Professional Documents
Culture Documents
SELF-CONSOLIDATING CONCRETE
Kamal H. Khayat
Editor
PROCEEDINGS OF
Organizing Committee
Chairman
Kamal H. Khayat
Universit de Sherbrooke, Canada
Secretary
Rabih Fakih
Grey Matters, UAE
Treasurer
Fouad Yazbeck
Readymix Abu Dhabi, UAE
Advisory Committee
Osama Abdulbari
Sodamco, UAE
James Aldred
GHD, UAE
Hussein Basma
PCFC-Civil Engineering Department, UAE
Tarek Fransawi
Sodamco, UAE
Abdel Kader Trabulsi
Al Falah Readymix, UAE
Sponsored by:
Organized by:
ACI International
ii
PREFACE
Kamal H. Khayat
Editor and Conference Chair
Universit de Sherbrooke, Canada
November 2009
iii
CONTENTS
iv
Keywords: plastic viscosity, pressure loss, reactive powder concrete, selfcompactability, self-compacting concrete, self-consolidating concrete,
superplasticizer, viscosity- agent, yield value
INTRODUCTION
Recent years have seen large number of concrete structures with increasing height
and overall size in Japan, resulting in an increasing demand for a wider diversity
of types of highperformance concrete. In efforts to meet these needs, highstrength, high-durability and high-fluidity concretes, among others, have been
studied.
In Japan, self-compacting concrete/self-consolidating concrete (referred to as
SCC) had been receiving particular scrutiny since 1980s [1], and SCC was put
to practical use in 1990s. With SCC, the use of a vibrator for compacting at site
is not necessary. The development of SCC was initiated from an increasing
demand for improving the reliability of concreting work in Japan. The
background of SCC development in Japan can be seen from the following
information.
The topography of Japan comprises many precipitous mountains and rugged
valleys; thus it is often the case that concrete placement must be carried out under
extremely difficult circumstances. Further, the increasingly complex shape of
concrete structure is making it more difficult to use a vibrator, while the more
densely arranged reinforcing bars resulting from the increasing high-rise concrete
buildings make consolidation more difficult to carry out.
In addition, the following reasons have been cited for the increasing demand for
SCC:
1) There have not been enough workers to carry out compacting work at
construction sites.
2) Vibrating compaction of concrete is extremely noisy and deleterious to the
health of workers, as well as an annoyance to people in the surrounding
neighborhood.
3) Vibrating compaction of concrete is costly and time-consuming.
SCC not only alleviates these problems, but also improves the efficiency of
construction. In other words, the use of SCC requires only the setting up of forms
and placement of concrete by pumping, and eliminates the previously required
hard, labor-intensive work of compacting and the related setting up of scaffolding.
For these reasons, it had been expected that the use of SCC would become
widespread.
Firstly, this paper introduces classification of SCC and verification of selfcompactability prescribed in Recommendation for practical application of selfcompacting concrete established by Japan Society of Civil Engineers (JSCE).
Secondly, this describes technology transition of materials and mixture
proportions, and shows typical examples of SCC applications. Finally, future
research on the estimation of self-compactability of SCC is described.
PAST AND PRESENT OF SCC
Recommendation for self-compacting concrete by JSCE
Details to publish the recommendation In 1990s, SCC increased its practical
results gradually in Japan, therefore, the Concrete Committee of the Japan Society
of Civil Engineers (JSCE) set up Research Subcommittee on Self-Compacting
Concrete in April 1994 and published State of the Art-Report on SelfCompacting Concrete [2] in December 1996. The subcommittee published
Recommendation for practical application of self-compacting concrete based on
the report and accumulated research results in July 1998 [3]. The article includes a
recommendation, a manual for mixture proportioning, a manual for production
and placement, standard test methods and collected data for practical application.
Classification of SCC
In this recommendation, SCC is classified into three types, i.e., Powder-type,
Viscosity agent-type and Combination-type.
Application of SCC
Technology transition of materials and mixture proportions
The most different feature between SCC and ordinary concrete is that SCC has
sufficient self-compactability in the fresh state as before mentioned. Principal
technologies which have been developed to realize the self-compactability of SCC
in terms of materials and mixture proportions are introduced here.
Superplasticizer: The most remarkable development of material for SCC is the
invention of superplasticizer [5]. Dr. Hattori developed formaldehyde condensates
of beta-naphthalene sulfonates with the primary aim of significantly reducing the
water demand of concrete to produce high-strength concrete [6]. Water reduction
of up to 30 percent was achieved with the use of this superplasticizer called
Mighty 150. This admixture was introduced into the Japanese concrete industry as
a nominal name of high-range water-reducing admixture in the early 1960s.
Since then, it has greatly contributed to production of high-strength concrete.
In Germany, Dr. Aignesberger and his colleagues developed the melamine-based
superplasticizer having nearly the same performance as the beta-naphthalenebased one [7]. These two superplasticizers, however, have one common defect
that the loss in slump is considerably large.
Therefore, these superplasticizers were mainly used for factory production where
the problem of loss in slump during transportation is negligible. Another
approach was to add superplasticizer right before placing concrete at site.
To solve this problem, new technology of slump control with a reactive
polymeric dispersant [8], was studied. As a result, superplasticizer called airentraining and high-range water-reducing admixture was developed. This
superplasticizer is mixed into concrete at the mixing plant even for in-situ casting
because the loss in slump during transportation is relatively small.
In terms of rheological aspect, by adding conventional superplasticizers, SCC
shows a small yield stress but it shows a relatively high plastic viscosity due to a
low water-cement (binder) ratio. This too much high plastic viscosity generally
makes it difficult to achieve the easy placement.
A new type of polycarboxylate-based air-entraining and high-range waterreducing admixture has been developed on the basis of steric hindrance theory in
Japan. This polycarboxylate-based superplasticizer imparts the SCC with high
deformability and adequate plastic viscosity in the period immediately after
initial mixing until placement, even when the water-binder ratio is 30% or lower.
Under the circumstances, airentraining and high-range water-reducing
admixture was specified in JIS A 6204 (Chemical admixtures for concrete) in
2)
3)
4)
Slump flow: 550 - 650mm at mixing plant and 450 - 600mm at placing spot.
cases, Rank 1 SCC (see Section 1.3) is often utilized instead of ordinary concrete
to attain the required quality. Mixture proportions depend on the requirements.
Coarse aggregate with maximum size of 15mm is preferred and often used to
enhance the self-compactability.
Application to CFT: Concrete-filled tubular (referred to as CFT) structure was
firstly adopted in 1990s in Japan. The development of the high-strength SCC
made it possible to fill the concrete completely into a steel tubular column having
a number of diaphragms inside (see Fig.8). CFT structure has rapidly been
prevailing in Japan with the strength requirements for concrete becoming more
demanding. In this section, an application record of CFT structure with highstrength SCC to a high-rise building [14] is reported.
Building A is a 47-story high-rise condominium located in Tokyo. Columns up to
90m above ground were designed in CFT structure using SCC with a specified
design strength of 100MPa. The cross-section of the main columns was a square
measuring 900 by 900mm. Square columns were strengthened with internal
diaphragms, each having a round opening in the center with a diameter of 200 to
400mm, through which SCC passed. SCC was pumped from the ground (see
Fig.9) to the allowable level at which the lateral pressure was expected to reach
the permissible stress of the welds. After aforementioned SCC hardened, concrete
buckets were used for filling SCC into the remaining upper parts of square
columns. Circular columns with a diameter of 800mm were partially used. Since
circular columns were found strong against the lateral pressure, SCC was placed
into circular columns by pumping from the ground to a height of 90m in a single
lift.
Table 5 shows the mix proportions of the high-strength SCC with water binder
ratio of about 20% and slump flow of 650mm. The binder was 70% ordinary
portland cement, 20% blast-furnace slag and 10% silica fume.
Application to high-strength steel fiber reinforce concrete: In Europe, highperformance steel fiber reinforced cementitious composite referred to as Reactive
Powder Concrete (RPC) was developed [15]. The RPC is a sort of SCC since it
needs no vibrating compaction during placement. Steam curing at 90oC and the
densest packing design enable to produce precast concrete having highperformance and ultra high-strength of around 200MPa. The actual applications
of RPC have been done by 35 projects in the world. Recently some applications
have been done in Japan and related recommendations have been established by
JSCE [16]. In this section, main features as to mixture proportions, properties in
fresh and hardened states, mechanical properties and construction method of the
first application to pre-stressing concrete bridge in Japan is reported.
10
Table 6 shows the mixture proportions. Water-cement ratio is around 23% and
water-binder ratio is 12%. Cement, quartz and silica fume particles are well
balanced to obtain the densest packing described in Fig. 10. This packing design
helps achieve high-strength and high durability. RPC shows good selfcompactability for casting into thin or complicated shaped mold. It takes 8-14
minutes of mixing to get the specified flowability even by the revolving-blade
mixer.
The compressive strength of RPC is 200-240MPa and the tensile strength is
around 9MPa after 90oC heat curing for 48hours. The high tensile strength
combined with enough ductility makes conventional reinforcement unnecessary.
The 15mm long steel fibers inside of the mortar matrix act as reinforcement to
resist tensile stress.
RPC premix, water and a superplasticizer were mixed as a primary mixing. After
checking the flow value, steel fibers were added to the mixture and it was mixed
as a secondary mixing for 7 min. Then, the flow value was checked again (see
Fig. 11). Mixed RPC was placed through a tremie pipe attached to the hopper
outlet to manufacture the precast blocks described in Fig. 12. After placing, the
block was cured by sheets to prevent water evaporation as a primary curing. For
secondary curing, 90oC steam curing was conducted for 48 hours in a house. The
rate of rising and dropping temperature were controlled to15oC/h and 7 to 10oC /h,
respectively to prevent the cracks due to temperature difference in the blocks.
Transverse wet jointing of blocks was conducted in the plant prior to installation.
After blocks were fixed on both sides, RPC was poured from the bottom slab to
the top slab. The joints were heat-cured by electrical heaters and insulation. The
strength was achieved as temperature between 70 and 90oC was maintained at
joints. A general view of Sakata Bridge constructed with PRC is shown in Fig. 13.
FUTURE OF SCC
Future research on SCC
In Japan, recently some researches have been working to estimate the selfcompactability, especially the pressure loss between obstacles when SCC is
flowing in formworks [17].
Background of research Recently in Japan, ordinary concrete can not be applied
to construction of concrete structures on some conditions that densely reinforcing
bars are arranged. Concrete cannot flow through these reinforcing bars. In these
cases, it is difficult to use vibrators and to check state of concrete compaction
directly. Therefore, in case of SCC, a perfect concrete compaction inside
11
12
2D
D L
Where,
c :
P :
D:
13
(1)
L:
14
15
Viscosity agent-type
Combination-type
Rank 1
0.28-0.30
0.28-0.31
0.28-0.30
Rank 2
0.30-0.33
0.30-0.33
0.30-0.33
Rank 3
0.32-0.35
0.30-0.36
0.30-0.35
Placed in situ
28%(2006)
24%(2005)
5%(2003)
No information
Factory
product
50%(2005)
50%(2003)
70%(2005)
W/C
(%)
s/a
(%)
Air
(%)
LP
SP
55.8
45
2.5
145
260
150
769
965
6.355
55.8
36
2.5
145
260
150
609
1121
7.8
16
Fc
(MPa)
100
Slump
flow
(mm)
600700
s/a
(%)
Air
(%)
LP
Va
SP
49.6
2.5
162.5
260
210
823
880
4.06
7.8
B: Ordinary Portland cement 70%, GGBF Slag 20%, Silica fume 10%
Flow
value
(mm)
200-240
240260
Steel fiber
Dia.
(mm)
Length
(mm)
Vol.
(%)
0.2
15
180
17
1523
157
22
18
PC-based
BNS-based
MM-based
Japan
Korea
China
Vietnam
Thailand
India
Singapore
Malaysia
NewZealand
Australia
USA
Canada
Germany
Share
Ratio
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
RMC SP
Fig. 3 - Share of superplasticizers in each country
19
20
21
Porosity
Densest
packing
Quartz
Cement
Cement
Silicafume
Fig. 10 - The densest packing design
22
Tank A partition
gate
Tank A
Tank A
Tank B
concrete
Tank B
Tank B
ha
Steel bars
hb
concrete
concrete
2000
1500
Clearance: L
30mm
35mm
40mm
45mm
Blockage
1000
500
0
0.26
0.28
0.30
0.32
0.34
Unit volume of coarse aggregate Xv (m3/m3)
23
Shearing stress:
Coarse
aggregate
Sheer stress:
Concrete flow
Steel bar
Shearing Stress :t
By increasing of volume
of coarse aggregate
Sheer Stress :s
D+L
P
Steel bar
SCC
P(D+L)
2D
24
250
Dosage of
viscosity agent
Vis.=12.5%
Vis.=15.0%
Vis.=17.5%
200
150
100
50
0
25
26
GUIDELINES FOR DESIGN AND CONSTRUCTION OF SELFCONSOLIDATING CONCRETE FOR PRECAST, PRESTRESSED
CONCRETE MEMBERS
Kamal H. Khayat, Soo-Duck Hwang, Gillaume Lemieux, and Wu Jian Long
27
INTRODUCTION
The competitive situation in the precast concrete construction market is
significantly affected by price and cost factors as well as by productivity and
quality. This environment is characterized by ever-shorter construction times,
rising labor costs, as well as greater demand for high workability, strength, and
durability. Technological developments and methods of production that can lead
to improved concrete quality and savings are therefore becoming increasingly
important. Self-consolidating concrete (SCC) represents a significant
advancement in concrete technology that provides great potential for efficiency
and economy in concrete construction.
28
SCC is a highly workable concrete that can flow through densely reinforced or
geometrically complex structural elements under its own weight without
mechanical consolidation and adequately fill the formwork with minimum risk of
segregation. The flowability of SCC is higher than that of normal highperformance concrete typically used in precast, prestressed concrete plants. This
characteristic of SCC, coupled with the absence of the noise associated with
vibration, make SCC a desirable material for fabricating prestressed bridge
elements. The use of SCC in the precast, prestressed applications can therefore
enable complex precast concrete members to be prefabricated with greater ease,
speed, economy, and higher surface quality. This can be achieved even in tightlyspaced areas or congested reinforcement, such as columns, cap beams and
superstructure elements and lead to providing uniform and aesthetically pleasing
surfaces. The quality control and quality assurance measures used for producing
SCC will also help achieve structures with the desired durability and service life.
Properly designed SCC is expected to provide similar properties as the
conventional counterparts except for the high workability. However changes in
mix design and fluidity of SCC can result in SCC with hardened properties and
performance that are different from that commonly expected from conventional
concrete. Proper selection of material constituents and proper proportioning are
necessary for achieving the desired workability and performance of SCC. The
factors that significantly influence the design, constructability, and performance
of precast, prestressed bridge elements with SCC need to be researched. It is also
necessary to develop guidelines for the use of SCC in bridge elements and to
recommend changes to AASHTO LRFD bridge specifications. These guidelines
will provide highway agencies with the information necessary for considering
concrete mixtures that are expected to expedite construction and yield economic
and other benefits that are associated with SCC (surface finish, labor cost, etc.).
For successful design of SCC, some factors require greater attention than that
required for conventional concrete, including type and size and grading of coarse
aggregate, composition and content of binder, and w/cm. Proper selection of
material constituents is necessary for workability and performance of the
hardened concrete.
A number of test methods have been used to characterize workability of SCC. No
single test method has been found to fully characterize all relevant workability
aspects of SCC. Selection of proper combined test methods can facilitate
workability testing protocol and provide means for quality control of field
applications.
Knowledge of the compressive strength, elastic modulus, and flexural strength of
concrete is required for estimating camber of prestressed members at the release
of the prestressing load, and for determining elastic deflections caused by dead
29
and live loads, axial shortening and elongation, and prestress losses. Literature
review showed that a loss of up to 20% in the modulus of elasticity could be
obtained compared to the high-performance concrete of normal consistency
because of the lower coarse aggregate volume of SCC [Holschemacher and Klug,
2002]. However, under air-drying conditions, the elastic modulus of SCC can be
higher than that of normal concrete at long term.
Typically, SCC mixtures are proportioned with higher binder content and lower
coarse aggregate volume and maximum size which increase the risk of thermal,
autogenous and drying shrinkage, and creep leading to loss of prestress and
excessive deflections and elastic shortening. Therefore, creep and drying
shrinkage characteristics of SCC need to be considered in the design of precast,
prestressed bridge elements to account for losses in prestress and avoid cracking
of concrete.
According to the literature survey, there seems to be some discrepancy regarding
the visco-elastic properties of SCC because of differences in mix design (w/cm),
type and content of coarse aggregates, type of chemical admixture, and testing
exposure. It is reported that the creep potential of SCC appears to be slightly
higher than that of conventional concrete made with the same raw materials and
having the same 28-day compressive strength [Attiogbe et al., 2002; Pons et al.,
2003; Byun et al., 1998]. Depending on the selected binder, w/cm, and ambient
temperature at the precasting plant, the use of new generation HRWRA may
eliminate the need to use radiant heat or steam curing.
SCC used in precast, prestressed applications is typically proportioned with a low
w/cm (0.32 to 0.36) to enhance stability of the plastic concrete. Relatively low
w/cm values, coupled with high content of binder lead to greater degree of
autogeneous shrinkage than conventional concrete. Such type of shrinkage also
increases with the fineness of the binder and fillers in use. Therefore, drying
shrinkage, autogeneous shrinkage, and thermal contraction have to be managed in
the mix design process and in the structural detailing of the prestressed element.
Studies have shown that the scatter between measured and predicted drying
shrinkage values is greater in the case of SCC than that for normal concrete.
Experimental shrinkage strains for SCC were found to be larger than those
estimated by various prediction models [Byun et al., 1998]. Also, comparison of
experimental creep data to those obtained from major creep-prediction models
indicated differences. Work is required to compare creep and shrinkage data of
SCC mixtures made with representative mix designs and the material constituents
available in the United States to those obtained from prediction models.
The stability of SCC is a key property in ensuring uniform mechanical properties
and adequate performance of precast, prestressed bridge girders. Properly
30
31
32
33
2004]. In some cases, higher level of fly ash replacement may reduce the ability
of SCC to flow. The replacement rate of fly ash also affects strength and
durability. Contribution of fly ash delays the hydration process and strength
development. Fly ash can also affect air entrainment since the carbon present in
fly ash can absorb air-entraining admixture and adversely affect the ability to
entrain air. Therefore, state specific limits on LOI indicative of the carbon
content. Fly ash shall not be used with Type IP or IS cements.
Silica fume
Silica fume conforming to AASHTO M 307 or ASTM C 1240 can be used as
supplementary cementitious material in the proportioning of SCC for improved
strength and durability. Silica fume also improves resistance to segregation and
bleeding. Special care should be taken to select the proper silica fume content. In
some cases, high level of silica fume addition besides increasing cost can cause
rapid surface crusting that leads to cold joints or surface defects if delays occur in
concrete delivery or surface finish. According to Florida DOT [2004], the
quantity of cement replacement with silica fume should be 7% to 9% by mass of
cementitious materials.
Ground granulated blast- furnace slag
Ground granulated blast-furnace slag (GGBFS) meeting AASHTO M 302 or
ASTM C 989 may be used as supplementary cementitious materials. GGBFS
provides reactive fines and due to large replacement rate usually about 40%
enables a low heat of hydration. Cement replacement by GGBFS is based on the
severity of the environment to which the concrete is exposed. The level of
GGBFS addition is 25% to 70% for slightly and moderately aggressive
environments, and 50% to 70% by mass when used in extremely aggressive
environments. A high proportion of GGBFS exceeding 40% may however affect
stability of SCC resulting in reduced robustness with problems of consistency
control while delayed setting can increase the risk of static segregation. When
used in combination with silica fume and/or metakaolin, GGBFS content should
be limited to 50% to 55% of the total cementitious content, by mass of binder
[Florida DOT, 2004]. However, in precast prestressed members, the amount of
slag is usually 40%. GGBFS shall not be used with Type IP or Type IS cements.
Fillers
The particle-size distribution, shape, and water absorption of fillers may affect the
water demand/sensitivity and suitability for use in the production of SCC. Calcium
carbonate-based mineral fillers can enhance workability and surface finish. The fraction
below 0.005 in. (0.125 mm) shall be of most benefit to SCC flow properties. Contents
of fillers should be evaluated to ensure adequate performance of concrete, including
strength development and durability.
34
35
Fine aggregate
For normal weight concrete, fine aggregates conforming to AASHTO M 6 are
appropriate for the production of SCC. Fine aggregate component should be wellgraded concrete sand. It may be beneficial to blend natural and manufactured sand
to improve plastic properties of SCC. Common concrete sand, including crushed
or rounded sand, siliceous or calcareous sands, can be used in SCC. Particle size
fractions of less than 0.005 in. (0.125 mm) should be considered as powder
material in proportioning SCC. Such fine content can have marked effect on
rheology. Fine aggregates for SCC should conform to the gradation requirements
of AASHTO M 6 or ASTM C 33, as presented in Table 1.
Chemical Admixtures
Chemical admixtures are used in precast, prestressed concrete to reduce water
content, improve filling ability and stability, provide air entrainment, accelerate
strength development, enhance workability retention, and retard setting time.
Because chemical admixtures can produce different results with different binders,
and at different temperatures, the selection of the admixtures should be based on
the plant materials and conditions that will be utilized in production. For
prestressed concrete, chloride-ion content in chemical admixtures should be
limited to 0.1%, by mass of the admixture [AASHTO LRFD Bridge Design
Specifications, 2004]. Incompatibility of admixtures with binders can lead to
improper air void system and delayed or accelerated setting time. Therefore,
before the start of the project, concrete with the job materials, including the
admixtures, should be tested to ensure compatibility. Such testing should be
repeated whenever there is a change in the binder and admixtures.
High-range water-reducing admixtures
High-range water-reducing admixtures (HRWRA) shall conform to the
requirements of ASTM C 494 Type F (water-reducing, high range) or G (waterreducing, high range, and retarding) or ASTM C 1017. The admixture should
enable the required water reduction and fluidity during transport and placement.
The use of Type F or G HRWRA is essential to achieve SCC fluidity. Such
HRWRA can be used in combination with regular water-reducing admixtures or
mid-range water-reducing admixtures. There are mid-range water-reducing
admixtures that may be classified under ASTM C 494 as Type A or F depending
on dosage rate. The required consistency retention will depend on the application.
Precast concrete is likely to require a shorter retention time than cast-in-place
concrete.
36
Viscosity-modifying admixtures
The use of a viscosity-modifying admixture (VMA) for SCC proportioned with
w/cm higher than 0.40 is recommended to ensure stability of the fresh concrete.
VMA should not be added to SCC as a means for improving a poor mix design or
poor selection of materials. High dosage of VMA may lead to increased HRWRA
demand and in some cases, some delay in setting, and development of early-age
mechanical properties.
VMAs are used in SCC to enhance segregation resistance and to enhance
robustness by minimizing the effect of variations in aggregate moisture content,
temperature, etc. This can make the SCC less sensitive to small variations in the
proportioning and characteristics of material constituents. There are currently no
ASTM specifications for VMA. Producers should confirm by trial mixtures that
VMA does not adversely affect the hardened concrete properties.
Air-entraining admixtures
Air-entraining admixtures shall conform to the requirements of AASHTO M 154
or ASTM C 260. Air-entraining admixtures are used in concrete primarily to
increase the resistance of the concrete to freeze-thaw damage. Proper selection of
air-entraining admixture that can stabilize small bubbles and properly formulated
HRWRA that does not cause a large number of coarse air bubbles are needed to
design the SCC with adequate air-void system.
In some cases, high dosage of HRWRA coupled with the high fluidity of the
mixture can make it difficult to ensure the entrainment of a fine, stable air-void
system in the concrete. HRWRA can also entrain coarse air bubbles.
Compatibility evaluation between the air-entraining admixture and HRWRA is
therefore needed to achieve the targeted air-void characteristics.
Set-retarding and set-accelerating admixtures
An ASTM C 494 Type D set-retarding admixture may be used during hot weather
concreting or when a delay in setting is required, subject to acceptance by the Engineer.
Some water-reducing admixtures at high dosage rates can act as retarding admixtures.
They should be used with caution. Set-accelerating admixture (Type C) shall be used to
decrease setting time and increase the development of early-age mechanical properties.
The admixture is particularly beneficial in precast concrete construction to facilitate
early form removal and release of prestressing [PCI Bridge Design Manual, 1997]. In
the absence of accelerated radiant heat or steam curing, the use of set-accelerating
admixture in SCC may be beneficial in precast applications when using HRWRA,
especially the poly-naphthalene- or melamine-based products.
Shrinkage-reducing admixtures
37
38
SCC. In general, test methods include the components required for evaluating
simultaneously the filling ability, passing ability, and static stability. Table 2
summarizes some of the main test methods that proposed for the evaluation of
workability of SCC. The filling capacity combines the filling and passing abilities
of SCC and can be tested using the caisson filling capacity [Yurugi et al., 1993].
Filling Ability
The ability of SCC to flow into and fill completely all spaces within the
formwork, under its own weight, is of great importance to SCC casting, distance
between filling points, etc. [ACI Committee 237, 2007]. Slump flow test [ASTM
C 1611] is used to assess the horizontal free flow of SCC in the absence of
obstruction. The test method is based on the test method for determining the
slump of a normal concrete. The diameter of the concrete circle is a measure of
the flowability of the SCC. In general, slump flow varies between 23.5 and 29 in.
(600 and 735 mm) for SCC used in precast, prestressed applications [NCHRP
Report 628, 2009]. When slump flow test is performed, the time needed for the
concrete to spread 20 in. (500 mm) is also noted. This test is called T-50 flow
time. Advantages and precautions of slump flow and T-50 flow test methods are
presented in Table 3.
Passing Ability
The passing ability tests evaluate the ability of concrete to pass among various
obstacles and narrow spacing in the formwork without blockage that can rise from
local aggregate segregation in the vicinity of the obstacles that give rise to
interlocking and blockage of the flow in the absence of any mechanical vibration
[ACI Committee 237, 2007].
The J-Ring test [ASTM C 1621] can be used to assess the restricted deformability
of SCC through closely spaced obstacles [Bartos, 1998]. In general, the maximum
difference between slump flow and J-Ring flow varies between 2 to 3 in. (50 to 75
mm) depending on the filling ability (slump flow) of the mixture. A difference
between slump flow and J-Ring flow less than 1 in. (25 mm) indicates good
passing ability and no visible blocking of the concrete. Difference greater than 2
or 3 in. (50 or 75 mm), depending on the slump flow value, reflects blocking of
the concrete.
In the L-box test, the vertical part of the box is filled with concrete and left at rest
for one minute. The gate separating the vertical and horizontal compartments is
then lifted, and the concrete flows out through closely spaced reinforcing bars at
the bottom. The time for the leading edge of the concrete to reach the end of long
horizontal section is noted. The heights of concrete remaining in the vertical
section and at the leaving edge are determined. The blocking ratio (h2/h1) is
calculated to evaluate the self-leveling characteristic of the concrete. A blocking
39
ratio of 0.5 and higher is indicative of adequate passing ability. Higher values are
necessary in densely reinforced and thin sections.
The V-funnel apparatus consists of a V-shaped funnel with an opening of (2.55
3.0 in.) 65 75 mm at its bottom. The funnel is filled with concrete, then after
one minute, the gate is opened and the time taken for concrete to flow through the
apparatus is measured. In the case of structural applications, the V-funnel flow
time lower than 8 seconds indicates good passing ability [Hwang, 2006].
Advantages and precautions of the slump flow and J-Ring flow test, L-box, and
V-funnel methods are presented in Table 4
Filling Capacity
The property to completely fill intricate formwork or formwork containing closely
spaced obstacles is critical for SCC to achieve adequate in-situ performance. SCC
with high filling and passing abilities can achieve good filling capacity and spread
into a predetermined section to fill the formwork under the sole action of gravity
without segregation and blockage [ACI Committee 237, 2007].
Filling capacity test provides a small-scale model of a highly congested section
and is suitable to evaluate the filling capacity and its self-consolidating
characteristics [Ozawa et al., 1992; Yurugi et al., 1993]. For the caisson test, the
maximum size aggregate (MSA) is limited to in. (19 mm). In general, a filling
capacity higher than 70% is recommended for SCC used in precast, prestressed
applications. Advantages and precautions of the caisson filling capacity test are
presented in Table 5.
Static Stability
Static stability refers to the resistance of concrete to bleeding, segregation, and
surface settlement after casting while the concrete is still in a plastic state [ACI
Committee 237, 2007]. Surface settlement test method can be used to evaluate the
surface settlement of SCC from a plastic state until the time of hardening [Manai,
1995]. In general, a maximum surface settlement lower than 0.5% or a rate of
settlement after 30 minutes lower than 0.27% per hour is recommended for SCC
used in precast, prestressed bridge elements. Surface settlement test enables the
quantification of the effect of mixture proportioning on static stability. The
settlement is monitored until achieving a constant value.
The static stability of SCC can also be determined using column segregation test
[ASTM C 1610]. The coefficient of variation of the aggregate among the column
section can be taken as a segregation index (Iseg) [Assaad et al., 2004]. Another
index consisting of the percent static segregation (S) can be obtained by
measuring the difference between aggregate mass at the top and bottom sections
of the column. Column segregation test consists of casting concrete in a column
40
divided into four sections along the concrete sample. From each section, the
concrete is weighed and washed out. Then, the coarse aggregate content is
determined for each section. In general, a segregation index (Iseg) lower than 5%
or a percent of static segregation (S) lower than 15% is recommended for SCC
used in precast, prestressed bridge elements.
The visual stability index (VSI) involves visual examination of SCC prior to
placement and after performance of the slump flow test. It is used to evaluate the
relative stability of batches of the same or similar SCC mixtures. The VSI
procedure assigns a numerical rating of 0 to 3, in 0.5 increments. The VSI test is
most applicable to SCC mixtures that tend to bleed [Daczko and Kurtz, 2001].
The test can be considered as a static stability index when it is observed in a
wheelbarrow or mixer following some period of rest time (static condition). VSI
value of 0 to 1 is recommended for SCC for precast, prestressed concrete bridge
elements. Advantages and precautions of surface settlement and column
segregation tests are presented in Table 6.
Dynamic Stability
Adequate resistance of concrete to separation of constituents upon placement and
spread into the formwork is required for SCC when flowing through closely
spaced obstacles and narrow spaces to avoid segregation, aggregate interlock, and
blockage [ACI Committee 237, 2007]. The caisson test measures the filling
capacity indicative of the filling and passing abilities; therefore, it is a good
indicator of the dynamic stability. Concrete with high filling ability
(deformability) and good passing ability can achieve adequate filling capacity in
restricted and congested sections that are typical precast, prestressed applications.
Adequate combination of filling and passing ability tests can be used to evaluate
the filling capacity of the concrete which is indicative of the dynamic stability.
Rheology
Generally, two key parameters are determined when a rheology measurement test
is performed: the yield stress, 0 and plastic viscosity, p. Below the yield stress
value, the mixture does not undergo any deformation and behaves as an elastic
material. In SCC, the yield stress should be maintained low enough to ensure
good deformability. The plastic viscosity of concrete affects its ease of placement
and speed of flow. In practice, good balance between yield stress and plastic
viscosity should be achieved to ensure both good deformability, ease of
placement, and flow rate of SCC.
Rheological parameters of concrete can be determined using a concrete
rheometer. In general, the test involves recording the shear stress response to
maintain a given rate of shear at different shear rate values. A linear regression of
the data is usually used to determine the rheological parameters (0 and p)
41
42
43
This approach involves the combination of a high content of powder materials and
low dosage of VMA. Such mixtures are typically more robust than those
proportioned with high powder content and low w/cm. Robust mixture can react
less sensitively to fluctuations in the mixture composition, characteristics of the
raw materials, water content, and concrete temperature.
Cementitious Materials Content and Water-Cementitious Material Ratio
The concrete supplier shall determine the cementitious materials content and
w/cm required to satisfy the specified concrete category. In general, the
cementitious materials content recommended for SCC ranges between 650 and
800 lb/yd3 (386 and 475 kg/m3) [ACI 237R-07]. The w/cm ranges from 0.32 to
0.45. Special care should be taken to select the binder composition of the SCC
made with low w/cm to limit the compressive strength to the target value.
Otherwise, high strength and stiffness could lead to cracking given the high
degree of restrained shrinkage that can take place.
Nominal Size of Coarse Aggregate
Select MSA based on mix requirements and minimum clear spacing between the
reinforcing steel, clear cover to reinforcing steel, and thickness of the member.
Use coarse aggregate with MSA of to in. (12.5 to 19 mm), unless otherwise
specified in the contract documents. Coarse aggregate of 3/8 in. (9.5 mm) MSA
shall be used for casting highly reinforced and restricted sections.
Air-Entrainment and Air-Void Stability
Generally, SCC made with polynaphthalene sulfonate (PNS)-based HRWRA can
exhibit a relatively stable air-void system. The use of polycarboxylate ether
(PCE)-based HRWRA can lead in some cases to entrapment of large air bubbles,
especially if the SCC is subjected to prolonged mixing or agitation after the
introduction of such HRWRA. The introduction of shrinkage reducing admixture
(SRA) may also have significant effect on the air-void system since it makes it
more difficult to properly air-entrain the concrete.
