4. Tóm Tắt Lats Tiếng Anh (Ncs Lvh).Pdf

MINISTRY OF EDUCATION AND TRAINING
HANOI UNIVERSITY OF CIVIL ENGINEERING

Le Viet Hung
Study on the production of high-strength lightweight concrete using
hollow microspheres from fly ash (cenospheres)
Major: Material engineering
Code: 9520309

SUMMARY OF DOCTORAL DISSERTATION

Ha Noi - 2023


The work was completed at: Ha Noi University of Civil Engineering (HUCE)

Academic supervisor:
1. Prof. Dr. Nguyen Van Tuan – HUCE
2. Prof. Dr. Le Trung Thanh – VIBM

Peer reviewer 1: Prof. Dr. Luong Duc Long - VIBM
Peer reviewer 2: Prof. Dr. Nguyen Duy Hieu - HAU
Peer reviewer 3: Prof. Dr. Nguyen Thanh Sang - UTC

The doctoral dissertation will be defended at the level of the University Council of
Dissertation Assessment's meeting at the Hanoi University of Civil Engineering
at ..... hour .....', day ..... month ..... year 2022

The dissertation is available for reference at the libraries as follows:
- National Library of Vietnam;
- Library of Hanoi University of Civil Engineering;

1
INTRODUCTION
1. NECESSARY OF THE STUDY
Research and development of lightweight concrete for load-bearing structures in construction projects
are being carried out in many places around the world. This type of concrete ensures both strength
and durability like conventional concrete while providing various benefits such as reducing the
structural load, decreasing the size of structures, enhancing soundproofing, insulation, earthquake
resistance, fire resistance, easy transportation, construction, installation, and more. The concrete used
for load-bearing structures in construction increasingly demands high strength and long-term
durability. The concrete used for prestressed structures requires higher quality compared to concrete
used for conventional structures, specifically, the compressive strength often requires more than 40
MPa, rapid strength development, and the criteria for impermeability, water absorption, and longterm durability are also higher than those for conventional concrete.
In the past decade, there has been significant research interest in using hollow spherical particles
derived from fly ash (Fly Ash Cenospheres - FAC) for the production of lightweight concrete in
construction. Utilizing FAC as a lightweight material in concrete manufacturing offers several
advantages, such as achieving compressive strength over 40 MPa and low water absorption,
comparable to conventional concrete. This type of lightweight concrete can be classified as highstrength lightweight concrete with many superior characteristics compared to traditional lightweight
aggregate concrete. However, the research and development of high-strength lightweight concrete
using FAC still face limitations worldwide, especially in Vietnam. FAC can be recovered from fly
ash generated by coal-fired thermal power plants in Vietnam, with a recovery rate of 80-85% from
an annual total fly ash production of approximately 17 million tons (in 2021). With an average FAC
content in fly ash ranging from 0.3% to 1.5%, the theoretical amount of recoverable FAC could be
(32,640-163,200) tons per year.
Based on practical requirements and scientific challenges in developing high-strength lightweight
concrete, the chosen research direction is to explore the production of lightweight concrete using
hollow spherical particles derived from fly ash in thermal power plants in Vietnam. With this
objective in mind, the proposed dissertation topic is "Study on the production of high-strength
lightweight concrete using hollow microspheres from fly ash (cenospheres)”.

2. PURPOSE OF THE STUDY
The study focuses on manufacturing high-strength lightweight concrete for load-bearing structures in
construction projects using hollow spherical particles derived from fly ash. The objective is to achieve
a compressive strength greater than 40 MPa and a density not exceeding 2000 kg/m3, based on the
available materials in Vietnam, with a specific emphasis on a density range of 1300-1600 kg/m3.
3. OBJECTIVE AND SCOPE OF THE STUDY
3.1 Objective of the study
The type of high-strength lightweight concrete using hollow spherical particles derived from fly ash
cenospheres (FAC-HSLWC) has a compressive strength greater than 40 MPa and a density not
exceeding 2000 kg/m3, based on the available materials in Vietnam. The research focuses on
investigating the mechanical and physical properties and applications of this type of concrete, with a
specific emphasis on the range of density between 1300-1600 kg/m3.
3.2 Scope of the study
✓ Selection of materials and mix proportions for high-strength lightweight concrete using hollow
spherical particles derived from fly ash (FAC-HSLWC) with a compressive strength greater than
40 MPa and a bulk density not exceeding 2000 kg/m3 based on domestic materials. The main
materials include pozzolanic cement (XM) and mineral admixtures consisting of silica fume (SF)
and finely ground granulated blast furnace slag (GGBFS); Aggregates: natural sand and hollow
spherical particles derived from fly ash (FAC), in combination with superplasticizers and
polypropylene fibers (PP fibers).
✓ Development of a predictive model for the compressive strength of FAC-HSLWC.


2
✓ Development of a method for calculating the mixture proportions of FAC-HSLWC.
✓ Technical properties of FAC-HSLWC: properties of the concrete mix, mechanical and durability
characteristics.
✓ Performance of a reinforced concrete floor slab using FAC-HSLWC.
4. SCIENTIFIC BASIC
✓ The study on production of FAC-HSLWC is based on the theoretical principles of enhancing the

strength and durability of concrete, which include the following principles: optimizing the particle
composition to achieve the highest packing density of the material mixture; enhancing structural
homogeneity by selecting the appropriate maximum aggregate size; increasing the strength of the
cementitious matrix and the transition zone between aggregate and cement paste; improving
flexural/tensile strength and crack resistance through dispersed fiber reinforcement.
✓ The predictive model for compressive strength is built on the relationships between concrete
compressive strength, cement compressive strength, water-to-cement ratio (w/c ratio), and key
factors of mixture proportions. The predictive model for FAC-HSLWC is established using
nonlinear regression functions derived from experimental results.
✓ The mixture design method for FAC-HSLWC is developed based on the optimal composition of
aggregate particle sizes, the optimal ratio of binder to aggregate (binder/aggregate), the formula
for calculating density of concrete based on the replacement ratio of FAC for sand, and the
predictive model for compressive strength based on the key parameters of concrete mixture
proportions obtained from the research.
5. RESEARCH METHODOLOGY
The dissertation utilizes the following research methods:
✓ Theoretical research: Gathering relevant technical literature to synthesize, analyze, and provide a
basis for establishing the research program.
✓ Experimental research: Conducting experiments using both standard and non-standard methods.
Standard methods are primarily performed according to Vietnamese technical standards (TCVN)
and some commonly used international standards. Non-standard methods are commonly applied
in material research, concrete, and concrete structure fields, such as scanning electron microscopy
(SEM), differential thermal analysis/thermogravimetric analysis (DTA/TGA), determination of
compactness of material mixtures, and determination of viscosity of cement mortar mixtures.
✓

✓
✓

✓

6. NEW CONTRIBUTIONS
Developed a predictive model for the compressive strength of FAC-HSLWC, considering key
factors that influence its compressive strength, including the binder compositions (through binder
strength), binder/aggregate ratio, cenospheres to sand ratio, maximum aggregate size (Dmax), and
PP fiber content.
Established a mixture proportion design method for FAC-HSLWC, ensuring the target
compressive strength range of 40-80 MPa and a density of 1300-2000 kg/m3.
Investigated several physical and mechanical properties of FAC-HSLWC applicable to structural
elements, including: (1) workability characteristics of the fresh concrete mixture, (2) mechanical
properties of the hardened concrete (compressive strength, flexural strength, modulus of
elasticity, Poisson's ratio), and (3) durability properties of the concrete (drying shrinkage, water
absorption, chloride ion permeability, resistance to sulfate attack).
Evaluated the performance of structural slabs using FAC-HSLWC in comparison to slabs made
with conventional concrete of the same compressive strength grade.

7. STRUTURE OF THE THESIS
The thesis consists of an Introduction, 6 chapters, General Conclusion and Recommendations, 36
tables, 97 figures, presented within 151 pages excluding the appendix.


3
1

CHAPTER 1: INTRODUCTION OF LIGHTWEIGHT CONCRETE AND
CENOSPHERE-BASED LIGHTWEIGHT CONCRETE

1.1 LITERATURE REVIEW ON STRUCTURAL LIGHTWEIGHT CONCRETE
1.1.1 Concept and classification on lightweight concrete
1.1.2 Situation on research and application of structural lightweight concrete

For structural lightweight concrete used in construction, the minimum specified compressive strength
is usually 17 MPa. In practice, lightweight concrete is commonly used with compressive strengths
ranging from 21 to 35 MPa. High-rise buildings and bridge structures often employ high-strength
lightweight concrete, with compressive strengths typically ranging from 35 to 41 MPa and a density
in the range of 1680-1920 kg/m3.
The lightweight aggregates used in the production of such lightweight concrete are typically artificial
lightweight aggregates made from clay, shale, and expanded shale. Depending on the quality and
density of the lightweight aggregates, the compressive strength of lightweight structural concrete can
range from 20 to 55 MPa, with densities ranging from 1440 to 1920 kg/m3. The corresponding specific
strength would be in the range of 20 to 30 kPa/kg.m-3. One characteristic of expanded lightweight
aggregates is their porous structure, which results in a high water-absorption capacity (typically 1025%). This poses challenges in controlling the workability of the concrete mix and the properties of
the concrete, such as changes in density and volume when exposed to a moist environment.
1.1.3 High-strength lightweight concrete and its application
According to ACI 213-14, high-strength lightweight concrete (HSLWC) is a type of lightweight
structural concrete with a compressive strength greater than 40 MPa
1.1.4 Situation research and application of lightweight concrete in Viet Nam
In Vietnam, there have been studies and applications of common lightweight concretes such as
cellular concrete, lightweight aggregate concrete using keramzit, fly ash, and polystyrene beads.
However, research on using cenospheres for lightweight concrete is a new issue in Vietnam, and
currently, there have been no studies conducted. Cenospheres in Vietnam have the potential for largescale recovery from coal-fired thermal power plants.
1.2 LIGHWEIGHT CONCRETE USING CENOSPHERE
1.2.1 Introduction of lightweight concrete using cenospheres
Fly Ash Cenosphere Lightweight Concrete (FAC LWC) refers to a type of lightweight concrete that
utilizes cementitious material and hollow fly ash cenospheres. This type of lightweight concrete has
a lower density compared to conventional concrete.

