shape factor j = 4.00, together
for 24 h, and then in a drier. with the typical hypotheses of the methodology of this
The MIP tests were done on a Quantachrome Autoscan test [35,36].
33 porosimeter, which subjected the samples to a maximum
pressure of up to 226 MPa. The following parameter was 2.5.2. Corrections applied to the test values [37]
measured: the theoretical pore radius (rpt), within the range The test values obtained based above were submitted to
27 –59,000 Å (the Washburn equation was used for the three corrections to improve their accuracy.
calculation). The process included the typical intrusion and
extrusion cycles in this type of test. Finally, to improve the 2.5.2.1. Blank test. A test with the dilatometer without a
profile of the test curve, the sample results were filtered sample was carried out to obtain the volume of Hg that
using a moving average with Base 9. causes the dilatometer to expand, and the compression that

Fig. 3. General distribution (age 28 days).

1306 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311

Fig. 4. General distribution (age 90 days).

the Hg undergoes to fill the dilatometer. The following above variables are taken into account, it is possible to
00
correlation was obtained from this test (see Eq. (1)): determine VcHg using the following equation (Eq. (2)):

Vblank ¼ 2E07 x2 þ 7E05 x  0:0049: ð1Þ 00

VcHg ¼ 0:175BVsample log10 ð1 þ ðx=1820 MPaÞÞ: ð2Þ
The result of the above correlation (Vblank) was subtracted
from the uncorrected Hg volume of the tests (Vo) when the 2.5.2.3. Compression of the sample. This estimates the
test pressure (x) was greater than or equal to 92.56 MPa. The compressibility factor of a volume of the sample that was
blank test and the proposed correlation are shown in Fig. 1. not intruded (Vcs). Given the compressibility coefficient of
the material (ysample, proposed =5.0 E10 m2/N) and the
2.5.2.2. Differential compression of Hg. The sample dis- volume of the sample without intruded Hg (UVsample), it is
places a volume of Hg equal to the mass of the sample feasible to calculate the compression of the sample using the
(BVsample). The blank test, therefore, includes the compres- following equation (Eq. (3)):
sion of a volume of mercury (equal to BVsample) that is not
seen in the volume of the experimental test (VcHg 00
). If the Vcs ¼ xysample UVsample : ð3Þ

Fig. 5. Total porosity.

 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311 1307

Table 6
Critical radius for the different RCs
rcritical (nm)
dv/dp VS pore radius aa > 10 10 < aa < 50
Method factor 7 days 28 days 90 days 7 days 28 days 90 days 7 days 28 days 90 days
b b
OC 44.70 – – 63.64 – – 57.34 – –
r = 0.00 40.31 51.61 38.53 70.39 83.93 62.67 59.76 68.75 51.94
r = 0.15 36.50 73.63 40.60 65.17 68.68 57.26 56.21 47.63 46.46
r = 0.30 53.88 91.62 40.96 65.50 69.50 67.04 60.15 60.98 56.95
r = 0.60 44.45 51.00 48.75 70.01 79.48 97.39 60.24 55.78 93.36
r = 1.00 45.87 49.63 43.68 68.78 74.29 78.78 58.40 65.66 64.48
Maximum 53.88 91.62 48.75 70.39 83.93 97.39 60.24 68.75 93.36
Minimum 36.50 49.63 38.53 63.64 68.68 57.26 56.21 47.63 46.46
Average 44.28 63.50 42.50 67.25 75.71 72.63 58.68 59.76 62.64
a
Angle between two consecutive points (pore radiusvolume intruded Hg).
b
At 262 days of age.