Mixture Robustness
During the mixture qualification process, it is recommended to investigate the
robustness of the particular design of SCC to fluctuations in the characteristics of
concrete constituents. In addition, it is desirable to investigate the effect of slump
flow variation on stability for a particular mix design and set of materials. A well
designed and robust SCC can typically accept a change of 8.5 to 17 lb/yd3 (5 to 10
L/m3) in water content without falling outside the specified classes of
performance [European Guidelines, 2005].
44
SCC mixtures are more sensitive to the variations in the properties and conditions
of constituent materials and quantity fluctuations during production. Fluctuations
in raw materials, such as gradations and moisture contents of aggregates and
batching fluctuation can have dramatic influence on the flowability and the
stability of the concrete. Well designed SCC can give acceptable tolerance to
daily fluctuations in ingredients characteristics and environmental changes, such
as temperature. This tolerance is usually termed robustness and is controlled by
good practice in the selection and proportioning of ingredients, storage and
handling of basic constituents, and by appropriate content of the fine powders
and/or by use of VMA [European Guidelines, 2005].
Trial Batches
SCC mix design shall require minimum of four trial batches for varying
cementitious materials or w/cm to establish the proportions that can achieve
workability ranges and robustness: two water contents above and two below the
target value. The following information shall be included in the trial batch data:
source of all materials; specific gravity and gradation results for sand and coarse
aggregate; design slump flow range; target air content and design strength; details
of mixture proportioning, including admixture dosage rates for design slump flow
range; SCC trial mixture test results for Q/C testing; and mixer used for the mix
design, mixing sequence, charging sequence, and mixing time.
Recommended Range of Workability Characteristics
The use of proven combinations of test methods and performance-based
specifications is necessary to reduce time and effort required for the development
and quality control of high-performance SCC. A set of performance-based
specifications of SCC is summarized in Table 8. Such specifications also include
test methods recommended for material selection and mix design that can be
performed when developing the concrete mixture as well as quality control (Q/C)
test methods that can be performed for concrete acceptance at the precasting
plant.
Specific requirements for SCC in the fresh state may change depending on the
type of application and especially on:
confinement conditions related to the element geometry, reinforcement density,
cover, etc.;
placing equipment (e.g. bucket, pump, direct from truck-mixer, skip, tremie);
placing method (e.g. number and position of delivery points);
finishing method.
45
46
Setting
Typically, SCC used in precast, prestressed applications proportioned with low
w/cm requires high dosage of HRWRA. The setting time increases with the
increase in HRWRA dosage. Set-accelerating admixtures or heat (steam or
radiant) curing may be needed to decrease the setting time and increase the early
strength development. SCC made with Type I/II cement is shown to have lower
HRWRA demand than that with Type III cement with 20% Class F fly ash. The
latter concrete can then exhibit longer setting time. The use of VMA increases the
HRWRA demand and may lead to some set retardation.
The initial and final setting times can be as low as 4 to 6 hours and 5 to 7 hours,
respectively. These values depend on the materials in use, including HRWRA
type and dosage, binder composition, as well as temperature. Greater setting times
can be obtained when using naphthalene- or melamine- based HRWRA. The
difference between initial and final setting time (ASTM 403-05) can range
between 1 to 3 hours for SCC used in precast prestressed applications
proportioned with w/cm of 0.34 and 0.40, and Type I/II cement or Type III
cement with 20% of fly ash replacement [NCHRP Report 628, 2009]. Setting time
of SCC can be determined by using AASHTO T 197.
Temperature Development
In general, SCC proportioned with high cement content or with Type III cement
can lead to considerable temperature rise. SCC made with Type III cement with
20% Class F fly ash can develop comparable heat rise as that of SCC made with
Type I/II cement. SCC proportioned with the 0.34 w/cm has longer time to attain
maximum temperature than SCC made with 0.40 w/cm. This is mainly due to
higher HRWRA concentration of the former concretes. For a given w/cm, the use
of VMA delays cement hydration, thus extending time to attain peak temperature.
Typical temperature development of SCC proportioned with w/cm of 0.34 and
0.40, cement content of 742 and 843 lb/yd3 (440 to 500 kg/m3), and Type I/II and
Type III cement with 20% of fly ash lie in the range of 115 to 125oF (46 to 52oC)
after 48 hours under semi-adiabatic condition. The maximum temperature can
range from 126 to 145F (52.2 to 62.8C). The time to reach maximum
temperature is in the range of 17 to 28 hours [NCHRP Report 628, 2009]. Type
III cement with supplementary cementitious materials (for example 20% of fly
ash or 30% slag) is shown to attain greater release strength than SCC made with
Type I/II cement. Initial curing with heat (steam or radiant) may then be
necessary, especially for SCC proportioned with relatively low w/cm by reason of
high dosage of HRWRA demand causing retardation. The use of finely ground
limestone filler can also enhance compressive strength development at early age.
Finely ground fillers and supplementary cementitious materials can lead to a
47
denser hardened cement matrix and a denser interfacial transition zone with
aggregate and embedded reinforcement. This can lead to greater strength and
durability.
Release Compressive Strength
For precast applications, SCC mixtures are typically proportioned with 0.32 to
0.36 w/cm [ACI Committee 237, 2004]. The upper range may be increased to 0.40
depending on the concrete temperature and mixture compositions. Relatively low
w/cm can lead to higher compressive strength compared to conventional slump
concrete. The minimum specified compressive strength for prestressed concrete
bridge elements and decks is 4,000 psi (27.6 MPa) [AASHTO, 1998]. Typically,
compressive strength at release of the prestressing strands of structural AASHTO
type girders is on the order of 5,000 psi (34.5 MPa) after 18 hours of casting. The
typical 56-day compressive strength is set at 8,000 to 10,000 psi (55 to 69 MPa).
SCC made with polycarboxylate-based HRWRA can develop higher early
compressive strength and ultimate strength than similar SCC made with
naphthalene- or melamine-based HRWRA. The use of VMA can increase the
HRWRA demand and could lead to reduction in early strength development.
Release strength should be achieved within 18 hours after the concrete is cast into
place. The targeted release strength is selected so that the strength of the concrete
in the prestressed beam does not exceed 60% of the design concrete compressive
strength at the time of release (before any losses due to creep and shrinkage) [PCI
Bridge Design Manual, 1997]. This value is limited to 55% in the case of posttensioned members. Maturity testing can be considered as an effective way to
monitor strength development at early age whether accelerated heating is used
or not. In order to estimate f c ' , ACI 209 and CEB-FIP MC90 models can be used:
ACI 209
fc'
t
t
A Bt
fc'
28 d
'
48
exp s 1
28
t / t1
1/ 2
f cm
49
f r 0.97 f c
'
'
50
Ec
0.043
1.5
c
fc
'
'
t, ti
0.118
in which:
kvs
51
khc
1.56 0.008H
kf
35
, ktd
7 f ci '
t
61 0.58 f ci ' t
where:
H=
t=
ti =
A=
Autogenous Shrinkage
Autogenous shrinkage corresponds to the macroscopic volume reduction due to
cement hydration (chemical shrinkage) as well as self-desiccation of the cement
paste. The volume of the hydration products is less than the original volume of
unhydrated cement and water. Such reduction in volume can lead to tensile
stresses in the cement paste and microcracking. The reduction of relative humidity
in capillary pores due to cement hydration can also result in negative pressure in
the capillary pores, leading to the formation of meniscus and the development of
tensile stresses in the cement paste.
In the case of concrete proportioned with high w/cm (higher than 0.40),
autogenous shrinkage is low given the ample presence of water in capillary pores.
52
53
ks kh
sh
t
55 t
0.56 10
A (steam-cured)
t
26e
ks
0.0142(V / S )
1064 3.70(V / S )
923
t=
t
45 t
drying time (day)
ks =
size factor
kh =
humidity factor
V / S = volume-to-surface ratio
A=
cement factor: 0.918 for Type I/II cement and 1.065 for Type III + 20
percent
FA
binder
which
may be
used
for
P(SCC)
CEB-FIP 90
cso
( f cm )(
RH
(t tc )
2 Ac
350
100
s
( f cm ) [160 10
sc
(9 0.1 f cm )] 10
t tc
6
RH
1.55
ARH
cso
sc
RH
Ac =
perimeter (mm)
tc =
t=
54
ARH
1 ( RH /100)3
55
56
Laboratory trials should be used to validate the material selection and verify the
properties of the mix design to achieve the targeted properties. Once the optimum
properties are achieved, proper quality control for material properties should be
observed to eliminate fluctuations in fresh and hardened properties of the
concrete. Any changes in raw materials properties should be immediately
identified to allow necessary adjustments of the mix to meet the specified
properties.
Control of Raw Materials
Depending on the mix design, SCC may be less robust than conventional
concrete. SCC may therefore undergo greater changes in workability given small
variations in the physical properties of its constituents, especially in the moisture
content of the sand, fine particle content in sand, as well as grading and shape of
the sand and coarse aggregate. This would necessitate frequent controls to check
for any changes in material properties that could affect the performance of SCC.
The maximum deviation of the sand moisture should not exceed 0.2% in order
to minimize the variations in fresh properties of SCC. The water content of sand
should be determined just before production of SCC.
Changes in coarse aggregate physical characteristics (shape, texture, gradation)
can affect workability. Inspection at the storage location should be conducted on
coarse aggregate to characterize their physical characteristics for every aggregate
delivery.
The moisture content, water absorption, aggregate gradation, and variations in
fines content of the aggregate should be continuously monitored and must be
taken into account to produce SCC with constant characteristics. Changing the
source of supply for aggregate is likely having significant change to the concrete
properties and should be carefully and fully evaluated [European Guidelines,
2005].
It is preferable to control the moisture of sand before every batch of SCC. The
moisture content in coarse aggregate must be also taken into account and should
be determined at least twice a day, at the beginning of the first and second
production shifts.
Mixing Process and Sequence
The mixing process should be properly determined given the conditions at hand.
For example, the batch volume should be determined in consideration of the type
of SCC (consistency level), efficiency of the mixer to produce a well-dispersed
and homogeneous mix, and transportation rate from the plant to the casting site.
57
Just prior to mixing of the first batch, the mixer should be pre-wetted or
buttered with SCC of approximately similar consistency. Suitable mixing
sequence should be determined given the mixing and storage equipments
available at the plant. Cement particles should be wetted before contact with
HRWRA. Dry mixing before water introduction is not recommended, because it
may lead to build-up of fine materials in the mixer. All batching water should be
added at the same time. For example, it has been shown that the introduction of
VMA at the end of the mixing sequence and of air-entraining admixture at the
beginning can provide good performance [Khayat 1995; Khayat and Assaad,
2002]. The addition sequence of VMA should be evaluated given the mix design
and admixture in use.
The batch volume is typically limited to 80% to 90% of the maximum capacity of
the mixer to allow efficient mixing energy [JSCE, 1999]. When the mixer is
alternatively used for mixing normal concrete and SCC, testing should be
performed to verify that this does not result in any adverse effect on SCC
properties.
Mixing equipment and mixing sequence should be validated by testing
consistency and self-consolidation properties for a given mix design. Necessary
adjustments to time and speed of mixing should be carried out until consistent and
compliant results are obtained.
Transport
SCC should be delivered in a continuous and timely manner to ensure continuous
placement of precast members with the workability-retention period of the
mixture. This is necessary to avoid lift lines and other surface defects. Transport
method shall be confirmed in order to provide SCC at the casting location that is
sufficiently homogeneous to allow successful placement in the precast element
and to achieve the targeted properties. Mixer trucks have proven to be the best
method of delivery of SCC when transporting over rough terrain or long transport
distance [PCI Interim Guidelines, 2003].
Site Acceptance of Plastic Concrete
The producer should determine the frequency of performing quality control
testing based on available experience mixtures [PCI Interim Guidelines, 2003].
The quality control tests should include visual inspection of every batch of the
concrete and any specific tests and compliance parameters. For example, the
slump flow and VSI tests can be adopted. The T-50 can also be run at least once
on new mixtures and used to check the performance in the event of mix
performance problems.
58
59
60
Surface drying during finishing should be prevented. Fog misting to increase the
relative humidity would minimize rate of evaporation and reduce the risk of
plastic shrinkage. Because of the relatively higher content of fines and eventual
presence of VMA, SCC mixtures develop little or no bleed water compared to
conventional concrete. It is important to begin the finishing of the surface with
light vibrating screeds, or other manual equipment, as soon as the correct level of
the concrete in the formwork has been reached.
SCC exposed surfaces may dry faster than those of normal superplasticized
concrete. This can happen when casting at hot temperature or windy conditions.
Also, depending on the SCC mixture proportioning, stiffening can increase
rapidly in the period 10-40 minutes after casting. Setting time of the SCC mixture
should be adjusted to allow necessary time to carry out the placement process.
Curing
Membrane curing, matting, foils, or appropriate materials should be left in place
for at least 4 days for cast in-place concrete elements [Swedish Concrete
Association, 2002]. This measure should be applied to SCC with low w/cm and
that made with high fines content or VMA. During hot or windy weather
conditions, moisture should be added by watering or by protecting the surface
with wetted membranes for proper curing.
AASHTO LRFD Bridge Construction Specifications [1998] recommend that for
concrete cured other than by steam or radiant heat methods, whenever there is a
probability of air temperature below 36F (2C) during the curing period, the
concrete shall be maintained at a temperature of not less than 45F (7C) for the
first 6 days after placement. This period must be extended if pozzolans are used as
partial replacement of cement. If the compressive strength of 65% of the specified
28-day design strength is achieved in 6 days, an extended period of controlled
temperature may be waived [AASHTO LRFD Bridge Construction
Specifications, 1998].
Due to the specification in terms of early-age compressive strength, steam curing
or radiant heat curing can be used for precast concrete members. The initial
application of steam or heat shall be from 2 to 4 hours after the initial placement
of concrete to allow the initial set of the concrete to take place [AASHTO LRFD
Bridge Construction Specifications, 1998]. In the case of concrete incorporating a
set retarder, the waiting period shall be increased to between 4 and 6 hours after
casting. During the waiting period, the temperature within the curing chamber
shall not be less than 10C.
61
ACKNOWLEDGMENTS
The authors wish to thank the National Academies of the United States for the
support on NCHRP project 18-12, Report 628. The kindness and generosity of
Dr. Amir N. Hanna, Staff Officer at the Transportation Research Board, is greatly
appreciated.
REFERENCES
Acker, P., and Bazant, Z.P., Measurement of Time-Dependent Strains of
Concrete, RILEM Draft Recommendations. Material and Structures, RILEM,
Vol. 31, No. 8 (1998) pp. 507-512.
Acker, P., Mechanical Behavior of Concrete: Physico-Chemical Approach,
Etudes et Recherches des LPC, 152, Laboratoire Central des Ponts et Chausses,
(1988), 121 p. (In French)
American Association of State Highway Transportation officials (AASHTO),
AASHTO LRFD Bridge Construction Specifications, 1st edition, (1998).
American Association of State Highway Transportation officials (AASHTO),
AASHTO LRFD Bridge Design Specifications, 3rd edition, (2004).
American Association of State Highway Transportation officials (AASHTO),
AASHTO LRFD Bridge Design Specifications, 4th edition, (2007).
American Concrete Institute (ACI) Committee 237, Provisional Guidelines on
Self-Consolidating Concrete, (2007).
American Concrete Institute (ACI) Committee 318, Building Code Requirements
for Structural Concrete (ACI 318-05) and Commentary (ACI 318R-05), (2005).
Assaad, J., Formwork Pressure of Self-Consolidating Concrete Influence of
Thixotropy, Ph.D. Thesis, Department of Civil Engineering, Universit de
Sherbrooke, (2004) 453 p.
Assaad, J., Khayat, K.H., and Daczko, J., Evaluation of Static Stability of SelfConsolidating Concrete. ACI Materials Journal, Vol. 101, No. 3 (2004) pp. 207215.
Attiogbe, E.K., See, H.T., and Daczko, J.A., Engineering Properties of SelfConsolidating Concrete Proceedings, 1st North American Conference on the
Design and Use of SCC, Chicago, Illinois, USA (2002) pp. 371-376.
Bartos, P.J.M., An Appraisal of the Orimet Test as a Method for On-Site
Assessment of Fresh SCC Concrete Proceedings, International Workshop on
Self-Compacting Concrete, Japan (1998) pp. 121-135.
62
63
Concrete.
Concrete
Ozawa, K., Tangtermsirikul, S., and Maekawa, K., Role of Powder Materials on
the Filling Capacity of Fresh Concrete Supplementary Papers, 4th
64
65
Percent passing
(AASHTO M 6)
100
95 to 100
80 to 100
50 to 85
25 to 60
10 to 30
2 to 10
Percent passing
(ASTM C 33)
100
95 to 100
80 to 100
50 to 85
25 to 60
5 to 30
0 to 10
Applicable standard
Filling
ability
ASTM C 1611
Passing
ability
L-box
J-Ring
V-funnel
ASTM C 1621
Filling
capacity
Static
stability
Surface settlement
Column segregation
Visual Stability Index
66
ASTM C 1610
ASTM C 1611
Precautions
o Roughness and moisture of base plate
affect results
o Large base plate is required to perform
test
o Must be performed on level surfaces
T-50
o Sensitive to roughness and moisture of
base plate
o Poor single- and multi-operator
repeatability
o High error for low viscosity mixtures
o
o
Precautions
o Roughness and moisture of base
plate affect results
o Large base plate is required to
perform the test
o Must be performed on level
surfaces
L-box
Good repeatability
o Must be performed on level
Can be performed by a single
surfaces
operator
Flow time correlates to plastic
viscosity
V-funnel
Can be performed by a single
o Poor repeatability
operator
o Risk of flow interruption in high
Flow time correlates to plastic
viscosity mixtures
viscosity
67
Precautions
o Good repeatability
o Good indicator of filling capacity, which
combines filling ability and passing ability
of SCC
o Visual appreciation of filling capacity
through congested sections
Difficult to perform by
single operator
Requires some
calculation to evaluate
filling capacity
o
o
o
o
o
68
Advantages
Easy to perform in laboratory
Good repeatability, especially for
plastic viscosity
Provides fundamental flow
properties of SCC
Enables evaluation of structural
build-up of SCC at rest
Limitations
o Expensive apparatus, though
portable and more affordable
models are available
o Requires qualified personnel to
operate and interpret data
Property
Test method
Target values
Design
Q/C
Filling
ability
Slump flow
T-50 (ASTM C 1611)
J-Ring flow
(ASTM C 1621)
L-bx blocking ratio
(h2/h1)
Filling capacity
Slump flow and
J-Ring tests
Slump flow and
L-box tests
0.5 1.0
70% 100%
Passing
ability
Filling
capacity
Static
stability
Surface settlement
Column segregation
(ASTM C 1610)
VSI (ASTM C 1611)
Air
volume
AASHTO T 152
69
Slump flow
(ASTM
C1611/C16
11M-05 )
Element characteristics
Short
Moderate
Length
Long
Thin
Moderate
Thickness
Thick
Low
Medium
High
Coarse
agg.
content
70
90%
Deep
75%-90%
Depth
Caisson
filling
capacity
70%-75%
Moderate
0.7
Shallow
0.6-0.7
Congested
Element
shape
intricacy
0.5-0.6
Moderate
2 in.
Small
Reinforcement
density
2-3 in.
High
3- 4 in.
Medium
27.5-29 in.
Low
25-27.5 in.
23.5-25 in.
Relative
values
L-box
blocking
ratio (h2/h1)
Viscosity
Mechanical properties
Recommended values
Maximum surface settlement 0.5%
Column segregation index (Iseg) 5%
Percent static segregation (S) 15%
Plastic viscosity 0.0725 psi.s (500 Pa.s)
(Modified Tattersall two-point rheometer with
vane device)
Core-to-cylinder compressive strength 90%
(similar curing conditions)
Bond strength modification factor 1.4
Discharge
rate
Discharge
type
High
Medium/
High
High
Low/
Medium
Continuous
Single
discharge
volume
High
Continuous
Medium
Discontinuous
Low
Continuous
Medium
71
Relative
energy
delivered
High
High/
Medium
Medium
Low/
Medium
72
73
INTRODUCTION
Self-consolidating concrete (SCC) is undoubtedly the major advance of
concrete technology in the last two decades. Mainly due its flowability, SCC
differs from conventional vibrated concrete essentially when in fresh state.
Guidelines and Standards
Currently, a number of guidelines are available for the proportioning,
characterization, control and application of SCC [e.g. 1-7] which are accompanied
74
by national standard test methods [e.g. 8-13] and, in some countries, like Spain,
by code specifications [14].
The recommendations by the Japanese Society of Civil Engineers [1] define three
types of SCCs according to the use of fines (type P), viscosity modifying agents
(type V) or the combined used of both (type C). In type P, self-compactability is
obtained by minimizing the ratio water/fines through the use of superplasticizers
to provide appropriate resistance to segregation and bleeding. In the case of type
V, the use of Viscosity Modifying Admixtures (VMA) provides enough stability
to the concrete towards bleeding and segregation, since these admixtures fix the
water providing the appropriate rheology to the cement paste. Type C consists of
the combination of the other two types.
On the other hand, the European Guidelines [4] define different types of SCCs
according to its fresh properties, regardless the composition of the concrete.
Along these lines, there are three levels of self-compactability according to the
value of slump flow:
SF1: from 550 to 650 mm (21.7 to 25.6 in.)
SF2: from 660 to 750 mm (26.0 to 29.5 in.)
SF3: from 760 to 850 mm (29.9 to 33.5 in.);
two levels of viscosity according to the value of T50 or the time to pass through
the V-funnel, Tv:
VS1/VF1: T50 2 s or Tv 8 s
VS2/VF2: T50 > 2 s or 9 s Tv 25 s
and two levels of passing ability according to the passing ratio in the L-box:
PA1: 0.80 with 2 bars
PA2: 0.80 with 3 bars.
In spite of the fact that these two guidelines [1,4] involve the use of a wide range
of SCC both considering its composition, as well as its fresh properties, the reality
of the market indicates that SCC is mainly characterized by high content of fines,
usually in the range from 450 to 600 kg/m3 (28,1 to 37,5 lb/ft3), along with a low
volume of coarse aggregates. This high content of fines, which provides high
volume of cementitious paste (i.e., within 35-40% by volume), is needed in order
to provide flowability, passing-ability, and enough cohesion able to prevent
bleeding and/or segregation.
SCC rheology
Though historically with only academic interest, rheological studies using
concrete viscometers have become more important with the appearance of SCC,
75
up to the point that the material itself can be defined and characterized in terms of
its rheological parameters, such as the yield stress ( 0) and plastic viscosity ()
[15]. The rheological studies have permitted a deeper comprehension of the
material behavior, allowing in this way a more rational mixture proportioning.
SCC is characterized by having very low yield stress ( 0 ) and enough plastic
viscosity (), which redounds in the material flowability, passingability, and
segregation resistance during transport, casting and hardening; the engineering
properties of the material. In this manner, SCC requires an adequate combination
of 0 and to allow for mobility without segregation. Generally, very viscous
mixes require a 0 value close to zero, while a higher value of such parameter can
be convenient in the case of low viscosity mixes. A combination of extremely low
0 and could imply risks of segregation [15].
The composition of SCCs varies significantly from one country to another.
According to [16], the content of fines in Sweden and The Netherlands is usually
higher than 550 kg/m3 (34,3 lb/ft3) with yield values and plastic viscosity values
ranging from 0 to 30 Pa and from 50 and 120 Pas, respectively. These ranges
agree with those shown in a recent study by [16]. In Denmark, Norway and
Iceland, the content of fines is usually lower than 450 kg/m3 (28,1 lb/ft3) with
yield values and plastic viscosity ranging from 10 to 60 Pa and from 20 to 45 Pas,
respectively. Along these lines, the concrete with higher content of fines is
characterized by a higher plastic viscosity and is suitable for heavily reinforced
structures, whereas the concrete with lower content of fines show higher yield
values being more suitable for lightly reinforced structures.
Even when the use of rheometers or viscometers leads to a better characterization
of conventional concrete and SCC [17], their use is usually more expensive and
difficult in field applications. Progress is being made to address this issue and
some portable equipment has been recently developed [18,19]. In this way, the
relationships between rheological parameters and results from engineering type
tests turn to be of great practical importance. In this regard, it is shown in [20] that
SCC prepared with similar types of component materials show direct relationships
between the slump flow diameter and the yield stress and also between the plastic
viscosity and the flow times (T50 or TV), even when the concrete temperature, the
mixing energy, the environmental conditions or the time after mixing was varied,
Fig. 1.
The actual situation- The use of SCC in the precast industry has increased
significantly from its introduction in the market, in the mid 1990s. It is estimated
that, in Europe, more than 50% of the concrete used for structural precast
elements is SCC. However, the situation is quite different in the ready-mixed
concrete industry where less than 1% of the concrete produced is SCC. The
reasons of such scenario are mostly related to the higher cost of the material,
76
which, very often, is supplied with excessive mechanical performance due to the
higher cementiticious paste content necessary to achieve self-compactability
requirements. On the other hand, the use of additional fines (fillers) can also
imply new silos and more quality control testing that, along with the elevated
cement content, and the higher sensitiveness of SCC to variations of the mix
proportions, can consequently lead to an increase in production costs.
Evidently, it is technically difficult to prepare SCC with low content of fines
(cement + filler), even using currently available VMAs. On the other hand, the
characteristics of locally-available materials (i.e., cement, aggregates and filler)
for SCC production are not always the most appropriate, and consequently, a lot
of laboratory work is needed in order to adjust the mixture proportions with the
aim of obtaining stable concrete. Also, variations in the moisture content of the
aggregates, especially of the sand, can significantly influence the stability of the
SCC, implying continuous adjustments in the mixture proportions.
Consequently, the use of SCC as a common every-day concrete is nowadays
difficult, especially in the case of the ready-mixed concrete industry. Note that,
the required mechanical performance in most of the applications is significantly
lower than that provided by standard SCC (i.e., with high amount of fines) and,
therefore, SCC is supplied, very often, with suboptimal performance.
An innovative alternative
The above reality justifies the development of an innovative VMA, which is
allowing the design of SCC with the required stability (bleeding and segregation
resistance) but without incorporating extra fines and, consequently, having a total
content of fines (cement + filler) of about 380 kg/m3. The use of this new family
of VMAs, along with superplasticizers especially designed for SCC, leads to the
development of smart dynamic concrete [21-23], in other words, to an
improvement in the cost-effectiveness of SCC that can contribute expanding its
use in the ready-mixed concrete industry, as well as consolidate its use in the
precast concrete industry.
This study presents the results of industrial-scale testing in different European
countries, aimed at achieving the following objectives: proportioning SCC with
normal cement contents for everyday concrete, study the possible upgrade of
conventional concretes of high workability into concretes with self-compacting
characteristics, and the possibilities of replacing/reducing the filler content of
reference SCC maintaining the compressive strength and its evolution.
Results demonstrate the usefulness of the new VMA to overcome the common
drawbacks of current type of admixtures, and its capacity to increase the mix
robustness and decrease the necessity of filler materials, if desirable. Such
powerful capacities could certainly help self-compacting concrete increase its
77
market share in the ready-mix sector and further consolidate its use in precast
applications, leading to more productive, environmentally friendly, optimized and
smart construction processes.
MECHANISM OF ACTION
The preparation of SCC with low fines content requires the optimum control of
the rheological parameters providing the appropriate balance between the high
fluidity needed and the stability required (bleeding and segregation). In this
respect, a VMA able of assuring enough plastic viscosity without sacrificing the
slump flow of the concrete, that is, without modifying significantly the yield
value, is essential.
Some state of the art VMAs belonging to different chemical families (natural
gums, starches and synthetic polymers) modify the rheology of concrete mainly
by increasing the yield value without providing significant improvement of the
plastic viscosity as can be seen in Fig. 2. Note that these tests were done in
mortars from concretes that were sieved at 2 mm and tested with a Viskomat NT
rheometer from Schleibinger Gerte. Note, also, that the dosage of each VMA
was adjusted in order to obtain a slump flow of about 65 cm in each concrete.
Moreover, in other cases, the VMA does not affect the yield value but provides a
slight increase in the plastic viscosity. Note, additionally, that in other cases (see
Natural Gum A and B), both the yield value and the plastic viscosity are not
affected when compared to the reference mortar. Nevertheless, as can be seen in
the figure, the new and innovative VMA is able to increase significantly the
plastic viscosity of the mortar without influencing the yield value and,
consequently is able to provide stable concrete without affecting significantly its
slump-flow.
This innovative chemical admixture consists in a new ionic polymer of high
molecular weight whose molecules interact between them apart from interacting
with the water molecules. This leads to a tridimensional structure that embraces
both water molecules and fine solid particles (cement and the fines belonging to
the sand). This tridimensional network made up, not only by VMA-water
interactions, but also by VMA-VMA and VMA-solid particles interactions as
shown in Fig. 3, leads to a strong interconnected molecular network able to
provide, at the macroscopical level, enough stability in the concrete.
The use of this innovative VMA provides the appropriate stability in the concrete
towards segregation and/or bleeding in concretes with paste volume significantly
lower than those used usual for SCC. Along these lines, the rheological behavior
of this type of concretes depends on the optimized mixture proportions of the
concrete, as well as the use of chemical admixtures of last generation.
78
79
80
In plant C from Italy, the reference concrete has a consistency class S4 and
exposure class XC2 (characteristic compressive strength of 30 MPa or 4350 psi)
according to EN 206-1. The durability class fixes the amount of cement and the
water/cement needed. The mixture proportions of the reference concrete are
shown in Table 3 (Reference-C). Accordingly, 800 m3 (1046 yd3) of concrete with
the composition shown in Table 3 (SDC-C) were prepared in series of 10 m3
(13 yd3). The mixing was done directly in the truck since no mixer was available
in the plant. The chemical admixtures (superplasticizer and the new VMA) were
incorporated along with the mixing water. The concrete was transported one hour
before placing (Fig. 6), and the temperature at the job site was of 30C (86F).
In plant D from Turkey, the reference concrete was a conventional concrete with
characteristic compressive strength of 25 MPa (3600 psi) and S3 consistency that
had to be vibrated. The superplasticizer used was naphthalene-based. The
composition of this reference concrete, as well as the composition of the concrete
with the new VMA, is shown in Table 4. In this case, 8 m3 (10 yd3) of concrete
SDC-D with self-compacting characteristics were prepared in batches of 2 m3 (2.6
yd3) that were pre-mixed before discharging in the truck. The concrete, which
shows a slump flow of 71 cm (27.9 in.), was transported throughout 90 minutes at
38C (100F). After this period, the slump flow was measured again and was of 60
cm (23.6 in.).
Finally, in plant E from Germany, the corresponding reference mixture
proportioning was optimized and upgraded to self-compacting concrete. The
reference concrete at this plant had a consistency of type F5 according to DIN
18218 with a composition shown in Table 5, and need vibration. In this case,
24 m3 of concrete SDC-E, as described in Table 5, was prepared. The concrete
was transported for 30 minutes before being placed (Fig. 7).
In these cases, the use of the new VMA allowed transforming conventional
vibrated concretes of S3, S4, and F5 classes into truck-mixed medium-range selfcompacting concretes, capable of maintaining its self-compactability during
realistic periods of time, even at elevated temperatures. The robust SCCs achieved
also maintained the reference compressive strength.
Replacement of the filler used in a reference self-compacting concrete
maintaining the strength development
The objective of this test program was to eliminate completely the filler used in
the mixture proportioning of a reference SCC and to propose an alternative
composition able to provide similar strength development at plant F, in Spain.
This would permit to free a silo that is currently used to store filler and,
consequently, to improve the logistics at the plant. The sand available was natural
siliceous sand with about 1% of fine content ( < 0.125 mm or 0.005 in.). The
81
composition of the reference SCC is shown in Table 6. The cement and filler
content were of 330 and 170 kg/m3 (20.6 and 10.6 lb/ft3), respectively.
The use of the innovative VMA permitted to prepare a new mixture proportioning
(SDC-F) with similar strength development using 400 kg/m3 (24.9 lb/ft3) of
cement. The fines content (cement + filler) was reduced in 100 kg/m3 (6.2 lb/ft3)
compared to the reference SCC (Reference-F). Both concretes had similar
appearance (Fig. 8).
CONCLUSIONS
From the range of materials and conditions studied, the following conclusions can
be drawn from the results obtained:
The new and innovative viscosity modifying admixture tested in this study
permits obtaining concretes with self-compacting properties but with fines content
(cement + filler) lower than 380 kg/m3. This can contribute to extend the use of
SCC in the ready-mixed concrete industry as an everyday concrete and to extend
its use in the precast concrete sector.
Such VMA provides enough cohesion and stability in the concretes by increasing
its plastic viscosity without modifying significantly the yield value.
The concretes incorporating the new viscosity modifying admixture show
appropriate robustness toward changes in the composition, especially water
content, as well as variations in the characteristics of the materials. Allows a)
proportioning SCCs with normal cement contents for everyday concrete, b)
upgrading conventional concretes into robust concretes with self-compacting
characteristics, and c) replacing/reducing the filler content of SCC maintaining
the compressive strength and its evolution.
The achieved smart dynamic concretes with medium-range self-compactability
can be truck-mixed and are capable of maintaining its self-compacting
characteristics during realistic periods of time, even at elevated temperatures.
REFERENCES
1. Japan Society of Civil Engineers (JSCE), Recommendation for SelfCompacting Concrete, Tokyo, Japan, August 1999.
2. Swedish Concrete Association, Report No. 10 (E) Self-Compacting ConcreteRecommendations for Use, Stockholm, Sweeden, August 2002.