Figure 1.1 Cenosphere and typical cenosphere-containing concrete structure
The research on the use of FAC as a lightweight aggregate in cementitious binder systems began in
1984. However, it was not until the late 20th century that the role of FAC as a lightweight aggregate
for low-density, low-strength concrete, primarily fulfilling insulation requirements, was extensively

studied. Recently, several studies have successfully produced a type of lightweight concrete called
Ultra Lightweight Concrete (ULWC) with a density ranging from 1154 to 1471 kg/m3. The
compressive strength at 28 days ranges from 33.0 to 69.4 MPa, and the flexural strength is around 8
MPa. The thermal conductivity coefficient of ULWC typically ranges from 0.3 to 0.8 W/m·K, which
is significantly lower than that of conventional concrete, approximately 1.9 W/m·K.


4
1.2.2 Hollow microsphere from fly ash (Cenosphere)
Cenospheres are hollow spherical particles composed mainly of alumino-silicates, similar to fly ash
particles. Their bulk density typically ranges from 0.4 to 0.9 g/cm3, with particle sizes ranging from
1 to 400 μm. The majority of cenospheres fall within the range of 20 to 300 μm, with wall thicknesses
ranging from 1 to 18 μm. Cenospheres possess good compressive strength and high resistance to gas
and water permeability. Therefore, they are suitable for use in lightweight concrete to enhance
strength and reduce bulk density. The cenosphere content in fly ash is approximately 0.3 to 1.5%, and
considering an estimated annual fly ash generation of around 17 million tons, the theoretical
recoverable amount of cenospheres would be between 32,640 to 163,200 tons per year. The chemical
composition and mineralogy of cenospheres are similar to those of fly ash particles. FAC particles
contain amorphous silica minerals, which have the ability to react pozzolanically and contribute to
the formation of calcium-silicate-hydrate (C-S-H) gel in the cementitious binder system. However,
this reactivity is relatively low at room temperature and increases with the curing temperature of the
concrete.
1.2.3 Properties of the concrete using cenosphere
1.2.3.1 Fresh concrete
There have been very few studies determining the properties of FAC concrete mixtures. The
workability of FAC lightweight concrete is typically assessed using the flowability test method for
mortar. The flowability of FAC lightweight concrete mixtures is usually controlled within the range
of 150-220 mm. The air void content in FAC lightweight concrete mixtures is higher compared to
conventional concrete.
1.2.3.2 Concrete Properties

1.2.3.2.1 Density and Compressive Strength
The volume mass of FAC lightweight concrete depends on the FAC content, the ratio of fine
aggregate to cementitious material, and the presence of coarse and fine aggregates. It can range from
1075 to 2000 kg/m3. For FAC lightweight concrete with a volume mass lower than 1600 kg/m3, most
studies do not use any aggregates other than FAC. The current compressive strength and bulk density
properties of FAC lightweight concrete typically range from 30-68 MPa and 40-47 kPa/kg.m-3,
respectively.
1.2.3.2.2 Flexural Strength and Flexural/Tensile Strength: Similar to other lightweight concrete types,
FAC lightweight concrete exhibits relatively low flexural and tensile strength compared to its
compressive strength (indicating the brittle nature of the concrete). Therefore, studies on FAC
lightweight concrete often incorporate fiber reinforcement such as PVA, PE, or PP to improve its
flexural resistance. The flexural strength of FAC concrete, when combined with fiber reinforcement
in some studies, generally falls within the range of 5-8 MPa with a volume mass of 1200-1900 kg/m3.
1.2.3.2.3 Elastic Modulus: The elastic modulus of concrete primarily depends on its compressive
strength and volume mass. Due to its low density, similar to other lightweight concretes, FAC
concrete has a lower elastic modulus compared to its compressive strength and decreases
proportionally with its volume mass. With compressive strengths ranging from 33-69.4 MPa, the
elastic modulus of FAC lightweight concrete is typically between 10.4-17.0 GPa.
1.2.3.2.4 Durability: The water resistance of FAC lightweight concrete has not been extensively
studied. Research has shown that FAC particles have a water absorption capacity 18 times greater
than sand, but the difference in water absorption is not significant while the water permeability rate
is higher than that of conventional concrete. The ability to resist water, liquids, and gases is an
important factor related to the durability of concrete in aggressive environments. Therefore, in-depth
research is needed to clarify these characteristics regarding FAC lightweight concrete.
1.2.3.2.5 Shrinkage: To date, there have been very few studies on the shrinkage of FAC lightweight
concrete. However, since FAC lightweight concrete is a cementitious binder system with a high
cement content and does not use a dense reinforcement framework like conventional concrete, its
shrinkage is expected to be greater than that of conventional concrete.

5
2

CHAPTER 2: SCIENTIFIC BASIC OF MATERIAL SELECTION, COMPRESSIVE
STRENGTH MODELING, AND MIX DESIGN FOR FAC-HSLWC

Unlike conventional concrete, including commonly used lightweight aggregate concrete such as
expanded clay aggregate (keramzit), coarse aggregates are not typically used in high-strength
lightweight concrete using FAC (FAC-HSLWC) to ensure low density as well as to reduce the risk
of segregation. These characteristics lead to several challenges in the mix design of FAC-HSLWC:
1) Increased surface area leading to higher water demand and W/B ratio: Incorporating a large amount
of FAC hollow particles into the concrete mixture to achieve the desired density results in a significant
increase in the interfacial contact area between phases in the system. This reduces the workability of
the concrete mixture, leading to compromised product quality. Moreover, the higher water absorption
capacity of FAC compared to sand also contributes to an increase in the water demand.
2) Weak transition zone and poor adhesion between cement paste and FAC particles: FAC particles
have a rough surface texture, resulting in a low bond strength between the particle surface and the
cement paste. This diminishes the bond strength within the concrete matrix, affecting its strength and
durability.
3) Brittle behavior and significant shrinkage of FAC-HSLWC: To ensure the compressive strength
and long-term durability of the concrete, a high cement content and low water-to-cement ratio are
necessary. However, these conditions increase the likelihood of generating internal stresses that
exceed the tensile capacity of the concrete, leading to cracking. Additionally, the lack of a solid
reinforcing framework in FAC-HSLWC, as found in conventional concrete, and the high cement
content contribute to its brittle behavior.
4) Lack of a unified mix design method for FAC concrete: Due to the absence of coarse aggregates
in FAC-HSLWC, common mix design methods for lightweight concrete, such as ACI 211.2 or
CEB/FIP, cannot be directly applied. Developing a mix design method specifically for FAC-HSLWC
requires identifying the factors that influence the concrete's properties, with strength and density
being the fundamental characteristics. The following sections present the scientific foundations to

address the aforementioned challenges.
2.1 SCIENTIFIC BASIS FOR MATERIAL SELECTION IN FAC-HSLWC
2.1.1 Scientific Basis for Aggregate Selection in FAC-HSLWC
To mitigate segregation in concrete mixtures, several principles have been identified: (1) increasing
the packing density of aggregate mixtures reduces segregation; (2) continuous gradation of aggregates
leads to less segregation compared to discontinuous gradation; (3) reducing the maximum size of
aggregates reduces segregation compared to using aggregates with the same particle size distribution
but larger Dmax; (4) increasing the proportion of fine particles in the mixture decreases the degree of
segregation; (5) minimizing the difference in thermal conductivity of the aggregates reduces
segregation. Based on these principles, the aggregates for FAC-HSLWC are selected to have a high
compaction factor, small Dmax, increased proportion of fine particles, and materials with minimal
variation in thermal conductivity. The FAC particles primarily have sizes ranging from 45-250 μm.
Therefore, in this study, the aggregates for FAC-HSLWC are chosen to include FAC combined with
natural sand aggregates with a maximum particle size of 5.0 mm to ensure a continuous particle size
distribution and limit segregation.
2.1.2 Scientific Basis for Using Pozzolanic Materials in FAC-HSLWC
Pozzolanic materials are used in this study to improve adhesion and enhance the strength
characteristics of the interfacial transition zone (ITZ) between the FAC particles and the cement paste
in FAC-HSLWC. In this study, pozzolanic materials, specifically silica fume (SF) and ground
granulated blast furnace slag (GGBFS), with particle sizes ranging from fine to ultrafine, are oriented
in the CKD component of FAC-HSLWC. These pozzolanic particles, along with cement, can fill the
voids created by larger-sized particles such as sand and FAC particles with an average particle size
of approximately 100-120 μm, thereby creating a dense structure for the concrete.


