Construction and Building Materials 181 (2018) 94–105

Contents lists available at ScienceDirect

Construction and Building Materials

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

Adequacy of carbonate aggregates as an alternative for chert aggregates

in concrete manufacture
A.M.A. Abd-Allah ⇑, A.R. Baghdady, Y.H. Dawood
Geology Department, Faculty of Science, Ain Shams University, 11566 Cairo, Egypt

h i g h l i g h t s

 Diagenesis controls the physical and mechanical properties of carbonate rocks.

 Chert intact rock have higher strength than carbonate intact rock.
 Angularity and porosity of carbonate aggregates increase the concrete strength.
 Chemical deterioration of carbonate aggregates decreases their concrete durability.

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

Article history: Carbonate and chert aggregates are extensively used in concrete manufacture. The present study assesses
Received 13 September 2016 the effectiveness of both aggregates in different concrete mixtures. The diagenetic processes control the
Received in revised form 16 March 2018 dissolution of carbonate aggregates and their physical and mechanical properties. Although the mechan-
Accepted 2 June 2018
ical properties of the chert aggregates increase the concrete strength, the physical properties of the car-
bonate aggregates benefit largely the concrete strength. Sewer water is more effective than the rain and
tap water in carbonate dissolution, so the chemical deterioration of the carbonate aggregates could
Keywords:
decrease the concrete durability.
Aggregates properties
Concrete strength
Ó 2018 Published by Elsevier Ltd.
Carbonate degradation
Water/cement ratio

1. Introduction depletion of chert aggregates has motivated construction agencies

Aggregates constitute the fundamental component in the con-
crete that used in diverse engineering applications. The demand
for aggregates increases day by day due to extensively use of con-
crete in the world [1]. The aggregates must be clean, sound, dur-
able, hard and resistant to abrasion; uniform in quality,
properties and free of friable, thin, elongated or laminated pieces,
disintegrated materials, alkalis substances [2]. Petrology, texture,
particle shape, porosity, rock materials control the mechanical
degradation of aggregates [3]. The naturally fragmented sources
of these aggregates are clastic alluvial sediments, whereas artifi-
cially fragmented sources include crushed sedimentary, igneous,
and metamorphic rocks [4].
Natural coarse chert aggregates and quartz sand are quarried
from the Oligocene and Miocene rock units around the Greater
Cairo area (Egypt), which are composed of cherty gravels and
quartz sand (Fig. 1). After more than 150 years of quarrying, the
⇑ Corresponding author.
E-mail address: aliabdallah62@sci.asu.edu.eg (A.M.A. Abd-Allah).
to opt for artificial sources. Popular amongst these are the crushed
carbonate aggregates, widely exposed in the North Egypt (Fig. 1),
they are cheaper than chert aggregates. Consequently, the demand
for carbonate aggregates has increased progressively with time.
The carbonate aggregates are commonly derived from the Middle
Eocene outcrops, east of Greater Cairo at Jabal Ataqa (Fig. 1). These
outcrops consist mainly of hard, massive limestone, dolomitic
limestone, dolomite and siliceous limestone beds, with chert
bands. Argillaceous limestone and chalky limestone appear higher
in the stratigraphic sections of the Middle Eocene unit.
The aim of the present contribution is to determine the chemi-
cal and mechanical disadvantages of chert and carbonate aggre-
gates and their effects on concrete strength and deterioration.
This study reports the petrographic characteristics of the carbonate
aggregates, and explores the dissolution effects of rain, sewer and
tap waters on the carbonate aggregates to evaluate their chemical
stability.
In order to investigate the properties of the coarse carbonate
and chert aggregates in experimental concrete mixtures, represen-
tative samples were collected from Ataqa area and from Cairo-Suez

https://doi.org/10.1016/j.conbuildmat.2018.06.019
0950-0618/Ó 2018 Published by Elsevier Ltd.
 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105
Fig. 1. (a) Simple geologic map of the Cairo-Suez district showing the outcrops of coarse and fine aggregates, as well as the sampling sites (Jabal Ataqa for carbonate
aggregates). Oligocene sediments are the source of natural fine sand and coarse chert aggregates. (b) Location of map (a) in Egypt.
district, respectively. In addition, quartz sands as fine aggregates
that were collected from Jabal Ahmer area.
2. Materials and methods
2.1. Materials and sources
Primary megascopic investigation of the carbonate rocks at
Jabal Ataqa revealed the presence of four carbonate types that
are recorded also in the aggregates products. For measuring the
physical (density, effective porosity, aggregates shape) and
mechanical (strengths, hardness, abrasion, p-wave velocity) prop-
erties of aggregates, four samples (about ½ m3, each) of coarse car-
bonate aggregate were collected from the Ataqa area (two
samples) and the client stockpiles (two samples), as well as four
coarse chert samples (about ½ m3, each) were collected from the
client stockpiles. Four samples (about ½ m3, each) of quartz sand
were collected from a stockpile derived from Jabal Ahmer
(Fig. 1). The sand samples were mixed together and then split into
two large samples for using in experimental concrete mixtures.
2.2. Properties and tests of aggregates
The carbonate and chert aggregates were investigated in hand
specimen and in thin sections. The thin sections were examined
using polarizing microscope before and after staining the carbon-
ate with Alizarin Red-S. The grain size distribution in the aggre-
gates was determined using sieve analysis. Materials less than
5 mm are termed as fine aggregates, while those more than
95
5 mm are coarse aggregates. The fractions were washed to remove
fine sand, silt and dust. The physical and mechanical properties of
the aggregate materials were tested using mainly the ASTM tech-
niques and methods. X-ray diffraction analysis was conducted on
the bulk samples using a Philips diffractometer (model PW/1840)
with Ni filter, Cu–Ka radiation (k = 1.542 Å). Instrument settings
were 40 Kv and 30 mA potential, scanning speed of 0.02°/second
and 2h ranged between 2° and 60°. The various mineral species
were identified using ASTM index and files. Their proportions were
semi-quantitatively determined, based on the intensities of the
strongest diffraction peaks.
Rain, sewer and tap water were collected to study their possible
dissolution effects on the carbonate aggregates. Three batches of
aggregates, each about 10 g, representing different types of carbon-
ate rock, were immersed separately into 500 ml of these three
water types for 14 months. The water was then chemically ana-
lyzed, and the aggregate particles were dried and weighed to calcu-
late the weight loss due to dissolution. Chemical analyses of the
carbonate aggregates and water samples were performed using
SP-ICP and ICP-OES, respectively.
2.3. Properties and tests of concrete
Concrete mixtures (8000 cm3 for each test) were separately pre-
pared for chert and carbonate aggregates. The coarse aggregate to
fine aggregate ratio of the mixtures was measured as a volume
ratio. For each aggregate type, three water/cement ratios (w/c)
were used in these mixtures. Water/cement ratio is measured in
ml/g, and the three ratios used were 0.3, 0.4, and 0.5 w/c. Portland
96 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105

