TECHNICAL PAPER

Characterising the segregation

of self‑consolidating concrete
Journal of the South African
Institution of Civil Engineering

using ultrasonic pulse velocity

ISSN 1021-2019
Vol 61 No 1, March 2019, Pages 26–37, Paper 0208

GRINI ABDELOUAHAB is a PhD student in the

Laboratory of Civil and Hydraulics G Abdelouahab, B Abdelhalim, D F Laefer
Engineering at the University of Guelma,
Algeria. He received his BS, his Diploma of
Engineering in Civil Engineering from the
University of Annaba, Algeria, and his Segregation is the unintentional separation of the fresh components of concrete or mortar,
Diploma of Magister in Civil Engineering and
which can have negative impacts on the mechanical, transport and durability properties of
Hydraulics from the University of Guelma,
Algeria, in 1985, 1990 and 2008 respectively. His research interests the cured product. The problem is acute in self-consolidating concrete (SCC), because of its
include materials structures. high fluidity level. To help evaluate segregation, this paper investigates the potential of using
Contact details:
ultrasonic pulse velocity (UPV) as a means to identify and characterise segregation in traditional
Laboratory of Civil and Hydraulics Engineering and SCC mixes. Fourteen different concrete mixes were tested using standard techniques
University of 8 May 1945 (sieve and column) in comparison with the UPV-based test proposed herein. Six of the 14
BP 401
concrete mixtures were stable, as indicated by having sieve segregation indices lower than
Guelma
Algeria 15% and segregation resistances (f) higher than 95%. These six stable samples displayed UPV
T: +213 37 21 58 48 segregation index values (fu) approaching 100%. The remaining samples were found to be
E: abdelouahab.grini@yahoo.fr unstable concretes with sieve segregation index values higher than 15% and resistance index
values lower than 65%. These concretes could also be clearly identified as unstable by a UPV
PROF BENOUIS ABDELHALIM is a researcher
segregation index lower than 80%. The UPV method provides a clean, quick and easy non-
in the Department of Civil Engineering and a
member of the team working in the destructive alternative for testing segregation of both fresh and hardened concrete.
Materials Division of the Laboratory of Civil
and Hydraulics Engineering at the University
of Guelma, Algeria. He received his BS from
the University of Constantine, and his MS
INTRODUCTION as those containing limestone filler. The
and PhD from ENPC–PARIS, France, in 1988, Segregation is the separation of the con- problem is also a common risk in self-
1991 and 1995 respectively. His research interests include durability and stituents of a fresh concrete or mortar. consolidating concrete (SCC) (Khayat &
NDT of concrete and reinforced concrete structures. When segregation occurs in concrete, there Guizani 1997; Bauchkar & Chore 2014;
Contact details: is a concentration of coarse aggregates Vakhshouri & Nejadi 2016).
Laboratory of Civil and Hydraulics Engineering in some areas and fine aggregates in oth- Limestone filler is frequently used in
University of 8 May 1945
BP 401
ers. Segregation results in non-uniform SCC technology as a means to minimise
Guelma concrete with non-uniform distributions segregation (Grzeszczyk & Podkowa
Algeria of engineering properties like strength, 2004). The air content influences con-
T: +213 37 21 58 48
stiffness and time-dependent deformation crete bleeding, as well as segregation.
E: benouis_h@yahoo.fr
with respect to substandard durability Larger volumes of air decrease viscosity
PROF DEBRA LAEFER is a Professor of Urban
and structural performance (Panesar & and influence the volume of paste avail-
Informatics at New York University’s Centre Shindman 2012), which are likely to con- able to improve the flow (Beaupre 1994;
for Urban Science and Progress. She is also an tribute to higher maintenance costs and/or Benaicha et al 2015). Furthermore, SCC,
Adjunct Faculty Member in Civil Engineering
a shorter life expectancy. which has gained worldwide popular-
at the University College Dublin. She holds
degrees from Columbia University, the Segregation may occur due to a con- ity since its first introduction in 1990
University of Illinois at Urban-Champaign, crete’s mix or its handling. For example, (Okamura & Ouchi 2003), is character-
and New York University. water quantity is known to influence the ised by its ability to spread in place under
Contact details: separation between mortar and aggregate its own weight without the need of exter-
Center for Urban Science and Progress and (Li & Kwan 2013). To counter this risk of nally applied compaction energy. The idea
Department of Civil and Urban Engineering
impacting long-term concrete strength, is that SCC is inherently better able to
Tandon School of Engineering
New York University water-reducing admixtures are often achieve reinforcement bar coverage and
370 Jay Street, 12th Floor used to minimise the quantity of mixing reduce honeycombing without the risk
Brooklyn water required to produce concrete of a of vibrator-induced segregation. Thus,
New York 11201
United States of America
certain workability. Unfortunately these SCC can significantly improve the long-
T: +1 646 997 0504 admixtures (plasticisers and fluidis- term performance of in-situ concrete [as
E: debra.laefer@nyu.edu ers, among others) may cause excessive extensively discussed elsewhere, e.g. Xie
bleeding or segregation (Uysal et al 2012; et al (2002)]. With SCC, the segregation
Hattori 1979). The problem occurs in risk stems from its generally high fluidity;
regular concretes and those with highly typically, an SCC mix has a water-cement
Keywords: self-consolidating concrete, ultrasonic pulse velocity, cohesive concrete mixtures (Ghafoori & ratio of about 0.4. The segregation gener-
concrete segregation, sieve stability test, column test Diawara 2010; Job & Harilal 2014), such ally occurs during placement (Khayat &