6
2.1.3 Scientific Basis for Using Polypropylene Fiber
Polypropylene (PP) fibers are commonly used in concrete. Concrete reinforced with PP fibers is
known to improve crack resistance by controlling crack propagation within the concrete structure. PP
fibers in concrete act as bridging elements for cracks formed under load, thus impeding crack

development. Additionally, the use of PP fibers has been shown to be an effective measure in reducing
concrete shrinkage.
2.2 SCIENTIFIC BASIS FOR BUILDING A COMPRESSIVE STRENGTH PREDICTION
MODEL FOR FAC-HSLWC
2.2.1 Some Models for Predicting Concrete Strength
Several models for predicting the compressive strength of concrete have been developed. One of the
earliest models is Feret's model (1892). This concrete strength model has been further developed by
researchers such as Abrams (1919), Bolomey (1935), De Larrard (1993), and Popovics (1965).
Bolomey (1935) simplified Feret's formula into a linear model:
c
f′c = 24,6 [ − 0,5]
w

(2.1)

The model by De Larrard (1993) takes into account multiple factors influencing concrete strength,
such as the influence of cement paste (through the compressive strength of cement paste, fcp) and
aggregates in concrete (through the maximum paste thickness, MPT):
f′c = fcp × MPT −r

(2.2)

It can be observed that De Larrard's strength prediction model is comprehensive, as it considers the
influence of cement paste (based on the strength and composition of cement paste using Feret's
formula), the volume and Dmax of aggregates (through the MPT parameter). For the FAC-HSLWC
system, when applying De Larrard's compressive strength prediction model, the aggregates will
consist of a mixture of FAC and sand, the Dmax of aggregates will be the Dmax of sand, and CKD
will be a multiple-component system composed of OPC cement and pozzolanic materials. These
factors will affect the coefficients in the concrete strength prediction model.
2.2.2 Some Models for Predicting Lightweight Aggregate Concrete Strength

Several models for predicting the compressive strength of lightweight aggregate concrete (LWAC)
have been proposed, such as the CEB/FIB model (1983). Other models have also been developed
based on improvements to the CEB/FIB formula or in the form of logarithmic functions of the strength
of lightweight aggregate and mortar. Generally, the current models for predicting LWAC strength
take into account various factors such as the W/C ratio, compressive strength of cement or mortar,
the thermal conductivity of lightweight aggregates (LWA), compacted LWA strength, and the volume
of LWA. However, applying these strength prediction models to the FAC-HSLWC system is
challenging due to the difficulty in determining the compressive strength of small-sized particles like
FAC particles.
2.2.3 Proposed Approach for Building a Compressive Strength Prediction Model for FACHSLWC
The proposed compressive strength prediction model for FAC-HSLWC is based on key factors
influencing concrete strength, including the strength of binder, paste volume, type and content of
aggregates (sand and FAC), and type and content of dispersed fiber reinforcement. The 28-day
compressive strength (R28) of FAC-HSLWC is a function of these factors. The quantification of
factors influencing R28 of FAC-HSLWC is established based on some concrete strength prediction
formulas mentioned earlier in the introduction. Details regarding the construction of the compressive
strength prediction model for FAC-HSLWC are presented in Chapter 4.
2.3 SCIENTIFIC BASIS FOR DEVELOPING MIX PROPORTIONING METHOD FOR FACHSLWC
2.3.1 Methods for Mix Proportioning of Concrete and Lightweight Concrete
For the FAC-HSLWC system, due to the distinct characteristics of FAC particles compared to
lightweight sand particles and the absence of large aggregates, conventional mix proportioning
methods for lightweight concrete like the ACI 211.2-98 (2004) method cannot be used. The mix
proportioning methods for high-performance concrete nowadays are mainly based on selecting


7
appropriate materials and optimizing the particle size distribution. For a concrete mixture with the
same W/B ratio, increasing the compactness of the material mixture will increase the amount of free
water in the system. Conversely, for a concrete mixture with the same cementitious content,
increasing the compactness of the larger aggregates will increase the amount of excess water in the

system. Some mix proportioning methods for LWAC have been established based on this principle.
With this approach, the designed concrete mix will have a good W/B ratio and compactness,
minimizing the voids between particles, thereby enhancing the structural integrity of the LWAC
system.
2.3.2 Proposed Approach for Developing Mix Proportioning
Method for FAC-HSLWC The proposed approach for developing a mix proportioning method for
FAC-HSLWC ensures two factors: compressive strength and workability of concrete, based on the
following principles. The mix proportioning for FAC-HSLWC is based on optimizing the CKD
(binder) component, including cement (XM), SF, and GGBFS, through experimental work using
compaction or calculations from De Larrard's compaction model. The relationship between the B/A
ratio and workability, compressive strength is established. The relationship between compressive
strength and key influencing factors such as W/B ratio, binder/aggregate (B/A ratio), FAC/aggregate
(FAC/A) ratio is established. The relationship between the workability of lightweight concrete,
concrete with 100% sand aggregate, is established to determine the amount of lightweight aggregate
to replace sand aggregate.
3

CHAPTER 3 MATERIALS USED AND RESEARCH METHODS

3.1 MATERIALS USED IN THE STUDY
3.1.1 Cement: The cement used in the study is PC50 Nghi Son Portland cement, with an average
particle size of 15 μm.
3.1.2 Silica fume: The silica fume (SF) used in the study is a loose uncompacted microsilica product,
with an average particle size of 0.151 μm.
3.1.3 Ground granulated blast furnace slag: The ground granulated blast furnace slag (GGBFS) used
in the study is of type S95, with the main particle size ranging from 1-45 μm and an average particle
size of 7.8 μm.
3.1.4 Cenosphere: The study utilizes cenospheres (FAC) obtained from the fly ash of the Pha Lai 2
thermal power plant. The FAC used has a particle size range of 10-300 μm, with a significant
concentration in the range of 45-250 μm and an average particle size of 117 μm.

Figure 3.1 Cenosphere and its particle shape by SEM
3.1.5 Sand aggregate: The aggregate used in the study is river sand, specifically sand suitable for
concrete. The sand is categorized into different types based on the largest particle size (Dmax), which
are 0.315, 0.63, 1.25, 2.5, and 5.0 mm, determined through sieving.
3.1.6 Superplasticizer admixture: The chemical admixture used for the experimental samples is
ViscoCrete 3000-20 superplasticizer from Sika. It has a water-reducing capacity of 36.5%.
3.1.7 Polypropylene fibers (PP fibers): Polypropylene fibers (PP fibers) are non-water-absorbent
fibers that are alkali-resistant and chloride-resistant. The type of fiber used in the study is
monofilament fibers with a length of 12-18 mm.
3.1.8 Mixing water: The water used for concrete mixing in the research is tap water from Hanoi city's
domestic supply. The mixing water meets the requirements specified in the TCVN 4506:2012
"Concrete and Mortar Mixing Water - Technical Requirements" standard.


8
3.2 EXPERIMENTAL METHODS
3.2.1 Standard Test Methods: The study utilized both Vietnamese and international standard test
methods to experimentally determine the mechanical and physical properties of the materials used,
as well as the properties of cement paste, fresh mixed concrete, and concrete.
3.2.1.1 Load-bearing capacity test of precast FAC-HSLWC components: Precast concrete
components made with FAC-HSLWC were subjected to load-bearing capacity tests following the
TCVN 9347:2012 standard, but using a continuous loading method with a hydraulic jack.
3.2.2 Non-standard Test Methods: Modern physical and chemical analysis methods were employed,
including X-ray diffraction (XRD), X-ray fluorescence (XRF), laser diffraction, and scanning
electron microscopy (SEM). The viscosity of fresh concrete was determined using a viscometer, and
the cement hydration degree was analyzed using thermogravimetric analysis (TGA) to measure the
CH content. The compaction degree of the FAC-HSLWC mixture was evaluated using the
Compressive Packing Model (CPM) proposed by De Larrard (1999) and verified through
experimental methods using vibration and compaction with a pressure of 10 kPa.

3.2.2.1 Concrete Mixing Method: Hobart mixers with capacities of 5 L and 20 L, as well as a
horizontal shaft mixer with a capacity of 60 L, were used in the study.
3.2.2.2 Concrete Curing Methods: The maintenance regimes for the FAC-HSLWC specimens
included: (1) standard curing regime, and (2) thermal and moisture curing regimes at 70°C and 90°C,
as well as autoclave curing at 210°C.
4

CHAPTER 4 RESEARCH ON SELECTING COORDINATION COMPONENTS FOR
FAC-HSLWC

4.1 RESEARCH ON SELECTING SUITABLE CKD COMPONENTS FOR FAC-HSLWC
4.1.1 Selection of CKD (binder) components based on optimal compaction
At the first step, the binder components, including XM, SF, and GGBFS, are selected based on the
optimal packing density of binder. The calculated packing density (PD) of the binder with different
types and proportions of admixtures is shown in Figure 4.1. The results indicate that the compaction
of CKD mainly depends on the proportion of SF, reaching the highest value of 0.767 with an SF
proportion of 30% and an XM proportion of 70% by volume, corresponding to SF and XM
proportions by weight of 23.3% and 76.7%, respectively. When the SF proportion exceeds 30%, the
compaction of CKD decreases.

Figure 4.1 Packing density of the binder with various
SMCs

Figure 4.2 Response ssurface and contour
plot of the packing density with the binder
ternary system (XM-SF-GGBFS)

4.1.2 Selection of CKD components based on optimal workability and compressive strength
From the CKD components optimized using theoretical compaction calculations, the rational
composition of CKD is selected based on ensuring the optimal workability and maximum

compressive strength of CKD using the experimental design method.
Table 4.1 Mix proportions and experimental results of binder for D-optimal design
Binder ingredients (% wt)
No.
1
2

OPC

SF

GGBFS

0,90
0,90

0,00
0,00

0,10
0,10

Binder mortar
Sand/binder Superplasticizer
Flow
ratio (by wt) (%wt of CKD)
(mm)
3,0
0,44
198

3,0
0,44
195

R3
(MPa)
37,5
37,4

R28
(MPa)
58,9
59,2


9
Binder ingredients (% wt)
No.
3
4
5
6
7
8
9
10
11
12
13
14

15
16

Design-Expert® Software
Component Coding: Actual
Độ chảy vữa (mm)

OPC

SF

GGBFS

0,85
0,75
0,70
0,70
0,64
0,60
0,51
0,51
0,46
0,45
0,30
0,30
0,30
0,30

0,15
0,10

0,30
0,30
0,00
0,15
0,00
0,00
0,30
0,16
0,30
0,30
0,10
0,10

0,00
0,15
0,00
0,00
0,36
0,25
0,49
0,49
0,24
0,39
0,40
0,40
0,60
0,60

Binder mortar
Sand/binder Superplasticizer

Flow
ratio (by wt) (%wt of CKD)
(mm)
3,0
0,44
180
3,0
0,44
188
3,0
0,44
148
3,0
0,44
145
3,0
0,44
195
3,0
0,44
182
3,0
0,44
215
3,0
0,44
216
3,0
0,44
162

3,0
0,44
185
3,0
0,44
168
3,0
0,44
166
3,0
0,44
212
3,0
0,44
210

Design-Expert® Software

Component Coding: Actual
Rn28 (MPa)

