Using the simplex centroid design approach, concrete mixture design can be optimized based on workability and key strength metrics such as compressive, flexural, and splitting strengths.
Introductions
Since the introduction of high-performance concrete (HPC) [1], its excellent mechanical properties and durability have attracted the attention of scholars in the field of engineering materials [2] Its application in engineering has significantly reduced the self-weights of structures, improved their spans, strengths, and durability, and extended their service lives [3, 4] Because of these advantages, HPC has considerable potential for the construction of buildings [5], bridges [6], and tunnels[7], as well as for marine engineering [8] And military engineering The mechanical behavior and durability of HPC affect its service life in engineering applications Moreover, the flowability of HPC significantly affects its hardened properties[9, 10]
High-strength concrete is a critical material in modern construction, prized for its enhanced durability, structural performance, and versatility In the pursuit of sustainable construction practices, there is an increasing demand for optimizing the use of raw materials and reducing the environmental impact associated with concrete production One significant aspect of this endeavor is the management of waste generated during the production of high-strength concrete, particularly the waste generated from Residue Fluid Catalytic Cracking (RFCC) catalysts
An ideal optimization method achieves consistent and comprehensive test results using the minimum number of experiments, effectively optimizing the mix proportion of Ultra-High-Performance Concrete (UHPC) in a thorough manner Orthogonal experimental design is a commonly employed optimization method in such experiments [11, 12] that satisfies these requirements and is widely used for optimizing the mix proportions of various types of concrete[13] Researchers have conducted orthogonal experiments to investigate the fundamental mechanical properties and flow characteristics of UHPC They varied parameters such as water-
2 to-binder ratios, silica fume content, fly ash content, and steel fiber content to identify effective combinations of factors and levels that optimize these properties [14-16]
Research has been conducted on optimizing the amount of waste UHPC One study found that using recycled brick powder (RBP) as a replacement for silica fume (SF) in UHPC can improve mechanical strength and reduce the autogenous shrinkage of the concrete mixture [17] Another study focused on the use of waste glass powder (WGP) as a cementitious material in UHPC It summarized previous works that used milled waste glass to replace cement, quartz powder, silica fume, and sand in UHPC mixtures [18] Additionally, the effective utilization of glass sand (GS) to replace quartz sand in UHPC was evaluated, and it was found that the compressive strengths of UHPC can be improved by the addition of GS [19] However, there was no specific mention of optimizing the amount of RFCC waste in UHPC in the abstracts provided
The RFCC process plays a vital role in the petrochemical industry, facilitating the conversion of heavy hydrocarbons into valuable products such as gasoline and diesel fuel However, this process generates a substantial quantity of spent RFCC catalyst waste, which, if not managed efficiently, poses environmental challenges and impacts the overall sustainability of construction projects
This research aims to address this critical issue by focusing on the optimization of RFCC waste utilization in high-strength concrete By strategically incorporating this waste material into concrete mixes, we can potentially reduce the demand for traditional raw materials, such as cement and aggregates, while simultaneously providing an environmentally responsible solution for managing RFCC waste
The interactions among multiple factors affecting the fluidity, compressive strength, and flexural strength of HPC remain unclear Therefore, it is crucial to elucidate the effects of these multifactor interactions during the optimization of UHPC mix proportions Additionally, research on the mechanical properties, durability, and microstructure of HPC after self-compaction without vibration requires enhancement Consequently, experimental investigations are necessary to
3 analyze the impacts of various factors on slump flow, compressive strength, flexural strength, and splitting strength of HPC This research aims to identify the most significant factors and understand the influence of multifactor interactions on these properties
Therefore, it is crucial to quantify and incorporate these by-products in order to maintain the mechanical properties and durability of concrete Professor Hwang at Taiwan Tech has utilized the Densified Mixture Design Algorithm (DMDA) method, applying its results to notable projects in Taiwan, such as Taipei 101 and 85 Sky Tower The DMDA can optimize the proportion of of high-performance concrete to effectively fill voids in fine aggregate, thereby enhancing both the durability and mechanical properties of concrete [20, 21]
In this study, we will investigate various aspects of incorporating RFCC waste into high-performance concrete, including its effects on concrete properties, mechanical performance, and sustainability Through comprehensive laboratory experiments and analysis, we intend to develop guidelines and recommendations for the effective utilization of RFCC waste in high-performance concrete production
The outcomes of this research not only hold the potential to reduce the ecological footprint of construction but also contribute to the sustainable management of RFCC waste, aligning with the broader goals of environmental responsibility in the construction industry As we embark on this journey to optimize the quantity of RFCC waste in high-strength concrete, we aim to promote a more sustainable and resource-efficient approach to construction engineering
In this study, based on previous research on UHPC and HPC, the proportion of closest-packed aggregates was determined, along with the influence ranges of various factors on the basic mechanical properties and fluidity of UHPC Subsequently, using Simplex-Centroid Design Methodologies , the effects of various parameters on the fluidity, compressive strength, and flexural strength, splitting strength of UHPC were examined via a range and variance analysis, along with the
4 maximum packing density of aggregates According to the specific requirements for compressive strength, flexural strength, splitting strength, and slump flow, the optimized mix proportion of UHPC was obtained To verify the optimization effect in the orthogonal design, the flowing and mechanical properties of UHPC with and without the optimized mix proportion were compared.