The final equation that was used to calculate the porosity at stability and influence the mechanical properties of the
each point of the tests according to the above corrections concretes [38].
was the following (see Eq. (4)):
2.5.3.2. Critical pore radius. Characteristic radius or crit-
00 ical pore (rcritical) is the term given to the corresponding
Vintrusion ¼ Vo  Vblank þ VcHg  Vcs : ð4Þ
radius that causes the beginning of the maximum slope in
the curve of the radius versus volume of the intruded Hg
2.5.3. Results graph. This pore radius is usually an indicator of the
The following sections present the results and analyses microstructure of the material and it is used to detect a
derived from the MIP tests performed on the various variety of materials. For the determination of this parameter,
concretes in the study. two different methods were used as follows:
(i) Calculate the angle between two consecutive points
2.5.3.1. Distribution of the pore radius. Figs. 2– 4 show (a) of all the values that take in all of the curves in the study.
the graphs for the tests carried out at 7, 28 and 90 days for The criteria for establish distinction were the following: (a)
the various concrete samples. To avoid confusion when pore radius that causes the first angle that is greater than 1
interpreting the graphs, only the curves of the Hg intrusion of elevation in the slope of the curve under study; and (b)
stages are presented. the average pore radius range that generates an angle of
It may be seen from the three graphs that the increase in elevation between 1 and 5 of the beginning of elevation of
the r factor of the RCs shows a correlation with total the curve under study.
porosity, as the latter increases by 5.9% when r goes from (ii) Point maximum peak of the named curve of frequen-
r= 0.00 to r= 1.00 at 7 days, 2.3% for 28 days and 3.8% for cies density dv/dp (cm3/MPa cm3) VS pore radio (nm).
90 days (taken from the three graphs as average values of
two tests for each variable).
It is also seen in Fig. 5 that the total porosity
decreases appreciably for all the concretes as a function
of age. The evolution of total porosity of these concretes
amounts to a decrease of 0.5% from 7 to 28 days with
the exception of r= 1.00, which shows a greater reduction
(3.0%). Total porosity drops by 0.42% on average upon
going from 28 to 90 days, the maximum standing at
r= 1.00 (1.9%).
It should be pointed out that the main difference in the
total porosity between the different RCs (r= 1.00 and the rest
of the concretes) is seen in the area with pore radius less
than 30 nm (the zone of maximum percentage of pore
volume in the concretes and originating in the cement
mortar). It is also in this zone that the decrease in porosity
with age is seen. The above pore range is commonly
associated with the power to damage the dimensional Fig. 6. Specific surface area for RC.
1308 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311

Fig. 7. Maximum pore radius.

Table 6 shows the critical pore radius calculated as 2.5.3.3. Specific surface area. The MIP technique ena-
explained above. As is seen, the values in the tests are bles the determination of the surface area of the material
approximately 36 – 97 nm (with a population average of from direct sampling of the pressure versus volume of
60.46 nm), which locates these pores in the range of large intruded Hg (Fig. 6). The method does not require
capillary or macropores (50 –10,000 nm); at this capillary previous knowledge of the geometry of the pore, which
range, negative effects that can affect the strength and the means that its results are reliable. However, the maximum
impermeability of a concrete can be attributed to the pore level of available pressure of the porosimeter used cannot
size. However, the difference between the different rcritical of guarantee intrusion of all the pore sizes that cover the
the concretes studied is negligible as the variation between specific surface of the samples and it is not possible to
them is small. With respect to the two methods for its evaluate whether high levels of pressure cause crushing of
determination, the most exact is critical pore radius; how- sealed pores.
ever, Case (i,b) is of greater importance due to the ease with The observed general behavior is an increase (import-
which it determines rcritical and the fact that it presents a mean ant for the case of r = 1.00) in agreement with the
variation with respect to critical pore radius of only 15%. increase of the r factor. For this property, the results do

Fig. 8. Average pore radius.