82
83
84
Reference-A
330 (20,6)
760 (47,4)
700 (43,7)
412 (25,7)
175 (10,9)
SDC-A
350 (21,8)
840 (52,4)
700 (43,7)
310 (19,4)
190 (11,9)
7,6 (0,5)
5,5 (0,34)
---
0,5 (0,03)
63 cm (24,8 in.)
57 cm (22,4 in.)
Slight bleeding.
Good, without
Presence of fly ash in
Appearance
bleeding or
the surface of the
segregation
concrete
Compressive strength, 7 days
39,9 MPa (5790 psi)
25,6 MPa (3710 psi)
* Composition of Portland cement CEM II/B-M according to
European Standard EN-197: 65-79% of clinker + 21-35% of blends of
slag, silica fume, natural pozzolan, fly ash or limestone + 0-5%
minority components
85
Reference-B
558 (34,8)
596 (37,2)
604 (37,7)
400 (25,0)
260 (16,2)
SDC-B
558 (34,8)
646 (40,3)
604 (37,7)
350 (21,8)
255 (15,9)
7,0 (0,44)
7,0 (0,44)
----
1,0 (0,06)
77 cm ( 30,3 in.)
Good, without
bleeding or
segregation
73 cm (28,7 in.)
Good, without
bleeding or
segregation
Compressive strength,
7 days
38,0 MPa (5510 psi)
22,0 MPa (3190 psi)
28 days
47,1 MPa (6830 psi)
33,2 MPa (4820 psi)
* Composition of Portland cement CEM II/B-M (V-S-LL) according to
European Standard EN-197: 65-79% of clinker + 21-35% of blends of
fly ash, slag and limestone + 0-5% minority components
86
Reference-C
992 (61,9)
396 (24,7)
596 (37,2)
30 (1,9)
280 (17,5)
165 (10,3)
2,6 (0,16)
---
SDC-C
1090 (68,0)
820 (51,2)
--30 (1,9)
300 (18,7)
180 (11,2)
3,2 (0,20)
0,4 (0,02)
S4 20 cm (7,9 in.)
68 cm (26,7 in.)
slump
T50
-->2 s
Compressive strength, 1 day
NA
10,5 MPa (1520 psi)
7 days
NA
26,4 MPa (3830 psi)
28 days
NA
34,1 MPa (4950 psi)
* Composition of Portland cement CEM II/A-LL according to
European Standard EN-197: 80-94% of clinker + 6-20% of
limestone + 0-5% minority components
Slump Flow
87
Reference-E
753 (47,0)
368 (23,0)
630 (39,3)
120 (7,5)
280 (17,5)
175 (10,9)
3,5 (0,22)
---
Properties
Slump flow
T50
Compressive strength,
SDC-E
774 (48,3)
172 (10,7)
774 (48,31)
80 (5,0)
300 (18,7)
185 (11,5)
4,5 (0,28)
0,6 (0,04)
88
89
900
SCC-12
800
SCC-20
700
600
500
400
300
0
40
80
120
6
SCC-12
SCC-20
5
4
3
2
1
0
0
40
80
120
40
SCC-12
SCC-20
30
20
10
0
0
40
80
120
160
90
12
Starch A
Synthetic Polymer B
Starch B
Natural Gum D
NEW VMA
Torque (Nmm)
8
Synthetic Polymer A
Natural Gum C
REFERENCE
Natural Gum B
Natural Gum A
0
0
40
80
Speed (rpm)
120
160
91
92
a) Reference-A
b) SDC-A
Fig. 4 - Appearance of (a) reference concrete and (b) smart dynamic concrete
in plant A
93
a) Reference-B
b) SDC-B
94
95
a) Reference-F
b) SDC-F
96
97
INTRODUCTION
General
Self- compacting concrete (SCC) has been gaining in popularity since its
development in Japan in the late 1980s. Much work has been undertaken to
establish mixture proportions methods to quantify and optimise its properties.
SCC is differentiated by its properties in the fresh state and has found many insitu
and precast applications due to speed of placement, improved form filling,
enhanced surface finish and noise and vibration reduction.
Production of SCC for the Gulf Region favours the use of performance-based
rather than prescriptive specifications. This paper compares prescriptive and
performance based approaches and discusses the use of performance-based
specifications for the fresh and hardened properties of SCC, particularly with
regard to durability design of reinforced concrete structures in extreme
environments.
Tools, such as deterministic and probabilistic deterioration modelling and
reliability analysis, which provide the designer and specifier with a robust process
for determining performance limits are introduced. An example of the techniques
and process of modelling and performance design is included.
98
99
100
101
that require careful consideration are chloride ingress, sulfate attack and salt
crystallisation resulting in surface scaling and weathering.
Durability design needs to be part of an integrated and structured process that is
typically driven by the clients requirements for in-service performance (often
with zero-maintenance requirements throughout the design life), which takes
account of both structure-specific operational and environmental loads. Fig. 1
illustrates an example of an integrated durability design process. The design
process should be iterative, selecting and optimising for appropriate available
materials, performance characteristics and structural function.
Specification routes for durability design
Two principal routes for undertaking durability design are presented in Fig. 1,
either a prescriptive or performance-based approach, or more commonly a
combination (hybrid) of the two:
The prescriptive (or deemed-to-satisfy) approach typically relies on
current national standards and guidance to arrive at specific limiting
proportions to concrete mixture constituents, proportions and cover for a
given service life and exposure, commonly stated in published tables (e.g.
BS EN 206(7), BS8500(8), BS 6349(9), ACI 318(10), Concrete Society CS
163(11)). The requirements in published tables are defined on the basis of
practical experience and engineering judgement.
The performance approach aims to achieve parameters that demonstrate
the suitability of the concrete and its constituents for a given environment
and for an explicitly given design life. The durability performance
parameters are often derived by performance testing and track record,
often accompanied by durability modelling. In its purest form,
performance-based designs are not restricted by limiting the proportions of
concrete constituents, unless required for performance relating to fresh
properties.
Typically, current practice is a combination of mainly prescriptive elements
combined with certain criteria for which a performance based approach is applied
(such as compressive strength, cover and consistence (i.e. workability or flow)).
Regardless of whether prescriptive or performance approaches are implemented,
the design method should provide a structured, cohesive and defensible approach
to the design for concrete durability for SCC.
102
103
suitable for use in specifying durability in the Gulf environment, and should be
avoided. Furthermore, chloride ingress modelling indicates that, on a prescriptive
basis, high covers would be needed for long design lives in extreme
environments.
Performance based durability design
From benign to severe exposure conditions, for structures with normal service
life, appropriate detailing, concrete quality and minimum cover from prescriptive
guidance is likely to be sufficient. However, the importance of maintaining
durable structures in extreme conditions or with extended service lives may
require the designer to consider alternative methods to enable durability of the
structure to be determined on a more realistic basis.
Performance based specification for durability can provide a greater level of
comfort to all stakeholders, that the structure will achieve its intended design life,
thus reducing the risk of costly maintenance and remedial works in later life.
The definition of a performance-based approach is simply that concrete materials
shall be designed and specified on the basis of their demonstrated ability to
adequately perform in a given environment for a specified time period. The
design life should not be less than that required by the client but may be longer to
account for uncertainties, or to increase the probability of achieving the required
service life.
The concept of design life requires the end of the design life to be identified. The
deemed to satisfy approach of prescriptive designs does not ascertain what
the failure of a structure (i.e. its limit state) actually corresponds to, whether it is
intervention by repairs or complete collapse.
The end of a design life in performance based specification should usually be
characterised by some level of damage caused by the deterioration process (e.g.
corrosion initiation, cracking or spalling of concrete). This is generally
understood as the Serviceability Limit State (SLS), indicating service failure or a
time of major intervention, as opposed to the Ultimate Limit State (ULS) which
implies failure or collapse of the structure.
The following guidance is available for performance-based durability design.
The FIB Model Code(15) attempts to establish a durability design approach
similar to that used for structural design and outlined in Eurocodes (EC2)(16),
104
105
(1)
where:
Z>0 the concrete is in a desired state
Z=0 the concrete has reached the limit state
Z<0 the concrete is in the adverse state
Fig. 3 illustrates the input distributions for load, S, and resistance, R. Random
sampling of these input distributions is undertaken, often by using techniques
such as Monte Carlo or Latin Hypercube techniques. The output of the model is a
service life distribution, Z, which than then be assessed statistically to give
estimates of the service period (tp) for a given limit state failure probability.
106
If R and S have a normal distribution, Z will also be linear, the reliability index
and the probability of failure (Pf) can be calculated using the standard normal
distribution(19).
Pf
PR
( R)
2
( R)
(S )
2
(S )
(2)
where: (.) = standard normal distribution function; (R,S) = mean of the R and
S distributions; and (R,S) = standard deviation of the R and S distributions
The value of (- ) can be found in tables for the standard normal distribution.
The reliability of a structure can be expressed as the failure probability (- t), or
directly as .
The value of (- t) or is given in Codes of Practice. The current Eurocode EC2
gives the serviceability limit state (SLS) for a 100 year design life as between
1.5 1.8, which corresponds to a (Pf) of 6.7 x 10-2 to 3.6 x 10-2, that is, between 7
and 4% failure probability.
EXAMPLE: DETERMINING DURABILITY PERFORMANCE
PARAMETERS FOR MARINE WORKS USING SCC
General
Performance based designs are currently being used in the Gulf for major
structures to determine the durability requirements using parameters and
distributions appropriate to the environment. The following example is based on
a case study of using SCC in the UK, but modified and compared with the Gulf
Region environment using relevant input data.
Background
The construction and maintenance of coastal defence structures, such as
breakwaters and harbour walls, represents a considerable challenge to provide
long-serving, durable structures that are sympathetic to the character of the
surroundings. SCC concrete with appropriate use of pigments and patterned
formwork provided a solution to this challenge.
In order to develop patterned pigmented SCC, laboratory and field work was
undertaken to match the colour and texture with the local stonework. In addition
107
108
5% failure probability. Each scenario assumes the same concrete quality, only
environmental conditions were changed.
Scenario 1: UK conditions, 13 C mean ambient temperature, 50mm cover.
Scenario 2: Gulf conditions, 33 C mean ambient temperature, reduced longterm hydration, 50mm cover.
Scenario 3: Gulf conditions, 33 C mean ambient temperature, reduced longterm hydration, 110mm cover.
The concrete under evaluation is considered to be a high quality triple blend
concrete containing both GGBS and silica fume, with a chloride diffusion
coefficient of 1.36 x 10-12 m2/second measured at 28 days from casting.
Predictive modelling
Predictive modelling was undertaken using a chloride ingress model developed
in-house at Halcrow. Established techniques based on Ficks second law of
diffusion and empirical relationships (TRL(20) and DETR(21)) were adapted to
predict chloride ingress and corrosion. The model enables the user to study the
effect of a number of boundary conditions on concrete mixtures and formulate the
most suitable basic mixture proportions that are durable over its specified service
life.
The model has been developed to provide the capability of running probabilistic
scenarios using @Risk, a MS Excel add-in allowing random sampling of input
distributions. The three scenarios were run for 5000 sampling iterations to derive
output for further analysis.
Output of probabilistic modelling
The output of the modelling produced probability distribution functions for the
amount of chloride ingress to the depth of the reinforcement cover. The three
scenarios were analysis statistically and expressed as failure probability with time
and are presented in Fig. 4.
Fig. 4 illustrates that in UK conditions, the concrete is capable of achieving a 100
year design life at below the 5% probability of failure (i.e. for onset of corrosion)
at 50mm mean cover. This represents a saving of 10mm cover when compared
with the requirements of BS 6349, which may be further reduced if the limit state
was relaxed (e.g. to onset of cracking due to corrosion).
109
However, if the same concrete, with the same cover, were to be used in a similar
marine environment in the Gulf Region, modelling estimates that the specified
failure probability would be exceeded after only 22 years, corresponding to a 92%
chance of failure at 100 years. By changing cover depth only, it is estimated that
an increase in mean cover to 110mm for the same concrete would be necessary to
achieve the required failure probability.
Therefore in basic form, a performance specification can be derived from the
modelling to specify that; for concrete to perform with a 5% probability of
failure (to onset of corrosion), a mean cover of 110mm or greater is required, and
the chloride diffusion coefficient of the concrete should not be greater the 1.36 x
10-12 m2/sec at 28 days.
Additional analysis could be undertaken to value engineer the durability of the
structure by improving the mixture further and reducing the diffusion coefficient,
or by considering additional protective measures to reduce cover (e.g. use of
corrosion inhibitors, controlled permeability formwork or stainless steel
reinforcement).
The above requirements can then be added to the specification to provide the
contractor/ concrete producer with a set of requirements (including for SCC in the
fresh state) that are to be demonstrated during mixture proportioning and
development. This defensible approach gives greater confidence to clients and
owners that appropriate concrete for their structure has been identified by
demonstrating its performance, rather than rely on prescribed values in guidance
that not only are inappropriate for any demanding structure requirements
proposed, but also potentially producing savings in less severe environments.
CONCLUSIONS
Self compacting concrete (SCC) is differentiated from conventionally placed
concrete by its ability to flow and level without additional vibration, without
segregation. It is generally regarded to have similar properties to conventional
concrete in the hardened state.
Performance specifications are applicable to SCC in both fresh and hardened
states. When performance specifications are applied for SCC in the fresh state, it
is prudent to apply broad limits and rely on trials to determine placing and
compaction performance using established test methods.
Prescriptive or deemed to satisfy specifications are historically employed for
specifying durability. However, in extreme environments and for structures
110
111
112
113
Remarks
SCC comparable or
slightly higher
Tensile strength
SCC comparable
Elastic modulus
Creep
Shrinkage
Thermal expansion
coefficient
Broadly similar
Dependent on aggregate
type
Reinforcement bond
SCC comparable or
slightly higher
SCC comparable
Compressive strength
Durability
114
Table 2 - Comparison of prescriptive durability limits for marine cyclic wetting and drying exposure classes (i.e.
splash/tidal zone)
Code/Standard
Country/
Region
Minimum
Strength
(MPa)
Maximum
water/cement
ratio
BS8500 - 1
UK
50
0.35
60% GGBS,
40% PC
380
45
100
BS 6349
UK
50
0.40
60% GGBS,
40% PC
370
60
100
UK
Concrete
Soc. CS 163
ME
40
0.40
Triple blend
360
65
30
UK
Concrete
Soc. CS 163
ME
60
0.35
60% GGBS,
40% PC
400
65
30
ACI 318
USA
34.5
0.40
Not given
Not given
64
Not given
UFGS
USA
42
0.40
Not given
400
75
Not given
Australia
50
0.40
Not given
400
65
25
CorrPred
(Halcrow Grp)
ME
60
0.35
60% GGBS,
40% PC
400
60
30
CorrPred
(Halcrow Grp)
ME
60
0.35
60% GGBS,
40% PC
380
90
100
AS 4997
115
Cement
Type
Minimum
Minimum
cementitious cover (mm)
content
(kg/m3)
Design
Life
(Years)
Test Type
Standard / Reference
(informative)
Precision
Porosity
Hydrostatic weighing
1.5 %
Water absorption
BS 1881-122
Chloride diffusion
Accelerated migration
15%
Chloride diffusion
Accelerated migration
Non-steady-state (Nordtest
NTB492)
15%
Gas permeability
Oxygen / CO2
Concrete Society TR 31
30%
CO2 diffusion
Accelerated diffusion
DURACRETE
Water permeability
Water
BS EN 12390-9
Cover
Covermeter
BS 1881-204
Temperature
CIRIA C660
Thermal expansion
Isothermal calorimeter
CIRIA C660
Surface absorption
ISAT
BS 1881-208
Electrical resistivity
Two/four probe
116
10%
Table 4 - Input variables and distributions used in probabilistic modelling of chloride ingress and cover values required to
achieve 100 years service life at around 5% probability to initiation of corrosion
Variable/ Input
Parameter
Cement content
(kg/m3)
Water cement ratio
Background
chloride
Mean temperature
Surface chloride
Age factor
Remarks
Total cementitious content (50%
GGBS, 8% silica fume)
Free water cement ratio
Initial chloride content of concrete
(% by mass cement)
Mean ambient over 1 year ( C)
By mass cement (%)
Age factor for chloride diffusion
coefficient
Chloride diffusion
Apparent value at 28 days age, with
coefficient
variation coefficient of 30%
( 10-12 m2/s)
Threshold to initiation (by mass of
Chloride threshold
cement (%))
Mean cover (mm)
Characteristic
Distribution
Value
Type
UK Conditions
Characteristic Distribution
Value
Type
Gulf Region Conditions
400
Deterministic
400
Deterministic
0.38
Deterministic
0.38
Deterministic
0.1
Deterministic
0.1
Deterministic
13
4.5
Deterministic
Pert(2.5, 4.5, 6.5)
33
6.0
0.58
0.45
Deterministic
Pert(3, 6, 9)
Pert(0.2, 0.45,
0.65)
1.36
N(1.36, 0.41)
1.36
N(1.36, 0.41)
0.4
LN(0.4, 0.15)
0.4
LN(0.4, 0.15)
50
115
4.1%
1.75
117
Operational Loading /
Functional requirements
Prescriptive
H yb rid
Flowability/Viscosity
Chloride diffusion coefficient
CO2 diffusion coefficient
Water / air permeability
Porosity
Compressive strength
Minimum/ nominal cover
Tensile / flexural strength
Elastic modulus
Coefficient of thermal expansion
Peak temperature rise
AAR reactivity
Abrasion resistance
Iterative process
Materials Selection
Cement types
Aggregates
Additives
Admixtures
Iterative process
Design Capability Review
Design Compatibility between Structure and Durability
Buildability
Health and Safety
Cost
Operations Phase
118
Distribution of R
R,S
(t)
R ( t)
Distribution of S
(t )
S (t )
Time
Failure probability
Service period
( tp )
119
Strength grade
60
Concrete density
Total cementitious content
Binder type (pc, ggbs, pfa, sf )
Percentage Binder
Water/binder (w/b)
Background Chlorides
MPa
2350
kg/m3
400
GGBS
60
0.35
0.20
kg/m3
% wt of binder
Est by w/c
m2/s
1.50E-13
Y
Y
20
% wt of binder
Carbon Steel
Bar diameter
25
mm
Ambient Temperature
33
Dca at 35 days
1.38E-11
Dca at 20 years
3.28E-13
Dca at 20 years
Age Factor
3.28E-13
0.70
35
m2/s
Exposure Details
Y
8.1
3. Cyclic wet/dry
80
Coating used?
C
Surface chloride level
7.00
% cement
8.05
% cement
0.09
% cement
No
N
N
Minimum
Cover Depth
(mm)
Chloride
threshold >
Time to
initiation
plus cracking
60
28
120
51
0.9
UK conditions, 50mm cover
Gulf conditions, 50mm cover
Gulf conditions, 110mm cover
0.8
0.7
0.6
0.5
In-service performance
achieved
0.4
0.3
0.2
0.1
0
0
20
40
60
80
100
120
140
160
180
200
Fig. 4 - Failure probabilities (SLS=5% probability to initiation of corrosion) for the same concrete in three modelling scan
121
122
Harald Beitzel
123
INTRODUCTION
The concrete-technological aspects for the use of self-compacting concrete (SCC)
are the optimization and increase of compressive strength, porosity, density, the
resistance to the aggressive action of freeze with de-icing salt, creep and
shrinkage. From these concrete-technological requirements result a number of
economic advantages. SCC can be installed even in densely reinforced areas
without the aid of compaction, retaining at the same time the properties of normal
concrete (NC).
OVERVIEW AND PROBLEM
SCC is distinguished through its advantageous rheological and self-dearating
properties. At the same time, SCC has good resistance to segregation. Other
characteristics of correctly processed SCC are high surface density and improved
durability. The processing properties of SCC differ from the processing properties
of NC as defined in DIN 1045 [4]. The main reason for this is the changed
constitutive composition with a high powder content and the use of innovative
superplasticizers which create the extraordinarily good flow properties. Constant
processing properties are requisite for a uniform concrete production process.
Here, the superplasticizers are of great significance (Fig. 1). Only through the
addition of these admixtures does SCC achieve its temporary flow properties. The
time factor thus plays an important role with regard to the workability and quality
assurance of SCC.
The following criteria are used for assessing the mixing times:
mixture quality test results
homogenization test results
processing test results
124
125
126
that a random elementary particle of the mixture can occupy a specific position in
the mixing space only with a certain probability. A stationary distribution state is
reached after an interminably long mixing time, as the result of the mixture
particle's individual probability of movement. A state described as uniform
random mixture is attained only in case of the same probability of movement.
The assessment of the quality of the mixture is regulated in DIN 459 Part 2 [3]
and in RILEM Final Report TC-150 [2]. In these, the scope of testing for
determining the concrete mixture quality is described for defined experimental
concretes.
The weight of the individual specimens should be selected so high that the
components of the mixture that are available in only small proportions can be
captured through the individual specimens. The maximum particle size should, if
possible, also be included in the probability approach.
Based on a survey conducted by the author and the investigations carried out,
approx. 30% of the manufacturers of concrete mixers have obtained a certificate
in compliance with DIN 459 [3] and the RILEM Final Report TC-150 [2] on the
production of SCC and other concretes. The remaining manufacturers do not have
equivalent, objective proof of the performance capability of their concrete mixers.
To better assess the velocity effect on the homogeneity of the mixture, the
coefficient of variation v is compared as a function of the Froude number Fr.
PRESENTATION AND DISCUSSION OF THE RESULTS
Fig. 3 presents, by way of example, the results of four of the concrete mixers
investigated. The mixture efficiency and the mixture quality for the individual
mixer types are presented by the coefficients of variation as a function of the
mixing time. All of the concrete mixers shown have nominal contents ranging
from 1 m to 3 m and different Froude numbers. Standard mixer No. 1 and
No. 2 and mixers No. 1 and No. 2, which were modified by changing the mixing
speed, are presented in comparison.
The results show that shorter mixing times for producing quality-assured SCC are
possible by changing the mixing speed. This shows that economically favorable
mixing times are possible through appropriate adjustments on the concrete mixer,
also for the production of SCC.
The assessment of the individual coefficient of variation show that the modified
concrete mixers investigated - on the basis of Table 1 - can be classified as highperformance mixers (Fig. 4).
With regard to the processing of SCC, a slump flow of over 700 mm was
127
moreover attained for both modified concrete mixers in the investigations (Fig. 5).
The mean compressive strengths WN after 28 d reached with the modified
concrete mixers, at 63 N/mm and 65 N/mm after a mixing time of 35 seconds
each, can be well classified as a high performance concrete.
SUMMARY AND OUTLOOK
The processing properties of SCC differ from the processing properties of NC as
defined in DIN 1045 [4]. The main reason for this is the changed constitutive
composition with high powder content and the use of innovative superplasticizers
which create the extraordinarily good flow properties. For a reliable concrete
production process, constant processing properties must be ensured. Here, the
superplasticizers are of great significance. Only through the addition of these
admixtures does SCC achieve its temporary flow properties. The time factor thus
plays an important role with regard to the workability and quality assurance of
SCC.
The assessment of the quality of the mixture is regulated in DIN 459 Part 2 [3]
and in RILEM Final Report TC-150 [2]. In these, the scope of testing for
determining the concrete mixture quality is described for defined experimental
concretes. Only approximately 30% of the manufacturers of concrete mixers have
obtained a certificate in compliance with DIN 459 [3] and the RILEM Final
Report TC-150 [2] on the production of SCC and other concretes.
The fine-grained composition of the SCC means for the homogenization in the
mixing process that the movements of the mixing tools - due to a higher energy
input - must produce many shear planes in the mixture. Here, thorough
distribution and mixing of the superplasticizers, in low quantitative proportion
compared to the remaining mixture constituents, is also of special significance.
With regards to concrete raw materials, as well as process and plant technology,
the following developments are being observed in concrete technology:
Concrete raw materials:
New binder combinations
New superplasticizer combination
New aggregate combination
Process technology:
New mixing systems
New materials preparation technologies
128
'DAfStb-Richtlinie
2. Beitzel, H.; Charonnat, Y.; Beitzel, M., 'RILEM TC-1 50 ECM - Efficiency of
Concrete Mixers, Final Report', in Materials and Structures, Vol. 26, No. 256
(RILEM Publications, Bagneux, 2003)
3. Deutsches Institut fr Normung (DIN), 'DIN 459 Teil 1/2 - Mischer fr Beton
und Mrtel' (Beuth Verlag, Berlin, 1995/96)
4. Deutsches Institut fr Normung (DIN), 'DIN EN 206-1 / DIN 1045-2 Festlegung, Eigenschaften, Herstellung und Konformitt' (Beuth Verlag, Berlin,
2001)
129
0,8
WP:
P:
SP:
HSP:
0.8
0.7
0,6
0.6
0,5
0.5
0,4
0.4
0,3
0.3
0,2
0.2
0,1
0.1
0
0.0
w/c ratio
0,7
WP
1940
WP
1950
1960
SP
1970
1980
1990
Without plasticizer
Plasticizer
Superplasticizer
High performance
superplasticizer
HSP
today (2007)
Mixer:
Geometry:
Mixing quality
Mixing chamber:
Operational influences
Process :
Diameter
Height
Form
Breadth
Height
Angle of cut
Cutting angle
Form
Number
Working rate
Acceleration of gravity
Filling height
Mixing time
Order of feed
Filling time
Filling variation
Humidity variation
Servicing
Discharge time
Height of fall
130
10
9
8
7
6
5
4
3
2
1
0
30
M1 Fr 0.60
M1Fr 0.30
M2 Fr 0.60
M2 Fr 0.30
40
50
60
70
variation vv [%]
of variation
Coefficientof
[%]
Coefficient
Limit HP M
M1
M2
10
8
6
Limit HP M
4
2
0
w
w/f
< 0.25
2/16
Mixture ingredients
M2: Fr = 0.6
74
72
76
M2: Fr = 0.3
70
68
66
64
62
72
70
68
66
64
62
60
60
30
35
40
45
50
55
Mixing time t [s]
60
65
30
35
40
45
50
55
Mixing time t [s]
60
65
Fig. 5 Influence of the mixing time and the processing speed on the slump
131
132
133
INTRODUCTION
With the growing popularity of specialist concretes such as self-consolidating
concrete, the requirements for information about the raw materials used has
increased considerably for manufacturers. The processes used to produce these
concretes have also had to be considered to ensure they allow the correct
proportioning, mixing and placing on site. The National Cooperative Highway
Research Programme in the USA Report 628 states that the need for adequate
quality control is much more critical with SCC than in the case of conventional
concrete. The same report goes on to say that successful production of SCC
requires greater competence and proper control of materials and equipments used
for production.
THE IMPORTANCE OF WATER
The control of the water content in the production of any concrete is essential but
as will be proven in this paper it is even more crucial in SCC production. Its direct
effect is measurable in terms of strength, fresh properties and durability. These
effects can be broken down in more detail into the following main characteristics
of concretes:
w/c and w/b a change in water content of the concrete mix will directly
affect the w/c or w/b (water/binder)
a/c changes in the water content of the aggregates will, if left unchecked,
also lead to variations in the weights of dry aggregates actually being loaded
into the concrete mix
w/p or water/powder ratio, arguably more applicable to self-consolidating
concretes, is the relationship between water and all cementitious and filler
materials in the concrete mix.
Flowability, Viscosity and Passing ability variation of water contents can
directly and indirectly (by changes in aggregate dry weight proportioning for
example) lead to changes in the concretes flowability
134
Stability unless the mix design is adhered to, for which accurate water
control is required, mix segregation, bleeding and surface settlement can
occur accompanied by their effects on appearance, durability and in the
extreme structural integrity
SOURCES OF VARIATION
Water content changes can originate from different parts of the concrete
production process with differing effects:
1.
2.
accuracy of water feeding system into the concrete mixer, not usually an
issue in modern concrete plants
3.
Clearly the most significant source of moisture variation is the natural aggregate
moisture content.
THE EFFECT OF WATER VARIATION
Variation of water affects the production of concrete in many ways. For a greater
understanding of these effects it is helpful to consider the influence on the
individual constituents and concrete properties that are affected by it. To produce
a consistent SCC, the producer requires consistent materials and processes.
Knowledge of the moisture will therefore help him achieve his mix design.
Moisture will have an effect on:
Dry aggregate weights
A variation of 1% moisture content in a dry aggregate by weight results in a
change in 10 kg of aggregate loaded into the mixer for every 1000 kg of dry
aggregate weighed. Therefore the greater the variation of moisture, and the
greater the weight of aggregate used, the more serious this condition becomes.
As can be seen from Table 1 above, the range of moistures which different size
aggregates can hold as free water, increases with the fineness of the aggregate.
This occurs due to the increased aggregate surface area by weight. Due to
proportionally high use of fine aggregates in the production of SCCs (usually 4855% of total aggregate weight, source: EGSCC), the potential effects of moisture
variation increases in relation to ordinary concretes.
135
136
Mix stability
Maintaining the proportions envisaged and tested in the mix design is important to
avoid bleeding, segregation and plastic settlement. Tight water control would help
control the occurrence of these issues.
COST BENEFIT OF MOISTURE MEASUREMENT
As well as improvements in quality, the use of moisture control also allows
improved profitability for a concrete production operation.
By avoiding over- or under-cementing the concrete producer will reduce cement
costs and increase production yield. Although more difficult to quantify, the
reduced number of sub-standard batches increases profitability. The improvement
is such that the general consensus from Hydronix customers is that the
equipments return on investment period is of between 3 to 6 months.
OPTIONS FOR MOISTURE MEASUREMENT IN CONCRETE
PRODUCTION
Today there are products available in the market that allow the measurement of
the moisture content of aggregates and fresh concretes with a high degree of
accuracy. The microwave technique has emerged as the most suitable for
measurement of moisture in concrete production, mainly due to its accuracy
(usually in the range of +-0.2% to +-0.4% depending on the quality of the
calibration), its lack of influence by dust or colour, and the competitive costbenefit relationship of this type of equipment.
Measuring moisture in aggregate bins, hoppers and conveyor belts
Aggregate moisture measurement is usually carried out using a digital microwave
moisture sensor at the mouth of the aggregate hopper in question. Once the
sensor has been set up correctly, moisture readings can be taken as an average for
each batch, and the plant control system can adjust the correct dry weight of each
aggregate in real time.
Digital microwave sensors may also be placed on a feeder belt under the hopper
as shown below in Figure 5.
Measuring moisture in mixers
Measuring the moisture of fresh concrete in a mixer allows the concrete producer
to have full control over the final moisture content before the mix is discharged,
ensuring that variations due to aggregate wetting/drying during transport from the
hopper to the mixer, or variable water pressure, are eliminated.
137
The digital microwave technique is also being used to accurately measure the
moisture content of materials while they are inside the mixer. The measurements
from mixer floor moisture sensors and in-mix sensors such as the Orbiter are used
to calculate the correct volume of water required to achieve constant water
contents inside the mixer, batch after batch.
HOMOGENEITY CONTROL USING MOISTURE
Using moisture as an indicator of the degree of homogeneity achieved by the
concrete mixer at any stage of mixing is also possible using microwave mixer
sensors. Based on the use of water as an indicator of material dispersion, when
materials are loaded onto the mixer the sensors see variable moisture contents
over time as usually aggregates will be wet, cementitious materials will be dry. As
these materials mix with water and admixtures, gradually a flatter trace is
achieved as the water disperses evenly over the entire mix. A flat trace output
from the mixer sensor indicates that homogeneity has been reached (see Figure 7).
Using homogeneity control also allows the optimization of mix cycles. Once the
producer can evaluate whether the mix is homogenous or not, he can then adjust
the mixing time to ensure that only the time required to homogenize the mix is
used. This in turn has large benefits in terms of reduced power consumption,
reduced wear of parts, and increased production output.
CONCLUSION
The requirement for adequate quality control is more critical with SCC than in the
case of a conventional concrete. A simple and cost-effective improvement to
reduce fluctuations in the fresh and hardened concrete is to install accurate, digital
moisture measurement equipment in the aggregate hoppers and inside mixers.
According to the NCHRP Report 628, the moisture content,aggregate gradation
and variations in fines content of the aggregate should be continuously
monitored and must be taken into account to produce SCC with constant
characteristics. It goes on to say that it is preferable to control the moisture of
sand before every batch of SCC. The moisture content in coarse aggregate must
also be taken into account.
As well as the moisture itself being important to SCC production, it is also
straightforward to use the same equipment to control mix homogeneity before it is
discharged from the mixer.
The use of digital microwave moisture sensors allows high degrees of accuracy,
repeatable results and simple integration to the control system.
138
Finally, not only is there a high degree of improvement in quality derived from
the use of moisture measurement in SCC production but the increase in
profitability is such that the equipments return on investment is made in a matter
of months.
REFERENCES
National Cooperative Highway Research Program (NCHRP), Self-Consolidating
Concrete for Precast, Prestressed Concrete Bridge Elements, NCHRP Report
628, Transportation Research Board of The National Academies, (2009).
Su J.K. et al.: Effect of Sand Ratio on the Elastic Modulus of Self-Compacting
Concrete, Journal of Marine Science and Technology, Vol. 10, No. 1, pp. 8-13
(2002).