Literature review
Foreign research
Ultra-high-performance concrete (UHPC) mix design and optimization have been the focus of several research papers Various additives and modifications have been explored to enhance the properties of UHPC for specific applications, such as nuclear waste storage facilities[22] Some UHPC mixes require special curing methods and unusual materials, making them uneconomic for practical construction The particle packing theory and optimal mineral proportion have been used to design UHPCs with coarse basalt aggregate, and the effects of mineral composition, aggregate size, and powder content have been analyzed A stepwise optimization approach based on particle packing density has been proposed for UHPC mix design, resulting in improved materials saving and performance Additionally, a steel mix and connect based on UHPC has been developed for steel-reinforced concrete in bridge engineering, simplifying construction and reducing concrete quantity
D Jiao et al [23] The primary objective of concrete mixture design is to achieve a harmonious balance among workability, compressive strength, durability, cost-effectiveness, and sustainability In this research, for a given strength grade, we optimized the ideal paste composition, consisting of cement, fly ash, and slag, as well as the optimal ratio between paste, fine aggregate, and coarse aggregate, using the simplex centroid design method based on rheological properties (Figure 1.1) The results revealed that the optimal content of total cementitious materials in concrete could be determined by considering the relationships between workability, yield stress, plastic viscosity, and the paste volume fraction Likewise, the optimal substitution of supplementary cementitious materials could be determined based on
5 the rheological properties and compressive strength of concrete with ternary cementitious components Utilizing the simplex centroid design method based on rheological properties proves to be an effective approach for optimizing concrete mixture design
Figure 1.1: Optimization of cementitious materials composition
Z Sun et al [24] discusses the optimization design of ultrahigh-performance concrete (UHPC) through a multi-factor interaction analysis The study aims to identify the optimal mix proportion of UHPC (Figure 1.2) and evaluate its properties, including mechanical properties, microstructure, and bond performance The analysis includes range analysis and orthogonal design, and the results provide insights into the effects of different factors on UHPC performance The study found that the optimal mix proportion of UHPC includes a water-binder ratio of 0.18, a sand ratio of 1.15, and a cementitious material content of 20% The addition of steel fiber can improve the mechanical properties of UHPC, while the addition of silica fume can improve its microstructure The optimized UHPC had high compressive and flexural strength, as well as low shrinkage The study provides valuable insights into the effects of different factors on UHPC performance and can be used as a reference for future research on UHPC
Figure 1.2: Optimization results for the UHPC mix proportion.[24]
P.-K Chang [25] discusses an approach to optimizing mix design for high- performance concrete (HPC) using the densified mixture design algorithm (DMDA)
It explores the significant influence of the water-to-solid weight ratio on concrete volume stability (Figure 1.3) and the effects of the W/S ratio on strength and durability at both fresh and hardened states utilization of fly ash and slag has also proven beneficial in enhancing the rheology of HPC
Figure 1 3: The relationship between concrete resistivity and W/S ratio at N 1.6[25]
M Sohail et al [26] provides a comprehensive review of the methods used to produce high-performance and ultrahigh-performance concretes based on particle size (Figure 1.4), including their constituents, mixture proportions, mixing protocols, and particle packing models The paper covers a range of topics, including the effects of water- cement ratio, supplementary cementitious materials, and superplasticizers on concrete properties, as well as the use of fibers and nanoparticles to enhance strength and durability Additionally, the paper discusses various testing methods for measuring concrete properties, such as porosity and compressive strength, and presents several models and theories for achieving a higher packing density in concrete and reducing defects
Figure 1 4: Size distribution of constituents of UHPC [26]
E Ghafari et al [27] present a statistical mixture design approach for eco-efficient
UHPC mixtures The method allows for the assessment of the influence of different parameters on the performance of UHPC The results show that the incorporation of quartz flour improves both compressive strength and flowability (Figures 1.5 and 1.6) The study also indicates that it is possible to design UHPC mixtures with high compressive strength and low cement content without heat curing
Figure 1 5: The counterplots for compressive strength (MPa): (a) Plain-UHPC; (b)
UHPC with 1 vol.% steel fibers.[27]
Figure 1 6: Two response optimization in the function of cement, silica fume, and quartz flour [28]
E Ikponmwosa et al [28] investigate the optimization of high-performance concrete using local materials Twelve different concrete mixes with varied water-cement ratios and particle sizes of coarse aggregates were studied The results showed that the slump value increased with an increase in the water-cement ratio while the
9 strength decreased A statistical model was proposed to predict the strength of the concrete (Figure 1.7)
Figure 1.7: Compressive strength at different w/c ratios for (a) Sample A, (b)
S Math and P Rangaraju [29] discuss the Application of Simplex-Centroid Design
Methodologies to Optimize the Proportions of Ternary cementitious blends in High- Performance Concretes (Figures 1.8, 1.9, and 1.10) The methodology involves using statistical mixture design techniques to achieve multiple outcomes with fewer test runs and includes figures and tables that show the performance parameters of concrete mixtures and how they can be optimized using this methodology Overall, provides insights into how this methodology can be used to improve the strength and durability of high-performance concrete mixtures
10 a) regular simplex triangle (b) augmented simplex triangle Figure 1 8: Simplex-Centroid design triangle with constrained design points[29]
Figure 1 9: Superimposed multiple response surface contours[29]
Figure 1.10: Compressive strength and rapid chloride-ion permeability isocontours overlap[29]
Domestic research
D.Dung et al [30] investigated and assessed Non-Ground RFCC (Figure 1.11)
Fias an Additive for Commonly Used Cement Types in the Vietnamese Market: OPC, PBC30, PBC40, and PCB50 Through experiments on mortar samples (Figure1.12) with specific proportions tailored to each cement type, it was concluded that Non- Ground RFCC exhibits similar physical and chemical properties to fly ash (Table 1.1), a widely used additive in cement production
(a) After cracking (b) After burning carbon