Design points above predicted value

Design points below predicted value
216

R28
(MPa)
65,6
61,8

68,3
68,9
57,6
61,4
56,4
55,9
64,4
61,52
60,3
60,6
55,75
54,7

a) Consistency of the binder mortar
Based on the experimental results presented in Table 4.1, applying the D-Optimal design yields a
second-order experimental model for the workability of binder as follows:
Flow = 194,6*A +19,4*B + 227,0*C + 112,1*AB -47,0*AC + 113*BC

Design points above predicted value
145

R3
(MPa)
42,2
36,8
44,5
44,6
34,7
35,7
27,2

26,8
39,0
35,3
35,8
35,5
26,5
26,4

Design points below predicted value

54.25

65.9

220

66

X1 = A: OPC

X1 = A: OPC

X2 = B: SF

64

X2 = B: SF

X3 = C: GGBFS

X3 = C: GGBFS

62

(MPa)
(MPa)
R28Rn28

Flow
vữa (mm)
Độ chảy(mm)

200

180

160

140

A (1)

60
58
56
54
52

A (1)

B (0)

B (0)

C (0.7)
C (0)

C (0.7)
C (0)

A (0.3)
B (0.7)

A (0.3)
B (0.7)

Figure 4.3 Response Surface and contour plot of
Hình 4.4 Response Surface and contour plot
the workability model for CKD.
of the R28 and ingredients of the binder
From the response surface and contour plot in Figure 4.3, obtained from the experimental model, it
can be observed that increasing the SF content from 0% to 30% significantly reduces the workability
of the binder from approximately 200 mm to below 160 mm. When the binder consists only of OPC
and SF, there exists a maximum SF content that ensures a workability of the mortar mix of 180 mm
or above, which is approximately 12.5% based on the experimental model. When combining SF with
GGBFS, the workability of CKD is improved.
b) Compressive strength of binder
Based on the experimental results presented in Table 4.1, applying the D-Optimal method yields an
experimental model for the compressive strength at 28 days as follows:
R28 = 59*A + 56,8*B + 50,1*C + 32,4*AB + 12,5*AC + 30,8*BC

The results depicted in Figure 4.4 show that the compressive strength of CKD increases as the SF
content increases from 0% to 30%. However, the rate of increase in compressive strength tends to
decrease when the SF content exceeds approximately 20%. Conversely, with the presence of SF,
increasing the GGBFS content from 0% to 60% decreases the early-age strength of CKD. However,
there exists an optimal GGBFS content (approximately 36%) to achieve the maximum compressive
strength (R28) of CKD at 28 days, after which the strength decreases with further increases in GGBFS
content. These results are consistent with the experimental findings regarding the dependence of
packing density on the CKD mixture's components, SF and GGBFS, as mentioned earlier.
c) Optimization of CKD components containing OPC-SF-GGBFS


10
Two CKD compositions were selected as the base components to investigate the properties of FACHSLWC:
1) CKD system consisting of XM-SF: XM/CKD ratio = 90%; SF/CKD ratio = 10% (ratios
calculated by weight).
2) CKD system consisting of XM-SF-GGBFS: XM/CKD ratio = 54%; SF/CKD ratio = 10%;
GGBFS/CKD ratio = 36% (ratios calculated by weight).

R28 (MPa)

Segregation (%)

4.2 DESIGN OF MIXTURE COMPONENTS FOR FAC-HSLWC
4.2.1 Selection of Sand Aggregate Particle Size for FAC-HSLWC
The influence of the maximum particle size (Dmax) of the sand was evaluated based on workability,
segregation of fresh mixture with W/B ratios of 0.5, 0.4, 0.3, and binder/aggregate ratio of 0.667, and
FAC/aggregate ratio of 0.5 by volume. From the experimental results, it was found that the required
PSD content to ensure the workability of concrete mixtures in the range of 180-200 mm increased as
the Dmax of the aggregate decreased from 5 mm to 0.315 mm. The experimental results determined
the segregation rate of the fresh concrete, as shown in Figure 4.5, indicating that as the Dmax of the

aggregate increased from 0.315 mm to 5 mm, the stratification level of SCC increased significantly.
The stratification levels were in the range of (1.35-7.29)%, (3.26-10.67)%, (6.22-18.21)% for the
W/B ratios of 0.3, 0.4, and 0.5, respectively. Based on the research findings regarding the influence
of sand aggregate's Dmax on workability, segregation, and compressive strength of FAC-HSLWC, this
study selected sand aggregate with a Dmax of 0.63 mm as the base aggregate for investigating the
properties of FAC-HSLWC.

Dmax (mm)

Figure 4.5 Effect of Dmax of
aggregate on segregation of the
fresh FAC-HSLWC

Dmax (mm)

Figure 4.6 Effect of Dmax of
aggregate on R28 of the fresh FACHSLWC

4.2.2 Selection of Aggregate/CKD Ratio for FAC-HSLWC
Based on the rational selection of binder ingredients discussed in section 4.1, with binder composed
of 90% XM and 10% SF, further research is conducted to optimize the ratio of material mixtures for
the production of FAC-HSLWC, which consists of three components: sand, FAC, and binder. To
optimize the particle composition of this mixture, the study utilizes an experimental method to
determine the packing density of the material mixture, following the approach proposed by De
Larrard. This method determines the packing density of the dry particle mixture, taking into account
the influence of compression pressure.
4.2.3 Selection of Aggregate and CKD Ratio based on Optimal Packing Density
From the experimental results presented in Table 4.2, a second-order function model of the packing
density of the CKD-CL mixture for FAC-HSLWC is represented as follows:
PD = 0,63*Sand+ 0,66*FAC + 0,58*CKD + 0,65*Sand*CKD+ 0,54*FAC*CKD

Using the optimization tool of the Design-Expert software to find the optimal proportions of Sand,
FAC, and CKD for FAC-HSLWC based on maximum compactness. For the aggregate system
consisting of Sand+FAC, the ratio of Sand:FAC:CKD is 0.20:0.39:0.41. For the aggregate system
consisting of only FAC, the ratio of FAC:CKD is 0.60:0.40. Therefore, two mixtures are selected as
the base mixtures to investigate the properties of FAC-HSLWC, converted into the following
parameters: Sand/A, FAC/A, and CKD/A as follows:
1) In the FAC-HSLWC system with only FAC as the aggregate, the ratio of FAC/A is 1, and the
ratio of CKD/A is 0.667 (by volume).


11
Table 4.1 Mixture ratios and experimental results based on the D-Optimal design.
Input variable:

PD

B (FAC)

C (CKD)

0
0.75
0.5
0.1875
0.5
0
1
0.6875
0
0

0.4375
0
1
0.5
0.5
0.1875

1
0
0
0.5625
0
0.75
0
0.1875
1
0.5
0.1875
0.5
0
0.5
0.5
0.6875

0
0.25
0.5
0.25
0.5
0.25

0
0.125
0
0.5
0.375
0.5
0
0
0
0.125

Design points below predicted value
0.776

X1 = A: Cát
X2 = B: FAC
X3 = C: CKD

0.75
0.7
0.65
0.6

A (1)

B (0)
C (1)

C (0)
A (0)

B (1)
A: Cát
1

Design-Expert® Software
Component Coding: Actual

0.642

PD

Design Points

0.636

0.776

X1 = A: Cát

X2 = B: FAC

0.66

X3 = C: CKD

0

2

2

0.68 0.7
0.72

0

0.74
0.76

0.76
2

2

1
B: FAC

0

PD

1
C: CKD

2) In the FAC-HSLWC system with Sand and FAC as the aggregate, the ratio of Sand/A is 0.333,
the ratio of FAC/A is 0.667, and the ratio of CKD/A is 0.667 (by volume).
4.2.4 Study on the selection of binder and aggregate ratios using the experimental method
To investigate the influence of binder content on the workability and compressive strength of FACHSLWC, the study examines the binder content with a fixed W/B ratio (fixed binder composition).
Therefore, for each binder with a specified W/B ratio, the influence of material proportions on the
concrete properties mainly depends on three variables: Sand/Total Volume of Materials (Sand/VLK),

FAC/VLK, and CKD/VLK, where VLK includes binder, sand, and FAC.
The compressive strength of FAC-HSLWC is represented as a second-degree polynomial function,
which is constructed based on an experimental plan. From the surface plots showing the influence of
the three components and the graph showing the influence of two components on the compressive
strength (R28) in Figure 4.7a, b, c for different W/B ratios (0.5, 0.4, and 0.3), it can be observed that
R28 is most influenced by the CKD/VLK ratio. As the CKDVLK ratio increases from 0.25 to 0.5, the
compressive strength (R28) tends to increase up to a certain value and then decrease.
Based on the experimental results and the surface plots of R28, for each W/B ratio, there exists an
optimal binder content to achieve the maximum compressive strength. When the amount of
cementitious material surrounding the aggregate particles is thinner or thicker than the optimal value,
it will reduce the compressive strength of the concrete.
Design-Expert® Software

Design-Expert® Software

Component Coding: Actual

Component Coding: Actual

Design-Expert® Software

Component Coding: Actual

Design points above predicted value
59.4

R28 (MPa)

R28 (MPa)

Design points below predicted value
44.6

0.8

Design points above predicted value

PD

A (Cát)

0.636

R28 (MPa)

Response surface

Component Coding: Actual

No.
1
2
3
4
5
6
7
8
9
10

11
12
13
14
15
16

Objective
function:
Packing
density (PD)
0.667
0.742
0.767
0.751
0.764
0.748
0.639
0.698
0.667
0.758
0.776
0.753
0.636
0.659
0.652
0.695

Design-Expert® Software

(a) W/B=0,5

Design points below predicted value
57.1

Design points above predicted value

(b) W/B=0,4

Design points above predicted value
69.4

(c) W/B=0,3

Design points below predicted value
69.3

78.6

X1 = A: Cát/VLK
X2 = B: FAC/VLK

X1 = A: Cát/VLK

X3 = C: CKD/VLK

X2 = B: FAC/VLK

X1 = A: Cát/VLK

70

X3 = C: CKD/VLK

60
55

X3 = C: CKD/VLK

76

A (0.75)
45

C (0.25)