Abdelouahab G, Abdelhalim B, Laefer DF. Characterising the segregation of self-consolidating concrete using ultrasonic pulse velocity.
26 J. S. Afr. Inst. Civ. Eng. 2019:61(1), Art. #0208, 12 pages. http://dx.doi.org/10.17159/2309-8775/2019/v61n1a3
Feys 2010), and the resulting heterogene- construction industry, intense research in 2013; Kundu 2014) that can be caused by
ities may compromise strength, deforma- this field has led to improved NDT equip- indirect factors, such as the origin of the
tion, and durability in those areas. ment, which has made data collection aggregates, the mixture’s ingredients, and
Segregation tests are usually performed quicker and easier. With proper analysis problems of consolidation (vibration) dur-
on samples of fresh concrete or at the of these NDT results more is possible ing concrete placement. The work herein
beginning of hardening (Okamura & beyond simple quality checks, including proposes that concrete density variation
Ouchi 2003; Shindoh & Matsuoka 2003; possible prediction of important material through the height of an element can
Rakesh 2015). A current test for this is the parameters (Sanish & Manu 2012). While be determined by UPV, noting that the
sieve stability test, which measures the effective, deployment issues remain that elastic modulus can be influenced by the
portion of the fresh SCC sample passing preclude rapid, widespread usage of these large aggregates in concrete versus only
through a 5 mm sieve. If the SCC has a NDT techniques. the small aggregates in the mortar. Thus,
poor resistance to segregation, it can eas- As a work-around to those challenges, the objective of this experimental study
ily pass through the sieve. Therefore, the the ultrasonic pulse velocity (UPV) is to analyse correlations between vari-
sieved portion indicates the SCC stability method was introduced for both field ous segregation indices and UPV outputs
­– likelihood to segregate (De Schutter and laboratory work and is the most in SCC mixes to assess the viability of
2005). Another segregation test widely widely used NDT test for the inspection UPV for on-site segregation evaluation.
used with SCC is the column stability test and evaluation of concrete structures Ultimately, this work aims to propose a
(ASTM International 2017), in which the today. UPV is mainly deployed for the rapid, non-destructive method of charac-
difference between the percentages by determination of the dynamic modulus of terising segregation in fresh SCC mixes.
weight of coarse aggregate in the bottom elasticity and Poisson’s ratio (Huet 1982).
and top sections are considered (Ambroise Using UPV, Naik and Malhotra (1991),
et al 1999; Rols et al 1999). In contrast, the and Malhotra and Carino (1991) dem- RESEARCH SIGNIFICANCE AND
column stability test involves the analysis onstrated that the relationship between EXPERIMENTAL PROCEDURE
of samples from the top and bottom parts compressive strength and pulse velocity This paper presents comparative experi-
of the column, to determine the propor- was non-linear. They concluded that sev- mental work on 14 different self-compact-
tion of the coarse aggregate (Cussigh et al eral parameters can interfere, including ing concrete samples using two standard
2003). This apparatus has been employed the composition of the concrete and its techniques (sieve and column) in com-
for over 35 years (Sidky et al 1981), evalu- moisture content. While there are exist- parison to the herein proposed UPV-based
ated in several European countries, and ing standards that propose correlations to test using a testing procedure introduced
considered sufficiently sensitive to the overcome these difficulties (e.g. RILEM by Abdelouahab and Abdelhalim (2016).
variations of SCC design for widespread 1973; ASTM International 2009; British Correlations were performed with respect
adoption. Finally, the penetration test, in Standards Institute 2004), heterogene- to various segregation indices to test
which a cylinder of pre-specified dimen- ity creates dispersions of the pulsations whether the UPV method is efficient and
sions is allowed to penetrate a fresh (Xiong et al 2011), which complicates reliable compared to traditional methods.
SCC sample, is another commonly used matters, as these dispersions are caused The UPV was used to determine differ-
method. The penetration depth indicates by indirect factors such as the origin of ences between the samples’ top and bottom
a SCC’s stability level (De Schutter 2005). the concrete mixture and the problems sections as cast. The study also concerns
These tests allow estimating concrete seg- of casting in-situ (e.g. over-vibration, the exploration of correlations between the
regation in the material’s fresh state. or placement from too high a height). different parameters studied amongst self-
There are also non-destructive testing However, early work in applying UPV to compacting concrete samples.
(NDT) approaches for hardened concrete concrete, such as the work by Zülfü et al To understand the potential applica-
(e.g. Kumavat et al 2014; Mesbah et al (2008) on the correlation between ultra- bility and reliability of UPV to identify
2011; Silva & Brito 2013). Of particular sonic velocity and compressive strength, segregation problems in SCC mixes, an
relevance is the work by Breul et al (2008) by Abdelhalim and Abdelouahab (2011) experimental programme was devised.
who used image analysis to estimate con- on the estimation of a concrete’s porosity, The scope of that work involved the fol-
crete segregation on-site. As part of that by Kumavat et al (2014) on general condi- lowing: sieve stability tests, as described
work, experiments on SCC and crushed tion assessment, and by Abdelouahab and by EFNARC (2005), column stability tests
particle concrete were conducted in Abdelhalim (2016) on the segregation of (e.g. Cussigh et al 2003; British Standards
16 cm diameter columns. Measurements ordinary concrete, show the potential Institute 1986; Sonebi 2005; Rooney &
were compared with results acquired on of the technique, but do not assess its Bartos 2001), and UPV testing (newly
the same columns using a video-counting robustness when applied to a variety of developed). All are described below.
method. Concerning the segregation SCC mixes, which is the focus of the new
characterisation, image analysis seems research presented herein. Sieve stability test as per
well adapted and provides a fast map- Specifically, the aforementioned initial EN 12350‑11 (EN 2010)
ping of the structure by interpolation. efforts indicated UPV as a promising Principle: The test aims to investigate the
Gamma densitometry has also been technology for studying concrete segrega- resistance of an SCC mix to segregation
used to evaluate segregation in SCC tion, a topic to which UPV has not been by allowing a 10 L sample to undergo
mixes (Schwhdenmann 2005; Li et regularly employed. The opportunity to static segregation for 15 minutes (in a
al 2011; Kundu 2014). As many NDT use it arises from the heterogeneity in bucket). Then the top layer of the sample
methods have found application in the concrete pulse dispersions (Silva & Brito (4.8 kg ± 0.2) is poured into a 5 mm

Journal of the South African Institution of Civil Engineering  Volume 61  Number 1  March 2019 27
sieve. Some mortar then passes through
the sieve. An index π (the mass percent- (a)
age of the sample passing through the 3.94 in
VA A
sieve) is determined using Equation 1 (100 mm)
and expressed in terms of the nearest 1% Direct transmission
(AFGC 2000). The amount of material
passing through the sieve indicates the
propensity towards segregation.

Mcs 19.68 in
π= ∙ 100(1)
Mc (500 mm)

3.94 in i i
Where: (100 mm)
Mcs = mass of concrete collected
through the 5 mm sieve opening
size 3.94 in
(100 mm)
Mc = initial mass of the top layer. 3.94 in
VB B
(100 mm)
Column stability test Cut i-i
Principle: In this test, fresh concrete
is placed in a tube (cross-section =
(b)
100 × 100 mm, height = 500 mm).
Concrete is then taken from the top
(A = 100. 100. 100 mm) and bottom
(B = 100. 100. 100 mm) parts of the
column. After being washed through a
sieve, samples are analysed to determine
the proportion of coarse aggregate. Only
aggregates greater than 5 mm in size are
analysed. An index f is determined as per
Equation 2 to an accuracy of ±1%.