Figure 1 11: RFCC powder from Binh Son refinery[30]
Figure 1.12: Sample for an experiment based on Vietnam standard 6016-2011 Table 1.1: Chemical composition Non-Ground RFCC based on Vietnam standard
CKT SO 3 MKN Al 2 O 3 Fe 2 O 3 CaO MgO SiO 2 CaO td
N D Hung [31] researched incorporating local materials and industrial waste, such as fly ash or FCC (Fluid Catalytic Cracking), into construction materials, which can reduce costs and contribute to environmental protection A control mortar sample
(without using fly ash and FCC) with a local sand volume ratio below ASTM standards of 0.5 can achieve a compressive strength of up to 44 MPa at 28 days For mortars with a W/C ratio of 0.4 or 0.42, when cement is replaced with fly ash or FCC at a rate of 5 to 10%, it can increase the compressive strength of the mortar by 10.8% to 28.5% at 28 days and reduce the workability of the mortar compared to the control sample while still meeting construction requirements
L.A Thang [32] has commented that RFCC (Residue Fluid Catalytic
Cracking) is a waste product generated from oil refineries, and currently, there is an excess of this waste material in considerable quantities RFCC was utilized in the study of fatigue resistance of asphalt concrete under cyclic loading, following the EN
12697-24 (2012) standard The results revealed that RFCC enhanced the fatigue resistance of asphalt concrete by up to 20% compared to the conventional type
D.C Phan et al [33] have noted that adding RFCC to asphalt concrete
(Figure1.13) enhances properties like elastic modulus and abrasion, reducing wheel ruts and increasing fatigue resistance, based on laboratory comparisons with conventional asphalt
Research Objectives
The research illustrates that RFCC can effectively produce HPC concrete by partially substituting cement or Q-powder
The overall goal of the project is to propose an effective method of optimizing concrete mixtures with strong robustness for variable raw materials, wide applicability, and sustainability
Discuss various methods of concrete composition optimization and focus on the simplex centroid design method as a potential solution
Highlight the importance of considering both workability and mechanical properties, durability, economy, and ecology of engineering applications when designing concrete mixtures.
Research content
We are focused on the mixture design of concrete using the simplex centroid design method
Discuss the challenges of optimizing concrete mixtures to meet various performance requirements, such as workability, mechanical properties, durability, economy, and ecology of engineering applications
Provide an overview of various methods of concrete composition optimization and highlight the advantages and disadvantages of each method Then, introduce the simplex centroid design method and explain how it can be used to optimize concrete mixtures In this study, a detailed case analysis is presented to demonstrate the effectiveness of the proposed method
The potential benefits of utilizing the simplex centroid design method for concrete mixture design will be discussed These benefits include its strong robustness in handling variable raw materials, its wide applicability, and its contribution to sustainability.
Approach method
Research based on general theory and laboratory experiments
The proposed approach is to optimize the mixture design of concrete based on rheological properties using the simplex centroid design method
Provide detailed explanations of each step and provide examples to illustrate the process Design experiments for samples in the laboratory
Refer to domestic and international studies on optimizing the mixture design of concrete.
Research method
Materials and Mix Design: Collect RFCC samples from the petroleum refining industry Prepare UHPC mixes with varying percentages of RFCC as a cement replacement Design mixes according to established UHPC standards and guidelines
Testing and Analysis: Conduct a series of tests to evaluate UHPC properties with varying RFCC content, including compressive strength, flexural strength, and durability tests Perform microstructural analysis using scanning electron microscopy (SEM) to understand the impact of RFCC on the material's structure
Data Analysis: Statistical analysis, including regression analysis and the use of
Minitab software, will be performed to determine the optimal RFCC content for each UHPC property under investigation.
16
Microstructural properties of high-performance concrete
Because of its superior packing density and limited utilization of larger particles, HPC displays a higher level of uniformity compared to NSC Unreinforced UHPC tends to be quite brittle due to the minimal disparity in size and strength between the matrix and the tiny aggregates, causing cracks to frequently propagate when exposed to external pressure Nevertheless, when HPC is reinforced with minuscule or microfibrous materials, it demonstrates ductile material characteristics, potentially significantly improving its tensile performance
The dense microstructure and high material strength of high-performance concrete (HPC) can be attributed to three key factors Firstly, HPC typically has a very low water/binder ratio, often 0.25 or less, where the binder includes cement, silica fume, and other reactive substances such as nanoparticles This low water/binder, water-to-cement ratio results in a matrix with very low permeability ( Figure 2.2) Secondly, the packing density of HPC is optimized Ultrafine particles such as cement, quartz powder, and silica fumes are combined in an extremely compact manner Lastly, the high pozzolanic content, including silica fume, reacts
19 with free calcium hydroxide (CH) in the matrix to form additional calcium silicate hydrate, which significantly enhances compressive strength
Figure 2 2: Effect of packing-optimized ultrafine particle combinations on HPC compressive strength[41]
Besides the mass-based water/binder ratio, the volume-based water/ultrafine particles ratio has also been established [42] for further improving the HPC strength, as shown in Figure 2.2 It has been proven that given the same water/binder ratio, HPC with volume-based optimization can get even higher strength enhancement
The workability of UHPC presents challenges in optimizing packing density Increased packing density results in smaller pores between particles, making water less effective in lubricating them Moreover, a higher concentration of ultrafine particles leads to a significant increase in surface area requiring wetting Additionally, inter-particle forces among ultrafine particles tend to cause agglomeration, hindering their effectiveness as optimal fillers These issues can be addressed with high- performance superplasticizers capable of dispersing and liquefying not only cement but also all other fine particles, thereby mitigating conflicts arising from particle interactions [43].
Material compositions
The material composition of UHPC diverges from that of conventional concrete in several aspects In recent years, there has been a growing body of research aimed at comprehending how these material compositions impact the performance of UHPC.