62
60
58

B (0)

R28 (MPa)

64

R28 (MPa)

R28 (MPa)

78

66

50

40

80

X2 = B: FAC/VLK

68

B (0.000)

A (0)

B (0.700)

C (1)

A (0.65)
B (0)

68

C (0.35)

C (0.300)

B (0.75)

72
70

A (0.700)

56

74

A (0.000)

C (1.000)

B (0.65)

A (0)

C (1)

Figure 4.7 Response surface R28 and material ratios in the mixture of FAC-HSLWC
Using the optimization tool in the Design-Expert software to determine the suitable material
composition for making FAC-HSLWC. Based on the criteria set forth to determine the optimal
material composition, the software has selected the following proportions for maximum compressive
strength (R28): with W/B ratio of 0.5, the proportions of Sand/VLK: FAC/VLK: CKD/VLK are


12

0.60:0.00:0.40; with W/B ratio of 0.4, the proportions are 0.58:0.00:0.42; with W/B ratio of 0.3, the
proportions are 0.55:0.00:0.45.
Experimental results indicate a parabolic relationship between the CKD/VLK ratio and R28.
Therefore, there exist optimal combinations of the sand, FAC, and CKD/VLK components to achieve
the highest 28-day compressive strength for FAC-HSLWC. Based on the analysis from the
experimental figure, the optimal CKD/VLK ratio (for maximum compressive strength) ranges from
0.40 to 0.45 when the W/B ratio is between 0.5 and 0.3. It can be observed that the optimal CKD/VLK
ratio slightly increases as the W/B ratio decreases, but the effect is not significant.

Figure 4.9 Relationship between paste
volume and R28 of FAC-HSLWC

Figure 4.8 Relationship between
CKD/VLK and R28 of FAC-HSLWC

4.2.5 Verification experiment of the base mix proportion for FAC-HSLWC
Based on the objectives and scope of the research project: concrete strength > 40 MPa; density range:
1300-1600 kg/m3. The base mix for FAC-HSLWC was selected according to the research results
mentioned above, and calculations were performed based on the proposed optimized mix parameters,
followed by experimental verification. The oriented base mix composition was tested as follows:
binder with a composition of 90% XM and 10% SF; two types of aggregates: (1) consisting of only
FAC; (2) consisting of sand and FAC with a FAC/A ratio of 0.667. The dosage of superplasticizer
was adjusted to achieve a slump flow of 180-200 mm, and an assumed air entrainment of 4.5%.
Absolute volume method was used to determine the component mass of the materials. The calculated
base mix composition results are shown in Table 4.2.
Table 4.2 The base mix proportion of FAC-HSLWC by the optimized particle composition method
Mix

W/B
(%wt)

CP1
CP2

0.4
0.4

Mix parameters
FAC
S/A
B/A
/A
(%vol)
(%vol)
(%vol)
0
1.0
0.667
0.333
0.667
0.667

Mix proportion for one cubic meter (kg/m3)
Binder
(kg/m3)

XM

SF

FAC

A

SP

Water

764
761

687
685

76
76

297
197

338

4.5
4.5

306
304

The experimental results of the properties of the two base mixes are presented in Table 4.2. Both
mixes exhibited the desired workability of the concrete (within the range of 180-200 mm slump flow),

and the compressive strength at 28 days for mix CP1 and CP2 were 62.6 MPa and 67.7 MPa,
respectively (> 40 MPa). The corresponding density values were 1322 kg/m3 and 1568 kg/m3, which
fell within the range of 1300-1600 kg/m3.
Table 4.3 Experimental results of the base mix for FAC-HSLWC
Mix parameters
Ký hiệu
cấp phối
CP1
CP2

W/B
(%wt)

S/A
(%vol)

0.4
0.4

0
0.333

FAC
/A
(%vol)
1.0
0.667

B/A
(%vol)

0.667
0.667

Flow
(mm)
182
191

Compressive strength
(MPa)

Density at (kg/m3)

3 days

28 days

3 days

28 days

42.5
46.5

62.6
67.4

1348
1621

1312
1608


13
5
6 CHAPTER 5: DEVELOPMENT OF A MODEL FOR PREDICTING COMPRESSIVE
STRENGTH AND MIX DESIGN METHOD FOR FAC-HSLWC
For FAC-HSLWC, due to the differences in composition and material properties compared to
conventional concrete, the mix design used for regular concrete or some common lightweight
concretes cannot be directly applied. This is a scientific issue that needs to be addressed to support
the mix design of FAC-HSLWC.
6.1 FACTORS AFFECTING THE COMPRESSIVE STRENGTH OF FAC-HSLWC
Theoretical research and general research results have shown that the compressive strength of
concrete is influenced by several key factors, including: (1) CKD content, (2) CKD strength, (3)
Maximum aggregate size, (4) Packing density and properties of aggregates, (5) Influence of dispersed
fibers (when used), (6) Curing conditions. In this study, for simplification purposes, the development
of FAC-HSLWC strength is considered under standard curing conditions. The compressive strength
of FAC-HSLWC will then be a function of the following factors:
Rn = f(Rckd, Vckd, Dmax, Vfac/cl, Vs)
where Rn and Rckd are the compressive strengths of FAC-HSLWC and CKD, Vckd represents the CKD
content or the ratio Vckd/Vcl, Dmax is the maximum aggregate size, Vfac/cl is the volume replacement of
sand by FAC to meet the required density, and Vs is the fiber content. The quantification of the
influence of these factors on Rn of FAC-HSLWC is performed according to the principle of
considering concrete with only sand aggregate (FAC=0%) as the reference concrete, and gradually
replacing the sand with FAC (in terms of volume) to form the designed FAC-HSLWC mix. Based on
this approach, the influence coefficients of each factor on the compressive strength of the reference
concrete are established. Finally, the compressive strength of FAC-HSLWC is calculated by
multiplying the compressive strength of the reference concrete by the influence coefficients of each
factor. It should be noted that the influence coefficient of each factor will be a function depending on

that factor and the W/B ratio.
6.1.1 Influence of CKD Strength
The influence of CKD strength on FAC-HSLWC is evaluated through the effects of CKD activity
and the W/B ratio. Three CKD systems are investigated, including (1) CKD system with XM and SF,
(2) CKD system with XM+10%SF+GGBFS, and (3) CKD system with XM and GGBFS. Based on
the compressive strength survey results and using non-linear regression analysis (NLRA), the
following formula is derived to predict the CKD strength based on cement strength and the P/X ratio
(fly ash/XM):
(1) For system CKD = XM+SF then:
𝑃
2,47𝑃
R 28𝑐𝑘𝑑 = R 28opc (−7,08( )2 +
+ 1)
𝑋
𝑋

(6.1)

(2) For system CKD=XM+10%SF+(20-60)% GGBFS then:
𝑃
0,289𝑃
R 28𝑐𝑘𝑑 = R 28opc (−0,133( )2 +
+ 1)
𝑋
𝑋

(6.2)

(3) For system CKD=XM+(20-60)% GGBFS then:
𝑃

0,185𝑃
R 28𝑐𝑘𝑑 = R 28opc (−0,153( )2 +
+ 1)
𝑋
𝑋

(6.3)

At the same age, the strength of CKD is mainly influenced by CKD activity and the W/Bratio.
Therefore, the relationship between concrete strength and CKD strength can be represented by the
Bolomey formula as follows:
R 28 = AR 28𝑐𝑘𝑑 (

1
𝑁
𝐶𝐾𝐷

− 𝐵)

(6.4)

The influence of the CKD composition on its strength is determined using the standard mortar method
with ISO sand. The CKD strength is determined in the range of 44-70 MPa with a W/B ratio ranging
from 0.3 to 0.5. It should be noted that these mixtures only include the same type of sand with a


14
maximum particle size (Dmax) of 0.63 mm. By using nonlinear regression analysis, the compressive
strength formula of FAC-HSLWC is obtained as follows:
R 28 = 0,26R 28𝑐𝑘𝑑 (

1
𝑁
𝐶𝐾𝐷

+ 1,56)

(6.5)

KKd coefficient

Compressive strength R28 (MPa)

Compressive strength R28 (MPa)

The correlation coefficient R2 in equation (5.5) is 0.80.
6.1.2 Influence of CKD content
The influence of CKD content on the properties of FAC-HSLWC in this study is evaluated through
the following parameters: (1) thickness of the CKD coating layer around the aggregate particles
(denoted as CPT), (2) Excess CKD paste coefficient (denoted as Kd), (3) maximum thickness of the
CKD paste (denoted as MPT).
6.1.3 Evaluation of the influence of CKD paste content on the compressive strength of FACHSLWC
Experimental concrete mixtures are evaluated to assess the influence of the parameters BPT, Kd, and
MPT on the properties of FAC-HSLWC, based on the selected mix proportions of CKD and
aggregates as described in section 4.3.2. A total of 48 experimental mixtures are considered,
corresponding to three water-to-binder ratios (W/B) of 0.5, 0.4, and 0.3 (16 experiments for each W/B
ratio).