Mα A
f= ∙ 100(2)
Mα B

Where:
Mα A = coarse aggregate mass in the top
part (retained on 5 mm sieve
­opening size)
MαB = coarse aggregate mass in the
­bottom part (retained on 5 mm Figure 1 M
 easurement procedure: segregation resistance by ultrasound pulse velocity:
sieve opening size). (a) velocity propagation on Parts A and B, (b) tube steel moulds

Proceedings ultrasonic test Where: In general, the pulse velocity of cement

Principle: Based on the techniques of VA = VfA – VeA : Propagation velocity ratio paste is lower than that of aggregate.
segregation characterisation used in the in the top section (A) between the According to several studies (e.g. Bullock
column stability test (British Standards filled mould and the empty mould & Whitehurst 1959; Kaplan 1959; Jones
Institute 1986), UPV can be used for V B = VfB – VeB : Propagation velocity ratio 1962), at the same strength level concretes
material in its fresh state or at the begin- in the bottom section (B) between with higher aggregate content give higher
ning of hardening (Hamidian et al 2012) the filled mould and the empty pulse velocities.
to determine an ultrasonic segregation mould, respectively (Figure 1). The samples were tested using ultra-
index f u and to establish correlations Vf = is the propagation velocity in a zone sound for determination of the velocities
between the last index, the sieve seg- of the filled mould of the longitudinal ultrasonic waves. UPV
regation index π, and the segregation Ve = is the propagation velocity in a zone tests were conducted for each sample
resistance f. The ultrasonic coefficient of of the empty mould. using a portable ultrasound model E58
segregation resistance will be proposed as and pulses with a 54 kHz frequency
per Equation 3 to an accuracy of ±1%. The pulse velocity is known to be affected (Figure 2).
significantly by the type and amount An average of three readings per
VA of aggregate (Bullock & Whitehurst sample was taken, as reported herein. The
fu = ∙ 100(3)
VB 1959; Jones 1962; Popovics et al 1990). actual testing procedures are described

28 Volume 61  Number 1  March 2019  Journal of the South African Institution of Civil Engineering
Table 1 Physical and chemical properties of the materials used in ISO1920-7 (ISO 2004). The principle
Factor Cement* Limestone filler* Superplasticiser* of the test is that the pulse of the lon-
gitudinal vibrations is produced by an
CaCO3 (%) – 98.00 –
electro-acoustical transducer (transmit-
CaO (%) 55–65 56.03 – ter). After traversing a known path length
SiO2 (%) 22–28 0.04 – in the concrete, the pulse vibrations were
AL2O3 (%) 5–6 0.08 – converted into electrical signals by a sec-
ond transducer “receiver” (IAEA 2002).
Fe2O3 (%) 3–3.6 0.02 –
Electronic timing circuits enable the
MgO (%) 1–2 0.17 – transit time of the pulse to be measured
K 2O (%) 0.3–0.6 0.02 – and, thus, the velocity of the pulse to
NaO2 (%) 0.1–0.16 0.05 – be calculated. This involved developing
a series of mixes (as will subsequently
SO3 (%) 1.8–2.5 0.0021 –
be described) and applying two com-
CaOL (%) 0.8–1.8 – –
mon tests (sieve and column), as well
cl- 0–0.01 0.0033 <1 as the proposed UPV method. For each
Loss on ignition (%) – 43 – sample, a segregation index f was gener-
ated according to the three procedures.
Density (kg/l) 3.15 2.7 1.2
The UPV test method and the column
Blaine (cm2/g) 3 300–4 000 – –
test method measured both the segrega-
pH – 9 8.2 tion and plastic sedimentation; plastic
Beginning of setting time (mn) ≥ 60 – – sedimentation is affected by segregation,
bleeding and setting time. In contrast,
End of setting time (mn) 150–250 – –
the sieve stability method only measured
– Items not measured  * Values are from manufacturers
segregation.

Materials
The experimental materials were sourced
locally in Algeria, including an ordinary
Portland cement (CEM II-A, 42.50), a
limestone filler (a 0/80 μm) to modify the
viscosity, and a polycarboxylate type based
superplasticiser. The crushed fine aggre-
gates had a maximum grain size of 5 mm,
a fineness modulus of 2.56, and a specific
(a) (b)
gravity of 2.53. The coarse aggregates had
Figure 2 T he UPV equipment used: (a) ultrasonic device, (b) transducers which used 54 kHz a maximum size of 15 mm and a specific
gravity of 2.67. The chemical composition
of the local Portland cement and mineral
Particle size distribution admixtures are given in Table 1 (the values
Inferior to 63μ: 85% Inferior to 125μ: 95% Inferior to 2 mm: 100% are from manufacturers), along with a few
100 performance characteristics of various
90 components. The particle size distributions
of the limestone filler, fine aggregate and
80
coarse aggregate are shown in Figures 3
70 and 4.
Cumulative undersize

60
Mixture proportions
50
This work aims to assess the applicability of
40 ultrasonic velocity measurements to various
SCC mixtures to determine segregation. As
30
such, the concrete formulations were based
20 on methods described by Okamura and
10 Ouchi (2003) and Bensebti (2008). In the
research herein, the proportion of coarse
0
0.1 1 10 100 1 000 3 000 aggregate was fixed at 50% with sand as 40%
Particle size (μm) of the total volume of the mixture (cement,
sand, filler and water). Ideally, the volume
Figure 3 P
 article size distribution of limestone of filler (permission for re-use granted by of the SCC paste should allow the concrete
Entreprise Nationale des Granulats, Unité El Khroub) to flow, while minimising the cost of the

Journal of the South African Institution of Civil Engineering  Volume 61  Number 1  March 2019 29
 100 Sand Gravel Stones
Sand 0.5 mm
90 Coarse aggregate 5/15 mm

80

70
Passing percentage (by mass)

60

50

40

30

20

10

8.0000

40.000

60.000
80.000
25.000
20.000

50.000
ground

16.000
10.000
12.500

31.500
0.800

2.000

4.000
5.000
6.300
0.400
0.200
0.080

0.500

2.500
0.250

1.000
0.630

1.600
1.250
0.100

0.315
0.125

3.150
0.116
Back-

86.6
0.22
0.58

18.2

99.4
7.34

71.6
41.2

Sand

54.96
11.25
6.25
0.04

0.25

0.95

2.11
Gravel

Figure 4 P
 article size distribution of fine and coarse aggregate (permission for re-use granted by the Laboratory of Civil and Hydraulic Engineering
at Guelma University)

raw materials. The compositions of the 14 superplasticiser (1.7% to 2%) and water (32% resistance index values lower than 65%) and
mixtures tested are presented in Table 2. to 48%). The 14 mixes represented two a set of stable mixes (C9-C14) unlikely to
In all of the mixes, the binder quantity was groups of material: a set of unstable mixes segregate. Within each group, the percent-
held constant, and the ratios of the other (C1-C8) likely to segregate (with sieve seg- age of filler to binder was changed starting
constituents were varied: filler (0% to 20%), regation index values higher than 15% and with 0% and eventually reaching 20%.