Overview of RFCC
RFCC stands for "Residue Fluid Catalytic Cracking" (Figure 2.3) This is a process in the petroleum industry used to convert residual oil products from oil distillation into higher-value products such as gasoline, diesel, and petroleum gas
The RFCC process helps better use remaining oil components to produce higher- value products
Figure 2 3: Residue Fluid Catalytic Cracking (RFCC) Technology and Catalysts
Principle of operation: During the RFCC process, the remaining oil is passed through a series of catalyst tanks at high temperatures and pressure Catalyst helps separate and convert oil molecules into higher-value petroleum products through chemical conversion
Main products: The main products of the RFCC process include gasoline and diesel, which are important fuels for transportation and industry (Figure 2.4) Petroleum gas is also produced during this process
Process improvement: RFCC technology is continuously researched and improved to optimize performance and minimize environmental impact
Wide Application: RFCC is widely used in refineries and crude oil processing plants around the world to create higher-value petroleum products from raw materials (Table 2.1)
The RFCC process plays an important role in converting crude oil into higher-value products, contributing to the global oil, gas, and energy industry
Figure 2 4: Residue Fluid Catalytic Cracking (RFCC)[44]
Table 2 1: Comparison of Residue Fluid Catalytic Cracking (RFCC) versus Fluid
(RFCC) Fluid Catalytic Cracking (FCC)
• The feedstock is Atmospheric Residue (AR)
• Lower API gravity feedstock compared to vacuum gas oil (VGO) (API < 20)
• Higher Conradson carbon residue (CCR) (> 4 wt%)
• Heavy metal contaminants such as nickel and vanadium are well concentrated in the feedstock
• The feedstock is vacuum gas oil (VGO) and heavy gas oil (HGO)
• Higher API gravity and hydrogen content compared to atmospheric residue feedstock (API > 30)
• Lower Conradson carbon residue and lower heavy metal contents (< 4 wt%)
• Contaminant Coke is exclusive to RFCC unit operations
• Conradson carbon in the feedstock predominantly winds up as coke
• Heavy metals in the feedstock typically accumulate on the catalyst particles catalysing the formation of coke
• A higher temperature in the regenerator
5kg per kg of feedstocks)
• Lower coke yield of about 5 wt% on feed compared to residue feedstock with a coke yield higher than 10 wt% on feed
• A lower temperature in the regenerator
• Higher catalyst circulation rate (> 5kg per kg of feedstock)
(RFCC) Fluid Catalytic Cracking (FCC)
• Two-stage regenerator with catalyst coolers
• Combustor-type regenerator for complete combustion, thereby eliminating the cost of installing a
• Single-stage regenerator without catalyst cooler
• Bubbling bed regenerator for partial combustion, thereby necessitating the need for a CO incinerator
• The catalyst must have a hierarchical structure to accommodate the large polycyclic aromatic molecules
• Catalyst must be tolerant of the heavy metal contaminants in the feed, notably vanadium and nickel
• Catalyst requires a low delta (∆) coke character to stall the high ∆ coke tendencies of the impurities in the feedstock
• Microporous catalyst structure is sufficient to accommodate the molecules of the feedstocks
• Catalyst needs a relatively higher delta (∆) coke to drive their action
27
Application of Simplex-Centroid Design Methodologies to mix the proportions
In mixture experiments , the factors are the components of ingredients of a mixture, and consequently, their levels are not independent
In an experiment where each component in the mixture can range from 0.0% to 100.0% of the total (Figure 3.1), a 23-factorial experiment would involve all possible combinations of proportions ranging from 0.00 to 1.00, resulting in the corners of a cube However, the constraint 𝑥1 + 𝑥2 + 𝑥3 = 1 reduces the three- dimensional experimental region to a two-dimensional shaded equilateral triangular plane
Figure 3 1: Coordinate system for Three-component mixture [58]
= in a three-component mixture experiment
0 1 1 2 2 3 3 y= + x + x + x + and the constraint x 1 + + =x 2 x 3 1 It makes one of the four coefficients in this model redundant
1 and 2 * do not represent the effects of components x 1 and x 2 as they would in a factorial experiment, but they represent the effects of x 1 and confounded with the opposite of the effect of the slack variable 𝑥3
3.1.2 Experimental designs to fit Scheffé models
Simplex-lattice designs (SLD) are used to study the effects of the mixture components on the response variable
• 𝑚 + 1 equally spaced values from 0 to 1
• All possible combinations (mixtures) of the proportions are used
( , , ) (1,0,0),(0,1,0),(0,0,1)x x x Only the pure components are required for a linear design
The coefficients 𝛽𝑖 in the linear model can be estimated as an average of all the response data at the pure component where 𝑥𝑖 = 0
In the linear model, the effect of blending two or more components is assumed to be linear, and no intermediate points are necessary in the design
The 50/50 mixtures of each pair of components are required to estimate the coefficients 𝛽𝑖𝑗 of the quadratic blending effects in the model
2 2 2 2 2 2 are required in addition to the pure components for a quadratic design
Simplex-centroid design (SCD) is an alternate design that allows the estimation of all coefficients in the special cubic model
• SCD in the three-mixture components (Figure 3.4)
Pure component blends: ( , , ) (1,0,0),(0,1,0),(0,0,1)x x x 1 2 3 Binary mixtures : ( , , 0), ( , 0, ), (0, , )1 1 1 1 1 1
Figure 3 4: Simplex-centroid design (SCD)[58]
Simplex-Lattice Design (3,2): 6 runs (Figure 3.5)
Simplex-Centroid Design (3,2): 7 runs (Figure 3.6)
3.1.3 Design mixed component RFCC replacement for cement
Based on previous studies and utilizing the Simplex-Centroid Design Methodology,
I designed the mixture using Minitab software (see Figure 3.7 and Table 3.1) The results are presented in Table 3.2
Figure 3 7: Design mix component RFCC replacement for cement
Table 3 1: Bounds of Mixture Components
Amount Proportion Pseudo component Comp Lower Upper Lower Upper Lower Upper
C 0.000000 0.035000 0.00000 0.10542 0.00000 0.26515 Table 3 2: Mixture composition in cubic meters for cement is replaced by RFCC