139
Moisture %
Range
0 to 16
Fine Sand
Fig. 1
Course Sand
0 to 12
8mm
0 to 10
10mm
0 to 4
12mm
0 to 3
20mm
0 to 2
Filling height and slump flow vs. S/A ratio (source Su J.K. et al)
140
Fig. 2
Fig. 3
141
Fig. 4
Fig. 5
142
Fig. 6
Fig. 7
143
144
145
146
pump in the middle of the rotor case produces low pressure, which serves to
restore the original cross-section of the rotor-hose. This system is distinctive,
because neither valves nor reductions are required. Further, at the point where the
pressure travels from one hub to another, the concrete columns hold is
considerably shorter when compared to other piston pumps. However, the
relatively short delivery pressure is a disadvantage.
The described conveyance methods for concrete pumps are depicted in Fig. 1. On
the left the piston pump is visible and on the right the rotor pump is visible.
At present there is insufficient knowledge of how SCC is transmitted with
concrete pumps. The efficient transportation of SCC with concrete pumps raises
many technical questions. Some publications assume a higher concrete supply
pressure compared to normal concrete whereas other publications maintain the
opposite. There also is no comparative data available with regards to the possible
hoisting height and transport distances, as well as the temperature development
within the concrete amongst others.
To pump SCC over long distances, several questions are of fundamental
importance. Little is known about whether the quality of the concretes
composition is maintained during a casting process without segregation and if its
quality is affected under extreme conditions. It is not yet clear whether or not the
fresh concretes rheological behaviour, while being pumped, will block the
delivery pipe of the pumping cylinders during the induction process. Furthermore,
the danger of segregation over long pumping distances, the fresh concretes
temperature development and the deformation of the delivery pipes has not been
closely examined yet. Finally, both the theoretical pump and hydraulic pressures
during vertical and horizontal transmissions and the expected abrasion rates
during the transmission of concrete over long distances are of interest.
Experimental investigations were conducted in order to answer these questions
and the results have been evaluated from a machine-technical point of view.
EXPERIMENTAL INVESTIGATIONS AND RESULTS
Composition of the SCC
When preparing a concrete mixture to pump, the following criteria must be
considered for the concrete strength classes and for its workability:
Aggregate configuration
Cement content
Cement paste content
147
Grain content
Consistency
When composing the grading curve of the sieve test for concrete aggregates, the
existing debris pores must be considered. At first they contain air and then
become more or less filled with the developing cement paste and/or fine mortar
during the preparation of the fresh concrete.
Further important aspects regarding the SCCs composition are:
The specified volumes of the sands fine material content shouldnt be sharpedged in cement depleted mixtures.
The granulometric composition should be in a convenient range (A/B) and in
the range of sand near the grading curve B according DIN 4226.
The cement content 270 kg/m, improved 300 kg/m, should be cement with a
good ability to retain water.
The mortar content should be determined in subject to the top sized aggregate:
standard values:
148
149
Due to the fine material content and good flow ability of SCC, it can be pumped
over extreme distances, heights and heights of fall. The experiment did however
demonstrate that over greater delivery lengths, the delivery pressure did increase.
150
50
100
40
80
30
20
10
A
B
C
60
40
20
0
0
0/4
4/8
SP
Mixture components
Fig. 2 Mixture proportions (left) and grading curve (right) of the SCC
Cement
Water
Additive
Aggregate
SP
Mixing
Slump flow
700 mm
No
Yes
Concrete pump
151
16
Single-shaft mixer
Mixer platform
Fig. 4 Experimental mixer setup, with mixer platform used for SCC
Measuring equipment
152
90 m
125 mm
Pump
Strain gauges
Flexible hose
Pressure sensors
800
80
700
60
600
40
500
Delivery pressure
20
400
Slump flow
0
300
0
30
60
90
100
Temperature
[%]
change tt [%]
Temperaturechange
change
Pressure
p
p [%]
change
Pressure
30
60
Time t [min]
Time t [min]
153
90
154
Synopsis: The objective of the investigation reported in this paper was to extend
the pumpability of a high-performance concrete with a targeted compressive
strength of 90 MPa to meet stringent requirements of the Dubai Tower project in
Doha, Qatar measuring 250 m in height requiring a pumping rate of 30m3/hr. This
objective was achieved via an extensive mixture optimization and the proper
selection of a combination of polycarboxylate-based superplasticizer and
viscosity-modifying admixture. The project also aimed at improving the
understanding of the behavior of concrete as a fluid under pressure by monitoring
several parameters, including workability, hydraulic pressure during pumping,
concrete pressure, and friction coefficient to extrapolate the various pumping
requirements to achieve the targeted height of 250 m. The investigation aimed
also at correlating workability characteristics of the concrete (T50 and V-funnel
flow) to the pumpability range of the concrete and hence the power required by
the selected pump (Putzmeister BSA1400HP) to meet the project needs.
Two mixtures were developed and tested. Mix I refers to the initial C90 mixture,
and Mix II refers the modified pumpable mixture. Key differences between them
consisted of reducing the maximum size of aggregate from 20 to 10 mm,
increasing the content of cementitious materials, enhancing the aggregate
size distribution, replacing a standard polycarboxylate-based superplasticizer by a
blend of polycarboxylate-based superplasticizer and viscosity-modifying
admixture, and finally increasing the W/C from 0.25 to 0.27.
The investigation showed that the pressure required to pump 30 m/hr of Mix I to
a height of 250 m exceeds the pumps maximum pressure as well as the pipelines
maximum allowable pressure. Therefore, while Mix I can easily meet the
physical requirements of the project (strength, durability, etc.), it failed to meet
the delivery/placement requirements. The proposed modifications of Mix I to Mix
II and the addition of the new combined admixture resulted in more than 80%
reduction of the concretes friction coefficient (), while maintaining the physical
properties of hardened concrete. Field trials proved that Mix II could be pumped
at 250 m with the chosen piston pump functioning at a safe value of 75% of the
capacity of the pump.
Keywords: High-strength concrete, polycarboxylate-based superplasticizer,
pumping, viscosity-modifying admixture, self-consolidating concrete.
155
156
BACKGROUND
To build as fast as possible, concrete is being pumped through pipes to make it
easily reach its final destination. In this pumping stage, some problems could
occur which cause delays and issues difficult to overcome. Sometimes, the
concrete is not suitable for pumping or it was designed to be suitable, but due to
setting or loss of workability, the concrete was no more pumpable at the time of
pumping. This could cause blocking in the pipeline. Therefore, proper quality
management of the concrete mixture must be carried out to anticipate and resolve
these issues. The major idea of prior quality management is the application of the
concrete composition and other properties of the concrete to predict long-term
properties. For high-rise pumping of high-strength concrete, prior quality
management was used to assess the properties of conventional slump concrete
that can influence pumpability. The expected pumping pressure and pump
performance were calculated by determining the average friction factor and the
required motor power for a certain pumping height.
Basic knowledge about pumping concrete
The pumping process can be executed with a variety of pumps and pipes.
Differences in pumps and pipes can influence the pumpability of the concrete due
to variations in pumping conditions. In the first part of this section, the equipment
used to pump concrete will be discussed. This will be followed by a discussion of
the concrete behavior inside the pipes during the pumping.
The most commonly used pump to deliver concrete in high-rise buildings is the
piston pump for which the working principle consists of pushing a piston back
and forward to deliver concrete. While going backward, the piston creates a
vacuum which pulls the concrete in the piston, and during forward movement it
pushes the concrete through the pipe.
When selecting the pump, the important factors to consider are the maximum
pressure, the length of a stroke, and the duration of a stroke. The last two factors
are important because they determine the flow rate of the concrete leaving the
hopper. The maximum pressure is important because it determines the friction
that could be overcome with the pumping equipment. The pressure applied for
pumping can vary from 40 to 300 bars.
The most significant factor for the pipe is its diameter. Commonly, a pipe
diameter of 100 to 200 mm is used. The difference in diameter is important
because it is one of the factors influencing the velocity of the concrete in the pipe.
Smaller pipe diameter results in higher velocities, hence resulting in an increase in
friction leading to greater pumping difficulties.
157
In order to better understand the pumping process, the behavior of concrete during
the pumping process is explained. Although direct observation is impossible, the
flow of concrete in straight pipes in stationary (steady-state) conditions can be
described by two parts. a plug and a lubricating layer (Figure 1), as can be seen in
centre part, which is called the plug flow, consists of aggregate, sand, cement, and
water. It is important to notice that the velocity over the full width of the plug is
equal. This means that there is no relative velocity within the plug, and that the
forces acting in this plug are small. The outer part, called the lubricating layer, has
a thickness of 1 to 2,5 mm and consists of water, cement and fine sand particles.
The velocity profile in the lubrication layer drops from the value of the flow
velocity of the plug to zero at the pipe wall. This means that in this layer, some
particles are moving faster than other particles, and that several forces are acting
within this layer. This behavior of concrete is only correct in a stationary (steadystate) flow in straight pipes. The flow pattern in non-straight sections, such as
bends, reducers, or bad connections, will be turbulent inducing changes in the
flow pattern and the velocity profile. These changes result in a higher sensitivity
for blocking (Browne & Bamforth, 1977 and Neville, 1997).
Describing pumpability
The main characteristic of concrete for determining its pumpability is workability.
According to Neville (1997), the definition of workability is that property of
freshly mixed concrete or mortar which determines the ease and homogeneity
with which it can be mixed, placed, consolidated, and finished. This means that
the workability is not a fundamental property, but it is related to the method of
placing, consolidating, and finishing. This makes it difficult to measure the
workability of concrete. But, if a concrete has adequate workability, it should also
exhibit good pumpability.
Focusing on the definition of the workability again, it becomes clear that it is a
composite property. The two main components are the consistency and
cohesiveness of the concrete, which are described by the fluidity and segregation
resistance. A good workability is obtained if both the consistency and
cohesiveness are sufficient, leading to a good pumpability as well. Consistency
and cohesiveness of the concrete are influenced by constituting materials in use.
The most important ingredients which influence the consistency and cohesiveness
are the fines and aggregates, cement content, water content, and the use of
admixtures and mineral additives. For better understanding of pumpability,
consistency and cohesiveness are discussed further below (Mehta & Monteiro,
2006, Neville, 1997).
158
COHESIVENESS
The cohesiveness of the concrete is the resistance to bleeding and segregation,
which is also named the stability of the mixture. This property is important
because it describes the homogeneity of the concrete. Bleed water needs to
transfer shear forces to the other materials in concrete. If the cohesiveness is not
sufficient, these forces could not be transferred to the other materials, and
bleeding would occur. The process of bleeding is illustrated in Figure 2. Due to
particle interlock, high friction occurs, and because of a lack of cohesiveness the
forces on the water could not be transferred to the other materials. This results in
water migration between the aggregate particles, resulting in an unsaturated state.
The bleeding of concrete is a problem because concrete in the unsaturated state
will result in high friction flow. As can be seen in Figure 3, the flow resistance
will increase when a concrete is in the unsaturated state. This flow resistance will
certainly be larger than the maximum pressure that the pump could deliver with
blocking as a consequence.
The other property that is important for pumpability is the cohesiveness that
affects the resistance to segregation. Segregation means that the aggregate
particles sink in the mixture. During pumping, aggregates can move further ahead
of the concrete mass in the pipe, relative to the mortar. If the cohesiveness is not
sufficient, aggregate particles could move to the front of the concrete causing a
plug in front of the concrete. This plug will cause high friction in the pipe and
may lead to blocking. To prevent segregation or separation of coarse aggregate
particles from the mortar, the pumping pressure may have to be lower to a value
lower.
The cohesiveness of concrete is not often measured. In most cases, it is estimated
by the normal bleeding test or visual inspection of the slump test. The normal
bleeding test is a time consuming test, as it takes several hours, and the visual
inspection is very subjective as it is dependent on the experience of the operator.
Another test which can be used is the pressure bleeding test (Mehta & Monteiro,
2006, Neville, 1997).
CONSISTENCY
The consistency is defined as the ability to flow, which means the effort needed to
make concrete flow. A concrete with good consistency has low viscosity. The
consistency is an important property for pumpability because it is related to the
friction in the pumpline. Consistency is evaluated by the slump test. According to
Neville (1997), the slump needs to be between 50 and 150 mm for the concrete to
be pumpable. In case the slump is lower than 50 mm, the concrete would be too
stiff resulting in high friction and more difficulties in pumping. A higher pressure
159
is needed to push the concrete through the pipes which will increase the risk for
bleeding or segregation.
To create a mixture composition which will result in good consistency, special
attention must be paid to the grading curve of the aggregate. When there will be a
high amount of coarse aggregates, the aggregates can increase the friction against
the pipe wall if the consistency will is too low. When the mixture has too many
fines, flow friction will also increase. As mentioned previously, the fine sand
particles are part of the lubricating layer in the pipe (Anderson, 1977, (Neville,
1997).
TEST METHODS
This section offers an overview of test methods and their requirements that can be
carried out to ensure successful pumping. The slump flow test is easy to carry out
and can provide an indication of filling ability of SCC. The requirements on final
slump flow and T50 values should be different for mixtures with different
maximum aggregate size and aggregate shape and when using different
admixtures. It is difficult to assess the segregation/settlement tendency of the
concrete using this test. To perform a slump flow test, the fresh concrete is poured
into a mould in the shape of a frustum of a cone. When the cone is lifted upwards,
the distance the concrete has spread provides a measure of the consistency of the
concrete.
The V-funnel flow time can b used to estimate the apparent viscosity of SCC.
However, many factors play a key role in influencing the V-funnel test results,
including the amount, shape and size distribution of the aggregate and the
viscosity and volume of the cement paste. A concrete rheometer can be used to
measure the yield stress and plastic viscosity of the concrete. However, for the
case study discussed in this paper, a concrete rheometer was not used available for
use.
The average friction factor, the expected pressure, and required motor power for a
given pumping height can be assessed by means of measuring the pumping
pressure and calculating the performance of the concrete pump.
TEST RESULTS
Concrete composition
The mixture composition for the C90 concrete that was used for the main core
wall of the high rise structure was optimized to ensure proper pumpability.
Mixture optimization necessitated changes in the water-to-binder ratio, sand-total
aggregate ratio, aggregate maximum size, and the type of superplasticizer. The
results of the mixture optimization are presented in Table 1 and discussed below.
160
The concrete compositions and properties are given in Table 2. The variation in
slump flow and V-Funnel flow time results over 150 minutes of the two tested
mixtures are plotted in Figures 4 and 5, respectively.
INFLUENCE OF SCC ADMIXTURE
The polycarboxylate-based superplasticizer that was incorporated was a blend of
superplasticizer and viscosity-modifying admixture specifically designed for use
in SCC and high-performance concrete. This specialty admixture helps achieve
very high concrete flow while maintaining a adequate segregation resistance
(Figure 6). The admixture also makes the concrete more tolerant to variations in
the water content of the concrete mixture in such way that plastic viscosity is
maintained and segregation is prevented. By making the concrete more cohesive,
the separation of the lubricating mortar fraction from the mixture is prevented
during pumping. SCC made with the special SCC admixture was shown to
maintain consistency for extended periods of time, especially under extreme
conditions where the concrete has to be pumped for long distance.
Pump trial measurements and flow rate calculations
Data Sheet for two pump trials of Mix II and Mix I are summarized in Tables 4
and 5, respectively. Based on the pump trial the following calculations can be
executed. The total concrete volume per stroke was calculated from the concrete
pump delivery piston dimensions and is equal to 0.054 m3. The number of strokes
per hour is calculated as 11 x 60 = 660 Stroke/h for Mix I and 19 60 =1140
Stroke/h for Mix II. Assuming a 90% filling ratio, the concrete flow rate can be
calculated from the following equation:
Q = V S F
(1)
where
V : volume of delivery line (m3)
S : number of strokes per hour
F : filling ratio
This results indicate that the flow rates for Mix I and Mix II were 32.1 m3/h and
55.4 m3/h, respectively.
Concrete pressure
The surface ratio ( i) between the delivery piston (green in the illustration in
Figure 7) and the hydraulic piston (red) determines the hydraulic transmission
ratio. The output Q and concrete pressure Pc depend on this ratio.
161
With a rod side pressurization of the hydraulic system, the transmission is more
efficient compared to the piston side pressurization. This results in a higher
delivery performance but a lower concrete pressure, while for piston side
pressurization, a higher concrete pressure is attained but lower delivery
performance can be achieved.
According to the following physical formula:
Phy A1/2
Phy
i
the concrete pressure Pc can be calculated from the hydraulic pressure Phy, as
follows:
Phy A1 / 2
Pc A3
Phy A1 / 2
Phy
A3
i
The hydraulic pressure can be read from the pump display and control panel.
There is always a system operation pressure of approximately 30 bar, which has
to be deducted from the pressure reading of the manometer at the pump before
calculating the concrete pressure.
Pc
Phy
(2)
A3
30
(3)
(4)
where
Pc : concrete pressure (bar)
i: hydraulic transmission ratio
This results in a concrete pressure of 155 bar for Mix I and only 92 bar for Mix II.
The pressure calculation is based on two positions of the pumping system. The
first position is the current high (Y1+Y2 = 113 m), while the second position is
the target high (Y1+Y2 = 250 m) as can be seen in Figure 8.
In position 1, the friction coefficient is calculated based on site trial data for both
concrete mixtures (Mix I and Mix II), while for position 2, the concrete pressure
is estimated based on the calculated friction coefficient for both concrete
mixtures, as illustrated below.
162
According to the Energy Equation (Bernoullis law) for a fluid, the total energy
can be summarized as the elevation energy, velocity energy, and pressure energy.
The Energy Equation can then be expressed as:
(5)
where
p = pressure in fluid (Pa (N/m2), psi (lb/ft2))
ploss = pressure loss (Pa (N/m2), psi (lb/ft2))
= density of the fluid (kg/m3, slugs/ft3)
v = flow velocity (m/s, ft/s)
g = acceleration of gravity (m/s2, ft/s2)
ploss = p1 - g h2
The pressure loss is divided in major losses due to friction and minor losses due
to changes of velocity in bends, valves and similar situations. The minor losses
are considered as negligible.
The pressure loss in pipes depends on the flow velocity, pipe length, pipe
diameter, and a friction factor based on the roughness of the pipe, and whether the
flow is turbulent or laminar, expressed by the Reynolds Number of the flow. The
pressure loss in a tube due to friction, called the major loss, can be expressed as:
163
(7)
ploss = (l / dh) ( v2 / 2)
where
ploss = pressure loss (Pa, N/m2)
= friction coefficient
l = length of duct or pipe (m)
dh = hydraulic diameter (m)
Equation (7) is also called the Darcy-Weisbach Equation, valid for fully
developed, steady incompressible flow. By applying the above equations in both
positions, using different concrete compositions, the results are summarized
below.
The field trials carried out on Mix I and Mix II with a total pipe length of 195 m, a
vertical length of 113 m, a pipe diameter of 125 mm using a Putzmeister
BSA14000HP concrete pump can be summarized below.
T500
Mix 1
9.2 s
Mix 2
3.4 s
Variance
-62%
Flow
735
mm
720
mm
-2%
VFunnel
30.2 s
6.4 s
-79%
164
Hydraulic
Pressure
300-310
bar
180-200
bar
Strokes /
min
Pump
output (Q)
11
32.1m3/hr
19
55.4m3/hr
-38%
+73%
+73%
The field trial results are extrapolated below for a pipe vertical length of 250
m.
Mix I
Pump output
Concrete
(Q)
Pressure
Values from
trials at
Height=113m
Calculated
values for
Height=250m
Hydraulic
Pressure
required to
pump 30m3/hr
at
Height=250m
32.1m3/hr
Mix II
Pump output
Concrete
(Q)
Pressure
55.4m3/hr
92 bar
250 bar*
30m3/hr
(min.
requirement)
91.5 bar
30m3/hr
(min.
requirement)
155 bar
480 bar
(37% more than pump
maximum capacity)
195 bar
(75% of pump maximum
capacity)
Mix I concrete pressure exceeds safe pipeline operating pressure of 200 bar.
CONCLUSIONS
Based on the field trials to evaluate the pumpability of C90 SCC to enable
concrete delivery to a height of 250 m, the following conclusions can be stated:
1. The pressure required to pump 30m/hr of Mix I to a height of 250m exceeds
the pumps maximum pressure as well as the pipelines maximum allowable
pressure. Therefore, while Mix I easily meets the physical requirements of the
project (strength, durability, etc.) it fails to meet the delivery/placement
requirements.
2. The modification of the mixture proportioning of Mix I to develop Mix II that
included the incorporation of a novel SCC admixture that combines a lowviscosity polycarboxylate-based superplasticizer and a special viscositymodifying admixture has resulted in more than 80% reduction of concretes
friction coefficient (), while maintaining the required physical properties of
hardened concrete.
3. Field trials proved that Mix II can now be pumped successfully at 250m with
the chosen piston pump functioning at a safe value of 75% of its capacity.
165
REFERENCES
1. ACI Committee 304. (1996, January). Placing Concrete by Pumping Methods.
Retrieved September 26, 2007, American Concrete Institute: www.concrete.org.
2. Anderson, W. G. (1977). Analyzing Concrete Mixtures for Pumpability. ACI
Journal 74-42, 447-451.
3. Browne, R.D., Bamforth, P.B. (1977). Tests to Establish Concrete Pumpability.
ACI Journal, 74-19 , 193-203.
4. EuroLightCon. (2000). Stability and Pumpability of Lightweight Aggregate
Concrete Test Methods.
5. Ferraris, C.F. (1999). Measurement of the Rheological Properties of High
Performance Concrete: State of the Art Report. Journal of Research of the
National Institute of Standards and Technology , 461-478.
6. Kaplan, D., Larrard, F.D., and Sedran, T. (2005). Avoidance of Blockages in
Concrete Pumping Process. ACI Materials Journal, 102-M21, 183-191.
7. Khayat, K.H., Assaad, J., and Daczko, J. (2004, March-April). Comparison of
Field-Oriented Test Methods to Assess Dynamic Stability of SCC. ACI Materials
Journal , 168-176.
8. Koehler, E.P., Fowler, D.W. (2003, August). Summary of Concrete Workability
Test Methods. Retrieved August 22, 2007, International Center for Aggregates
Research: http://www.icar.utexas.edu/publications/105/105_1.pdf
9. Larrard, F.D., Hu, C., Sedran, T., Szitkar, J.C., Joly, M., Claux, F., et al.
(1997). A New Rheometer for Soft-to-Fluid Fresh Concrete. ACI Materials
Journal , 234-243.
10. Mehta, P. K., Monteiro, P. J. (2006). Concrete - Microstructures, Properties
and Materials. New York: McGraw-Hill.
11. Neville, A. (1997). Properties of Concrete. New York: John Wiley & Sons,
Inc.
166
12. Pashias, N., Boger, D.V., Summers, J., and Glenister, D.J. (1996). A Fifty
Cent Rheometer for Yield Stress Measurement. Journal of Rheology , 1179-1189.
13. Powers, T., Dahl, L. (1939). The Bleeding of Portland Cement Paste, Mortar
and Concrete Treated as a Special Case of Sedimentation. Chicago: Portland
Cement Association.
14. Zhang, M.Y. (2003). Coal-Water Paste Water-Seeping Properties under
Pressure and its Pumpability. Journal of Engineering for Thermal Energy &
Power, 78-81 (Chinese version).
15. Christian Stirm (2009). Dubai Tower Doha pump trial report.
16. Robert W., Alan T., and Philip J., Introduction to Fluid Mechanics. John
Wiley & Sons, Inc.
167
Mix I
500
0.25
20
Limestone
40
138
322
Mix II
550
0.27
10
Limestone
55
95
400
528
473
804
0
7.5
0
0
892
503
310
0
8.0
168
V-Funnel (Sec)
T500 (Sec)
Flow Slump
(mm)
Test
Air Content (%)
Concrete Temperature (C)
Visual Stability Index
Duration (min)
(Elapsed Time)
60
60
60
5
30
60
90
120
150
5
30
60
90
120
150
5
30
60
90
120
150
Mix I
Mix II
0.8
25
0
750
745
730
680
650
520
4.41
6.16
7.02
7.45
9.02
11.51
20.31
20.54
24.5
31.21
39.13
42.36
0.7
25
1
730
755
755
740
760
755
3.26
2.35
3.29
3.2
3.06
2.5
10.32
7.49
11.54
10.54
9.48
10.24
169
STANDARD
BSEN
BSEN
BSEN
BSEN
BSEN
ASTM
ASTM
ASTM
RESULT
39
75
95
104
8
0.01
0.39
42.0
Grade
Conc.
batch
size
(m3)
90/10
27
First
Concret
e mixer
arriving
Time
6:50:00
5:40 PM PM
Concrete
Batching
Time
T500
(s)*
Flow
Slump
*
(mm)
Concrete
temp.* (C)
V
Funnel*
(s)
3.4
720
31.4
6.4
Machine
No.
Pipeline
Horizontal length (m)
57.5
103
Placing boom
Type of placing boom
Placing boom
operation position
32m
Horizontal length (m)
Vertical length (m)
24
10
Average Strokes/Min
19
170
180 - 200
608656
609211
555
19
Concrete
Batching
Time
Concrete
arriving
Time
T500
* (s)
Flow
Slump*
(mm)
Concrete
temp.*
(C)
V
Funnel
* (s)
9.2
735
30.5
30.2
10:32 11:200:00
PM PM
Machine No.
Pipeline
Horizontal length (m)
57.5
103
Placing boom
Type of placing boom
Placing boom
operation position
32m
Horizontal length (m)
Vertical length (m)
24
10
Hydraulic pressure
in the middle of
stroke (bar)
171
300 - 310
609784
610184
400
172
Flow (mm0
20
40
60
80
100
120
140
160
Mix II
V Funnel Test
50
(Sec)
40
30
20
10
0
0
20
40
60
80
100
Mix II
173
120
140
160
174
175
176
CAN ORDINARY VIBRATED CONCRETE AND SELFCONSOLIDATING CONCRETE BE TREATED IN THE SAME WAY
CONCERNING PUMPABILITY?
177
INTRODUCTION
Pumping of concrete is one of the most used casting methods, well known for its
ease of application, the possibility to apply it in many different situations
(especially in the ready-mix industry), the high construction speed, etc. [1-2].
Many theoretical and practical studies have been performed on the pumping of
concrete in order to investigate several aspects such as different pumping systems
[1-4], the pressure required to pump [3-7], the configuration of the conveying line
178
[1-2] and the consequences of pumping on the concrete properties [4, 8-10]. With
the introduction of Self-Consolidating Concrete (SCC) on the market, the
practical rules for pumping of concrete are still applied. The question remains
whether the same phenomena influence the pumping parameters and whether
SCC shows similar changes in properties due to pumping as Ordinary Vibrated
Concrete (OVC).
This paper compares the results obtained in a research project on pumping of SCC
executed at Ghent University, verified by similar projects at other institutes, with
results available in literature dealing with pumping of OVC. After a brief
introduction into concrete rheology, the experimental setup is described and the
results for SCC are compared with literature. The velocity profiles of OVC and
SCC flowing in pipes are discussed in detail, followed by a study on the influence
of bends on the pumping pressures, the change in air content and air bubbles size
distribution due to pumping and possible variations in rheological properties.
where
(1)
179
From a rheological point of view, the main differences between OVC and SCC
are the lower yield stress in case of SCC, which can be observed by executing the
slump test, and, in general, a higher viscosity for SCC [16].
Transient behaviour
The transient behaviour can be described as the change of the steady-state
rheological properties with time. Theoretically, the transient behaviour can be
divided into three different parts: thixotropy, structural breakdown and loss of
workability [12].
Thixotropy is defined as the reversible breaking of physical connections between
small particles due to shear, and the rebuilding at rest [17]. The most interesting
part of thixotropy in practice is the rebuilding at rest because it can reduce the
formwork pressure exerted by the fresh concrete cast [18]. Thixotropy can be
visualised by monitoring the stress response when a sudden increase and decrease
in shear rate are applied (Fig. 1) [19]. When the shear rate suddenly increases, an
overshoot in the stress is observed and the stress decreases with time (at constant
shear rate) until the equilibrium value is achieved. In case of a sudden decrease in
shear rate, the opposite happens: the stress is too low and rebuilds as a function of
time until the equilibrium value is obtained. At rest, there is no equilibrium value
for the stress, and as a result, the static yield stress (yield stress at rest) increases
linearly with time [20].
As thixotropy is of physical nature, induced by inter-particle forces, structural
breakdown has a chemical nature [11-12]. At the first contact between cement and
water, chemical reaction products are formed which can bond some cement
particles. When the shearing forces are sufficiently high, these connections can be
broken. The effect of structural breakdown due to a sudden increase in shear rate
is similar as for thixotropy, but there is no rebuild when the shear rate decreases.
Loss of workability is regarded as the non-reversible increase in yield stress and
plastic viscosity with time. It can be both physical and chemical of nature, but
finally, the chemical reaction between cement and water transforms the material
from a liquid into a solid.
In practice, the distinction between thixotropy and structural breakdown is not
easy to make. In this paper, the decrease in shear stress at a constant shear rate
will be described as structural breakdown, but it includes both the thixotropic
breakdown and the pure structural breakdown described above. As a consequence,
there exists an equilibrium shear stress at each shear rate and the lower the
number of connections in the concrete, the lower the rheological properties. As a
result, the equilibrium yield stress and plastic viscosity are lower at higher shear
rates.
180
EXPERIMENTALS
Test setup
Full-scale pumping tests were performed inside the laboratory, using a truckmounted concrete piston pump (Fig. 2). This pump has two pistons, containing
theoretically 83.1 liter each, alternately pushing concrete inside the pipes and
pulling concrete from the reservoir. The operator can vary the discharge in ten
steps from 4-5 l/s (step 1) to 40 l/s (step 10). For safety reasons, a maximum
discharge of 20 l/s (step 5) was set during the experiments.
Behind the pump, two different loop circuits were installed. The short circuit
measured 25m and consisted of a straight horizontal section of 12m, a 180 bend
constructed with two 90 bends and a 1m straight pipe, and an inclined part
heading back towards the pump (Fig. 3). At the end of the circuit, a reservoir with
a valve at the bottom was suspended to a rolling bridge, which enabled sampling
and discharge calibration. During tests, the valve was open and the concrete
flowed back inside the reservoir of the pump. The long circuit measured
approximately 100m and was built as an extension of the short circuit with four
straight horizontal sections and 180 bends in between (Fig. 4). The principle is
similar: the circuit formed a loop and the concrete flowed back inside the
reservoir of the pump.
In the main measurement section, which is the only measurement section in the
short circuit and the last straight horizontal section in the long circuit, the pressure
loss was measured with two flush-mounted pressure sensors (Fig. 5). The pressure
loss per unit length was calculated by dividing the difference between the
upstream and downstream pressure by the length between the two sensors. In the
vicinity of the each pressure sensor, three strain gauges were attached to the outer
pipe wall, initially as a back-up for the pressure sensors (in the short circuit) and
finally as independent measuring systems in other sections in the long circuit [3,
6-7]. The deformation of the pipe is related to the local pressure. Close to the
strain gauges, a temperature sensor monitored the temperature evolution.
Discharge was not measured directly, but it was calculated by measuring the time
interval for a certain amount of pumping strokes. As verification, the content of a
full pumping cylinder was pumped inside the suspended reservoir with the valve
closed. The load cell between this reservoir and the rolling bridge registered an
increase in force, which is related to the discharge when density is known.
Comparison of these data with the time measurements shows a very good
agreement between both methods.
181
Testing procedure
As the volume of concrete needed was around 1.5m for the short circuit and
3.5m for the long circuit, the concrete was produced and delivered by a readymix company. In normal circumstances, the concrete arrived at the laboratory at
an age of 45 minutes (relative to the water adding time) and after insertion of the
concrete in the pipes, the first test could start at 1 hour of age. Two different
testing procedures were executed during the investigation. The regular testing
procedure consists of decreasing the discharge stepwise from step 5 to step 1,
maintaining each discharge during 5 full strokes. This tests takes around 4-5
minutes and was repeated each 30 min, until the concrete was removed from the
site. Before this regular procedure, a sample of the concrete was taken and
subjected to a rheometer test with the Tattersall Mk-II rheometer and to different
tests on fresh SCC, as slump flow, V-funnel, density, sieve stability, etc.
The special testing procedure consists of three tasks, repeated for each
discharge step (Fig. 6), after the insertion of the concrete at the lowest discharge
(step 1). The three tasks are as follows:
Maintaining discharge constant until equilibrium in pressure loss is
observed. In practice, the full amount of concrete was pumped through
the circuit. Especially at the low discharges, this task can take several
minutes.
Sampling and discharge calibration (with the reservoir and the load
cell). The concrete sample is used to determine the rheological
properties, slump flow, V-funnel flow time, air content, density and
sieve stability (segregation resistance).
Stepwise decreasing discharge curve, similar as the regular procedure,
but with the discharge from task one as maximum discharge.
Logically, at discharge step 1, there is no decreasing discharge curve.
After the decreasing discharge curve, the discharge is increased by one step and
the tasks are repeated. This procedure takes between 60 and 90 minutes, and due
to some blocking issues during the insertion of the concrete in the long circuit, the
concrete age was 3h 30 min at the end of the test in some cases. As a result, due to
high pressures or high concrete ages, discharge step 5 was cancelled sometimes.