cement produced by the Suez Cement Factory (25 N/mm2) and
quartz sand were used in the concrete mixtures. The cement
weights were 2800 g, 2400 g, and 2000 g for the 0.3, 0.4, and 0.5
w/c ratios, respectively. Concrete strength was measured 28 days
after preparation of the concrete mixtures. The measured uncon-
fined compressive strength (UCS) value for each concrete mixture
is an average of the three tested samples.
3. Results and discussion
3.1. Aggregates
3.1.1. Petrography of carbonate aggregates
The studied carbonate aggregates are classified into four car-
bonate types according to Folk classification [5]. These types are
dolomicrite, dolobiomicrite, dolosparite, and intrasparite (Fig. 2).
3.1.1.1. Dolomicrite. Dolomicrite is made up of dolomitized micrite.
Dolomite is microcrystalline forming a xenotopic to hypidiotopic
mosaics (Fig. 2A). Some idiotopic unzoned dolomite crystals are
rarely observed. The micritic groundmass is, in places, recrystal-
lized to microsparite (2%) or shows some clayey matrix. Moldic
(fabric selective) and vuggy (non-fabric selective) porosity are
abundant, some of the pores are lined with dolomite crystals or
are partially filled with iron oxides. In addition to dolomitization,
this rock shows evidences of diagenetic dissolution and aggrading
neomorphism.
3.1.1.2. Dolobiomicrite. Dolobiomicrite consists of skeletal allo-
chems (10%) embedded in dolomitized, clayey, partially recrys-
tallized micritic groundmass. Planktonic foraminiferal tests are
the dominant allochems. They are filled with sparitic low-
magnesian calcite or show micrite envelopes. Macrocrystalline

A B

s
m
150µm 500µm

C D

250µm 250µm

E F

100µm 150µm

Fig. 2. Photomicrographs of the recognized carbonate rock types (crossed polars). (A) Dolomicrite. The microcrystalline carbonate groundmass contains abundant secondary
vuggy porosity. (B) Dolobiomicrite. Skeletal grains (s) are set in a partially recrystallized micritic groundmass with dolomite rhombs (arrows). Note the complete filling of
tests and fractures by sparite and the presence of a micrite envelope (m) around one of these tests. (C) Dolobiomicrite: Well developed cloudy sparitic dolomite rhombs
constitute a major part of a partially recrystallized micritic groundmass. The cores of some dolomite rhombs (arrows) are made up of calcite as a result of dedolomitization.
(D) Dolosparite. The rock is made up of dolomitic sparitic groundmass showing xenotopic and rare idiotopic dolomite crystals. (E) Dolosparite. Partial dissolution of the
relatively less stable cores (arrows) of the sparitic dolomite crystals (rock type). (F) Intrasparite, allochems are represented mainly by fragments of fossiliferous micritic
limestone (f), intergranular and drusy calcite (arrow) constitutes a part of the groundmass, and porosity is mainly intercrystalline and moldic.
 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105 97

cloudy dolomite rhombs, some of which have calcite cores, are
common. Cementation with calcite is more common than with sil-
ica or iron oxides. Recrystallization patches of micro- to pseudo-
sparite are found in the groundmass or filling fractures. In many
cases, the calcite filling has obliterated the original structure of
tests and sealed the fracture planes (Fig. 2B). Rarely, microcrys-
talline silica intertongues with calcite. The existing of floating tex-
tures in the dolobiomicrite types and the scarcity of fractures may
be attributed to inhibiting effect of the early phase of cementation.
The micro-fractures, skeleton/matrix contacts and crystal bound-
aries represent rock weakness planes. Iron oxides nodules and
staining of the rock constituents are commonly observed.
Dedolomitization was recorded in the dolobiomicrite rock type
(Fig. 2C). The effect of dedolomitization in this rock type is marked
by the presence of calcite pseudomorphic crystals with the rhom-
bohedral form of precursor dolomite.
Dolomite crystals rich in iron oxides are more susceptible to
dedolomitization than those poor in these oxides. The growth of
the pseudomorphic calcite crystals began in the relatively less
stable core of dolomite crystals and proceeded toward their
peripheries as indicated by [6,7]. Al-Hashimi [8] reported break-
down of ferroan dolomite occur under subaerial conditions by
the action of circulating sea water along permeable zones.
3.1.1.3. Dolosparite. Dolosparite is made up of dolomitized sparite
or pseudosparite; the latter constitutes  5% of the rock ground-
mass and occurs as scattered patches (Fig. 2E). Dolomite is mainly
mesocrystalline forming a xenotopic mosaics. Some idiotopic
cloudy dolomite rhombs with leached cores are observed. The rock
groundmass is, in places, pigmented with iron oxides and contain-
ing a few concealed fractures. Dolomite replacement is based on:
(i) the absence of evaporites; (ii) the occurrence of micritic calcite
relics in the rock groundmass; and (iii) the existence of cloudy
dolomite rhombs, which are known to originate by transformation
from low Mg-calcite during prolonged diagenesis [9]. Mountjoy
et al. [10] reported that dolomitization may be controlled by frac-
turing, which seems to have played a much more prominent role in
paleo-fluid flow. On the other hand, dolomitization of the studied
carbonates was accompanied and/or followed by dissolution. This
fact is confirmed by the presence of dissolution-centered crystals.
Folk and Siedlecka [11] reported that the partial dissolution of
the relatively less stable cores of the dolomite crystals were pro-
duced by the effect of salinity reduction.
3.1.1.4. Intrasparite. Allochems in this rock type are represented
mainly by terrigenous grains (25%) and very rare echinid frag-
ments (Fig. 2F). The former are fine to coarse sand-sized, subangu-
lar to subrounded and dominated by fragments of fossiliferous
micritic limestone. Monocrystalline quartz grains displaying unit
extinction are rarely recorded. The orthochemical constituents of
intrasparite are represented by sparite and relics of micrite. Sparry
calcite exists as intergranular and drusy cements. Intercrystalline
and moldic (fabric selective) porosity are more dominant than frac-
tures. These secondary porosity forms are observed also in the
hand samples. Cementation, dissolution and aggrading neomor-
phisms are more common than compaction. A high aggrading neo-
morphism in a rock has generally a positive effect on its strength.
3.1.2. Petrography of chert aggregates
The investigated chert aggregates are made up mainly of micro-
quartz, megaquartz, and chalcedonic quartz. Microquartz is fine to
very fine equant crystals, sometimes recrystallized to megaquartz
(Fig. 3A, B). Megaquartz crystals are larger in size, they are display-
ing unit extinction and often possess good crystal shapes and
terminations (Fig. 3C). Chalcedonic quartz is a fibrous variety with
radiating arrangement (Fig. 3D). Sometimes, chert aggregates