3.1.4 Design mixed component RFCC replacement for Q-Powder
Similar to the design of the RFCC replacement for cement, I also created mixtures where q-Powder is replaced by RFCC (see Figure 3.8 and Table 3.3)
These mixtures are named CI, CII, CIII, CIV, and CV and display in Table 3.4
Figure 3 8: Design mix component RFCC replacement for Q-Powder
Table 3 3: Bounds of Mixture Components
Amount Proportion Pseudo component Comp Lower Upper Lower Upper Lower Upper
C 0.041000 0.110000 0.20918 0.56122 0.00000 0.66346 Table 3 4: Mixture composition in cubic meters for Q-Powder is replaced by RFCC
Mixtures CI CII CIII CIV CV
Raw materials 34 1 Residue Fluid Catalytic Cracking (RFCC)
3.2.1 Residue Fluid Catalytic Cracking (RFCC)
RFCC waste is a catalyst in the petroleum cracking process, and after multiple processes, it loses its activity and becomes RFCC waste (Figure 3 9 and 3.10) The composition and properties of RFCC are presented in Tables 3.5 and 3.6
Table 3 5: Chemical composition of RFCC [30]
RFCC SiO2 Al2O3 Fe2O3 CaO
Table 3 6: Physical properties of RFCC [30]
Number Parameter Unit Outcome standards
2 Surface area ratio cm2/g 2980 TCVN4030-85
3 Residue on a sieve with a hole size of 0.09mm
Figure 3 9: Residue Fluid Catalytic Cracking (RFCC)
Figure 3 10: (a) SEM Image; (b) EDX Spectrum of the sFCC Sample[59]
SF is a by-product from the silicon and ferro-silicon alloy industries To lower the oxygen concentration, quartz is heated in an electric furnace to a temperature of 2000C using wood chips, coal, or coke At the top of the furnace, the fume condenses and oxidizes to produce silicon oxide (SiO2), which is composed of approximately 75% SiO2 and has an ultrafine particle size (Table 3.7) The amount of silica content is related to the quality of silica fume.[60]
Because SF is so ultra-finely ground (Figure 3.12), using it in concrete increases the amount of water needed; thus, a super-plasticizer must be added to achieve the necessary workability Reduced heat release, narrow pore size distribution, higher compressive strength, and lime-consuming activity are all positively impacted by SF's strong pozzolanic reactivity [61]
Project using Silica Fume (SF) in dispersed particle form provided by Elkem Company, which is available in the Vietnamese market see Figure 3.11
Figure 3 11: A sample of Silica fume
Table 3 7: Technical characteristics of SF[61]
Number Properties Unit Outcome Request
2 Volume weight or Bulk density kg/m 3 250 TCVN 4030:2003
3 Activity strength toward cement % 112,5 TCVN 8827-2011
4 Average particle size àm 0,15 Laser analysis
The dissertation utilized PC50 cement (Figure 3.13) for research purposes The results of the basic properties of cement, cement composition, and particle composition are presented in Table 3 8 below
Table 3 8: Basic properties of PC50 cement[62]
Number Properties Unit Outcome Request Testing method
3 Median particle size àm 16,68 Laser analysis
In this topic, the fine aggregate used is natural sand from the Lô River (see Figure 3.14), which is a common type of aggregate in Vietnam According to various research studies worldwide [63] To UHPC, coarse sand with a size modulus greater than 3.0 is necessary, and properties are shown in Table 3.9
Table 3.9: Basic properties of Sand[63]
Number Properties Unit Request Testing method
2 Bulk volume of pores or Pore volume kg/m 3 1510 TCVN 7572:2006
3 Porosity in the porous state
4 Surface-dry saturated moisture content % 1,05 TCVN 7572:2006
5 Modulus of fineness of sand
6 Content of dust, mud, and clay
Calcite (calcium carbonate: CaCO3) is the main mineral that makes up limestone, a sedimentary rock, as seen in Figure 3.15 and Figure 3.16 High concentrations of calcium and/or magnesium carbonate, as well as tiny amounts of other minerals, are found in large quantities in limestone, a sedimentary rock that occurs naturally and is widely distributed The addition of limestone filler to Portland cement alters the kinetics and mechanism of cement hydration in several aspects
The filler effect of limestone filler accelerates the hydration of Portland Clinker grains, particularly the C3S, at early ages, enhances cementitious system particle packing, creates new calcium hydroxide nucleation sites, and results in the formation of calcium carbon aluminates through the reaction of C3A from Portland Clinker and CaCO3 from limestone [64]
Figure 3 15: Microstructure of limestone powder (SEM) [64]
Figure 3 16 A sample of calcium carbonate (CaCO3) Quartz I/Limestone filler
Prepare for the experiment
To carry out the experiment and ensure consistency across all mixtures, I developed a mixing process (see Figure 3.17) Using this process, all sample mixtures were prepared in the laboratory (see Figures 3.18 to 3.23)
RFCC partially replacing cement in mixtures
To evaluate the feasibility of the above option, experiments test were conducted in the laboratory (see Figure 3.24)
Figure 3 25 Mixture C1 Figure 3 26 Mixture C2 Figure 3 27 Mixture C3
Figure 3 28 Mixture C4 Figure 3 29 Mixture C5 Figure 3 30 Mixture C6
The slump flow test was conducted (see Figures 3.25 to 3.30), and the results are displayed in Table 3.10
Table 3 10: The result for Slump flow (cm) option replacement Cement
Figure 3 31: Slump Flow for option replacement Cement
The graph in Figure 3.31 gives information about slump flow for option replacement cement Overall, components C1, C3, C4, and C5 have fluctuated from
54cm to 60 cm We can see that component 2 gets a high value of 60 cm; otherwise, component 6 does not have slump flow, meaning slump flow equals 0cm
Regression Equation in Coded Units
Slump flow (cm) = 66.84 + 65.26 Silicafume + 63.78 Cement
- 37.20 Silicafume*Silicafume + 25.89 Cement*Cement - 0.04206 SFCC*SFCC
Figure 3 32: Main effects for Slum flow (cm)
(a) Contour Plots of slump flow
(b): Contour plot of Slump Flow (c): 3D Surface
Figure 3 33: Model graphs for slump flow of composite RFCC replacement for cement
The results of the slump flow were modeled (see Figure 3.33) and analyzed to assess the effects of RFCC on slump flow (see Figure 3.32)
The flexural test was conducted (see Figures 3.34 to 3.39), and the results are displayed in Table 3.17