Table 2 Mix proportions

Proportions in kg/m3

Concrete Super-
Gravel Sand Cement Fillers Water a* = F/B b* = SP/B d* = W/B
plasticiser
(5–15 mm) (0–5 mm) (C) (F) (W) (%) (%) (%)
(Sp)

C01 775 736 495 0.0 237.8 9.91 0.00 2.00 48

C02 775 736 477 24 235.2 9.71 4.76 1.94 47

C03 775 736 460 46 232.5 9.50 9.09 1.88 46

C04 775 736 457 55 230.2 9.49 10.71 1.86 45

C05 775 736 454 64 227.8 9.48 12.28 1.83 44

C06 775 736 452 72 225.3 9.47 13.79 1.81 43

C07 775 736 450 81 222.8 9.46 15.25 1.78 42

C08 775 736 429 107 219.8 9.11 20.00 1.70 41

C09 775 736 618 0.0 198.0 12.35 0.00 2.00 32

C10 775 736 577 29 200.0 11.76 4.76 1.94 33

C11 775 736 550 55 200.0 11.37 9.09 1.88 33

C12 775 736 528 84 196.0 11.06 13.79 1.81 32

C13 775 736 509 102 195.0 10.75 16.67 1.76 32

C14 775 736 488 122 195.0 10.36 20.00 1.70 33

* [B; binder, (a=F/B); limestone filler-to-binder ratio, (b=Sp/B); superplasticiser-to-binder ratio, (d=W/B); water-to-binder ratio]

30 Volume 61  Number 1  March 2019  Journal of the South African Institution of Civil Engineering
Testing procedure
Firstly, a perforated plate sieve with 90
square holes of 5 mm, a frame diameter
80 78.0
of 300 mm, and a height of 40 mm were 75.5 74.0
73.5
70.5 71.5 71.0 71.5 71.0 70.5 71.5
used for testing (ISO 2004). The test 70 69.0
67.0
column used to evaluate the segregation 61.5
resistance was a steel mould measuring 60

Slump flow (mm)

500 mm in height and 100 × 100 mm 50
in cross-section. This mould was also
used for the UPV test, where transduc- 40
ers were attached to the outside of the
30
steel column, which were coupled to the
surface through a suitable medium (e.g. 20
grease), at each of the two ends (Figure 1)
(Kumavat et al 2014). These procedures 10
were applied to the mixtures listed in
0
Table 2. The temperature during mixing C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14
and testing ranged from 18°C to 22°C. Mixture
The experimental work conducted
on the 14 mixes in Table 2 involved the Figure 5 Slump flow (Sf) for all SCC mixes
following:
■■ Sieving method using Equation 1
■■ Column method using Equation 2 30
y = 0.8962x – 50.215
■■ Ultrasonic velocities method proposed
25 R 2 = 0.2825
using Equation 3
Sieve stability (%)

20
Characterisation of fresh concretes
SCC mixes require validation through 15
all three of these tests (for “ordinary”
concrete only one is required). A spread 10
ranging between 640 mm and 720 mm, an
5
H2/H1 ratio higher than 0.80 for the “L”
box, and a sieve segregation index π rang-
0
ing between 0 and 15% are all necessary. 60 65 70 75 80
Additionally, the tests should be conducted Slump flow (mm)
in-situ, as well as in the laboratory.
Figure 6 Sieve stability vs slump flow
Characterisation of segregation
At the end of mixing, tests are immediately ■■ f > 0.95 corresponds to a good segrega- after casting. The process used a sensor
conducted to assess resistance to segrega- tion resistance and a nominal frequency transmitter of
tion. This was done firstly by the sieve ■■ f < 0.90 corresponds to a tendency 54 kHz (Figure 2a), both standard in the
segregation test, secondarily by UPV meas- towards segregation. industry. This nominal frequency limited
urement and, finally, by the column test. In the tests herein, tube forms were also the depth of propagation and the mini-
For the column test, several segrega- used to determine a similar coefficient mum thickness of concrete that could be
tion characterisation techniques have based on UPV ( f u). probed.
been reported in the literature (Kumavat
et al 2014; Lowke et al 2003). In this Ultrasonic velocity measurements
case, the fresh concrete was placed in The ultrasonic coefficient of segrega- EXPERIMENTAL RESULTS
a tube (cross-section 100 × 100 mm, tion resistance proposed will be as per AND DISCUSSION
height 500 mm). The concrete was taken Equation 3. In the temperature range from The data on the slump flow test are given
from the top (A = 100 × 100 × 100 mm) 22°C ± 2°C and one hour after casting, in Figure 5. Slump flow test results differed
and bottom (B = 100 × 100 × 100 mm) the measurement consisted of determin- slightly from the allowable range; specifi-
parts of the column. After being washed ing the propagation time of sound waves cally, EFNARC (2005) suggested a slump
through a sieve, the samples were ana- through the fresh concrete at the upper flow value ranging from 600–750 mm for
lysed to determine the proportion of (A) and lower (B) parts of the samples a concrete to be an SCC. At more than 750
coarse materials within the aggregate. (Figure 1a, b). For this, a pair of transduc- mm, the concrete might segregate, and at
Segregation resistance f was expressed as ers was used (Figure 2b), one serving as a less than 600 mm, the concrete might have
the ratio between the coarse aggregate source (emitter) and the other as a receiv- insufficient flow to pass though highly
mass in the top part and the coarse er. Guided by EN 12504‑4 (EN 2010), congested reinforcement. All the concretes
aggregate mass in the bottom part where: direct transmission mode was conducted studied herein had a slump flow higher