Concrete compositions
In total, 19 concrete mixtures were produced at the mixing plant, of which the
compositions can be found in table 1. Mixtures LM 0, OVC 1, LM 15 and LM 17
were commercial products of the mixing plant, while the other mixtures were
designed by the laboratory. The laboratory mixes contained Ordinary Portland
Cement (CEM I), limestone filler, sand, rounded aggregates, water and a
182
RESULTS
Pumping of ordinary vibrated concrete literature review
Saturated versus unsaturated concrete
When pumping ordinary vibrated concrete, two different situations can occur, as
described by Browne and Bamforth [6]. Either the concrete remains saturated or it
becomes unsaturated due to internal water migration in the concrete caused by the
applied pumping pressure. In the case of saturated concrete, the stress transfer is
of the liquid type and the fundamental laws of rheology can be applied [3,6]. As a
consequence, the pressure loss per unit length does not depend on the local
pressure and the pressure decreases linearly over the length of the pipe [3]. In case
the concrete becomes unsaturated, some zones with a shortage of water are being
created, which results in a frictional type of stress transfer instead of a liquid type.
In this case, Coulombs friction law applies and the frictional stress is dependent
on the local stress [56]. As a result, the pressure loss per unit length does depend
on the local pressure and the total pressure decreases exponentially with the
length of the pipe. The case of unsaturated concretes is the least advantageous
one, as shown by Browne and Bamforth [6]. For a certain setup and certain
concrete properties, they calculated the maximum length of the conveying line
and obtained 250m in case the concrete remains saturated and only 1.1m if a part
becomes unsaturated.
The main pumpability issue for OVC is thus avoiding the concrete to become
unsaturated and avoiding any friction type of stress transfer to occur. As a result,
the amount of coarse aggregates is reduced and some tests are developed to test
the forced bleeding of concrete under pressure [13,6]. Blocking during start-up is
also the consequence of concrete becoming unsaturated. Kaplan described
blocking during start-up as the consequence of the coarse aggregates moving
ahead of the concrete in the water/cement mixture pumped in before to grease
the pipes [3]. Furthermore, the pipe walls are covered by a small layer of cement
paste and the rubber connection seals are filled with cement paste. As a result, the
total cement paste in the first meters of concrete pumped is reduced.
183
184
185
blocking occurs behind a section where the streamlines needed to change (bends)
[3].
Flow in straight sections
Recalling Fig. 7, the flow of SCC is situated in part 2 of the curve, as the yield
stress of the concrete is sufficiently low to allow shearing of the concrete itself.
Nevertheless, a lubrication layer must be present, as the Buckingham-Reiner
equation predicts too elevated pressures at a certain discharge compared to the
experimental results. The SCC pumped during the experiments showed shear
thickening behaviour, which is reflected in the pressure loss discharge curve. As
can be seen in Fig. 8 the pressure loss discharge curve is not linear but it shows
an upward curvature. When plotting the pressure loss per unit of length as a
function of the tangential (or differential) viscosity at 10s-1, a very clear
relationship is obtained for each discharge step, as can be seen in Fig. 9. As a
result, the viscosity of SCC is, for a certain geometry of the conveying line and a
certain discharge, the determining factor for the pressure losses. Reducing the
viscosity of the concrete will logically lead to a reduction in pressure losses.
Flow in bends
The flow in bends is certainly not negligible in case of SCC. The pressure loss
over a 90 bend was found to be equivalent to a pipe length between 0.5m and
5.5m, with an average of 2m, which corresponds very well to the practical
statements in [2]. A 180 bend has an equivalent length between 3.5m and 11m,
with an average of 6m. Due to the large scatter in the results, no conclusions for
the equivalent length can be taken. Furthermore, for different rheological
properties of the concrete, different aggregate content and aggregate size and a
different bend configuration, different values of the equivalent length will be
obtained.
In this respect, one must realize that in case of pure water flow the pressure losses
in bends show a large scatter in the measurement results.
Influence of pumping on concrete properties
By means of the special testing procedure, the influence of pumping on the
concrete properties was monitored carefully. As a general trend, a decrease in
plastic viscosity, a decrease in V-funnel flow time and an increase in air content
were observed, as can be seen in Table 2. The yield stress and segregation
resistance sometimes increased and sometimes decreased. The physical causes of
these changes will be discussed in the next section.
Air void analysis on hardened samples of pumped concrete shows a decrease in
air bubbles size compared to samples of similar non-pumped concrete.
186
A temperature increase during pumping was observed during all experiments, but
it did not have a significant influence on the concrete properties or pumping
pressures in these experiments, except for some cases. The temperature increase
per unit of time (dT/dt) is linearly related to the pressure loss per unit of length
(dp/dx), but the rate of increase is different for each concrete [24]. Further
investigation is needed to clarify this phenomenon.
where
p R2
(2)
When the value of the shear stress at the boundary of the lubrication layer is lower
than the yield stress of the concrete, there is no shearing flow and the flow can be
categorized in part number 1 of Fig. 7. In the other case, shearing flow occurs. As
a result, a low yield stress of the concrete or a smaller pipe diameter enhances the
shearing flow at equal discharge. SCC shows generally shearing while OVC does
generally not. This has significant consequences on the flow behaviour of
concrete in pipes.
Flow in bends
As shown in the previous sections, the flow in bends is a very complicated flow
phenomenon, which is very difficult to understand. As far as the results are
reliable, the pressure loss in bends, relative to the pressure loss in straight
sections, caused by SCC is higher compared to OVC, which can indicate more
complex flow behaviour for SCC. Further investigation on the influence of
different parameters on the flow in bends must be executed.
187
188
explained by the increase in air content and as a result, a second theory must be
applied onto the results.
The second theory is the structural breakdown theory introduced when discussing
transient behaviour. Due to structural breakdown, the equilibrium values of both
yield stress and plastic viscosity decrease when a higher shear rate is applied. By
performing the special testing procedure, as equilibrium is awaited, the concrete is
at a new, lower equilibrium stage at each discharge step. As a result, the
rheological properties and the segregation resistance measured on the sampled
concrete, decrease with increasing maximum discharge applied.
Depending on which effect is dominant, the yield stress increases or slightly
decreases. If the SCC has initially a rather high slump flow, it is more likely that
the segregation resistance will decrease due to structural breakdown, while for
SCC with lower initial slump flows, the yield stress will increase due to the
increase in air content. Independently of which theory is dominating, the viscosity
always decreases.
During the experiments performed, the concrete has passed the pump seven or
eight times. In practice, the concrete only passes the pump once and it is expected
that these effects will be less pronounced. On the other hand, it appeared that
these effects had a more significant influence at high pumping velocities, resulting
in the advice to pump SCC at rather low velocities to reduce these effects [24].
Note that the structural breakdown theory and the increase in air content are not
applicable in case of OVC, as the concrete is not sheared in the pipes.
CONCLUSIONS
Based on experimental results for self-consolidating concrete (SCC) and a
literature review for ordinary vibrated concrete (OVC), a comparison between the
flow behaviour of these concrete types in pipes and the consequences of pumping
on the fresh properties has been investigated.
The velocity profiles of SCC and OVC during pumping are different. Both are
composed of a plug flow in the centre of the pipe and a lubrication layer near the
pipe wall, but in case of SCC, a part of the concrete is also sheared. The
properties of the lubrication layer have a large influence on the pumping pressure,
but in case of SCC, a good relationship between the pressure loss and the
viscosity of the concrete has been obtained at each discharge step applied. As a
result, in order to decrease the pumping pressure, the viscosity of SCC must be
reduced.
189
REFERENCES
[1]
[2]
190
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
de Larrard F., Ferraris C.F., Sedran T., Fresh concrete: A HerschelBulkley material, Mat. Struct. 31 (1998), 494-498.
[14]
191
Feys D., Verhoeven R., De Schutter G., Fresh self compacting concrete: a
shear thickening material, Cem. Conc. Res. 38 (2008), 920-929.
[16]
Wallevik O.H., Rheology a scientific approach to develop selfcompacting concrete, Proc. of the 3rd Int. RILEM Symp. on SCC,
Reykjavik (2003), 23-31.
[17]
[18]
[19]
[21]
Buckingham E., On plastic flow through capillary tubes. Proc. Am. Soc.
Testing Mat. 21 (1921), 11541161.
[22]
Aldred J., Pumping concrete to 601 metres in a single lift, Proc. of the
Ready-mixed concrete congress Beton2008, Istanbul (2008), 456-462.
[23]
Petit J.-Y., Wirquin E., Vanhove Y., Khayat K., Yield stress and viscosity
equations for mortars and self-consolidating concrete, Cem. Conc. Res.
37 (2007), 655-670.
[24]
Feys D., Interactions between rheological properties and pumping of selfcompacting concrete, Ph-D-thesis, Ghent University, Ghent (2009).
[25]
[26]
192
Amount (m)
Composition (kg/m)
Gravel 8/16
Gravel 3/8
Sand 0/5
CEM I 52.5 N
Limestone filler
Water
SP 2
Powder content (kg/m)
W/C-ratio (-)
W/P-ratio (-)
SP/C-ratio (%)
LM 0
1.5
6/03/2007
LM 1
1.25
LM 2
1.25
LM 3
1.5
17/04/2007
LM 4
1.25
LM 5
1.25
LM 6
1.5
5/06/2007
LM 7
1.25
LM 8
1.25
434
263
853
360
239
165
11
434
263
853
360
239
165
11
434
263
853
360
239
165
15.22
459
278
901
300
200
165
12.16
434
263
853
360
239
165
20.95
434
263
853
360
239
165
13.33
434
263
853
360
239
165
12.69
434
263
853
360
239
165
14.44
599
0.458
0.275
4.60
599
0.458
0.275
4.60
599
0.458
0.275
6.37
500
0.550
0.330
6.08
599
0.458
0.275
8.77
599
0.458
0.275
5.58
599
0.458
0.275
5.31
599
0.458
0.275
6.04
690
710
710
720
650
680
640
0.551
0.314
1.98
Amount (m)
Composition (kg/m)
Gravel 8/16
Gravel 3/8
Sand 0/5
CEM I 52.5 N
Limestone filler
Water
SP 2
Powder content (kg/m)
W/C-ratio (-)
W/P-ratio (-)
SP/C-ratio (%)
Slump flow at plant (mm)
Remarks
Plant-Mix
OVC 1
1.55
328
0.538
0.521
1.31
18/09/2007
LM 9
1.25
LM 10
1.25
LM 11
1.5
4/12/2007
LM 12
1.5
LM 13
1.5
410
248
805
400
300
165
18.15
434
263
853
360
239
165
11
410
248
805
400
300
165
unknown
434
263
853
360
239
165
unknown
434
263
853
360
239
165
unknown
434
263
853
360
239
160
21.9
700
0.413
0.236
6.05
599
0.458
0.275
4.60
700
0.413
0.236
unknown
599
0.458
0.275
unknown
599
0.458
0.275
unknown
599
0.444
0.267
9.16
581
0.452
0.324
2.47
599
0.458
0.275
unknown
581
0.452
0.324
unknown
700
650
700
675
700
640
650
700
700
target SF
target SF
Plant-Mix
target SF
Plant-Mix
target SF
193
12/02/2008
LM 14
LM 15
3.25
3.25
6/05/2008
LM 16
LM 17
3.25
3.25
434
263
853
360
239
165
unknown
Table 2 - Results of tests on fresh SCC for all mixes subjected to the special
testing procedure.
Age (hour)
Tests on fresh SCC
Slump flow (mm)
V-Funnel (s)
Sieve Stability (%)
Air content (%)
Age (hour)
Tests on fresh SCC
Slump flow (mm)
V-Funnel (s)
Sieve Stability (%)
Air content (%)
Age (hour)
Tests on fresh SCC
Slump flow (mm)
V-Funnel (s)
Sieve Stability (%)
Air content (%)
Age (hour)
Tests on fresh SCC
Slump flow (mm)
V-Funnel (s)
Sieve Stability (%)
Air content (%)
Age (hour)
Tests on fresh SCC
Slump flow (mm)
V-Funnel (s)
Sieve Stability (%)
Air content (%)
Age (hour)
Tests on fresh SCC
Slump flow (mm)
V-Funnel (s)
Sieve Stability (%)
Air content (%)
Q1
1:03
Q2
1:12
SCC LM 12
Q3
1:21
Q4
1:33
Q5
1:42
735
3.16
12.3
2.4
725
3.07
8.1
2.7
755
2.34
12.3
2.8
735
2.23
12.9
2.8
645
2.09
9.4
3.2
Q1
2:30
SCC LM 14
Q2
Q3
2:45
3:00
Q4
3:30
818
5.23
10.9
1.6
758
6.1
11.0
1.8
658
5.82
7.5
1.5
Q1-1
1:30
745
3.65
14.2
1.6
SCC LM 15 - test 1
Q1-2
Q1-3
1:45
2:00
660
3.77
6.6
3.2
Q1-4
2:10
645
5.43
4.2
2.1
625
4.18
7.0
2.4
570
3.42
4.0
4.2
Q2-1
2:50
SCC LM 15 - test 2
Q2-2
Q2-3
Q2-4
3:00
3:10
3:20
Q2-5
3:30
525
3.54
3.4
3.7
543
3.06
4.5
3.9
505
3.29
1.9
4.6
445
3.74
0.3
6.2
Q1
2:35
Q2
2:45
SCC LM 16
Q3
Q 3 bis
3:00
3:10
670
5.24
8.7
1.1
675
4.02
12.7
1
655
4.78
6.9
1.4
Q1
1:20
Q2
1:30
785
3.39
10.5
1.4
780
3.08
1.9
498
3.46
0.8
5.0
Q 3 ter
3:20
Q4
3:30
585
3.72
6.8
1.3
620
3.76
7.8
2.2
535
3.89
5.7
3.9
SCC LM 17
Q3
1:40
Q4
1:50
Q5
2:00
750
2.66
11.7
3.1
765
2.35
15.6
3.9
750
2.22
18.5
4.9
194
Shear stress
Shear rate
shear rate
shear stress
Equilibrium shear stress
Time
195
Fig. 3 - Short circuit with a length of 25m, the measurement section is located
in the straight horizontal part of the circuit, on the left side of the picture
196
Strain gauges
Temperature sensor
Pre
ssu
re
sen
so
Sampling
Time
197
40
35
30
25
20
15
10
5
0
0
10
15
20
25
Discharge (l/s)
Fig. 8 - Pressure loss discharge curve for SCC and OVC, reflecting the
rheological properties of the concretes. The OVC is a perfectly Bingham
material, while the SCC shows shear thickening behaviour
198
90
80
y = 1.6093x + 6.9824
2
R = 0.9271
70
60
y = 0.9427x + 1.2378
50
R2 = 0.8899
40
30
y = 0.3234x + 1.4083
20
R2 = 0.6898
10
0
0
10
20
30
40
50
60
70
Fig. 9 - The relationship between pressure loss and the differential viscosity
at a shear rate of 10/s, shows, for each discharge, good to very good
correlations
199
200
Synopsis: Self compacting concrete (SCC) is a material with a short history and a
large impact. After its invention considerable research has been carried out and
practical applications have been realised. An important question has always been
whether SCC should be considered as a regular concrete or if its needs a separate
set of recommendations. In this paper it is argued why the properties in the
hardened state can be taken from regular codes for structural concrete. Contrary to
this the behaviour in the fresh state cannot be taken from existing
recommendations, since other measuring methods than used for conventional
concrete are required. This refers especially to the concrete consistence. In this
respect the latest rules according to the new Eurocode are discussed. For practical
applications it is important to know the relation between the type of concrete
structure or member considered and the SCC-properties required. As an example
the properties are classified by the slump flow diameter and the funnel time,
resulting in 9 subclasses for the consistence, linked to typical structural
prototypes. In future defined performance design will be an important approach
both for structures and structural materials. Within this scope some special SCCmixtures are treated. Especially self compacting fibre concrete turns out to be an
attractive option. Suggestions for further research and development are given.
Keywords:
Self
Compacting
Concrete,
201
properties,
consistence,
codes
INTRODUCTION
Self compacting (or self consolidating) concrete (SCC) was first developed in
Japan, in the early nineties of the previous century, under the stimulating
leadership of Prof. Okamura. The main idea behind self compacting concrete was,
that such a concrete is robust and relatively insensitive to bad workmanship. In
Western Europe the idea was picked up at the end of the last century. The main
drive to develop self compacting concrete was the option to improve the labour
conditions at the building site and in the factory (noise, dust, vibrations). During
recent years self-compacting concrete developed to an important research item. A
large number of research projects were carried out, followed by recommendations
for potential users. Especially for the precast concrete industry self compacting
concrete was a revolutionary step forward. Contrary to that, casting of SCC at the
construction site was regarded with some reservation. The variable conditions at
the construction site, the more complicated control of the mixture composition
and disagreement with regard to the question how the properties should be
measured at the site were retarding factors. In spite of a number of successful
examples, some problems due to unsuitable use of SCC in initial applications
generated further scepticism. Since then substantial effort has been done in
developing more stable and robust mixtures. Moreover a better insight has been
gained with regard to SCC composition and consistence criteria in relation to
types of application. For a better understanding of relations between SCCproperties and SCC-composition, at first a short treatment is given of the question
why basically a concrete can be self-compacting, and what influence can logically
be expected on the properties in the fresh and the hardened state.
PROPERTIES OF SELF COMPACTING CONCRETE
The Japanese way of composing the optimum mixture composition of SCC
consists of a number of steps. At first, in a simple test, the optimum ratio of water
to powder is determined. Then a number of general criteria have to be met, the
most important of which are that the coarse aggregate volume should be 50% of
202
the solid volume of the concrete without air, and that the fine aggregate volume
should be 40% of the mortar volume, where particles finer than 0.09 mm are not
considered as aggregate, but as powder. If the composition of the mixture,
obtained in this way, is mathematically analyzed, it is found that this procedure
leads to a concrete composition with a little bit of excess paste. That means that
there is slightly more paste in the mixture than necessary to fill all the holes
between the particles: this implies that around any particle a very thin
lubricating layer exists, by virtue of which the friction between the particles in
the fluid mixture is greatly reduced in comparison to conventional mixtures, Fig.
1. The optimum thickness of those layers lies between narrow limits. If the
thickness is too small, there is too much friction to achieve self compactibility. If
the thickness is too large, the coarse aggregate particles sink down and
segregation occurs. The rheologic properties of the excess paste layers are
determined by the choice of the superplasticizer. Furthermore, in the fresh state
around the cement and powder particles thin layers of water are formed [1]. In
this way a three phase system (coarse particles, fine particles and powder) with
intermediate layers of paste and water is obtained which minimize the internal
friction in the fresh state.
Midorikawa [2] carried out tests in order to find the optimum thickness of the
excess paste layer. Fig. 2 shows the optimum thickness of the layer for a varying
ratio Vw/Vp (volume of water to volume of powder) for different grading curves.
It is seen, that the thickness of the paste layer, for which the concrete is still self
compacting, increases with decreasing volume of water. Below V w/Vp = 0.8 the
appropriate thickness increases overproportionally. For practical application,
however, this area is not relevant. The optimum ratio in this case is in the range
Vw/Vp = 0.8 0.9. The mean thickness of the excess paste layer is then in the
order of magnitude of 0.05mm.
An important question is, to which extent the lubricant layer influences the
properties of the concrete in the hardened stage. It seems to be obvious, that for
instance the modulus of elasticity of SCC is smaller than that of a conventional
concrete of the same strength, as a result of the effect of the relatively soft
lubricant layers. An evaluation by Holschemacher [3] showed, that the E-modulus
of an SCC is indeed somewhat smaller, Fig. 3. It should, however, not be
forgotten, that also the E-moduli of normal concretes are subjected to scatter,
most of all in dependence of the stiffness of the aggregate used. For practical
applications it can therefore be assumed, that the E-modulus of self-compacting
concrete is not outside the region of scatter of conventional concretes. Of course it
makes sense to carry out suitability tests in the case of special applications, such
as for structures in high speed railway lines. In that case, however, also the creepand shrinkage properties should carefully be investigated.
203
Another aspect important for the behavior in the hardened state is the concrete
tensile strength. When the axial tensile strength of a SCC would substantially
differ from that of the tensile strength of conventional concrete, this should have
large implications for design, since the tensile strength is a governing aspect in the
design for shear, punching, anchorage, crack width control and the minimum
reinforcement. It is obvious to expect that the tensile strength of SCC is higher
than for a conventional concrete, because of the more homogeneous interface
between the aggregate particles and the cement past (no direct contact between
the aggregate particles). An evaluation of test results [3] confirms this, see Fig. 4.
However, also here the results are in the range of scatter of conventional
concretes, so that no complicating exception for SCC has to be made.
Another important aspect is the formwork pressure of self compacting concrete.
Many measurements have been carried out, but the results were often conflicting.
Often the role of the rising speed of the concrete in the formwork was
disregarded. Fig. 5 shows the results of a number of Swedish [4] and Dutch [5]
tests, collected in one diagram. It is visible that the rising speed of the concrete in
the formwork influences the formwork pressure. For the concretes tested, from a
rising speed of about 2m/hour the distribution of the pressure corresponds
approximately to the hydrostatic pressure. This, however, does not implicate, that
for lower rising speeds a reduction of the formwork pressure is a reliable
assumption. According to its rheologic behavior SCC is a Bingham fluid. Such a
fluid is characterized by two parameters: the yield value and the plastic viscosity.
The yield value is a measure for the force, necessary to get the concrete moving.
The plastic viscosity is a measure for the flow rate (toughness) of the mixture.
When the yield value is high and the plastic viscosity is low, it may happen that
the formwork pressure is initially very low, but suddenly increases due to a shock
against the formwork. It is therefore advisable to work always with the
hydrostatic formwork pressure.
PROPERTIES OF SELF COMPACTING CONCRETE AND CODE
RECOMMENDATIONS
Behaviour of SCC in the hardened state
The most recent large scale international building code is the Eurocode. The
Eurocode consist of many parts. The basic code for the design of concrete
structures is EN 1992-1-1 Design of concrete structures: general rules and rules
for buildings. Complementary, the properties and specifications of the material
204
205
called Bingham Fluid, characterized by two parameters: the yield value and the
plastic viscosity.
Therefore EN206-1 offers an appendix for SCC, giving additional methods to
classify SCC. These are:
EN 12350-8: Slump-flow test for self-compacting concrete (flow diameter test)
EN 12350-9: V-funnel test for self-compacting concrete
EN 12350-10: L-box test for self compacting concrete
EN 12350-11: Sieve segregation test for self compacting concrete
EN 12350-12, J-ring test for self-compacting concrete
Where the flowability of self-compacting concrete is to be determined, it shall
be measured as a slump-flow test in accordance with EN 12350-8
Where the viscosity of self-compacting concrete is to be determined, it shall be
measured by means of:
- t500 test conforming to EN 12350-8
- V-funnels test conforming to EN 12350-9
Where the passing ability of self-compacting concrete is to be determined, it
shall be measured either by means of:
- L-box test conforming to EN 12350-10
- J-Ring test conforming to EN 12350-12
Where the resistance to segregation of self-compacting concrete is to be
determined, it shall be measured by means of:
- Sieve segregation test according to EN 12350-11
For the various tests classes have been defined, given in Tables 2-7.
Some further requirements have been given with regard to the time and procedure
of testing, like:
- where the consistence of self-compacting concrete is to be determined, it shall
be tested at the time of use of the concrete or in the case of ready mix concrete,
at the time of delivery.
- If concrete is delivered in a truck mixer or agitating equipment, the consistence
may be measured using a spot sample obtained from the initial discharge.
206
207
For successful production of SCC it is essential that the basic constituents, like
sand, gravel, fillers and the third generation of superplasticizers, have a constant
quality. This is not always the case. Moreover, not all cement producers supply a
constant quality. So, there should be good agreements between the concrete
producers and the suppliers of the constituents on the quality control. The step
from a traditional concrete production to the production of SCC is not a big one.
Installations with an age of say 5-10 years are generally suitable. Further to the
traditional equipment a high intensity mixing machine and an installation to dose
the fillers are needed.
As a result of the introduction of SCC the formwork is hardly loaded anymore: it
has only a retaining function. So, the wall can be made of other materials than
timber, like polystyrene. Also steel formwork with magnetic couplers is possible.
The time for demoulding and re-installing the formwork has been reduced by
50%. There is no need for the installation of vibration isolators anymore. Rubber
joint sealings can be omitted, since by virtue of SCC no leakage through the joints
occurs anymore.
Fig. 10 shows an example of an architectural balcony element of SCC. The
element does not only show a beautiful shape with very sharp profiles, it has also
a homogeneous white colour. Fig. 11 shows the assembly of a precast prestressed
concrete girder of SCC for the new metro station at the Amsterdam Arena, the
stadium of soccer club Ajax. The girder has a length of 22,5m. The concrete
strength class is C55 (characteristic cylinder compressive strength of 55 MPa
(7850 psi)). The metro station has a length of 350m with 4 tracks. This means that
60 girders had to be produced with a total length of 1.4 km. If the girder would
have been compacted in the traditional way, heavy vibrating machines would
have been necessary. Due to that, the formwork would have had to be replaced
after a relatively small number of casts. By virtue of the use of SCC the life of the
formwork was very long. Another important reason to choose for SCC was the
improvement of the labour conditions in the factory.
Fig. 12 shows a set of foundation piles. The production of such type of piles in the
firm was 70 000 piles a year. For an average length of 15m a total production
length of 1000 km a year is obtained. Until recently the piles were produced with
the so-called shock procedure. This means that the formwork was forced to
repeatedly fall down from a height of 50mm (2 inches), which created a shock
208
effect. By virtue of the change to SCC the necessary casting time was reduced
from 7.5 minutes to 1.5 minutes. Since mechanical compaction was not necessary
anymore 12 further minutes were gained. Taking also into account the advantages
with regard to the reduction of noise and dust, energy consumption and wear, it is
clear that SCC yields considerable advantages. Fig. 13 shows a number of
concrete arches. Every arch has a length of 65 meter and is composed of 5 pieces
of 13m. The cross section has a box-shape, with a foam core. Producing such an
element with conventional concrete does not make sense, since the foam core
would move due to vibration. A production in parts could be an alternative but is
by far too time consuming, and therefore too costly. With SCC perfect elements
could be made. Meanwhile many precast concrete firms have changed their
production to SCC, some even for 100%.
APPLICATIONS OF SCC IN SITU
The introduction of SCC for in-situ applications was slower than in the precast
concrete industry. There are a number of reasons for this:
- In case of failure the consequences for an in-situ application are much more
severe than in the precast concrete industry. In the latter case the unsuitable
elements can be simply rejected, whereas in the first case demolition might
be the ultimate consequence.
- There was often no agreement on the way in which the properties at the
building site have to be controlled.
- Self compacting properties can be more easily reached with higher strength
than with lower concrete strength. In a number of practical applications the
concrete strength was therefore higher than actually necessary, which has
cost consequences. For many applications a concrete strength class C25 is
sufficient. However, especially for the lower strengths classes it is more
difficult to obtain robust and reliable self compacting concretes.
Meanwhile, however, a lot of barriers have, or are being, removed. There is now a
better insight into the required properties of SCC for particular applications, like
previously shown in Fig. 9. Furthermore qualifying test methods have been
evaluated. Finally a new generation of superplasticizers has been introduced.
Nevertheless, a number of convincing examples exist, which proof that SCC, if
applied in an appropriate way, can give excellent results. The first application of
SCC in The Netherlands according to modern principles was such an example,
Fig. 14. In 1998 a large faade was made for the National Theatre in The Hague,
which, for architectural reasons, was provided with fine triangular ribs with a side
length of 8mm. In this case an SCC with relatively high flowability was used
209
(flow diameter 730 mm) and a low viscosity (funnel time 8-9 seconds). Fig. 15
shows the sports stadium at Tor Vergata University (Roma II), under
construction. The architect is Santiago Calatrava, who required self compacting
concrete for all the concrete used. The contractor succeeded in realizing excellent
surface properties.
There are many practical problems where SCC gives a suitable solution. An
example is the retrofitting of the Ketelbridge, a glued segmental bridge in The
Netherlands. At the time of retrofitting in the year 2002 the bridge was 45 years
old. During the years the bridge deck was renovated several times, but the old
deck was often not totally removed. So, finally the bridge deck was 180 mm thick
in stead of 50 mm. Since as well the traffic load had increased, the joints between
the segments opened. Therefore it was decided to increase the load bearing
capacity by external prestressing. A difficulty was the provision of the deviators
inside the box girder. Because the lower flange of the girder had not been
designed for the transport of heavy materials, casting concrete inside the girder
was no realistic option, even regardless of the technological difficulties involved.
As a solution therefore SCC was used. The formwork with the reinforcement was
built up in the interior of the girder (Fig. 16), and the SCC was cast from the
outside through a little window in the upper flange. The concrete strength class
was C35. By a suitable use of the rheological properties an excellent result was
obtained.
Another interesting case for which SCC gave a solution was the provision of the
end walls in elements for a submerged tunnel, Fig. 17. Those end walls had a
temporary character and served only for enabling floating transport and
submerging. After the elements had been coupled under water, the walls were
demolished. In order to facilitate easy demolishing, SCC in a strength class C20
was used. For casting the concrete between the tunnel walls through small
windows in the formwork SCC appeared to the most appropriate solution.
SPECIAL SELF COMPACTING CONCRETES
A remarkable development occurred with regard to the workability of fibre
reinforced concrete. For a very long period it was noted that the addition of fibres
to concrete decreased the workability. However, in his PhD-thesis Grnewald [7]
showed that this is not necessary at all. He proved that self compacting fibre
concretes are very well possible, even up to fibre contents of 140 kg/m3, if the
right combination of fibres and mixture composition is chosen. Fig. 18 shows the
maximum possible fibre content for which mixtures are still self-compacting
(defined as having a flow circle with a diameter of at least 600 mm, a round shape
and a homogeneous fibre distribution). At the vertical axis the fibre content in
210
kg/m3 is given. At the horizontal axes the fibre type (aspect ratio/length) and the
mixture type (with the sand/gravel vol. ratio) are given.
Fig. 19 gives an impression of the excellent flowing properties during casting of a
concrete with 125 kg/m3 fibres. Fig. 19 shows the measurement of the flowability
of an ultra high performance fibre reinforced concrete in a U-shaped formwork.
The concrete had a mean cube strength of about 180 MPa (25000 psi). It
contained 235 kg/m3 steel fibers 20/0.3mm. It was used in a factory to produce
prestressed beams for a bridge. In another paper the topic of optimizing the selfcompacting properties of fibre concrete was treated more in detail [8].
The self-compacting properties of fibre reinforced self compacting concrete can
be of great use for practical applications. An example is the repair of the Huisne
Bridge in France [13], Fig. 21. This bridge should be strengthened in order to be
able to carry a larger traffic load. A solution was found by widening the width of
the webs of the prestressed I-beams with high performance fibre reinforced
concrete. As a result of the large E-modulus of the HPFRC (about 60.000
N/mm2), the normal force, exerted by additional external prestressing, is
predominantly sustained by the high performance concrete strips, in stead of
loading the old concrete, the capacity of which was nearly fully utilized according
to the original design. Casting concrete against the webs of I-beams under the
existing bridge decks is hardly possible with conventional concrete. With the high
performance fibre concrete used, this was no problem, by virtue of the selfcompacting properties. A little window in the deck was sufficient for remote
casting of the concrete under the deck.
Recently it was shown that not only concrete with medium or high strength
concrete can be self-compacting. At TU Delft a research program was carried out
with the aim to fill large underground corridors, used for the storage of nuclear
waste, with self compacting concrete by remote casting. For this application no
strength requirements applied for the concrete. Therefore an economic solution
was strived at: developing a low strength self compacting concrete. The grout,
which was composed with 10% cement, 30% limestone powder, 0,03%
superplasticizer en 59,95% sand, with a water cement ratio of 1.30-1.35, showed a
28 days strength of 19-20 N/mm2. Fig. 22 shows a laboratory test where the
mixture flows through a pipe investigating the suitability for remote casting [14].
Another interesting option is self-compacting lightweight concrete. High
performance lightweight concrete could allow significant savings in reinforcingand prestressing steel and foundations. Self compacting properties would even
increase the attractiveness of such a material. With regard to the production
technology, there is a major difficulty. The lightweight aggregate particles are
porous and therefore influence the mixture composition by sucking water from
the mixture in the fresh state. As self compacting concrete is sensitive to the right
211
212
SCC with regard to chloride ingress, carbonation and frost-thaw cycles. It was
shown by many research projects that SCC is approximately equivalent to
conventional concretes with regard to the majority of its mechanical properties in
the hardened state. However, with regard to the microstructure of hardened SCC
and its significance to durability there are still quite a number of open questions.
In this respect the interface between matrix and aggregates plays an important
role. Furthermore the role of (combinations of) additives (superplasticizers, air
entraining agents, viscosity agents) on the microstructure, including porosity and
permeability should get due attention. In this respect special attention has to be
devoted to the average and low strength concretes used in in-situ structures, if
exposed to more severe environmental conditions.
CONCLUSIONS
1. In spite of its short history, self compacting (or consolidating) concrete has
confirmed itself as a revolutionary step forward in concrete technology.
2. For the application of SCC in situ, it is necessary that SCCs are designed
(tailor-made) for any particular case. General rules are available on the basis of
experience.
3. It can be shown by cost analysis, that SCC in precast concrete plants can be
more economically produced than conventional concretes, in spite of the slightly
higher material price. Cost comparisons should always be made on the basis of
integral costs.
4. There is a considerable future for tailor-made self compacting fiber reinforced
concretes
5. The most important task for research is to develop SCCs with decreased
sensitivity to variations in constituents and environmental influences. This holds
particularly true for in-situ concretes, with medium and low strengths.