A B

250µm 250µm

C D

250µm 250µm

Fig. 3. Photomicrographs of the recognized chert types (crossed polars). Note in all photos, the interlocking of silica crystals and absence of porosity. (A) Equant crystals of
microquartz silica with very rare cryptocrystalline opal (arrow). (B) Microquartz silica in the rock groundmass contains some patches of megaquartz silica (arrow). (C)
Megaquartz silica crystal displays unit extinction and well-defined crystal boundaries. (D) Micro- to mega-quartz silica with fibrous, radiating chalcedonic quartz.
98 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105

contain rare cryptocrystalline silica (opal). All the examined thin Table 1
sections clarify the interlocking of crystals and absence of porosity. Chemical composition of the carbonate aggregates.

Dolosparite Dolobiomicrite Dolomicrite Intrasparite

3.1.3. Chemistry Major oxides (%)
X-ray Diffraction analysis of the carbonate aggregates (Fig. 4) CaO 30.22 32.32 25.33 25.89
showed that dolomite is the most dominant mineral (about 36%– MgO 22.03 21.22 18.57 17.91
SiO2 0.78 0.67 1.02 1.05
90%) with a subordinate amount of calcite (about 8%–62%). These
mineral components control mainly the chemical composition of Elements (ppm)
Al 187 354 266 164
the studied carbonate samples.
Ti 6.85 8.72 7.16 4.68
The chemical composition of the carbonate aggregates (Table 1) Fe 256 519 488 76.8
clarifies that Ca+2 and Mg+2 constitute the major elements of the Na 452 326 1800 470
aggregates, the sum of CaO and MgO ranges between 43.8% and K 110 265 159 105
53.5%. These values agree with El-Kammar et al. [12] for the Middle Mn 41.8 23.1 10.6 20.1
P 96.2 24.2 12.5 19
Eocene carbonates at Jabal Abu Trifeiya (Fig. 1). The high abun- Sr 118 125 98 112
dance of dolomite and, to a much lesser extent, calcite in the car- Ba 20 23 9 80
bonate aggregates (Fig. 2) controls the concentrations of Ca+2 and Zn 13 20 23 9
Mg+2. On the other hand, Sr+2 substitutes for Ca+2 as indicated from Ni 3 6 2 8
V 5 13 9 7
their strong positive correlation (R = 0.89, Fig. 5). This substitution
has also been reported from carbonates in other regions [13,14]. It
is reasonable to assume that Al+3 is essentially related to the non-
carbonate terrigenous fraction which consists mainly of clay min- values than those of the massive dolosparite. It was found that
erals and feldspars. This is evident from the strong positive corre- higher moisture content of carbonate aggregates (6 samples for
lation between Al+3 and K+ (R = 0.97, Fig. 5). Also, Zn+2 shows a each type) increases their density by 0.74% and 3.9%. On the other
positive correlation with Al+3 (R = 0.8, Fig. 5) which strongly sug- hand, 6 samples of the massive, fracture-free chert materials
gests that it is primarily related to the non-carbonate terrigenous (Fig. 6A) show a density value of 2.58 ± 0.13 g/cm3. Hence the den-
fraction. Zn+2 is known to be adsorbed on natural clay particles sity of the chert aggregates is based mainly on chert material itself.
[15]. The contents of Fe+3 and Mn+2 may be related to their incor- Porosity of the aggregates particles controls the water and
poration into dolomite structure, perhaps as proxies for Mg+2 [16]. cement absorption in fresh concretes [22]. Effective porosity values
measured for 6 carbonate aggregates samples and range between
3.1.4. Physical properties of aggregates 0.34 and 7.9%, with an average of 5.17%. This marked variation is
Table 2 presents the measured values of the physical and related mainly to difference in the degree of dissolution and
mechanical properties of the carbonate aggregates as compared dolomitization that responsible for pores formation particularly
with those of chert aggregates. in dolomicrite, dolobiomicrite, and intrasparite. On contrary, chert
aggregates consist of interlocked crystal of silica (Fig. 6A) that
show no any measurable porosity value.
3.1.4.1. Density and porosity. Density is mainly related to rock
porosity, saturation, mineral composition, and texture. On the
other hand, porosity in aggregate materials increases their mois- 3.1.4.2. Grain shape, angularity, and surface texture. More than 100
ture content [21]. Therefore, density of the studied carbonate particles from each aggregate type were used in the determination
aggregates was measured in both oven-dry and fully saturated of shape, angularity, and surface texture (Table 2). Chert particles
samples. Density values for the different types of carbonate aggre- have more uniform shapes and rounded than the carbonate parti-
gates (6 samples each) vary between 2.43 and 2.79 g/cm3 with an cles (Fig. 6B and C, respectively). Also, the surfaces of chert parti-
average value of 2.54 g/cm3, depending mainly on their porosity cles are almost smooth to glazy and contain some pits (Fig. 6A
and dolomite content. All the intrasparite samples and some of and B). The carbonate aggregates, on the other hand, have rough
the dolomicrite and dolobiomicrite samples have lower density surfaces and contain pits and occasionally dissolution cavities