Gross error is a type of error that arises from a lack of care during the measurement process Besides, raw error is also known by other names such as error or error error Therefore, it is essential to eliminate errors in the experiment as shown in Table 3 11 to Table 3.16
Table 3 11: Flexural test on mortars (MPa) for mixtures C1
Flexural test on mortars (MPa)
Flexural test on mortars (MPa)
(b) After processing gross error Table 3 12: Flexural test on mortars (MPa) for for mixtures C2
Flexural test on mortars (MPa)
Mixtures C2 Version Flexural test on mortars (MPa)
Average 12.456 (b) After processing gross error Table 3 13: Flexural test on mortars (MPa) for mixture C3
Flexural test on mortars (MPa)
Flexural test on mortars (MPa)
18.673 (b) After processing gross error Table 3 14: Flexural test on mortars (MPa) for mixture C4
Flexural test on mortars (MPa)
Flexural test on mortars (MPa)
15.829 10.195 (b) After processing gross error Table 3 15: Flexural test on mortars (MPa) for mixture C5
Version Flexural test on mortars (MPa)
Flexural test on mortars (MPa)
Table 3 16: Flexural test on mortars (MPa) for mixture C6
Flexural test on mortars (MPa)
Flexural test on mortars (MPa)
6.647 (b) After processing gross error Table 3 17: Flexural test on mortars (MPa) for component replacement cement
Mixture Flexural test on mortars (MPa)
Figure 3 40: The result of the Flexural test on mortars ( MPa) regarding Cement is replaced by RFCC
The graph in Figure 3.40 illustrates the compressive strength results of cement replaced by RFCC at 28 days Overall, all components fluctuate between 5.294 MPa and 18.673 MPa Component C3 reaches a maximum value of 18.673 MPa, while component C1 drops to a minimum of 5.294 MPa
Regression Equation in Coded Units
Flexural test on mortars (MPa) = 20.57 + 10.12 Silicafume + 10.34 Cement
- 13.07 Silicafume*Silicafume - 2.482 Cement*Cement + 4.950 SFCC*SFCC
MixturesFlexural test on mortars ( MPa)
Figure 3 41: Main effects plot Flexual test on montars
(a): Contour Plots of Flexural test on montars
Figure 3 42: Model graphs for Flexual test on montars of composite RFCC replacement for cement
The results of flexural test were modeled (see Figure 3.42) and analyzed to assess the effects of RFCC on slump flow (see Figure 3.41)
Determine the tensile cross-sectional area of the cylindrical samples when pressed and split On the cylindrical test specimen, draw a frame made up of two generating lines and two diameters lying on the same plane Measure with accuracy up to 1mm each pair of parallel sides and calculate the average values The tensile area is the area of the frame calculated according to the average values of the edges
Clean the surface of the compression plate, transfer pad, and sample pellet in the parts that will be in contact during the split pressing test
Similarly, place the second transfer pad on top of the test specimen along the generating line The framing lines drawn on the cylindrical sample must coincide with the longitudinal axis of the transmission pad (see Figure 3.43) For material samples using organic adhesives, the time to perform operations from the time the sample is taken out of the curing place to the time the sample is placed in the compressor must not exceed 2 min
Figure 3 43: Diagram of cylindrical test specimen placement
Forcing the test specimen to split by increasing the load continuously and
56 evenly until the specimen is destroyed (see from Figure 3 43 to Figure 3.49) The deformation rate (speed of moving the machine's compression table) when pressing to split the material sample using organic adhesive is equal to (50.5) mm/min For material samples using inorganic adhesives, use an increasing load rate so that the tensile stress, when pressed and split, increases steadily in the range from 0.10 MPa/min to 0.70 MPa/min proportional to the strength of the test sample so that the sample destruction time is not less than 30 s Record the maximum destructive load for each test specimen
The tensile strength R kc, when pressed and split for each cylindrical test piece, is calculated to the nearest 0.01 MPa according to the formula:
R kc: Tensile strength when pressed and split (MPa) P: Load when destroying a cylindrical sample (N) H: Height of the cylindrical sample (mm)
D: Diameter of the bottom of the cylindrical sample (mm)
Table 3 18: Information splitting tensile test for RFCC replacement for cement
Figure 3 50: Result in Splitting strength at age 28 regarding cement replaced by
The graph in Figure 3.50 and Table 3 18 illustrate the splitting strength at 28 days for the option where cement is replaced by RFCC Overall, all components range between 1.46 MPa and 4.63 MPa Notably, component C6 reaches a minimum value of 1.53 MPa, while component C3 achieves a peak of 4.63 MPa
Regression Equation in Coded Units
Figure 3 51: Main effects plot for Splitting
Sp li tt ing Str eng th ( MP a)
Figure 3 52: Graphs for Splitting tensile test on the mortar with RFCC replacement for cement
The Splitting tensile test results were modeled (see Figure 3.52) and subsequently analyzed to evaluate the impact of RFCC on flexural strength (see Figure 3.51)
The compressive strength test was conducted (see Figures 3.53 to 3.58), and the results are displayed in Table 3.19
Figure 3 53: Compressive strength sample test Mixtures C1
Figure 3 54: Compressive strength sample test Mixtures C2
Figure 3 55: Compressive strength sample test Mixtures C3
Figure 3 56: Compressive strength sample test Mixtures C4
Figure 3 57: Compressive strength sample test Mixtures C5
Figure 3 58: Compressive strength sample test Mixtures C6
Table 3 19: Result in Compression strength for option replacement cement
Figure 3 59: Compressive Strength of mixture with cement replaced by RFCC
The graph in Figure 3 59 and Table 3 19 illustrate compressive strength at age 28 for option replacement cement Overall, all components have varied between 27.5
MPa and up to 105 MPa As we can see, component C6 reaches a low point at an average of 36 MPa, whereas component C2-2 reaches a high value of 105 MPa In conclusion, component C3 reaches a peak at 94 MPa
Regression Equation in Coded Units
Compressive strength (Mpa) = 99.68 + 54.42 Silicafume + 44.04 Cement
Figure 3 60: Main effects for compressive strength
Figure 3 61: Model graphs for Compressive Strength test on the mortar of mix with
