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)[.]
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: Prof Dr Nguyen Van Tuan – HUCE 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; INTRODUCTION 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)” 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 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 ✓ 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 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 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 ✓ ✓ ✓ ✓ 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 STRUTURE OF THE THESIS The thesis consists of an Introduction, chapters, General Conclusion and Recommendations, 36 tables, 97 figures, presented within 151 pages excluding the appendix 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 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 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 to 400 μm The majority of cenospheres fall within the range of 20 to 300 μm, with wall thicknesses ranging from 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 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 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 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 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 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 CHAPTER 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 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 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 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 CHAPTER 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 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 11 Table 4.1 Mixture ratios and experimental results based on the D-Optimal design Input variable: PD B (FAC) C (CKD) 0.75 0.5 0.1875 0.5 0.6875 0 0.4375 0.5 0.5 0.1875 0 0.5625 0.75 0.1875 0.5 0.1875 0.5 0.5 0.5 0.6875 0.25 0.5 0.25 0.5 0.25 0.125 0.5 0.375 0.5 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 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 2 0.68 0.7 0.72 0.74 0.76 0.76 2 B: FAC PD 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 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) 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 0.333 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) days 28 days days 28 days 42.5 46.5 62.6 67.4 1348 1621 1312 1608 13 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 ( 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 ( 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 ( 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) (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 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 + 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 to 1.5% by volume, to the compressive strength R28 of mix proportions with Vs = 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 ( 𝑁 𝐶𝐾𝐷 + 1,56) Κ 𝐾𝑑 Κ 𝐹𝐴𝐶 Κ 𝐷𝑚𝑎𝑥 Κ 𝑠 (6.13) where: 𝑅28𝑐𝑘𝑑 , Κ 𝐾𝑑 , 𝐾𝐹𝐴𝐶 , Κ 𝐷𝑚𝑎𝑥 , 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+(2060)% 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 to 0,7 The correlation coefficient R2 of formula (6.17) is 0,98 c) For CKD consisting of (XM+(2060)% 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 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 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 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 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% R =0,98 -15% +15% 60 50 40 R2=0,98 +15% 30 ngày 20 ngày 10 91 ngày 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 20 28 ngày 91 ngày 10 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 20 28 ngày 91 ngày 10 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 or 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 Chapter 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) xòe (mm) Độ chảy 250 200 -5 150 -10 100 -15 50 -20 -25 FAC0 Flow Độ chảy-nhóm PGK (b) 200 control Change (%)(%) số soto ĐC Sai % so ĐC % so ĐC 190 -4 180 -8 170 160 -12 150 -16 control Change (%)(%) số sotoĐC Sai Flow Độ chảy-nhóm thay cát xòe (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 HHBT FAC-HSLWC 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 hours and 30 minutes for initial setting and 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 hour and 20 minutes and 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 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 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 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 and 28 days of testing The effectiveness in enhancing the compressive strength through moisture curing conditions is as follows: autoclave curing (200°C, 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 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 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 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, 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 CONCLUSION I CONCLUSION Based on the research results of the dissertation, the following conclusions can be drawn: 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 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 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 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 to 91 days of age for FACHSLWC can be forecasted using the established models 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 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 24 higher creep, caution should be exercised in the structural design calculations when using FACHSLWC The use of mineral admixtures, including 10% SF, SF combined with 20-60% GGBFS, improves the compressive strength, flexural strength, splitting tensile strength, modulus of elasticity, and bond strength of steel reinforcement in FAC-HSLWC The most effective improvement is achieved with a combination of 10% SF and 40% GGBFS The improvement in flexural strength and modulus of elasticity of FAC-HSLWC using mineral admixtures, SF, and GGBFS is superior to the improvement in compressive strength at 28 days The durability of FAC-HSLWC, as indicated by properties such as water absorption, water permeability, chloride ion resistance, and sunlight resistance, is improved when OPC is used in combination with supplementary cementitious materials (SMC) such as SF and GGBFS Among them, the best effectiveness in improving the durability properties is achieved with a SMC consisting of 10% SF + 60% GGBFS For concrete with the same compressive strength grade, the allowable deflection load, the crack width of 0.3 mm, and the ultimate load of the slab using FAC-HSLWC are equivalent to those of traditional reinforced concrete slabs The different behavior of FAC-HSLWC slabs compared to traditional slabs is that the appearance of the first crack occurs much earlier, but the crack width and crack development (in terms of length and width) are significantly smaller The deflection at the point of slab failure for FAC-HSLWC slabs is also much smaller compared to traditional slabs II RECOMMENDATIONS Continue studying and testing the mechanical properties and long-term durability of large-scale structural elements using FAC-HSLWC under different environmental conditions such as hot and humid climates, aggressive environments (marine environment, chemical exposure) Study the use of FAC-HSLWC for structures with special requirements, such as fire-resistant structures in high-rise buildings (fire walls), floating structures in marine environments, large span structures, and lightweight composite reiforcement components for thin-walled structures 25 LIST OF PUBLISHED WORKS BY THE AUTHOR Journal papers: Le Viet Hung, Le Trung Thanh and Nguyen Van Tuan (2021), Experimental Study on Drying Shrinkage of Structural Lightweight Concrete using Fly Ash Cenosphere, International Journal of GEOMATE, 21(87) 95–101; ISSN 2186-2982 (P), 2186-2990 (O), Japan, Geotechnique, Construction Materials and Environment DOI: https://doi.org/10.21660/2021.87.j2337 (SCOPUS) Lê Việt Hùng, Lê Trung Thành, Nguyễn Văn Tuấn (2021), Nghiên cứu chế tạo bê tông nhẹ cường độ cao sử dụng hạt vi cầu rỗng từ tro bay, Tạp chí Khoa học Cơng nghệ Xây dựng, ĐHXDHN, 15 (6V) 146–157; ISSN 2615-9058, DOI: https://doi.org/10.31814/stce.huce (nuce)2021-15(6V)13 Lê Việt Hùng, Lê Trung Thành, Nguyễn Văn Tuấn (2021), Nghiên cứu sử dụng hạt vi cầu rỗng từ tro bay thay phần cốt liệu nhỏ cho chế tạo bê tơng nhẹ chịu lực, Tạp chí Vật liệu Xây dựng, Viện Vật liệu xây dựng, 11(6) 21–27; ISSN: 1859-381X DOI: https://doi.org/10.54772/jomc.6.2021.205 Scientific conference papers: L V Hung, N V Tuan, and L T Thanh (2021), Experimental Investigation of High-Strength Lightweight Concrete Using Fly Ash Cenosphere, Conference paper CIGOS 2021, Emerging Technologies and Applications for Green Infrastructure, pp 637–645; ISSN 2366-2557 (P), 23662565 (O) DOI: https://doi.org/10.1007/978-981-16-7160-9_64 (Indexed by SCOPUS)