Paste excess coefficient

Figure 6.1 Relationship between
Kd and R28 with various W/B
and FAC/A ratios

Figure 6.2 Relationship between
Kd and R28 with various W/B and
FAC/A=0

Figure 6.3 Relationship between
Kd and KKd with various W/B and
FAC/A=0

The relationship between Kd and R28 of FAC-HSLWC with W/B ratios of 0.5, 0.4, and 0.3, in the
case of aggregate consisting of sand and FAC, is shown in Figure 6.1. Similarly, in the case of
aggregate consisting of sand, the relationship is shown in Figure 6.2. From the graphs, it can be
observed that, similar to the CKD/VLK ratio, the relationship between Kd and R28 can be represented
by a parabolic curve. This implies that there exists an optimal value of Kd for maximizing the
compressive strength (R28) of the concrete. The optimal Kd value depends on the W/B ratio, but the
variation level is not significant.
The influence of CKD content through the Kd parameter on the compressive strength of FACHSLWC is established using the following formula:
R 28 = 0,26R 28𝑐𝑘𝑑 (

1
𝑁
𝐶𝐾𝐷

+ 1,56) Κ 𝐾𝑑

(6.6)

where KKd represents the influence coefficient of the excess CKD paste coefficient.
Through regression analysis of experimental results, the influence coefficient KKd in the above formula
can be expressed as follows:
𝑁

𝐾𝐾𝑑 = 0.913𝐶𝐾𝐷 (−0.69𝐾𝑑2 + 2.61𝐾𝑑 − 1.46)
2

(6.7)

The correlation coefficient R is 0.94.
6.1.4 5.1.4 Influence of FAC/A Ratio
The coefficient Kfac is calculated as the ratio of the compressive strength R28 of various mix designs
to the FAC/CL ratio ranging from 0 to 1, with R28 of mix designs at FAC/CL=0. The relationship
between Kfac and the FAC/A ratio is shown in Figure 6.5. Through regression analysis of experimental
results, the influence coefficient of FAC/A (in volume) in the formula can be expressed as follows:


15
𝑁

𝑉𝑓𝑎𝑐
)
𝑉𝑐𝑙

(6.8)

Kfac coefficient

Compressive strength R28 (MPa)

𝐾𝑓𝑎𝑐 = 0,994𝐶𝐾𝐷 (1 − 0,143

(by volume)

(by volume)

Figure 6.4 Relationship between FAC/CL ratio
and R28 of FAC-HSLWC

Figure 6.5 Relationship between Kfac ratio and
Fac/A of FAC-HSLWC

Since FAC is used to partially or completely replace sand to reduce the unit weight of concrete, the
relationship between the unit weight of lightweight concrete and the ratio of lightweight aggregate in
the aggregate (Vfac/Vcl) needs to be established. The formulas for calculating the volume of aggregate
(Vcl) and the volume of FAC (Vfac) in lightweight concrete are as follows:
𝑉𝑐𝑙 =

𝑉𝑓𝑎𝑐

1000 − 𝑉𝑝𝑔𝑠𝑑 − 𝑉𝑠 − 𝑉𝑘
(6.9)

𝑁

1 + Γ𝑐𝑘𝑑 +
∙ 𝜌𝑐𝑘𝑑 ∙ Γ𝑐𝑘𝑑
𝐶𝐾𝐷
𝛾𝑏𝑡𝑛 − 1,2 ∙ 𝑉𝑐𝑙 ∙ Γ𝑐𝑘𝑑 ∙ 𝜌𝑐𝑘𝑑 − 𝑉𝑐𝑙 ∙ 𝜌𝑐á𝑡

=
𝜌𝑓𝑎𝑐 − 𝜌𝑐á𝑡

(6.10)

KDmax coefficient

Compressive strength R28 (MPa)

Therefore, Kfac in formula (5.8) can be calculated when the W/B ratio and the Vfac/Vcl ratio are known,
where Vfac/Vcl can be determined when the target unit weight of FAC-HSLWC (γbtn), the ratio
Vckd/Vcl, ρsand, and ρfac are known according to formulas (5.31) and (5.32).
6.1.5 5.1.5 Influence of Aggregate Maximum Size (Dmax)
The relationship between KDmax and the aggregate maximum size (Dmax) is depicted in Figure 6.7.

Dmax of aggregate (mm)

Figure 6.6 Relationship between Dmax and R28 of
FAC-HSLWC

Dmax of aggregate (mm)

Figure 6.7 Relationship between KDmax and Dmax
of FAC-HSLWC

The coefficient KDmax is established based on the influence of the W/B ratio and the ratio of Dmax to
the minimum aggregate size (Dmin). Through nonlinear regression analysis of experimental results,
the influence coefficient of aggregate size on the R28 strength of FAC-HSLWC, KDmax, can be
expressed as follows:
𝑁

𝐾𝐷𝑚𝑎𝑥 = 0,999𝐶𝐾𝐷 [1,06 − 0,37 (

𝐷𝑚𝑎𝑥 −1,14
)
]
𝐷𝑚𝑖𝑛

(6.11)

The correlation coefficient in formula (5.11) is R2 = 0.88.
6.1.6 Study on the influence of PP fiber content
The coefficient Ks is calculated as the ratio of the compressive strength R28 of various mix
proportions with different PP fiber contents (Vs) ranging from 0 to 1.5% by volume, to the
compressive strength R28 of mix proportions with Vs = 0. Through nonlinear regression analysis of


16
experimental results, the influence coefficient of PP fiber content (by volume) in the formula can be
expressed as follows:
𝑁

𝐾𝑠 = 0,962𝐶𝐾𝐷 (1,02 − 0,07 ∙ 𝑉𝑠 )

(6.12)

In summary, the compressive strength R28 of FAC-HSLWC can be predicted using the following
formula:
R 28 = 0,27R 28𝑐𝑘𝑑 (

1
𝑁
𝐶𝐾𝐷

+ 1,56) Κ 𝐾𝑑 Κ 𝐹𝐴𝐶 Κ 𝐷𝑚𝑎𝑥 Κ 𝑠

(6.13)

where: 𝑅28𝑐𝑘𝑑 , Κ 𝐾𝑑 , 𝐾𝐹𝐴𝐶 , Κ 𝐷𝑚𝑎𝑥 , và 𝐾𝑠 are determined using the corresponding formulas (5.5),
(5.13), (5.15), (5.19), and (5.21) respectively.
6.1.7 Study on the rate of compressive strength development over time
The influence of time on the compressive strength of FAC-HSLWC can be represented as follows:
R (𝑡) = 𝑅28 Φ(𝑡)

(6.14)

where R28 is the compressive strength at 28 days of FAC-HSLWC (in MPa), and Φ(t) is the ratio of
compressive strength at a specific time to its strength at 28 days. Φ(t) mainly depends on the time and
the activity of cement or CKD when combined with PGK.
According to fib 2010, Φ(t) can be expressed using the following formula:
Φ(𝑡) = 𝐸𝑋𝑃 [𝑠 ∙ (1 − (

28 0,5
) )]
𝑡

(6.15)

where s is the slope coefficient of the curve representing the relationship between time and the
strength development of concrete (slope of the strength development curve), and t is the age of

concrete (in days). From experimental results, through nonlinear regression analysis (NLRA), Φ(t)
can be expressed using the corresponding formulas for CKD with different compositions as follows:
a) For CKD consisting of (90% XM+10% SF):
Φ(𝑡)

𝑁 0,083
28 0,5
= 𝐸𝑋𝑃 [((
)
− 0,693) ∙ (1 − ( ) )]
𝐶𝐾𝐷
𝑡

(6.16)

The correlation coefficient R2 of formula (6.16) is 0,98.
b) For CKD consisting of (XM+10% SF+(2060)% GGBFS):
0,885

Φ(𝑡) = 𝐸𝑋𝑃 [(1 + 𝑅𝑝 )

𝑁 0,081
28 0,5
)
−
0,745)
∙
−
(
) )]

((
(1
𝐶𝐾𝐷
𝑡

(6.17)

where 𝑅𝑝 is ratio of (SF+GGBFS) in CKD by mass. 𝑅𝑝 is in range of 0 to 0,7. The correlation
coefficient R2 of formula (6.17) is 0,98.
c) For CKD consisting of (XM+(2060)% GGBFS):
Φ(𝑡) = 𝐸𝑋𝑃 [(1 + 𝑅𝑔𝑠 )

0,98

((

𝑁 0,059
28 0,5
)
− 0,736) ∙ (1 − ( ) )]
𝐶𝐾𝐷
𝑡

(6.18)

where 𝑅𝑔𝑠 is ratio of GGBFS in CKD by mass. 𝑅𝑔𝑠 is in range of 0 to 0,6. The correlation coefficient
R2 of formula (6.18) is 0,98.
6.1.8 Verification of the proposed model
The suitability of the proposed model for predicting the 28-day compressive strength of FACHSLWC with different W/B ratios is assessed in Figure 6.8a. Additionally, the proposed model's
prediction of the compressive strength development over time (from 3 to 91 days) is shown in Figure

6.8b. The results indicate that the proposed models allow for predicting the 28-day strength and the
compressive strength development from 3 to 91 days of FAC-HSLWC within a deviation of no more
than 15% compared to experimental results. The suitability of the Figure model for the compressive
strength from 3 to 91 days of FAC-HSLWC with different CKDs is evaluated in Figure 6.9.


17
90
80

70

(MPa)
đoán
Cường độ dự R
Predicted
28 (MPa)

Predicted R28 (MPa)

+15%
2

R =0,98

-15%

+15%

60

50
40

R2=0,98

+15%

30

3 ngày

20

7 ngày

10

91 ngày

0
0.00

20.00

40.00

60.00

80.00

Cường độ thí nghiệm
(MPa)
Tested compressive
strength
(MPa)

Tested compressive strength (MPa)

Figure 6.8 The 28-day compressive strength of FAC-HSLWC as determined by the proposed
model and experimental data
90

CKD: OPC+SF+GGBFS
80
70
60
50
40

30

3 ngày

7 ngày

20

28 ngày
91 ngày

10
0

0.0

20.0

40.0

60.0

Cường độ thí nghiệm
(MPa)
Tested compressive
strength
(MPa)

80.0

độ dự đoán (MPa)
strength (MPa)
PredictedCường
compressive

Cường độ dự đoán (MPa)
strength (MPa)
Predicted compressive

90

CKD: OPC+GGBFS
80
70
60
50

40
30

3 ngày
7 ngày

20

28 ngày
91 ngày

10
0

0.0

20.0

40.0

60.0

80.0

Cường độ thí nghiệm
(MPa)
Tested compressive
strength
(MPa)

Figure 6.9 The compressive strength at various ages of FAC-HSLWC as determined by the
proposed model and experimental data.
6.2 5.2 DEVELOPING MIX DESIGN METHOD FOR FAC-HSLWC
6.2.1 General principles
The proposed mix design method in this study is based on the following principles: Optimization of
the CKD composition, including 2 or 3 constituents such as cementitious materials (XM), silica fume
(SF), and ground granulated blast furnace slag (GGBFS); Selection of CKD/CL ratio and W/B ratio;
Achieving the target lightweight concrete properties by using lightweight aggregate, specifically
cenospheres, to partially or fully replace the fine aggregate; Verification of the W/B ratio based on
the predicted compressive strength of FAC-HSLWC to ensure the required strength; Incorporation of
dispersed polypropylene (PP) fibers when there is a demand for reduced shrinkage and enhanced
flexural strength and crack resistance of the concrete; Utilization of superplasticizers to adjust the
workability of the lightweight concrete mix as required (typically in the range of 180-200 mm slump).
6.2.2 Steps of FAC-HSLWC mix design
The steps of the proposed FAC-HSLWC mix design method in this research are presented in Figure
5.10.