Journal of the South African Institution of Civil Engineering  Volume 61  Number 1  March 2019 31
than 600 mm. Thus, these concretes pre- Table 3 Ultrasonic testing results Table 4 Column testing results
sent an acceptable fluidity with no blockage
In m/s In kg
risk, and conducting L-box tests to check Mixture Mixture
the passing ability of the SCC was not nec- Side A Side B σ* Top Top
σ
essary. However, when the sieve stability is part A part B
between 15% and 30% (Figure 6), the stabil- C1 2 198 2 421 157.7
C1 0.46 0.78 0.23
ity is considered critical, and the specific
C2 2 229 2 515 202.2
testing of segregation is necessary (RILEM C2 0.51 0.80 0.21
1973; ASTM International 2009; Sonebi C3 2 183 2 533 247.5
2005; AFGC 2000). The tests of self-com- C3 0.43 0.80 0.26
pacting practised do not have any obvious C4 2 232 2 520 203.6
C4 0.40 0.75 0.25
relationship between them (Figure 6); they
C5 2 239 2 493 179.6
are not redundant, but complementary to
C5 0.44 0.95 0.36
one another and highlight different aspects C6 2 165 2 514 246.8
of self-compacting. So practising only the C6 0.55 0.96 0.29
on-site slump test is not indicative of an C7 2 166 2 444 196.6
SCC (Mouret et al 2003). Daczko (2002) C7 0.49 0.86 0.26
C8 2 132 2 536 285.7
compared some mixtures with the same
fluidity and obtained different levels of C8 0.40 0.84 0.31
C9 2 560 2 565 3.5
stability.
C9 0.71 0.74 0.02
The amalgamated results for each C10 2 567 2 580 9.2
of the three samples for the 14 mixes C10 0.79 0.83 0.03
C11 2 591 2 597 4.2
are presented for the UPV testing
(Table 3), the column testing (Table 4), C12 2 560 2 561 0.7 C11 0.75 0.79 0.03
and the sieve testing (Table 5). Results
of the ultrasonic segregation index are C13 2 559 2 561 1.4 C12 0.85 0.89 0.03
expressed as the ultrasonic velocity ratio
C14 2 533 2 557 17.0 C13 0.71 0.73 0.02
of the upper part A to that of the lower
part B. Tables 3–5 clearly show a division C14 0.71 0.74 0.02
* σ = the average standard deviation
between the unstable and stable mixes
(C1–C8 vs C9–C14). The results of the
UPV test can be used to assess the segre- Table 5 Summary of test results: ultrasonic, column and sieve coefficients
gation resistance and to control the qual-
UPV coefficient Column coefficient Sieve coefficient
ity of concrete products, as will be further
Concrete
discussed in the section below. fu = VA/VB f = A/B π
(%) (%) (%)
Discussion C1 90.80 ± 3.5 58.30 ± 0.1 24.08
As described in the results, the tests were
only done on fresh mixes. As was noted, C2 90.70 ± 0.4 64.30 ± 4.3 16.49
the first eight concrete mixes (C1–C8)
were unstable, while the subsequent six C3 88.20 ± 1.8 53.60 ± 5.1 16.84

mixes (C9–C14) were stable (60 mm > Sf

C4 88.60 ± 0.8 52.00 ± 4.2 16.12
< 75 mm, π < 15% and f < 95% following
this f u < 98%). Each of the three testing C5 89.50 ± 0.0 46.10 ± 4.6 16.08
methods confirmed the characterisation.
The UPV remained almost constant C6 87.20 ± 1.2 58.50 ± 3.8 16.91
in the case of the stable mixes (C9–C14)
C7 87.20 ± 2.3 57.40 ± 1.8 16.23
(Table 3). For these concretes, the
average standard deviation σ was 6.01 C8 84.10 ± 0.3 47.70 ± 2.8 23.53
m/s, with a minimum of 0.7 m/s and a
maximum of 17 m/s. For the unstable C9 99.80 ± 0.0 95.30 ± 0.7 06.60
concretes (C1–C8), the average standard
C10 99.50 ± 0.3 95.20 ± 0.2 06.35
deviation σ was 215.0 m/s, and ranged
from a minimum of 157.7 m/s to a maxi- C11 99.80 ± 0.2 95.10 ± 0.5 03.30
mum of 285.7 m/s.
Similarly, the column test (Table 4) C12 99.90 ± 0.2 95.10 ± 2.2 08.90
demonstrated that the coarse aggregate
C13 99.90 ± 0.1 96.40 ± 1.6 04.67
mass remained almost constant in the
stable cases (C9–C14). For these con-
C14 99.00 ± 0.6 95.30 ± 2.5 11.45
cretes, the average standard deviation σ

32 Volume 61  Number 1  March 2019  Journal of the South African Institution of Civil Engineering
 was 0.025 kg and ranged only from
50 0.02 kg to 0.03 kg. For the unstable
y = –0.912x + 124.31
concretes (C1–C8), the average standard
R 2 = 0.7382
45 deviation σ was an order of magnitude
higher at 0.27 kg and ranged from 0.21 kg
to 0.36 kg.
W/B (%)

40
In the sieve test, results (Table 5) for
C9–C14 were stable with sieve segrega-
35 tion indices lower than 15% and resis-
tance stability indices higher than 95%.
Furthermore, the coefficients of segrega-
30
85 90 95 100 tion resistance remained almost constant
(a) UPV segregate index fu (%) in the stable cases (C9–C14). Each of the
mixes C1–C8 was shown to be unstable,
50 with a sieve segregation index exceeding
y = –0.2739x + 59.121
15%. The average standard deviation σ
R 2 = 0.8277
45 was 18.28%, with a minimum of 16.08%
and a maximum of 24.08%. For the stable
concretes (C9–C14), the average standard
W/B (%)

40
deviation σ was 7.22% and ranged from
3.3% to 11.45%.
35 In the case of the stable mixes
(C9–C14) the variation between their
values did not exceed 0.9% for UPV
30
45 50 55 60 65 70 75 80 85 90 95 100 [(C13 = 99.90) – (C14 = 99.00)], 1.30%
(b) Column segregate index f (%) for the column test [(C13 = 96.40) –
(C11 = 95.10)] and 8.15% for the sieve test
50 [(C14 = 11.45) – (C11 = 3.30)], respec-
y = 0.8211x – 28.235
tively. Based on the stability test results,
R 2 = 0.6984
45 samples C9–C14 can be considered highly
satisfactory compared to other concretes,
with respect to their stability. Notably,
W/B (%)