RFCC replacement for part of cement
The results of the compressive strength test were modeled (see Figure 3.61) and analyzed to assess the effects of RFCC on slump flow (see Figure 3.60)
3.4.5 Summary of results from experiment testing
The summary of experimental results is presented in Table 3.20 and illustrated in Figures 3.62, 3.63, 3.64, and 3.65
Table 3 20: Summary result Splitting, Compressive, Slump test for Mixtures cement replaced by RFCC
Figure 3 62: The relationship between Compressive Strength and Splitting Strength at age 28 days
Figure 3 63: The relationship between compressive strength, splitting strength, and slump flow
Figure 3 64: The relationship between compressive strength, splitting strength, flexural and slump flow
Figure 3 65: The relationship between Compressive Strength and Slump flow
The graphs in Figure 3.64, Figure 3 65, and Figure 3 66 illustrate the relationship between compressive strength (MPa) and slump flow (cm) for different mixtures of RFCC From the data provided in Table 3.20 and the corresponding graph, several observations can be made:
Mixture C2 (0% RFCC) stands out with the highest compressive strength, reaching
88.991 MPa and a slump flow of 60 cm However, this mixture also exhibits a high standard deviation of 22.593, indicating significant variability in its compressive strength measurements This variability suggests that while the mixture can achieve high strength, its performance may be inconsistent, potentially affecting its reliability in practical applications
Mixture C3 (5% RFCC) presents the second-highest compressive strength at
94.3445 MPa, slightly surpassing that of C2 The slump flow for this mixture is 59 cm Notably, the standard deviation is relatively low at 4.940, suggesting more consistent and reliable results This balance of high strength and consistency makes C3 a promising option for applications requiring both durability and predictability
Mixture C1 (4% RFCC) demonstrates strong performance with a compressive strength of 87.4255 MPa and a slump flow of 57 cm The low standard deviation of 2.495 indicates minimal variability in its strength measurements, reinforcing its potential as a dependable mixture Although slightly lower in strength than C2 and C3, its consistency makes it a viable choice for many construction projects
Mixture C5 (13% RFCC) has a lower compressive strength of 76.937 MPa, accompanied by the lowest slump flow of 55 cm among the RFCC mixtures The standard deviation is 1.882, reflecting consistent results despite its lower strength This consistency is beneficial, but the reduced compressive strength might limit its use in applications requiring higher durability
Mixture C4 (0% RFCC), serving as the control mixture, shows a compressive strength of 61.3085 MPa and a slump flow of 59 cm With a very low standard deviation of 1.024, this mixture demonstrates highly consistent compressive strength results While its strength is lower compared to the other mixtures, its reliability makes it a useful benchmark for evaluating the performance of other mixtures
Mixture C6 (9% RFCC) presents the lowest compressive strength at 35.3725 MPa and a slump flow of 0 cm, indicating no workability The high standard deviation of 11.160 suggests considerable variability in its strength measurements, highlighting its inconsistency The combination of low strength, lack of workability, and high variability makes C6 the least favorable option among the tested mixtures
In summary, as the RFCC content increases, there is a trend of decreasing slump flow, indicating reduced workability The highest compressive strengths are observed in mixtures with lower RFCC content (C2 and C3), while the mixture with the highest RFCC content (C6) shows the lowest compressive strength and no workability A notable compressive strength and slump flow drop for the mixture with 9% RFCC (C6) significantly affects its overall performance The mixture with 5% RFCC (C3) appears to provide the best balance of high compressive strength and reasonable workability
3.4.6 The Impact of RFCC Content on Slump Flow and Compressive
Table 3 21: Compressive strength and Slump flow test results based on percentage
Figure 3 66: Slump flow based on RFCC percentage in the mixture
90
108
OPTIMIZATION OF RFCC IN HIGH- PERFORMANCE CONCRETE (HPC)
Concrete mix optimization involves determining the proportions of concrete constituents to enhance specific desirable properties in fresh or hardened concrete This process entails selecting suitable ingredients and determining their relative quantities to produce concrete that economically meets the minimum required properties, such as strength, durability, and consistency Given the variability of ingredients and the challenge of quantifying material properties accurately, selecting concrete proportions is also about finding the optimal combination of these ingredients based on empirical data from relevant standards, practical experience, and general guidelines
High-performance concrete (HPC) has been found to have practical applications in engineering due to its excellent workability and impressive resistance to cracking To enhance its performance, it's essential to conduct a thoughtful design process Therefore, in this study, we performed a comprehensive analysis of multiple factors to determine the optimal HPC mix proportion using an orthogonal design Subsequently, we assessed the characteristics of the refined HPC
Our approach incorporated multiple performance parameters and levels within the orthogonal design, with each factor's optimal range determined through a combination of range and variance analyses and multi-factor interaction analysis [24] The mixed proportions for our tests were then established within these optimized parameter ranges to meet performance requirements We further examined the workability, mechanical properties, durability, and microstructure of the optimized UHPC
The findings demonstrated that the HPC, based on the optimized mix