Fig. 4. X-ray Diffractogram of a representative carbonate aggregate sample showing the dolomite as a main mineral constituent.
 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105 99

Fig. 5. Bivariate plots between some chemical components in the studied carbonate aggregates. The positive correlations between Ca+2 and both Sr+2 and Mg+2 indicate
atomic substitutions whereas, the positive correlations between Al+3 and both Zn+2 and K indicate a possible association in the terrigenous clay and feldspars.

Table 2
The measured physical and mechanical properties of coarse carbonate aggregates compared with those of chert.

Property Test method Measured values

Density
Effective porosity
Aggregates shape, surface
texture, and roundness
Uniaxial compressive strength
Rock hardness
Carbonate aggregates Chert aggregates
ASTM C127 [17] 2.43–2.79 g/cm3 2.58 ± 0.13 g/cm3
6 samples of each carbonate type and 6 chert
sample
0.34–7.9% Silica materials show no measurable
value
ASTM D4791 [18] more than 100 grain sample Elongated to triangular Elongated with L/W ratio varies from
for each aggregates type Elongation ratio varies from 5/1.3 to 5/4
5/1.7 to 1
Stress rate 1–1.2 Mpa/s (ASTM D2938-95) [19] 51–81 Mpa It is difficult to prepare core samples from
for carbonate. the largest particles
Tests were measured for 3 samples for each
carbonate type
ISRM 2015, correction and calculation for chert
materials 77–164 Mpa
14 large chert particles strike by using
adaptor jacket
Schmidt hammer (N-type). rebound number between 42 and 61 rebound number between 59 and 73

The lowest and highest rebound numbers were More than 200 hammer strikes on 14 large chert particles strike by using
omitted. the carbonate rocks. adaptor jacket
Tensile strength Brazilian test load rate 200 N/s 3.9–6.9 Mpa It is difficult to prepare core samples
Tests were measured for 3 samples for each
carbonate type
Abrasion Los Angeles 21–31% 14–17%
ASTM C131 [20]
Tests were repeated 3 times for each aggregates
type
P-wave velocity PUNDIT PLUS, 50 KHz 2.2–6.3 km/s It is difficult to prepare cubic samples
Cuboidal samples with 70 mm side and
smoothed surface
7 samples for each carbonate aggregate type

particularly in intrasparite, dolobiomicrite and dolosparite. The 3.1.5. Mechanical properties of aggregates
chert particles are rounded while the carbonate ones are of angular 3.1.5.1. Compressive strength and hardness. The uniaxial compres-
periphery (Fig. 6 B and C, respectively). sive strength (UCS) was measured for 3 core samples of each car-
100 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105
Fig. 6. Photographs of the chert (A and B) and carbonate (C) aggregates, showing their textural characteristics.
bonate type, yielding values between 51 and 81 Mpa with an aver-
age value of 57.9 Mpa. Dolosparite and intrasparite samples have
the highest UCS values (between 68, 72 and 81 Mpa), whereas
those of dolomicrite and dolobiomicrite have the lowest values
(between 51, 52 and 56 Mpa). The highly cemented coarse crystals
in the dolosparite and intrasparite control their high UCS values. By
contrast, dissolution and fine crystals of the dolobiomicrite and
dolomicrite are the reasons for their lower UCS values. Dolomitiza-
tion increases rock porosity and probably has a major effect on the
UCS variance as indicated by the variable content of dolomite
(36%–90%). On the other hand, 14 large chert particles were stroke
using the adaptor jacket of Schmidt Hammer to determine UCS val-
ues (Table 2). The measured UCS values vary from 77.24 to 194.28
Mpa with an average of 94.56 Mpa.
As the carbonate aggregates are mixed with water in the con-
crete mixtures, the saturated UCS property was used to measure
the differences between dry and fully saturated UCS values. The
saturated UCS values for 3 core samples of each carbonate type
were found to be less than their dry values, by amounts ranging
between 6.9% and 11.2% with an average of 8.9%. Dolobiomicrite
and dolomicrite samples which have the highest porosity values
(1.9–5.3%) showed the lowest saturated UCS values (47–69 Mpa).
Limestone and dolomitic limestone studied by Kahraman and
Fener [23] show correlation between UCS and both of porosity%
and point load index.
Hardness was measured for both carbonate and chert aggre-
gates (Table 2). The Schmidt hammer rebound numbers range
between 42 and 61, with an average value 54 for the carbonate
intact rocks and between 59 and 73 with an average value of 64
for chert particles. Good correlations exist between the Schmidt
hammer rebound numbers and the UCS and P-wave velocity values
[24,25]. Based on Eq. (15) in Kahraman [24], the Schmidt hammer
rebound numbers and rock density can be used to estimate the UCS
for the carbonate rocks. Comparison between the measured and
estimated UCS values for the carbonate aggregates show that the
estimated UCS values derived from rebound numbers represent
62% to 86% of the measured UCS values. It is noticed mostly that
the carbonate crystal size plays the main role in UCS variation, so
that intrasparite and dolosparite have estimated UCS values rang-
ing between 74% and 86% of their measured UCS values, whereas
those of dolomicrite and dolobiomicrite represent 62% to 76% of
their measured UCS values.
3.1.5.2. Tensile strength. The tensile strength values were measured
for 3 samples of each carbonate type, they range from 3.9 to 6.9
Mpa (Table 2) with an average value of 5.58 Mpa. These variable
values are mostly attributed to the crystal sizes and dissolution
features that also control the measured UCS values.
3.1.5.3. Abrasion. The measured abrasion loss values for both car-
bonate and chert aggregates (test repeated 3 times each), range
from 21% to 31% and from 14% to 17%, respectively (Table 2).
Homogenous composition and absence of porosity of chert parti-
cles are responsible for their low abrasion values. There is a nega-
tive correlation between abrasion loss values and UCS values
measured for the carbonate aggregates. This correlation is con-
trolled mostly by the carbonate compositional and porosity charac-
teristics. Investigation of the aggregates after abrasion testing
shows that the highly porous and fine crystalline carbonate types
(micrite) have high loss percentages, whereas low porosity and
coarse crystalline carbonate types (sparite) have relatively low loss
percentages.
The Los Angeles abrasion values for the carbonate aggregates
are controlled by their bulk density, Schmidt hardness, shore hard-
ness, P-wave velocity, UCS, point load index, and indirect tensile
strength properties [1,26].
3.1.5.4. P-wave velocity. The mechanical and physical properties of
carbonate rocks can be estimated using P-wave velocity [27]. P-
wave velocity values were measured for 7 cubes of each carbonate
type, they vary between 2.2 and 6.3 km/s with an average of 3.8
km/s. Crystal size and porosity of aggregates control the P-wave
velocity as reported by Yasar et al. [27], so the present dolobiomi-
crite samples show the lowest velocity values (2.2–2.8 km/s).
 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105 101