 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311 1309

Table 7
Correlation between total porosity and other properties of the RCs
Property Equationa R2
Absorption (%) y = 0.3501x + 4.0460 .7641
Porosity to water (%) y = 0.5714x + 11.036 .7000
Ds (t/m3) y = 0.0165x + 2.3514 .7267
Dsss (t/m3) y = 0.0191x + 2.5836 .7745
Permeability (cm)
Average y = 0.1564x  1.3955 .7652
Maximum y = 1.0006x  1.3819 .6076
ft (MPa) y = 0.0185x2  0.6527x + 8.9077 .5727
fc0 (MPa) y = 0.3744x2  12.906x + 141.05 .7350
E (GPa) y = 0.1984x2  6.395x + 77.316 .6510
esh drying (mm/m) y = 2E08z2  9E05z + 0.4607 .8367
j (90 days, t0) y = 3E07z2 + 0.0019z + 1.0833 .9347
Fig. 10. Total porosity versus compressive strength.
a
x = total porosity in (%) MIP; z = pore threshold (nm).

not present a logical correlation with the age of the speci- 2.5.4. Correlation between porosity and properties of the
mens, which maybe related to the fact that the samples were concretes
submitted to two types of environmental conditions and two The correlation process was carried out on the properties
drying processes. obtained from the MIP tests, the physical and mechanical
properties of the RCs (7 and 28 days), and shrinkage and
2.5.3.4. Pore threshold. This name is given to the max- creep (90 days). The processed data were examined with
imum pore radius found (either in the intrusion or extrusion different lines of tendency, in which different correlation
stage) in the MIP test [39]. Fig. 7 shows the maximum coefficients (R2) were used as indicators of greater recipro-
radius for the concretes studied. city between the variables.
Correlations for the three curves are fairly high. How- Table 7 shows the correlation equations for the total
ever, the interpretation and usefulness of this parameter can porosity data and the physical properties of the RCs. The
only serve as a guideline, as its determination involves a correlations only include MIP tests for 28 days, as the
high degree of uncertainty. physical tests were done at this time. The table also shows
the correlation for total porosity and mechanical properties
2.5.3.5. Average pore radius. Pore radius (raverage) is the of the RCs (the correlations are an average of tests at 7 and
name given to the radius that corresponds to 50% of the 28 days). Finally, the correlation equations between shrink-
total volume of Hg intruded in the test. In Fig. 8, the raverage age and the creep coefficient with the MIP tests for 90 days
calculated through the Lagrange interpolation of the two are also presented.
nearest points to 50% of the total volume for each sample Figs. 9 – 11 show the graphs that gave rise to the
is presented. correlations of the mechanical properties and the total
The presented behavior of raverage is correlative with the porosity obtained in the MIP tests.
age of the samples, as well as with the r factor (in this last Total porosity, obtained from the MIP tests, is the
case in smaller proportion). property that showed the best correlation for the behavior

Fig. 9. Total porosity versus tensile strength. Fig. 11. Total porosity versus Young’s modulus.
1310 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311

of the studied concretes, followed by the specific surface It would also be advisable to include parameters that
area, rcritical, the pore threshold radius and, lastly, raverage. evaluate the formation and propagation of microcracks in
As regards the types of lines of tendency used, the the interface zones, as these zones (mechanically weaker
physical properties follow the pattern of linear correlation than the mortar) increase in the RCs due to the replacement
better than the mechanical properties, while the mechanical of NA by RCA.
properties fit quadratic-type equations better. This latter fact
may indicate that these equations could be improved if
other factors (in addition to total porosity) of similar Acknowledgments
importance are taken into account, such as the increase in
the interface zones (which favors the formation and pro- The author would like to express their gratitude to the
pagation of microcracks) and a major variation in the company Cervezas Finas de Cd. Serdán, Puebla, Mexico,
distribution of the pore radius (especially in the zone of for the partial financing of this research; to the Technical
pore radius < 30 nm). In both cases, the cause of these new University of Catalonia, Barcelona, Spain, for the use of
factors to be considered in RCs is the old mortar that their facilities; and to Drs. L. Agulló and E. Vázquez.
adheres to the NA of which the RCA is composed.