91 proportion, exhibited significant enhancements in flowability, compressive strength, and flexural strength (Figure 4.1) Additionally, it displayed superior resistance to chloride ion penetration and long-term drying shrinkage when compared to conventional concrete Furthermore, the application of hot water curing notably accelerated the hydration process of UHPC As the curing period extended, hydration products filled internal voids, resulting in denser UHPC paste The pores situated in the interfacial transition zone between steel fibers and the paste gradually filled with hydration products, leading to improved bonding performance between steel fibers and paste
Figure 4 1: Optimization results for the UHPC mix proportion[24]
An experimental program was devised to achieve an optimal balance of key mix design parameters for economically producing ultra-high-performance concrete (UHPC) The program systematically evaluated various mix design factors, including superplasticizer content, coarse-to-fine aggregate ratio, and steel fiber volume fraction Important metrics such as packing density, water film thickness, and excess paste film thickness were calculated during this process [65]
The study investigated how packing density, water film thickness, and excess paste film thickness influenced both compressive strength and the flow properties of
92 fresh UHPC mixes These findings were crucial in establishing practical ranges for these parameters within the UHPC domain
Through response surface analysis of fresh mix flow and hardened concrete compressive strength tests, optimal values for mix design parameters were identified The optimized mix achieved a desirable balance between the flow characteristics of fresh UHPC and the compressive strength of hardened concrete, thereby meeting the project's objectives effectively
Adil M Jabbar et al [66] comment that achieving ultra-high strength in concrete is facilitated by optimizing packing density through high cement content, silica fume, fine aggregate, a low water-to-cementitious materials ratio (w/cm), and a high dosage of high-range water-reducing admixture (HRWRA) These factors contribute significantly to enhancing the material's compressive strength
D Jiao et al.[23] comment that the central aim of concrete mixture design is to achieve a harmonious blend of workability, compressive strength, durability, cost- effectiveness, and environmental sustainability In this study, with fixed raw materials and a specified concrete strength grade, we used the simplex centroid design method based on rheological properties to optimize the ideal paste composition, comprising cement, fly ash, and slag, as well as the optimal ratio between paste, fine aggregate, and coarse aggregate The flowchart illustrating the new method for concrete mix proportion is depicted in Figure 4.2, and the subsequent steps are elaborated as follows:
Figure 4 2: Flow chart for the proposed mixture design method [23]
4.2 Optimization of the mix RFCC replacement for cement
(c) Overlay Plot Figure 4 3: Optimal prediction point number 1 for mixture with RFCC replacing a part of cement
Figure 4.4: Optimal prediction point number 2 for mixture with RFCC replacing a part of cement
When evaluating the optimal prediction points presented in the two overlay plots, the primary focus is on the mixture's ability to achieve key performance metrics such as compressive strength, slump flow, flexural strength on mortars, and splitting tensile strength Both Figure 4.3 and Figure 4.4 offer viable models for concrete mixtures incorporating RFCC, but the subtle differences in the proportions of the components may influence the final choice of the optimal model
Figure 4.3 presents the first optimal prediction point, where the mixture proportions of silica fume, cement, and SFCC are balanced to achieve a high compressive strength, adequate slump flow, and strong flexural and splitting tensile strengths The yellow region in this plot clearly indicates a feasible area where all the performance criteria are met, and the red design points within this region represent the most effective mixture composition
Figure 4.4 offers a second optimal prediction point with similar goals in mind The proportions in this plot are slightly adjusted, yet the resulting performance remains within acceptable parameters Like in Figure 4.3, the yellow area identifies the feasible region where the mixture meets the desired performance criteria, and the red design points indicate the optimal mixture
When comparing the two overlay plots, both provide strong models for the mixture composition with RFCC, but Figure 4.3 appears to offer a more robust optimization The first plot’s design point ensures a balance of high compressive strength, slump flow, and flexural and splitting tensile strengths, making it a potentially more reliable choice Although Figure 4.4 presents an alternative, the slight variations in the proportions suggest that Figure 4.3 might be better suited for situations where maximizing all performance criteria is critical
4.3 Optimization of the mix RFCC replacement for Q-Powder
(c) Overlay Plot Figure 4 5: Optimal prediction point number 1 for mixture with RFCC replacing a part of Q-Powder
The graph in Figure 4 5 for silica fume shows that its optimal amount is set at the maximum range value of 0.12 Silica fume is a crucial additive that enhances the mechanical properties and durability of concrete Its high content in the mix helps to fill micro-voids within the concrete matrix, thereby increasing density and reducing permeability This leads to higher strength and better resistance to aggressive environmental factors, which are essential characteristics for high-performance applications.`
The optimal level of SFCC in the mix is very low, around 0.018501, indicating its effectiveness even in small quantities SFCC contributes significantly to the hydration process, leading to the formation of additional calcium silicate hydrate (C- S-H) gel This gel is vital for the concrete's strength and durability The minimal requirement of SFCC highlights its efficiency in enhancing the overall properties of the concrete mix without necessitating a large amount
The optimal quantity for Quartz L/Limestone is approximately 0.057499, indicating a mid-range value Quartz and limestone act as fillers that can improve the mechanical strength and workability of the concrete By providing a more homogenous and dense mix, these fillers enhance the concrete's overall performance, making it suitable for high-performance applications