18

Figure 6.10 The flowchart of the steps for the mix design of FAC-HSLWC.
7 Chapter 6. STUDY ON THE MECHANICAL PROPERTIES OF FAC-HSLWC
This chapter presents the research results on the influence of the replacement ratio of FAC for sand,
the influence of supplementary cementitious materials (SCMs) replacing cement, and the influence

of curing conditions (standard and moist curing) on the properties of FAC-HSLWC. The replacement
ratios of FAC for sand were 0%, 50%, 70%, and 100% by volume, corresponding to the lightweight
concrete density range of 1300-1600 kg/m3, along with a control sample using 100% sand (FAC ratio
= 0%). The SCMs used to replace cement included 10% SF, 20%, 40% GGBFS, and 10% SF + 20%,
40%, 60% GGBFS (by weight percentage of CKD). The mix proportions of FAC-HSLWC were
calculated based on the selected mix design parameters presented in Chapter 3, with a CKD/CL ratio
of 0.667 and a water-to-binder ratio (W/B) of 0.4 by mass.


19
7.1 CHARACTERISTICS OF FAC-HSLWC MIXTURE
7.1.1 6.1.1 Workability
The experimental results, as shown in Figure 7.1a, indicate that as the FAC content increases, the
flowability of the fresh FAC-HSLWC mixture tends to decrease. For mix designs using XM
combined with SF and GGBFS, the flowability of the fresh FAC-HSLWC mixture decreases when
10% SF is used. The workability of the fresh FAC-HSLWC mixture is further improved when SF is
combined with GGBFS at ratios of 20%, 40%, and 60% (Figure 7.1b).

Flow (mm)
xịe (mm)
Độ chảy

250
200

0
-5

150

-10

100

-15

50

-20

0

-25
FAC0

Flow
Độ chảy-nhóm PGK

(b)
200
control
Change
(%)(%)
số soto ĐC
Sai

% so ĐC

% so ĐC

190

0
-4

180
-8
170
160

-12

150

-16

control
Change
(%)(%)
số sotoĐC
Sai

Flow
Độ chảy-nhóm thay cát

xịe (mm)
chảy(mm)
ĐộFlow

(a)

FAC50 FAC70 FAC100
CấpMix
phối

Cấp phối
Figure 7.1 Workablity of fresh FAC-HSLWC

Flow (mm)

Segregation (%)

Segregation (%)

7.1.2 Viscosity
The viscosity of the FAC-HSLWC concrete mixture is determined for mix designs with only FAC as
the aggregate. The experimental results show that the presence of SF and GGBFS in the CKD
component reduces the viscosity of the concrete mixture. However, if the SF content is increased
excessively, it significantly increases the amount of water in the system due to the large surface area
of SF particles compared to cement. Additionally, it can be observed that the effectiveness of reducing
the viscosity of the CKD paste is significantly evident when increasing the GGBFS content from 0%,
20%, 40%, to 60%.
7.1.3 Bleeding
The experimental results determine the bleeding for different concrete mix designs, and it is found
that all tested FAC-HSLWC samples show no signs of bleeding on the surface. Therefore, the
bleeding of these mixtures can be considered as zero.
7.1.4 6.1.4 Paste segregation
Segregation
% Control
Segregation

Flow
(b)
(a)

Figure 7.2 Độ phân tầng của HHBT FAC-HSLWC khi chịu tác động rung
The experimental results for segregation with different FAC-HSLWC mix designs in Figure 7.2a
show a clear decrease in segregation tendency when replacing sand aggregate with FAC. Regarding
the influence of PGK on the segregation of FAC-HSLWC mixtures, the experimental results in Figure
7.2b demonstrate that segregation is similar to the flowability of the mixtures. The use of 10% SF in
the CKD reduces stratification, and stratification increases when further replacing cement with 2060% GGBFS.
7.1.5 Air Content
The air content tends to increase from 3.2% for the 100% sand mixture (FAC0) to 4.3% when
replacing 100% sand with FAC (FAC100). With the CKD using 10% SF and (20-60)% GGBFS, the
air content slightly increases from 4.2% (OPC100) to 4.5% when replacing cement with 10% SF in


20
the CKD (SF10GS0). When continuing to replace cement with 20%, 40%, and 60% GGBFS in the
CKD, the air content tends to slightly decrease from 4.5% to 4.2% (60% GGBFS sample).
7.1.6 Setting Time
The setting time of the FAC-HSLWC with 100% sand aggregate is 5 hours and 30 minutes for initial
setting and 7 hours and 20 minutes for final setting, increasing gradually when sand is replaced with
FAC. The setting time increases proportionally with the degree of sand replacement by FAC.
Compared to the 100% sand mixture (FAC0), the initial and final setting times increase by 1 hour
and 20 minutes and 1 hour and 10 minutes, respectively, when sand is completely replaced by FAC
(FAC100). When using 10% SF to replace cement in the CKD, both the initial and final setting times
of the FAC-HSLWC show negligible changes. However, when further replacing cement with (2060)% GGBFS in the CKD, the setting time of the FAC-HSLWC gradually increases.
7.2 HYDRATION DEGREE AND MICROSTRUCTURE
The hydration degree and microstructure of FAC-HSLWC in this study are evaluated through the
calcium hydroxide (CH) content using thermogravimetric analysis (TGA) and microstructure analysis

of FAC-HSLWC samples using scanning electron microscopy (SEM).
7.2.1 CH Content
From the TG/DTG analysis results, the calculated CH content from the experimental results shows a
significant decrease in CH content at both 3 and 28 days for mix designs using SCMs to partially
replace cement in the CKD. The highest reduction is observed in the sample using CKD containing
10% SF + 60% GGBFS, with CH content of only 1.03% and 1.05% at 3 and 28 days, respectively.
Additionally, the CH content calculated based on the mix design tends to decrease when using 10%
SF and further decrease when using (20-60)% GGBFS to replace cement. The experimental results
on the effect of curing conditions at 70°C, 90°C, and autoclave on the CH content of FAC-HSLWC
samples with different SCM contents show that higher temperature curing significantly reduces the
CH content in the analyzed samples. The CH content decreases rapidly with increasing curing
temperature, and almost no CH content is observed in all samples cured in an autoclave.
7.2.2 Microstructure of FAC-HSLWC
Through the observation of SEM images, it can be seen that there are not many hydration products
on the surface of FAC particles with cement stone at 3 days of age. However, at 28 days of age, it is
possible to observe a better bond in the ITZ (Interfacial Transition Zone) between FAC particles and
cement stone due to the formation of hydrated minerals. Additionally, the thickness and density of
the hydration products on the surface of FAC particles also increase corresponding to the curing
temperature at 70°C, 90°C, and autoclave.
(a)

Hạt FAC

Hạt FAC

Đá CKD

(b)

Đá CKD

Vùng ITZ

Đá CKD

SF10GS0-3d

Hạt FAC

SF10GS0-28d

Đá CKD

SF10GS0-90 oC-28d

Hạt FAC

SF10GS0-AC-28d

Figure 7.3 SEM images capture the microstructure of the FAC-HSLWC samples
7.3 MECHANICAL PROPERTIES
7.3.1 Density and Compressive Strength
The density of FAC-HSLWC decreases correspondingly from 2180 kg/m3 to 1656 kg/m3, 1505
kg/m3, and 1322 kg/m3, representing a decrease of 24%, 30.9%, and 39.4%, respectively, as the
replacement of sand with FAC increases to 50%, 70%, and 100%. It should be noted that the
compressive strength of FAC-HSLWC is determined on 40x40x160 mm specimens. The conversion
factor from 40x40x160 mm specimens to 150x150x150 mm specimens is 0.83, as determined through
experimentation. The specific strength (the ratio of strength to bulk density of the material) of the
lightweight concrete specimens increases proportionally with the volume of FAC replacing sand.
Specifically, the specific strength increases from 34 kPa/kg.m-3 for the FAC0 sample to 41.8

21
kPa/kg.m-3, 45.6 kPa/kg.m-3, and 47.9 kPa/kg.m-3, corresponding to an increase of 12.3%, 34.1%, and
40.9%, respectively, when the FAC replacement of sand is 50%, 70%, and 100%. The use of 10% SF
increases the compressive strength of the concrete at 7, 28, and 91 days, with respective increases of
6.6%, 5.6%, and 6.8% compared to the control sample using only cement (OPC100). When partially
replacing OPC with a combination of SF and GGBFS at GGBFS ratios of 20%, 40%, and 60%, the
compressive strength at 7, 28, and 91days decreases compared to the sample containing 10% SF
(FAC40).
Change to Control (%)

(c)
Compressive strength (MPa)

Compressive strength (MPa)

Density (kg/m3)

Compressive strength (MPa)

Specific strength (KPa/kg.m-3)

(b)

(a)

Figure 7.4 Density, compressive strength and specific strength of FAC-HSLWC
7.3.2 Compressive Strength at Different Moisture Curing Conditions
Moisture curing at 70°C, 90°C, and autoclaving at 200°C are effective in improving the compressive

strength of FAC-HSLWC at 3 and 28 days of testing. The effectiveness in enhancing the compressive
strength through moisture curing conditions is as follows: autoclave curing (200°C, 2 MPa) > moist
curing at 90°C > moist curing at 70°C > standard curing at 27°C and RH ≥ 95%.
7.3.3 Flexural Strength
Similar to compressive strength, the flexural strength decreases as the volume fraction of FAC
increases. The flexural strength at 28 days decreases from 8.57 MPa for the control sample (FAC0)
to 6.88 MPa, 6.12 MPa, and 5.87 MPa, corresponding to reductions of 19.4%, 28.3%, and 31.3%
compared to the control sample. When replacing OPC with GGBFS, the flexural strength tends to
increase with the increase in the GGBFS ratio. The highest flexural strength is achieved with a mix
design using 40% GGBFS. The relationship between flexural strength and compressive strength at
28 days for FAC-HSLWC can be represented by the following equation:
0,86
𝑅𝑢28 = 0,21 ∙ 𝑅28
∙

(7.1)

Figure 7.5 The relationship between the elastic
modulus and bulk density of FAC-HSLWC

Density (kg/m3)

Poissons coefficient

Density (kg/m3)

Elastic modulus (GPa)

where Ru28 is the flexural strength and compressive strength at 28 days of FAC-HSLWC (MPa) and
R28 is the compressive strength on 15x15x15 cm cubic specimens of FAC-HSLWC (MPa).