References
3. Conclusions
[1] J. Bjfrn-Jakobse, M. Elle, E.K. Lauritzen, On-site use of regenerated
demolition debris, in: Y. Kasai (Ed.), Demolition and Reuse of Con-
Based on the research and results presented in this paper, crete and Masonry, Reuse of Demolition Waste, vol. 2, E&FN SPON,
the following conclusions are reached. London, 1988, pp. 537 – 546.
. The replacement factor r of the RCs shows a correla- [2] W. Grübl, Environmentally Friendly Construction Technology—Inter-
tion with total volume and pore size, its influence being action Between Construction and Environment (1999) (http:www.b-i-
m-de/public/Tudmassiv/dacon98gruebl.htm).
more important at lower ages and diminishing as the [3] T.C. Hansen, Recycled of demolished concrete and masonry, Report
concrete ages. This influence is attributed to the crystalliza- of Technical Committee 37-DRC Demolition and Reuse of Concrete,
tion of new products that reduce both the number and size Part 1, E&FN SPON, London, 1992, pp. 1 – 160.
of the pores. [4] T.C. Hansen, E. Bfegh, Elasticity and drying shrinkage of recycle-
. The most significant differences of the studied samples aggregate, ACI J. (5) (1995) 648 – 652.
[5] T.C. Hansen, M. Marga, Strength of recycle made from coarse and
are seen in two parameters: (1) the greater pore radius fine recycled concrete aggregates, in: Y. Kasai (Ed.), Demolition and
threshold as the replacement of NA by RCA is increased; Reuse of Concrete and Masonry, Reuse of Demolition Waste, vol. 2,
and (2) the detection of zones of major quantitative changes Chapman & Hall, London, 1988, pp. 605 – 612.
seen in the increase of the pore volume from pores with [6] M. Kakizaki, M. Harada, T. Soshiroda, S. Kubota, T. Ikeda, Y. Kasai,
radius < 30 nm. Strength and elastic modulus of recycled aggregate concrete, in: Y.
Kasai (Ed.), Demolition and Reuse of Concrete and Masonry, Reuse
. The effect of the r factor does not seem to influence the
of Demolition Waste, vol. 2, Chapman & Hall, London, 1988, pp.
rcritical values, and therefore, this parameter may be used 565 – 574.
only in cases in which determination of the variation of [7] E.K. Lauritzen, M. Jannerup, Guidelines and experience from the
structural components of the concrete that was used for the demolition of houses in connection with the resound Link between
RCA is desired, and whose probable cause of variation Denmark and Sweden, Demolition and Reuse of Concrete and Ma-
sonry, in: E.K. Lauritzen (Ed.), Guidelines for Demolition and Reuse
would be the modification of the physical constants of the of Concrete and Masonry, E&FN SPON, Denmark, 1993, pp. 35 – 46.
concretes (W/C, type of cement, etc.). [8] P.J. Wainwright, A. Trevorrow, Y. Yu, Y. Wang, Modifying the per-
. The results for specific surface area, together with total formance of concrete made with coarse and fine recycled concrete
porosity, are those which better describe and correlate the aggregates, in: E.K. Lauritzen (Ed.), Demolition and Reuse of Con-
crete and Masonry, Guidelines for Demolition and Reuse of Concrete
results of the properties of the studied concrete. However, it
and Masonry, E&FN SPON, Denmark, 1993, pp. 319 – 330.
will be necessary to supplement this information with tests [9] A. Müeller, A. Winkler, Characteristics of processed concrete rubble,
of gas absorption, which would enable quantification of the in: K.R. Dhir, N.A. Henderson, M.C. Limbachiya (Eds.), Uses of
surface area in the micropore zone and thus be able to Recycled Concrete Aggregate. Sustainable Construction, Tomas Tel-
better correlate these tests with other properties such as ford, London, 1998, pp. 1109 – 1119.
shrinkage and creep; both are linked to the distribution and [10] P.J. Wainwright, J.G. Cabrera, Use of demolition concrete to produce
durable structural concrete, in: J.J.J.M. Goumans, H.A. Van Der Sloot,
quantity of micropores. T.G. Aalbers (Eds.), Environmental Aspects of Construction with
. Finally, correlation between the properties of the RC
Waste Materials, Elsevier, The Netherlands, 1994, pp. 553 – 562.
and total porosity is difficult to determine. However, it [11] B. Zhang, Relationship between pore structure and mechanical proper-
would be feasible to improve it if the distribution of the ties of ordinary concrete under bending fatigue, Cem. Concr. Res. 28
pore radius was included. This distribution could be ascer- (5) (1998) 699 – 711.
[12] H. Uchikawa, S. Hanehara, Recycling of concrete waste, in: R.K.
tained by classifying boundaries between different ranges of Dhir, T.D. Dyer (Eds.), Concrete for Environment Enhancement
pore size, taking as a classification criterion the effect of and Protection, E&FN SPON, London, 1996, pp. 163 – 172.
size on the behavior of the property of the RC under study. [13] C.F. Hendriks, Certification system for aggregates produced from
 J.M.V. Gómez-Soberón / Cement and Concrete Research 32 (2002) 1301–1311 1311