3.2. Concrete

3.2.1. Concrete strength and failures

3.2.1.1. Aggregates grading. Aggregates grading for both fine and
coarse types was determined to select three size gradations that
used in the concrete mixtures (Fig. 7). The standard size distribu-
tion of coarse aggregates was selected based on ASTM C33 [28].
Fineness Modulus (FM) controls the workability, surface texture,
and air content of the concrete [21]. The used fine aggregates have
FM value of 2.39, whereas the small coarse, large coarse, and stan-
dard coarse gradations of coarse aggregates of each carbonate and
chert type have FM values 2.86, 2.43, and 3.55; respectively.

3.2.1.2. Strength and failures. The effects of the three size gradations
of each coarse carbonate and chert aggregates on the concrete
strength were investigated (Fig. 8). The carbonate aggregates were
oven dried to eliminate any initial moisture contents due to their
porosity. The effect of particle shape of the coarse aggregates on
the concrete strength was investigated at three water/cement
(w/c) ratios. It was found that at each w/c ratio, the coarse carbon-
ate aggregates produced concrete of higher strength than that pro-
duced using coarse chert aggregates (Fig. 8). The high strength
values of the carbonate aggregate are attributed to the good inter-
locking (packing) of their rough surfaces, angular, irregular parti-
cles (Fig. 6C); as well as to their relatively high porosity (0.34–
7.9%) and large surface areas. The angularity of the carbonate
aggregates results in large surface areas of particles in the concrete
mixtures, which in turn absorb more water and cement than the
smooth, rounded and non-porous chert aggregates (Fig. 6A and
B). This absorption required slightly addition of water and cement,
to balance the w/c ratio for the carbonate aggregates. Surface
roughness of carbonate particles resulted in a strong bond between
the particles and cement paste. The properties of these mixtures
Fig. 8. Comparison between UCS values at different w/c ratios for the concrete
promote the tightening of bonds between the cement paste and made up of coarse chert and carbonate aggregates. (a) Small coarse aggregates, (b)
the carbonate aggregates and produce a strongly compacted con- Large coarse aggregates, (c) Standard coarse aggregates.
crete. Conversely, the smooth and rounded chert aggregates have
relatively small surface areas and have not required the addition
of extra amount of water and cement to maintain the w/c ratio. grade, the UCS values of the concrete significantly increase with
Strength of the coarse aggregates controls the concrete the decrease of the w/c ratio. At low w/c ratios, the increased
strength, particularly for high-strength concrete [29]. Using car- cement in the concrete mixture strengthens the bonds between
bonate aggregates in the concrete results in higher UCS values than cement paste and the aggregate surface, compared to high w/c
those obtained upon using chert aggregates (Fig. 8). At all w/c ratios. In mixtures having small cement proportions, the shape
ratios, the UCS values of the carbonate concrete are 5.25% to and strength of the aggregates control the UCS values of the con-
20.39% higher than the corresponding values for the chert con- crete. The angular carbonate aggregates produced more tightly
crete. This difference in strength grows with increasing w/c ratio, interlock concrete skeleton that resists the applied stress and leads
where the cement portion decreases (Fig. 8). For each aggregate to high UCS values. The effect of the aggregate packing on concrete

Fig. 7. Cumulative curves illustrating the grain size distributions of the fine and coarse aggregates.
102 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105
strength is more significant than that related to the strength of the
aggregate materials, for carbonate materials (average of 57.9 Mpa)
which have lower UCS values than chert materials (average of 94.5
Mpa). The other properties of carbonate aggregates, such as poros-
ity and surface areas, also affect the strength of the produced con-
crete. The effect of the aggregates fineness has been ignored
because the same fine aggregates have been used in all the con-
crete mixtures. According to Wu et al. [29], in high-strength con-
crete with a w/c less than 0.4, the strength of the mortar and the
bonds at the aggregates particles interface may be like the strength
of the coarse aggregates. The using of high strength coarse aggre-
gates with low brittleness may improve the mechanical properties
of the produced concrete.
The influence of particle sizes of coarse aggregates on the con-
crete strength was investigated using three size grades. For each
chert and carbonate coarse aggregates, the small size gradation
produced concrete having higher UCS than those resulted by larger
size grades (Fig. 8). The combination of different grain sizes from
both fine aggregates and coarse aggregates controls the concrete
strengths. A mixture including both fine aggregates and coarse
aggregates produced a well-interlocked and low-porosity concrete,
which required the addition of small proportions of cement. At all
w/c ratios, the large and standard aggregate grades of carbonate
and chert, along with fine aggregates, required lower quantities
of cement and produced concrete having lower strengths when
Fig. 9. Photographs illustrating (A) the spalling failure in the chert concrete, (B) failures between chert particles and the cement paste, (C) failure of carbonate aggregates and
cement paste, (D) shear failure in carbonate concrete.
compared with that produced using small aggregate grade
(Fig. 8). The small coarse aggregate grade produced concrete with
larger particle surface areas, which required an additional amount
of cement for coating these surface areas. Increasing of the cement
on using the small coarse aggregates produced stronger concrete
mixtures than those obtained by using the large coarse aggregates.
Later concrete has small surface areas and low porosity as reported
by ACI [21].
The compressive strength of the concrete produced using lime-
stone aggregates depends upon the w/c ratios and aggregate sizes,
whereas the microstructure of concrete depends upon the amount
and structure of coarse aggregates and mortar [27]. Vu et al. [30]
reported that the increasing in Young’s modulus, ultimate stress
and Poisson’s ratio in concrete due to decrease in w/c ratios.
Distinction between the carbonate and chert concretes is also
observed in the failure modes and geometry of the produced con-
cretes. In the carbonate concrete, failures under compressive stress
comprised the whole concrete materials, i.e. the carbonate parti-
cles and paste (Fig. 9C and D). Few failure planes were observed
between the carbonate particles and the cement paste. Failures
of chert concrete occurred mainly between the cement paste and
the outer surfaces of the chert particles (Fig. 9A and B). Microscopic
examination of some of these failure planes revealed that the fail-
ure took place particularly along the weakness planes that affect
the carbonate and chert particles and/or by the development of
 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105 103