7.3.4 Elastic Modulus and Poisson's Ratio
Increasing the FAC/CL ratio leads to a decrease in bulk density (KLTT) and a reduction in elastic
modulus. The elastic modulus decreases from 32.7 GPa for the concrete sample without FAC (FAC0)
to 19.4 GPa to 13.73 GPa, corresponding to decreases of 40.7% to 58.0% as the FAC/CL ratio
increases from 50% to 100%. For the mix designs using SF and GGBFS, the elastic modulus slightly
decreases at 7 days and is higher or equivalent when the GGBFS ratio is between 20% and 60%. The
elastic modulus of the concrete primarily depends on the compressive strength and bulk density of
the concrete.

Figure 7.6 The elastic modulus and Poisson's ratio
of FAC-HSLWC


22
When comparing the experimental results with the predicted formula for elastic modulus based on
compressive strength and bulk density of concrete according to ACI 318-14, it is found that this
formula can be applied to FAC-HSLWC with a similarity coefficient of R2=0.98.
7.4 DURABILITY
7.4.1 Drying Shrinkage
The drying shrinkage of FAC-HSLWC tends to decrease when sand is replaced with FAC. The drying
shrinkage of concrete after 182 days with a mix design containing 100% sand aggregate decreased
by 3%, 8%, and 26% corresponding to FAC/aggregate ratios of 50, 70, and 100%. The drying
shrinkage of FAC-HSLWC improves when using supplementary cementitious materials such as SF
and GGBFS to replace OPC. Compared to the sample with 100% OPC (CKD), the drying shrinkage
after 182 days decreased by 36.3% when using 10% SF in CKD, and further replacing OPC with
20%, 40%, and 60% GGBFS resulted in drying shrinkage reductions of 40.7%, 41.4%, and 47.1%
respectively. Additionally, the drying shrinkage of FAC-HSLWC also decreases with a decrease in
the W/B ratio and the addition of PP fibers. The drying shrinkage at 182 days decreased from 940 με
for the sample without PP fibers (FAC40W0.4) to 900 με for the sample containing 0.3% PP fibers
(FAC40PP0.3) and 872 με for the sample containing 0.5% PP fibers (FAC4PP0.5), corresponding to

reductions of 4.3% and 7.3% respectively.
7.4.2 Water Absorption
The water absorption of FAC-HSLWC increases with the increase in FAC content replacing sand.
The water absorption at 28 days increased from 3.61% for the control sample (FAC0) to 4.62%,
5.05%, and 6.21% corresponding to increases of 28.0%, 39.9%, and 71.7% when the FAC/sand ratio
was 50, 70, and 100%. When using SF and GGBFS to replace OPC, the water absorption at 7 and 28
days both decreased. The best reduction in water absorption was achieved with a GGBFS ratio of
60%, where the water absorption decreased from 7.15% for the control sample (OPC100) to 5.35%,
corresponding to a reduction of 25.2%.
7.4.3 Chloride Ion Permeability
Increasing the FAC ratio significantly reduces the chloride ion penetration through the concrete by
decreasing the ionic migration coefficient and increasing the concrete resistivity. The Rapid Chloride
Penetration Test (RCPT) decreased by 61.9%, 68.9%, and 77.8% from 1590 coulombs for the 100%
sand sample (FAC0), while the Bulk Electrical Resistivity Test (BERT) of the control sample
increased by 89%, 108.9%, and 153.7% respectively, when the FAC/CL ratio was 50, 70, and 100%.
Using a combination of 10% SF and GGBFS in CKD at a ratio of 20-60% further reduced the chloride
ion penetration. From the experimental results, the correlation between RCPT and BERT for FACHSLWC had a correlation coefficient of R2=0.92.
7.4.4 Sulfate Attack Resistance
The expansion due to sulfate attack of FAC-HSLWC decreases as the FAC/CL ratio increases from
0 to 100%. The expansion of mortar prisms at 12 months for the 100% sand sample (FAC0) decreased
by 45.6%, 53.4%, and 59% respectively, when the FAC/aggregate ratios were 50, 70, and 100%.
When using 10% SF, the expansion due to sulfate attack of the FAC-HSLWC prism specimens
decreased at all test ages, with a reduction of 50.1% at 12 months compared to the CKD sample of
100% OPC (OPC100). When continuing to replace cement with GGBFS at a content of 20-60%, the
expansion due to sulfate attack at various test ages tended to increase compared to the 10% SF sample,
although it remained lower than the 100% OPC sample. Therefore, using 10% SF and 10% SF + 60%
GGBFS was the most effective in reducing the expansion due to sulfate attack of FAC-HSLWC.
7.5 LOAD-BEARING CAPACITY OF FAC-HSLWC SLAB FLOORS
The study was conducted on prestressed reinforced concrete (PRC) slab floors with dimensions of
3280x1060x140 mm, using lightweight FAC-HSLWC and normal concrete. The FAC-HSLWC used

had densities of 1400 kg/m3 (D 1.4) and 1600 kg/m3 (D 1.6), and the normal concrete (D 2.4) had a
compressive strength grade of B35 (actual compressive strengths of the three types of concrete D 1.4,
D 1.6, and D 2.4 were 45 MPa, 48 MPa, and 52 MPa respectively). The PRC slab floors were tested


23
to evaluate their behavior under uniformly distributed bending loads (4 concentrated loads, 2
supports), as shown in Figure 7.7.
P/4

P/4

P/4

P/4

Figure 7.7 Experimental setup diagram and the loading system and test equipment for the slab floor.
The research results showed that for concrete with the same compressive strength grade, the allowable
deflection load, the load at the 0.3 mm crack width, and the ultimate load of the slab floor using FACHSLWC were equivalent to those of the reinforced concrete slab floor. The difference in behavior
between the FAC-HSLWC slab floor and the reinforced concrete slab floor is as follows: the
appearance of the first crack occurs much earlier in the FAC-HSLWC slab floor, but the crack width
and the development of cracks (in terms of length and width) are significantly smaller; the deflection
at the point of failure of the FAC-HSLWC slab floor is also much smaller than that of the reinforced
concrete slab floor.
8

CONCLUSION

I. CONCLUSION
Based on the research results of the dissertation, the following conclusions can be drawn:

1. It is entirely possible to produce high-strength lightweight concrete using lightweight aggregates
such as fly ash hollow spheres from fly ash of thermal power plants (FAC) and other available
materials in Vietnam (FAC-HSLWC), with mechanical properties such as compressive strength
ranging from 40 to 70 MPa, density ranging from 1300 to 1600 kg/m3, and water absorption below
6.5%, meeting the technical requirements for structural use in construction projects.
2. Binder (CKD) can be used in the production of FAC-HSLWC by combining cement with SF
and/or GGBFS. Binder combined with 10% SF (by mass) is effective in improving the
development of strength, water absorption, and other mechanical properties of FAC-HSLWC.
Binder with a combination of 10% SF and 20-60% GGBFS is effective in enhancing durability,
such as the water resistance of FAC-HSLWC.
3. The relationship between the binder/aggregate ratio (B/A ratio) by volume and the compressive
strength of FAC-HSLWC can be represented by a parabolic curve. Therefore, there exists an
optimal value for the CKD/A or CKD/A ratio to achieve the highest compressive strength of FACHSLWC. Depending on the W/Bratio, the optimal CKD/A ratio increases as the W/B ratio
decreases. For W/B ratios ranging from 0.5 to 0.3, the optimal CKD/A ratio falls within the range
of 0.4 to 0.45.
4. The relationship between the key factors affecting the compressive strength of FAC-HSLWC,
including the strength of binder, binder content, replacement ratio of FAC for sand, maximum
aggregate size, and dispersed fiber content, has been established in the form of nonlinear
functions. Through these influencing factors, a prediction model has been developed to estimate
the 28-day compressive strength of FAC-HSLWC with a high coefficient of determination, R2 =
0.98. Furthermore, the development of compressive strength from 3 to 91 days of age for FACHSLWC can be forecasted using the established models.
5. A mix design method has been developed for the FAC-HSLWC system, capable of providing
preliminary proportions of FAC-HSLWC to achieve compressive strengths ranging from 40 to 80
MPa and bulk densities ranging from 1200 to 2000 kg/m3.
6. Substituting sand with FAC significantly reduces the bulk density and compressive strength,
flexural strength, modulus of elasticity, and drying shrinkage of FAC-HSLWC, but increases the
specific strength and Poisson's ratio. The decrease in flexural strength and modulus of elasticity
is greater than the decrease in compressive strength. Due to the lower modulus of elasticity and