building waste and demolished buildings, in: J.J.J.M. Goumans, H.A. [26] UNE 83.133, Áridos para hormigones, Determinación de las densi-
Van Der Sloot, T.G. Aalbers (Eds.), Environmental Aspects of Con- dades, coeficientes de absorción y contenido de agua en el árido fino,
struction with Waste Materials, Elsevier, Amsterdam, The Nether- AENOR (Asociación Española de Normalización y Certificación),
lands, 1994, pp. 821 – 843. 1990 (in Spanish).
[14] J. Kasai, Y. Kasai, Guidelines and the present state of the reuse [27] UNE 83.134, Áridos para hormigones, Determinación de la densidad,
of demolished concrete in Japan, in: E.K. Lauritzen (Ed.), Demoli- porosidad, coeficiente de absorción y contenido en agua del árido
tion and Reuse of Concrete and Masonry, Guidelines for Demoli- grueso, AENOR (Asociación Española de Normalización y Certifica-
tion and Reuse of Concrete and Masonry, 1993, pp. 93 – 104 ción), 1990 (in Spanish).
(Odense, Denmark). [28] UNE 83.304, Ensayos de hormigón, Rotura por compresión, AENOR
[15] RILEM Recommendation, 121-DRG Guidance for demolition and (Asociación Española de Normalización y Certificación), 1984 (in
reuse of concrete and masonry. Specifications for concrete with re- Spanish).
cycle aggregates, J. Mater. Struct. 27 (1994) 557 – 59. [29] UNE 83.306, Ensayos de hormigón, Rotura por tracción indirecta
[16] J. Vyncke, E. Rousseau, Recycling of construction and waste in Bel- (Ensayo Brasileño), AENOR (Asociación Española de Normalización
gium: actual situation and future evolution, in: E.K. Lauritzen (Ed.), y Certificación), 1985 (in Spanish).
Demolition and reuse of concrete and masonry, Guidelines for Dem- [30] UNE 83.310, Ensayos de hormigón, Determinación de la permeabili-
olition and Reuse of Concrete and Masonry, E&FN SPON, Denmark, dad, AENOR (Asociación Española de Normalización y Certifica-
1993, pp. 57 – 69. ción), 1990 (in Spanish).
[17] J.M. Gómez, E. Vázquez, L. Agulló, Strength and deformation proper- [31] UNE 83.312, Ensayos de hormigón. Hormigón endurecido, Determi-
ties of recycled aggregate concrete, Fifth CANMET/ACI International nación de la densidad, AENOR (Asociación Española de Normaliza-
Conference on Recent Advances in Concrete Technology, Singapore, ción y Certificación), 1990 (in Spanish).
2001, pp. 103 – 120 (Supplementary papers; M. Venturino (Compiler)). [32] UNE 83.316, Ensayos de hormigón, Determinación del módulo de
[18] J.M. Gómez, L. Agulló, E. Vázquez, Cualidades fı́sicas y mecánicas elasticidad en compresión, AENOR (Asociación Española de Normal-
de los agregados reciclados de concreto. Construcción y Tecnologı́a. ización y Certificación), 1996 (in Spanish).
Magazine of the Mexican Institute of the Cement and the Concrete, [33] UNE-EN 933:3, Ensayos para determinar las propiedades mecánicas y
AC (IMCYC), Mexico, DF, Mexico, Vol. XIII (157) (2001) pp. 10 – fı́sicas de los áridos: Parte 3. Determinación de la forma de las partı́c-