6. Further research into the potential role of viscosity agents and their interaction
with superplasticizers is worthwhile
7. Since in the near future service life design (SLD) of concrete structures will be
as important as design for safety and serviceability, increased attention should be
given to the role of the microstructure of the various types of available SCCs and
its role for durability.
8. With regard to its mechanical properties it is not necessary to make a
distinction between SCC and conventional concretes. Therefore also in modern
codes like the Eurocode for Concrete Structures EN 1992-1-1, SCC is treated like
any other regular concrete.
213
REFERENCES
1.
2.
Midorikawa, T., Pelova, G.I., Walraven, J.C., Application of the water layer
to self compacting concrete with different size distribution of fine aggregate,
Proceedings of the Second International Symposium on Self-Compacting
Concrete, Tokyo, Japan, 23-25 October 2001, pp. 237-246.
3.
4.
5.
6.
www.concretecentre.com.
7.
8.
9.
10. Haist, M., Mechtcherine, V., Beitzel, H., Mller, H.S., Retrofitting of
building structures using pumpable self-compacting lightweight concrete,
Proceedings of the 3rd International RILEM Symposium on SelfCompacting Concrete, pp. 776-795.
11. Grnewald, S., Walraven, J.C., The effect of viscosity agents on the
characteristics of self-compacting concrete, Conference SCC 2005, Chicago,
USA, Oct. 30th Nov. 2nd, 2005.
214
12. Takada, K., Influence of admixtures and mixing efficiency on the properties
of self compacting concrete, PhD-Thesis, TU Delft, May 11th, 2004.
13. Thibault, Th. Strengthening of Huisne Bridge using Ultra-High-Performance
Fibre-Reinforced Concrete, Proceedings fib Congress 2008, Amsterdam, pp.
331-334.
14. Bakker, P., Grout backfill process of nuclear waste disposal gallery, Masters
Thesis, TU Delft, Faculty CiTG, Group of Concrete Structures, June 9th, 2009
215
Compaction
Class
C0
1.46
C1
1.45-1.26
C2
1.25-1.11
C3
1.10-1.04
Slump
Class
mm
Flow diameter
Class
mm
S1
S2
S3
S4
S5
F1
F2
F3
F4
F5
F6
10-40
50-90
100-150
160-210
220
Flow diameter in mm
(Limit values for individual batches)
550 650
660 750
760 - 850
Table 3. Viscosity classes t500
Class
VS1
VS2
VS3
T500 (seconds)
2
36
>6
Table 4. Viscosity classes V-funnel
Class
VF1
VF2
216
340
350-410
420-480
490-550
560-620
630
Class
PJ1
PJ2
217
Inspection/test
Purpose
Minimum frequency
Consistence
Visual inspection
For comparison
with normal appearance
To assess the
achievement of the
specified values of
consistence and to check
e.g. possible changes of
water content
V-funnel test
L-box test
Sieve segregation
test
J-ring test
To assess the
achievement of the
specified values of
consistence
- If consistence is specified
as EN 206-1:2000, Table
13 for compressive
strength
- In case of self compacting
concrete at least once day
- when testing air content
- In case of doubt
following visual
inspections
- When performing initial
test
- Before using a new
concrete composition
- In case of a change in the
constituents. In case of
doubt following visual
inspections or slump-flow
test.
218
Vwater/Vpowder
219
220
Height
[m]
hydrostati
c
1,
0
0,
8
1,3
0
0
1
0
2
0
1,
6
1,
4
3
0
rising speed in
m/h
10
2
4
0
5
0
6
0
7
0
Pressure [kPa]
Slump
Flow diameter
Degree of compaction
221
270
3
0
70
60
flow
cone
r0=100
paste
r1
240
r2
120
60
30
5-9
3-5
Slump flow
1
2
3
Consistency classes
470-570
4
>
540-660
630-800 [mm]
222
9-25
High &
Ramps
Slender
Walls
5-9
Floors
3-5
Consistency class
470-570
540-660
630-800
223
224
Fig. 16. Remote casting of a wall with openings in the interior of a box girder
bridge for creating deviation points for additional external prestressing
tendons, aiming at increasing the bearing capacity (Ketelbridge in The
Netherlands, 2002).
225
Fig. 17 Casting the end wall of an element for a submerged tunnel in SCC.
140
120
100
Fibre
content
80
[kg/m3]
60
Reference
mixture
40
20
Mix 4 (68/39.0)
Mix 3 (68/36.5)
Mix 2 (57/39.0)
Mix 1 (57/36.5)
0
45/30
Fibre type
65/40
80/30
80/60
226
Fig. 19 Self-compacting
concrete, with 125 kg/m3
fibres, in strength class C115
227
Fig. 22. Remote casting test for low strength SCC for nuclear waste disposal
[14].
Fig. 23. Lightweight aggregate particle with skin of with cement paste [9].
228
INTRODUCTION
Self Compacting Concrete (SCC) has gained wider acceptance since its
development in the 80s in Japan. Today, especially in the precast industry, it is
an option for the production managers who search for production costs savings
and improvements of the quality of the concrete structures at the same time.
SCC utilization in the precast industry is nowadays a consolidated reality also in
Italy, even if there is still room for growth to get to the same share of utilization of
other European countries (1)
Moreover the utilization of Self Compacting Concrete represents a technological
innovation that involves both on the material and casting procedure sides. As a
matter of fact to implement SCC in an industrial process requires a careful study
of the mix formulation to achieve the desired technical requirements, but also a
thorough evaluation of the mixing and pouring processes that are usually designed
and meant for lower (usually S5) consistency class concrete mixes. (2)
This work has also the scope of underlining how, on the one hand SCC
production is easily feasible in a modern precast yard even if not designed for
SCC, on the other to draw attention to some necessary precautions to adopt in
order to have satisfactory results (3).
Technical requirements for the structural elements
Different types of beams were considered for the present case study: TI2 shaped,
TI shaped and rectangular section shaped. The first two types being produced
with prestressed reinforced concrete.
The above mentioned elements were originally designed for 55 N/mm2
characteristic strength and to be cast with fluid concrete with consistency class of
S5, determined by the slump value of the Abrams cone method. The maximum
diameter of the aggregate used was set to 25mm.
The request was to improve the appearance of the elements that were originally
characterized by excess of blow holes on the surface, and to speed up the
production operations at the same time.
229
It is known that the utilization of SCC sharply improves the surface appearance of
the elements bringing in addition a better ability to resist to the aggressive agents
of the environment, and thus to an improved durability of the service life of the
whole structure.
The choice was then to design again the concrete mix to switch from S5
consistency class to SCC. In addition to that an experimental program to improve
early strength by selecting among different admixtures for precast concrete was
decided, in order to gain productivity at the same time.
The study of the mix design was carried out according to the technical
requirements without changing the raw materials already used for normal
production with the only exception of adding a limestone filler and, of course,
selecting the most appropriate admixture combination.
MATERIALS
Limestone Filler: Commercially available limestone filler from Rezzato, Italy was
used. Its particle size distribution is reported in Fig. 1.
Superplasticizers: Two PCE High Range Water Reducing admixtures specifically
designed by MAPEI for precast concrete production (Type HRWR + HAA pr.
3+7 according to EN 934/2) were selected. Their characteristics are listed in Table
1.
Aggregates: Particle size distribution of aggregates is reported in Table 2.
Cement : A commercially available Cem I 52,5R from Italcementi was used.
TEST METHODS
Concrete Tests: The mixes were designed to have Slump Flow value of about 750
mm measured according to prEN 12350-8 (4). V-funnel tests were done according
to prEN 12350-9 (5) and L-Box test according to prEN 12350-10 (6) to better
define the mix design at laboratory scale. The proportions of the selected mixes
are reported in Table 3.
SCC mixes were prepared in a forced action mixer with interchangeable mixing
plastic tubs (Collomix mod. XM3-900). All materials (aggregates, cement and
water) were dried at ambient temperature overnight prior to use.
At industrial scale several mixes were produced in a 1cubic meter forced action
mixer. Different mix designs were tested in the above mentioned elements
evaluating both rheological and aesthetical aspects.
230
231
CONCLUSIONS
The study, was conducted according to the new European Normative on Self
Compacting Concrete (7), has lead to the conclusion that the technical
requirements of the final compressive strength, the filling of the moulds and the
surface finishing of the elements were met by adopting the self compacting
concrete technology. In addition, when using a recently developed admixture
designed for the precast concrete production, it is possible to improve the
compressive strength results at early ages allowing thus a faster mould release
with a positive effect on productivity.
The most recent precast yards are designed to have little or even no transportation
of the fresh SCC and are equipped with moulds on rails that are able to be filled
directly under the concrete mixer. The precast yard involved in the study was
instead not originally designed to produce SCC, and is equipped with a
production line for traditional vibrated concrete. Therefore SCC production and
casting led to a substantial change in the production methods.
Even if the 2 m3 bucket was not possible to eliminate, the casting procedures were
conducted with a modification able to let the SCC flow in the mould without
falling. By means of this precaution it was possible to take the most of the
rheological properties of this highly flowable material.
Picture 2 shows the modified bucket with a flexible hose allowing SCC to be cast
from the bottom of the element without free falling. Moreover, in order to further
allow entrapped air to easily flow away, the pouring was done from one side of
the element without moving the bucket, allowing the SCC to flow from the one
end to the other.
This result was only possible by an accurate selection and proportioning of all the
ingredients (aggregate skeleton, cementitious systems and admixture) in order to
have an highly flowable material with no segregation and optimum strength
development combined with a critical evaluation of the mixing and pouring
processes. The precast concrete producers presently adopting traditional vibrated
concrete can consider to improve the quality of their elements and the overall
productivity by moving to Self Compacting Concrete production after a careful
study of the mix and of the production conditions.
1.
2.
REFERENCES
J. Walraven, Self Conpacting Concrete: Challenge For Designers And
Researchers, SCC Symposium, 2005 Chicago.
K. Juvas, European Experiences with Self Compacting Concrete in precast
Concrete Industry, SCC Symposium, 2008 Chicago.
232
3.
Cotic, Koglot, Zarnic, Bokan Bosiljkov, Experiences For Scc Mixes For
Precast Elements And Field Applications In Slovenia, SCC symposium,
2008 Chicago.
4.
5.
6.
7.
233
Admixture
Dynamon SP1
Solid Density
(%)
(g/cm3)
30
1.09
30
1.07
26
1.06
Passing %
0,1
1,0
10,0
Sieve opening in mm
S1
G1
G2
234
G3
Risultante aggregati
100,0
0.50
0.50
w/p
0.36
0.33
Cement kg/m3
400
400
Limestone Filler
kg/m3
150
200
798
877
197
Gravel 6-10 mm
kg/m3
461
526
235
Dos. l/m
l/m3
Lbox (%)
V funnel (sec)
T50 (sec)
Air (%)
SCC
1
Dynamon
SP1
0.90
5.5
201
0.36
2373
26 C
100
2.2
0.5
780
SCC
2
Dynamon
NRG 1020
0.56
3.3
198
0.36
2347
26 C
<80
22
3.5
2.4
740
SCC
3
Dynamon
NRG 1012
0.75
4.5
201
0.36
2377
26 C
95
12
2.8
1.0
770
Admixture
Dos. % vol/pwd
Water
a/p
S.G.
kg/m
3
Water
Dos. % Dos.
vol/pwd l/m
l/m3
w/p
16h
18h
20h
24h
7gg
28gg
SCC1
0.90
5.5
77.4
SCC2
0.56
3.3
75.8
SCC3
0.75
4.5
76.4
236
Water
Cs (N/mm2) 20C
in water
Dos.
%
vol/
pwd
Dos
.
l/m
l/m3
w/p
Slump
Flow
(mm)
T50
(sec)
16h
20h
7d
28d
Dynamon
SP1
1.00
6.0
203
0.34
730
3.0
23.2
30.5
54.3
62.6
Dynamon
NRG1012
0.85
5.0
201
0.34
780
4.2
29.6
53.9
63.2
Admixture
237
238
239
240
241
242
243
244
volume of 370 l/m3, w/cm of 0.42, and polycaroxylate-based high-range waterreducing agent (HRWRA) and viscosity-modifying agent (VMA) dosages of 3.6
and 2.8 l/m3, respectively. The casting rate was set at 10 m/hr. The column was then
sealed, and air pressure was applied in steps equivalent to one-meter of concrete
head at the same casting rate of 10 m/hr until a concrete head corresponding to 13
m was reached. The corresponding lateral pressure exerted on the column wall was
recorded using the sensor at the base of the UofS2 pressure column, as shown in
Fig. 3. Between each consecutive increment, a decrease in lateral pressure profile
was obtained reflecting the restructuring (or structural build-up at rest) of the
concrete. This was more apparent at the end of casting where pressure decay is
observed. The pressure device is usually demolded before concrete hardening. The
maximum lateral pressure (Pmax) recorded at each increase of simulated concrete
height is plotted in Fig. 4 indicating that the SCC developed 40% lower relative
lateral pressure (K0) than the equivalent hydrostatic pressure at concrete height of
13 m. K0 refers to the ratio between Pmax and equivalent hydrostatic pressure (Phyd)
and is plotted against concrete height in Fig. 5.
In order to determine the repeatability of the UofS2 pressure column, an SCC
mixture was prepared and tested four times. The results of the relative errors of K0
at various heights are given in Table 1 and indicate high precision of pressure
measurements.
The lateral pressure characteristics of three different SCC mixtures determined
using the UofS2 device were also compared to measurements obtained from a
PVC column measuring 0.2 m in diameter and 3 m in height. Good agreement
was obtained between both systems in terms of initial lateral pressure and
pressure drop in time [2]. Further validation of the UofS2 pressure device clearly
demonstrated that the device is capable of adequately reflecting the effect of
placement rate, initial slump flow, and mix design (including paste volume, w/cm,
maximum size aggregate (MSA), and VMA dosage) on lateral pressure
characteristics [2].
EMPIRICAL TESTS TO EVALUATE STRUCTURAL
BUILD-UP AT REST
A total of six field-oriented test methods were developed and used to evaluate the
structural build-up at rest of concrete-equivalent mortar (CEM) and SCC mixtures.
The portable vane, inclined plane, and undisturbed spread test methods showed good
repeatability and low relative error [2]. The response obtained with the portable vane
and inclined plane were compared to rheological measurements obtained using a
modified Tattersall-Type MK-III concrete rheometer [3] using 22 SCC mixtures
prepared with various mix designs and material constituents. The comparison clearly
245
indicates that the empirical methods can capture the structural build-up of SCC at rest
in terms of static yield stress and its rate of gain in time [2].
Portable vane test
The portable vane test is inspired from a field test for in-situ measurement of
shear strength of soil (in particular clay soils). Four-blade vanes of different sizes
(Table 2) were manufactured from stainless steel to enable the use of one torquemeter to capture shear strength of the plastic concrete after various times of rest
(Figs. 6 and 7). The largest vane is used for the weakest structure, i.e., shortest
resting time, and vice versa. A torque-meter measuring with high precision was
employed to capture the torque values.
Immediately after mixing, the four vanes are centered vertically in the containers.
The containers are filled with SCC to a given height (h) indicated in Table 2. The
rested materials are covered. After 15, 30, 45, and 60 minutes of rest, the torquemeter is attached to the axis of the vane and turned slowly (10 to 15 s for a quarter
turn). The maximum torque needed to break down the structure is then noted. The
torque values are converted to static shear stress (0rest) using Eqs. 1 and 2 [3] as
follows:
0rest
T
G
G 2 r2 h
(1)
1
r
3
(2)
where: T = measured torque (N.m), and the other terms are defined in Fig. 7.
Variations of static yield stress (0rest) determined using the portable vane test
(PV0rest) with resting time for typical SCC mixtures designed with different
thixotropic characteristics are shown in Fig. 8. The 0rest obtained from the
portable vane at 15 min (PV0rest@15min) corresponding to the initial response (Ri)
is used as a structural build-up index. Similarly, the rate of change of static yield
stress with time [PV0rest(t)] or [R(t)] and the coupled effect of RiR(t) can also be
for structural build-up indices.
Inclined plane test
The inclined plane method [4] involves casting concrete in a cylindrical mould
onto a horizontal plate of a given roughness, then lifting the plate to initiate flow
of the material, as shown in Figs. 9 and 10. The corresponding angle the
necessary to initiate flow is used to determine the static yield stress, 0rest (Pa), as
follows:
0rest = .g.h.sin
246
(3)
where is the density of the sample (in g/cm3), g is the gravitation constant (=
9.81 m/s2), h is the characteristic height (in mm) of the slumped sample, and is
the critical angle of the plane (in degree) when the sample starts to flow. The
characteristic height (h) is determined by calculating the mean value of five
heights of the slumped sample near the center of the spread. Four tests are
performed after different periods of rest (after 15, 30, 45, and 60 min) to evaluate
the rate of increase in 0rest at rest.
Variations of 0rest obtained with the inclined plane test (IP0rest) with resting time
for typical SCC mixtures of different structural build-up, or thixotropic, properties
are shown in Fig. 11. Similar to the portable vane test, three structural build-up
indices can be obtained using the inclined plane method: (Ri), [R(t)], and
[RiR(t)].
247
thixotropic indices shown in Table 3 determined from the empirical test methods
are discussed. The first model was derived to predict K0 as function of concrete
height (H in m), casting rate (R in m/hr), concrete temperature (T in C), and
structural build-up (Eq. 4). The structural build-up is expressed in terms of static
yield stress at rest after 15 min of rest (referred as initial value) determined using
the portable vane test (PV0rest@15min in Pa). The influence of maximum size
aggregate (MSA) is incorporated in the model (fMSA) when using MSA other than
14 mm. The effect of different waiting times between successive lifts (fWT) is also
taken into consideration. PV0rest@15min measurements in Eq. 4 are measured at 22
2 C. Thus, a separate factor expressing the influence of concrete temperature is
introduced.
K0 = [112.5 - 3.8H + 0.6R - 0.6T +
0.01Dmin - 0.021 PV0 rest@15 min] fMSA fWT
(4)
where
fMSA is a correction factor for MSA other than 14 mm.
For relatively low thixotropic SCC [PV0rest@15 min 700 Pa] and
H<4m
fMSA = 1
H = 4 -12 m
fMSA = 1
fMSA = 1 +
1.26 H - 5.04
100
when MSA = 20 mm
. when MSA = 10 mm
For high thixotropic SCC [PV0rest@15 min > 700 Pa], fMSA = 1
fWT is a correction factor reflecting the effect of waiting time (WT)
between successive lifts and ranges between 0.85 and 1.0 for a waiting
time of 30 min, depending on the thixotropy of the concrete.
When the structural build-up varies with the temperature of the SCC at casting, a
modified prediction model for K0 values is shown in Eq. 4'. The fMSA and fWT
factors are the same as those defined for Eq. 4.
K0 = [98 - 3.82H + 0.63R + 0.011Dmin - 0.021 PV 0rest@15min] fMSA fWT
(4')
Correlations between measured and predicted K0 values are shown in Figs. 14 and
15. The predicting K0 from the models given in Eqs. 4 and 4' gives 1:1
relationship with coefficient of correlation (R2) of 1, which indicates excellent
agreement between both prediction models for K0.
Other two predicting models using the structural build-up indices of the portable
vane test were developed. The time-dependant change of static yield stress, i.e.,
rate, (PV0rest(t) in Pa/min) and the couple effect of initial static yield stress and its
248
(5)
(6)
Similar predicting models for K0 derived using the three structural build-up
indices of the inclined plane test are shown in Eqs. 7 to 9.
K0 = [112 - 3.83H + 0.6R - 0.6T + 0.01Dmin - 0.023 IP0rest@15min] fMSA fWT
(7)
(8)
(9)
The decreasing rate of concrete spread [US(t)] obtained from the undisturbed
spread test is used for the structural build-up index in the predicting equation of
K0 (Eq.10).
K0 = [106 - 3.95H + 0.6R - 0.62T + 0.016Dmin 1.762 US(t)] fMSA fWT
(10)
It is imperative noting down that the prediction models (Eqs. 4 to 10) are valid for
the ranges of tested parameters shown in Table 4 in terms of mix design, casting
conditions, and formwork geometry. These limitations are as follows:
H= 1 13 m
PV0rest@15min= 0 2000 Pa
R= 2 30 m/hr
T= 12 32 C
IP0rest@15min= 0 2000 Pa
IP0rest(t) = 0 30 Pa/min
US(t) = 0 22 mm/min
249
[2]
[3]
[4]
[5]
250
0.7
2.4
2.3
12
4.0
15
30
45
60
H
250
200
149
100
h
Vary according
to the filling
height of up to
total height (H)
(0rest)
Ri, (Pa)
(0rest)
Ri, (Pa)
R(t),
(Pa/min)
RiR(t),
(Pa2/min)
R(t),
R(t), (mm/min)
(Pa/min)
RiR(t), -(Pa2/min)
251
(average spread)
--
Range
252
63.5 mm
Sensor
19 mm
577 mm
704 mm
500 mm
Sensor
19 mm
63.5 mm
190 mm
Fig. 3 - Variations of lateral pressure with time for typical SCC mixtures
253
254
PV o rest (Pa)
4000
3500
3000
2500
2000
1500
1000
500
0
0
15
30
45
60
75
90
105
ISOMETRIC VIEW
10 mm
Bottom plate
Inclined plane
Angle
fixing rod
5 mm
FRONT VIEW
610 mm
400
400
mm mm
PLAN
255
D2
D1
256
257
Figure 16 - Correlation between K0 and static yield stress Figure 17 - Correlation between K0 and couple effect of static yield
after 15 min resting time obtained using portable vane test stress after 15 min resting time and its rate obtained using portable
for SCC cast, R=10 m/hr, resulted from Eqs. 4 or 4'
vane test for SCC, R=10 m/hr, resulted from Eq 6
258
259
- 260 -
INTRODUCTION
Self consolidating concrete (SCC) is considered as a revolution in concrete
technology. Though its advantages are many, its usage is still limited worldwide.
The two main reasons behind this slow growth and acceptance are: the higher
cost of SCC compared to conventional concrete and the lack of knowledge about
SCC in construction industry. This is true worldwide and the market in Saudi
Arabia is no exception.
In 2003 Saudi Readymix Concrete Company took a responsibility to raise the
awareness about SCC in Saudi Arabia by sponsoring and conducting many public
and private seminars about SCC. Since then Saudi Readymix managed to pour
SCC in several projects throughout the kingdom. However, Alturki Business Park
(ABP) was a unique project and, to the best of our knowledge, the only building
in the Gulf Area that was constructed completely using SCC.
SCC history and technology, its properties, advantages and testing methods are
well documented in many other papers, guidelines and technical reports. It is not
within the scope of this paper to talk about these points. The main objective of
this paper is to share the experience and knowledge gained during the
construction of ABP and to present the challenges encountered and what can be
expected on projects where SCC is specified.
The project
Al Turki business Park project (Figure 1) is located in Al Khobar city in the
Eastern Province of Saudi Arabia. The project consists of one office tower and 10
office lofts brought together by spacious surroundings, harmonized architectural
treatment, serene landscaping, and unified image. The development provides
34,000 m2 of leasable space with a project site area of 8,600 m2. The lofts are
three-floor buildings. The tower is a 16 floor building with two underground
floors providing parking for 300 cars connecting directly to the building lobby.
The design is based on a 9-meter structural grid with a variable cantilever to form
a cylindrical building with a 38-meter diameter. The structure is a multiframe
system. The slabs are designed as post-tensioned slabs with drop panel at column
locations. Vertical elements were founded on a thick reinforced concrete mat
foundation supported on soil. Vertical loads transmission is done by concrete
- 261 -
columns and concrete walls, lateral stability is given by concrete shear walls
around the staircases and elevators. The building is fully modern and will
incorporate all of the latest features of intelligent building design, sophisticated
communications, security, computer and other low voltage systems.
The project was designed by Pei Partnership Architects; a New York based design
office, and built by a local Saudi contractor (Alsaad). A local engineering office
(THA) was hired to manage the project while a Lebanese engineering firm (SKP)
was appointed as the project consultant.
Construction commenced on June 2008 and expected to be completed on June
2010.
Why SCC?
Though it was not necessary to use SCC for the whole structure, the decision to
build ABP using only SCC achieved three objectives.
1. Technical advantages: SCC generally provides better quality concrete for the
structure. This was observed in ABP project.
Some elements were highly congested with steel reinforcement (Figure 2) and
SCC was the right choice to prevent segregation and honeycombing. SCC
passed though congested steel and filled the forms without any problem.
Moreover, the use of SCC assured a uniform concrete consolidation.
Compacting conventional concrete using poker vibrators usually results in non
uniform compaction where one can find well compacted areas, partially
compacted areas and areas that were not compacted at all. Placing and
finishing SCC was much easier than normal concrete. It was possible to pour
concrete for a 7 m high wall continuously in a very short time without any
placing difficulties and concrete surface was very smooth.
The surface finish was excellent and almost blemish free and there was no
need to do any surface repair (Figure 3). The contractor acknowledged that
this was one of the best surface finishes they had ever produced.
2. Raising the construction industry awareness. This was done by giving a real
full scale live example about the constructability and many advantages of
SCC. As a matter of fact, many interested engineers, consultants, students and
university teachers visited the site during construction to see SCC being
placed and learn from this experience. Technical leaflets will be published
about the project and distributed to potential clients, engineers, designers and
consultants.
- 262 -
- 263 -
The following tests were performed on raw materials throughout the project
duration:
- Tests on aggregates: gradation, moisture content and absorption, specific
gravity, flakiness and elongation, abrasion as well as chemical analysis.
-
Pumpability of SCC
Pumping high strength, SCC in high rise buildings is a challenge. Despite the fact
that SCC is flowable, pumping this type of concrete requires more pressure1 and
more care and attention. SCC is very sensitive and may exhibit segregation and
slump flow loss due to the change in pumping pressure or discharge rate.
- 264 -
Resistance to change. This was expected and it was a direct result of the
previous point.
- 265 -
and mixtures ingredients had to be changed and tweaked with the variation
of aggregates quality and moisture condition. There was a shortage in
good quality aggregates and it was not possible to maintain the same
quality of raw materials. The same problem was observed with cement
whose quality was oscillating and difficult to predict.
-
Loss of flow. This was confronted several times and in some cases it was
necessary to use vibrators to compact the concrete. One reason behind this
drop in flow was pumping. We knew that pumping is affecting the
rheological properties of SCC and that was causing the drop in flow from
700 50 mm before pumping to about 575 25 mm after pumping.
However, the exact scientific explanation was not clear to us. Recently, we
understood that pumping affects the viscosity and the yield stress, and
hence flowability, of SCC concrete due to structural breakdown and an
increase in the air content. Both factors decrease the viscosity. However,
yield stress increases due to structural break down and increases due to an
increase in air content1. In ABP project we measured the air content and
found it to increase from about 1.6% before pumping to about 2.5% after
pumping.
Loss of flow was also caused by the delay of trucks on site and due to high
temperature and drying of concrete.
Though it was not observed in lab trials and on station during production,
in some pours, bleeding water was noticed clearly on the top of concrete
after pouring. In one case, a paste layer of about 2.5 cm thick was formed
on the top of the slab and had to be removed. In those cases, the mix was
adjusted immediately. In other cases superpozz caused an unpleasant
stained surface on some columns due to the floatation of superpozz to the
surface with the bleeding water. As a result, we stopped using supepozz
and replaced it with silica fume. This adjustment gave more robust mixes
and the stains were never noticed again.
At early stage of the project and upon striking of the forms, pour lines
were observed in the soffit of one slab. However, further inspection and
investigation proved that these pour lines were superficial and did not
have any adverse results on the structure.
- 266 -
Occasionally, blow holes were noticed on the surfaces. This was attributed
to air migration or air popping where air migrates to the surface. In order
to minimize and eliminate this undesirable occurrence the viscosity and
flow of the mix were modified.
Benefits
The benefits obtained from using SCC for the project may be summarized as
follows:
- More understanding of SCC behavior, production, transporting and
placing. The project provided important learning topics.
-
- 267 -
and to assign tasks for the team members in order to make sure that the job
would go smoothly.
-
Conduct necessary trials and mock ups ahead of time and get good
knowledge about the behavior of SCC using different raw materials.
Modify and adjust and optimize the mix as necessary during the course of
the project.
SCC is sensitive to variation. Try to keep the raw materials uniform and
consistent and from a single source, if possible.
Select the most proper and suitable admixture. Different admixtures will
perform differently.
Rate of placing SCC should not be fast to allow for entrapped air to escape
Since SCC does not bleed, plastic shrinkage cracks are more likely to
occur during hot weather concreting and necessary preventive precautions
need to be taken.
There is still a lot of work to be done to optimize SCC mixes and reduce
its cost.
- 268 -
REFERENCES
1. Feys, D., De Schutter, G. and Verhoeven, R., Rheology and Pumping of SelfCompacting Concrete, Proceedings of the tenth ACI international conference on
recent advances in concrete technology and sustainability issues, Seville, Spain,
2009.
2. Kasten, K., Pump Trial for SCC at SRMCC (unpublished report),
Putzmeister, Germany, 2009.
- 269 -
Unit
Limits
Slump flow
Mm
Min
650
max
750
T50
Sec
J-Ring(flow
difference)
L-Box
Mm
50
h2/h1
0.8
1.0
V-funnel
Sec
12
Test
Unit
Limits
Slump flow
mm
Min
650
max
750
T50
Sec
(flow mm
50
h2/h1
0.8
1.0
Sec
12
J-Ring
difference)
L-Box
V-funnel
- 270 -
1A -Basements
1C - Almost completed
1B -First Floors
- 271 -
- 272 -
100
90
80
Percentage passinge
70
60
20 mm
50
10 mm
40
5 mm
Fine sand
30
20
10
0
0.075
0.75
7.5
- 273 -
- 274 -
- 275 -
90
80
70
60
50
40
30
20
Actual
10
Linear (Rheometer)
0
0
20
40
60
80
- 276 -
100
APPENDIX
- 277 -
- 278 -
- 279 -
Hani Nassif teaches bridge engineering and prestressed concrete at Rutgers, The
State University of New Jersey. He is a Fellow of the American Concrete
Institute (ACI) and member of its technical activities committee (TAC). His
research area is in structural concrete with emphasis on using SCC and HPC for
infrastructure applications. He holds a BS and MS from the Uuniversity of
Detroit, and Ph.D. from The University of Michigan-Ann Arbor.
Nakin Suksawang is an Assistant Professor at Florida International University,
Miami, Florida. His research area is in HPC, SCC, and cementitious materials.
He holds his BS, MS, and Ph.D. from Rutgers, the State University of New
Jersey.
Kagan Ates holds an MS Degree from Rutgers, The State University of New
Jersey and has performed research work on Restrained Shrinkage of HPC.
Husam Najm is an Associate Professor at Rutgers, The State University of New
Jersey. His area of research is fiber reinforced concrete, SCC, and advanced
materials.
INTRODUCTION
Self-consolidating concrete (SCC) is defined as concrete with high viscosity that
can flow freely without segregation. SCC is able to flow under its own selfweight into corners of formwork and through closely spaced reinforcement with
little or no vibration or compaction. This leads to lower energy cost, lower stress
on the formwork, reduced labor cost, and elimination of the potential human error
in consolidation of the concrete. The concrete becomes more consistent since
there are equal dispersions of the cementitious paste and aggregates. As a result,
both SCC mechanical properties and durability are improved over
normal/conventional concrete, which has increased its popularity in Europe and
Japan (1-5). However, the use of SCC in the United State (US) remains limited.
Part of the reason is because of the limited knowledge and experience in using it,
as well as its high initial cost. Nevertheless, there is an effort led by the Federal
Highway Administration (FHWA) to promote the use of SCC in transportation
structures in the US (6).
Fresh concrete can easily attain high flowability by simply increasing the waterto-binder (w/b) ratio. However, increasing the w/b ratio alone could lead to
concrete segregation and less durability. Thus, in order to successfully develop
- 280 -
SCC, mineral and chemical admixtures, e.g., pozzolans, lime stone filler,
superplasticizer, and viscosity-modifying admixture (VMA), need to be added to
the mix design to prevent segregation and enhance the durability of SCC. In
addition, the absolute volume of coarse aggregates also needs to be limited to
reduce inter-particle friction and allow the SCC to flow under its self-weight
without segregation (7-11). Also a reduction in the volume of coarse aggregate
would need to be balanced by an increase in the volume of cement paste, which
would result in higher material cost and an increase in the capillary pores. One
solution to decrease the paste volume is to use VMA that reduces interparticle
friction and increases flowability, and therefore, the volume of coarse aggregates
could be increased. Despite this solution, concrete producers and owners still
have questions on the use of VMA because there is little information on its longterm effects and the effects of various chemical admixtures on SCC. Therefore,
there is a need for research to study the effects of various chemical admixtures on
SCC and evaluate and compare alternative solutions.
One of the alternative solutions is to use Pozzolans, such as fly ash or dust
powder, to replace the cement content (12-14). The pozzolanic materials do not
only reduce the cement content, but also fill the capillary pores, which make the
concrete denser, thus increasing the concrete durability. Moreover, some of the
Pozzolans, such as fly ash and slag, can also increase the flowability of concrete,
which results in reducing the amount of superplasticizers and lowering production
cost. However, any decrease in the volume of coarse aggregate and any increase
in the volume of cementitious paste will greatly affect the mechanical properties
of SCC. Thus, in order for SCC to become a standard concrete mixture, the
mechanical properties need to be evaluated to ensure a level of comfort for the
bridge designers and owners.