40
the stable mixes C9–C14 presented an
ultrasonic index of segregation approach-
35 ing 100% (Table 5).
Figure 7 presents the variations of the
water-to-binder ratio (W/B) versus the
30
0 5 10 15 20 25 segregation level obtained for the three
(c) Sieve segregate index π (%) segregation tests for all mixes. In general,
the increase in W/B led to a decrease
Figure 7 ( a) UPV segregation index test vs ratio of water-to-binder, (b) column segregation index of the UPV and column values and an
test vs ratio of water-to-binder, (c) sieve segregation index test vs ratio of water-to- increase of the sieve stability ratio across
binder, for all SCC mixes all 14 mixes. When the W/B was around
0.33, the stability risk was clearly evident
( f u > 098, f > 095 and π < 0.15), and
110 increased when the W/B ratio exceeded
100 0.41 ( f u < 098, f < 0.95 and π > 0.15). The
90
W/B is the most significant parameter
80
Segregation (%)

influencing the rheological properties

70
60 of concretes. The relationship obtained
50 between W/B and the different stability
40 parameters used, show that this ratio
30 influences them differently (Figure 7).
20 Additionally, these relationships may not
10
all be proportional.
0
30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 The effect of W/B on the viscosity on
W/B (%) the sieve segregation of mixtures is more
fu f π dominant than other studied parameters.
The W/B affects viscosity exponentially,
Figure 8 E volution of the segregation index (fu, f and π) with W/B ratio  as indicated by Libre et al (2010). The

Journal of the South African Institution of Civil Engineering  Volume 61  Number 1  March 2019 33
effect of the water quantity is also illus-
trated in Figure 8. (a) Unstable mixes
110
Segregation appeared in all three
100 C1 C2
tests with W/B ratios higher than 0.32. C3 C4 C5 C6 C7
90 C8
A decrease of this ratio by 19.51% (from
80
a W/B of 41 to 33) caused a decrease

Segregation (%)
70
in the indicators of the segregation π
60
(sieve) of 51.33% and an increase of f
50
and f u to 99.79% and 17.71%, respectively
40
for the concretes with 20% fillers (C8,
30
C14). Figure 9 shows the evolution of
20
the three indicators of segregation ( f u, π
10
and f ), with the filler/binder (F/B) ratio
0
similar in both cases. The ultrasonic 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
coefficient f u was less sensitive to the F/B (%)
variation of the fines quantity in the
concrete than the sieve coefficient π (b) Stable mixes
110
and column coefficient f, especially for C9 C10 C11 C12 C13 C14
100
unstable concretes. The variations were
90
respectively, 6.6% [(C2, f u = 90.70) – (C8,
80
f u = 84.10)], 7.04% [(C2, π = 16.49) – (C8,
Segregation (%)

70
π = 23.53)], and 16.60% [(C2, f = 64.30)
60
– (C8, f = 47.70)] for a 15.24% variation
50
of the F/B ratio [(C2, F/B = 4.76) – (C8,
40
F/B = 20)]. This is because UPVs were
30
determined through the concrete (mor-
20
tar and gravel), whereas the resistance
10
segregation index f related only to
0
gravel. The pulse velocity was affected 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
significantly by the F/B ratio (consider- F/B (%)
ing the two concrete groups: unstable fu f π
and stable). The pulse velocity of the
binder pulse was lower than that of the Figure 9 E volution of the segregation index (fu, f and π) with F/B ratio: (a) unstable mixes, (b) stable mixes
aggregate, but the consistency of paste
affected the velocity more than the
aggregates when the concrete was stable. (a) SD column
3.0
In Figure 10, for the stable mixes the
Standard deviation (%)

standard deviation (SD) of each segrega- 2.5

tion test was plotted against the average 2.0
results. For the column resistance test,
1.5
at segregation ratios above 95% (the pro-
posed limit for non-segregating mixes), 1.0
the average SD was 1.4% (Figure 10a). 0.5
For the UPV column resistance test, at
0
segregation ratios above 99% (the pro- 94 95 96 97
posed limit for non-segregating mixes), Column average (%)
the average standard deviation was 0.28%
(b) SD UPV
(Figure 10b). In general, for all tests SDs 0.7
were higher in the concrete where segre-
0.6
Standard deviation (%)

gation was present.

0.5
The stable concretes (C9–C14) were
those having a sieve segregation index 0.4
lower than 15%, (Figures 11 and 12). All 0.3
of these concretes displayed an ultrasonic 0.2
index f u in excess of 98%. For these mixes, 0.1
when the difference of aggregate content
0
in the column ( f ) did not vary by more 98 99 100
than 5%, the segregation ultrasonic index UPV average (%)
f u did not change by more than 2%. These
results demonstrate the efficacy of using Figure 10 S tandard deviation vs average value (for stable concretes): (a) column average, (b) UPV average

34 Volume 61  Number 1  March 2019  Journal of the South African Institution of Civil Engineering
 the homogeneity of the concrete in terms
100 of segregation. Considering the two
concrete groups, unstable and stable, a
95 decrease in the W/B ratio led to distinc-
tive increases in the segregation indices
f u, f and π. The ultrasonic segregation
fu (%)

90
index f u was found to be less sensitive to
the variation of the fines ratio than the
85 segregation resistance index f. This was
especially true for unstable mixes. The
proportion of water was a major factor
80
40 45 50 55 60 65 70 75 80 85 90 95 100 in segregation for SCC, as expected. The
f (%) effect of the F/B ratio showed similar
results. This was because the UPVs were
Figure 11 R
 elations between UPV segregation index (fu) and column segregation resistance index (f) determined through the cement paste and
granular skeleton (the pulse velocity of
binder was lower than that of aggregate),
100 whereas the f index concerns gravel
mass only.
The stable concretes (those having a
95
sieve segregation index lower than 15%)
all displayed resistance index f values
fu (%)

90 higher than 95% and ultrasonic index f u

values higher than 99%. In this study, the
stable concretes (C9–C14) were highly
85 identifiable regardless of the measuring
methods and the water proportions used,
but identification was especially easy with
80
0 5 10 15 20 the UPV test results, which remained
π (%) almost constant. The results found by
UPV, and those found by traditional
Figure 12 R
 elations between UPV segregation index (fu) and sieve stability index (π) sieve and column tests were similar. As
such, the usability of non-destructive test
methods for evaluation of segregation of
110 concrete was proven. This study showed
the possibility to characterise concrete
100
segregation with a clean, rapid, and easy-
90 to-use non-destructive method at an
acceptable precision and level.
Segregation (%)