new failure planes, particularly in the carbonate particles. During
application of compressive stress, the failures in the tested con-
cretes occurred generally at the peak of maximum compression.
The majorities of failures in the carbonate and chert concretes
are parallel or sub-parallel to the long axes of the samples (spalling
to axial splitting failures, as in Fig. 9A and B). Some of mesh-like
failure planes (web failure) were developed in the tested carbonate
concrete samples.
Generally, the carbonate concrete shows failure surfaces that
have a more regular planar geometry, while those in the chert con-
crete are un-even planes commonly coinciding with the outer sur-
faces of the chert particles (Fig. 9A and B). This difference may be
related to the relative strength between aggregates and cement
paste, lower in the carbonate concrete. In some carbonate concrete,
two conjugate failure planes were observed embracing a cone-like
feature (shear failure, Fig. 9D). The variation in the geometry of the
failure planes is attributed mainly to the cement proportion and
the shape, type and size of the coarse aggregates. For the carbonate
and chert aggregates, the spalling and axial splitting failures
occurred mainly at 0.3 and 0.4 w/c ratios and partially at 0.5 w/c
ratio. Three web failures were formed at 0.4 and 0.5 w/c ratios in
the small grade of carbonate aggregates, whereas shear failures
formed at 0.5 w/c ratio in two carbonate concrete samples having
standard coarse aggregate grade.
3.3. Chemical stability and deterioration of aggregates
3.3.1. Chemical stability
Chert (SiO2) has higher hardness and strength values than those
of the carbonate materials (Table 2). The measured physical and
geotechnical parameters favor the higher strength chert concrete
rather than carbonate concrete. Despite these parameters, the
UCS values of the carbonate concrete are higher than those for
chert concrete (Fig. 8). In this comparative study, chemical stability
and dissolution of carbonate aggregates are investigated using
three water types.
Two main chemical interactions have been previously identified
among the concrete components: the Alkali-Silica reaction (ASR)
and the Alkali–Carbonate reaction (ACR). The ASR occurs between
alkalis (Na+, K+) in the concrete pores and the silica dissolved from
aggregates, forming alkali–silica gel. This gel absorbs water, creat-
ing an expansive force, which may be sufficient to disrupt the
hardened concrete [31,32]. In the ACR reaction, the acidic solutions
slowly or rapidly disintegrate Portland cement in concrete by
attacking and leaching the calcium compounds of both cement
paste and carbonate aggregates [33]. This reaction was subdivided
by Katayama [34] into: i) the (ACR) of dolomitic limestones that
results in dedolomitization; ii) the ACR of non-dolomitic lime-
stones that produces reaction rims, and iii) the alkali–silica reac-
tion (ASR) for various carbonate rocks. The chemical
compositions of the present carbonate and chert aggregates allow
the occurrence of both ASR and ACR reactions in the produced
concrete.
3.3.2. Chemical deterioration
Chemical deterioration experiment of aggregate particles has
been performed to imitate the actual effects of particles deteriora-
tion on concrete durability. The results of chemical analyses show
the difference between the chemical composition of rain, sewer,
and tap water before and after immersing of carbonate aggregates
in these water types for 14 months (Tables 3 and 4). In general, the
aggregates dissolution in sewer water is the highest, followed by
rain water and then tap water, dissolving capacity is deduced from
the variation in weight loss. Most probably, this is attributed to the
low pH of sewer water. In addition, calcium carbonate (intraspar-
ite) has higher dissolving capacity in sewer water than the other
three magnesium-rich carbonate types (Fig. 10).
Ca+2, Mg+2, Sr+2 and Ba+2 have higher solubility in sewer water
than the other cations (Figs. 11 and 12). The weight loss of the
aggregates is associated with decreasing volume and/or density.
This implies that if the aggregates in concrete are in contact with
water for a long time, the volume and/or density of the aggregates
may decrease, leading to lower compressive strength and conse-
quent engineering problems.
Conversely, silica solubility is little in natural water and deter-
mined mainly by temperature and pH; while ionic strength and

Table 3
Chemical composition (mg/l) of the different water types before immersing the carbonate aggregates.

Element Ca Mg Sr Ba Zn Al Fe Ni V pH Eh (mv)
Water type
Tap 0.2 0.013 0.322 0.019 0.122 0.437 b.d 0.044 b.d 7.46 31
Sewer 0.35 0.02 1.36 0.019 0.135 0.06 b.d. 0.016 0.015 7.14 19
Rain 0.016 0.009 0.153 0.029 0.15 0.022 b.d. b.d. 0.029 7.37 32

b.d.: below detection limit in (ppm); Fe (0.01), Ni (0.005), V (0.005).