There are different methods to produce SCC and we need to evaluate these
methods for its proper use in construction projects, specifically for drilled shafts.
Before SCC can be incorporated in New Jersey Department of Transportation
(NJDOT) projects, mix designs and specifications need to be developed. These
items should reflect the state-of-art in design of SCC while incorporating factors
specific to New Jersey and the availability of SCC producers and need in various
projects. Factors to be considered in developing specifications include: aggregate
gradation requirements, maximum volume of coarse aggregates, appropriate filler
materials and other admixtures, flow and segregation testing methods, durability,
and material handling and placement requirements.
The main objective of this paper is to provide results from a laboratory performed
to compare the mechanical properties of SCC, HPC and NC and to construct and
monitor the performance of three concrete drilled shafts that was cast using SCC
- 281 -
(15). Three out of five shafts were instrumented with strain and temperature gages
to monitor the behavior of SCC during hydration and in curing phases. Fresh
concrete testing was also performed to evaluate existing testing procedures and
identify those that are reliable and easily implementable under field conditions.
Moreover, concrete cylinder specimens were taken to verify the design strength
requirements. Finally cross-hole sonic logging tests were performed to test the
continuity and integrity of the drilled shaft profile and cross section.
EXPERIMENTAL PROGRAM
This study consisted of a laboratory testing program of SCC, HPC and NC
specimens as well as constructing drilled shafts with SCC on I-280 Interchange
Project located in Newark, New Jersey.
Laboratory Tests
The laboratory tests included compressive, tensile, elastic modulus, drying
shrinkage, and rapid chloride permeability. Table 1 shows the various mixes and
the results of compressive, tensile, and elastic modulus tests.
Figure 1 illustrates the drying shrinkage of normal/conventional concrete, SCC,
and HPC mixes. It is observed that the HPC mixes had the lowest drying
shrinkage, whereas the SCC mixes had the highest drying shrinkage. The reason
for this is because the SCC mixes had higher cement paste and higher capillary
pores that lead to higher drying shrinkage. It is also observed that the addition of
pozzolans to SCC reduced its drying shrinkage. Among all the SCC mixes
evaluated in this study, the mix containing both SF and F has the lowest drying
shrinkage with a 20% reduction in comparison with the SCC1 or regular SCC
mix. Mix SCC2 (or SCC mix containing 20%F) also had a 10% reduction,
whereas, mix SCC3 (or SCC containing 30%SL) did not have a significant
reduction.
Figure 2 shows the rapid chloride permeability test (RCPT) results of all mixes. It
is observed that there is a significant variation in the RCPT charge passing rates
among normal/conventional concrete, SCC, and HPC. The SCC mixes had
significantly higher charge-passed values at all testing days. This can be
attributed to the increase in paste volume that leads to higher capillary voids. The
added pozzolanic materials (with the exception of SL) also had a significant effect
on the RCPT of SCC.
Also, Figure 2 shows that SCC with and SF Pozzolans had about 30% reduction
in the charge-passing rate compared to the regular SCC. The combination of SF
- 282 -
and F seems to have the most influence on the permeability with a 57, 74, and
81% reduction at 28, 56, and 90 days, respectively, over regular SCC. The added
SF also reduces the permeability of SCC containing F by approximately 50%.
The reason for this is that SF is highly reactive and has a high surface area,
allowing it to fill the capillary pores and making the concrete denser and less
permeable. In addition, as expected, both HPC mixes outperform all other mixes
in RCPT test.
Field Instrumentation and Testing
The field work started in November of 2006 and the drilled shaft construction was
completed in mid 2007. The authors worked with the contractor and the NJDOT
resident engineer on coordinating the instrumentation and collection of field
samples during the construction of the drilled shafts. Two SCC Shafts from the
project and a demonstration shaft were instrumented with vibrating wire strain
gages to monitor load effects and temperature. Results from drilled shaft testing
(including the standard cross-holes sonic logging (CSL) performed by the
Contractor) were used to evaluate shaft integrity similar to the other shafts of the
project.
The drilled shaft was instrumented with five Geokon Model 4200 Concrete
Embedment Gages to monitor the temperature and strains during and after
construction. In addition to measuring the concrete strain, the gage is also
equipped with a thermistor to measure the temperature. Table 2 and Figure 3
illustrate the sensor locations inside the drilled shaft. Figure 4 shows the concrete
embedment gage attached to the cage of the drilled shafts. The data acquisition
system was used to collect data at 5-minutes interval from all sensors starting
from the time of pour.
FRESH CONCRETE PROPERTIES
Various fresh concrete tests were performed on the SCC sampled from the truck
(Figure 5). Due to space limitations at the location of the drilled shaft concrete
pours, not all tests were performed for each truck. However, the slump flow test
was performed for every truck arriving at the site.
CONCRETE FIELD SAMPLING AND LABORATORY TESTING
Twenty 4 x 8 inch and three 6 x 12 inch concrete cylinders were taken at the day
of the concrete pour from the second of five trucks that were assigned for the job.
The specimens taken were to be used to test the compressive strength as well as
the modulus of elasticity of the mix at different ages. The 6 x 12 cylinders were
also used to confirm the results from the other cylinders at 28 days.
- 283 -
RESULTS
Fresh Concrete Properties
Table 3 and 4 summarize the results obtained from the spread, j-ring, and l-box
tests for drilled shafts #3 and #1, respectively. For drilled shaft #3, the value of
the test from the first truck was higher compared to the other trucks. According to
the results of the J-Ring test the concrete from the first truck will have minimal to
noticeable blocking and the concrete from Truck 3 will have no visible blocking.
L Box test performed on the concrete from Truck 4 indicated that blocking may
be experienced in the concrete from the 4th truck since h2/h1 was observed to be
less than 0.9.
For drilled shafts #1, the slump flow values from both trucks varied between 19
and 21.5 inches, which is less than the limit of 24 inches specified by NJDOT
Specifications. The JRing test performed on the second truck (measured flow is
21.5 inch.) indicated that blocking should be observed in the concrete from this
truck. The LBox readings correlated well with the J-Ring test results showing
that blocking might be obtained due to the h2/h1 ratio of 0.632, which is lower
than the LBox test limit 0.9.
MECHANICAL PROPERTIES
Figure 6 shows a comparison of mixes obtained from the field including the
demonstration shaft. It is clear that concrete from drilled shafts 1 and 3, both
having lower slump flows, outperformed the concrete from the demonstration
shaft. Figures 7 and 8 compare the tensile splitting and modulus of elasticity of
drilled shafts, respectively. Both shafts have comparable mechanical properties,
with the exception of the modulus of elasticity where shafts 1 was stiffer. This
could be attributed to the lower slump flow in shaft 1.
STRAIN AND TEMPERATURE PROFILE
Figures 9 and 10 show the change in strains and temperature from the time of
pour (December 11, 2006) to 28 days (10:25 am January 8, 2006), respectively.
Sensors 3 and 4 recorded temperature values around 63 0F. This is expected since
these sensors are in the middle of the shaft. Dissipation of the heat of hydration is
therefore much slower than the concrete near the top surface (which is exposed to
ambient temperatures) and the concrete that is close to bottom (which dissipates
heat through contact with the surrounding soil). The initial temperature recorded
by the gages is the temperature of the concrete at the time of arrival of the truck to
- 284 -
the site. This value is around 75 0F for all sensors. Peak temperature of Sensors 1
and 2 are 106 0F and 94 0F, respectively. For Sensors 3 and 4 the values were
126.5 0F and 128 0F, respectively, while the value for Sensor 5 was 111 0F.
Strains in all sensors are in compression indicating that the concrete is shrinking
freely.
Cross-Hole Sonic Logging Evaluation of Drilled Shafts
The project Contractor performed Cross-hole sonic logging evaluation on each
drilled shaft including the demonstration shaft. The results showed no (major) air
pockets or a discontinuity in the integrity of any of the shafts. This means that the
SCC mix used was successful in passing through the dense reinforcement layers
in the 6 feet diameter shafts. Appendix A shows typical results from the analysis
of the CSL of the Drilled Shafts in I-280 Interchange.
RECOMMENDATIONS AND CONCLUSIONS
The following conclusions and recommendations could be made from this study
1. Performance of the SCC obtained from the drilled shafts was found
satisfactory. The only problem that was encountered was during the pour of
the demonstration shaft. The slump flow measured was over 28 inches and
segregation was observed in the cylinders during testing which adversely
affected the strength of this mix. This problem was overcome by reducing the
slump flow requirement to in between 20 inches and 24 inches. Therefore, it
is suggested to keep the slump flow in SCC applications within these limits to
prevent segregation and strength loss.
2. SCC in drilled shaft should not be dropped from higher distances. In this
project, dropping SCC into the drilled shafts from heights more than 6 ft,
resulted in segregation. For higher drops, a tremie should be used to avoid
segregation.
3. ASTM Standard J-Ring Test (C1621/C1621M-06 Standard Test Method for
Passing Ability of Self-Consolidating Concrete by J-Ring) was successfully
used to test the passing ability of SCC. It is suggested that this test be used as
a part of quality control measure along with the regular slump flow test when
a more qualitative result is needed. This is especially true for mixes with
superplasticizer only and high coarse aggregate content where the slump flow
may pass the slum flow test but may segregated when passing through the Jring. Visual rating could also be used in lieu of the J-ring test but required
trained eyes.
- 285 -
- 286 -
7. Saak, A.W., H.M. Jennings, S.P. Shah, New Methodology for Designing SelfCompacting Concrete, ACI Materials Journal, 98 (6), 2001, pp. 429 439.
8. Sedran, T., F. De Larrard, Optimizing of Self-Compacting Concrete,
Proceedings Of The 1st RILEM Symposium on Self-Compacting Concrete, 1999,
pp. 321 332.
9. Edamatsu, Y., N. Nishida, A Rational Mix-Design Method for SelfCompacting Concrete Considering Interaction Between Coarse Aggregates And
Mortar Particles, Proceedings of The 1st RILEM Symposium on Self-Compacting
Concrete, 1999, pp. 309 320.
10. Jacobs, F., F. Hunkeler, Design of Self-Compacting Concrete For Durable
Concrete Structures, Proceedings of The 1st RILEM Symposium on SelfCompacting Concrete, 1999, pp. 397 407.
11. Xie, Y., B. Liu, J. Yin, S. Zhou, Optimum Mix Parameters of High-Strength
Self-Compacting Concrete with Ultrapulverized Fly Ash, Cement And Concrete
Research, Vol. 32, No. 3, 2002, pp. 477-480.
12. Bouzoubaa, N., M. Lachemi, Self-Compacting Concrete Incorporating High
Volumes of Class F Fly Ash: Preliminary Results, Cement and Concrete
Research, Vol. 31, No. 3, 2001, pp. 413-420.
13. Ho, D.W.S., A.M.M. Sheinn, C.C. Ng, C.T. Tam, The use of Quarry Dust for
SCC Applications, Cement and Concrete Research, Vol. 32, No. 4, 2003, pp. 505
511.
14. Suksawang, N., and Nassif, H.H., Effect of Modulus of Elasticity on Creep
Prediction of High Strength Concrete Containing Pozzolans, Shrinkage and Creep
of Concrete, ACI, SP-227, Farmington Hills, M.I., 2005, pp. 261-284.
15. Nassif, H., Aktas, K., Suksawang, N., Najm, H., (2008) "Self-Compacting
Concrete in New Jersey (Phase I & II)," FHWA-NJ-2007-10, Draft Final Report
Submitted to NJDOT, 51 pp.
- 287 -
Mixtures Identification
NC
SCC1 SCC2
477
515
412
-----103
---1022
817
817
701
885
885
186
201
201
0.39
0.39
0.39
2339
5647
3850
94
51
77
---127
699
629
6.00
6.50
5.75
SCC3
386
26
103
-817
885
201
0.39
4620
51
-635
5.00
SCC4
360
--154
817
885
201
0.39
2631
51
-648
6.00
SCC5
335
--85
860
801
148
0.35
3094
445
774
635
6.00
HPC1
382
-95
-1022
701
186
0.39
1476
498
-83
5.00
HPC2
358
24
95
-1022
701
186
0.39
2214
498
-79
5.00
19.8
28.9
36.0
40.8
42.4
45.2
43.7
32.6
43.5
48.2
60.6
60.0
62.1
61.8
19.4
29.3
36.0
39.0
44.2
51.0
56.1
22.3
29.3
32.9
41.1
46.0
49.6
53.6
22848
25407
27303
30395
33336
33636
34888
18332
22572
20893
26146
31271
31018
29489
25209
26078
29556
29035
31637
34656
38322
23722
24013
25177
26716
27719
28871
30270
3.7
3.9
4.2
4.4
5.3
5.4
5.5
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
- 288 -
Slump
Flow
(in)
J-Ring
Flow
(in x in)
1
2
3
4
28.5
24.5
18.5
19.5
27
18.25
-
Difference
from
Slump
Flow (in)
1.5
0.25
-
L Box
h2 (in)
h1 (in)
h2/h1
3.25
4.75
0.684
Truck
Slump
Flow
(in)
J-Ring
Flow
(in x in)
1
2
19
21.5
21.5
Difference
from
Slump
Flow (in)
0
- 289 -
L Box
h2 (in)
h1 (in)
h2/h1
4.75
0.632
200
0
-200
-400
NC: w/b=0.39, 100%OPC
SCC1: w/b=0.39, 100%OPC
SCC2: w/b=0.39, 20%F
SCC3: w/b=0.39, 5%SF 20%F
SCC4: w/b=0.39, 30%SL
SCC5: w/b=0.35, 20%SL
HPC1: w/b=0.39, 20%F
HPC2: w/b=0.39, 5%SF 20%F
-600
-800
-1000
1
10
Time (Days)
- 290 -
100
- 291 -
6000
4000
HPC2: w/b=0.39
5%SF 20%F
7000
HPC1: w/b=0.39
20%F
SCC5: w/b=0.35
20%SL
SCC4: w/b=0.39
30%SL
SCC3: w/b=0.39
5%SF 20%F
SCC2: w/b=0.39
20%F
SCC1: w/b=0.39
100%OPC
NC: w/b=0.39
100%OPC
28 Days
56 Days
90 Days
5000
NJDOT RCPT
Specified Passing
Rate
3000
2000
1000
Sensor 1
Sensor 2
Sensor 4
Sensor 3
Sensor 5
- 292 -
- 293 -
69
8000
55
6000
41
4000
28
2000
0
Demo Shaft
Shaft #3
Shaft #1
0
10
20
30
40
Time (Hours)
50
14
10000
0
60
800
600
400
200
1
Shaft #3
Shaft #1
10
20
30
40
Time (Hours)
50
0
60
- 294 -
41
5000
34
4000
28
3000
21
Shaft #3
Shaft #1
2000
10
20
30
40
Time (Days)
6000
14
60
50
100
Sensor 1
Sensor 2
Sensor 3
Strain (
Sensor 4
Sensor 5
-100
January 8, 2006
-200
-300
December 11, 2006
-400
10
15
20
Time (Days)
25
- 295 -
30
140
Sensor 1
Sensor 2
Sensor 3
Sensor 4
Sensor 5
Temperature (F )
120
100
80
60
40
December 11, 2006
20
January 8, 2006
10
15
20
25
30
Time (Days)
Figure 10- Temperature profile of the demonstration shaft
- 296 -
- 297 -
James Aldred, Ph.D, CPEng, LEED AP, FACI has over 25 years in the concrete
industry. Dr Aldred is Principal Engineer with GHD Pty Ltd and Manager of the
Independent Verification and Testing Authority for the Burj Dubai. James is a
member of ACI 130, 201, 212, 233, 234, 305 and 365 and authored over 40
technical papers on various aspects of concrete technology.
INTRODUCTION
According to the Council on Tall Buildings and the Urban Habitat by October
2007, there were over 55 super-tall buildings of 300 metres or greater under
construction, over half of which are in the Middle East and being constructed
primarily on reinforced concrete piles and raft foundations (1). However, the
Middle East is not a benign environment for concrete and great care and attention
is necessary to achieve durable concrete in this region. The flowing properties of
self-consolidating concrete (SCC) have been used to minimize the potential for
honeycombing in these elements, particularly in situations where there is a highly
congested reinforcing steel.
One concern regarding the use of SCC has been the requirement for a relatively
high binder and admixture content to achieve the required combination of
flowability and segregation resistance. In the case of foundation elements in the
Middle East, the durability requirements of the specification would normally
require a low water/cementitious ratio and a high binder content, often with
significant replacement with supplementary cementitious materials. Therefore
modification of aggregate grading and the use of suitable viscosity modifying
agent may be the only changes to achieve self-consolidating characteristics at
minimal additional cost. The fact that SCC would greatly reduce the size of the
concrete placing team means that SCC could provide significant cost savings in
these applications.
While there have been numerous investigations on laboratory and mock-up scale
specimens, there is little published information about the in-situ properties of SCC
from actual construction, particularly with respect to penetrability and durability
parameters. This paper presents the data from cores taken from a large raft after
construction.
SPECIFICATION REQUIREMENTS
The required compressive strength for the raft concrete was 50 N/mm2. The
penetrability requirements were 2.0% water absorption in accordance with BS
1881: Pt 122, 1000 coulombs in accordance with ASTM C1202 and 15 mm
penetration in accordance with BS EN 12504-1. The temperature limit was 710C
and the specification called for the use of SCC in the raft with a minimum
slumpflow of 600 mm.
- 298 -
252 kg/m3
168 kg/m3
30 kg/m3
368 kg/m3
368 kg/m3
552 kg/m3
552 kg/m3
60%
4 - 6 L/m3
1 - 1.5 L/m3
0.32
- 299 -
considering usual bleed effects. The compressive strength of the cores tested to
ASTM C39 were only slightly lower and also showed limited strength reduction
towards the top surface (Figure 7).
The indirect tensile or tensile splitting test results showed greater variability as
would be expected but there was also a greater tendency towards reduced tensile
strength for cores taken from higher levels within the raft (Figure 8). Static
modulus values appeared relatively consistent at around 40 GPa until the top 400
mm where the value dropped by approximately 30 percent (Figure 9). This
suggests that there had been a reduction in aggregate content in the uppermost
samples which was apparent from visual examination.
Penetrability Properties:
Figure 10 shows a considerable variation in the measured coulomb values from
the limited number of tests conducted and a trend towards higher values at the top
surface. The negligible values measured below the top surface were profoundly
less than those for the trial mix and , even the highest value obtained from the
core at the surface was still significantly lower than the specification requirement
and the value for the trial mix.
The water absorption showed a similar trend of lower values below the upper
surface. However the surface value was well within the specification requirement.
Cores were also tested for chloride diffusion using a modified NT Build 443
procedure. The concrete in the cores was tested approximately 3 months after
casting and therefore the chloride diffusion coefficient would have reduced by
approximately 30% to 50% compared to the value measured after 28 days. The
original modeling of service life by GHD had been based on a 28 day chloride
diffusion coefficient (Dce) of 3 x 10-12 m2/s and a mature Dce of 0.5 x 10-12 m2/s.
The average Dce for the five core samples tested was 1.54 x 10-12 m2/s which is
approximately the value anticipated by GHD. The variation in Dce with depth was
only 0.92 1.50 x 10-12 m2/s which is within the range expected.
CONCLUSIONS
A wide range of mechanical and penetrability properties were measured on cores
taken from a thick raft cast with SCC in Dubai. The data presented in this paper
demonstrate that the properties of the hardened concrete exceeded the
specification requirements. There was a trend towards reduced performance in the
surface layer but this was still well within the requirements. The reduction in
many of the parameters was probably less than would be expected due to bleed
effects in normally placed concrete.
REFERENCES
1. Oldfield, P. (2007) The Tallest 20 in 2020 Council on Tall Buildings and
Urban Habitat, November, 2007.
- 300 -
80
70
60
Strength MPa
50
40
30
20
10
0
0
10
15
20
25
30
Age - days
Figure 1: Compressive strength development for C50-20 - SCC (Trial mix)
50
45
40
Modulus GPa
35
30
25
20
15
10
5
0
0
10
15
20
25
Age days
Figure 2: Static modulus for C50-20 - SCC (Trial mix)
- 301 -
30
800
Slumpflow (mm)
775
750
725
700
675
650
625
600
0
10
15
20
25
30
35
Hours
Strength MPa
20.0
10.0
0.0
No of results
- 302 -
80
A
m
b
i
e
n
70
60
40
30
20
10
0
10
20
30
40
Days
50
60
70
80
Date : Time
Temperature (Degree C)
50
4-800 (5)
8-1200 (9)
12-1600 (2)
16-2000 (2)
20-2400 (0)
24-2800 (4)
28-3000 (3)
0
10
20
30
40
50
60
70
80
90
Sterength MPa
- 303 -
4-800 (3)
8-1200 (1)
12-1600 (4)
16-2000 (0)
20-2400 (1)
24-2800 (2)
28-3000 (2)
0
10
20
30
40
50
Strength MPa
60
70
80
Shear
0-400 (4)
4-800 (1)
8-1200 (1)
12-1600 (1)
16-2000 (0)
20-2400 (3)
24-2800 (0)
28-3000 (1)
0
1000
2000
3000
4000
5000
Shear kPa
- 304 -
6000
Modulus
0-400 (2)
4-800 (3)
8-1200 (5)
12-1600 (0)
16-2000 (5)
20-2400 (0)
24-2800 (1)
28-3000 (1)
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
Modulus N/mm2
100
200
300
400
500
Permeability - coulomb
- 305 -
600
Wate r Absorption
0-400 (4)
4-800 (1)
8-1200 (1)
12-1600 (1)
16-2000 (0)
20-2400 (0)
24-2800 (0)
28-3000 (0)
0
0.2
0.4
0.6
0.8
1.2
Water absorption - %
- 306 -
1.4
Keywords: Fly ash, Ground Granulated Blast Furnace Slag (GGBS), Self
Compacting Concrete, Sustainability
- 307 -
- 308 -
In UAE the bulk source of both coarse and fine aggregates are from crushed
limestone mainly from Ras Al Khaimah and Fujairah limestone deposits. The 05mm crushed rock fine (CRF) contains a very good source of filler passing below
75micron sieve. Traditionally, large amount of this filler would be removed from
the CRF by washing to satisfy ASTM C33 which allows maximum of 5 to 7
percent passing 75 micron sieve. Most of the local specification would also put a
limit of 5% maximum allowable fine content for CRF. Generally unwashed
CRF would contain 10 % of fine limestone dust which is below 75micron, a
good source of filler material contributing to the aggregate packing improving
strength, durability and the flowability of concrete. All these attributes are
equally important for SCC. An attempt has been made to compare the
performance of all these three criteria of SCC made of washed and unwashed
CRF. Industry would save a significant amount of resources should the
acceptable minimum dust content in CRF be raised to a much higher level such as
10% without compromising basic requirement of concrete characteristics. This
would also add value to its environmental credential as a significant amount of
waste would be reduced and expensive energy would be saved by not requiring of
washing CRF.
Finding novel materials alternative to portland cement to reduce CO2 emission has
become an essential field of research interest to the scientific community. While
the doors of possibilities are wide open, actual available alternatives are limited.
Ground Granulated Blast Furnace Slag (GGBS), Fly ash and Microsilica have
been well proven to be partial alternative to portland cement. In the Middle East
they have been widely used. It is not necessarily due to Environmental point of
view, rather to meet durability requirements of the concrete. Traditionally for
GGBS, portland cement are being replaced from 55 to 70% by mass depending on
the durability requirements. 30% replacement for Fly ash is widely accepted.
Microsilica depending on both durability and strength could range from 4 to 12%
replacement of portland cement. As none of these materials are locally produced
an inherent carbon footprint due to shipment is unavoidable. Reprocessing works
to make them useable materials also add additional carbon footprint.
Nevertheless, considering from current scenario GGBS, Fly ash and Microsilica
are the only viable alternative to portland cement in the Middle East.
Strength development of alternative cementitious materials
One of the inherent problems with the alternative cementitious materials is loss of
strength when using beyond their usual prescribed limits. The chemical reaction
resulting subsequent development of strength enhancing hydrate phases depends
on the amount of portland cement present in the system. Secondly, presence of
Ca(OH)2 is a requirement for any pozzolanic reaction to occur, which is being
- 309 -
reduced with the reduction of the mass of portland cement from the cement paste,
hence loss of strength.
An alkaline activator such as NaOH or waterglass (SiO2.nNa2O.mH2O + NaOH)
with pH higher than 13 can be used to activate slag or Fly ash. The high alkaline
environment of the activator will break the bonds of Ca-O, Si-O and Al-O from
surface of slag creating reactive silicate and aluminate species. When the
concentration of these species reaches the product of solubility of the different
solid compounds, they precipitate creating CSH and CAH phases.[1]
The exact mechanism, however, of the setting of alkali activating slag/Fly ash is
not fully understood. Most proposed mechanism consists of a dissolution,
transportation or poly-condensation. Temperature is also considered as one of the
factors that affects to any polymerization. [2] One of the major issues of alkali
activation is the requirement of higher temperature for the strength development,
which often is not practical for actual application. In this study a special
proprietary admixture from Chryso has been used to boost the hydration process
and enhance the strength gain.
OBJECTIVES
The objective of this case study to explore the strength development and
subsequent durability performance of concrete containing very high replacement
of alternative cementitious materials such as GGBS and Fly ash while
maintaining their self compacting capabilities. The impact of excess filler present
in the unwashed CRF in the physical properties of the SCC is also of interest.
This study will demonstrate the environmental credential of SCC using alternative
cementitious materials and unwashed sand to save energy and reduce waste.
TRIALS
GGBS and Fly ash have been considered to test their maximum allowable limit to
replace portland cement. Microsilica has been added in to both GGBS and Fly ash
mixes to observe the improvement in performance, if any.
Each GGBS and Fly ash series examines the effect of filler content in the CRF
sand by using unwashed and washed 0-5mm sand made of limestone fines.
GGBS Mix:
A mix containing 70% GGBS + 30% portland cement has been chosen to be the
control mix as this proportion is being widely used in the Middle East. The
proportion of GGBS content has been increased by 10% for next 3 mixes
- 310 -
containing 80%, 90% and 100% GGBS, reducing the portland cement content at
the same rate.
A 5% of Microsilica has been added for additional 3 mixes replacing same
amount of GGBS while keeping the portland cement content at the same rate.
This is to observe whether Microsilica could help to boost the strength as
increasing GGBS would naturally reduce the compressive strength.
Both unwashed and washed sands have been used for above two sets of concrete
mix designs creating total 4 sets of concrete to compare their performance with
each other.
Total cementitious content of each mix is 400 kg with 152kg of water providing
w/b of 0.38. This is same for both GGBS and Fly ash series.
The concrete has been designed to be self compacting with having initial flow of
700mm. The expected flow at 90 minutes to be more than 600mm.
Total 14 mixes have been prepared for GGBS series.
Table 1 shows the GGBS containing concrete mix designs.
Fly Ash Mix
A 30% replacement by Fly ash with 70% portland cement is the control mix for
the Fly ash series. Similar to the GGBS mixes, a 10% increment of Fly ash has
been chosen for up to maximum 80% replacement.
Microsilica has been added at a 5% rate replacing same amount of Fly ash to
observe whether there is any significant boost of strength mainly due to particle
packing phenomenon.
Similar to the GGBS series both unwashed and washed CRF have been used,
creating 4 sets of concrete. All concrete have been designed to be self
compacting with similar flow requirement of GGBS mixes. Total 22 numbers of
concrete mixes using Fly Ash have been produced.
Table 2 shows the Fly Ash series concrete mix designs.
Cement and cementitious materials
Among all cementitious materials only portland cements are locally produced.
GGBS are locally ground while classified Fly ash and densified Microsilica are
imported.
The chemical compositions of all cementitious materials are illustrated in Table 3.
- 311 -
- 312 -
- 313 -
Figure 7 & 8 compares the strength gain of GGBS mixes with and without
Microsilica for both unwashed and washed sand respectively.
Figure 9 & 10 compares the effect of sand on the strength gain at 28 days. Mixes
containing unwashed crashed rock fine perform better than mixes containing same
sand but washed to satisfy ASTM C33. The unwashed sand contains
approximately 9-10% fines smaller than 75 micron. Higher compressive strength
at all age levels for both mixes with and without Microsilica have been observed
for mixes containing unwashed sand compared to washed sand mixes.
Table 13 displays the results of compressive strength of GGBS mixes for all ages.
Fly Ash Mixes: Fly ash series produced very similar results like the GGBS mixes.
The control mix with 70% Portland cement + 30% Fly ash is having the best
compressive strength. The strength reduction is almost linear to the rate of
reduction of portland cement from the concrete. However, mix containing 80%
Fly ash with 20% portland cement (unwashed sand) shows a significantly
improved strength of 30 MPa at 28 days. There is no significant impact on
Microsilica additions was found, perhaps due to the same reasons as mentioned
earlier.
Similar to GGBS mix, concrete containing unwashed sand shows better strength
enhancement compare to concrete with washed sand in the Fly ash mixes.
Figs. 11, 12, 13, 14 and Table 14 illustrate and compare the compressive strength
development of Fly ash mixes.
Durability Tests:
Concrete containing Microsilica for both GGBS and Fly Ash mixes were not
tested for durability due to disappointing compressive strength results.
Rapid Chloride Permeability Test: RCPT values as per ASTM 1202 are excellent
for both GGBS and Fly ash mixes especially for GGBS mixes. No significant
differences have been observed among washed and unwashed sand mixes, though
overall unwashed mixes performed better than washed mixes. All values are
within 1000 coulomb indicating excellent chloride resisting capabilities with
concrete containing very high replacement of GGBS or Fly ash. GGBS concrete
containing 100% slag and unwashed sand is having RCPT value of 257 coulombs
which is the lowest. Most of the RCPT values for GGBS mixes are below 500
coulombs. For Fly ash mixes, the RCPT values hover around 600 coulombs.
Water Permeability Test: Water permeability test according to DIN 1048 method
shows no sign of water penetration for Fly Ash mixes. No graph could be plotted
due to zero penetration of water through the specimen. For GGBS mixes except
the two control mixes (30% Portland cement + 70% GGBS, washed and
- 314 -
unwashed), all other results are below 10mm, indicating excellent water-tightness
of the concrete. There is no significance of the influence of unwashed/washed
sand have been observed.
Water Absorption (BS 1881: Part 122): Fly ash mixes are having better water
absorption values mostly less than 1% compared to GGBS mixes which stay
between 1% and 1.2%. Like Permeability test, there is no influence of the
presence of filler provided by unwashed sand.
In general it has been seen that higher replacement of portland cement gives better
durability values with no significance in durability if unwashed sand is used.
Tables 15 & 16 and Figs. 15, 16, 17, 18 & 19 present the durability values of both
GGBS and Fly ash mixes.
ENVIRONMENTAL BENEFITS
Results from the concrete trials using high volume GGBS and Fly ash show that
structural grade concrete where very high strength is not a requirement can be
produced readily with no or very little requirement of ordinary portland cement.
Enhanced durability performance would effectively increase the service life of
structures providing opportunities to reduce the demand of raw materials.
Producing better concrete in terms of strength and durability using unwashed
crushed sand raises questions on the validity of the requirement of ASTM C33 on
the limitation on maximum fines. A huge amount of energy could be saved and
fewer or no waste would be generated as the filler materials would be turned into
useful concrete ingredient.
Use of SCC made of higher alternative cementitious materials and unwashed CRF
as sand would reduce the net carbon footprint of projects.
FUTURE WORKS
The strength gain of concrete without or minimum portland cement can be further
enhanced should the w/c can be reduced further for a given consistency. Concrete
with 0.34 w/c or less need to be investigated to produce concrete with higher
grade.
The scope of physical tests should also be widened. Tensile and flexural strength,
E-modulous and shrinkage of concrete should be investigated.
- 315 -
Actual chloride diffusion tests beside indirect test (RCPT) should be included.
Change in length in sulphate solution and other aggressive elements such as sea
water should also be investigated.
Effect of Microsilica should be re-examined by incorporating undesified
Microsilica for lab mixes. Other pozzolanic materials such as reactive rice husk
ash could be of interest.
Effect of recycled aggregate with high volume alternative slag or other pozzolanic
materials should also be investigated.
Actual carbon footprint needs to be calculated to measure the net gain in terms of
the environmental sustainability.
CONCLUSIONS
Two sets of self compacting concrete trials have been conducted using GGBS and
Fly ash as replacement of portland cement for each. For GGBS mixes the
maximum replacement was 100% and minimum 70% and for Fly ash mixes it
was 80% and 30% respectively. A new proprietary admixture from Chryso has
been used to boost the hydration of the cementitious materials.
Two types of CRF sand have been used for each set. Washed CRF to remove
excessive fine materials to meet ASTM C33 requirement with less than 5% fines
and unwashed limestone CRF with approximately 10% fines smaller than 75
micron.
Microsilica has been added additionally to boost the strength gain, which was not
materialized perhaps due to the usage of densified Microsilica.
Self compacting concrete with an average flow of more than 700mm and L-Box
passing ability index of more than 0.90 have been achieved even without any
presence of portland cement.
Compressive strength of 30 MPa has been achieved for concretes with 100%
GGBS + 0% Portland cement and 80% Fly ash + 20% Portland cement mixes.
Durability test results of RCPT (ASTM 1202), Water Permeability (DIN 1048)
and Water Absorption (BS1881:Part 122) positively indicate that concrete
containing little or no portland cement with higher GGBS or Fly Ash provide
better durability performance.