22 (ISSN 0187-7895) (In Spanish) http://www.imcyc.com/cyt/junio/ ulas. Índice de lajas, AENOR (Asociación Española de Normalización
cualidades.htm. y Certificación), 1997 (in Spanish).
[19] J.M. Gómez, E. Vázquez, L. Agulló, Hormigón con áridos reciclados. [34] ASTM D 4404, Determination of pore volume and pore volume dis-
Una guı́a de diseño para el material. International Center of Numerical tribution of soil and rock by mercury intrusion porosimetry. P.A.,
Methods in Engineering (CIMNE). Monograph M60 (2001) 1 – 137 United States of America, 1998.
(ISBN 84-89925-80-1), Barcelona, Spain (in Spanish). [35] F.G. Rodrı́guez, Porosimetrı́a por intrusión de mercurio: Fundamentos
[20] J.M. Gómez, L. Aguyó, E. Vázquez, Repercussions on concrete per- de la técnica y aplicación a la caracterización microestructural de
meability due to recycled concrete aggregate, in: V.M. Malhotra (Ed.), hormigones, Ing. Civ. (1997) 21 – 37 (in Spanish).
Third CANMET/ACI International Symposium on Sustainable Devel- [36] S. Diamond, Methodologies of PSD measurements in HCP: Postu-
opment of Cement and Concrete, San Francisco, USA, 2001, pp. lates, peculiarities, and problems, in: L.R. Roberts, J.P. Skalny (Eds.),
181 – 195 (ISBN 0-87031-041-0; SP 202 – 13). Pore Structure and Permeability of Cementitious Materials, Symposi-
[21] UNE 7083, Determinación del peso especı́fico y de la absorción en um Proceedings vol. 137, Materials Research Society, Pittsburgh,
gravas y arenas, AENOR (Asociación Española de Normalización y United States of America, 1989, pp. 83 – 89.
Certificación), 1954 (in Spanish). [37] A. Raymond, A. Cook, K.C. Hover, Mercury porosimetry of cement-
[22] UNE 7140, Determinación de los pesos especı́ficos y absorción de based materials and associated correction factors, ACI Mat. J. (1993)
agua en áridos finos, AENOR (Asociación Española de Normaliza- 152 – 161.
ción y Certificación), 1958 (in Spanish). [38] D.A. St. John, A.W. Poole, I. Sims, Concrete petrography, A Handbook
[23] UNE 7238, Determinación de coeficiente de forma del árido grueso of Investigative Techniques, Arnold, London, 1998, pp. 1 – 474.
empleado en la fabricación de hormigones, AENOR (Asociación Es- [39] R.F. Felddman, The porosity and pore structure of hydrated Portland
pañola de Normalización y Certificación), 1971 (in Spanish). cement paste, in: L.R. Roberts, J.P. Skalny (Eds.), Pore Structure and
[24] UNE 83.116, Áridos para hormigón, Determinación del coeficiente Permeability of Cementitious Materials, Symposium Proceedings, vol.
Los Ángeles, AENOR (Asociación Española de Normalización y Cer- 137, Materials Research Society, Pittsburgh, United States of Amer-
tificación), 1990 (in Spanish). ica, 1989, pp. 59 – 73.
[25] UNE 83.131, Áridos para hormigones, Determinación del equivalente
de arena, AENOR (Asociación Española de Normalización y Certif-
icación), 1990 (in Spanish).