Table 4
Chemical composition (mg/l) of the different water types after immersing the carbonate aggregates.

Element Ca Mg Sr Ba Zn Al Fe Ni V
Aggregate Water type
Dolosparite Tap 4362.18 312.2 1.84 0.045 0.183 0.205 b.d 0.017 0.055
Sewer 4794.24 413.2 2.18 0.104 0.185 0.015 0.013 b.d. 0.023
Rain 3072.8 167.8 2.27 0.047 0.189 0.124 0.016 0.042 0.009
Dolobiomicrite Tap 864 225 0.576 0.026 0.176 b.d. 0.014 0.051 0.064
Sewer 4280 313.3 1.62 0.03 0.172 b.d. b.d. 0.042 0.045
Rain 3078 267.9 2.14 0.034 0.175 b.d. 0.015 0.009 0.019
Dolomicrite Tap 947 123.2 0.622 0.02 0.158 0.013 b.d. 0.036 0.02
Sewer 3089 387.4 1.79 0.044 0.16 b.d. b.d. 0.029 0.057
Rain 4323 244.56 2.11 0.025 0.168 b.d. 0.015 0.031 0.029
Intrasparite Tap 911.5 110.11 0.646 0.021 0.156 b.d. b.d. 0.03 0.014
Sewer 5179.8 668.6 1.65 0.045 0.149 b.d. b.d. 0.047 0.038
Rain 3112.7 231.5 2.05 0.029 0.172 b.d. b.d. 0.013 0.028

b.d.: below detection limit (ppm); for Al (0.01), Fe (0.01), Ni (0.005).