It is possible to produce structurally sound self compacting concrete without using
portland cement. The compressive strength achieved is 30MPa with much
improved durability performance without removing important additional filler
- 316 -
from the crushed fine aggregates. This enhances the environmental credential of
these concrete tremendously.
REFERENCES
1.
2.
- 317 -
Mix
No.
Concrete Grade
Cementitious
Materials, kg/m3
OPC GGBS MS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
UW(30%OPC+
70%GGBS)
UW(20%OPC+
80%GGBS)
UW(10%OPC+
90%GGBS)
UW(0%OPC+
100%GGBS)
UW(20%OPC+
75%GGBS+5%MS)
UW(10%OPC+
85%GGBS+5%MS)
UW(95%GGBS+
5%SF)
W(30%OPC+
70%GGBS)
W(20%OPC+
80%GGBS)
W(10%OPC+
90%GGBS)
W(0%OPC+
100%GGBS)
W(20%OPC+
75%GGBS+5%MS)
W(10%OPC+
85%GGBS+5%MS)
W(95%GGBS+5%
MS
Coarse
Fine
Admixture
Aggregate, Aggregate, Water
(Chryso)
kg/m3
kg/m3
20
5 Dune
Dosage,
10 mm
kg/m3
mm
mm Sand
L/m3
120
280
241
565
530
530
152
6.0
80
320
240
563
528
528
152
5.0
40
360
239
562
526
526
152
5.0
400
239
560
525
525
152
4.5
80
300
20
239
561
526
526
152
4.0
40
340
20
238
560
525
525
152
4.0
380
20
238
559
524
524
152
4.8
120
280
241
565
530
530
152
4.2
80
320
240
563
528
528
152
3.9
40
360
239
562
526
526
152
3.0
400
239
560
525
525
152
3.8
80
300
20
239
561
526
526
152
4.0
40
340
20
238
560
525
525
152
380
20
238
559
524
524
152
4.5
- 318 -
UW(60%OPC+ 40%FA)
240
160
238
559
524
524
152
4.0
UW(50%OPC+ 50%FA)
200
200
236
555
520
520
152
2.72
UW(40%OPC+ 60%FA)
160
240
235
551
517
517
152
3.2
UW(30%OPC+ 70%FA)
120
280
233
547
513
513
152
2.5
UW(20%OPC+ 80%FA)
UW(20%OPC+
7
75%FA+5%MS)
UW(30%OPC+
8
65%FA+5%MS)
UW(40%OPC+
9
55%FA+5%MS)
UW(50%OPC+
10
45%FA+5%MS)
UW(60%OPC+
11
35%FA+5%MS)
12 W(70%OPC+ 30%FA)
80
320
231
543
509
509
152
2.5
80
300
20
231
543
509
509
152
2.8
120
260
20
233
547
513
513
152
3.1
160
220
20
234
551
516
516
152
3.4
200
180
20
236
555
520
520
152
3.6
240
140
20
238
558
523
523
152
3.8
280
120
239
563
527
527
152
3.5
13 W(60%OPC+ 40%FA)
240
160
238
559
524
524
152
3.3
14 W(50%OPC+ 50%FA)
200
200
236
555
520
520
152
2.8
15 W(40%OPC+ 60%FA)
160
240
235
551
517
517
152
2.1
16 W(30%OPC+ 70%FA)
120
280
233
547
513
513
152
1.8
17 W(20%OPC+ 80%FA)
W(20%OPC+
18
75%FA+5%MS)
W(30%OPC+
19
65%FA+5%MS)
W(40%OPC+
20
55%FA+5%MS)
W(50%OPC+
21
45%FA+5%MS)
W(60%OPC+
22
35%FA+5%MS)
80
320
231
543
509
509
152
1.8
80
300
20
231
543
509
509
152
2.5
120
260
20
233
547
513
513
152
160
220
20
234
551
516
516
152
200
180
20
236
555
520
520
152
240
140
20
238
558
523
523
152
4.6
- 319 -
SiO2
Al2O3
Fe2O3
CaO
MgO
SO3
K2O
Na2O
TiO2
Mn2O3
Cl
IR
LOI
C3 A
AE
m2/kg
H2O
Portland
Cement
20.12
4.84
3.53
63.44
1.36
2.95
0.51
0.20
0.02
0.48
2.84
6.86
0.54
334
Fly
Ash
57.66
30.35
4.92
1.71
0.70
0.16
1.14
0.01
1.64
0.04
0.01
0.63
0.97
GGBS
31.6
16.58
1.28
37.20
10.34
0.24
0.35
0.24
0.48
0.01
0.60
1.0
Microsilica
90.53
2.38
0.76
444
0.3
0.03
0.87
37.50
20.00
14.00
10.00
5.00
0.075
Pan
419.4
2026.6
1200.5
105.1
6.9
1.5
419.4
11.2
88.8
2446
65.1
34.9
3646.5
97
3
3751.4
99.8
0.2
3758.5
3760
- 320 -
20.00
14.00 10.00
385.9
385.9
17.6
82.4
5.00
1665.6
2051.5
93.5
6.5
2.36
121.5
2173
99.0
1.0
0.075
9.1
2182.1
99.4
0.6
Pan
12.9
2195
5.00
59.1
59.1
4.4
95.6
2.36 1.18
0.60
0.30 0.150 0.075
490.8 323.4 160.1 87.6
44.4
59.3
549.9 873.3 1033.4 1121 1165.4 1224.7
40.8 64.8
76.7
83.2
86.5
90.9
59.2 35.2
23.3
16.8
13.5
9.1
Pan
122.6
1347.3
5.00
58.5
58.5
5
95
2.36
454
512.5
43.5
56.5
1.18
280.6
793.1
67.3
32.7
0.60
155.9
949
80.6
19.4
0.30
81.3
1030.3
87.4
12.6
0.150
55.1
1085.4
92.2
7.8
0.075
42.3
1127.7
95.7
4.3
0.60
0
0
0
100
0.30
4.8
4.8
0.8
99.2
- 321 -
0.150
296.9
301.7
53.1
46.9
0.075
262
563.7
99.2
0.8
Pan
Pan
20 mm
100
88.8
0.2
0
0
10 mm
100
100
6.5
1
0
5 mm
100
100
95
23.3
13.5
D/SAND COMBINED
100
100
100
99
100
57
100
35
46.9
17
20 mm
100
88.8
0.2
0
0
10 mm
100
100
6.5
1
0
5 mm
100
100
95
19.4
7.8
- 322 -
D/SAND COMBINED
100
100
100
99
100
57
100
34
46.9
16
Table 11: Slump flow retention and L-box test results GGBS mixes
Mix
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
T0
725
710
745
735
T30
715
680
730
725
T60
700
650
710
680
T90
620
610
665
610
LBox
h1/h2
0.82
0.80
0.89
0.84
760
740
700
680
0.9
730
720
700
680
0.88
750
750
710
720
745
730
740
700
700
720
660
660
680
680
680
600
640
650
650
620
0.87
0.86
0.89
0.88
0.85
720
700
660
610
0.9
730
705
675
610
0.88
735
710
665
620
0.9
Slump Flow, mm
Concrete Grade
UW(30%OPC+ 70%GGBS)
UW(20%OPC+ 80%GGBS)
UW(10%OPC+ 90%GGBS)
UW(0%OPC+ 100%GGBS)
UW(20%OPC+
75%GGBS+5%MS)
UW(10%OPC+
85%GGBS+5%MS)
UW(95%GGBS+5%MS)
W(30%OPC+ 70%GGBS)
W(20%OPC+ 80%GGBS)
W(10%OPC+ 90%GGBS)
W(0%OPC+ 100%GGBS)
W(20%OPC+
75%GGBS+5%MS)
W(10%OPC+
85%GGBS+5%MS)
W(95%GGBS+5%MS)
- 323 -
Table 12: Slump flow retention and L-box test results Fly ash mix
Mix
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Concrete Grade
UW(70%OPC+ 30%FA)
UW(60%OPC+ 40%FA)
UW(50%OPC+ 50%FA)
UW(40%OPC+ 60%FA)
UW(30%OPC+ 70%FA)
UW(20%OPC+ 80%FA)
UW(20%OPC+
75%FA+5%MS)
UW(30%OPC+
65%FA+5%MS)
UW(40%OPC+
55%FA+5%MS)
UW(50%OPC+
45%FA+5%MS)
UW(60%OPC+
35%FA+5%MS)
W(70%OPC+ 30%FA)
W(60%OPC+ 40%FA)
W(50%OPC+ 50%FA)
W(40%OPC+ 60%FA)
W(30%OPC+ 70%FA)
W(20%OPC+ 80%FA)
W(20%OPC+
75%FA+5%MS)
W(30%OPC+
65%FA+5%MS)
W(40%OPC+
55%PFA+5%MS)
W(50%OPC+
45%FA+5%MS)
W(60%OPC+
35%FA+5%MS)
T0
720
740
760
750
750
760
Slump Flow, mm
T30
T60
720
710
740
735
745
730
740
740
745
730
760
760
T90
700
730
710
720
710
745
L-Box
h1/h2
0.9
0.9
0.9
0.9
0.9
0.86
740
730
720
700
0.98
750
730
725
700
0.99
740
715
700
690
0.97
750
720
695
690
0.92
730
700
690
670
0.98
710
770
770
790
770
750
710
760
760
770
770
730
695
750
745
740
750
710
690
730
715
730
730
670
0.99
0.99
0.99
0.98
0.98
0.99
750
750
730
710
0.98
750
740
710
690
0.99
730
705
680
665
0.98
720
715
710
700
0.97
715
710
700
660
0.98
- 324 -
Concrete Grade
UW(30%OPC+ 70%GGBS)
UW(20%OPC+ 80%GGBS)
UW(10%OPC+ 90%GGBS)
UW(0%OPC+ 100%GGBS)
UW(20%OPC+ 75%GGBS+5%MS)
UW(10%OPC+ 85%GGBS+5%MS)
UW(95%GGBS+5%MS)
W(30%OPC+ 70%GGBS)
W(20%OPC+ 80%GGBS)
W(10%OPC+ 90%GGBS)
W(0%OPC+ 100%GGBS)
W(20%OPC+ 75%GGBS+5%MS)
W(10%OPC+ 85%GGBS+5%MS)
W(95%GGBS+5%MS)
- 325 -
Compressive Strength,
MPa
3d
7d
28d
32
48.0
58.5
27
43.0
56.0
22
39.5
47.0
16
22.5
30.0
27
40.5
54.0
22
36.5
43.0
15
19.5
26.5
28
43.0
56.0
28
45.0
54.5
24
31.0
46.0
16
19.0
25.5
25
34.5
47.0
20
34.0
39.0
16
18.0
23.5
UW(70%OPC+ 30%FA)
UW(60%OPC+ 40%FA)
UW(50%OPC+ 50%FA)
UW(40%OPC+ 60%FA)
UW(30%OPC+ 70%FA)
UW(20%OPC+ 80%FA)
UW(20%OPC+ 75%FA+5%MS)
UW(30%OPC+ 65%FA+5%MS)
UW(40%OPC+ 55%FA+5%MS)
UW(50%OPC+ 45%FA+5%MS)
UW(60%OPC+ 35%FA+5%MS)
W(70%OPC+ 30%FA)
W(60%OPC+ 40%FA)
W(50%OPC+ 50%FA)
W(40%OPC+ 60%FA)
W(30%OPC+ 70%FA)
W(20%OPC+ 80%FA)
W(20%OPC+ 75%FA+5%MS)
W(30%OPC+ 65%FA+5%MS)
W(40%OPC+ 55%PFA+5%MS)
W(50%OPC+ 45%FA+5%MS)
W(60%OPC+ 35%FA+5%MS)
- 326 -
Compressive Strength,
MPa
3
7
28
32.5
42.0
58.0
32.5
40.0
53.0
27.0
35.5
54.5
22.0
29.0
49.0
14.0
19.5
38.5
11.0
16.0
30.0
13.5
19.0
28.5
16.5
21.0
38.0
20.5
27.5
42.0
19.0
30.5
45.0
27.0
40.0
51.0
27
37.0
53.0
27.5
35.5
51.0
26
33.0
49.0
21.0
23.0
42.0
15.0
19.2
33.0
7.0
11.0
20.5
9.5
12.0
24.0
17
21.0
37.5
16.5
24.0
40.0
22.5
31.0
44.0
24.0
36.0
48.0
Concrete Grade
UW(30%OPC+
70%GGBS)
UW(20%OPC+
80%GGBS)
UW(10%OPC+
90%GGBS)
UW(0%OPC+
100%GGBS)
W(30%OPC+
70%GGBS)
W(20%OPC+
80%GGBS)
W(10%OPC+
90%GGBS)
W(0%OPC+
100%GGBS)
Water
Water
Permeability Absorption
Coulomb
mm
%
RCPT
377
15
1.1%
315
10
1.2%
513
0.9%
257
3.5
1.0%
297
15
1.0%
442
0.7%
560
1.2%
355
1.2%
Concrete Grade
UW(70%OPC+ 30%FA)
UW(60%OPC+ 40%FA)
UW(50%OPC+ 50%FA)
UW(40%OPC+ 60%FA)
UW(30%OPC+ 70%FA)
UW(20%OPC+ 80%FA)
W(70%OPC+ 30%FA)
W(60%OPC+ 40%FA)
W(50%OPC+ 50%FA)
W(40%OPC+ 60%FA)
W(30%OPC+ 70%FA)
W(20%OPC+ 80%FA)
RCPT
Coulomb
806
578
542
579
670
502
970
864
574
528
452
516
- 327 -
Water
Permeability
mm
0
0
0
0
0
0
0
0
0
0
0
0
Water
Absorption
%
1
0.2
1
0.8
0.8
0.9
1.1
0.9
0.3
0.6
0.8
1.4
PASSING %
10
100
B.S SIEVE (m m )
PASSING %
10
B.S SIEVE (m m )
- 328 -
100
Flow Retention
GGBS - Unwashed CRF
800
UW(30%OPC+
70%GGBS)
750
UW(20%OPC+
80%GGBS)
UW(10%OPC+
90%GGBS)
mm
700
UW(0%OPC+
100%GGBS)
650
UW(20%OPC+
75%GGBS+5%MS)
600
UW(10%OPC+
85%GGBS+5%MS)
550
T0
T30
T60
T90
UW(95%GGBS+5%
MS)
Time
Flow retention
GGBS - Washed CRF
800
W(30%OPC+
70%GGBS)
750
W(20%OPC+
80%GGBS)
W(10%OPC+
90%GGBS)
700
mm
W(0%OPC+
100%GGBS)
W(20%OPC+
75%GGBS+5%MS)
650
W(10%OPC+
85%GGBS+5%MS)
600
W(95%GGBS+5%
MS)
550
T0
T30
T60
T90
Time
- 329 -
Flow Retention
Fly Ash - Unwashed CRF
780
760
740
mm
720
700
680
660
640
620
600
T0
T30
T60
T90
UW(70%OPC+
30%FA)
UW(60%OPC+
40%FA)
UW(50%OPC+
50%FA)
UW(40%OPC+
60%FA)
UW(30%OPC+
70%FA)
UW(20%OPC+
80%FA)
UW(20%OPC+
75%FA+5%MS)
UW(30%OPC+
65%FA+5%MS)
UW(40%OPC+
55%FA+5%MS)
UW(50%OPC+
45%FA+5%MS)
UW(60%OPC+
35%FA+5%MS)
Time
Fig. 5: Flow retention for Fly ash mixes with unwashed CRF
Flow Retention
Fly Ash - Washed CRF
800
W(60%OPC+ 40%FA)
780
W(50%OPC+ 50%FA)
760
W(40%OPC+ 60%FA)
740
W(30%OPC+ 70%FA)
720
mm
W(70%OPC+ 30%FA)
W(20%OPC+ 80%FA)
700
W(20%OPC+ 75%FA+5%MS)
680
660
W(30%OPC+ 65%FA+5%MS)
640
W(40%OPC+ 55%PFA+5%MS)
620
W(50%OPC+ 45%FA+5%MS)
600
T0
T30
T60
T90
W(60%OPC+ 35%FA+5%MS)
Time
Fig. 6: Flow retention for Fly ash mixes with washed CRF
- 330 -
Compressive Strength
GGBS - Unwashed CRF
With & without MS
UW(30%OPC+
70%GGBS)
60
55
UW(20%OPC+
80%GGBS)
50
45
UW(20%OPC+
75%GGBS+5%
MS)
UW(10%OPC+
90%GGBS)
MPa
40
35
30
UW(10%OPC+
85%GGBS+5%
MS)
UW(0%OPC+
100%GGBS)
25
20
15
UW(95%GGBS+
5%MS)
10
0
14
21
28
Days
60
W(30%OPC+
70%GGBS)
W(20%OPC+
80%GGBS)
50
W(20%OPC+
75%GGBS+5%
MS)
W(10%OPC+
90%GGBS)
mm
40
30
W(10%OPC+
85%GGBS+5%
MS)
W(0%OPC+
100%GGBS)
20
10
W(95%GGBS+5
%MS)
14 Days 21
28
35
- 331 -
70.0
60.0
58.5
W(30%OPC+ 70%GGBS)
47.0 46.0
50.0
MPa
UW(30%OPC+
70%GGBS)
UW(20%OPC+
80%GGBS)
40.0
W(20%OPC+ 80%GGBS)
30.0
30.0
25.5
20.0
UW(10%OPC+
90%GGBS)
W(10%OPC+ 90%GGBS)
10.0
UW(0%OPC+
100%GGBS)
0.0
W(0%OPC+ 100%GGBS)
GGBS
MPa
40.0
UW(20%OPC+
75%GGBS+5%MS)
54.0
47.0
W(20%OPC+
75%GGBS+5%MS)
43.0
39.0
UW(10%OPC+
85%GGBS+5%MS)
26.5
30.0
23.5
W(10%OPC+
85%GGBS+5%MS)
20.0
UW(95%GGBS+5%MS)
10.0
W(95%GGBS+5%MS)
0.0
GGBS
- 332 -
Compressive Strength
Fly Ash - Unwashed CRF
With & Without MS
60
UW(70%OPC+ 30%FA)
55
UW(60%OPC+ 40%FA)
50
UW(60%OPC+
35%FA+5%MS)
UW(50%OPC+ 50%FA)
45
MPa
40
UW(50%OPC+
45%FA+5%MS)
UW(40%OPC+ 60%FA)
35
30
UW(40%OPC+
55%FA+5%MS)
UW(30%OPC+ 70%FA)
25
UW(30%OPC+
65%FA+5%MS)
UW(20%OPC+ 80%FA)
20
15
UW(20%OPC+
75%FA+5%MS)
10
0
14
21
28
Days
Fig. 11: Compressive strength development of Fly ash mixes with unwashed
sand and with & without Microsilica
Compressive Strength
Fly Ash - Washed CRF
With & Without MS
60
W(70%OPC+ 30%FA)
W(60%OPC+ 40%FA)
50
W(60%OPC+
35%FA+5%MS)
W(50%OPC+ 50%FA)
40
MPa
W(50%OPC+
45%FA+5%MS)
W(40%OPC+ 60%FA)
30
W(40%OPC+
55%PFA+5%MS)
W(30%OPC+ 70%FA)
20
W(30%OPC+
65%FA+5%MS)
W(20%OPC+ 80%FA)
10
W(20%OPC+
75%FA+5%MS)
0
0
10
15
20
25
30
Days
Fig. 12: Compressive strength development of Fly ash mixes with unwashed
sand and with & without Microsilica
- 333 -
60.0
UW(70%OPC+ 30%FA)
W(70%OPC+ 30%FA)
58.0
53.0
53.0
54.5
51.0
UW(60%OPC+ 40%FA)
49.0
50.0
49.0
W(60%OPC+ 40%FA)
42.0
UW(50%OPC+ 50%FA)
38.5
MPa
40.0
W(50%OPC+ 50%FA)
33.0
30.0
UW(40%OPC+ 60%FA)
30.0
20.5
20.0
W(40%OPC+ 60%FA)
UW(30%OPC+ 70%FA)
W(30%OPC+ 70%FA)
10.0
UW(20%OPC+ 80%FA)
W(20%OPC+ 80%FA)
0.0
Fly Ash
Fig. 13: Compressive strength at 28 days: unwashed vs. washed Fly ash
without Microsilica
48.0
45.0
UW(50%OPC+ 45%FA+5%MS)
44.0
42.0
40.0
40.0
38.0
W(50%OPC+ 45%FA+5%MS)
37.5
MPa
UW(40%OPC+ 55%FA+5%MS)
28.5
30.0
W(40%OPC+ 55%PFA+5%MS)
24.0
UW(30%OPC+ 65%FA+5%MS)
20.0
W(30%OPC+ 65%FA+5%MS)
10.0
UW(20%OPC+ 75%FA+5%MS)
W(20%OPC+ 75%FA+5%MS)
0.0
Fly Ash
Fig. 14: Compressive strength at 28 days: unwashed vs. washed Fly ash with
Microsilica
- 334 -
600
513
Coulomb
500
400
UW(30%OPC+ 70%GGBS)
442
377
355
297
315
UW(20%OPC+ 80%GGBS)
257
300
W(30%OPC+ 70%GGBS)
W(20%OPC+ 80%GGBS)
200
UW(10%OPC+ 90%GGBS)
100
W(10%OPC+ 90%GGBS)
UW(0%OPC+ 100%GGBS)
W(0%OPC+ 100%GGBS)
GGBS
Fig. 15: RCPT comparison of GGBS mixes for unwashed vs. washed sand
15
15
14
12
UW(30%OPC+ 70%GGBS)
10
W(30%OPC+ 70%GGBS)
mm
10
8
UW(20%OPC+ 80%GGBS)
6
3.5
4
2
W(20%OPC+ 80%GGBS)
UW(10%OPC+ 90%GGBS)
W(10%OPC+ 90%GGBS)
UW(0%OPC+ 100%GGBS)
W(0%OPC+ 100%GGBS)
GGBS
Fig. 16: Permeability comparison of GGBS mixes for unwashed vs. washed
sand
- 335 -
1.4%
1.2%
1.2%
1.2%
W(30%OPC+ 70%GGBS)
1.1%
1.2%
1.0%
1.0%
0.9%
1.0%
UW(20%OPC+ 80%GGBS)
0.7%
W(20%OPC+ 80%GGBS)
0.8%
0.6%
UW(10%OPC+ 90%GGBS)
0.4%
W(10%OPC+ 90%GGBS)
0.2%
UW(0%OPC+ 100%GGBS)
0.0%
W(0%OPC+ 100%GGBS)
GGBS
Fig. 17: Water absorption comparison of GGBS mixes for unwashed vs.
washed sand
RCPT ASTM 1202
Unwashed vs Washed
1200
UW(70%OPC+ 30%FA)
W(70%OPC+ 30%FA)
970
1000
UW(60%OPC+ 40%FA)
864
806
W(60%OPC+ 40%FA)
Coulomb
800
UW(50%OPC+ 50%FA)
670
600
578
542
574
579
W(50%OPC+ 50%FA)
528
452
502
516
UW(40%OPC+ 60%FA)
W(40%OPC+ 60%FA)
400
UW(30%OPC+ 70%FA)
W(30%OPC+ 70%FA)
200
UW(20%OPC+ 80%FA)
W(20%OPC+ 80%FA)
0
Fly Ash
Fig. 18: RCPT comparison of Fly Ash mixes for unwashed vs. washed sand
- 336 -
1.4
1.4
1.2
W(70%OPC+ 30%FA)
UW(60%OPC+ 40%FA)
1.1
1
0.9
0.8
0.8
0.8
0.6
0.6
0.4
UW(70%OPC+ 30%FA)
0.2
0.8
0.9
W(60%OPC+ 40%FA)
UW(50%OPC+ 50%FA)
W(50%OPC+ 50%FA)
UW(40%OPC+ 60%FA)
0.3
W(40%OPC+ 60%FA)
0.2
UW(30%OPC+ 70%FA)
W(30%OPC+ 70%FA)
UW(20%OPC+ 80%FA)
Fly Ash
W(20%OPC+ 80%FA)
Fig. 19: Water absorption comparison of GGBS mixes for unwashed vs.
washed sand
- 337 -
- 338 -
- 339 -
INTRODUCTION
General
The main intention by the Eco-SCC project is to develop concrete that contains
less cement (up to one third less) than the concrete traditionally used in the
industry today, without loss in quality. The production of 1 metric ton of cement
releases almost 1 ton of CO2 to the atmosphere. Thus a significant reduction in
release of CO2 to the atmosphere (due to cement production) will be attained.
Additionally, the concrete is to maintain the properties of normal SCC[1-4] and its
superiority to conventional vibrated concrete (CVC). Among other benefits is less
material cost, less shrinkage in concrete and thereby reduced risk for crackformation (due to reduced paste volume in the concrete). The approach used is
new, based on rheology. Other known methods, like replacing cement with fly ash
or blast furnace slag have been used.
In short[14], Eco-SCC is an economical and environmentally advantageous
alternative to conventional vibrated concrete, in which the total powder content
(cement, fly ash, silica fume, lime stone filler) is 315 kg/m3 or below.
Eco-SCC compared to conventional SCC
Eco-SCC has very low content of cementitious material compared to traditional
SCC. For instance it is common to use about 600 kg/m3 of cementitious materials
in SCC in Germany (and most of Asia) whereas it is as low as 360 kg/m 3 in
Denmark. Danish SCC is generally stiff (according to our definition it is a highyield SCC), but on the other hand it has acquired some 30% market share in
Ready-Mix concrete.
As Eco-SCC has so low cement content its rheology is quite different from other
SCCs, in particular as it has very low plastic viscosity. This makes Eco-SCC very
easy to place; on the other hand it is difficult to ensure stability (in respect of
segregation) with such a low viscosity.
In a survey by Domone[3] on sixty-eight case studies of applications of selfcompacting concrete (SCC) the average powder content was about 500 kg/m3
which is well over 50% higher than in conventional vibrated concrete, see figure
1. It is noteworthy that relative to distribution, the mixes containing stabilizers
(VMA) contains only 20 kg/m3 less powder than mixes without stabilizers, see
Figure 1a.
- 340 -
According to figure 1 the lowest content of powder in SCC is about 450 kg/m 3.
The Eco-SCC has much less or only 2/3 of that. The ICI Rheocenter
classification[14] of SCC relative to its powder content is as follow:
Rich SCC
550
kg/m3
415 35 kg/m3
Lean SCC
Green SCC
350 35 kg/m
Eco-SCC
315
kg/m3
up to 315 kg/m3
Actually, the definition should be based on clinker content or CO2 per produced
ton of cement rather than powder content, but it is more complicated (also as the
clinker content in cement is so variable). In general SCC that has less than 380
kg/m3 of powder should be considered green alternative, in particular if or more
of it is not based on clinker and thus the Eco-SCC is very green.
The aggregates is the key
When the cement content in SCC is reduced the concrete is apt to become
unstable. In Eco-SCC the stability is accomplished by employing aggregate
grading which enhances the so-called particle-lattice effect. This grading is
characterized by high fines content, although preferably without colloidal
particles. Simply put, this is continuous grading with relatively high content of
mid size aggregate and relatively low content of stone near the maximal aggregate
size. The maximum aggregate size is generally 16 mm but can be reduced to 12
mm if necessary to obtain stable mix in respect of segregation, alternatively
enlarged to 22 mm in case of compliant aggregates.
In figure 2 are given particle size distributions of typical Eco-SCC in use, X-crete
(which is a SCC with 220 kg/m3 of cement) as well as theoretical distribution for
Eco-SCC with 16 and 22mm maximum aggregate size.
The principle of the particle-lattice concept (see Figure 2) implies that the content
of particles by volume in a given particle size range shall always be alike or
higher than the content in the corresponding upper adjacent range.
The theoretical grading would be a straight line across the diagram in figure 3 at
about 11%. The added powder shall be included (except for pure Portland cement
or the clinker in the cement) in that volume. As mentioned the aggregates shall
preferably not contain a significant amount of colloidal particles. If needed, these
can be removed from the sand/aggregates by simple floating process.
- 341 -
Naturally, it is not only the particle size distribution of the aggregates which is of
importance, but also their shape and surface properties, but they will not be
commented here.
In addition the Eco-SCC may also contain stabilizers (VMA) to increase its
stability and to reduce bleeding. Silica fume as powder added to the mix is
recommended as it adds volume to the binder and is at least as effective as most
chemical stabilizers.
A general comment regarding the aggregates; one can surely use worse grading
or quality but that will necessitate higher cement content. So the choice between
suitable grading or more cement can be a question of economy.
THE RHEOLOGY WAS VITAL PART OF THE DEVELOPEMENT
The flow characteristics
Fresh concrete requires application of force (or shear stress) in order to flow. In
other words, it has a certain resistance to flowing. The force required to initiate
flow is called the yield value (0), see figure 4a. Increasing the rate of flow (rate
of shear) causes the resistance (shear stress) to increase, as shown in figure 4b.
The slope () of the line is an expression of the plastic viscosity of the mass.
- 342 -
- 343 -
- 344 -
FINAL REMARKS
The key to make the Eco-SCC is the particle-lattice effect of aggregates. That
effect depends upon grading where the coarser particles are suspended in blend of
progressively smaller particles thus reducing the driving force of segregation.
The grading is characterized by relatively continuous grading, rich in mid size and
contains relatively low amount of stone near the maximum size. It is also
beneficial to use high fines content. Theoretically, the volume of aggregate in
given particle size range shall be alike or higher than the content in the
corresponding upper adjacent range.
As the total powder content is so very low (cement + fly ash + silica fume +
limestone filler is totally 315 kg/m3 or less) and concrete is so low viscous, the
Eco-SCC is environmentally friendly both in respect to greenhouse gasses and
hard labor work.
In addition, the concrete may contain stabilizers (VMA) or silica fume to increase
its stability in respect to segregation and to reduce bleeding.
ACKNOWLEDGEMENT
The authors gratefully acknowledges the support of (in alphabetical order):
CEMEX Research Group AG (Swiss), Icelandic Housing Fund, Landsvirkjun
(Iceland), MEST Steypustin (Iceland), Norstone (Norway) and Icelandic
Research Foundation Rannis for the project Eco-SCC. In addition, the
colleagues at the Institute and in particular S. Oesterheld are acknowledged for
their contributions to the project.
REFERENCES
1. Wallevik. O., Nielsson, I., Self Compacting Concrete A Rheological
Approach, International Workshop on SCC, Kochi, Japan {JSCE Concrete
Engineering Series no. 30}, August, 1998, pp. 136-159.
2. Wallevik, O.H., Rheology A Scientific Approach to Develop SelfCompacting Concrete, paper submitted to 3rd Int. Symp. on SCC, Rilem,
Reykjavik, Iceland, August, 2003, p. 10.
3. Nielsson, I., Wallevik, O.H., Mix Design of HS-SCC and Practical
Application, paper submitted to 3rd Int. Symp. on SCC, Rilem, Reykjavik,
Iceland, August, 2003, p. 8.
- 345 -
- 346 -
[kg/m3]
Silica fume
Fly ash
Limestone filler
Sum powder
Water
Glenium
51
PC
Gl.
stream
Stabilizer
[kg/m3]
[kg/m3]
GER
SWE
DK
NO
ECO
DK rapid
SWE bygg
DK rapid
NO standard
NO-FA
315
330
315
11
27
330
27
235*
353
150
1,5
357
175
2,6
275
[kg/m3]
[kg/m3]
[kg/m3]
[kg/m3]
[kg/m3]
185
515
180
3,5
590
165
5,3
0,34
50*
285*
185
2,9
3,42
Eco-SCC
X-crete
Theory-16
Theory-22
100
80
60
40
20
0
0,063
0,125
0,250
0,500
1,000
2,000
4,000
- 347 -
40
CVC
Theoretical Eco-SCC
30
Practical Eco-SCC
20
10
0,51
12
24
48
816
1632
1.0
16
32
c)
b)
- 348 -
160
120
80
550 mm
40
650 mm
700 mm
0
0
30
60
90
120
- 349 -
Figure 7- Eco-SCC is easy to place due to its very low plastic viscosity
(Lj on the book means poem)
40%
Paste volume (%)
40%
35%
30%
30%
20%
10%
0%
25%
NO DK Eco SE
NO DK Eco SE
Start
3 sec. after start
Figure 10- At start of the slump-flow measurements on four of the five mixes
(left) and the slump-flow three seconds later (right). The Eco-SCC (upper left
corners) is fairly advanced in three seconds [13]
- 350 -
- 351 -
2,5
.
2,0
120
G-Yield (A)
160
80
40
1,5
1,0
0,5
0,0
0
0
40
80
120
160
Plastic viscosity (Pa s)
200
8
12
16
H-Viscosity (A s)
20
24
2,0
120
G-Yield (A)
2,5
160
1,5
80
1,0
40
0,5
0
400
500
600
700
800
Slum p flow (m m)
0,0
400
900
500
600
700
800
Slum p flow (m m )
900
80
1,2
60
Shrinkage [mm/m]
28 d
2d
40
20
0
0,4
0,5
0,6
w/b-ratio [%]
0,7
+ 10 l
+5
0,8
Ref
-5l
- 10 l
0,4
IS-CVC
s_1
0,0
0
56
112
168
Time [d]
Figure 12- Compressive strength and shrinkage data of 5 mixes from the robusttest having water content of 178, 181, 186, 191 and 200 kg/m3 [15]
- 352 -
430m
high
Figure 12- A sketch of the 430m long vertical pipes (penstock) at Krahnkar
hydro-power plant
- 353 -
- 354 -