80
Future work should focus on more
70 precise analyses of how the weight, nature
of the gravel (the effect of geometry on the
60 gravel segregation) and shape of the test
50 piece (edge effect), as well as the frequency
of the ultrasonic transducers, impact the
40 outputs. Additional work is also needed
to compare segregation measurements on
30
0 5 10 15 20 25 fresh SCC with real in-situ segregation.
π (%)
fu f
REFERENCES
Figure 13 R
 elations between sieve segregation index (π) and segregation resistances index (fu and f) Abdelhalim, B & Abdelouahab, G 2011. Estimation of
concrete’s porosity by ultrasound. Physics Procedia,
the ultrasonic pulse velocity method the UPV segregation resistance index ( f u) 21: 53–58.
for segregation characterisation in fresh at 99%, as per Figure 13. Abdelouahab, G & Abdelhalim B 2016. Investigation
concrete. Therefore, it is necessary to of concrete segregation by ultrasonic pulse velocity.
suggest that the sieve stability index (π) Journal of Architectural Engineering Technology,
should be reduced to 12%, while keeping CONCLUSIONS 5: 169. DOI: 10.4172/2168-9717.1000169.
the column segregation resistance index In this experimental study, a non- AFGC (Association Française de Génie Civil) 2000.
( f ) at 95%, and proposing the addition of destructive method was used to diagnose Recommandations provisoires sur les bétons

Journal of the South African Institution of Civil Engineering  Volume 61  Number 1  March 2019 35
 auto-plaçants, Documents Scientifiques et EFNARC (Experts for Specialized Construction and International Journal of Research in Engineering and
Techniques. Paris: AFGC. Concrete Systems) 2005. European Guidelines for Technology, 3(9): 63–67.
Ambroise, J, Rols, S & Péra, J 1999. Self-leveling Self-Compacting Concrete: Specification, Production, Kundu, T 2014. Ultrasonic and electromagnetic
concrete: Design and properties. Concrete Science and Use. Available at: http://www.efnarc.org/pdf/ waves for nondestructive evaluation and structural
and Engineering, 1(3): 140–147. SCCGuidelinesMay2005.pdf. health monitoring. Proceedings, 1st International
ASTM International 2009. ASTM C 597 2009. EN (European Standard) 2010. EN 12350-11. Testing Conference on Structural Integrity, 4–7 February,
Standard Test Method for Pulse Velocity through Fresh Concrete. Part 11: Self-compacting Concrete – Tucson, AZ, pp 395–405.
Concrete. West Conshohocken, PA: ASTM Sieve Segregation Test. Pilsen: European Committee Li, B, Duan, Y, Zhang, Y & Liu, S 2011. Electromagnetic
International. for Standardization (CEN). wave absorption properties of cement-based
ASTM International 2017. ASTM C1610/C1610M-17 Ghafoori, N & Diawara, H 2010. Evaluation of fresh composites filled with porous materials. Materials
2017. Standard Test Method for Static Segregation properties of self-consolidating concrete under Design, 32(5): 3017–3020.
of Self-Consolidating Concrete Using Column long transportation time and extreme temperature. Li, L G & Kwan, A K H 2013. Concrete mix design based
Technique. West Conshohocken, PA: ASTM Proceedings, 4th North American Conference on on water film thickness and paste film thickness.
International. the Design and Use of Self-consolidating Concrete, Cement and Concrete Composites, 39: 33–42.
Bauchkar, S D & Chore, H S 2014. Rheological SCC 2010, 26–29 September, Montreal, Canada, Libre, N A, Khoshnazar, R & Shekarchi, M 2010.
properties of self-consolidating concrete with pp 139–152. Relationship between fluidity and stability of self-
various mineral admixtures. Structural engineering Grzeszczyk, S & Podkowa, P 2004. Effect of micro consolidating mortar incorporating chemical and
and mechanics, 51(1): 1–13. filler on the properties of self-compacting concrete mineral admixtures. Construction and Building
Beaupre, D 1994. Rheology of high performance mixture. Proceedings, 19th Scientific Conference Materials, 24: 1262–1271.
shotcrete. PhD Thesis, University of British “Concrete and Prefabrication”, Jadwisin, Poland, Lowke, D, Wiegrink, K H & Schiessl, P 2003. A
Columbia. Available at: http://hdl.handle. pp 257–262. simple and significant segregation test for SCC.
net/2429/6975 [accessed on 17 September 2017]. Hamidian, M, Shariati, A, Khanouki, M M A, Sinaei, Proceedings, 3rd International RILEM Conference
Benaicha, M, Jalbaud, O, Roguiez, X, Hafidi Alaoui, A H, Toghroli, A & Nouri, K 2012. Application of on SCC, Reykjavik, Iceland, pp 356–368.
& Burtschell, Y 2015. Prediction of self-compacting Schmidt rebound hammer and ultrasonic pulse Malhotra, V M & Carino, N J 1991. Handbook on
concrete homogeneity by ultrasonic velocity. velocity techniques for structural health monitoring. Nondestructive Testing of Concrete. Boca Raton, FL:
Alexandria Engineering Journal, 54: 1181–1191. Scientific Research and Essays, 7(21): 1997–2001. CRC Press.
Bensebti, S 2008. Essai de caractérisation Hattori, K 1979. Experiences with Mighty Mesbah, H A, Yahia, A & Khayat, K H 2011. Electrical
expérimentale de la ségrégation verticale des superplasticizer in Japan. In: Superplasticizers in conductivity method to assess static stability of
bétons autoplaçants. PhD Thesis, Université de Concrete, Special Publication 62, Farmington Hills, self-consolidating concrete. Cement and Concrete
Constantine, Algérie. Available at: http://revue. MI: American Concrete Institute (ACI), pp 37–66. Research, 41: 451–458.
umc.edu.dz/index.php/b/article/view/205 Huet, C 1982. Propriétés acoustiques. In: Le béton Mouret, M, El Barrak, M, Bascoul, A, Sbartai, M,
[accessed on 20 September 2017]. hydraulique. Paris: Presses de l’École Nationale des Picot J 2003. Caractérisation des grandeurs
Breul, P, Geoffray, J M & Haddani, Y 2008. On site Ponts et Chaussées, pp 423–52. rhéologiques et d’ouvrabilité des pâtes cimentaires
concrete segregation estimation using image analysis. IAEA (International Atomic Energy Agency) 2002. en fonction des paramètres de composition.
Advanced Concrete Technology, 6(1): 171–180. Guidebook on non-destructive testing of concrete Validation sur béton. Recherche de corrélation
British Standards Institute 2004. Testing Concrete structures. Available at: http://www-pub.iaea.org/ multi-échelle. PN BAP 2002 Projet National Bétons
Part 4: Determination of Ultrasonic Pulse Velocity, mtcd/publications/pdf/tcs-17_web.pdf. Autoplaçants, Groupe 2 Axe 22, Toulouse, France:
BS EN 12504-4. London: British Standards ISO (International Standards Organization) 2004. LMDC (Laboratoire Matériaux et Durabilité Des
Institution. ISO1920-7 2004. Testing of concrete. Part 7: Constructions).
British Standard 1986. BS 1881-203:1986. Testing Nondestructive Tests on Hardened Concrete. Naik, T R & Malhotra, V M 1991. The ultrasonic pulse
Concrete. Recommendations for Measurement of Geneva, Switzerland: ISO. velocity method. In: Malhotra, V M, & Carino,
Velocity of Ultrasonic Pulses in Concrete. London: Job, T & Harilal, B 2014. Fresh and hardened properties N J (Eds). Handbook on Nondestructive Testing of
British Standards Institution. of concrete containing cold bonded aggregates. Concrete. Boca Raton, FL: CRC Press.
Bullock, R E & Whitehurst, E A 1959. Effect of certain Advances in Concrete Construction, 2(2): 77–89. Okamura, H & Ouchi, M 2003. Self-compacting
variables on pulse velocities through concrete. Jones, R 1962. Non-Destructive Testing of Concrete. concrete. Advanced Concrete Technology, 1(1): 5–15.
Highway Research Board Bulletin, 206: 37. London: Cambridge University Press. Panesar, D & Shindman, B 2012. The effect of
Cussigh, F, Sonebi, M & De Schutter, G 2003. Kaplan, M F 1959. The effects of age and water-to- segregation on transport and durability properties
Project testing of SCC-segregation test methods. cement ratio upon the relation between ultrasonic of self-consolidating concrete. Cement and Concrete
Proceedings, 3rd International RILEM Symposium pulse velocity and compressive strength of concrete. Research, 42(2): 252–264.
on Self-compacting Concrete, 17–20 August, Magazine of Concrete Research, 11(32): 85. Popovics, S, Rose, J L & Popovics, J S 1990. The
Reykjavik, pp 311–322. Khayat, H K & Guizani, Z 1997. Use of viscosity- behavior of ultrasonic pulses in concrete. Cement
Daczko, J A 2002. Stability of self-consolidating modifying admixtures to enhance stability of fluid and Concrete Research, 20: 259.
concrete assumed or ensured? In: Shah S P, Daczko concrete. ACI Materials Journal, 94(4): 332–340. Rakesh, K 2015. Self-compacted concrete mix design
J A & Lingscheit J N (Eds). Proceedings, 1st North Khayat, K H & Feys, D (Eds) 2010. Production and its comparison with conventional concrete
American Conference on the Design and Use of and Placement of Self-Consolidating Concrete. (M-40). Journal of Civil Environmental Engineering,
Self-consolidating Concrete, 2–13 November, (Proceedings of the 4th North American 5(3): 176.
Chicago, IL, pp 245–252. Conference on the Design and Use of Self- RILEM 1973. Tests on Concrete by the Method
De Schutter, G 2005. Guidelines for Testing Fresh consolidating Concrete, SCC 2010.) Berlin: Springer. of Ultrasonic Testing. RILEM Technical
Self-compacting Concrete. European Union Growth Kumavat, H R, Tapkire, G, Patil, P S & Chitte, C J Recommendations. Series: Tests and Measurements,
Contract No. GRD2-2000-30024 – Testing of SCC. 2014. Condition assessment of concrete with NDT. No. 142.