104 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105
Fig. 10. Percentages of weight loss upon immersing the different rock types of
carbonate aggregates in different water types for 14 months.
Fig. 11. The effects of immersing the carbonate aggregates in different water types
on the concentration of Ca+2.
Fig. 12. The effects of immersing the carbonate aggregates in different water types
on the concentrations of Mg+2, Sr+2 and Ba+2.
pressure are of secondary importance under the geothermal condi-
tions [35]. The rate of silica dissolution is slow at ordinary temper-
atures; but at a temperature near the boiling-point, the rate is
much higher. The solubility is little affected by pH between values
of 1 and 9 but rises rapidly at pH’s over 9 [36].
4. Summary and conclusions
Recently, the depletion and high price of coarse chert aggre-
gates in Egypt has prompted the construction agencies to consider
crushed carbonate aggregates as an alternative. The present study
discriminates between the physical and mechanical properties of
these two coarse aggregate materials and their impacts on con-
crete strength. Petrographically, the investigated carbonate aggre-
gates are classified into four carbonate types, which are
dolomicrite, dolobiomicrite, dolosparite and intrasparite. The car-
bonate rocks were subjected to cementation, neomorphism, disso-
lution, compaction, dolomitization and dedolomitization
processes. These diagenetic processes influence the physical and
mechanical properties of the carbonate materials. The chert aggre-
gates are made up of chemically stable silica in the forms of micro-
quartz, megaquartz, and chalcedonic quartz. Recrystallization is
the sole diagenetic process observed in the chert materials. The
chemical and recrystallization characteristics controlled the phys-
ical and mechanical properties of chert. Generally, the mineralogi-
cal and physical properties of chert aggregates enhance their
mechanical properties compared to carbonate materials.
Three size gradations of each chert and carbonate coarse aggre-
gates were used, with the same quartz sand as fine aggregates, to
assess the concrete strengths at 0.3, 0.4, and 0.5 w/c ratios. The
rough, angular and irregular particles of the carbonate aggregates
consumed more cement than the smooth, rounded and regular
particles of the chert aggregates. The produced concrete strength
depends upon strength of the cement paste, cement/aggregates
bond, aggregates gradations, particle geometry and aggregate
materials strength. Under compressive stress, the geometry of fail-
ure planes in the concrete samples are controlled by the w/c ratios,
aggregates gradations and aggregate/cement paste relative
strength.
Although the UCS values of the carbonate concrete are higher
than those of chert concretes, the carbonate materials have a high
dissolution potentiality when exposed to water. The dolosparite is
more resistant to dissolution in sewer water rather than the other
carbonate types, but generally the dissolution decreases from
sewer water to rain water to tap water for all carbonate types.
The dissolution is accompanied by variable weight loss of the
aggregates and cation release into water. Ca+2 and Mg+2 constitute
the major components of the carbonate aggregates and Ca+2, Mg+2;
Sr+2 and Ba+2 represent the highly dissolved cations. In conclusion,
while carbonate aggregates are less expensive and produce higher
strength concrete, the water saturated carbonate concretes may be
susceptible to easier damage and reduced concrete life. Thus, chert
concrete is recommended to be used in wet ground zones and high
humidity areas, whereas the carbonate concrete is suggested for
dry zones and areas.
Conflict of interest
The authors declare no conflict of interest.
Acknowledgments
The authors thank the Suez Canal and United Arab Emirates
Universities for providing the various facilities especially lab anal-
yses and tests. The authors wish to thank Prof. Mohamed Abu Zeid
 A.M.A. Abd-Allah et al. / Construction and Building Materials 181 (2018) 94–105
(Ain Shams Univ.) for revising the manuscript and providing valu-
able comments and Eng. Mohamed Abdel Kader (consultant of
water treatment) for his assistance about the sewer water.
References
[1] I. Ugur, S. Demirdag, H. Yavuz, Effect of rock properties on the Los Angeles
abrasion and impact test characteristics of the aggregates, Mater. Charact. 61
(2010) 90–96.
[2] V. Sravika, G. Kalyan, A study on the partial replacement of coarse and fine
aggregate by coconut shell and quarry dust mix, IOSR J. Mech. Civil Eng. 14
(2017) 1–7.
[3] M. Korkanc, Evaluation of selected basalts from Nigide, Turkey, as source of
concrete aggregate, Eng. Geol. 75 (2004) 291–307.
[4] A. Kazi, Z. Al Mansour, Influence of geological factors on abrasion and
soundness characteristics of aggregates, Eng. Geol. 15 (1980) 195–203.
[5] R.L. Folk, Practical classification of limestone, Am. Assoc. Petrol. Geol. 43 (1959)
1–38.
[6] A.R. Baghdady, Diagenetic evolution of the Middle Eocene carbonates
(Mokattam Formation), Eastern Cairo-Suez District, Egypt, J. King Abdul-Aziz
Univ. Earth Sci. 23 (2011) 43–69.
[7] J. Frank, Dedolomitization in the Taum Sauk limestone (Upper Cambrian),
southeast Missouri, J. Sediment. Petrol. 37 (1981) 760–773.
[8] W. Al-Hashimi, J. Hemingway, Recent dedolomitization and the origin of the
rusty crusts of Northumberland, J. Sediment. Petrol. 43 (1973) 82–91.
[9] A.V. Carozzi, Sedimentary Petrography, Prentice-Hall Inc., Englewood Cliffs,
New Jersey, 1993.
[10] E.W. Mountjoy, H.G. Machel, D. Green, J. Duggan, A.E. William-Jones, Devonian
matrix dolomite and deep burial carbonate cements: a comparison between
the Rimbey-Meadowbrook reef trend and the deep basin of west-central
Alberta, Bull. Can. Petrol. Geol. 47 (1999) 487–509.
[11] R.L. Folk, A. Siedlecka, The schizohaline environment: its sedimentary and
diagenetic fabrics as exemplified by Late Paleozoic rocks of Bear Island,
Svalbard, Sediment. Geol. 11 (1974) 1–15.
[12] A.M. El-Kammar, A.A. Dabous, A.M.A. Moustafa, The mineralogy and
geochemistry of Eocene carbonates from Cairo-Suez district, Egypt, Middle
East Research Center, Earth Science Series, Ain Shams University, 1990, pp. 68–
80.
[13] S.M. McLennan, B.J. Fryer, G.M. Young, The geochemistry of the carbonate-rich
Espanola Formation (Huronian) with emphasis on the rare earth elements,
Can. J. Earth Sci. 16 (1979) 230–239.
[14] H.K. Wood, G.L. Macpherson, Sources of Sr and implications for weathering of
limestone under tallgrass prairie, northeastern Kansas, Appl. Geochem. 20
(2005) 2325–2342.
[15] S. Veli, B. Alyuz, Adsorption of Copper and Zinc from aqueous solutions by
using natural clay, J. Hazard. Mater. 149 (2007) 226–233.
105
[16] G.M. Young, Origin of carbonate-rich early proterozoic Espanola formation,
Ontario, Canada, Geol. Soc. Am. 84 (1973) 135–160.
[17] ASTM C127-88, Standard test method for density, relative density (specific
gravity), and absorption of coarse aggregates, 1993.
[18] ASTM D4791, Standard test method for unconfined compressive strength of
intact rock core specimens, 1995.
[19] ASTM D2938, Flat and elongated particles in coarse aggregate, 1995.
[20] ASTM C131-03, Standard test method for resistance to degradation of small-
size coarse aggregates by abrasion and impact in the Los Angeles machine,
2003.
[21] ACI, Education Bulletin E1-07, Aggregates for concrete, Developed by
Committee (E-701), 2007.
[22] A.M. Ferrero, M.R. Migliazz, Theoretical and numerical study on uniaxial
compressive behavior of marl, Mech. Mater. 41 (2009) 561–572.
[23] S. Kahraman, M. Fener, The effect of porosity on the relation between uniaxial
compressive strength and point load index, Int. J. Rock Mech. Min. Sci. 42
(2005) 584–589.
[24] S. Kahraman, Evaluation of simple methods for assessing the uniaxial
compressive strength of rock, Int. J. Rock Mech. Min. Sci. 38 (2001) 981–994.
[25] M.H. Abdel Aal, A.M.A. Abd-Allah, A. Khafagy, Compressive strength of some
carbonate rocks: problem and measurements, J. Appl. Geophys. 3 (2004) 27–
41.
[26] S. Kahraman, M. Fener, Predicting the Los Angeles abrasion loss of rock
aggregates from the uniaxial compressive strength, Mater. Lett. 61 (2007)
4861–4865.
[27] E.I. Yasar, Y. Erdogan, A. Kilic, Effect of limestone aggregates type and water–
cement ratio on concrete strength, Mater. Lett. 58 (2004) 772–777.
[28] ASTM C33-03, Standard specifications for aggregates, 1997.
[29] K. Wu, B. Chen, W. Yao, D. Zhang, Effect of coarse aggregates type on
mechanical properties of high-performance, Cem. Concr. Res. 31 (2001) 1421–
1425.
[30] X.H. Vu, Y. Malecot, L. Daudeville, E. Buzaud, Effect of the water/cement ratio
on concrete behavior under extreme loading, Int. J. Numer. Anal. Meth.
Geomech. 33 (2009) 1867–1888.
[31] F. Bektas, T. Topal, M.C. Goncuoglu, L. Turanli, Evaluation of the alkali reactivity
of cherts from Turkey, Constr. Build. Mater. 22 (2008) 1183–1190.
[32] J. Lindgard, O. Andic-Cakir, I. Fernandes, T.F. Ronning, M.D.A. Thomas, Alkali-
silica reactions (ASR): literature review on parameters influencing laboratory
performance testing, Cem. Concr. Res. 42 (2012) 223–243.
[33] A.M. Lopez-Buendia, V. Climent, P. Verdu, Lithological influence of aggregates
in the alkali-carbonate reaction, Cem. Concr. Res. 36 (2006) 1490–1500.
[34] T. Katayama, How to identify carbonate rock reactions in concrete, Mater.
Charact. 53 (2004) 85–104.
[35] B. Fleming, D. Crerar, Silica acid ionization and calculation of silica solubility at
elevated temperature and pH application to geothermal fluid processing and
reinjection, Geothermics 11 (1982) 15–29.
[36] B. Krauskopf, Dissolution and precipitation of silica at low temperatures,
Geochim. Cosmochim. Acta 10 (1956) 1–26.