36 Volume 61  Number 1  March 2019  Journal of the South African Institution of Civil Engineering
Rols, S, Ambroise, J & Péra, J 1999. Effects of different Shindoh, T & Matsuoka, Y 2003. Development of chloride ion permeability and impermeability
viscosity agents on the properties of self-leveling combination-type self-compacting, concrete of self-compacting concrete. Construction and
concrete. Cement and Concrete Research, and evaluation test methods. Advanced Concrete Building Materials, 27: 263–270.
29(2): 261–266. Technology, 1(1): 26–36. Vakhshouri, B & Nejadi, S 2016. Mix design of self-
Rooney, M & Bartos, P J M 2001. Development of the Sidky, M, Legrand, C & Barrioulet, M 1981. Influence compacting light-weight concrete. Structural
settlement column segregation test for fresh self- of the concentration in aggregate and vibration time Engineering and Mechanics, 58: 143–161.
compacting concrete. Proceedings, 2nd International on internal segregation in fresh concrete. Materials Xie, Y, Liu, B, Yin, J & Zhou, S 2002. Optimum mix
Symposium on Self-Compacting Concrete, Tokyo, and Structures, 14(83): 367–377. parameters of high-strength self-compacting
Japan, pp 109–116. Silva, P & Brito, J 2013. Electrical resistivity and concrete with ultrapulverized fly ash. Cement and
Sanish, K B & Manu, S 2012. Characterization of capillarity of self-compacting concrete with Concrete Research, 32: 477–480.
strength development of concrete the using incorporation of fly ash and limestone filler. Xiong, G X, Deng, M, Huang, H Q & Tang, M S 2011.
ultrasonic method. Proceedings, 18th World Advances in Concrete Construction, 1(1): 65–84. Absorbing and mechanical properties of cement-
Conference on Non-destructive Testing, 16–20 Sonebi, M 2005. Evaluation of the segregation resistance based composites with nano-titanic oxide absorbent.
April, Durban, South Africa. of fresh self-compacting concrete using different test Advances In Materials Research, 177: 558–561.
Schwendenmann, G 2005. Study of segregation in self- methods. Proceedings, 1st International Symposium on Zülfü, Ç, Ulucan, Türk, K & Karataş, M 2008. Effect
compacting concrete walls using gamma densitometry. the Design, Performance and Use of Self-Consolidating of mineral admixtures on the correlation between
Proceedings, 2nd North American Conference on the Concrete, SCC 2005, Changsha, China, pp 301–308. ultrasonic velocity and compressive strength
Design and Use of Self-consolidating Concrete, 10–12 Uysal, M, Yilmaz, K & Ipek, A M 2012. The effect for self-compacting concrete. Russian Journal of
November, Chicago, IL. of mineral admixtures on mechanical properties, Nondestructive Testing, 44(5): 367–374.

Journal of the South African Institution of Civil Engineering  Volume 61  Number 1  March 2019 37